Está en la página 1de 439




Other Titles in this Series:
Y. Chisti: Airlift Bioreactors


Edited by

Laboratory of General and Industrial Microbiology,
State University of Ghent, Belgium




Crown House, Linton Road, Barking, Essex IG11 BJU, England

Sole distributor in the USA and Canada

655 Avenue of the Americas, New York, NY 10010 USA


Sof tcover reprint of the hardcover 1st edition 1989

British Library Cataloguing in Publication Data
Biotechnology of vitamins, pigments and
growth factors
1. Industrial chemicals: organic compounds
I. Vandamme, Erick J.
Library of Congress Cataloging-in-Publication Data
Biotechnology of vitamins, pigments, and growth factors/edited by
Erick J. Vandamme.
p. cm.-(Elsevier applied biotechnology series)
Includes bibliographies and index.
ISBN-13: 978-94-01 0-6991-5
e-ISBN -13: 978-94-009-1111-6
001: 10.1007/978-94-009-1111-6
1. Vitamins-Biotechnology.
2. Growth promoting substances-Biotechnology.
3. Pigments (Biologyl--Biotechnology.
I. Vandamme, Erick J., 1943II. Series.
[DNLM: 1. Biotechnology.
2. Growth Substances.
3. Pigments.
4. Vitamins.
QT 34 B61545]
for Library of Congress
No responsibility is assumed by the Publisher for any injury and/or damage to persons or
property as a matter of products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions or ideas contained in the material herein.
Special regulations for readers in the USA
This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem,
Massachusetts. Information can be obtained from the CCC about conditions under which
photocopies of parts of this publication may be made in the USA. All other copyright
questions, including photocopying outside the USA, should be referred to the publisher.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, or otherwise, without the prior written permission of the publisher.

To Mireille


Vitamins and related growth factors belong to the few chemicals with a
positive appeal to most people; the name evokes health, vitality, fitness,
strength .... each one of us indeed needs his daily intake of vitamins, which
should normally be provided via a balanced and varied diet. However, current
food habits or preferences, or food processing and preservation methods do
not always assure a sufficient natural daily vitamin supply, even for a healthy
human being; this is all the more true for stressed or sick individuals. Although
modern society is seldom confronted with the notorious avitaminoses of the
past, they do still occur frequently in overpopulated and poverty- and
famine-struck regions in many parts of the world. Apart from their in-vivo
nutritional-physiological roles as growth factors for man, animals, plants and
micro-organisms, vitamin compounds are now being introduced increasingly as
food/feed additives, as medical-therapeutical agents, as health-aids, and also
as technical aids. Indeed, today an impressive number of processed foods,
feeds, cosmetics, pharmaceuticals and chemicals contain extra added vitamins
or vitamin-related compounds, and single or multivitamin preparations are
commonly taken or prescribed.
These reflections do indicate that there is an extra need for vitamin supply,
other than that provided from plant and animal food resources. Most added
vitamins are indeed now prepared chemically and/or biotechnologically via
fermentation/bioconversion processes.
Similarly, other related growth factors, provitamins, vitamin-like compounds, i.e. special unsaturated fatty acids, gibberellins and certain pigmentssome of which are increasingly used in agriculture, food/feed production or
processing and as health aids-are equally important biotechnological products, where production via microbial fermentation or micro-algal bioconversion is now applied industrially. Indeed, biotechnology-based on bacteria,
yeasts, fungi and micro-algae-here again has been very instrumental in
procuring sufficient amounts of several of these valuable complex molecules via



natural processes, although for certain products there is fierce competition with
chemical synthesis. Abundant and excellent literature is available on the
chemical properties, biochemistry, nutritional aspects and clinical aspects of
vitamins and related products, while that on microbial synthesis and production methods is rather scarce or difficult to find.
In this respect, this book intends to assemble useful information on the
(potential) industrial synthesis of economically important vitamins, growth
factors and pigments, with emphasis on biotechnological aspects including
microbiology, genetics, biochemistry and bioprocess technology; so far, such
information is scattered widely in the scientific literature: for some products,
only secrecy and sparse data are available, other excellent volumes deal with
only one specific vitamin compound, while others then stress the chemical
synthesis processes. Therefore, I felt that there was a scientific need for a
biotechnological survey of the world of vitamins and related compounds
The help of several colleagues and friends in suggesting potential authors for
difficult-to-get chapters has been invaluable in constructing a rather comprehensive volume; so was the positive interaction with all my contributors. In
this respect, I am particularly indebted to:
Dr S. Anderson, Genentech, USA; Dr G. C. Barrere, Rhone-Poulenc,
France; Dr E. Cerda-Olmedo, University of Seville, Spain; Dr D. De Buyser,
N. V. Vandemoortele, Belgium; Dr. D. Defterdarovic, Pliva, Yugoslavia;
Professor Dr A. L. Demain, MIT, USA; Dr J. Florent, Rhone-Poulenc,
France; Dr A. Furuya, Kyowa Hakko Kogyo Co., Ltd, Japan; Dr H. Nelis,
University of Ghent, Belgium; Dr L. Segers, Orffa, Belgium; Dr S. Shimizu,
Kyoto University, Japan; Dr G. Smits, Tiense Suikerraffinaderij, Belgium;
Professor Dr Y. Tani, Kyoto University, Japan.
The editor also thanks the staff of Elsevier Science Publishers Ltd.
I suspect that my wife, Mireille, could only have withstood my 'mental
absence', strengthened with multivitamin preparations, although my sole
vitamin-shot was her encouraging and moral support during this biotechnological enterprise.

E. J .



Preface . . . . . .


List of Contributors .



Vitamins and Related Compounds via Micro-organisms: a

Biotechnological View (E. J. Vandamme) . . . . . . .

Fat-Soluble Vitamins and Pigments


p-Carotene (Provitamin A) Production with Algae (L. J.

Borowitzka & M. A. Borowitzka) . . . . . . . . . .
Production of Carotenoids with Fungi (E. Cerda-Olmedo
Microbial Production of Carotenoids other than p-Carotene (H.
J. Nelis & A. P. De Leenheer) . . . . . . . . . . . .
Vitamin D: The Biotechnology of Ergosterol (P. Margalith) . .
Algal and Microbial Production of Vitamin E (Y. Tani) . . . .
Microbial Production of Polyunsaturated Fatty Acids (VitaminF Group) (S. Shimizu & H. Yamada) . . . . . . . . .
Microbial Production of Vitamin K2 (Menaquinone) and
Vitamin Kl (Phylloquinone) (Y. Tani) . . . . . . . .


Water-Soluble Vitamins

Microbial Synthesis of Vitamin Bl (Thiamine) (A. Iwashima)

Microbial Production of Vitamin B2 (Riboflavin) (T. Kutsal &
M. T. Ozbas) . . . . . . . . . . . . . . . . . . . . .
Microbial Production of D-Ribose (K. Sasajima & M. Yoneda)
Pantothenic Acid (Vitamin B s), Coenzyme A and Related
Compounds (S. Shimizu & H. Yamada). . . . . . . . . .
Microbial Production of Vitamin B6 and Derivatives (Y. Tani)




Microbial Production of Biotin (Y. Izumi & H. Yamada)

Microbial Production of Vitamin B12 (C. Spalla, A. Grein, L.
Garofano & G. Feroi) . . . . . . . . . . . . . .
Microbial Production of Orotic Acid (Vitamin B 13) (K.
Takayama & A. Furuya) . . . . . . . . . . . . .
Microbial Reactions for the Synthesis of Vitamin C (L-Ascorbic
Acid) (V. Delic, D. Sunic & D. Vlasic) . . . . . . . . . . .


Other Growth Factors



Microbial Production of ATP (Y. Tani) . . . . . . . . .

Adenosylmethionine, Adenosylhomocysteine and Related
Nucleosides (S. Shimizu & H. Yamada) . . . . . . . . .
Other Vitamin-Related Coenzymes (S. Shimizu & H. Yamada)
Fungal Gibberellin Production (B. Brueckner, D.
Blechschmidt, G. Sembdner & G. Schneider)




Friedrich Schiller University Jena, Department of General Microbiology,

DDR-6900 Jena, Neugasse 24, GDR
Western Biotechnology Ltd, 2-6 Railway Parade, Bayswater, WA 6063,
Algal Biotechnology Laboratory, School of Biological and Environmental
Sciences, Murdoch University, Murdoch, WA 6150, Australia


Friedrich Schiller University lena, Department of General Microbiology,

DDR-6900 lena, Neugasse 24, GDR



Departmento de Genetica y Biotecnia, Universidad de Sevilla, Apartado

1095, Sevilla, Spain
Laboratoria voor Medische Biochemie en voor Klinische Analyse,
Rijksuniversiteit Ghent, Harelbekestraat 72, B-9000 Ghent, Belgium
V. DELle,
PLIVA Pharmaceutical, Chemical, Food and Cosmetic Industry, Research
Institute, Zagreb, Yugoslavia

Farmitalia Carlo Erba, Via dei Gracchi 35, 20146, Milano, Italy



Kyowa Hakko Kogyo Co. Ltd, Tokyo Research Laboratories, 3-6-6

Asahi-machi, Machida-shi, Tokyo 194, Japan

Farmitalia Carlo Erba, Via dei Gracchi 35, 20146, Milano, Italy

Farmitalia Carlo Erba, Via dei Gracchi 35, 20146, Milano, Italy





Department of Biochemistry, Kyoto Prefectural University of Medicine,

Kawaramachi-Kirokoji, Kamigyo-ku, Kyoto 602, Japan
Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan

List of Contributors


Hacettepe University, Chemical Engineering Department, 06532 Beytepe,

Ankara, Turkey



Department of Food Engineering & Biotechnology, Technion-Israel

Institute of Technology, Haifa, 32000, Israel
Laboratoria voor Medische Biochemie en voor Klinische Analyse,
Rijksuniversiteit Ghent, Harelbekestraat 72, B-9000 Ghent, Belgium
M. T. (hBAS,

Hacettepe University, Chemical Engineering Department, 06532 Beytepe,

Ankara, Turkey


Central Research Division, Takeda Chemical Industries Ltd, 2-17-85

Jusohonmachi, Yodogawa-ku, Osaka 532, Japan



Institute of Plant Biochemistry, Academy of Sciences of the GDR, DDR4050 Halle, Weinberg 3, GDR

Institute of Plant Biochemistry, Academy of Sciences of the GDR, DDR4050 Halle, Weinberg 3, GDR


Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan





Farmitalia Carlo Erba, Via dei Gracchi 35, 20146, Milano, Italy

PLIVA Pharmaceutical, Chemical, Food and Cosmetic Industry, Research

Institute, Zagreb, Yugoslavia


Kyowa Hakko Kogyo Co. Ltd, Tokyo Research Laboratories, 3-6-6

Asahi-machi, Machida-shi, Tokyo 194, Japan


Research Center for Cell and Tissue Culture, Faculty of Agriculture, Kyoto
University, Kyoto 606, Japan
Laboratory of General and Industrial Microbiology, Faculty of Agricultural
Services, State University of Ghent, Coupure Links 653, B-9000 Ghent,
PLIVA Pharmaceutical, Chemical, Food and Cosmetic Industry, Research
Institute, Zagreb, Yugoslavia
Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan

Corporate Strategy,
Takeda Chemical Industries
Jusohonmachi, Yodogawa-ku, Osaka 532, Japan



Chapter 1


Laboratory of General and Industrial Microbiology, Faculty of Agricultural
Sciences, State University of Ghent, Coupure Links, 653, B-9000 Ghent,

The start of the history of vitamins can be traced back to 400 Be, when
Hippocrates reported that eating liver could cure night-blindness. Much later,
in the 16th century, the therapeutical effects of lemon juice against scurvy or
scorbut became known; scorbut had inter alia caused the loss of 100 crew
members on Vasco da Gama's journey around Cape Hope. The English ship
doctor James Lind studied this disease further and described in 1757 in his
book A Treatise of Scurvy the beneficial effect of eating fresh vegetables and
fruits in preventing it. For another nutritional deficiency disease (already
mentioned in 1762 by Oviedo), the Italian doctor Francesco Frapoli used the
name pellagra (pella = skin; agra = rough). In the 19th century in Japan, the
Hikan child disease (keratomalacia and xerophthalmia) was successfully
treated by including ale-fat, cod liver oil or chicken liver in the diet. Trousseau
discovered that cod liver oil and also, direct sunlight, had a curing effect on
rickets, a disease already well described by Whistler in 1645. In the Far-East,
when hulled rice was replaced by de hulled or polished rice, a sharp increase in
the occurrence of beriberi was observed. In 1897 Eijkman observed that
poultry fed with polished rice developed polyneuritis, a disease very similar to
human beriberi. This disease could also be prevented and cured by feeding
rice and the silver fleece of the rice kernel. Grijns in 1901 hypothesised that
beriberi was caused by a protecting factor, which was obviously lacking in
dehulled rice. We now know that all these diseases are a result of nutritional
vitamin deficiencies (Machlin, 1984).
Around 1910, Hopkins in the UK and Osborne & Mendel in the USA
initiated modern vitamin research and developed a theory, stating that diseases
such as scurvy, pellagra, rickets, beriberi, etc., are the result of a lack of
certain essential food components. The Polish chemist Casimur Funk isolated
in 1912 from rice bran a beriberi-preventing compound, displaying chemical

E. J. Vandamme

properties of an amine; this led him in 1912 to coin the name vitamin for this
type of compounds.
In 1913, McCollum & Davis demonstrated a liposoluble factor A in butter
fat and egg yolk, and in 1915, a water-soluble factor B was found in
wheat-germ. It was Drummond in 1920 who named the fat-soluble factor,
vitamin A; the water-soluble anti-beriberi factor was named vitamin B; the
water-soluble anti-scorbut factor vitamin C. In 1925, the fat-soluble anti-rickets
factor was named vitamin D. After 1930, discovery and isolation of several
other vitamins followed quickly and their structure, nutritional and chemical
properties and synthesis were studied in great detail in the following decades.
These aspects of vitamins and growth factors are compiled in several
excellent standard references (Goodwin, 1963; Sebrell & Harris, 1971; De
Luca, 1978; Machlin, 1984; De Leenheer et al., 1985; Diplock, 1985; Chytil &
McCormick, 1986; Adrian, 1988; Friedrich, 1988).


Vitamin nomenclature was initially based on the use of letter symbols,

alphabetically arranged to time of discovery. Soon, it appeared that one-letter
named vitamins were multiple complexes, and this led to the addition of an
index to the original letters (Bb B 2 , . ). Often, when the function of the
vitamin became known, an appropriate letter symbol was chosen, i.e. vitamin
K, with K as the first letter of the German word Koagulation; other names
reflected deficiencies, i.e. aneurin (B 1) for anti-polyneuritis vitamin; vitamin
PP = "pellagra-preventing" vitamin. Letter names or trivial names are generally more in use rather than the IUPAC names. The division into fat- and
water-soluble vitamins as introduced by McCollum & Davis, is still universally
A listing of well-recognised vitamin compounds is presented in Table 1.
From a chemical point of view, vitamins are a very heterogeneous mixture of
compounds, yet they can be considered as a single group, since they are all
organic compounds which are essential for a healthy development of humans
and animals and need to be present in their food, since their body is not able at
all--or not sufficiently able-to synthesize them. Indeed, certain vitamins can
be formed partially or indirectly in the body:
(a) compounds--often called provitamins--with no apparent vitamin activity can be converted within the body into a vitamin, i.e.
Jj-carotene (in vegetables, fruits) - - vitamin A
tryptophan (in protein-rich food) - - nicotinic acid (niacin)
7-dehydrocholesterol (in skin) uv) vitamin D3 (cholecalcifenol)
(b) other vitamin compounds are formed by the intestinal bacterial flora i.e.
vitamin K, some B-vitamins, i.e. B 12 , etc.
Most vitamin compounds thus have to be provided via daily food/feed intake.

Vitamins and Related Compounds via Micro-organisms

Table 1
Survey of Vitamins and Growth Factors
Vitamin group

Provitamin A group

Important examples


Important natural

Vitamin A group

Provitamin D
Vitamin D group

I IV = 0,6 pg
I pg = 334 IV

24 mg,B-carotene

Fish oil, liver,

milk and daily
products, eggs

I IV = 030 pg

075 mg retinol

Fish oil, liver,
egg yolk, milk,
dairy products
Plant oils,
liver, eggs,
dairy products
Fish oil, algae,

I IV = 0025 pg
vitamin D,
Illg = 40 IV
I IV= IOmg

cholecalciferol (D,)
a- Tocoferol,
Linolic acid,
Vitamin F group
y-linolenic acid,
fatty acids, PVFAs), eicosa-pentaenoic
acid (EPA),
w-3-fatty acids)
Liver, vegetables
Vitamin K group
Phylloquinone (K.)
bacterial flora,
or phytomenadion,
phamoquinone (K 2 ) or
plant oils
menadion (K,),
menadiol (K.)
Cereals, liver,
Vitamin B.
1 IV = 31lg
meat, vegetables,
dairy products,
lllg = 0333 IV
Liver, eggs,
Vitamin B2
(vitamin G)
dairy products,
vegetables, meat,
bacterial flora
Liver, cereals,
Vitamin PP
meat, potatoes,
(vitamin B 3 )
nicotinic acid
yeast, tryptophane-protein
Pantothenic acid,
Liver, eggs, molasses
Pantothenic acid
coenzyme A
(vitamin B 5 )
Liver, potato,
Vitamin B6 group
meat, yeast
Liver, egg yolk,
vegetables, cane
(vitamin H, B 8 )
bacterial flora

Vitamin E group

Recommended adult
daily intake



Retinol (vitamin A.),
(vitamin A 2 ),
retinal (vitamin A.aldehyde)
Ergocalciferol (D2 ),

Units (IU)

25 pg vitamin D3




10-15 mg





E. J. Vandamme

Table l---contd.
Vitamin group

Important examples

Important natural

Folic acid
(B.-group, Bc ' M)

Vitamin B12 group

Vitamin B13
Vitamin C group

Pteroyl-glutamic acid,
Cyanobalamin (B 12 ),
(B I2b) ,
(B 12c )
Orotic acid
Ascorbic acid
(vitamin q,
dehydro-ascorbic acid

Units (IU)

Cereals, liver,
meat, milk,

400 I'g

Meat, liver, milk,

bacterial flora

Citrus fruit, fruit,


1 IU = 50 I'g
L-ascorbic acid

Yeast, meat
Yeast, meat,
Egg yolk, meat,
Plants, fungi

Lipoic acid
Choline (B7' J)
Para-aminobenzoic acid
(PABA) (Bx' H', H 2 )

Recommended adult
daily intake

,(not for animals

except, primates,
cavia, etc.)


The vitamin content was originally and still often is expressed in International units (IU), which relate to the biological activity of a certain amount of
pure vitamin towards a specific test animal.
Presently, it is common practice to express the vitamin content as mg or f-tg
per 100 g of material; conversion factors for IU values into weight units are also
given in Table 1. Recommended daily intake per person per day (FAO/WHO
data) is also given.
Quantitative assays for vitamin content, in foodstuffs, concentrated mixtures, synthetic formulations, tissues, blood, fermentation broths, etc., are
very important and ever more sophisticated techniques (e.g. HPLC) are
introduced; for several vitamins, i.e. B 12 , biotin, etc., a microbiological
bioassay is still the method of choice (Freed, 1966; Berg & Behagel, 1972; De
Leenheer et al., 1985).


Vitamins have a catalytic role in the body in enabling optimal synthesis,

conversion and degradation of macromolecules such as nucleic acids, proteins,
lipids and carbohydrates.
The biochemical (physiological) function of most water-soluble vitamins is
well known: they are part of co-enzymes and thus co-responsible for specific
biochemical reactions (Machlin, 1984; Diplock, 1985). A survey is given in
Table 2.

Vitamins and Related Compounds via Micro-organisms

Table 2

Water-Soluble Vitamins and their Corresponding Coenzymes


(B3 or PP)



Folic acid


acid (Bs)
(B 1 )

(B 12)


Group Transfer

dinucleotide (NAD+)
dinucleotide phosphate
Flavine adenine mononucleotide (FMN)
Flavine adenine dinucleotide (FAD)
Biotin (biocytin,
Coenzyme A


Thiaminepyrophosphate (TPP)

Cz-aldehyde, decarboxylation

Amino-group, decarboxylation

As to the function of fat-soluble vitamins and vitamin C, much controversy

still exists; their involvement can often be traced down to specific biochemical
processes (Table 3), although their exact function is not yet known in all cases
(De Luca, 1978; Diplock, 1985).
Much debate also exists about the positive effects of high (mega) doses of
certain vitamins, (e.g. vitamin C) on human and animal physiology; on the
other hand, several hypervitaminoses, (e.g. A, D, K) are well known
(Machlin, 1984).
Some vitamins display different activity toward man or animal, i.e. ergocalciferol (D 2 ) is poorly active in poultry, while in other animals and man it is
equally active as cholecalciferol (D3)' This observation has great practical
repercussions on feed formulations. Others are essential solely for man, (i.e.
vitamin C), while most animals (except primates, cavias) can synthesise them.
Compounds which specifically counteract the functioning of vitamins are also
known and are named antivitamins or vitamin antagonists (Machlin, 1984).
Their negative action can be based on degradation of the vitamin (thiaminase,
ascorbase, etc.), or on the complexation of the vitamin into a non-resorbable
complex, (i.e. avidin plus biotin). N5-HydroxY-L-arginine is a vitamin B12
antagonist (Perlman et al., 1974). Dicoumarin excludes vitamin K from the
prothrombin synthesis system and amethopterin is an antagonist of folic acid;
however, both antivitamins are medically important for the treatment of

E. I. Vandamme

Table 3
Important Biochemical Actions of Fat-Soluble Vitamins and of Vitamin C
Vitamin A
active form:
Vitamin D3
active form:
Vitamin E:
( a-tocoferol)
Vitamin K j :
Vitamin C

Physiological functions
Retinol is part of rhodopsin, the light sensitive
molecule in the eye
Biosynthesis of mucopolysaccharides;
synthesis and maintenance of epithelial cells
Regulation in Ca2 + and phosphorous metabolism

Antioxidant action towards unsaturated compounds, i.e. fatty acids, etc.; membrane
Essential for formation of prothrombine,
a blood coagulation factor
Cofactor for carboxylation of protein bound
glutamate residues to form y-carboxyglutamates
Cosubstrate of monooxygenases;
Role in redox reactions and in
hydroxylation reactions of amino acids and
amines (i.e. proline conversion into
hydroxyproline in collagene);
Role in hormone-synthesis, iron
absorption, etc.

thrombosis and leukemia, respectively. Antivitamins present in our daily food

are usually destroyed during processing or cooking.
Excellent information on detailed nutritional, clinical and physiological
aspects of vitamins has been assembled by Machlin (1984), Diplock (1985) and
Friedrich (1988).



In addition to their nutritional, physiological and medical importance, vitamins

and vitamin-like compounds have also found large-scale technical applications
as antioxidants (vitamin C, E), as acidulant (vitamin C), and as pigments
(f3-carotene and analogues) in the food, feed, cosmetic, chemical and
pharmaceutical industry (Machlin, 1984; Diplock, 1985; Florent, 1986).
Apart from the synthetic pigments, most of the natural colorants are now
being extracted from plants and animals, e.g. annatto, grapes, red beets,
paprika, female insects (Coccus cacti). Micro-organisms could be an excellent
source of non-toxic food colorants, but except for the fungal Monascus
pigments (monascin), used in the Orient, and for astaxanthin from yeast
few attempts have been made to produce and introduce them (Lin, 1973;

Vitamins and Related Compounds via Micro-organisms

Palleroni et a/., 1978; Institute of Food Technologists, 1980; Wong & Koehler,
1981). Indeed, in addition to the well-known provitamin-carotenoids, a range
of anthraquinone pigments, chlorophylls and several others have been demonstrated in bacteria, yeasts, fungi and algae and are attractive as natural
In this respect, more emphasis should be given to screening and research on
other natural pigment synthesis via safe micro-organisms; this would open up a
wider application area in agriculture, food, feed, chemicals, cosmetics and
Fungal gibberellins are already widely used as plant growth-regulators in
agriculture and horticulture and in the brewing industry (Jefferys, 1970;
Palmer, 1974; Brueckner et a/., this volume).
Details about technical applications of certain vitamins, pigments and
gibberellins are given in the corresponding chapters in this volume.


The staple food of man, including cereals, rice, potato, vegetables, fruits, milk,
fish, meat, eggs, forms his basic source of vitamins and growth factors. Adequate
nutrition should thus supply this daily vitamin need, which however increases
with physical exercise, pregnancy, lactation, active growth, reconvalescence,
drug abuse, stress, air pollution, etc. Pathological situations (intestinal
malresorption, stressed intestinal flora, liver/gall diseases, drug, antibiotic or
hormone treatment, enzyme deficiencies), can also lead to vitamin shortages
despite a sufficient intake. Malnourishment in many countries of the world also
asks for direct medical remediations, combined with diet adjustment. Vitaminenriched and medicated feed is used worldwide to procure healthy livestock.
However, derived concentrates or extracts from these vitamin-rich natural
food products find relatively little use in the food, feed, pharmaceutical or
cosmetic industry (except for w-3-fatty acids, vitamin E, etc.). Some of the
reasons are:
(a) the average daily intake of vitamins does not always seem to be supplied
solely via these natural products;
(b) the level of the natural plant/ animal source vitamins is usually relatively
low and fluctuates drastically;
(c) their organoleptic presentation and shelf-life is often far from optimal;
(d) vitamins are labile molecules during the process of preservation, storage
or preparation of foodstuffs and are generally very sensitive to pH, heat,
light, oxygen, etc.; water-soluble vitamins are easily lost by aqueous
extraction or manipulation of foods.
These factors have led to the industrial manufacturing of most vitamins and
growth factors. Current world production of important vitamins is given in
Table 4.
At the moment, several vitamins are produced only or mainly chemically

E. I. Vandamme

Table 4
Survey of Vitamins and Growth Factors-Production and Application

(see also
Table 1)

Provitamin A
Vitamin A
Provitamin D
Vitamin D
Vitamin E
Vitamin F
Vitamin K
Vitamin 8 1
Vitamin 8 2
Vitamin PP
(vitamin 8 3 )
(vitamin 8 s)
Vitamin 8 6
(vitamin H,
8 8)
Folic acid
(89 )
Vitamin 8 12

(E = extraction;
B = biotechnology)

(F = food/feed;
T = technical)

8 (fermentation, algal)


8 (fermentation)
8 (fermentation)
8 (algal)
8 (fermentation, algal)
8 (fermentation)
8 (fermentation)
8 (fermentation)
8 (fermentation)

World production


or country a



2, 5, 6, 6a, 15a,
18, 20, 22, 25











8, 14a, 18, 34



3, 6, 6a, 10, 18,
9,11,17, 18,














18,31, 39, 41



2, 18, 24, 27,








14, 15, 16,21,23,


Vitamins and Related Compounds via Micro-organisms

Table 4--contd.
World production

(see also
Table 1)

(E = extraction;
B = biotechnology)

(F = food/feed;
M = medical;
T = technical)


Vitamin B13

B (fermentation)
C plus
B (fermentation)



Lipoic acid



(B., H', Hz)


B (fermentation)

Vitamin C







or country a




a Listing of vitamin and growth factor producing companies or countries: 1. Abbott (USA); 2. BASF
(FRG); 3. Biocrops Ltd (UK); 4. China; 5. Coors Biotech Inc. (USA); 6. Cyanotech Inc (USA); 6a.
Dainippon Ink & Chemicals (Japan); 7. Degussa (FRG); 8. Duphar (The Netherlands); 9. EastmanKodak (USA); 10. Efamol Holdings Ltd (UK); 11. Eisa Co. (Japan); 12. Eli Lilly (USA); 13. E. Merck
(FRG); 14. Farmitalia-Carlo Erba (Italy); 14a. Fujisawa (Japan); 15. Genex Corp. (USA); 15a. Gist
Brocades (The Netherlands); 16. Glaxo (Sefton Chern) (UK); 17. Henkel (FRG); 18. HoffmannLaroche (Switzerland); 19. ICI (UK); 20. Koor Foods (Israel); 21. Kyowa Hakko Kogyo (Japan); 22.
Martek Inc. (USA); 23. Medimpex (Yugoslavia); 24. Merck S & D (USA); 25. Microbial Products Inc.
(USA); 26. Nippon Oil & Fats Co. Ltd (Japan), 27. Nippon Zeon (Japan); 28. Pao Yeh (Taiwan); 29.
Pierrel (Italy); 30. Pfizer Inc. (USA); 31. Pliva (Yugoslavia); 32. Q. P. Corporation (Japan); 33.
Rhone-Poulenc (France); 34. Riken Vitamin (Japan); 35. RMC (Evening Primrose Oil Company Ltd)
(UK); 36. Roussel-UcLaF (France); 37. Shiseido Co. Ltd (Japan); 38. Sturge Biochemicals (UK); 39.
Sumitomo (Japan); 40. Synergen Inc. (USA); 41. Takeda (Japan); 42. Tanabe-Kongo (Japan); 43.
Tanabe (Japan); 44. USSR; 45. Vanetta (Italy); 46. Western Biotechnology Ltd (Australia); 47. Yuki
Gosei (Japan); 48. Yodo Gowa-Duphar (Japan).

(vitamins A, D 3 , E, K, Bh H, etc.), although microbiological methods exist or

emerge; others are produced (exclusively) via fermentation (D2' B 6 , B 12 , F) or
micro-algal culture (fJ-carotene, F) and for others both chemical and microbial
processes are run industrially (vitamin B2 ), while vitamin C is produced via a


E. J. Vandamme

combination of chemical and microbiological reactions (Sakaguchi et al., 1971;

Yamada et al., 1971; Ogata et al., 1976; Periman, 1978; Florent, 1986;
Borowitzka & Borowitzka, 1988). Details about these bioprocesses form the
backbone of this volume!
A broad range of applications now exists for these vitamins and related
compounds in food, feed, cosmetics, technical and pharmaceutical
revitamination: restoring the original vitamin level of a foodstuff;
standardisation: addition of vitamin(s) to compensate for natural
vitamin-enrichment: extra addition of vitamin(s) to a level higher than the
natural one (health food, diet food industry);
vitamination: addition of vitamins to products lacking them (feed,
cosmetics) ;
technical additive (in food, feed, cosmetics), e.g. p-carotene as pigment,
vitamin C and E as antioxidants, vitamin C as acidulant, gibberellins as
growth hormones for plants;
medical applications: to alleviate hypo- or even avitaminoses.
Apart from obtaining these vitamins via a natural way-what microbial and
algal biotechnology is all about-fermentation or bioconversion reactions yield
the desired enantiomeric vitamin compound, while products from organic
synthesis are often racemic mixtures, each displaying a differing biological
activity. Furthermore, vitamin yields in fermentation broths are often very
high-especially when using mutants or genetically engineered strains-as
compared with their natural levels in plants or animals.
In this respect, continued efforts should be made to further explore the
potential and power of microbial and algal biotechnology in the field of
economic natural vitamin, pigment and growth factor production.
Adrian, J. (1988). Parat Dictionary of Food and Nutrition, VII. Ellis Horwood Series in
Food Science and Technology. Ellis Horwood, Chichester, UK.
Berg, T. M. & Behagel, H. A. (1972). Semi-automated method for microbiological
vitamin-assays. Appl. Microbiol., 23,531-42.
Borowitzka, M. A. & Borowitzka, L. J. (eds) (1988). Micro-algal Biotechnology.
Cambridge University Press, Cambridge, UK.
Brueckner et al. (1989). Fungal gibberellin production. In Biotechnology of
Vitamins, Pigments and Growth Factors, ed. E. J. Vandamme. Elsevier Science
Publishers, London, pp. 383-429
Chytil, P. & McCormick, D. B. (eds) (1986). Vitamins and Coenzymes. Methods in
Enzymology, Vols 122, 123. Academic Press, New York, London.
De Leenheer, A. P., Lambert, W. E. & Deruyter, M. G. M. (eds) (1985). Modern
Chromatographic Analysis of the Vitamins. Marcel Dekker, New York, Basel.
De Luca, H. F. (ed.) (1978). The fat-soluble vitamins. In Handbook of Lipid Research,

Vitamins and Related Compounds via Micro-organisms


Vol. 2. Plenum Press, New York, London.

Diplock, A. T. (ed.) (1985). Fat-Soluble Vitamins. Their Biochemistry and Applications. Heineman, London.
Florent, J. (1986). Vitamins. In Biotechnology, Vol. 4, Microbial Products II; ed. H.
Pape & H. J. Rehm. VCH, Weinheim, pp. 115-58.
Freed, M. (ed.) (1966). Methods of Vitamin Assay, 3rd edn. John Wiley & Sons, New
Friedrich, W. (1988). Vitamins. Walter de Gruyter, Berlin, New York.
Goodwin, T. W. (1963). The Biosynthesis of Vitamins and Related Compounds.
Academic Press, London.
Institute of Food Technologists (1980). Food colors. Food Technology, 34,77.
Jefferys, E. G. (1970). The Gibberellin Fermentation. Adv. App/. Microbio/., 13,
Lin, C. F. (1973). Isolation and cultural conditions of Monascus sp. for the production
of a pigment in a submerged culture. 1. Ferment. Techno/., 51,407-9.
Machlin, L. J. (ed.) (1984). Handbook of Vitamins: Nutritional, Biochemical and
Clinical Aspects. Marcel Dekker, New York, Basel.
Ogata, K., Kinoshita, S., Tsunoda, T. & Aida, K. (1976). Microbial Production of
Nucleic Acid Related Substances. Kodansha, Tokyo.
Palleroni, N. S., Reichelt, K. E., Mueller, D., Epps, R., Tabenkin, B., Bull, D. N.,
Schnep, W. & Berger, J. (1978). Production of a novel red pigment, rubrolone, by
Streptomyces echinoruber sp. 1. Antibiotics, 31, 1218-25.
Palmer, G. J. (1974). The industrial use of gibberellic acid and its scientific basis-a
Review. 1. Inst. Brewing, 80, 13-30.
Perlman, D. (1978). Vitamins. Economic Microbiology, 2, 303-26.
Perlman, D., Vlietinck, A. J., Matthews, H. W. & Lo, F. F. (1974). Microbial
production of vitamin B12 anti metabolites. I. N5-hydroxY-L-arginine from Bacillus
cereus 439. 1. Antibiotics, 27, 826-32.
Sakaguchi, K., Uemara, T. & Kinoshita, S. (1971). Biochemical and Industrial Aspects
of Fermentation. Kodansha, Tokyo.
Sebrell, W. H. & Harris, R. S. (eds) (1971). The Vitamins, 2nd edn. Academic Press,
New York, London.
Wong, M. C. & Koehler, P. E. (1981). Production and isolation of an antibiotic from
Monascus purpureus and its relationship to pigment production. 1. Food Sc., 46,
Yamada, K., Nakahara, T. & Fukui, S. (1971). Petroleum microbiology and vitamin
production. In Biochemical and Industrial Aspects of Fermentation, ed. Sakaguchi,
T. Uemura & S. Kinoshita. Kodansha, Tokyo, pp. 61-90.


Chapter 2
Western Biotechnology Ltd, 2-6 Railway Parade, Bayswater, WA 6063,


M. A.


Algal Biotechnology Laboratory, School of Biological and Environmental

Sciences, Murdoch University, Murdoch, WA 6150, Australia

Dunaliella is a unicellular, biflagellate, naked green alga (Chlorophyceae,

Dunaliellales), and the type species of this genus, Dunaliella salina (Dunal)
Teodoresco is often found in natural hypersaline waters where it colours the
brines red (Teodoresco, 1905). This algal species was first recognised as
containing high intracellular concentrations of p-carotene by Mil'ko (1963) and
Aasen et al. (1969). Initial research on the potential of using this alga as a
commercial source of p-carotene began in the Ukraine in the 1960s (cf.
Massyuk, 1966; Massyuk & Abdula, 1969) and it was later also proposed as a
commercial source of glycerol (Ben-Amotz; 1980; Chen & Chi, 1981;
Ben-Amotz & Avron, 1982).
A number of commercial-scale developments on the production of pcarotene from D. salina have commenced in Australia (Curtain et al., 1987;
Borowitzka & Borowitzka, 1988a), Israel (Rich, 1978) and the USA (Klausner, 1986), and a number of open-pond and closed reactor experimental units
are in development stages. Pilot projects are also under way in China and
Chile. Commercial quantities of extracted algal p-carotene and dried D. salina
powder rich in p-carotene have been marketed by companies from the US and
Australia since 1985. All are targeting the markets for 'natural' food and
animal feed colourings, and 'natural' p-carotene (provitamin A) nutritional
p-Carotene is accumulated as droplets in the chloroplast stroma of Dunaliella
salina cells, particularly when culture conditions include high light intensities,

T. Kutsal & M. T. Ozbas





















fo 105931














.- U




Time (min)

Fig. 1. Typical HPLC analysis of D. salina J3-carotene extract.

high temperature, high salinity and deficiency of nitrogen source (Lerche,

1937; Mil'ko, 1963; Mironyuk & Einor, 1968; Semenko & Abdullayev, 1980;
Ben-Amotz & Avron; 1983; Borowitzka et at., 1984). p-carotene contents of
up to 14% of dry weight have been reported for D. salina (Ben-Amotz et at.,
1982; Borowitzka et at., 1984). The p-carotene does not appear to act as a
light-harvesting photo-accessory pigment, but rather as a photo-protective
'sunscreen' to the chlorophyll and cell DNA in the high light intensities which
characterise the normal salt-lake environments where this alga grows
(MacKinney & Chichester, 1960; Ben-Amotz, 1980). Borowitzka & Borowitzka (1988a) have also proposed that p-carotene acts as a 'carbon sink' in D.
salina for excess combined carbon produced when photosynthesis continues
under conditions where growth is limited.
The p-carotene in the chloroplast droplets is a mixture of the cis- and
trans-isomers. A typical analysis from D. salina (also called D. bardawil) gives
the following composition as percentages of total p-carotene: 15-cis-pcarotene, 10%; 9-cis-p-carotene, 41%; all-trans-p-carotene, 42%; other isomers, 6% (Ben-Amotz et at., 1982).
For analytical purposes, total p-carotene may be calculated from its
extinction in acetone extracts (m~::. at 452 nm is 2592; Bauernfeind, 1981).
HPLC is required to separate the isomers, and to characterise the small
quantities of other carotenoids also present in typical D. salina extracts (Nelis
& de Leenheer, 1983). Figure 1 shows an HPLC analysis of a typical
commercial batch of 20% p-carotene suspension in peanut oil produced by
Western Biotechnology Limited from D. salina.
Dunaliella salina production systems, like those used by Western Biotechnology Ltd (Fig. 2), have large open ponds situated in, or near, salt lakes or solar

{3-Carotene Production with Algae


Fig. 2. Aerial photographs of Western Biotechnology's Dunaliella production facility at

Hutt Lagoon, WA, Australia.

salt works. These ecosystems and the ponds are inhabited by wild-type D.
salina, non-carotenogenic Dunaliella species, predatory protozoa, halophilic
bacteria and brine shrimp (Borowitzka et al., 1986; Moulton et al., 1987b). In
the presence of this mixture of organisms, maintenance of cultures dominated
by the carotenogenic D. salina is the first priority of pond management. In an
environment dominated by the wild-type D. salina, introduction of an


L. J. Borowitzka & M. A. Borowitzka

'improved' D. salina strain is extremely difficult, and has not been reported on
a commercial scale. In the large, open-air ponds used for the commercial
culture of D. salina the major control of the culture available to the operator is
salinity (Borowitzka et al., 1986), although some additional control may be
exercised by manipulating nutrient concentrations.
In more isolated systems, as for example, concrete or plastic-lined raceway
ponds and tubular photobioreactors, wild-type D. salina and other competing
organisms may be excluded and the introduction of improved strains can be
achieved, at the cost of higher investment in plant and equipment.
Strain improvement, both in terms of growth characteristics and product
yield, can be achieved using mutagenesis and selection programmes. Improved
strain breeding using mating strains as in Chlamydomonas, a closely related
alga, is also possible (Craig et al., 1988). Further into the future, and awaiting
additional research, is the use of cell fusion techniques or the genetic
engineering manipulations developed for yeast cells (e.g. Spencer & Spencer,
In the laboratory, D. salina may be isolated into unialgal and axenic culture
either by repeated streaking on agar plates or by isolating clonal cultures from
single cells. The most commonly used medium for D. salina and related algae
is modified Johnson's medium (Borowitzka, 1988), although a range of other
media can also be used, as long as the salinity is appropriately adjusted with
Laboratory cultures are usually maintained on agar slants or in liquid
cultures, and sub-cultured every 1-2 months. Care must be taken to avoid
excessive drying out of the high salinity media. We have also successfully
maintained Dunaliella species, frozen in liquid nitrogen in our laboratory for
periods in excess of 12 months. In the field, stock cultures are best maintained
at saturating salinities where the likelihood of contamination by other
organisms is minimised.
There is an inverse relationship between growth rate and carotenogensis in D.
salina; conditions which enhance accumulation of p-carotene include high light
intensity, high temperature, high salinity, nitrogen deficiency, and these same
conditions tend to decrease growth rate. This is a typical pattern for secondary
metabolite production in stationary phase cultures of algae. Table 1 summarises the environmental factors which influence growth and carotenogenesis
in D. salina.
The effect of salinity on carotenogenesis is best shown in experiments which
investigated the induction of carotenogenesis. In these experiments the NaCI
concentration was increased from 15 to 25% (w/v) and the p-carotene content
of the D. salina cells increased linearly from <10 to 260 mg (g cell protein)-t,
over 5 days, while growth ceased (Borowitzka et al., 1984).


fJ-Carotene Production with Algae

Table 1
Summary of the Influence of Various Environmental Factors on
the Biomass and Jj-Carotene Content of Dunaliella salina. + =
Stimulating Effect; - = Inhibitory Effect; 0 = No Effect


Increase in salinity
Decrease in salinity
N deficiency
P deficiency
Increase in [C02] supply
Increase in light
Decrease in light
Increase in temperature
Decrease in temperature
Increase in [0 2]






Rate of decrease is very slow.

Fortuitously, if salinity is reduced by a similar amount, as for example when

rain enters open ponds, the decline in p-carotene content of the cells is very
slow (Borowitzka et al., 1984) as the p-carotene is not broken down, but
rather the cell content is reduced as the cells divide.


Open-pond D. salina production is perhaps more closely related to farming

than to fermentation techniques. Open cultures are influenced by weather
(temperature, light, rainfall, wind), seasons, and input of contaminants from
the environment. Enclosed ponds, tank or tubular photobioreactors and
fermentor systems can reduce the influence of these external factors on the
culture; however, these have, as yet, not been applied to D. salina culture on a
commercial scale.
Dunaliella salina cultures may be operated either in batch or continuous
mode. A two-stage process, using lower salinity conditions for optimum
growth rate and biomass production and then changing to high salinity stress
conditions favouring carotenogenesis has also been piloted and the disadvantages of this two-stage process are discussed in Borowitzka et al. (1984). A
single-stage, semi-continuous process at an intermediate salinity appears to be
the best at this stage, and this is the process used by Western Biotechnology
Ltd (Fig. 3).
In all D. salina culture systems the following are necessary for good growth
and carotenogenesis:
(a) Light. Light is the only energy source for metabolism as D. salina is an
obligate photoautotroph. The light source may be sunlight or high
intensity artificial white light. Growth and carotenogenesis respond

L. 1. Borowitzka & M. A. Borowitzka




(50 Ha)











Dunaliella salina
powder containing
3% fJ-Carotene


Suspensions in
Vegetable oil
(2%, 10%,20%,30%)
Water Dispersibles
Water Solubles
-Quality control



Fig. 3. Flow chart of the Dunaliella salina fJ-carotene production process used by
Western Biotechnology Ltd.

fJ-Carotene Production with Algae







differently to particular wavelengths of light (Mil'ko, 1963), and high

photon-flux densities favour fj-carotene accumulation.
Temperature. Dunaliella salina is tolerant of a wide range of temperatures for growth, from below OC (Siegel et al., 1984) to above 40C
(Wegmann et al., 1980). The optimum growth temperature lies between
21 and 40C, and is dependent on the strain and light intensity (Gibor,
1956; Federov et al., 1968; Borowitzka, 1981). Temperatures close to
40C and higher stimulate carotenogenesis at the expense of growth
(Mil'ko, 1963). In practice, the combination of light intensities, day
length and temperatures between 30 and 40C give optimum growth and
fj-carotene production at Western Biotechnology's Australian plant in
the summer months, although growth and fj-carotene production are
sustained throughout the year.
Carbon source. Dunaliella salina is a photoautotroph; it can use only CO2
and possibly bicarbonate (HC03 ) as carbon source. Growth of D. salina
is stimulated by addition of NaHC0 3 or bubbling with COz, provided
there is no precipitation of carbonates (Massyuk, 1965; Drokova &
Dovhorouka, 1966). Limited availability of inorganic carbon (C0 2 and
HC0 3 ) is the most common growth limiting factor in the high salinity,
high pH and temperature conditions used for D. salina culture.
Inorganic carbon equilibria and uptake by D. salina under typical
growth conditions are discussed by Borowitzka & Borowitzka (1988a).
Nitrogen and phosphorus sources. The best nitrogen source for D. salina
is NaN0 3 (Mil'ko, 1962; Grant, 1968); ammonium salts, urea and
nitrites are less effective (Gibor, 1956; Mil'ko, 1962; Borowitzka &
Borowitzka, 1988b). Care must be taken to maintain culture pH; despite
natural buffering conferred by the inorganic carbon equilibrium
(CO~-~HC03~C02)' nitrate uptake can cause alkalinisation, and
ammonium uptake acidification of the growth medium, leading to cell
death (Borowitzka & Borowitzka, 1988b). The optimum phosphate
concentration for growth is approximately 002 j.lg litre- 1 K 2HP0 4
(Gibor, 1956); higher concentrations can inhibit growth.
NaCI. All known marine and halophilic species of Dunaliella have a
specific requirement for sodium; D. salina has the broadest reported
NaCI tolerance range, from 1 to 35% (w/v) NaCI (saturation) (Loeblich,
1972). Selection of the appropriate NaCI concentration for the commercial culture depends on growth or carotenogenesis rate required, and the
need (if any) to control predators, competitors or bacterial numbers
(Borowitzka et al., 1986). Normal salinities used for growth range from
20 to 30% (w/v) NaCI, i.e. about 9-10 times the salinity of sea-water.
Other requirements. Dunaliella salina requires Mg2+, Ca2+, Na+, CI-,
SO~-, chelated iron and trace elements for growth. Optimum pH is
between 7 and 9. A discussion of the literature reports on concentration
and ratios of these components is given in Borowitzka and Borowitzka
(1988a). No requirement for vitamins has been demonstrated.


L. 1. BOTowitzka & M. A. BOTowitzka


Commercial D. salina production is a young industry, only 3 years old, and

companies in development and production phases guard the confidentiality of
their processes. In particular, because harvesting and extraction represent
major cost areas, technical advantages in these areas are eagerly sought and
strongly guarded. Harvesting D. salina is more difficult and costly compared to
most commercial algae. Dunaliella salina is a single cell, with no protective cell
wall, approximately 20 x 10 /lm in size, and neutrally buoyant in a high specific
gravity, high viscosity brine. Cell densities in large cultures tend to be about
1 g litre- 1 and very large volumes therefore have to be processed. Efforts to
centrifuge or filter the algae from the brine generally shear-damage the cells
leading to tJ-carotene loss by oxidisation. The cells also distort and pass
through filters with pore sizes less than 10 /lm. Corrosion of all metal
equipment by the brine is also a major problem.
These problems have led to development and patenting of several biological
and chemical methods for harvesting the cells, or for pre-concentration of cells
prior to use of more conventional harvest methods. Patents issued for
Dunaliella harvesting include high pressure filtration using diatomaceous earth
(Ruane, 1974a), exploitation of salinity-dependent buoyancy properties in
stationary and moving gradients (Bloch et al., 1982), exploitation of the
phototactic and gyrotactic responses of the algae (Kessler, 1982, 1985)
salinity-dependent hydrophobic adhesion properties of Dunaliella cells (Curtain & Snook, 1983) and flocculation (Sammy, 1987). The harvesting of
micro algae is reviewed by Mohn (1988).
The process for extraction of tJ-carotene from the biomass depends in part
on the harvesting procedure used, and on the market requirements. Extraction
using conventional organic solvents is efficient (Ruane, 1974b), but may not be
acceptable to customers seeking a 'natural' product. More acceptable alternative extraction methods use hot vegetable oil (Nonomura, 1987; Potts, 1987) or
liquid or supercritical gas solvents.
Harvested biomass may be dried, (i.e. spray-drying) rather than extracted,
and is marketed as a tJ-carotene-rich supplement for human health or animal
feed (Ben-Amotz et al., 1986; Avron et al., 1987; Curtain et al., 1987).
By-products of tJ-carotene production could include glycerol, which can
make up 30% of biomass dry weight (see Chen & Chi, 1981, for process
proposal and cost analysis) and high quality protein meal remaining after
extraction. The amino acid analysis of several Dunaliella spp is summarised in
Table 2.2. in Borowitzka and Borowitzka (1988a).
tJ-Carotene is marketed mainly as an orange colouring, for foods such as
margarine, baked goods, and beverages (KUiui, 1981). A smaller market exists

fJ-Carotene Production with Algae


for use of {:1-carotene as a pigment supplement in the commercial feeds for

crustaceans, such as prawns, which convert it to astaxanthin, the natural red
colouring of the shell and flesh.
{:1-Carotene is provitamin A, and provides the main source of vitamin A in
human diets. More recently, interest has increased in the role played by
{:1-carotene as an anti-oxidant in human nutrition and correlations have been
reported between high dietary {:1-carotene intake and low incidences of some
cancers (Mathews-Roth, 1982; Schwartz et aZ., 1986). Large-scale trials are
underway in the US (National Cancer Institute) and in Europe to confirm the
'anti-cancer' activity of {:1-carotene.
Trials with cattle have also indicated an increase in fertility attributable to
supplements of {:1-carotene.


The first commercial synthesis of f3-carotene was established by Hoffman-La

Roche in 1956. Since then, the process has been improved by Roche and other
companies; patents describe the key steps in the synthesis. Syntheses of
{:1-carotene are reviewed by Mayer & Isler (1971).


{:1-Carotene is a lipid- and oil-soluble product. As suspensions or solutions in

vegetable oil (Fig. 2) it finds applications in colouring margarine, baked goods
and some prepared foods. Conversion to water-soluble or water-dispersible
formulations, by forming emulsions or microencapsulated beadlets extends the
food applications to beverages including orange drinks, and to confectionery
and further prepared foods.
Nutritional supplements can be prepared by encapsulation of oil suspensions
or solutions, or by tableting using beadlet forms. Current recommended daily
doses for {:1-carotene are around 6 mg; 20-25 mg doses are in use in the
'anti-cancer' trials in the US and Finland. Dried Dunaliella powder is also
prepared for use in animal feed supplements.
Just as the companies in the new industry guard the confidentiality of their
processes, costs of commercial production are not available. Moulton et aZ.
(1987a), however, present a bioeconomic model for the {:1-carotene production
process using D. salina.
Synthesised {:1-carotene is sold for a minimum of US$300 per kg {:1-carotene,
and at higher prices depending on the formulation. 'Natural' {:1-carotene



I. Borowitzka & M. A. Borowitzka

commands higher prices, with the highest price attainable for the nutritional
supplement application. Market price for 'natural' IJ-carotene is volatile and
in 1987-88 world market demand exceeds supply.
If a price range for formulations of natural IJ-carotene of US$500-1000 is
assumed, and IJ-carotene represents 10% of D. salina biomass, production
costs for D. salina should not exceed US$50-100 per kg. In fact, costs must be
significantly lower, to account for losses at each processing step, capital
expenses, marketing, packaging and distribution costs.
Aasen, A. J., Eimhjellen, K. E. & Liaaen-Jensen, S. (1969). An extreme source of
p-carotene, Acta Chim. Scand., 23,2544-5.
Avron, M., Ben-Amotz, A. & Edelstein, S. (1987). Feed supplement. UK patent
application 2 189 675A.
Bauernfeind, J. C. (ed.) (1981). Carotenoids as Colourants and Vitamin A Precursors.
Academic Press, New York, pp. 938.
Ben-Amotz, A. (1980). Glycerol production in the alga Dunaliella. In Biochemical and
Photosynthetic Aspects of Energy Production, ed. A. San Pietro. Academic Press,
New York, pp. 191-208.
Ben-Amotz, A. & Avron, M. (1982). The potential use of Dunaliella for the production
of glycerol, p-carotene and high protein feed. In Biosaline Research: A Look to the
Future, ed. A. San Pietro. Plenum Publishing Corporation, New York, pp. 207-14.
Ben-Amotz, A. & Avron, M. (1983). On the factors which determine the massive
p-carotene accumulation in the halotolerant alga Dunaliella salina. Plant Physiol.,
Ben-Amotz, A., Katz, A. & Avron, M. (1982). Accumulation of p-carotene in
halotolerant algae: purification and characterisation of p-carotene-rich globules
from Dunaliella bardawil (Chlorophyceae). J. Phycol., 18,529-37.
Ben-Amotz, A., Edelstein, S. & Avron, M. (1986). Use of the p-carotene rich alga
Dunaliella bardawil as a source of retinol. Br. Poult. Sci., 27,613-9.
Bloch, M. R., Sasson, J., Ginzburg, M. E., Goldman, Z., Ginzburg, B. Z., Garti, N.
& Perath, A. (1982). Oil products from algae. US patent no. 4.341 038.
Borowitzka, L. J. (1981). The microflora. Adaptions to life in extremely saline lakes.
Hydrobiologia, 81, 33-46.
Borowitzka, M. A. (1988). Algal growth media and sources of algal cultures. In
Micro-algal Biotechnology, ed. M. A. Borowitzka & L. J. Borowitzka, Cambridge
University Press, Cambridge pp. 456-65.
Borowitzka, L. J., Borowitzka, M. A. & Moulton, T. P. (1984). The mass culture of
Dunaliella salina for fine chemicals: from laboratory to pilot plant. Hydrobiologia,
116/117, 115-2l.
Borowitzka, L. J., Moulton, T. P. & Borowitzka, M. A. (1986). Salinity and the
commercial production of beta-carotene from Dunaliella salina. Nova Hedwigia,
Beih., 83, 224-9.
Borowitzka, M. A. & Borowitzka, L. J. (1988a) Dunaliella. In Microalgal
Biotechnology, ed. M. A. Borowitzka & L. J. Borowitzka. Cambridge University
Press, Cambridge, pp. 27-58.
Borowitzka, M. A. & Borowitzka, L. J. (1988b) Limits to growth and carotenogenesis
in laboratory and large-scale cultures of Dunaliella salina. In Algal Biotechnology,
ed. T. Stadler, J. Mollion, M-C. Verdus, Y. Karamanos, H. Morvan & D.
Christiaen. Elsevier Applied Science Publishers, Barking, pp. 371-82.

fJ-Carotene Production with Algae


Chen, B. J. & Chi, C. H. (1981). Process development and evaluation for algal glycerol
production. Biotech. Bioeng., 23, 1267-87.
Craig, R., Reichelt, B. Y. & Reichelt, J. L. (1988). Genetic engineering of
micro-algae. In Micro-algal Biotechnology, ed. M. A. Borowitzka & L. J.
Borowitzka. Cambridge University Press, Cambridge, pp. 415-55.
Curtain, C. C. & Snook, H. (1983). Method for harvesting algae. US patent no. 511
Curtain, C. c., West, S. M. & Schlipalius, L. (1987). Manufacture of p-carotene from
the salt lake alga Dunaliella salina; the scientific and technical background. Aust. J.
Biotechnol., 1, 51-7.
Drokova, I. G. & Dovhorouka, S. I. (1966). Carotene-formation in Dunaliella salina
Teod. under the effect of some carbon sources. Ukransk. Bot. Zhour., 21, 59-62.
Federov, V. D., Maksimov, V. N. & Kromov, V. M. (1968). Effect of light and
temperature on primary production of certain unicellular green algae and diatoms.
Fiziol. Rast., 15, 640-5l.
Gibor, A. (1956). The culture of brine algae. Bioi. bull., Woods Hole, 3,223-9.
Grant, B. R. (1968). The effect of carbon dioxide concentration and buffer system on
nitrate and nitrite and assimilation by Dunaliella tertiolecta, J. Gen. Micro., 54,
Kessler, J. O. (1982). Algal cell harvesting. US patent no. 4324067.
Kessler, J. O. (1985). Hydrodynamic focusing of motile algal cells. Nature, 313,
KI1tui, H. (1981). Industrial and commercial uses of carotenoids. In Carotenoid
chemistry and Biochemistry, ed. G. Britton & T. W. Goodwin. Pergamon Press,
Oxford, pp. 309-28.
Klausner, A. (1986). Algaculture: Food for thought. Biotechnology, 4,947-53.
Lerche, W. (1937). Untersuchungen iiber die Entwicklung und Fortpflanzung in der
Gattung Dunaliella. Arch. fur Protistenk., 88, 236-9.
Loeblich, L. A. (1972). Studies on the brine flagellate Dunaliella salina. PhD thesis,
University of California, San Diego.
MacKinney, G. & Chichester, C. O. (1960). Biosynthesis of carotenoids. In
Comparative Biochemistry of Photoreactive Systems, ed. M. B. Allen. Academic
Press, New York, pp. 205-15.
Massyuk, N. P. (1965). Carbonate and bicarbonate as stimulators of growth and
carotene accumulation in Dunaliella salina Teod. Ukransk. Bot. Zhour., 22, 18-22.
Massyuk, N. P. (1966). Mass culture of the carotene-bearing alga Dunaliella salina
Teod. Ukransk. Bot. Zhour., 23, 12-19.
Massyuk, N. P. & Abdula, E. G. (1969). First experiment of growing carotenecontaining algae under semi-industrial conditions. Ukransk. Bot. Zhour., 26,21-7.
Mathews-Roth, M. M. (1982). Medical applications and uses of carotenoids. In
Carotenoid Chemistry and Biochemistry, ed. G. Britton & T. W. Goodwin.
Pergamon Press, Oxford, pp. 297-307.
Mayer, H. & Isler, O. (1971). Total syntheses. In Carotenoids, ed. O. Isler.
Birkhauser, Basle, pp. 328-575.
Mil'ko, E. S. (1962). Study the requirement of two Dunaliella spp. in mineral and
organic components if the medium, Moscow University Vestnik. Biologya, 6, 21-3.
Mil'ko, E. S. (1963). Effect of various environmental factors on pigment production in
the alga Dunaliella salina. Mikrobiologiya, 32,299-307.
Mironyuk, V. I. & Einor, L. O. (1968). Oxygen exchange and pigment content in
various forms of Dunaliella salina Teod. under conditions of increasing NaCl
content. Gidrobiol Zhournal, 4,23-9.
Mohn, F. H. (1988). Harvesting of micro-algal biomass. In Micro-algal Biotechnology,
ed. M. A. Borowitzka & L. J. Borowitzka. Cambridge University Press, pp.


L. J. Borowitzka & M. A. Borowitzka

Moulton, T. P., Borowitzka, L. J. & Vincent, D. J. (1987a) The mass culture of

Dunaliella salina for p-carotene: from pilot plant to production plant,
Hydrobiologia, 151/152,99-105.
Moulton, T. P., Sommer, T. R., Burford, M. A. & Borowitzka, L. J. (1987b).
Competition between Dunaliella species at high salinity. Hydrobiologia, 151/152,
Nelis, H. J. C. F. & De Leenheer, A. P. (1983). Isocratic nonaqueous reversed-phase
liquid chromatography of carotenoids. Analyt. Chem., 55,270-75.
Nonomura, A. M. (1987). Process for producing a naturally derived carotene-oil
composition by direct extraction from algae. US patent no. 4,680,314.
Potts, W. T. (1987). Extraction of carotenoid pigments from algae. Australian patent
application no. 69260/87.
Rich, V. (1978). Israel's place in the sun. Nature, 275, 581-2.
Ruane, M. (1974a). Recovery of algae from brine suspensions. Australian patent no.
Ruane, M. (1974b). Extraction of caroteiniferous materials from algae. Australian
patent no. 487 018.
Sammy, N. (1987). Method for harvesting algae. Australian patent application no.
Schwartz, J., Suda, D. & Light, G. (1986). Beta-carotene is associated with the
regression of hamster buccal puch carcinoma and induction of tumor necrosis factor
in macrophages. Biochem. biophys. Res. Commun., 136, 1130-5.
Semenko, V. E. & Abdullayev, A. A. (1980). Parametric control of p~carotene
biosynthesis in Dunaliella salina cells under conditions of intensive cultivation.
Fiziol. Rast., 27,31-41.
Siegel, B. Z., Siegel, S. M., Speitel, T., Waber, J. & Stoeker, R. (1984). Brine
organisms and the question of habitat-specific adaption. Origins of Life, 14, 757-70.
Spencer, J. F. T. & Spencer, D. M. (1983). Genetic improvements of industrial yeasts.
Ann. Rev. Microbiol., 37, 121-42.
Teodoresco, E. C. (1905). Organisation et developpement du Dunaliella nouveau genre
de Volvocacee-Polyblepharidee. Bot. Zentralblatt, Beih., 18,215-32.
Wegmann, K., Ben-Amotz, A. & Avron, M. (1980). Effect of temperature on glycerol
retention in the halotolerant algae Dunaliella and Asteromonas. Plant. Physiol., 66,

Chapter 3


Departamento de Genetica y Biotecnia, Universidad de Sevilla, Apartado 1095,
E-41080 Sevilla, Spain


Humans, like most other animals, need carotenoids but cannot synthesize
them. Our main suppliers are the fruits and vegetables, with lesser, mostly
indirect contributions from fungi, algae, and bacteria. Carotenoid production
is rarely the main objective of a culture, but colourful products rich in
carotenoids are obtained from various plants and algae. The oldest example is
saffron (Crocus sativus) , still planted in La Mancha, Spain, and elsewhere for
the stigmas of its flowers. Many fields of central Mexico become bright orange
in the summer with the flowers of cempasuchil (Tagetes ten uifolia , a kind of
marigold), whose dried petals are fed to chicken for meat and egg yolk colour.
Annatto is extracted from the seeds of a tropical tree, Bixa orellana. The most
familiar carotenoid-rich product is probably paprika, a powder made by
grinding red peppers (Capsicum annuum).
The carotenoids may be scarce in the diets of humans and animals for
various reasons. Many staple foods and feeds, including most of the cheap
sources of carbohydrates and fat, are poor in carotenoids, as witnessed by their
lack of colour. The long winters of many countries halt the natural production
of carotenoids, and these are relatively unstable and may not survive in stored
foods and feeds.
Organic chemists have devised several complete syntheses, which now
supply considerable amounts of f3-carotene and other purified carotenoids. The
current trend towards natural food colours and the discovery of salutary effects
of the carotenoids, beyond their important role as provitamin A, stimulate the
demand and press towards a biological production.
The fungi can hardly be considered as traditional food colourants, but the
relative ease of cultivation and the many possibilities for physiological,
genetical, and industrial manipulations make them attractive to biotechnologists. Several species are potential sources of carotenoids, and a largescale process has been based on Blakeslea trispora.


E. Cerda-Olmedo

It seems that the Soviet Union is the only country that uses Blakeslea
trispora for the industrial production of carotenoids; the yearly production is
estimated at about 200 kg purified fJ-carotene and another 4 Mg fJ-carotene in
the form of mycelia used as an additive to animal feed (A. A. Dmitrovskii,
pers. comm., 1988).
The abundant literature on carotenoids has been the subject of many books
and reviews. The monumental and indispensable treatise edited by Isler (1971)
is complemented by newer reviews (Feofilova, 1974; Goodwin, 1976, 1980,
1988). The practical aspects are explained in great detail by Bauernfeind
(1981). Updates are provided by the International Symposia on Carotenoids,
published every 3 years as a special issue of Pure and Applied Chemistry.
There are specialized and recent reviews on the influence of external factors on
carotenogenesis in the Mucorales (Lampila et al., 1985a), on Phycomyces
(Cerda-Olmedo & Lipson, 1987), on the regulation of carotenogenesis
(Bramley & Mackenzie, 1988), on the metabolism and metabolic effects of the
carotenoids (Goodwin, 1986), and many other subjects.
Most fungi produce no carotenoids at all and many of the others contain a
single major carotenoid. Although the pathway intermediates may be present
in amounts widely variable with the strain and the culture conditions, the
carotenoid composition of the fungi is usually much simpler than that of the
photosynthetic organisms. Here are a few selected examples of fungal
carotenoids. fJ-Carotene predominates in the Mucorales Blakeslea trispora,
Phycomyces blakesleeanus and Choanephora cucurbitarum, in the yeast
Rhodotorula aurea, in Aspergillus giganteus (EI-Jack et al., 1988) and various
species of Penicillium; neurosporaxanthin is the end-product of Fusarium
aquaeductuum (Bindl et al., 1970); neurosporaxanthin and fJ-carotene, those
of Gibberella fujikuroi (Avalos & Cerda-Olmedo, 1987), Neurospora crassa
and Verticillium agaricinum (Valadon & Mummery, 1973); torularhodin and
fJ-carotene, those of the yeasts Rhodotorula rubra, R. minuta, and
Rhodosporidium diobovatum; fJ-carotene, lycopene, and fJ-zeacarotene are
found in different strains of the smut Ustilago violacea (Will et al., 1984);
canthaxantin in the mushroom Cantharellus cinnabarinus; astaxanthin, in the
yeast Phaffia rhodozyma. See Goodwin (1980) for other fungi and additional
Figure 1 shows the structure of the carotenoids that have just been
mentioned, with their likely biosynthetic pathways. The carotenoids, like all
terpenoids, are synthesized from hydroxymethylglutaryl-coenzyme A, which is
first converted to mevalonic acid. The specific part of the pathway begins with
the condensation of two molecules of geranylgeranyl pyrophosphate to form
phytoene, a colourless carotene. Four dehydrogenations transform phytoene
into lycopene, the pigment of red tomatoes. Two cyclisations convert lycopene



f) -








tions, the vertical arrows, cyclisations, and the tilted arrows, oxidative reactions.

Fig. 1. Some fungal carotenoids with their possible biosynthetic pathways. The horizontal arrows indicate dehydrogena-

----.. ------ ~ ~







E. Cerda-Olmedo

into p-carotene. In Phycomyces these reactions are carried out by an enzyme

aggregate that contains four copies of a dehydrogenase and two copies of a
cyclase (review, Cerda-Olmedo, 1987). In Giberellafujikuroi the production of
torulene requires a fifth dehydrogenation by the same enzyme; the final break
and oxidation to neurosporaxanthin appear to be carried out in a single step
(Avalos & Cerda-Olmedo, 1987). On the other hand, several dehydrogenases
have been postulated for Ustilago violacea (Will et al., 1984). The detection of
some minor intermediates indicates that the order of the reactions is
sometimes altered in these organisms, and may be different in others.
Canthaxantin, a likely intermediate in the biosynthesis of astaxanthin in the
crustaceans (Castillo, 1980), was not detected in Phaffia rhodozyma
(Andrewes et al., 1976).
Many carotenoids found in plants and algae are unknown in the fungi, but
most of the fungal carotenoids are also found in plants or algae, and this is
fortunate for fungal biotechnology, because the traditional ingredients of our
diet are more acceptable to the health authorities and to the consumers than
new products. A few carotenoids are exclusive of certain fungi: torulene,
neurosporaxanthin, torularhodin, plectaniaxanthin, phillipsiaxanthin, p-ycarotene. The astaxanthin from Phaffia rhodozyma has the opposite chirality
(3R, 3'R) to that of the astaxanthin responsible for the admired colours of
salmon and lobster (Andrewes & Starr, 1976).
Research towards industrial production has concentrated on the mucorales
Blakeslea trispora and Phycomyces blakesleeanus, whose main and nearly only
carotenoid is all-trans p-carotene. Commercial advantages of p-carotene are its
colour, its high provitamin A activity, its wide distribution in nature, and its
many current applications. A disadvantage is its inability to pigment some
avian and fish products.


Many fungi attracted the attention of carotenoid chemists because of their
bright colours, but the strains used in the laboratories may not be optimal for
carotenoid production.
Thus, an impressive polymorphism was found in natural isolates of Ustilago
violacea (Garber et al., 1978). Fourfold differences in carotenogenesis were
found in a few natural isolates of Phycomyces blakesleeanus (Goodwin &
Griffiths, 1952). A large systematic search for better starting strains should
be rewarding, as it was for many other biological products.
The carotenoid concentrations are usually too low for commercial application. The accumulation of carotenoids in the fungi is influenced by many
environmental variables; their study is a first step in yield improvement. While
so doing, one should bear in mind that the fungi are very distantly related
organisms, that the regulations of carotenogenesis are not universal, and
therefore, that observations on a species need not apply to others.


Production of Carotenoids with Fungi

3.1 Light
Blue illumination stimulates the production of carotenoids in many fungi, as it
does in various bacteria, algae, and plants (reviews, Harding & Shropshire,
1980; Rau, 1983; Rau & Schrott, 1987). This phenomenon (photocarotenogenesis) does not occur in many fungi, for example, in Blakeslea
trispora (Sutter, 1970).
Photocarotenogenesis presumably protects the cells against the deletereous
effects of very strong illuminations and avoids waste in the dark or in dim light,
but is of limited interest to biotechnologists, because bright illumination is not
practical in large cultures and the final concentrations of carotenoids are not
very high. The examples given in Table 1 prove this point, but should not be
compared with each other, because they were obtained under very different
conditions, most under non-saturating light exposures.

3.2 Sexual Interaction

In many species of the Mucorales the contact zone of mycelia of opposite sex
becomes bright yellow (Blakeslee, 1904). The sexual stimulation is easily
observed in mated cultures (mixed cultures of strains of opposite sex) (Barnett
et al., 1956; Anderson et al., 1958). The beta-carotene content increases 5-15
times over the level found in separate cultures, according to various reports.
Sexual stimulation in the Mucorales occurs in the absence of physical contact
when the interacting hyphae are separated by a membrane (Burgeff, 1924). An
enhanced carotenogenesis occurs in strains of Blakeslea trispora grown under
such conditions (Barnett et al., 1956). A race between chemists of several
major biotechnological companies led to the isolation of chemicals, the
trisporic acids, present in mated cultures and able to stimulate carotenogenesis
in single cultures (Caglioti et al., 1966). The trisporic acids are industrially
useless because the increases in p-carotene content do not compensate their
Table 1
Examples of Photocarotenogenesis in some Fungi

Name of
Aspergillus giganteus
Fusarium aquaeductuum
Gibberella fujikuroi
Neurospora crassa
Phycomyces blakesleeanus
Rhodosporidium diobovatum
Rhodotorula minuta
Verticillium agaricinum

Coloured carotenoids
in illuminated cultures
(Ilg/gt Main component




EI-Jack et al., 1988

Bindl et al., 1970
Avalos & Cerda-Olmedo, 1987
Harding et al., 1969
Bergman et al., 1973
Vaskivnyuk & Getman, 1984
Tada & Shiroishi, 1982
Valadon & Mummery, 1973

The numerical values are the total concentrations of the carotenoids detected in
illuminated cultures, in Ilg/g dry mycelial weight.

E. Cerd4-01medo


Since trisporic acids are made from p-carotene (Austin et a/., 1970), the
activation of carotenogenesis by trisporic acids in the Mucorales provides a
regulatory loop for increased sexual stimulation. Another likely role of the
newly synthesized p-carotene is to make sporopollenins, tough polymers found
in the cell walls of the zygospores (Gooday et a/., 1973) and of other
structures. Lack of trisporic acids and sporopollenins explains the sexual
impotence of mutant Mucorales devoid of p-carotene (Sutter, 1975; Khabrova
& Zhdanov, 1979), but not that of the carS mutants, which overproduce it. The
trisporic acids have no effect on carotenogenesis or morphogenesis in fungi
other than the Mucorales.

Culture Conditions

The accumulation of carotenoids is influenced by all sorts of variations in the

culture conditions, but usually only to a minor extent. The best temperatures
for carotenogenesis in B/akeslea trispora and Phycomyces b/akes/eeanus are
about 26C. Acetate, leucine, and other amino acids increase the p-carotene
content of Phycomyces, presumably because they can be converted directly to
hydroxymethylglutaryl-coenzyme A (Lilly et al., 1960; Cerda-Olmedo, 1987).
Many of the conditions tried by Lilly and co-workers allowed Phycomyces to
exceed 2 mg p-carotene per g dry weight.
The optimization of the media for the industrial production of p-carotene
with B/akes/ea trispora was the central concern of a series of successful
investigations (Anderson et a/., 1958; Ciegler et al., 1959a, b, 1962). The best
media are rich in various grain products, vegetable oils, and hydrocarbons
(kerosene, preferably deodorized).
Other fungi have received much less attention. A medium with tomato
pressings increased the astaxanthin content of Phaffia rhodozyma to 08 mg/g
(Johnson & Lewis, 1979). An appropriate carbon-to-nitrogen ratio and a
surface-active agent allowed Neurospora crassa to make over 25 mg carotenoids per g dry weight (Krzeminiski & Quackenbush, 1960).

Mycelial colour changes facilitate the detection of chemicals that influence
carotenogenesis in the fungi. The many inhibitors and the weak activators do
not interest us now. The accumulation of p-carotene in the Mucorales is
stimulated by p-ionone (Mackinney et a/., 1952, 1953; Reyes et a/., 1964) and
retinol (Eslava et a/., 1974; Feofilova & Bekhtereva, 1976). These molecules
are identical to p-carotene in the p-ring and the adjacent atoms, but they are
not incorporated into it (Engel et a/., 1953; Eslava et a/., 1974). Some
variations of these structures are active, and others are not (Mackinney et a/.,
1952; Engel et a/., 1953; De la Concha & Murillo, 1984; Bejarano et a/., 1988).
p-Ionone and retinol are kept away from industrial fermentors by their high

Production of Carotenoids with Fungi


prices. In the production of fJ-carotene by Blakeslea trispora, fJ-ionone may be

replaced by citrus pulps (alkali-heated peels, seeds, and rags) (Ciegler et al.,
1963b) and by limonene and citrus oils that contain limonene (Ciegler et al.,
Many aromatic compounds activate carotenogenesis in Phycomyces. These
compounds stimulate the pathway and inhibit phytoene dehydrogenase to
different extents. Predominance of the stimulation (observed with dimethyl
phthalate, 1,2-dimethoxy-4-propenylbenzene, 4-bromoveratrol, veratrol, and
others) results in vast accumulations of fJ-carotene. Mixtures of fJ-carotene and
partially desaturated acyclic intermediates result when both activities are
prominent (cinnamic alcohol, eugenol, safrol, 1,3-dimethoxybencene, 4methoxyphenol, etc.) (Cerda-Olmedo & Hllttermann, 1986). These intermediates are undesirable in the industrial production of fJ-carotene, but may be of
interest for research and other uses.
The simple comparison of the structural formulae is insufficient to establish
functional analogy between chemical activators. A genetic test for functional
analogy has been made possible by the availability of many mutants of
Phycomyces blakesleeanus with various defects in carotene synthesis and its
regulation. If mutants of a certain gene are activated by a chemical but not by
another, the mechanisms of action of the two chemicals differ in their need for
the product of that gene. The activators were thus classified into four groups:
light, trisporic acids, fJ-ring containing chemicals (retinol, fJ-ionone), and
phenols (dimethyl phthalate, veratrol) (Bejarano et aI., 1988; Bejarano &
Cerda-Olmedo, 1989). It is interesting to note that fJ-ionone, trisporic acids,
and phenols are not analogues of each other.
A more classical way to establish functional groups is to determine
synergisms. Thus, the joint application of activators belonging to different
groups leads to higher carotene contents than may be attained with either
alone (Govind & Cerda-Olmedo, 1986; Bejarano et al., 1988).
2,6,6-Trimethyl-1-acetylcyclohexene and some related compounds satisfactorily replace fJ-ionone in industrial media for Blakeslea. Other chemicals
(isoniazid, iproniazid) are synergistic with them (Ninet et aI., 1969). Many
herbicides are inhibitors of carotenogenesis, but propanil (N-(3,4dichlorophenyl)propanamide) is an activator (Feofilova et al., 1982).
The addition of microbial biomass to mated Blakeslea cultures stimulates
carotenogenesis (Ciegler et al., 1964). About equally effective are the spent
Blakeslea mycelia, after the extraction of fJ-carotene in previous fermentations, and cells of various bacteria, yeasts, and fungi.
Although over 100 chemical activators of carotenogenesis are known, most
are relatively ineffective, or too expensive, or too toxic for practical application. The search for activators is far from complete, and the spot test
(Cerda-Olmedo & Hllttermann, 1986) should permit the rapid screening of
new ones. The results for one fungus are often invalid for another: fJ-ionone
stimulates carotenogenesis in Blakeslea and Phycomyces, but inhibits it in
Gibberella (Avalos & Cerda-Olmedo, 1986); isoniazid is active for Blakeslea
but not for Phycomyces.

E. Cerda-Olmedo



5.1 Phycomyces
The genetic analysis of carotenogenesis in Phycomyces blakesleeanus (review,
Cerda-Olmedo, 1987) was made possible by the study of its sexual cycle and its
genetics (review, Eslava, 1987), the isolation of numerous car mutants with
alterations in the biosynthesis of carotene and its regulation (review, CerdaOlmedo, 1985) and the design of procedures to bypass their sexual disturbances (Roncero & Cerda-Olmedo, 1982).
Sexual stimulation can be obtained in a single mycelium with a simple
genetic trick. Mycelia containing a mixture of nuclei of opposite sex (intersexual heterokaryons) accumulate about 10 times more p-carotene than
mycelia containing one or the other kind of nuclei (homokaryons) (Murillo &
Cerda-Olmedo, 1976). Intersexual heterokaryons express full sexual stimulation of carotenogenesis: their p-carotene cannot be further increased by the
addition of trisporic acids (Govind & Cerda-Olmedo, 1986). When the
constituent nuclei have different genetic backgrounds, the intersexual heterokaryons tend to sector out, i.e. to produce mycelia of various nuclear
proportions and even pure homokaryons. Consaguineous strains, such as the
ones developed by Alvarez & Eslava (1983), reduce this danger, but the best
solution is a system of balanced lethals, that is, the induction of a recessive
lethal mutation in each of the nuclei. This is much easier than it sounds
(Murillo et al., 1978).
Attempts to block carotenogenesis in Phycomyces result in a general
activation of the pathway that tends to maintain the concentration of the final
product, p-carotene. Thus, the addition of an enzyme inhibitor (such as
diphenylamine, which blocks phytoene dehydrogenation) causes an accumulation of intermediates (phytoene and other partially desaturated acyclic
carotenes); the p-carotene content is affected only at high inhibitor concentrations. The same is true for inhibitors of lycopene cyclization (nicotine,
2-(4-chlorophenylthio)triethylamine), with accumulation of lycopene and
y-carotene, and for mutations that reduce either enzyme activity. Full chemical
or mutational blocks cause a 50-fold increase in carotene content; the product,
of course, is no longer fJ-carotene, but the untransformed intermediate,
phytoene or lycopene (Eslava & Cerda-Olmedo, 1974; Murillo & CerdaOlmedo, 1976; Murillo et al., 1981).
This end-product inhibition of the pathway is mediated by p-carotene and
two gene products (Cerda-Olmedo, 1987; Bejarano et al., 1988), as summarized in Fig. 2. Retinol and beta-ionone, which compete with beta-carotene,
and recessive carS mutations, which destroy a dispensable gene product,
abolish the inhibition and lead to high p-carotene contents. The best carS
mutants contain 100 times more fJ-carotene than the wild type.
Other superproducing strains, containing up to 20 times more p-carotene
than the wild type, carry recessive mutations in another gene, carD. Mutations


Production of Carotenoids with Fungi











" ~









Fig. 2. The genetics of carotenogenesis in Phycomyces blakesleeanus (Cerda-Olmedo,

1987). Four copies of phytoene dehydrogenase (the product of gene carB) and two
copies of lycopene cyclase (one of the products of gene carRA, closely linked to carB)
convert phytoene into beta-carotene (above). The product of gene carS and a product
of gene carRA combine with beta-carotene to repress the synthesis (center). Two other
genes, care and carD, regulate in opposite ways the action of carS (below).

in gene carS are epistatic over those in gene carD: the double mutants do not
surpass single carS mutants and are, therefore, of no practical value. There are
superproducing strains that considerably exceed the J3-carotene content of the
carS mutants, but they have not yet been published.
The highest published J3-carotene level in Phycomyces blakesleeanus
corresponds to intersexual heterokaryons with carS mutations and balanced
lethals (Murillo et aI., 1978). They contain 25 mg J3-carotene per g dry weight,
some 500 times more than the wild type. Their genetic make-up provides
constitutive sexual activation and removal of end-product inhibition, and
therefore, they do not respond to retinol or J3-ionone. They do not respond to
light or dimethyl phthalate either, although no provision is made for
endogenous activation of the mechanisms normally triggered by these agents.


E. Cerda-Olmedo

When certain Phycomyces strains are grown in the dark, most of the
fJ-carotene appears in a red complex, similar to the main carotenoprotein of
many plants, which stabilizes fJ-carotene and changes its colour in an attractive
way. These strains contain a carS mutation and a cytoplasmic carE mutation
(De la Concha & Murillo, 1984). They accumulate about 100 times more
fJ-carotene than the wild-type and no attempt has been made to improve
their yield.
Some genetic constructs contain considerable amounts of pathway intermediates. Lycopene is accumulated by carR mutants, defective in lycopene
cyclase. Constitutive sexual activation and partial removal of end-product
inhibition led to strains with about 15 mg lycopene per g dry weight (Murillo et
al., 1978). A high phytoene content (9 mg/g dry weight) was obtained by
growing a carB mutant, defective in phytoene dehydrogenase, with dimethyl
phthalate in the light (Bejarano & Cerda-Olmedo, 1989). Other intermediates can only be produced as part of carotene mixtures: phytofluene and
~-carotene in some leaky carB mutants (Eslava & Cerda-Olmedo, 1974;
Bejarano et al., 1987); y-carotene in certain heterokaryons (De la Guardia et
al., 1971).
5.2 Other fungi
The genetic analysis of carotenogenesis is feasible in many fungi, but has been
carried out in few, and only to limited extents (see, for example, Khabrova &
Zhdanov, 1979, for Blakeslea trispora; Harding & Turner, 1981, for
Neurospora crassa; Will et al., 1984, for Ustilago violacea; Avalos &
Cerda-Olmedo, 1987, for Gibberella fujikuroi). Certain mutants of Gibberella
are deep orange even when grown in the dark and contain up to 1.7 mg
neurosporaxanthin per g dry weight, but they are not analogous to Phycomyces
mutants because there is no end-product regulation in Gibberella.
6.1 Blakeslea
The only fungal production of carotenoids that has been developed to the
industrial stage is that of fJ-carotene with mated cultures of Blakeslea trispora.
After detailed studies of media and additives, Ciegler and his co-workers at the
US Department of Agriculture scaled up their cultures to a volume of 11 litres
(Ciegler et al., 1963a). The process has been reviewed in detail (Ciegler, 1965;
Hanson, 1967). The production phase (Ciegler et al., 1963a) consists in the
joint culture at 28C for 3 days of mycelia of two wild types of opposite sex in
liquid medium (20 g ground maize, 40 g cotton-seed meal, 30 g white grease or
vegetable oil, 30 g deodorized kerosene, 50 g citrus molasses, and 02 mg
thiamineHCI per litre). The fermented product contains about 17 mg fJ-

Production of Carotenoids with Fungi


carotene per g solids or about 1 g/litre medium. An antioxidant, ethoxyquin, is

added at the end to stabilize the carotene. The two strains are cultivated
separately before inoculation into the production vessels.
Several industry-affiliated research groups have improved the process
(review, Ninet & Renaut, 1979, and the patents referred to therein). The
procedure presented by a Rhone-Poulenc team in France (Ninet & Renaut,
1979) has two successive mated cultures. First, 04litres of separate cultures of
each sex are inoculated into 120 litres of medium (70 g corn steep, 50 g corn
starch, 05 g KH2 P04, 01 g MnS04, 10 mg thiamin HCI per litre tap-water)
and incubated for 40 h at 26C with agitation and aeration. Then, 32 litres of
the first culture are inoculated into 320 litres of another medium (70 g distillers
solubles, 60 g cornstarch, 30 ml soya oil, 30 ml cotton-seed oil, 20 ml kerosene,
035 g ethoxyquin, 06 g isoniazid, 02 g MnS04, 05 mg thiamineHCI, per litre
at pH 63) and incubated at 26C for 185 h with agitation and aeration. After
48 h, 1 g p-ionone, 5 ml kerosene, and 42 g glucose are added per litre (the
glucose solution is continuously injected to the end of the culture). The
production should be about 30 mg p-carotene per g dry weight or about
3 g/litre.
Other sources of carbon and nitrogen have received consideration. One of
them is whey, a subproduct of cheese manufacture (Lampila et al., 1985b,c).
Blakeslea allows the production of lycopene and other intermediate products
in the presence of appropriate enzyme inhibitors (Ninet et al., 1969; Hsu et al.,
1972; Ninet & Renaut, 1979).


Production has been studied only under laboratory conditions. The genetic
work was carried out on minimal agar; the wild type contains about 0040 mg
p-carotene per g dry weight and the superproducing strains described above,
about 25 mg (Murillo et al., 1978). The mycelial dry weight is 10 g/litre in both
cases. We have observed that various simple media, some composed of cheap
subproducts of biological industries, increase the biomass to about 40g dry
mycelia per litre and maintain its p-carotene concentration. Carotenogenesis
suffers if the cultures are agitated, whether they are wild type (Lilly et al.,
1960) or superproducers, and therefore surface cultures are the most


Industrial purification of p-carotene (Ninet et al., 1979) is complicated and

expensive, and may not be necessary in many practical applications. For use in
animal feeds, the carotene-rich mycelium may be dried, blended with various
polymers, extruded into filaments, and powdered; the p-carotene in such
preparations survives storage at 41C for 1 year (Gogek, 1968).


E. Cerda-Olmedo

The p-carotene content of Blakeslea and Phycomyces mycelia has been
enormously increased through optimization of culture conditions and genetic
breeding, respectively. Both lines of research are far from exhausted and
should lead to further production gains.
The production of xanthophylls (oxidated carotenoids) would be feasible
with certain fungi, particularly after yield improvements brought about by new
research. Phaffia rhodozyma offers a good start (Johnson et al., 1978; Johnson
& Lewis, 1979) and effectively gives the desired pigmentation to pen-reared
salmonids (Johnson et al., 1977).
Genes for carotenoid biosynthesis may eventually be inserted and expressed
in colourless, but genetically malleable organisms, such as Saccharomyces
cerevisiae. Carotenoids would then become important subproducts of wellestablished industries. More likely is the application of molecular genetics to
strain improvement in organisms that already produce carotenoids.
Neurospora crassa offers, of course, magnificent prospects for all sorts of
genetic manipulations. Efficient transformation of Phycomyces blakesleeanus
with exogenous DNA (Arnau et al., 1988; Suarez & Eslava, 1988) may now
be combined with the more traditional approaches.
Alvarez, M. I. & Eslava, A. P. (1983). Isogenic strains of Phycomyces blakesleeanw
suitable for genetic analysis. Genetics, 105, 873-9.
Anderson, R. F., Arnold, M., Nelson, G. E. N. & Ciegler, A. (1958). Microbiological
production of beta-carotene in shaken flasks. 1. Agric. Food Chem., 6,543-5.
Andrewes, A. G. & Starr, M. P. (1976). (3R, 3'R)-Astaxanthin from the yeast Phaffia
rhodozyma. Phytochem., 15,1009-11.
Andrewes, A. G., Phaff, H. J. & Starr, M. P. (1976). Carotenoids of Phaffia
rhodozyma, a red-pigmented fermenting yeast. Phytochem., 15, 1003-7.
Arnau, J., Murillo, F. J. & Torres-Martinez, S. (1988). Expression of Tn5-derived
kanamycin resistance in the fungus Phycomyces blakesleeanw. Molec. Gen. Genet.,
Austin, D. J., Bu'Lock, J. D. & Drake, D. (1970). The biosynthesis of trisporic acids
from beta-carotene via retinal and trisporol. Experientia, 26,348-9.
Avalos, J. & Cerda-Olmedo, E. (1986). Chemical modification of carotenogenesis in
Gibberella fujikuroi. Phytochem., 25, 1837-41.
Avalos, J. & Cerda-Olmedo, E. (1987). Carotenoid mutants of Gibberella fujikuroi.
Curro Genet., 11, 505-11.
Barnett, H. L., Lilly, V. G. & Krause, R F. (1956). Increased production of carotene
by mixed + and - cultures of Choanephora cucurbitarum. Science, 123,141.
Bauernfeind, J. C. (Ed.) (1981). Carotenoids as Colorants and Vitamin A Precursors.
Technological and Nutritional Applications. Academic Press, New York.
Bejarano, E. R. & Cerda-Olmedo, E. (1989). Inhibition of phytoene dehydrogenation and activation of carotenogenesis in Phycomyces. Phytochem., 28, 1623-6.
Bejarano, E. R, Govind, N. S. & Cerda-Olmedo, E. (1987). Zeta-carotene and other
carotenes in a Phycomyces mutant. Phytochem., 26,2251-4.

Production of Carotenoids with Fungi


Bejarano, E. R., Parra, F., Murillo, F. J. & Cerda-Olmedo, E. (1988). End-product

regulation of carotenogenesis in Phycomyces. Archs Microbiol., 150, 209-14.
Bergman, K., Eslava, A. P. & Cerda-Olmedo, E. (1973). Mutants of Phycomyces with
abnormal phototropism. Molec. Gen. Genet., 123, 1-16.
Bindl, E., Lang, W. & Rau, W. (1970). Untersuchungen tiber die lichtabhangige
Carotinoidsynthese. VI. Zeitlicher Verlauf der Synthese der einzelnen Carotenoide
bei Fusarium aquaeductuum under verschiedenen Induktionsbedingungen. Planta,
94, 156-74.
Blakeslee, A. F. (1904). Sexual reproduction in the Mucorineae. Proc. Am. Acad. Arts
Sci., 40,205-319.
Bramley, P. M. & Mackenzie, A. (1988). The regulation of carotenoid biosynthesis.
Curro Topics Cell Regul., 29, 291-343.
Burgeff, H. (1924). Untersuchungen tiber Sexualitat und Parasitismus bei Mucorineen.
Botan. Abhandl., 4, 1-135.
Caglioti, L., Cainelli, G., Camerino, B., Mondelli, R., Prieto, A., Quilico, A.,
Salvatori, T. & Selva, A. (1966). The structure of trisporic-C acid. Tetrahedron
Suppl., 7, 175-87.
Castillo, R. (1980). On the transformation of beta-carotene 15,15'-3H2 into astaxanthin
by the hermit crab Clinabarius erythropus Latreille (1818), Crustacea, Decapoda,
Anomoura. Compo Biochem. Physiol., 66A, 695-7.
Cerda-Olmedo, E. (1985). Carotene mutants of Phycomyces. Meth. Enzymol., 110,
Cerda-Olmedo, E. (1987). Carotene. In Phycomyces, ed. E. Cerda-Olmedo & E. D.
Lipson. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 199-222.
Cerda-Olmedo, E. & Htittermann, A. (1986). Forderung und Hemmung der Carotinsynthese bei Phycomyces durch Aromaten. Angew. Botanik, 60, 59-70.
Cerda-Olmedo, E. & Lipson, E. D. (Eds) (1987). Phycomyces. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY.
Ciegler, A. (1965). Microbial carotenogenesis. Adv. Appl. Microbiol., 7, 1-34.
Ciegler, A., Arnold, M. & Anderson, R. F. (1959a). Microbiological production of
carotenoids. IV. Effect of various grains on production of beta-carotene by mated
strains of Blakeslea trispora. Appl. Microbiol., 7,94-8.
Ciegler, A., Arnold, M. & Anderson, R. F. (1959b). Microbiological production of
carotenoids. V. Effect of lipids and related substances on production of betacarotene. Appl. Microbiol., 7, 98-1Ol.
Ciegler, A., Nelson, G. E. N. & Hall, H. H. (1962). Microbiological production of
carotenoids. VIII. Influence of hydrocarbon on carotenogenesis by mated cultures
of Blakeslea trispora. Appl. Microbiol., 10, 132-6.
Ciegler, A., Lagoda, A. A., Sohns, V. E., Hall, H. H. & Jackson, R. W. (1963a).
Beta-carotene production in 20-liter fermentors. Biotechnol. Bioengin., 5, 109-2l.
Ciegler, A., Nelson, G. E. N. & Hall, H. H. (1963b). Enhancement of beta-carotene
synthesis by citrus products. Appl. Microbiol., 11, 128-3l.
Ciegler, A., Nelson, G. E. N. & Hall, H. H. (1963c). Enhancement of carotenogenesis
in Blakeslea trispora by essential citrus oils. Nature, Lond., 198, 1305-6.
Ciegler, A., Pazola, Z. & Hall, H. H. (1964). Stimulation of carotenogenesis by
microbial cells. Appl. Microbiol., 12, 150-54.
De la Concha, A. & Murillo, F. J. (1984). Accumulation of a complex form of
beta-carotene by Phycomyces blakesleeanus cytoplasmic mutants. Planta, 161,
De la Guardia, M. D., Arag6n, C. M. G., Murillo, F. J. & Cerda-Olmedo, E. (1971).
A carotenogenic enzyme aggregate in Phycomyces: Evidence from quantitative
complementation. Proc. natn. Acad. Sci. U.S.A., 68,2012-15.
El-Jack, M., Mackenzie, A. & Bramley, P. M. (1988). The photoregulation of
carotenoid biosynthesis in Aspergillus giganteus mut. alba. Planta, 174, 59-66.


E. Cerda-Olmedo

Engel, B. G., Wiirsch, J. & Zimmermann, M. (1953). fIber den Einfluss von
beta-Jonon auf die Bildung von beta-Carotin durch Phycomyces blakesleeanus.
Helv. Chim. Acta, 36, 1771-6.
Eslava, A. P. (1987). Genetics. In Phycomyces, ed. E. Cerda-Olmedo & E. D. Lipson.
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 27-48.
Eslava, A. P. & Cerda-Olmedo, E. (1974). Genetic control of phytoene dehydrogenation in Phycomyces. Plant Sci. Lett., 2,9-14.
Eslava, A. P., Alvarez, M. I. & Cerda-Olmedo, E. (1974). Regulation of carotene
biosynthesis in Phycomyces by vitamin A and beta-ionone. Eur. J. Biochem., 48,
Feofilova, E. P. (1974). Pigmenty mikroorganizmov. Nauka, Moscow.
Feofilova, E. P. & Bekhtereva, M. N. (1976). Effect of vitamin A on carotene
biosynthesis by Blakeslea trispora (in Russian). Mikrobiologia, 45, 557-8.
Feofilova, E. P., Ushanova, A. E. & Ivanova, G. B. (1982). Effect of dimilin on
carotene production by Blakeslea trispora (in Russian). Mikrobiologia, 51,
Garber, E. D., Baird, M. L. & Weiss, L. M. (1978). Genetics of Ustilago violacea. II.
Polymorphism of color and nutritional requirements of sporidia from natural
populations. Bot. Gaz., 139,261-5.
Gogek, C. J. (1968). Stabilization of crude carotene-containing mycelium. J. Agric.
Food Chem., 16, 730-34.
Gooday, G. W., Fawcett, P., Green, D. & Shaw, G. (1973). The formation of fungal
sporopollenin in the zygospore wall of Mucor mucedo: a role for the sexual
carotenogenesis in the Mucorales. J. Gen. Microbiol., 74, 233-9.
Goodwin, T. W. (Ed.) (1976). Chemistry and Biochemistry of Plant Pigments, 2nd edn,
two vols. Academic Press, London.
Goodwin, T. W. (1980). The Biochemistry of the Carotenoids. 2nd edn., Vol. 1, Plants.
Chapman and Hall, London/Academic Press, London.
Goodwin, T. W. (1986). Metabolism, nutrition, and function of carotenoids. Ann. Rev.
Nutr., 6,273-397.
Goodwin, T. W. (Ed.) (1988). Plant Pigments. Academic Press, London.
Goodwin, T. W. & Griffiths, L. A. (1952). Studies in carotenogenesis. 5. Carotene
production by various mutants of Phycomyces blakesleeanus and Phycomyces nitens.
Biochem. J., 52, 499-501.
Govind, N. S. & Cerda-Olmedo, E. (1986). Sexual activation of carotenogenesis in
Phycomyces blakesleeanus. J. Gen. Microbiol., 132,2775-80.
Hanson, A. M. (1967). Production of pigments and vitamins. In Microbial Technology,
ed. H. J. Peppler. Van Nostrand-Rheinhold, Princeton, pp. 222-50.
Harding, R. W. & Shropshire, W. (1980). Photocontrol of carotenoid biosynthesis.
Ann. Rev. Plant Physiol., 31,217-38.
Harding, R. W. & Turner, R. V. (1981). Photoregulation of the carotenoid biosynthetic pathway in albino and white collar mutants of Neurospora crassa. Plant Physiol.,
Harding, R. W., Huang, P. C. & Mitchell, H. K. (1969). Photochemical studies of the
carotenoid biosynthetic pathway in Neurospora crassa. Archs Biochem. Biophys.,
Hsu, W. J., Yokoyama, H. & Coggins, C. W. (1972). Carotenoid biosynthesis in
Blakeslea trispora. Phytochem., 11,2985-90.
Isler, O. (Ed.) (1971). Carotenoids. Birkhauser Verlag, Basel.
Johnson, E. A. & Lewis, M. J. (1979). Astaxanthin formation by the yeast Phaffia
rhodozyma. J. Gen. Microbiol., 115,173-83.
Johnson, E. A., Conklin, D. E. & Lewis, M. J. (1977). The yeast Phaffia rhodozyma as
a dietary pigment source for salmonids and crustaceans. J. Fish. Res. Board Can.,

Production of Carotenoids with Fungi


Johnson, E. A., Villa, T. G., Lewis, M. J. & Phaff, H. J. (1978). Simple method for
isolation of astaxanthin from the basidiomycetous yeast Phaffia rhodozyma. Appl.
Environ. Microbiol., 35, 1155-9.
Khabrova, A. M. & Zhdanov, V. G. (1979). Study of sex defective mutants of
Blakeslea trispora. Mikrobiologia, 48, 1055-9.
Krzeminiski, L. F. & Quackenbush, F. W. (1960). Stimulation of carotene synthesis in
submerged cultures of Neurospora crassa by surface-active agents and ammonium
nitrate. Archs Biochem. Biophys., 88,64-7.
Lampila, L. E., Wallen, S. E. & Bullerman, L. B. (1985a). A review of factors
affecting biosynthesis of carotenoids by the order Mucorales. Mycopathologia, 90,
Lampila, L. E., Wallen, S. E., Bullerman, L. B. & Lowry, S. R. (1985b). The effect of
strain and type of whey on the production of beta-carotene and other parameters.
Lebensmittel-Wissenschaft-Technologie, 18, 366-9.
Lampila, L. E., Wallen, S. E., Bullerman, L. B. & Lowry, S. R. (1985c). The effect of
illumination on growth and beta-carotene content of Blakeslea trispora grown in
whey. Lebensmittel-Wissenschaft-Technologie, 18,370-73.
Lilly, V. G., Barnett, H. L. & Krause, R. F. (1960). The production of carotene by
Phycomyces blakesleeanus. West Virginia Univ. Agr. Exper. Sta., Bulletin 441T,
Mackinney, G., Nakayama, T., Buss, C. D. & Chichester, C. O. (1952). Carotenoid
production in Phycomyces. 1. Am. Chem. Soc., 74,3456-7.
Mackinney, G., Nakayama, T., Chichester, C. O. & Buss, C. D. (1953). Biosynthesis
of carotene in Phycomyces. 1. Am. Chem. Soc., 75, 236-8.
Murillo, F. J. & Cerda-Olmedo, E. (1976). Regulation of carotene synthesis in
Phycomyces. Molec. Gen. Genet., 148, 19-24.
Murillo, F. J., Calder6n, I. L., L6pez-Diaz, I. & Cerda-Olmedo, E. (1978).
Carotene-superproducing strains of Phycomyces. Appl. Environ. M icrobiol., 36,
Murillo, F. J., Torres-Martinez, S., Arag6n, C. M. G. & Cerda-Olmedo, E. (1981).
Substrate transfer in carotene biosynthesis in Phycomyces. Eur. 1. Biochem., 119,
Ninet, L. & Renaut, J. (1979). Carotenoids. In Microbial Technology, 2nd edn,
Vol. 1, ed. H. J. Peppler & D. Perlman. Academic Press, New York, pp.
Ninet, L., Renaut, J. & Tissier, R. (1969). Activation of the biosynthesis of carotenoids
by Blakeslea trispora. Biotechnol. Bioeng., 11, 1195-210.
Rau, W. (1983). Photoregulation of carotenoid biosynthesis. In Biosynthesis of
isoprenoid compounds, ed. J. W. Porter & S. L. Spurgeon. John Wiley, New York,
pp. 123-57.
Rau, W. & Schrott, E. L. (1987). Blue light control of pigment biosynthesisCarotenoid biosynthesis. In Blue Light Responses, Vol. I, ed. H. Senger. CRC
Press, Boca Rat6n, Florida, pp. 43-63.
Reyes, P., Chichester, C. O. & Nakayama, T. O. M. (1964). The mechanism of
beta-ionone stimulation of carotenoid and ergosterol biosynthesis in Phycomyces
blakesleeanus. Biochim. Biophys. Acta, 90, 578-92.
Roncero, M. I. G. & Cerda-Olmedo, E. (1982). Genetics of carotene biosynthesis in
Phycomyces. Curro Genet., 5, 5-8.
Suarez, T. & Eslava, A. P. (1988). Transformation of Phycomyces with a bacterial gene
for kanamycin resistance. Molec. Gen. Genet., 212, 120-23.
Sutter, R. P. (1970). Effect of light on beta-carotene accumulation in Blakeslea trispora.
1. Gen. Microbiol., 64, 215-21.
Sutter, R. P. (1975). Mutations affecting sexual development in Phycomyces blakesleeanus. Proc. natn. Acad. Sci. USA, 72, 127-30.


E. Cerda-Olmedo

Tada, M. & Shiroishi, M. (1982). Mechanism of photoregulated carotenogenesis in

Rhodotorula minuta. I. Photocontrol of carotenoid production. Plant Cell Physiol.,
Valadon, L. R. G. & Mummery, R. S. (1973). Effect of certain inhibitors of
carotenogenesis in Verticillium agaricinum. Microbios, 7, 173-80.
Vaskivnyuk, V. Y. & Getman, E. I. (1984). Biosynthesis of carotenoids in light in
continuous cultures of the yeast Rhodosporidium diobovatum (in Russian). Priklad.
Biokhim. Mikrobiol., 20, 480-83.
Will, O. H., Ruddat, M., Garber, E. D. & Kezdy, F. J. (1984). Characterization of
carotene accumulation in Ustilago violacea using high-performance liquid chromatography. Curro Microbiol., 10,57-64.

Chapter 4
Laboratories of Medical Biochemistry and Clinical Analysis, Faculty of
Pharmaceutical Sciences, State University of Ghent, B-9OOO Ghent, Belgium



Discovery of Carotenoids in Nature

The early discovery of carotenoids in living organisms is obviously attributable

to the conspicuous appearance of these yellow, orange or red pigments. In
1831 Wackenroder prepared 'carotene' in crystalline form (Isler, 1971),
whereas xanthophylls were first isolated from autumn leaves by Berzelius in
1837 (Isler, 1971). Following the main-classical chromatographic experiments
of Tswett at the beginning of this century (Isler, 1971), it was soon realized
that there existed a complex group of carotenoids in nature. The next decades
indeed saw the isolation and structural elucidation of a large variety of new
derivatives. More recently, the advent of modern spectrometric techniques and
the refinement of separation methods have resulted in the ongoing discovery of
more new structures, the determination of their absolute configurations and
the establishment of routes for their partial or total synthesis. Concurrently,
the biochemistry of carotenoids in plants and animals has gradually been
unravelled (Goodwin, 1980, 1984).


Early Observations on the Occurrence of Carotenoids in


The occurrence of carotenoids in micro-organisms was first reported about 100

years ago, when a number of pigmented non-photosynthetic bacteria (quoted
in Ingraham & Baumann, 1934) and fungi (quoted in Valadon, 1976) were
isolated. A more systematic survey of carotenoid-producing nonphotosynthetic bacteria was conducted in the 1930s (Ingraham & Baumann,

H. J. Nelis & A. P. De Leenheer


1934). Pioneering work on photosynthetic bacteria was carried out by Van Niel
& Smith (1935).


(Davies, 1976; Moss & Weedon, 1976)
Chemically, carotenoids are polyenes consisting of a number of isoprenoid
units (usually 8) arranged in such a way that the two central methyl groups (20
and 20', Fig. 1) are in the 1,6 position, while all remaining substituents are
positioned 1,5 relative to each other. A series of conjugated double bonds
constitutes a chromophore of variable length, resulting in the characteristic
yellow to red colors (absorption maxima 400-500 nm). Figure 1 presents the
five basic (no substituents) structures of which all compounds pertinent to this
chapter are derived. Full structures of the latter are depicted in Fig. 2, together
with their trivial (common), semi-systematic and systematic names (Table 1,
legend to Fig. 2).
Unsubstituted carotenoid hydrocarbons, e.g. lycopene, {:I-carotene, and
torulene, are commonly designated as 'carotenes'. The collective term xanthophylls covers oxygenated derivatives, including alcohols (e.g. lutein and











Fig. 1. Basic structures of carotenoids of which the compounds of interest (Fig. 2,

Table 1) are chemically derived. Systematic IUPAC names in parentheses. 1, lycopene
('IjI,'IjI-carotene); 2,y-carotene (p,'IjI-carotene); 3,p-carotene (p,p-carotene); 4,acarotene (P, e-carotene); 5, retrodehydro-p-carotene (4'5' -didehydro-4,5'-retro-p, pcarotene).

Microbial Production of Carotenoids














Fig. 2. Structural formulae of carotenoids of interest. See Table 1 for explanation.

zeaxanthin), epoxides (e.g. violaxanthin), ethers (e.g. spheroidene), ketones

(e.g. canthaxanthin, astaxanthin, spheroidenone) and acids such as torularhodin. Rhodoxanthin is an example of a retrocarotenoid, in which all single
and double bonds of the conjugated polyene system have shifted by one
position. The modern IUPAC systematic nomenclature (IUPAC, 1971) is
based on a designation of the end groups of the molecules, using such prefixes
as 1jJ (acyclic end-group), f3 or E (cyclohexene end groups) (indicated in Fig. 1).
Most natural carotenoids possess the all-trans configuration, although the
genuine presence of cis-isomers has been documented, e.g. in photosynthetic
bacteria (Koyama et al., 1983). If a compound contains chiral centers (as in
zeaxanthin and astaxanthin) the absolute configuration of the optical isomer(s)
found in nature is also mentioned in Table 1.

Trivial name

Basic structure
(Fig. 1)

Semi-systematic name

Systematic name (IUPAC)

2. Compounds of possible future biotechnological interest (produced by the micro-organisms listed in Table 2,11)
3' ,4' -Dehydro-y-carotene
3' ,4' -Didehydro-jJ, 1/J-carotene
16' -Carboxyl-3' ,4' -dehydro3' ,4' -Didehydro-jJ, 1/J-caroten-16'y-carotene
oic acid
1-Methoxy-1 ,2,7' ,8' -tetrahy1-Methoxy-3,4-didehydro-1 ,2,7' ,
8' -tetrahydro-1/J, 1/J -carotene
1-Methoxy-2-keto-1,2,7'8'-te1-Methoxy-3,4-didehydro-1,2,7' ,8'trahydro-3 ,4-dehydrolycopene
tetrahydro-1/J, 1/J -caroten-2-one

1. Compounds of immediate biotechnological interest (produced by the micro-organisms listed in Table 2, I)

A. Carotenes
1/J, 1/J -Carotene
B. Hydroxy-xanthophyUs
(3R, 3'R, 6'R)-jJ,E-Carotene4
3,3' -DihydroxY-ll'-carotene
3,3' -diol.
3,3' -Dihydroxy-{:j -carotene
(3R, 3'R)-jJ,jJ-Carotene-3,3' -diol
C. Monocyclic ketocarotenoids
D. Bicyclic ketocarotenoids
4,4' -Diketo-jJ-carotene
jJ,jJ-Carotene-4,4' -dione
3,3' -Dihydroxy-4 ,4' -diketo6
3,3' -Dihydroxy-jJ,jJ-carotene-4,4'Astaxanthin
dione (3S, 3'S or 3R, 3'R)
E. Retrocarotenoids
3,3' -Diketoretrodehydro4' ,5' -Didehydro-4,5' -retro-jJ,jJ-ca7
rotene-3,3' -dione


Table 1
Structures and Nomenclature of Carotenoids of Interest.




Microbial Production of Carotenoids



The extended chromophore in the carotenoid structure points to a biological

significance for these pigments in micro-organisms, by suggesting photosensitive and antioxidant properties. In algae and photosynthetic bacteria, carotenoids participate as secondary light-harvesting compounds in the photosynthetic
process (Mathis & Schenck, 1982) and exert photoprotective effects as well
(Burnett, 1976). As antioxidants they protect vulnerable tissues against the
deleterious effects of singlet oxygen and free radicals (Krinsky, 1979). Similar
(photo )-protective effects have been invoked in non-photosynthetic bacteria
(Mathews & Krinsky, 1965; Mathews-Roth & Krinsky, 1970; Mathews-Roth et
al., 1974), although conflicting evidence exists (Schwartzel & Cooney, 19740;
Dieringer et al., 1977).


The economic significance of carotenoids has been extensively reviewed (KHiui

& Bauernfeind, 1981; Marusich & Bauernfeind, 1981; Simpson et al., 1981;
KHiui, 1982). There is a continuously growing market for carotenoids as food
colorants (KHiui & Bauernfeind, 1981) and pigmenters in animal feeds
(Marusich & Bauernfeind, 1981; Simpson et al., 1981). Poultry feeds based on
agricultural waste products have to be supplemented with pigments in order to
impart a desirable yellow-orange color to egg yolks and/or the skin of broiler
chickens (Marusich & Bauernfeind, 1981). Optimal yolk color ensues from the
combination of a yellow base pigment (e.g. lutein, p-apo-8'-carotenoic acid
ethyl ester) with a small quantity of an orange-red one, preferably a
ketocarotenoid (canthaxanthin) or zeaxanthin.
The addition of ketocarotenoids (astaxanthin or canthaxanthin) to fish feeds
also intensifies the reddish colour of the flesh of the salmon and the trout
(Simpson et al., 1981). So far, the application of natural sources of carotenoids
remains confined to animal feeds (Marusich & Bauernfeind, 1981; Simpson et
al., 1981). Industrially produced natural carotenoids are mainly derived from
vegetable sources, including marigold flowers, alfalfa, red peppers and annatto
(Marusich & Bauernfeind, 1981).


There is a wealth of literature on the occurrence of carotenoids in algae, fungi

and bacteria, the emphasis being on structural studies and the elucidation of
the pertinent biosynthetic pathways. This information has been exhaustively
reviewed in an authoritative book by Goodwin (1980).
Our survey of carotenoid-producing organisms is limited to those species
which show biotechnological potential, i.e. which could be considered for the

H. J. Nelis & A. P. De Leenheer


development of a commercially exploitable, natural source of pigments. The

selection of useful organisms (Table 2) was made on the basis of three criteria:
(a) The presence in the organism of carotenoids of interest (cf. the
limitative list in Table 1, p-carotene excluded).
(b) The already established industrial use of a given strain. So far this has
only been the case for green algae. However, most organisms included
in Table 2 have not (yet) found such application, but do show promise to
be developed for this purpose. Specifically, this could mean that
attempts have already been made, are currently being made or are likely
to be made in the future to optimize the pigment levels in these
organisms. In order to justify the efforts, the carotenoids involved
should be particularly valuable, difficult to synthesize and be absent or
rare in cheaper and more accessible natural sources. The recent
commercial exploitation of micro-algae as a source of p-carotene proves
that there is a potential market for natural carotenoids, even if the
(cheaper) synthetic equivalents exist (cf. Chapter 2).
(c) The current interest in a number of carotenoid containing organisms for
reasons other than their pigment composition. Alternative applications
include the production of biomass, single cell oil or lipid, (other) fine
chemicals, or the purification of waste waters. However, the presence of
often high quantities of carotenoids in those particular organisms could
make them attractive for commercial production of these useful 'byproducts' as well.

BIOSYNTHESIS (for a review, see Spurgeon & Porter, 1983)

A first series of steps in the formation of carotenoids belongs to the

well-known mevalonate pathway, the general biosynthetic scheme of all
isoprenoid compounds. Starting from acetate (acetyl CoA), activated isoprene
units undergo a sequence of head-to-tail condensations, leading to geranylgeranyl pyrophosphate (Cw). From this point on, all subsequent steps are unique
to the biosynthesis of carotenoids. Head-to-head condensation of 2 C20 units
yields the key intermediate phytoene, the first colorless carotenoid. Through a
series of stepwise dehydrogenations, phytoene is converted to lycopene, from
which all other carotenoids can eventually be derived. The biosynthetic
interrelationship of the derivatives pertinent to this chapter is illustrated in
Fig. 3.


The biosynthesis of carotenoids by micro-organisms is under genetic control,

but is also influenced by various environmental factors (Spurgeon & Porter,

Microbial Production of Carotenoids

Table 2
Micro-organisms of Potential Practical Interest for the Production of

Carotenoid(s) produced

1. Micro-organisms specifically designed as pigment sources

1. Sources of carotenes
A. Fungi
Blakeslea trispora
B. Non-photosynthetic bacteria
Streptomyces chrestomyceticus,
subsp. rubescens
2. Sources of xanthophyllsa
A. Green algae
Spongiococcum excentricum
Chlorella pyrenoidosa
B. Fungi
Dacrymyces deliquescens
C. Non-photosynthetic bacteria
Flavobacterium sp.
Streptomyces chrestomyceticus,
Unidentified xanthophylls
var. aurantioideus
Mycobacterium phlei
Unidentified xanthophylls
3. Sources of monocyclic ketocarotenoids
A. Non-photosynthetic bacteria
Deinococcus radiophilus-radiodurans- Derivatives of 4-ketoradiopugnans
Mycobacterium smegmatis
Derivatives of 4-ketoy-carotene
4. Sources of bicyclic ketocarotenoids
A. Cyanobacteria
Anabaena variabilis
Aphanizomenon flos-aquae
Nostoc commune
B. Green algae (N-deficiency)
Dictyococcus cinnabarinus
Haematococcus pluvialis
C. Fungi/yeast
PhajJia rhodozyma
D. Non-photosynthetic bacteria
Brevibacterium KY -4313
Rhodococcus maris / Mycobacterium Canthaxanthin
brevicale 32-MCT
Mycobacterium lacticola
Brevibacterium 103
5. Sources of retrocarotenoids
Pseudomonas extorquens
II. Micro-organisms designed for applications other than
pigment production
A. Fungi/yeast
Rhodotorula, Rhodosporidium sp.
(red yeasts)
B. Non-photosynthetic bacteria
C. Photosynthetic bacteria
Rhodobacter capsulatus

Ketocarotenoids excluded.



H. J. Nelis & A. P. De Leenheer







Fig. 3. Biosynthetic relationship between the carotenoids of interest. 1, All-transphytoene (actual precursor of lycopene may be 15,15'-cis-phytoene); 2, lycopene; 3,
spheroidene; 4, spheroidenone; 5, y-carotene; 6, 4-keto-y-carotene; 7, torulene; 8,
torularhodin; 9, /J-carotene; 10, zeaxanthin (3R,3'R); 11, rhodoxanthin; 12, canthaxanthin; 13, astaxanthin (3S, 3'S); 14, a-carotene; 15, lutein (3R, 3'R, 6'R). Reactions
involved: A, stepwise desaturation (dehydrogenations). B, water addition at 1,2 double
bond + dehydrogenation (C-3 and C-4) + methylation of hydroxyl group at C-l. C, keto
group addition at C-2. Mechanism not established. 0, cyclization of 1 or 2 terminal
end(s) of lycopene, yielding 1/J-end group (/J, tp: y-carotene, 5), 2/J-end groups (/J,/J:
/J-carotene, 9) or 1/J- and Ie-end group (/J,e: a-carotene, 14), respectively. E,
dehydrogenation at C-3' and C-4'. F, terminal oxidation (-CH3--CH20H-CHO--COOH). G, keto group addition at C-4. Probably hydroxylation followed by
dehydrogenation. H, hydroxylation at C-3 and C-3'. Requires molecular oxygen.
Mechanisms unknown. I, Keto group addition at C-4 and C-4'. Probably hydroxylation
followed by dehydrogenation. J, Mechanism uncertain. Probably attack of OH+ on
zeaxanthin at C-5, loss of H+ at C-4', loss of water and dehydrogenation of hydroxyl
groups at C-3 and C-3'.


Environmental Factors

Both the pigment patterns and the amount of carotenoids in the cells may be
altered by changes in the composition of the growth medium and by the
influence of physical factors such as light, temperature and oxygen. Variation
of the growth conditions of an organism with the aim to optimize its carotenoid content sometimes has a rational basis (e.g. generation of biosynthetic
intermediates), but very often is rather empirical. From the numerous effects
reported in the literature (for a compilation, see Goodwin, 1980) only a few

Microbial Production of Carotenoids


examples have been selected to illustrate how the carotenoid biosynthesis in

certain algae, fungi and bacteria responds to changes in the environment.




Under optimal conditions for proliferation, green algae produce the typical
chloroplastidic carotenoids ~-carotene, lutein, zeaxanthin and various epoxides. However, nutritional imbalance, particularly the lack of nitrogen, often
induces the synthesis of secondary extraplastidic (keto )carotenoids, so that the
cultures turn orange or red.

Under autotrophic conditions, light has in general little or no effect on the

carotenoid production by algae. An interesting effect of oxygen is noted in
some organisms. Anaerobic conditions exclusively lead to carotene formation,
whereas in the presence of oxygen synthesis is primarily directed towards
7.1. 2

Fungi and yeast


The nature of the carbon and nitrogen source affects carotenoid formation in
fungi in an unpredictable manner. In general, a high carbon to nitrogen ratio
promotes carotenogenesis. The stimulatory effect of leucine and valine can be
explained in that these amino acids are precursors of ~-hydroxy-~-methyl
glutaryl-CoA, a key intermediate in isoprenoid synthesis. Fluoride is thought
to exert its positive influence at the enzymatic level: this toxic chemical
preferentially inhibits certain phosphatases, which catalyze the hydrolysis and,
hence, cause the depletion of essential phosphorylated intermediates from the
isoprenoid pathway.

A strange effect of temperature on the qualitative carotenoid composition has

been noted in certain red yeasts. At low temperatures, ~- and y-carotene
predominate, while a temperature increase reportedly favors the formation of
torulene and torularhodin. Strict photoregulation of carotenogenesis occurs in
a few, thoroughly investigated species (Rau, 1983). The latter only contain
negligible amounts of carotenoids when grown in total darkness. A brief
exposure to light, in the presence of oxygen, induces de novo synthesis of
carotenogenic enzymes and, accordingly, carotenoid production. The exact
nature of the photoreceptor is still uncertain and the whole process is under
genetic control (see below).


7.1. 3

H. J. Nelis & A. P. De Leenheer

Non-photosynthetic bacteria


Carotenoid formation is common among hydrocarbon assimilating bacteria.

However, the carotenoid pattern may be different according to whether they
are grown on hydrocarbon or non-hydrocarbon media. Like in fungi, high
carbon to nitrogen ratios commonly increase the pigment yield. Ammonium
salts are preferred nitrogen sources. The concentration of thiamine in the
medium may both affect the nature of the carotenoids formed (cyclic vs
acyclic) and their cellular concentration.

Photoinduction, as outlined for fungi (see Section 7.1.2(B has been reported
in a number of strains, particularly belonging to the genus Mycobacterium
(Rau, 1983). Illumination of cultures may also alter the qualitative carotenoid
composition, by converting carotenes, normally produced in the dark, into
Z 1. 4 Photosynthetic bacteria

Few data are available regarding the influence of nutrients in the medium on
the carotenoid formation by photosynthetic bacteria. A specific example of the
effect of iron will be discussed below (see Section 111,4).

Oxygen and light are key factors controlling carotenogenesis in photosynthetic

bacteria. In general, anaerobic conditions favor the formation of less-oxidized
derivatives (e.g. spheroidene), as opposed to ketocarotenoids (spheroidenone), which are produced in the presence of oxygen.
7.2 Inhibitors (Ninet et aI., 1969; Spurgeon & Porter, 1983)
Compounds inhibiting the normal biosynthesis of carotenoids can be classified
in two groups. The first group consists of inhibitors that cause the accumulation of early intermediates in the pathway, particularly the colorless phytoene.
These compounds, e.g. diphenylamine, are very useful academically to
elucidate the biosynthetic schemes, but obviously of no interest for practical
application. Other inhibitors preferentially block later reactions in the pathway, notably the cyclization of lycopene. Examples of these selective inhibitors
include nicotine and CPTA (2-(4-chlorophenylthio )-triethylamine) and other
substituted amines, as well as nitrogenous heterocyclic bases (e.g. imidazole).

Microbial Production of Carotenoids



Genetic Control

The biosynthesis of carotenoids is under the control of different genes that

code for the individual enzymes in the pathway. The basis of the photoregulation in fungi and non-photosynthetic bacteria is also genetic; the actual control
probably takes place at the transcription level in the protein synthesis (Rau,
1983). Illumination may either cause the inactivation of a repressor (derepression) or, alternatively the production of an inducer. Similarly, repression and
derepression may also be involved in the environmental control of


Various bacterial mutants have been prepared that have a defect in the
structural gene for one of the carotenogenic enzymes (Spurgeon & Porter,
1983). Three types of mutants with altered carotenoid biosynthesis are of
potential practical interest.

Blocked Mutants-Mutation to Non-production of Undesirable


Mutants blocked at the level of cyclization of lycopene to fJ-carotene could be

useful when overproduction of the former is aimed at. The undesired cyclic
compounds will not be formed and consequently lycopene will accumulate.

Mutational Biosynthesis-New Carotenoid Formation

Mutation may lead to an altered and, from a practical standpoint more

interesting pigment pattern (qualitative strain improvement). Saperstein et al.
isolated various colored mutants following treatment of a Corynebacterium
michiganense strain with uranium salts (Saperstein et al., 1954). While the
predominating pigments in the yellow parent strain were reportedly fJcryptoxanthin and lycopene, an orange mutant and a red back-mutant were
found to contain mostly canthaxanthin and lycopene, respectively. In both
cases the 'uninteresting' fJ-cryptoxanthin was replaced by more valuable

Mutants with Enhanced Pigment Levels

Few examples are known of systematic mutation programs directed towards

increasing the carotenoid content of a given species. This may be in part
attributed to the lack of objective criteria for the selection of mutants with
enhanced pigmentation. It may seem obvious to carry out such selection on the
basis of visual inspection. However, quantitative differences in carotenoid

H. 1. Nelis & A. P. De Leenheer


content may not always be so easy to recognize. The subjectivity of the

interpretation of color differences is exemplified by the results of Russian
workers who prepared carotenoid overproducing mutants of Mycobacterium
phlei (Voznyakovskaya & Daraseliya, 1972). The parent strain was described
as 'yellow-orange' (carotenoid content of 17 mg/ g), whereas two more
pigmented mutants were judged as 'orange' (44 mg/ g) and 'clear orange'
(79 mg/g), respectively.
Alternatively, it could be assumed that mutants with enhanced carotenoid
content are more resistant to photodynamic killing. If a correlation indeed
exists between pigmentation and the extent of photoprotection, these mutants
would supposedly show a higher survival rate upon illumination, e.g. in the
presence of a photosensitizer. However, as mentioned before (see Section 3), a
correlation does not necessarily exist between pigment content and the degree
of photoprotection supposedly conferred upon the organism by the carotenoid.
The production of 'interesting' carotenoids (cf. Fig. 2 and Table 1) through
fermentation of selected organisms (Table 2) will be surveyed in the light of
the criteria listed in Section 5: a basic distinction will be made between
organisms which have been or could be specifically designed for pigment
production, and those which are applied otherwise, but may become attractive
as pigment sources as well. How important fermentative production of a given
carotenoid really is may be judged from the availability of alternative natural
sources (briefly discussed in General). For each organism details will be
provided on fermentation conditions, optimization strategies to increase the
pigment content and the overall yields (cell mass, pigment level, fermentation
time). Most of the data are taken from the scientific or the patent literature,
except for part of our own unpublished work with hydrocarbon-utilizing
bacteria. The study of the carotenoids of Rhodobacter capsulatus is part of a
collaborative project coordinated at the State University of Ghent (Professor
Dr W. Verstraete, Faculty of Agricultural Sciences), in which the authors of
this chapter were responsible for the analytical monitoring of the carotenoid
levels in the course of the optimization work.




Lycopene, the major carotenoid in tomatoes is mainly used as a food colorant

(KHiui & Bauernfeind, 1981). However, there is evidence indicating that its

Microbial Production of Carotenoids


presence in poultry feeds enhances the color of the egg yolk (Marusich &
Bauernfeind, 1981).

1.2 Fungi
As discussed in this book (Chapter 3), the fungus Blakeslea trispora (Mucorales) synthesizes massive amounts of p-carotene when the two sexual forms of
this species are mixed in the same culture. However, there are several
approaches to block the p-carotene biosynthesis at the lycopene level.

1. 2.1 Composition of the Growth Medium

A medium with a neutral to alkaline pH favors lycopene production by
Blakeslea, whereas at more acidic pH's p-carotene is formed as the predominating pigment (Ciegler, 1965; Ninet & Renaut, 1979). In a 1963 patent, a
yield of 015 g lycopene per liter was claimed, using a broth containing glucose,
oleic acid, fish stick liquor and sodium carbonate (Ciegler, 1965; Ninet &
Renaut, 1979). The latter was added to maintain the pH above 66.
1. 2. 2 Use of Inhibitors
As mentioned above, substituted amines and various nitrogenous heterocyclic
compounds block the cyclization of lycopene to p-carotene (Ninet et al., 1969;
Spurgeon & Porter, 1983). The addition of CPTA to the growth medium
indeed results in the accumulation of lycopene in Blakeslea trispora (Hsu et al.,
1972). The maximum yield achieved so far, in a patented 5-day fermentation
process, was 07 g lycopene/liter (Ninet & Renaut, 1979). In a less efficient
process using the more readily available triethylamine as an inhibitor, 04 g
lycopene/liter was produced in a medium based on cotton-seed, cornflour and
soybean oil (Ninet & Renaut, 1979). Ninet et al. obtained about 1 g lycopene
per liter by adding 01 % piperidine to a medium containing, amongst others,
soybean oil, cotton-seed oil, kerosene and p-ionone (Ninet et aI., 1969).
1. 2. 3 Mutants
Mutants could be prepared which lack the ability to convert lycopene to
p-carotene. However, Blakeslea trispora lends itself reluctantly to mutation
(Ninet & Renaut, 1979).


Non.Photosynthetic Bacteria

Lycopene production (05 g/liter in 6 days) by a mutant of Streptomyces

chrestomyceticus, subsp. rubescens has been reported in the patent literature
(Ninet & Renaut, 1979). The medium used contained starch, soybean flour
and ammonium sulfate.




Among the xanthophylls, the hydroxylated derivatives lutein and zeaxanthin in

particular are of high interest as key components in natural pigmenters for


H. J. Nelis & A. P. De Leenheer

poultry (Marusich & Bauernfeind, 1981). Important vegetable sources of this

pigment include alfalfa meal, corn gluten meal and marigold flowers (Marusich & Bauernfeind, 1981). Zeaxanthin acts synergistically with lutein in that it
intensifies the egg yolk color obtained with the yellow base pigment. Some
micro-organisms containing xanthophylls of unspecified nature have also been
tested as pigmenters in poultry feed. However, legal considerations may
restrict their practical application.

2.2 Green Algae

In the early 1960s, alga meals were extensively studied for broiler and yolk
pigmentation, as alternatives to other xanthophyll sources like com and alfalfa
meal. So far the green alga Spongiococcum excentricum represents the only
example of a micro-organism industrially exploited for the production of
natural xanthophyll (lutein). For some time, the dried fermentation meal of
this species, containing 16-24 g of xanthophylls per kg dry mass was
marketed in the US under the name of A-Zanth (Grain Processing Co.,
Muscatine, Iowa) (Anonymous, 1962). The alga was grown under heterotrophic conditions, at 28C, in a medium containing glucose, com steep liquor,
urea and mineral salts (Ciegler, 1965; Hanson, 1967). Incremental feeding of
nutrients was required because Spongiococcum does not tolerate high sugar
concentrations. Another relative disadvantage is that this alga grows slowly in
liquid medium so that ample time must be allowed for inoculum development
(Hanson, 1967). Lutein was produced at a rate of 294 mg/liter in 9 days.
Of all other green algae tested Chlorella sorokiniana (pyrenoidosa, 7-11-05)
in particular showed great promise for further development. Cell yield and
xanthophyll content of this species, grown under heterotrophic conditions in
30-liter fermenters, were optimized by Theriault (Ciegler, 1965; Theriault,
1965; Hanson, 1967). Dry cell weights in excess of 100 g/liter and a total
xanthophyll content of 467-512 mg/liter were obtained from 230 to 260 g
glucose/liter, in 168 h illuminated batch-type fermentations (25C). Apart from
glucose, the medium constituents were urea, phosphate, MgS0 4 7H2 0 and
trace elements. The yield was further increased in continuous feed runs to
302 g cells/liter and 650 mg xanthophyll/liter, obtained from 520 g/liter of
glucose. Again, maximum concentrations of xanthophylls resulted from using
low initial levels of nutrients, followed by incremental feeding of additional
medium ingredients.
In a patent assigned to F. Hoffmann-La Roche, another, thermophilic strain
of Chlorella pyrenoidosa (A 119 Lagos, ATCC 14860) grown at 35C was used.
The culture medium consisted of cerelose, crude invert molasses, ethanol,
potassium nitrate, amino acids, guanidine nitrate, MgS04 7H2 0, phosphates,
acetate, EDTA salts and thiamine (Wendall et ai., 1964). Although the total
carotenoid content amounted to 74 mgt g dry weight, the cell yield (28 g/liter
in 4 days) was lower than in the process of Theriault.

Microbial Production of Carotenoids




Compared to algae, certain fungi belonging to the Dacrymycetaceae family are

inferior sources of xanthophylls (Ciegler, 1965; Ninet & Renaut, 1979). In a
patented process, the yield did not exceed 40 mg/liter broth (4 mg/g dry
biomass, fermentation time 5 days). A typical medium contained glucose,
glycerol and corn steep liquor as main ingredients. External illumination was
found obligatory for optimal pigment formation, the wavelength determining
the ratio of xanthophylls to total carotenoids and the yield of xanthophylls.


Non-photosynthetic Bacteria

In view of the relative shortage of zeaxanthin in higher plants, microbial

production of this valuable carotenoid is of high interest. Only few bacterial
species are known to produce free (non-glycosidic) zeaxanthin, among them
certain methylotrophs (Urakami & Komagata, 1986) and marine Flavobacteria
(McDermott et al., 1973). The latter in particular could be candidates for
industrial application. When no special measures are taken, fermentation of a
Flavobacterium sp. in a medium containing glucose and corn steep liquor
generates approximately 10-40 mg zeaxanthin per liter (Ninet & Renaut,
1979). The yield was increased to 335 mg/liter by supplementation with
palmitic esters, methionine, pyridoxine, ferrous salts, by continuous addition
of the F nutrients and reduction of the temperature (Ninet & Renaut, 1979).
Some actinomycetes produce high amounts of mostly unidentified xanthophylls. Streptomyces chrestomyceticus var. aurantioideus reportedly yields
05 g pigment/liter (Ninet & Renaut, 1979). Addition of the dry mycelium to
poultry feed was found to 'improve pigmentation' (Anonymous, 1969-1970a).
Even higher yields (up to 09 g/liter) were obtained with Streptomyces
mediolani, grown in a medium based on dextrin, casein and corn steep liquor
(fermentation time 5 days) (Ninet & Renaut, 1979). The predominating
carotenoids produced were allegedly isorenieratene (an aromatic carotene) and
its oxygenated derivatives. However, the efficiency of these compounds as
pigmenters in animal feeds has not been established.
Mycobacterium phlei deserves a special mention in that this species produces
carotenoids in quantities far exceeding those normally found in nonphotosynthetic bacteria. Unfortunately, there is some confusion about its exact
pigment composition, possibly because different strains have been investigated.
Schlegel reported the presence of mainly glycosides of oxycarotenoids in
amounts exceeding 10 mg/g dry cell weight (Schlegel, 1959). The media used
consisted of peptone, glucose, glycerol, sodium citrate, magnesium sulfate and
phosphates. Ingraham, in a 1935 study (Ingraham & Steenbock, 1935), found
that the addition of glycerol and various alcohols and glycols to a similar basic
medium markedly stimulated carotenogenesis, although the pigments isolated
were different. Other studies reported isorenieratene (leprotene), various
carotenes and some xanthophylls as the major pigments (Goodwin &

H. J. Nelis & A. P. De Leenheer


Jamikorn, 1956; Hochmannova et al., 1968). More recent patents describe the
production of 03 g 'oxycarotenoids' /liter in a medium containing beet
molasses and urea (Anonymous, 1966-1967, 1969-1970b; Ninet & Renaut,

3.1 General
Monocyclic, mono-ketocarotenoids have not been systematically studied so far
for application as pigmenters in animal feeds. One of the reasons may be that
they are not widespread in nature. However, the color (Amax 470-480 nm) and
chemical properties (keto group in conjugation with the polyene chain) of
these compounds resemble those of their bicyclic counterparts, e.g. canthaxanthin. Hence they could theoretically become useful if a sufficient supply could
be secured.

3.2 Non-photosynthetic Bacteria

Keto-derivatives of y-carotene and dehydro-y-carotene have been demonstrated in radio-resistant pink tetracocci, e.g. Deinococcus radiophilus and
radiodurans (formerly designated as Micrococcus) (Bamji & Krinsky, 1966;
Lewis & Kumta, 1973). Our own (unpublished) investigation of Deinococcus
radiopugnans ATCC 19172 (formerly Micrococcus roseus) also suggested the
presence of monocyclic ketocarotenoids in this related species. A systematic
study has been devoted to optimizing the growth during fermentation of
tetracocci (Shapiro et al., 1977). The iron content of the medium proved to be
a crucial factor in this regard. Various compounds increasing the availability of
this element, e.g. hydroxamic acids and hemin, acted as growth promoters.
Although no quantitative data on the pigments are available, the mere
presence of these unique ketocarotenoids and the exotic nature of the bacteria
appear to warrant further study.
Similar monocyclic ketocarotenoids have been reported in a Mycobacterium
smegmatis strain, grown on hydrocarbons (Tanaka et al., 1968a,b. The effect
of various nutrients and added compounds (amino acids, detergents) on
growth and pigmentation was examined. The optimal medium contained 2% of
an alkane (preferably n-hexadecane), a mixture of ammonium sulfate and
urea, or, alternatively, ammonium phosphate, as a nitrogen source, a number
of mineral salts and trace elements as well as histidine. However, the
maximum carotenoid yield was disappointingly low (08 mg/liter). A nonspecific part of this consisted of 4-keto-y-carotene and its mono-hydroxy- and
mono-methoxy derivatives.



Canthaxanthin and astaxanthin are widely used as pigmenters in poultry and

fish feeds (see above). Although both compounds are manufactured syntheti-

Microbial Production of Carotenoids


cally, their high market value makes any suitable natural source potentially
attractive. Canthaxanthin is found in substantial quantities in Crustacea
(Castillo et al., 1982) and the feathers of birds (Brush, 1981), but not in higher
plants. The fungus Cantharellus cinnabarinus, the first discovered natural
source of canthaxanthin (Haxo, 1950) can be dismissed from a practical point
of view. Astaxanthin occurs in the feathers of birds (Brush, 1981) and flower
petals of Adonis annua (Seybold & Goodwin, 1959) and is abundant in
Crustacea (Castillo et aI., 1982) and fish (Simpson et al., 1981). Apart from
crustacean products (shrimp meal, crayfish waste), no producing organism has
been used commercially (Marusich & Bauernfeind, 1981; Simpson et al.,
Although strictly speaking the retrocarotenoid rho do xanthin also belongs to
the ketocarotenoids, this compound is discussed in a separate paragraph (see
Section 5).



In view of the revival of interest in the biotechnological potential of

Cyanobacteria (formally named blue-green algae) (Lem & Glick, 1985) the
large-scale ketocarotenoid production from certain species belonging to the
Nostocaceae (Table 1) could theoretically be envisaged. No systematic study
of the usefulness of Cyanobacteria as industrial sources of canthaxanthin has
been conducted so far. However, the secretion of deadly toxins by certain
Nostocaceae (Lem & Glick, 1985) could obviously be prohibitive for any
practical application.


Green Algae

As mentioned in Section I, 7.1.1(A), many green algae, which normally

contain the xanthophylls typical of photosynthetic tissues, synthesize secondary
(keto)carotenoids at the end of their growth phase, when certain essential
nutrients, particularly nitrogen, become depleted (Czygan, 1968). The nature
of the ketocarotenoid( s) varies with the type of algae and not all species
synthesize them (Czygan, 1968). Astaxanthin is the most common secondary
pigment, mostly accompanied by canthaxanthin and echinenone. The obvious
practical potential of secondary carotenoid formation is probably outweighed
by the slow nature of the process and the low levels obtained. The two-phase
culturing on laboratory scale first involves proliferation of the algae in a
mineral medium containing an ample amount of a nitrogen source, e.g.
potassium nitrate. Production of ketocarotenoids starts after replacement of
the growth medium by a 'production medium' low in nitrate. Heterotrophic
conditions (addition of carbohydrates) accelerate the process indirectly by
stimulating growth (and primary carotenoid formation), thus causing the
nitrogen supply to be exhausted sooner.


H. 1. Nelis & A. P. De Leenheer

4.4 Fungi and Yeast

4.4.1 Importance of Phaffia rhodozyma
The red yeast Phaffia rhodozyma is unusual among the Deuteromycetes in
many respects. Unlike other genera like Rhodotorula, Rhodosporidium and
Sporobolomyces, it has the property of fermenting various carbohydrates
(Andrewes et al., 1976). In addition, it contains none of the pigments typical of
red yeasts (p-carotene, y-carotene, torulene, torularhodin), but rather astaxanthin (Andrewes et al., 1976). Apart from the genus Peniophora
(Hymenomycetes) (Arpin et al., 1966), this carotenoid has never been
demonstrated in fungi. The occurrence of astaxanthin is even more unique in
that its absolute configuration (3R, 3' R) is opposite to that of astaxanthin from
other natural sources (3S,3'S) (Andrewes & Starr, 1976). In general, the
chirality of a carotenoid is indeed considered not to be source dependent
(Andrewes & Starr, 1976).
Several observations indicating the efficiency of Phaffia rhodozyma as a
pigmenter in salmonid and crustacean diets (Johnson et al., 1977, 1980a) and
poultry feeds (Johnson et al., 1980b) created much excitement about this
organism. It proved superior to crustacean waste products as a feed additive
because of its higher nutritional value and pigment content. These promising
results appeared to warrant a thorough feasibility study of Phaffia rhodozyma
as a potential industrial source of natural astaxanthin.
4.4.2 Fermentation of Phaffia rhodozyma
The goal of such a feasibility study was 3-fold: optimization of pigment
production, a search for cheaper media and the improvement of the extractability of astaxanthin from the yeast cells.
(Johnson & Lewis, 1979)
A 'standard medium' for the propagation of Phaffia rhodozyma contained
cerelose, ammonium sulfate, KH2 P04 , MgS0 4 7H2 0, CaClz2H2 0 and yeast
extract. Essentially, astaxanthin formation is growth-associated, but its production still continues when growth has stopped upon exhaustion of glucose.
An optimum pH of 45 yielded maximum cell weight, growth rate and
astaxanthin levels. Of all carbon sources tested cellobiose afforded maximum
astaxanthin yield but less biomass and a lower growth rate than glucose.
However, increasing concentrations of the latter as well as reduced aeration,
both suggestive of fermentative conditions, were inhibitory to astaxanthin
production but rather caused less oxygenated precursors like echinenone and
p-carotene to accumulate. This is consistent with the fact that cellobiose,
which can only be utilized aerobically, is stimulatory. Unlike the carbon
source, the nature of the nitrogen source affected growth and carotenoid
formation only to a very limited extent. Light was totally without effect. The
addition of yeast extract to a vitamin-free medium increased pigmentation by a
factor of about 4. Carotenogenesis was strongly promoted (astaxanthin level of

Microbial Production of Carotenoids


8141lg/g dry weight) in the presence of tomato waste. This can be rationalized
in terms of the uptake by the yeast cells of carotenoid precursors. Typical
astaxanthin yields are of the order of 2 mg/liter, obtained in a 6O-h

Attempts to grow Phaffia rhodozyma on cheap food-processing waste products

have been restricted so far to the use of alfalfa residual juice as a substrate for
yeast propagation (Okagbue & Lewis, 1984a). This material supports appreciable growth but totally inhibits astaxanthin formation, which has been
attributed to the presence of saponins (Okagbue & Lewis, 1984b). Saponin
presumably disturbs the integrity of the plasma membrane of the cells. The
effect is indeed partially reversed by unsaturated acids and sterols.

A factor seriously limiting the efficient utilization of the astaxanthin from

Phaffia rhodozyma by fish and poultry is the high resistance of its cell wall. No
pigment is absorbed from intact cells (Johnson et al., 1978). Mechanical
rupture is not feasible for large-scale extraction and chemical hydrolysis of the
cell wall destroys the astaxanthin. Autolysis of the yeast cells by incubation in
distilled water or a citrate buffer has been proposed as a method to facilitate
extraction (Okagbue & Lewis, 1984c). Enzymatic digestion using the extracellular enzymes produced by Bacillus circulans is another useful alternative
(Johnson et al., 1978; Okagbue & Lewis, 1981, 1985). This can be accomplished in a 2-step method, involving the cultivation of Phaffia rhodozyma,
followed by heat-inactivation of the yeast cells, pH-adjustment and inoculation
of Bacillus circulans (Johnson et al., 1978). However, during heat treatment
part of the astaxanthin is lost and readjustment of the medium pH is laborious
and impractical. These drawbacks are overcome by growing the two organisms
in mixed culture (Okagbue & Lewis, 1981, 1985). Carbohydrates cheaper than
glucose, e.g. molasses can be used as carbon sources for this purpose
(Okagbue & Lewis, 1981). Strict pH control remains however critical to
survival and lytic activity of the bacillus (Okagbue & Lewis, 1985). After
growth, the biomass is directly amenable to extraction with acetone, resulting
in astaxanthin yields of over 90%. Unfortunately, the overall production of
astaxanthin is partly suppressed in mixed culture (maximum yield 1215 mg/liter) (Okagbue & Lewis, 1981, 1985). A practical advantage of the
mixed culture approach is the possibility for re-use of the cell-free culture
liquid (Okagbue & Lewis, 1985). After fortification with nutrients, this
medium supports growth of Phaffia rhodozyma and at the same time retains
the lytic activity required to modify its cell walls. Alternatively, this culture
fluid can also be used to digest the walls of yeast cells previously grown in pure
culture. A scheme both including mixed culture and recycling of culture filtrate
was developed for the enzymatic large-scale processing of Phaffia rhodozyma
for inclusion in animal diets (Okagbue & Lewis, 1981, 1985).

H. J. Nelis & A. P. De Leenheer



Non-photosynthetic Bacteria

The phenomenon of carotenoid production by members of the CMN complex

(Corynebacterium-Mycobacterium-Nocardia) grown on hydrocarbons has
been recognized for a long time (Haas et al., 1941; Haas & Bushnell, 1944;
Davis, 1967). The ketocarotenoids astaxanthin and canthaxanthin occur in a
limited number of those hydrocarbon utilizing bacteria.

Brevibacterium KY-4313: Canthaxanthin


Brevibacterium KY-4313 contains canthaxanthin and its biosynthetic precursors echinenone and p-carotene as the predominating carotenoids (Sakurai et
al., 1971). This organism has been used as a model to study the biosynthesis of
canthaxanthin (Hsieh et al., 1974). However, Japanese investigators, who
apparently had designed the bacterium mainly for petrochemical applications,
became aware of its industrial potential as a natural source of canthaxanthin.
Thus, a systematic optimization of the pigment level through variation of the
composition of the growth medium was undertaken (Tanaka et al., 1971).
Growth and pigment production were found to be intimately related. Optimal
growth (maximum cell yield about 35 g/liter) and maximum canthaxanthin
levels (06 mg/g, 1-2 mg/liter) resulted from using a basic medium containing
2-4% octadecane, 025% NH4 H 2 P04 , phosphates, magnesium sulfate and
minerals (pH 7), supplemented with vitamin B 12 , a natural nutrient, e.g. malt
extract (01 %) and a small amount of a non-ionic detergent (001 % Tween
40). High carbon to nitrogen ratios are known to favor the production of
cellular lipid in hydrocarbon fermentation (Ratledge, 1970). Accordingly,
formation of canthaxanthin was also promoted under these conditions. Light
did not affect carotenogenesis, while strong aeration was slightly inhibitory.
(Nelis & De Leenheer, unpublished)
Hydrocarbon fermentations suffer from a number of drawbacks (Einsele,
1983), including poor yields (low cell weights and slow growth rate), high costs
when pure alkanes are used and complex processing of the product (see
Section IV, 1). Pigment production by bacteria grown on paraffins is said to be
'very low' (Ninet & Renaut, 1979). Yet the figures cited above for canthaxanthin are not far below the astaxanthin levels obtained in Phaffia rhodozyma, an
organism which continues to be advocated as commercially attractive (Okagbue & Lewis, 1985). In our hands Brevibacterium KY-4313 tended to yield
more pigment in hydrocarbon than in non-hydrocarbon media. Therefore, a
systematic study was devoted to optimizing the canthaxanthin levels and
promoting the bacterial growth during hydrocarbon fermentation.


(a) Stimulation of carotenogenesis:

1. Optimization of the growth medium: Variation of the basic medium
constituents only marginally increased the canthaxanthin yield. Minor positive

Microbial Production of Carotenoids


effects were noted from combining the optimal carbon source, octadecane or
(cheaper) paraffin mixtures enriched (50-90%) in this compound, with certain
branched alkanes (pristane) or hydrocarbon mixtures rich in cycloalkanes.
Ammonium phosphate, ammonium acetate and sodium nitrate were nearly
equally effective as nitrogen sources. Yeast extract was slightly superior to malt
extract as a growth factor. The incorporation in the medium of chemicals
supposedly acting as precursors of carotenoid biosynthetic intermediates had
little or no effect. Examples of such compounds include acetate, pyruvate,
farnesol, geraniol, leucine and various sources of acetyl CoA. In addition,
Brevibacterium KY-4313 proved refractory to the action of a variety of
chemicals, which are reportedly carotenogenic in other micro-organisms,
particularly Blakeslea trispora: nitrogenous heterocyclic compounds, vegetable
oils, /3-ionone, retinol, aromatics. However, in non-hydrocarbon medium
(Brain Heart Infusion Broth) a number of substances were found to promote
substantially carotenoid biosynthesis, particularly alcohols (propanol and
isopropanol), glycerol and retinol. The latter caused accumulation of /3carotene rather than canthaxanthin. Vegetable oils greatly stimulated growth
but suppressed carotenoid formation.
2. Preparation of mutants: Following treatment of Brevibacterium KY-4313
with the mutagen NMU (Voznyakovskaya & Daraseliya, 1972), we isolated a
colony that distinguished itself from the bulk of the orange ones by its reddish
hue. This 'new', presumably mutant strain, contained about twice as much
canthaxanthin as the parent strain, but also more echinenone. Retreatment of
the 'new' strain with the same mutagen again permitted the isolation of a
slightly more pigmented colony. The latter was subsequently used for all
optimization work.

(b) Growth promotion: In view of the existing positive relationship between

growth and pigmentation, growth promotion was crucial to increase the overall
productivity of canthaxanthin. Two major factors are growth-limiting in
hydrocarbon fermentation, i.e. the availability (dispersion) of the insoluble
alkane (Einsele & Fiechter, 1971) and the secretion of toxic metabolites
(Einsele & Fiechter, 1971; Yamada et al., 1971). The effective dispersion of
octadecane, a liquid at the working temperature (30C) was facilitated by
thorough mechanical stirring and the addition of small amounts of a non-ionic
A major breakthrough came from removing the toxic metabolites, supposedly carboxylic acids, produced in the course of the bacterial growth. In
one fermentation on gaseous hydrocarbons, detoxification was accomplished
using continuous dialysis culture (Coty, 1968). As a result, the cell yield in this
particular example dramatically increased from 4 to 47 g/liter. However,
dialysis represents technical problems when liquid alkanes are used. We
mimicked its effect by regularly (each time when the pH of the medium
dropped to a constant minimum value) renewing the aqueous component of
the medium. The separation of the aqueous broth from the hydrocarbon layer,


H. 1. Nelis & A. P. De Leenheer

containing the cells, is very convenient because, after discontinuing the

stirring, the latter is floating on top of the medium. This semi-continuous
approach yielded dry cell weights of the order of 8-14 g/liter, as opposed to
3-4 g in the absence of the aqueous medium renewal. Concurrently, the
ketocarotenoid yield increased to 13 mg/liter (1-15 mg/g dry weight), of
which 70% was canthaxanthin and the remainder echinenone, obtained from
50 ml of hydrocarbon (working volume of the fermenter, 125 liters) in a 7-day
run. This figure compares favorably with the astaxanthin yield obtained from
Phaffia rhodozyma (13-1 5 mg/liter in 60 h using the mixed culture conditions
described above).
4.5.2 Rhodococcus maris/Mycobacterium brevicale
As part of an investigation of hydrocarbon-oxidizing marine bacteria (Koronelli et al., 1981), Russian workers isolated a pigmented Mycobacterium
brevicale strain (32-MCT) (Koronelli et al., 1982), later found to be synonymous with Rhodococcus maris (Koronelli et al., 1987). This organism contains
a dark orange peptidolipid with canthaxanthin as the main carotenoid
(Koronelli et al., 1987; Pachlavuni, 1987). Details on culturing conditions were
not available at the time of writing, except that the organism was grown in a
mineral medium containing hexadecane as a carbon source (Pachlavuni et al.,
1987). Rhodococcus maris / Mycobacterium brevicale 32-MCT is apparently not
equivalent with the Rhodococcus maris type strain described by Nesterenko
(Nesterenko et al., 1982). The latter also grows on various hydrocarbons and
produces canthaxanthin as well (personal observation, unpublished). However,
pigment levels were lower than in Brevibacterium KY-4313 and reproducibility
of carotenoid formation and growth on hydrocarbons was poorer (personal
observations, unpublished).

4.5.3 Miscellaneous Hydrocarbon-Utilizing Bacteria

Mycobacterium lacticola was the first bacterium reported to form astaxanthin in
hydrocarbon medium (Haas & Bushnell, 1944). Remarkably, the pigment was
absent when the organism was grown on nutrient agar. Brevibacterium 103 is
the second known example of a hydrocarbon utilizing bacterium that produces
astaxanthin (Iizuka & Nishimura, 1969). A mineral medium containing 8%
kerosene yielded approximately 3 g cells/liter, with a pigment content of
30 ",g/g dry cell weight. In view of this poor yield and the availability of Phaffia
rhodozyma, there appears to be no future for either of the above bacteria as a
useful source of astaxanthin.
4.5.4 Micrococcus roseus: a controversy
The presence of canthaxanthin and other ketocarotenoids has also been
reported, by one particular research group, in Micrococcus roseus (ATCC 516)
grown in a variety of (non-hydrocarbon) media (Cooney et al., 1966; Ungers &
Cooney, 1968; Schwartzel & Cooney, 1970, 1972, 1974a,b; Ascenzi & Cooney,
1975; Dieringer et al., 1977; Cooney & Berry, 1981). However, despite this

Microbial Production of Carotenoids


extensive evidence, we as well as others (unpublished results) have repeatedly

failed to confirm this finding. We re-investigated the pigment composition of
Micrococcus roseus ATCC 516 and of the original strain, donated by Cooney,
using HPLC and photodiode array detection (Nelis & De Leenheer, 1987).
Although a chromatographic peak corresponding in retention to canthaxanthin
was observed, its absorption spectrum, recorded with the aid of a photodiode
array detector (see Section IV, 2) was in no way consistent with this structure.
None of the pigments separated displayed an absorption spectrum indicative of
any ketocarotenoid. The reason for this extreme discrepancy of results has
remained obscure so far.



By virtue of its deep red color, rhodoxanthin, which has been approved as a
food additive in certain countries, might be envisaged as an alternative to
astaxanthin. The compound occurs in autumn leaves (Ida, 1981), fish (Katsuyama & Matsuno, 1979) and the feathers of birds (Brush, 1981).
5.2 Non-Photosynthetic Bacteria
Rhodoxanthin was tentatively identified as the main carotenoid in the
methylotroph Pseudomonas extorquens (Downs & Harrison, 1974) (obviously
equivalent with Protomonas extorquens (Urakami & Komagata, 1986. The
pigment was formed during growth in a mineral medium consisting of
ammonium sulfate as a nitrogen source and methanol or glycerol as a carbon
source. Oxygen and magnesium limitation caused a marked increase in cell
pigmentation, suggesting that conditions of growth restriction in general tend
to divert the available carbon to pigment synthesis. However, no attempts
have been made to quantitate or to optimize the actual carotenoid content.



Three selected examples will be briefly discussed of carotenoid-containing

micro-organisms whose primary biotechnological potential lies in a field other
than pigment production. The carotenoids present in those bacteria and yeasts
have in general not been approved yet as additives for feeds or foods.
However, this situation could obviously change in the future if these organisms
gain acceptance for large-scale biomass or single lipid production and, hence,
addition to animal feeds.

H. 1. Nelis & A. P. De Leenheer



Various red yeasts belonging to the genera Rhodotorula and Rhodosporidium
(Basidiomycetes) have been considered as industrial sources of single cell
protein (Lichtfield, 1979; Costa et al., 1984) and microbial fat (Ratledge,
1982). These organisms can be grown on cheap substrates, e_g. domestic
seawage, potato hydrolysate, whey, molasses, etc_ (Lichtfield, 1979; Ratledge,
Concurrently with fat accumulation, enhanced carotenoid synthesis can be
reasonably expected_ The major carotenoids in Rhodotorula and
Rhodosporidium are p-carotene, j'-carotene, torulene and torularhodin (Simpson et al., 1971; Goodwin, 1972). Supplementation of poultry feeds with
preparations of Rhodotorula mucilaginosa were found to promote egg yolk
pigmentation (Schwarz & Margalith, 1965). This yeast was propagated in
glucose-peptone-yeast extract media or, alternatively, in chemically defined
media containing a carbohydrate (5-10% glucose, beet molasses), ammonium
sulfate, urea and/or ammonium lactate as nitrogen sources, as well as minerals
and growth factors (yeast extract, thiamin). The total carotenoid yield on a dry
weight basis was as high as I-Smg/g (corresponding to 49mg/liter). Unlike in
Blakeslea trispora, the addition of p-ionone or kerosene inhibited carotenogenesis, whereas non-ionic detergents were only slightly stimulatory. However, a
disadvantage of using Rhodotorula mucilaginosa biomass as an additive for
poultry feed was the appearance of a violet hue in the egg yolk, caused by the
presence of torularhodin (Margalith & Meydav, 1968). Therefore, subsequent
efforts were directed towards reducing the torularhodin/torulene ratio. This
could be accomplished by adding diphenylamine, ethanol or isopropanol to a
glucose-ammonium lactate medium (Margalith & Meydav, 1968). Unlike
diphenylamine, the latter two additives also promoted the total carotenoid
yield, which however remained below 0-4 mg/g dry weight. The minimal
torularhodin/torulene ratio was 0-33, vs 1-00 in the absence of additives. Other
workers obtained a value of 048 using a glucose-asparagine medium
(Peterson et al., 1958). Both figures contrast with the unacceptably high ratio
of 6-12, resulting from the use of sucrose-ammonium sulfate (Villoutreix,


A number of methylotrophic bacteria are known vitamin B12 producers

(Toraya et al., 1975; Sato et al., 1977; Shimizu et al., 1982; Dumenil et al.,
1983) and potentially useful sources of single cell protein (Lichtfield, 1979,
1980). The commonly encountered pink color of these bacteria is due to the
presence of, mostly unidentified, carotenoids (Shimizu et al., 1982; Dumenil et
al., 1983; Urakami & Komagata, 1986), sometimes of a rhodoxanthin-like
nature (Downs & Harrison, 1974; Urakami & Komagata, 1986) (see also

Microbial Production of Carotenoids


Section 11,5). Attempts have been made to optimize the carotenoid yield
from selected methylotrophs by altering the culturing conditions. A
Corynebacterium sp. XG was found to form most pigment on methanol,
dimethyl amine and methylamine as carbon sources (Dumenil et al., 1983).
Other medium ingredients included ammonium sulfate, minerals and yeast
extract. Light, pH variation and the nature of the growth factor did not
appreciably affect pigmentation. Protaminobacter ruber, another facultative
methylotroph, produced 2 mg/liter of an unidentified pink carotenoid in a
medium containing 1,2-propanediol as a carbon substrate and methionine and
riboflavine as moderately effective carotenogenic agents (Shimizu et al., 1982).
Both examples demonstrate the interest in methylotrophic bacteria for their
carotenoid composition. Although their primary biotechnological potential
clearly lies in the production of vitamin B12 and biomass, the presence of
useful pigments again adds extra potential value to these organisms.




Purple photosynthetic bacteria, particularly Rhodobacter capsulatus (formerly

Rhodopseudomonas capsulatus, Imhoff et al., 1984) are useful sources of single
cell protein (Kobayashi & Tchan, 1973; Shipman et aI., 1977; Kobayashi &
Kurata, 1978; Lichtfield, 1979, 1980, 1983; Driessens et aI., 1987). Their
ability to grow on waste solutions makes them attractive for the purification
of polluted industrial effluents, giving highly nutritious cells as a valuable
by-product (Kobayashi & Tchan, 1973). Apart from its general nutritional
value, as evidenced from its amino acid profile (Kobayashi & Kurata, 1978;
Driessens et al., 1987) Rhodobacter capsulatus reportedly also promotes egg
production in chickens and increases the survival of fish (Kobayashi, 1972;
Kobayashi & Tehan, 1973).
Although the nature of carotenoids in Rhodobacter capsulatus has been
established for a long time (for reviews see Liaaen-Jensen, 1978; Schmidt,
1978; Goodwin, 1980), Driessens & Verstraete were the first to recognize the
potential of this organism as a useful natural source of pigments (Driessens et
al., 1987 and unpublished results). Spheroidene is the predominating carotenoid in Rhodobacter capsulatus when grown anaerobically, whereas under
aerobic conditions this compound is further oxidized to the ketocarotenoid
spheroidenone (Goodwin 1980; Spurgeon & Porter, 1983). Because of its
favorable color characteristics, the latter in particular is interesting from a
practical! economical standpoint. In the course of a thorough study of optimal
biomass production by Rhodobacter capsulatus, the co-production of carotenoids was also examined (Driessens et al., 1987).


H. J. Nelis & A. P. De Leenheer

4.2 Production of Carotenoids by Rhodobacter capsulatus

4.2.1 Batch culture
Initially, when Rhodobacter capsulatus was grown photo heterotrophically in
batch culture (lactic acid as carbon source, ammonium sulfate as nitrogen source,
anaerobic conditions and illumination) high pigment contents (5-10 mg/g)
were obtained. Spheroidene represented, as predicted, the major carotenoid
(>85%), whereas spheroidenone was only present in amounts ranging from 1
to 5%. Subsequent attempts were, for unknown reasons, less successful.
Under anaerobic conditions, the maximum levels of spheroidene and spheroidenone were 1 and 03 mg/g, respectively. Corresponding values obtained in
aerobic culture were 005 and 04 mg/g, respectively. Unexpectedly, in one
anaerobic fermentation cells containing 06 mg spheroidenone and virtually no
spheroidene were produced.
However, batch fermentations of Rhodobacter capsulatus are associated with
low cell yields (1-2 g/liter) (Kobayashi & Kurata, 1978; Lichtfield, 1983).
4.2.2 Culture in a continuous flow reactor (Driessens et aI., 1987)
Culture of Rhodobacter capsulatus in a continuous flow-through reactor, under
external illumination and microaerophilic conditions, yielded 5-10 times
higher biomass production rates. The photobioreactor used consisted essentially of a glass tube having a PVC decanter on top and two ports at the
bottom, one for the removal of excess biomass and one for the entry of
(recirculated) medium. The PVC decanter contained ports for effluent
removal, gas outlet, medium recirculation and nitrogen flushing (when
required). Two non-axenic feed solutions were supplied separately, by means
of a peristaltic pump. The first one contained the buffered carbon source
(preferably calcium lactate) and the second one the nitrogen source (ammonium sulfate) plus minerals, growth factors as well as selective inhibitors of
sulfate reducing bacteria (sodium molybdate) and algae (chloroxuron). The
bacteria readily flocculated in the reactor, which permits easy harvesting.
The quality of the resulting biomass did not differ from the one obtained in
batch culture. Under conditions of iron excess (4 mg Fe2 + /liter) about 1 mg
pigment per g dry weight was produced, of which 07 mg was spheroidene and
03 mg spheroidenone. However, because of their high toxicity, the compounds added to suppress competing micro-organisms had to be eliminated, in
order to make the end-product nutritionally acceptable. To maintain the
non-axenic conditions and yet to avoid contamination the iron content of the
medium had to be drastically lowered (down to 251lg Fe2 + /liter). This did not
appreciably affect the growth, but dramatically altered the carotenoid profile.
A number of unidentified, red lycopene-like compounds were now formed in
amounts ranging from 04 to 08 mgt g, but neither spheroidenone or spheroidene were present. A compromise in the iron concentration (with incremental
addition) still resulted in a similar carotenoid pattern. Although the cells, when
added to pOUltry feed, improved the color of the egg yolk without any signs of

Microbial Production of Carotenoids


toxicity (provided the sodium molybdate had been eliminated), the uncertainty
about the exact pigment composition precludes, for the time being, the use of
Rhodobacter capsulatus as a pigmenter for chickens.




Because carotenoids in micro-organisms are localized intracellularly, their

recovery essentially comes down to the isolation of the biomass. Wellestablished approaches for the harvesting of cells include filtration, centrifugation and coagulation/flocculation (Belter, 1979). However, hydrocarbon fermentations pose a number of additional recovery problems due to the presence
of residual alkanes (Levi et al., 1979; Einsele, 1983). After separation of the
aqueous and the organic phases by centrifugation or flotation, often in
conjunction with specialized techniques (Various patents, 1973), the residual
hydrocarbon, adsorbed on the cells, has to be removed. Solvent extraction or
treatment with surfactants have been suggested for this purpose (Einsele &
Fiechter, 1971). However, this purification will cause a very substantial part of
the carotenoids to leach from the cells and to dissolve in the hydrocarbon
(paraffin) or the extraction solvent (personal observations). The resulting
'clean' cells will not only be defatted, but also be devoid of most of the
carotenoids. The separation of the carotenoids from the hydrocarbon itself at a
later stage is unfeasible. Therefore, in hydrocarbon fermentation a basic
incompatibility appears to exist between the recovery of pure biomass and the
quantitative isolation of the intracellular carotenoids.
If the dry biomass does contain the pigments (as in non-hydrocarbon
fermentation) it could probably be directly added to animal feed, provided the
carotenoid concentration is sufficiently high and no toxic compounds are
present. In view of the instability of carotenoids, conditions for drying (sprayor flash-drying) should be as mild as possible, to avoid oxidation and cis /trans
isomerization. The addition of antioxidants (BHT, ethoxyquin) might be
desirable (Wood et aI., 1983). Alternatively, if the pigment quantity in the
biomass is too dilute, solvent extraction is required for its isolation (useful
organic solvents are mentioned in Section IV, 2).


Traditionally, the determination of carotenoids in biological materials, including micro-organisms, has always relied on spectrophotometry (Davies, 1976;
Britton, 1985). A typical sample pretreatment consists of a total lipid
extraction using such polar organic solvents as acetone or methanol, saponification to remove triglycerides, partition between hexane/methanol (separation


H. J. Nelis & A. P. De Leenheer

of non-polar epiphasic from more polar hypophasic derivatives) and open

column chromatography (Davies, 1976; Britton, 1985). Each colored band
eluting from the column is subjected off-line to spectrophotometric measurement at its Amax, with calculation based on the molar absorption coefficient of
the (tentatively) identified compound of interest. Often, this complex purification of total lipid extracts is omitted when only one pigment is presumably
present. That this may lead to errors is exemplified in the spectrophotometric
assay of astaxanthin in Phaffia rhodozyma, where the interference of colored
substances adsorbed from the medium onto the cells led to an overestimation
of the astaxanthin content (Okagbue & Lewis, 1984a). This shows that a
chromatographic fractionation is an essential step in carotenoid assays.
As a separation method, modern liquid chromatography on microparticulate
materials (HPLC) has now superseded open column chromatography because
of its superior resolving power, higher sensitivity and improved speed. HPLC
has been used for the profiling and quantitation of carotenoids in algae
(Braumann & Grimme, 1981; Mantoura & Llewelynn, 1983; Bowles et ai.,
1985) and bacteria (Gillan & Johns, 1983; Kester & Thompson, 1984; Korthals
& Steenbergen, 1985; Nelis & De Leenheer, 1987).
A recent development is the use of photodiode array absorbance detection
(Miller et ai., 1982) in connection with the HPLC of carotenoids (Gillan &
Johns, 1983; Mantoura & Llewelynn, 1984; Ruedi, 1985; Nelis & De
Leenheer, 1986). This versatile technique permits a (tentative) identification of
each chromatographic peak, based on its on-line recorded absorption
spectrum. The need for such a second identity criterion (aside from comparison with a standard compound) was clearly evidenced from the work with
Micrococcus roseus (see Section III, 4.5.4).
One issue that is poorly addressed in the literature is the problem of the
complete carotenoid extraction from micro-organisms. We observed that even
repeated treatment of several bacteria, e.g. Brevibacterium KY-4313 and
Micrococcus roseus with acetone or methanol failed to release all the pigment
(Nelis & De Leenheer, 1986, 1987). Also, from streptococci, carotenoids were
reportedly poorly extractable using a variety of organic solvents and even
alcoholic potassium hydroxide (Merritt & Jacobs, 1978). In the latter case,
efficient extraction was accomplished using a rather unusual mixture of
glycerol-fenol. Other investigators added trichloroacetic acid (Downs &
Harrison, 1974) or potassium hydroxide (Taylor, 1980) to promote the
liberation of carotenoids from bacteria. In our hands a brief pretreatment of
the cells with pure liquefied fenol was found to make them spectacularly more
amenable to extraction with methanol: this approach resulted in a total
bleaching of the residue, even for the most refractory species (Micrococcus
roseus). Fenol can be removed from the extract by partitioning between
diethyl ether and an alkaline water phase or, alternatively, by column
chromatography on microparticulate C 1S mini-columns (Bond Elut, Sep-Pak).
The ether or the column eluate is subsequently evaporated to dryness and the
residue is redissolved in the chromatographic solvent. An aliquot (50-100 Ill)

Microbial Production of Carotenoids


of the final mixture is injected on a reversed phase column, eluted with a

non-aqueous mobile phase (Nelis & De Leenheer, 1983, 1986, 1987). When, as
in the case of Brevibacterium KY-4313, hydrocarbons are co-extracted, a
chromatographic step on a Bond Elut Alumina mini-column is included to
separate the carotenoids (canthaxanthin) from the hydrocarbon.


The profitability of the biotechnological production of carotenoids has to be

judged in the light of the availability of alternative chemical routes for their
preparation. Virtually all derivatives can be prepared by total synthesis in
economically feasible processes (for an extensive review, see Mayer & Isler,
1971). This applies in particular to the carotenoids of high commercial interest,
i.e. f3-carotene, canthaxanthin, astaxanthin, f3-apo-8' -carotenal and f3-apo-8'carotenoic acid ethyl ester. Leading European manufacturers are F. HoffmannLa Roche (Basle, Switzerland) and BASF (Ludwigshafen, FRG).
Basically, the C 40 carotenoid skeleton is synthesized by combining two or
three building units in symmetrical (e.g. C 20 + C 20 , C 19 + C 2 + C 19 , CiS + C lO +
CiS) or asymmetrical schemes (e.g. ~i + C 19 , C 2S + CiS). The two most
frequently used methods are the Wittig reaction (condensation of a carbonyl
compound with a trimethylphosphonium halide) and the Grignard reaction
(condensation of a carbonyl compound with a metal acetylide). New developments continue to appear, particularly with respect to the total synthesis of
optically active carotenoids (Mayer, 1979; Muller et aI., 1982). This is a new

Fig. 4. Reduction reactions involved in the preparation of an optically active key

intermediate in the synthesis of (3R,3/R)-zeaxanthin. A, enantioselective hydrogenation of double bond, catalyzed by baker's yeast; B, chemical reduction; C, reduction of
keto group, catalyzed by baker's yeast (unwanted side reaction). 1, desired reaction
product (4R, 6R); 2, unwanted reaction product (4S, 6R).


H. J. Nelis & A. P. De Leenheer







Fig. 5. Enantioselective reduction catalyzed by baker's yeast as a first step in the

synthesis of (3S, 3'S}-astaxanthin.

area in which micro-organisms have found large-scale application. Synthetic

procedures for (3R,3'R)-zeaxanthin and (3S,3'S)-astaxanthin indeed involve
the use of micro-organisms as biocatalysts for the introduction of chiral
centers, by taking advantage of the stereoselective course of microbially
catalyzed reactions (Leuenberger, 1985). Thus the optically active synthon for
the synthesis of (3R,3'R)-zeaxanthin has been prepared by a combined
process consisting of enantioselective reduction of oxo-isophorone (hydrogenation of double bond) with baker's yeast (Fig. 4A) and chemical hydrogenation (Fig. 4B). However, special precautions have to be taken in the
microbial process to avoid reduction of the 4-keto group of oxo-isophorone
(Fig. 4C). This secondary reaction leads to the unwanted S configuration at
C-4. A synthetic route for the preparation of (3S,3'S)-astaxanthin also
involves the use of baker's yeast to carry out an enantioselective reduction. In
this particular reaction the 3-keto group of 4-hydroxy-6-oxo-isophorone is
specifically attacked to yield a tautomeric mixture of optically active ketodiols
(Fig. 5). Both reaction products can be converted chemically (acetonization) to
a ketal, a key intermediate in the further synthesis of (3S, 3'S)-astaxanthin.
Another type of useful reaction catalyzed by micro-organisms is microbial
resolution of racemates. In an improved synthetic process for (3S,3'S)astaxanthin a Bacillus sp. selectively hydrolyzes the 3R enantiomer of racemic
3-acetoxy-4-oxo-fl-ionone, but leaves the desired S-enantiomer untouched
(Fig. 6). The latter is subsequently isolated in high yield by physico-chemical

Fig. 6. Microbial resolution by a Bacillus sp. of racemic 3-acetoxy-4-oxo-tJ-ionone, a

key step in the synthesis of (3S,3'S}-astaxanthin. 1, unwanted reaction product (3S)
(hydrolyzed); 2, desired reaction product (3R) (untouched).

Microbial Production of Carotenoids


The efficiency of microbially catalyzed reactions can be enhanced by

immobilization of the cells, the use of two-phase biocatalysis (for water
insoluble substrates such as carotenoids) or via recombinant DNA technology.


So far there are no examples of microbiological fermentations of carotenoids

that are competitive with their total synthesis or their isolation from plants. A
limited number of processes, e.g. the production of lutein from green algae
have a mere historical value, as their development and use have been
discontinued since the sixties. The production of lycopene from fungi is, to the
best of our knowledge, not further developed at present. The most suitable
candidates for fermentation are carotenoids with chiral centers, which are
more difficult to synthesize than others. A process for the production of
(3R,3'R)-zeaxanthin from Flavobacterium sp. appears to approach the stage
of commercialization. The same presumably applies to Phaffia rhodozyma as a
source of astaxanthin, the absolute configuration of which is, however,
opposite to the desirable one (3R, 3'R instead of 3S, 3'S). To a lesser extent
there may be a future for a fermentative process for canthaxanthin preparation, although no really useful organisms have been found yet. Still ketocarotenoids remain valuable compounds because of their efficiency as pigmenters in
poultry and fish feeds. The future will tell whether a market for 'natural'
canthaxanthin and astaxanthin will develop as an alternative to their synthetic
Andrewes, A. G. & Starr, M. P. (1976). (3R, 3'R)-Astaxanthin from the yeast Phaffia
rhodozyma. Phytochem., 15,1009-11.
Andrewes, A. G., Phaff, H. J. & Starr, M. P. (1976). Carotenoids of Phaffia
rhodozyma, a red-pigmented fermenting yeast. Phytochem., 15, 1003-7.
Anonymous (1962). Algae-derived xanthophylls for pigmentation control. Feeds
Illustrated, November issue, 24-25.
Anonymous (1966-1967). Oxycarotenoid preparations. Microbiology Abstracts, Section
A, 2, A1056.
Anonymous (1969-1970a). Microbiological production of carotenoid-containing additives for animal feeds. Microbiology Abstracts, Section A, 5, A200l.
Anonymous (1969-1970b). Production of bacterial preparations with a high content of
oxy-carotenoids. Microbiology Abstracts, Section A, 5, A2292.
Arpin, N., Lebreton, J. & Fiasson, J.-L. (1966). Recherches chimiotaxinomiques sur
les champignons. II. Les caroteno"ides de Peniophora aurantiaca (Bres.)
(Basidiomycete). Bull. Soc. Mycol. France, 82,450-59.
Ascenzi, J. M. & Cooney, J. J. (1975). Action of visible light on enzymes in cell
envelopes of Micrococcus roseus. Photochem. Photobiol., 21, 307-1l.
Bamji, M. S. & Krinsky, N. I. (1966). The carotenoid pigments of a radiation-resistant
Micrococcus species. Biochim. Biophys. Acta, 115, 276-84.


H. I. Nelis & A. P. De Leenheer

Belter, P. A. (1979). General procedures for isolation of fermentation products. In

Microbial Technology, Vol. II, 2nd edn, ed. H. J. Peppler & D. Perlman. Academic
Press, New York, pp. 403-32.
Bowles, N. D., Paerl, H. W. & Tucker, J. (1985). Effective solvents and extraction
periods employed in phytoplankton carotenoid and chlorophyll determinations.
Can. J. Fish. Aquat. Sci., 42, 1127-31.
Braumann, T. & Grimme, L. H. (1981). Reversed-phase high-performance liquid
chromatography of chlorophylls and carotenoids. Biochim. Biophys. Acta, 637,
Britton, G. (1985). General carotenoid methods. Meth. Enzym., 111, 113-49.
Brush, A. H. (1981). Carotenoids in wild and captive birds. In Carotenoids as Colorants
and Vitamin A Precursors, ed. J. C. Bauernfeind. Academic Press, New York, pp.
Burnett, J. H. (1976). Functions of carotenoids other than in photosynthesis. In
Chemistry and Biochemistry of Plant Pigments, Vol. 1, 2nd edn, ed. T. W.
Goodwin. Academic Press, New York, pp. 655-79.
Castillo, R., Negre-Sadargues, G. & Lenel, R. (1982). General survey of the
carotenoids Crustacea. In Carotenoid Chemistry and Biochemistry, ed. G. Britton &
T. W. Goodwin. Pergamon Press, Oxford, pp. 211-24.
Ciegler, A. (1965). Microbial carotenogenesis. Adv. Appl. Microbiol., 7, 1-34.
Cooney, J. J. & Berry, R. A. (1981). Inhibition of carotenoid synthesis in Micrococcus
roseus. Can. J. Microbiol., 27,421-5.
Cooney, J. J., Marks, H. W. & Smith, A. M. (1966). Isolation and identification of
canthaxanthin from Micrococcus roseus. J. Bacteriol., 92, 342-5.
Costa, I., Martelli, H. L., da Silva, I. M. & Pomeroy, D. (1984). Produl;llo de
carotenoides por fermental;llo. I. Produc;ao de biomassa de Rhodotorula glutinis var.
glutinis. Rev. Microbiol., Sao Paulo, 15, 109-13.
Coty, V. F. (1968). Dialysis to improve cell growth. US patent 3,418,208. Cited in
Proteins From Hydrocarbons, Food Technol. Rev., no. 4, ed. S. Gutcho. Noyes
Data Corporation, Park Ridge, NJ, pp. 91-3, 1973.
Czygan, F.-C. (1968). Sekundar-Carotinoide in Griinalgen. I. Chemie, Vorkommen
und Faktoren welche die Bildung dieser Polyene beeinflussen. Arch. Mikrobiol., 61,
Davies, B. H. (1976). Carotenoids. In Chemistry and Biochemistry of Plant Pigments,
Vol. 2, 2nd edn., ed. T. W. Goodwin. Academic Press, New York, pp.
Davis, J. B. (1967). Petroleum Microbiology. Elsevier, Amsterdam, pp. 311-15.
Dieringer, S. M., Singer, J. T. & Cooney, J. J. (1977). Photokilling of Micrococcus
roseus. Photochem. Photobiol., 26, 393-6.
Downs, J. & Harrison, D. E. F. (1974). Studies on the production of pink pigment in
Pseudomonas extorquens NCIB 9399 growing in continuous culture. 1. Appl.
Bacteriol., 37,65-74.
Driessens, K., Liessens, J., Masduki, S., Verstraete, W., Nelis, H. & De Leenheer, A.
(1987). Production of Rhodobacter capsulatus ATCC 23782 with short residence
time in a continuous flow photobioreactor. Process. Biochem., 22, 160-4.
Dumenil, G., Laget, M., Cremieux, A., Millon, M. & Brousse, M. (1983). Etude des
conditions de pigmentation d'une souche bacterienne methylotrophe facultative.
Ann. Pharm. Franc., 41,427-35.
Einse1e, A. (1983), Biomass from higher n-alkanes. In Biotechnology, Vol. 3, ed. H.
Dellweg. VCH, Weinheim, pp. 44-81.
Einsele, A. & Fiechter, A. (1971). Liquid and solid hydrocarbons. Adv. Biochem.
Eng., 1, 169-94.
Gillan, F. T. & Johns, R. B. (1983). Normal-phase HPLC analysis of microbial
carotenoids and neutral lipids. 1. Chromatogr. Sci., 21, 34-8.

Microbial Production of Carotenoids


Goodwin, T. W. (1972). Carotenoids in fungi and non-photosynthetic bacteria. Progr.

Industr. Microbiol., 11, 29-88.
Goodwin, T. W. (1980). The Biochemistry of the Carotenoids, Vol. I, Plants. Chapman
& Hall, London.
Goodwin, T. W. (1984). The Biochemistry of the Carotenoids, Vol. II, Animals.
Chapman & Hall, London.
Goodwin, T. W. & Jamikorn, M. (1956). Studies in carotenogenesis. 17. The
carotenoids produced by different strains of Mycobacterium phlei. Biochem. J., 62,
Haas, H. F. & Bushnell, L. D. (1944). The production of carotenoid pigments from
mineral oil by bacteria. J. Bacteriol., 48, 219-3l.
Haas, H. F., Yantzi, M. F. & Bushnell, L. D. (1941). Microbial utilization of
hydrocarbons. Trans. Kansas Acad. Sci., 44,39-45.
Hanson, A. M. (1967). Microbial production of pigments and vitamins. In Microbial
Technology, ed. H. J. Peppler. Reinhold, New York, pp. 222-50.
Haxo, F. (1950). Carotenoids of the mushroom Cantharellus cinnabarinus. Bot. Gaz.,
111, 228-32.
Hochmannova, J., Kolmanova, A. & Malek, I. (1968). Comparison of pigments
produced by several strains of Mycobacterium phlei. Fol. Microbiol., 13, 505-9.
Hsieh, L. K., Lee, T.-C., Chichester, C. O. & Simpson, K. L. (1974). Biosynthesis of
carotenoids in Brevibacterium sp. KY-4313. J. Bacteriol., 118,385-93.
Hsu, W. J., Yokoyama, H. & Coggins, C. W. (1972). Carotenoid biosynthesis in
Blakeslea trispora. Phytochem., 11, 2985-90.
Ida, K. (1981). Eco-physiological studies on the response of taxodiaceous conifers to
shading, with special reference to the behaviour of leaf pigments. I. Distribution of
carotenoids in green and autumnal reddish brown leaves of gymnosperms. Bot.
Mag. (Tokyo), 94, 41-54.
Iizuka, H. & Nishimura, Y. (1969). Microbiological studies on petroleum and natural
gas. X. Carotenoid pigments of hydrocarbon utilizing bacteria. J. Gen. Appl.
Microbiol., 15, 127-34.
Imhoff, J. F., Truper, H. G. & Pfennig, N. (1984). Rearrangement of the species and
genera of the phototrophic "purple nonsulfur bacteria". Int. J. Syst. Bacteriol.. 34,
Ingraham, M. A. & Baumann, C. A. (1934). The relation of microorganisms to
carotenoids and vitamin A. J. Bacteriol., 28, 31-40.
Ingraham, M. A. & Steenbock, H. (1935). The relation of micro-organisms to
carotenoids and vitamin A. II. The production of carotenoids by Mycobacterium
phlei. Biochem. J., 29, 2553-62.
Isler, O. (1971). Introduction. In Carotenoids, ed. O. Isler. Birkhauser, Basle.
Switzerland, p. 13.
IUPAC Commission on the Nomenclature of Organic Chemistry and IUPAC-IUB
Commission on Biochemical Nomenclature (1971). Tentative rules for the nomenclature of carotenoids. In Carotenoids, ed. O. Isler. Birkhauser, Basle, Switzerland
pp. 851-64.
Johnson, E. A. & Lewis, M. J. (1979). Astaxanthin formation by the yeast Phaffia
rhodozyma. J. Gen. Microbiol., 115, 173-83.
Johnson, E. A., Conklin, D. E. & Lewis, M. J. (1977). The yeast Phaffia rhodozyma as
a dietary pigment source for salmonids and crustaceans. J. Fish. Res. Board Can.,
Johnson, E. A., Villa, T. G., Lewis, M. J. & Phaff, H. J. (1978). Simple method for
the isolation of astaxanthin from the Basidiomycetous yeast Phaffia rhodozyma.
Appl. Environ. Microbiol., 35, 1155-59.
Johnson, E. A., Villa, T. G. & Lewis, M. J. (1980a). Phaffia rhodozyma as an
astaxanthin source in salmonid diets. Aquaculture, 20, 123-34.


H. J. Nelis & A. P. De Leenheer

Johnson, E. A., Lewis, M. J. & Grau, C. R. (1980b). Pigmentation of egg yolks with
astaxanthin from the yeast Phaffia rhodozyma. Poultry Sci., 59, 1777-82.
Katsuyama, T. & Matsuno, T. (1979). Isolation and identification ofrhodoxanthin from
the fish Tilapia nilotica. Bull. lap. Soc. Scient. Fish., 45, 1045.
Kester, A. S. & Thompson, R. E. (1984). Computer-optimized normal-phase highperformance liquid chromatographic separation of Corynebacterium poinsettiae
carotenoids. I. Chromatogr., 310, 372-8.
Klaui, H. (1982). Industrial and commercial uses of carotenoids. In Carotenoid
Chemistry and Biochemistry, ed. G. Britton & T. W. Goodwin, Pergamon Press,
Oxford, pp. 308-17.
Klaui, H. & Bauernfeind, J. C. (1981). Carotenoids as food colors. In Carotenoids as
Food Colorants and Vitamin A Precursors, ed. J. C. Bauernfeind. Academic Press,
New York, pp. 48-317.
Kobayashi, M. (1972). Utilization of photosynthetic bacteria. In Fermentation Technology Today (Proceedings of the Wth International Fermentation Symposium), ed. G.
Terui, pp. 527-31.
Kobayashi, M. & Kurata, S. (1978). The mass culture and cell utilization of
photosynthetic bacteria. Process Biochem., 13,27-30.
Kobayashi, M. & Tchan, Y. T. (1973). Treatment of industrial waste solutions and
production of useful by-products using a photosynthetic bacterial method. Water
Res., 7, 1219-24.
Koronelli, T. V., Kalyuzhnaya, T. V. & Rozynov, B. V. (1981). Pigmented arctic
mycobacteria and their role in the oxidation of hydrocarbons. Mikrobiologiya, 50,
167-70; Microbiology (English translation), SO, 123-5.
Koronelli, T. V., Pahlavuni, I. K. & Rozynov, B. V. (1982). A pigmented amphiphilic
biopolymer from marine mycobacteria. Mikrobiologiya, 51,873-5.
Koronelli, T. V., Pakhlavuni, I. K., Voroybyeva, I. A. & Balashov, S. P. (1987). The
lipopeptidocarotenoid complex and cells of Rhodococcus maris studied by spectroscopy. Mikrobiologiya, 56,890-91.
Korthals, H. J. & Steenbergen, C. L. M. (1985). Separation and quantification of
pigments from natural phototrophic microbial populations. FEMS Microbiol. Ecol.,
Koyama, Y., Takii, T., Saiki, K. & Tsukida, K. (1983). Configuration of the carotenoid
in the reaction centers of photosynthetic bacteria (2). Comparison of the resonance
Raman lines of the reaction centers with those of 14 cis-trans isomers of f3-carotene.
Photobiochem. Photobiophys. 3, 139-50.
Krinsky, N. I. (1979). Carotenoid protection against oxidation. Pure Appl. Chem., 51,
Lem, N. W. & Glick, B. R. (1985). Biotechnological uses of Cyanobacteria.
Biotechnol. Adv., 3, 195-208.
Leuenberger, H. G. W. (1985). Microbiologically catalyzed reaction steps in the field of
vitamin and carotenoid syntheses. In Biocatalysts in Organic Syntheses, ed. J.
Tramper, H. C. van der Plas & P. Linko, Studies in Organic Chemistry, Vol. 22.
Elsevier, Amsterdam, pp. 99-118.
Levi, J. D., Shennan, J. L. & Ebbon, G. P. (1979). Biomass from liquid n-alkanes. In
Economic Microbiology, Vol. 4, Microbial Biomass, ed. A. H. Rose, Academic
Press, New York, pp. 361-419.
Lewis, N. F. & Kumta, U. S. (1973). Lycopene accumulation in pigmented radioresistant Micrococci grown in presence of nicotine. FEBS Lett., 30, 144-6.
Liaaen-Jensen, S. (1978). Chemistry of carotenoid pigments. In The Photosynthetic
Bacteria, ed. R. K. Clayton & W. R. Sistrom. Plenum Press, New York, pp.
Lichtfield, J. H. (1979). Production of single-cell protein for use in food and feed. In
Microbial Technology, Vol. I, 2nd edn, ed. H. J. Peppler & D. Perlman. Academic
Press, New York, pp. 93-155.

Microbial Production of Carotenoids


Lichtfield, J. H. (1980). Microbial protein production. BioScience, 30, 387-96.

Lichtfield, J. H. (1983). Single-cell proteins. Science, 219,740-6.
McDermott, J. C. B., Ben-Aziz, A., Singh, R. K., Britton, G. & Goodwin, T. W.
(1973). Recent studies of carotenoid biosynthesis in bacteria. Pure Appl. Chem., 35,
Mantoura, R. F. C. & Llewellyn, C. A. (1983). The rapid determination of algal
chlorophyll and carotenoid pigments and their breakdown products in natural
waters by reverse-phase high-performance liquid chromatography. Anal. Chim.
Acta, 151, 297-314.
Mantoura, R. F. C. & Llewellyn, C. A. (1984). Trace enrichment of marine algal
pigments for use with HPLC-diode array spectroscopy. J. High Resol. Chromatogr.
& Chromatogr. Commun., 7,632-5.
Margalith, P. & Meydav, S. (1968). Some observations on the carotenogenesis in the
yeast Rhodotorula mucilaginosa. Phytochem., 7, 765-8.
Marusich, W. L. & Bauernfeind, J. C. (1981). Oxycarotenoids in poultry feed. In
Carotenoids as Food Colorants and Vitamin A Precursors, ed. J. C. Bauernfeind.
Academic Press, New York, pp. 319-462.
Mathews, M. M. & Krinsky, N. I. (1965). The relationship between carotenoid
pigments and resistance to radiation in non-photosynthetic bacteria. Photochem.
Photobiol., 4, 813-17.
Mathews-Roth, M. M. & Krinsky, N. I. (1970). Studies on the protective function of
the carotenoid pigments of Sarcina lutea. Photochem. Photobiol., 11,419-28.
Mathews-Roth, M. M., Wilson, T., Fujimori, E. & Krinsky, N. I. (1974). Carotenoid
chromophore length and protection against photosensitization. Photochem.
Photobiol., 19,217-22.
Mathis, P. & Schenck, C. C. (1982). The functions of carotenoids in photosynthesis. In
Carotenoid Chemistry and Biochemistry, ed. G. Britton & T. W. Goodwin.
Pergamon Press, Oxford, pp. 339-47.
Mayer, H. (1979). Synthesis of optically active carotenoids and related compounds.
Pure Appl. Chem., 51, 535-64.
Mayer, H. & Isler, O. (1971). Total syntheses. In Carotenoids, ed. O. Isler.
Birkhauser, Basel, pp. 325-575.
Merritt, K. & Jacobs, N. J. (1978). Characterization and incidence of pigment
production by human clinical group B-streptococci. J. Clin. Microbiol., 8, 105-7.
Miller, J. c., George, S. A. & Willis, B. G. (1982). Multichannel detection in
high-performance liquid chromatography. Science, 218, 241-6.
Moss, G. P. & Weedon, B. C. L. (1976). Chemistry of the carotenoids. In Chemistry
and Biochemistry of Plant Pigments, Vol. 1, 2nd edn, ed. T. W. Goodwin.
Academic Press, London, pp. 149-224.
Muller, R. K., Bernhard, K. & Vecchi, M. (1982). Recent advances in the synthesis
and analysis of 3,4-oxygenated xanthophylls. In Carotenoid Chemistry and
Biochemistry, ed. G. Britton & T. W. Goodwin, Pergamon Press, Oxford, pp.
Nelis, H. J. C. F. & De Leenheer, A. P. (1983). Isocratic nonaqueous reversed-phase
liquid chromatography of carotenoids. Anal. Chem., 55,270-75.
Nelis, H. J. & De Leenheer, A. P. (1986). Carotenoid profiling and quantitation in
biological materials using non-aqueous reversed phase chromatography and photodiode array detection. Abstracts of the 10th International Symposium on Microchemical Techniques, Antwerp, 25-29 August, 1986, p. 133.
Nelis, H. J. C. F. & De Leenheer, A. P. (1987). Exhaustive extraction, profiling and
quantitation of bacterial carotenoids. Abstracts of the 8th International Symposium
on Carotenoids, Boston, 27-31 July 1987, p. 9.
Nesterenko, O. A., Nogina, T. M., Kasumova, S. A., Kvasnikov, E. I. & Batrakov, S.
G. (1982). Rhodococcus luteus nom. nov. and Rhodococcus maris nom. nov. Int. J.
Syst. Bacteriol., 32, 1-14.


H. J. Nelis & A. P. De Leenheer

Ninet, L. & Renaut, J. (1979). Carotenoids. In Microbial Technology, 2nd edn, Vol. I,
ed. H. J. Peppler & D. Perlman. Academic Press, New York, pp. 529-44.
Ninet, L., Renaut, J. & Tissier, R. (1969). Activation of the biosynthesis of carotenoids
by Blakeslea trispora. Biotechnol. Bioeng., 11, 1195-210.
Okagbue, R. N. & Lewis, M. J. (1981). Mixed culture of Bacillus circulans WL-12 and
Phaffia rhodozyma on different carbon sources: yeast-wall lytic enzyme production
and extractability of astaxanthin. Biotechnol. Lett., 5,731-6.
Okagbue, R. N. & Lewis, M. J. (1984a). Vse of alfalfa residual juice as a substrate for
propagation of the red yeast Phaffia rhodozyma. Appl. Microbiol. Biotechnol., 20,
Okagbue, R. N. & Lewis, M. J. (1984b). Inhibition of the red yeast Phaffia rhodozyma
by saponin. Appl. Microbiol. Biotechnol., 20,278-80.
Okagbue, R. N. & Lewis, M. J. (1984c). Autolysis of the red yeast Phaffia rhodozyma:
a potential tool to facilitate extraction of astaxanthin. Biotechnol. Lett., 6,247-50.
Okagbue, R. N. & Lewis, M. J. (1985). Influence of mixed culture conditions on
yeast-wall hydrolytic activity of Bacillus circulans WL-12 and on extractability of
astaxanthin from the yeast Phajfia rhodozyma. l. Appl. Bacteriol., 59,243-55.
Pachlavuni, I. K. (1987). A carotenoid-containing complex from Rhodococcus maris:
some properties and structure. Abstracts of the 8th International Symposium on
Carotenoids, Boston, 27-31 July, 1987, p. 12.
Peterson, W. J., Evans, W. R., Leece, E., Bell, T. A. & Etchells, J. L. (1958).
Quantitative determination of the carotenoids in yeasts of the genus Rhodotorula. l.
Bacteriol., 75, 586-9l.
Ratledge, C. (1970). Microbial conversions of n-alkanes to fatty acids: a new attempt to
obtain economical microbial fats and fatty acids. Chem. Ind., 843-54.
Ratledge, C. (1982). Microbial oils and fats: an assessment of their commercial
potential. In Progress in Industrial Microbiology, Vol. 16, ed. M. J. Bull. Elsevier,
Amsterdam, pp. 119-206.
Rau, W. (1983). Photoregulation of carotenoid biosynthesis. In Biosynthesis of
Isoprenoid Compounds, Vol. 2, ed. J. W. Porter & S. L. Spurgeon. Wiley, New
York, pp. 123-57.
Riiedi, P. (1985). HPLC-a powerful tool in carotenoid research. Pure Appl. Chern.,
Sakurai, H., Kato, K., Sakai, T., Masuda, Y. & Kuriyama, T. (1971). Studies on
microbial utilization of petroleum. I. Separation and characterization of carotenoids
produced by a species of Brevibacterium in hydrocarbon media. Bull. Chem. Soc.
lapan, 44,481-84.
Saperstein, S., Starr, M. P. & Filfus, J. A. (1954). Alterations in carotenoid synthesis
accompanying mutation in Corynebacterium michiganense. l. Gen. Microbiol., 10,
Sato, K., Veda, S. & Shimizu, S. (1977). Form of vitamin B12 and its role in a
methanol-utilizing bacterium, Protaminobacter ruber. Appl. Environ. Microbiol.,
Schlegel, H. G. (1959). Conversion of carotenoids to oxycarotenoids by Mycobacterium
phlei. l. Bacteriol., 77, 310-16.
Schmidt, K. (1978). Biosynthesis of carotenoids. In The Photosynthetic Bacteria, ed. R.
K. Clayton & W. R. Sistrom. Plenum Press, New York, pp. 729-50.
Schwartzel, E. H. & Cooney, J. J. (1970). Isolation and identification of echinenone
from Micrococcus roseus. l. Bacteriol., 104,272-74.
Schwartzel, E. H. & Cooney, J. J. (1972). Isolation of 4'-hydroxyechinenone from
Micrococcus roseus. l. Bacteriol., 112, 1422-24.
Schwartzel, E. H. & Cooney, J. J. (1974a). Action of light on Micrococcus roseus. Can.
l. Microbiol., 20, W15-2l.
Schwartzel, E. H. & Coon~y, J. J. (1974b). Isolation and characterization of
pigmentation mutants of Micrococcus roseus. Can. l. Microbiol., 20, 1007-13.

Microbial Production of Carotenoids


Schwarz, Y. & Margalith, P. (1965). Production of egg yolk coloring material by a

fermentation process. Appl. Microbial., 13, 876-8l.
Seybold, A. & Goodwin, T. W. (1959). Occurrence of astaxanthin in the flower petals
of Adonis annua L. Nature, Lond., 184, 1714-15.
Shapiro, A., DiLello, D., Loudis, M. c., Keller, D. E. & Hutner, S. H. (1977).
Minimal requirements in defined media for improved growth of some radio-resistant
pink tetracocci. Appl. Environ. Microbiol., 33, 1129-33.
Shimizu, S., Sato, K., Hiraoka, M., Yamashita, F. & Kobayashi, T. (1982).
Carotenoids formation by a facultative methylotroph, Protaminobacter ruber. J.
Ferment. Technol., 60, 163-6.
Shipman, R. H., Fan, L. T. & Kao, I. C. (1977). Single-cell protein production by
photosynthetic bacteria. Adv. Appl. Microbial., 21, 161-83.
Simpson, K. L., Chichester, C. O. & Phaff, H. J. (1971). Carotenoid pigments of
yeasts. In The Yeasts, Vol. 2, ed. A. H. Rose & J. S. Harrison. Academic Press,
New York, pp. 493-515.
Simpson, K. L., Katayama, T. & Chichester, C. O. (1981). Carotenoids in fish feeds.
In Carotenoids as Food Colorants and Vitamin A Precursors. Academic Press, New
York, pp. 463-538.
Spurgeon, S. L. & Porter, J. W. (1983). Biosynthesis of carotenoids. In Biosynthesis of
Isoprenoid Compounds, Vol. 2, ed. J. W. Porter & S. L. Spurgeon. Wiley, New
York, pp. 1-122.
Tanaka, A., Nagasaki, T., Inagawa, M. & Fukui, S. (1968a). Studies on the formation
of vitamins and their functions in hydrocarbon fermentation. (V) Production of
carotenoids by Mycobacterium smegmatis in hydrocarbon media (Part I). Studies on
the cultural conditions. J. Ferm. Technol., 46, 468-76.
Tanaka, A., Nagasaki, T. & Fukui, S. (1968b). Studies on the formation of vitamins
and their functions in hydrocarbon fermentation. (VI) Production of carotenoids by
Mycobacterium smegmatis in hydrocarbon media. (Part II) Isolation and characterization of several carotenoids produced by Mycobacterium smegmatis IFO-3080.
J. Ferm. Technol., 46,477-87.
Tanaka, A., Kato, K. & Fukui, S. (1971). Studies on the formation of vitamins and
their function in hydrocarbon fermentation. (IX) Production of carotenoids from
hydrocarbons by Brevibacterium. 1. Ferm. Technol., 49, 778-91.
Taylor, R. F. (1980). Chromatography of carotenoids and retinoids. Adv. Chromatogr.,
22, 157-213.
Theriault, R. J. (1965). Heterotrophic growth and production of xanthophylls by
Chlorella pyrenoidosa. Appl. Microbiol., 13, 402-16.
Toraya, T., Yongsmith, B., Tanaka, A. & Fukui, S. (1975). Vitamin B\2 production by
a methanol utilizing bacterium. Appl. Microbial., 30, 477-9.
Ungers, G. E. & Cooney, J. J. (1968). Isolation and characterization of carotenoid
pigments of Micrococcus roseus. 1. Bacteriol., 96,234-4l.
Urakami, T., & Komagata, K. (1986). Occurrence of isoprenoid compounds in
Gram-negative methanol-, methane- and methylamine-utilizing bacteria. J. Gen.
Appl. Microbiol., 32, 317-4l.
Valadon, L. R. G. (1976). Carotenoids as additional taxonomic characters in fungi: a
review. Trans Br. Mycol. Soc., 67, 1-15.
Van Niel, C. B. & Smith, J. H. C. (1935). Studies on the pigments of the purple
bacteria. Arch. Mikrobiol., 6,219-29.
Various patents (1973). Recovery techniques. In Proteins From Hydrocarbons, Food
Technology Review no. 4, ed. S. Gutcho. Noyes Data Corporation, Park Ridge,
NJ, pp. 144-99.
Villoutreix, J. (1960). Influence de divers agents chimiques sur la carotenogenese de
Rhodotorula mucilaginosa. Biochim. Biophys. Acta, 40, 434-4l.
Voznyakovskaya, Y. M. & Daraseliya, G. Y. (1972). Mutants of Mycobacterium phlei
with increased synthesis of carotenoids. Mikrobiologiya, 41, 886-90.

H. 1. Nelis & A. P. De Leenheer


Wendall, M., Farrow, F. & Tabenkin, B. (1964). Preparation of a lutein product.

French patent 1,367,027. CA 62 (1965) 9742.
Wood, J. F., Carter, P. M. & Savory, R. (1983). Investigation into the effect of
processing on the retention of the carotenoid fractions of Leucaena Leucocephala
during storage, and the effects of processing on mimosine concentration. Anim.
Feed Sci. Technol., 9, 307-17.
Yamada, K., Nakahara, T. & Fukui, S. (1971). Petroleum microbiology and vitamin
production. In Biochemical and Industrial Aspects of Fermentation, ed. K. Sakaguchi, T. Uemura & S. Kinoshita. Kodansha, Tokyo, pp. 61-90.


Several biotechnology companies are presently (May 1989) conducting extensive research to improve the production of astaxanthin from Phaffia rhodozyma. One of them (Gist Brocades) has now marketed the organism as a
pigmenter for salmonid diets. Recent papers on Phaffia rhodozyma are by N.
F. Haard (1988, Biotechnol, Lett., 10, 609-14) and G.-H. An (1989, Appl.
Environ. Microbiol., 55, 116-24). Maximum astaxanthin levels obtained by
these investigators range from 1.2 mg/g (Phaffia rhodozyma parent strain) to
2.5 mg/g (mutant strain). Corresponding yields on a volume basis were
15 mg/liter and not specified, respectively.

Chapter 5


Department of Food Engineering & Biotechnology, Technion-Israel Institute of
Technology, Haifa, 32000, Israel


Among the vitamins, vitamin D is outstanding because of its chemical nature,

it being the only vitamin with a steroid structure. Vitamin D is necessary for
the proper metabolism of minerals and bone formation, in mammals and birds.
Biotechnologically speaking, however, one usually refers to the provitamin,
ergosterol (1) the main sterol in yeasts and mycelial fungi. The provitamin may
be converted to the proper vitamin (D 2 ) which undergoes further metabolism
before exerting its hormone like activity in controlling the processes involved in
mineral metabolism.




Ergosterol is not the only provitamin suitable for antirachitic treatment.

7-dehydrocholesterol (2) as produced in the mammalian body, may also be
transformed into vitamin D (D3) with somewhat different chemical and
physiological characteristics. Since the latter vitamin has certain advantages in
use, ergosterol as a source for vitamin D no longer occupies the same position
in the manufacturing of this vitamin.
Under natural circumstances, the human body produces its vitamin D from
7-dehydrocholesterol upon exposure to sunlight. In many geographical areas

P. Margalith


where clouded skies prevail throughout most of the year not enough vitamin D
is biosynthesized. The migration into cities and the increase in indoor
professions led to a decline in the synthesis of the natural vitamin and created
a demand for a dietary supplement that would protect the human body from
In rickets, the organic matrix of new bone is not mineralized. This is due to
the body's inability to calcify the collagene matrix of the growing bone and
results in large areas of uncalcified bone (osteoid). The resultant lack in
rigidity of bones leads to the ends becoming entwisted and bent. Also, tooth
enamel formation may be affected. Children are specially prone to this disease.
Even with an optimal intake of calcium and phosphorus in vitamin D deficient
children, the concentration in the blood serum of these minerals is very low.
Vitamin D increases the calcium and phosphate absorption and restores the
mineral balance which results in the appropriate calcium phosphate deposition
in the bone.
The use of liver oil as a prophylactic against rickets, as well as the
importance of light as an antirachitic factor, have been recognized since before






Cholecalciferol (D3)

Ergocalciferol (D 2 )

Fig. 1.

The Biotechnology of Ergosterol


the end of the 19th century. Steenbock & Black (1924) showed that irradiation
of foods led to the generation of the vitamin. We now know that irradiation
may lead to two types of vitamin 0: (1) By irradiation of ergosterol,
ergocalciferol (or O 2 ) is formed while (2) irradiation of 7-dehydrocholesterol
leads to the production of cholecalciferol (or D 3 ), see Fig. 1. This chapter will
deal chiefly with the production of ergosterol, the provitamin O 2


Ergosterol is a fat-soluble vitamin with a typical steroid structure. It consists of

4 rings (the cyclopentanoperhydrophenanthrene system, comprising 3 fused
6-membered rings and 1 5-membered ring) and a side chain, in all 28 carbon
atoms. Ergosterol differs from cholesterol, the best known sterol of animal
origin, in additional carbon (methyl group) at C-24 in the J3-position (referring
to the position on the axis of the side chain), as well as additional double
bonds between carbons 7: 8 and 22: 23. Like most steroids, ergosterol also has
2 angular methyl groups at carbon 10 and 13 in the J3-position (perpendicular
to the plane of the ring). Both sterols have a hydroxyl group at carbon 3 in the
J3-position. Many of the chemical and physiological characteristics of these
sterols can be attributed to this group.
Contrary to the saponifiable lipids, hydrolysis of steroids with alkali will not
make them water-soluble. They are therefore classified as unsaponifiable
lipids. Like most 3J3-hydroxysteroids, ergosterol can also be precipitated with
digitonin, a steroidal saponin. Although digitonin may be employed for the
precipitation of such sterols, the quantitative determination by gravitational
methods is usually not sensitive enough. Colorimetric methods are now widely
employed, most of which are based on the Liebermann-Buchard reaction. A
sterol in chloroform solution is treated with acetanhydrid and sulfuric acid at a
suitable temperature. With time a color develops with characteristic absorption. Ergosterol, on the other hand, may be determined spectrophotometrically by absorption in UV. The characteristic spectrum of the 5,7 steroid in
absolute alcohol has peaks at 271, 282 and 293 nm. For the quantitative
determination of ergosterol it is advisable to employ a suitable separation
procedure (TLC or GLC) to avoid the interference of related compounds. The
molar extinction of ergosterol at 282 nm is 11 900.


Although yeasts are known as producers of ergosterol, the sterol was first
isolated from rye ergot (sclerotia of C/aviceps purpurea). It is from this plant
disease that the name ergosterol was derived (Tanret, 1889). Many organisms
were later screened and examined for their sterol content. While procaryotes
were found to be practically devoid of true sterols, eucaryotic micro-organisms


P. Margalith

could be divided into a number of categories with regard to the role sterols
play in their cell biology:
(1) Organisms in which sterols are an important constituent of the cell
(a) Organisms that synthesize their own sterols.
(b) Organisms unable to synthesize their sterols, their requirements
being met by environmental sources.
(2) Organisms which do not require sterols for vegetative growth, but
need such compounds for the differentiation of certain sexual or
asexual reproductive organs.
(3) Organisms which have no known requirements for sterols but which
may incorporate such compounds if suitably exposed to them (Margalith, 1986).
Screening for organisms that are good producers of ergosterol was obviously
directed to the first category of organisms that synthesize their own sterols.
Although sterols are widespread among photosynthesizing plants including the
fast-growing microalgae, phytosterols of green plants usually do not comprise
ergosterol. The search was therefore limited to eucaryotic heterotrophs,
amongst these the Mucorales (Zygomycotina), the Ascomycotina (including
the yeasts and yeastlike fungi), the Basidiomycotina (in the mushrooms) and
many of the Fungi Imperfecti.
In most cases ergosterol is not the only sterol to be found, it is usually
accompanied by other related sterols such as zymosterol (3) fungisterol (4) or
ergosta-5,7,22,24(28)-tetraenol (5). We shall return to these sterols when
discussing the biosynthesis of ergosterol. The quantitative aspects of ergosterol
formation and distribution have been also studied. There seems to be no great
variation in the sterol content of fungi. According to Appleton et al. (1955),
molds have an average of c. 06% sterols on a dry weight basis. Exceptional
higher values were recorded for Penicillium westlinghi (22%).
Not surprisingly, yeasts have been the object of many workers interested in
the biotechnology of ergosterol and its biosynthesis. The ease with which
yeasts can be grown and the familiarity of many microbiologists with these
organisms in various food industries, in addition to their being an excellent
source for most vitamin Bs, have made yeasts the organism of choice for the
production of provitamin D.
The first to study the ergosterol content of yeasts were Bills et al. (1930).
They examined 29 strains from 13 species of 5 genera: Endomyces, Nadsonia,
Mycoderma, Saccharomyces, and Zygosaccharomyces. Values of 02-03%
(dry weight) have been recorded. Saccharomyces carlsbergensis (strain ATCC
2345) had the highest ergosterol level (24%). Usdin & Burell (1952) found
27% in Rhodoturula gracilis. A systematic study for ergosterol in yeasts was
carried out by Dulaney et al. (1954) who examined 146 cultures of the
following genera:

The Biotechnology of Ergosterol





Saccharomyces, Schwanniomyces, Torulopsis, Cryptococcus, Endomyces,

Kloeckera, Pichia, Saccharomycodes, Sporobolomyces, Debaryomyces,
Endomycopsis, Nadsonia, Rhodotorula, Schizosaccharomyces, Taphrina. With
the exception of Saccharomyces, levels of ergosterol never exceeded 04%.
The best organism was Saccharomyces cerevisiae for which 39% was reported.
Additional studies were made by Appleton et al. (1955) who reached similar
conclusions. They confirmed the advantage of the genus Saccharomyces for the
formation of ergosterol. There seems to be considerable variation between
species or even strains. However, generally speaking, brewer's yeast had the
highest level of sterols, followed by baker's yeast. A method for the
production of high sterol yeasts (10% dry weight) was patented by Dulaney
Not all reports deal with the nature of the sterols found in yeast. In most
cases where this was done the ratio of ergosterol to other sterols was 8: 2.


Sterol production by yeasts is greatly affected by environmental conditions.

High concentrations of assimilable carbohydrates promote the production of
ergosterol, while nitrogen-rich media usually lead to decreased levels of sterol
biosynthesis (Dulaney et al., 1954; Gal'tsova et al., 1959).
Non-sugar carbon sources for the production of sterols have also been
examined. Acetate seems to have a promoting effect when added to sucrose
(Starr & Parks, 1962). Ethanol can also serve as a carbon source. An old
method for increased sterol production consisted of exposing yeast cells to
vapors of ethanol. It was claimed that by this method a sterol level of c. 10%

P. Margalith


could be obtained. The most important environmental factor in yeast growth

and sterol productivity is, however, the availability of oxygen.
Although fermenting yeasts have a mechanism for the attainment of
workable energy (ATP) under anaerobic conditions, it has been known for
some time that Saccharomyces yeasts would not proliferate under strictly
anaerobic conditions (Nordheim, 1966). An explanation for this incongruity
was first suggested by Andreasen & Stier (1953) in the classical papers on the
anaerobic nutrition of S. cerevisiae. It was established that sterols (as well as
unsaturated fatty acids) cannot be synthesized under anaerobic conditions and
only if added to the media may they replace in part oxygen for growth under
these conditions. Since growth cannot take place under anaerobic conditions,
these are totally unsuitable for sterol production. Growth under semianaerobic conditions, i.e. without aeration, yielded only 1/10 of sterols in
comparison to fully aerated yeasts (Klein, 1955). There seems to be a linear
relationship between oxygen availability and sterol biosynthesis. However,
there appear to be no detailed studies describing such a relationship and
whether there are any detrimental effects on sterol production due to


Earlier workers (Maguigan & Walker, 1940) suggested a relationship between

sterols and the total lipid composition, the former comprising 18-19% of the
total lipids. This was later found to be inaccurate since yeast strains with high
sterol content and low lipid composition were also found (Maugenet & Dupuy,
Since sterols belong to the lipid fraction of an organism, it is not surprising
that both sterols and fatty acids employ acetate as precursors. Indeed, label
studies showed that acetate, both the methyl and carboxyl carbons, serves as a
source for all the carbons of ergosterol, with the exception of C-28.
From acetate to lanosterol: the pathway by which the 2 carbon acetate
molecule is converted to the 5 carbon isoprene molecule, involves the
condensation of 3 acetate residues with the elimination of 1 carbon atom as
CO2 From 3-hydroxy-3-methylglutaric acid (HMG) , isopentenyl pyrophosphate (IPP) and dimethylallylpyrophosphate (DPP) are formed. A stepwise
addition of C s units leads eventually to the formation of farnesylpyrophosphate
(FPP)-C 1s ; 2 FPPs forming by tail-to-tail condensation squalene, the open
chain C 30 of sterols. The cyclization of squalene is the second major step in
sterol biosynthesis. This is a complex reaction involving an oxygenated
intermediate:2,3-oxidosqualene which is formed in presence of oxygen only.
Obviously, anaerobically grown yeasts cannot form this intermediate (Fig. 2).
Cyclization of the epoxide is initiated by a specific enzyme (squalene cyclase)
leading to a series of hydrogen and methyl migrations giving rise to a cyclic
triterpenoid having the sterol nucleus (Schechter et aZ., 1970) known as


The Biotechnology of Ergosterol






ERGOST A-5,7 ,22,24(28)TETRAENOL
Fig. 2.

lanosterol (6). Interestingly, this compound is also formed during sterol

synthesis in animal tissues and is not found in green plants which synthesize the
related cycloartenol (7) as the first cyclization product.




P. Margalith

Demethylations and alkylations: The tetracyclic lanosterol differs from a

true sterol in its 3 'extra' methyl groups at positions 4,4' and 14. In addition,
lanosterol shows 2 double bonds at C-24 (25) and C-8 (9). In order to become
ergosterol, the ultimate sterol of most fungi and yeasts, de methylation
reactions, reduction and rearrangements of double bonds in addition to the
alkylation of C-24 have to take place. Lanosterol as such cannot take the place
of ergosterol in anaerobic growth of yeast (Nes et al., 1978). This can be
explained by the bulky nature of lanosterol. The demethylation reactions
similar to squalene oxygenation require the presence of oxygen (for the
elimination in the form of CO2 ) and cannot proceed under anaerobic
conditions. The de methylation reactions lead to the formation of the first true
sterol compound:zymosterol (3).
Zymosterol differs from ergosterol in a number of features. The double
bond is still at 8: 9 and 24: 25, and C-24 must be alkylated with the C-28 methyl
group. We have already pointed out that the C-28 carbon is not derived from
acetate. The methyl group is derived from S-adenosylmethionine by a
transmethylation reaction (24-methyltransferase) yielding fecosterol (8) to be
followed by double bond migration from 8: 9 to 7: 8 giving episterol (9); the
introduction of an additional double bond (ergosta-7 ,22,24(28)-trienol); the
formation of the tetraenol (with the 5: 6 double bond) and the final reduction
of C-24(28) to ergosterol (1).



The pattern described above is not the only possible pathway for the
production of yeast ergosterol. Sterol synthesis, in general, does not follow a
rigid pattern but rather a more flexible framework where many alternative
pathways may take place. (Fryberg et al., 1973). Because of the complexity of
sterol biosynthesis and the diversity in the timing of certain key reactions (in
some strains (A~ B~ C, while in others A~ C~ B), the regulation of sterol
synthesis has not been sufficiently studied. Of the many enzymic reactions
potentially capable of a regulatory function, hydroxy-methyl-glutaryl-CoA
reductase is probably the rate-controlling enzyme of sterol biosynthesis
(Qureshi et al., 1980).
Very few publications deal with the ergosterol fermentation. In the 1950s a
number of papers were published dealing with the production of ergosterol by

The Biotechnology of Ergosterol


various yeasts and fungi. Since yeasts with strong fermentative metabolism
were found to be the best producers of ergosterol, most of the studies
concentrated on the fermentation of S. cerevisiae strains. Using a molassescornsteep medium, cells with an ergosterol content of 7-10% could be
obtained. Cell yields of 30-40 g/liter medium and a sterol content of
3-4 g/liter broth were found in shake flask experiments after 4 days of growth
at 28C (Dulaney et al., 1954). Good yields were also obtained with S.
fermentati (EI Refai & EI Kady, 1969). Little has been published about
fermentation in larger volumes. Nor has the patent literature revealed
important new information on the industrial fermentation of the provitamin.
Since the early patent of Dulaney (1957) a number of patents were granted to
several companies in various countries. Most of these deal with extraction
procedures for the isolation and crystallization of ergosterol from the biomass
(e.g. Nelson, 1959). Few patents describe a modification of the fermentation
procedure by the incorporation of precursors to the medium. The addition of
isopentenol (092%) and Na-polyphosphate (03%) to the growth of S. uvarum
UFO 1264 at 26C for 96 h, yielded 21 % ergosterol (dry weight) (Japanese
Patent 8158496, 1981). Organisms described in the patent literature comprise:
S. cerevisiae, S. carlsbergensis, Candida tropicalis, Fusarium spp and
Trichoderma spp. Yields (ergosterol on basis of dry weight or volume of
medium) described in the patent literature are generally far lower than those
reported in scientific publications. A number of patents have been granted for
the growth of yeasts in medium containing hydrocarbons as sole carbon source,
e.g. Candida petrophilum ATCC 20226 (Horiguchi et al., Japanese Patent
7392586, 1973). Also here ergosterol yields were rather low (see Table 1 for
list of patents).
Isolation of ergosterol from biological sources usually involves the extraction
of the fatty component, saponification and re-extraction with an ether. Direct
extraction of yeast is not sufficient, usually a preliminary digestion with hot
alkali is included prior to the extraction.
Provitamin D3

The equivalent of ergosterol in animal tissues is 7-dehydrocholesterol (2) which

can be produced commercially by the bromination of an esterified cholesterol
followed by dehalogenation to give the 7 -dehydrocholesteryl ester.
Both ergosterol and 7-dehydrocholesterol can be transformed into previtamin
D by UV irradiation (280-300 nm). Sufficiently low temperatures are maintained to prevent the formation of side products. This results in the opening of
the B ring. The 5,7 conjugated diene is activated yielding the previtamin (10),
which undergoes a thermal equilibration to vitamin D. Fission of the 9: 10

Cultivation of yeast
producing ergosterol
Production of




(UFO 1264)

S. uvarum








C. troJlicalis

C. JletroJlhilum
(ATCC 20226)
(FERM + P3496)
S. cerevisiae




Ergosterol from yeast

Extraction of sterols
Production of ergosterol
by yeasts from hydrocarbons
Ergosterol production from
Fermentative production of

S. cerevisiae
NRRL strains





Patent No.



Ergosterol containing yeast


Table 1
Some Patents for the Production of Ergosterol

Medium supplemented
with L-methionine

N-alkane as sole
carbon source
n-Paraffins and
acetone in medium






The Biotechnology of Ergosterol

bond leads to the formation of the secosteroid. Ergosterol thus becomes

Vitamin D2 (ergocalciferol) and 7-dehydrocholesterol becomes vitamin D3
(cholecalciferol). The biological activity of 0025 f.lg of pure cholecalciferol is
considered the international unit (IU) of vitamin D.



Vitamin D formed in the skin or absorbed by ingestion through the gut wall, is
transported by the bloodstream into the liver where it is hydroxylated at C-25
to yield 25-hydroxycalciferol which is the main circulating form of Vitamin D.
It is transported on a specific a-globulin to the kidney where further
hydroxylation takes place at C-1 or C-24, apparently in response to the calcium
levels in the blood. Low phosphate levels also stimulate 1,25dihydroxycholecalciferol (11) production which, in turn, enhances calcium
absorption. The C-1 hydroxylation is the critical step which enables the vitamin
to assume its hormonal activity in regulating the calcium transport as well as
bone calcium mobilization. This important reaction is carried out in kidney
mitochondria by a mixed function mono-oxygenase involving cytochrom P-450.




Regulation of vitamin D metabolism appears to be controlled by the levels
of calcium, phosphate and the parathyroid hormone in blood serum. Below
95 mg/ml of Ca2 +, 25-dihydroxy vitamin D is formed, above that level
24,25-dihydroxycalciferol is produced with the discontinuation of the lahydroxylase activity. The enzyme is also stimulated by the homeostatic
parathyroid hormone which causes phosphate diuresis in the kidney and
enhances 1,25-dihydroxycholecalciferol production. The parathyroid hormone
and 1,25-dihydroxycholecalciferol act at the bone site cooperatively in the
stimulation of calcium mobilization from bone. Other metabolites of vitamin D

P. Margalith


are also formed. However, they are less potent and their metabolic significance
is not yet fully understood.


While vitamin D is used for the fortification of many foodstuffs (margarine,

milk), large amounts are added to the animal feeds, mostly for swine, chicken
and dairy calves. Vitamin D2 and D3 have the same biological activity in rats.
However, in chicken, vitamin D2 is only 1/10 or less active in comparison to
vitamin D 3. In man cholecalciferol is apparently also more active than
ergocalciferol. The difference is probably due to the C-24 methyl group and
not the C-22 double bond since 22-dihydroergosterol is as effective as
provitamin D 2. Considering these facts led to the obvious preference of
cholecalciferol produced synthetically. Vitamin D3 is sold at about 1 $/1(t IU;
or 10 $/kg of a product of 400 000 IU/g(Hirsch, 1984). Considering the
worldwide demand for vitamin D at about 1750 X 1012 IU (1985), neither the
quantities required nor the price obtained on the market have given great
impetus to the improvement in the biotechnological production of provitamin
Nevertheless, some efforts have been made to produce biotechnologically a
precursor that could be employed for the production of vitamin D 3. Avruch et
al. (1976) suggested to inhibit the 24-methyltransferase, the enzyme
responsible for the C-24 methylation, characteristic to ergosterol. By growing
S. cerevisiae in the presence of 25, aza-24,25 dihydrozymosterol at 13 J.lM
concentration, it was possible to inhibit the transmethylation reaction without
affecting the growth of the yeast. Under these conditions little ergosterol was
formed but the accumulation of zymosterol (3) and cholesta 5,7,22,24(25)tetraenol (12) took place in considerable quantities. The cholestatetraenol was
suggested as a substitute for the more expensive 7-dehydrocholesterol provitamin. The conversion of this compound to the corresponding cholecalciferol by
irradiation would be expected. However, no additional information has
become available about this modified process.



The Biotechnology of Ergosterol


Andreasen, A. A. & Stier, T. J. B. (1953). Anaerobic nutrition of Saccharomyces
cerevisiae. I. Ergosterol requirement for growth in deficient medium. J. Cell.
Physiol., 41, 23-36.
Appleton, G. S., Kieber, R. J. & Payne, W. J. (1955). The sterol content of fungi II.
Screening of representative yeasts and molds for sterol content. Appl. Microbiol., 3,
Avruch, L., Fischer, S., Pierce, H. D. & Oehlschlager, A. C. (1976). The induced
biosynthesis of 7-dehydrocholesterols in yeast: potential sources of new provitamin
D3 analogs. Can. J. Biochem., 54,657-65.
Bills, C. E., Massengale, O. N. & Prickett, P. S. (1930). Factors determining the
ergosterol content of yeast. I. Species. J. BioI. Chem., 87, 259-64.
Dulaney, E. L. (1957). Preparation of ergosterol containing yeast. US Patent 2817624.
Dulaney, E. L., Stapley, E. O. & Simpf, K. (1954). Studies on ergosterol production
by yeasts. Appl. Microbiol., 2,371-9.
EI-Refai, A. E. M. & EI Kady, I. A. (1969). Utilization of some industrial by-products
for the microbiological production of sterols from Saccharomyces fermentati. J. Bot.
UAR, 12, 55-66.
Fryberg, M., Oehlshiager, A. C. & Uorau, A. M. (1973). Biosynthesis of ergosterol in
yeast. Evidence for multiple pathways. J. Am. Chem. Soc., 95,5747-57.
Gal'tsova, R. D., Novichkova, A. T. & Vakina, I. P. (1959). Effect of glucose on
ergosterol synthesis by yeasts. Mikrobiologiya, 28, 502-6.
Hirsch, A. (1984). Vitamin D. In Kirk Othmer's Encyclopedia of Chemical Technology, 3rd edn, Vol. 20. Wiley-Interscience, New York, pp. 186-213.
Klein, H. P. (1955). Synthesis of lipids in resting cells of Saccharomyces cerevisiae. J.
Bacteriol., 69, 620-7.
Klein, H. P., Eaton, N. R. & Murphy, S. C. (1954). Net synthesis of sterols in resting
cells of Saccharomyces cerevisiae. Biochim. biophys. Acta, 13, 591.
Maguigan, W. H. & Walker, E. (1940). Sterol metabolism of microorganisms. I. Yeast.
Biochem. J., 34,804-13.
Margalith, P. (1986). Steroid Microbiology. C. C. Thomas, Illinois.
Maugenet, J. & Dupuy, P. (1964). Synthese des sterols par la levure. Ann. Technol.
Agric., 13,329-54.
Nelson, H. A. (1959). Recovery of ergosterol. US Patent 2 874 171.
Nes, W. R., Sekula, B. C., Nes, W. D. & Adler, J. H. (1978). The functional
importance of structural features of ergosterol in yeast. J. Bioi. Chem., 253,
Nordheim, W. (1966). Metabolic behaviour of fermenting yeasts under cytostatic
conditions. Brauwiss., 19, 67-9.
Qureshi, A. A., Burger, W. C., Prentice, N., Bird, H. R. & Sunde, M. L. (1980).
Suppression of cholesterol and stimulation of fatty acid biosynthesis in chicken livers
by dietary cereals supplemented with culture filtrate of Trichoderma viride. J. Nutr.,
110, 1014-22.
Schechter, I., Sweat, F. W. & Bloch, K. (1970). Comparative properties of 2,3
oxidosqualene-Ianosterol cyclase from yeast and liver. Biochim. biophys. Acta, 220,
Starr, P. R. & Parks, L. W. (1962). Some factors affecting sterol formation in S.
cerevisiae. J. Bacteriol., 83, 1042-6.
Steenbock, H. & Black, A. (1924). Fat soluble vitamins XVII. The induction of growth
promoting and calcifying properties in a ration by exposure to ultra-violet light. J.
BioI. Chem., 61,405-22.
Taoret, C. (1889). Sur un nouveau principe immediat de L'ergot de seigle,
l'ergosterine. C. R. Seances Acad. Sci., 108,98.
Usdin, V. R. & Burrell, R. C. (1952). Preliminary study of the lipids of Rhodotorula
gracilis. Archs. Biochem. Biophys., 36, 172-7.

Chapter 6


Research Center for Cell and Tissue Culture, Faculty of Agriculture,
Kyoto University, Kyoto 606, Japan

The fat-soluble vitamin, vitamin E, has received great attention for its
usefulness as an antioxidant in clinical and nutritional applications. The
demand for the vitamin has increased year after year. The supply of
tocopherols has been limited to the chemically synthesized racemate of
a-tocopherol or a mixture of a-, f3(y)- and 6-tocopherols extracted from
vegetable oils.
Tocopherols are indispensable components of the lipid bilayer of biological
membranes; decrease in their content brings about structural and functional
damage to the membranes. It is generally accepted that the stabilizing effect of
tocopherols on the membranes is mainly related to three functions: (1)
reaction with lipid peroxide radicals, (2) quenching of reactive molecular
oxygen, and (3) ordering (i.e. restriction of molecular mobility) of the
membrane lipid bilayer by formation of tocopherol complexes with fatty acids.
The highest physiologically active form among the various tocopherols and
tocotrienols so far identified is D-a-tocopherol.
Since a preliminary survey on the occurrence of tocopherols in microorganisms (Green et al., 1959), its presence has been demonstrated especially
in chlorophyll-containing organisms, though not in all of them (Taketomi et
al., 1983). Early, a tocopherol-like antioxidant was found in yeasts (Forbes et
al., 1958) and a small amount of a-tocopherol was detected in cells of baker's
yeast (Diplock et al., 1961). a-Tocopherolquinol and a-tocopherolquinone
were found in the algae, Euglena gracilis, in bacteria and in yeasts (Hughes &
Tove, 1982). Ruggeri et al. (1985) reported enhancement of a-tocopherol
production in temperature-stressed cells of E. gracilis Z grown photoheterotrophically. However, a study on industrial production of these compounds has not been performed so far.
Recently, Tani & Tsumura (1989) have initiated a study aiming towards

Y. Tani


the development of new microbial processes to produce tocopherols, especially



Vitamin E was found as a dietary factor in animal nutrition, considered in the

1920s to be especially important for normal reproduction. a-- and f3Tocopherols were isolated from wheat germ oil and characterized in 1936.
Soon afterwards the third active factor, y-tocopherol, was found in cotton-seed
oil. In 1947, the fourth tocopherol, named ~-tocopherol, was isolated from
soybean oil. In the ensuing years, the investigation of several vegetable oils
disclosed the existence of four tocotrienols. Thus, four tocopherols and four
tocotrienols are now known to occur in nature (Fig. 1).
Tocopherols and tocotrienols are white to pale yellow oils. These are soluble
in organic solvents such as ethanol and ether, and miscible with other oils.
They are easily oxidized to a dark color. The esters are stable. The naturally
occurring form is the D-form, and D-a--tocopherol has the 2R, 4'R and 8'R
configuration. The absorption maxima are 290-298 nm.
Determination of tocopherols in cells has been described by Tani &
Tsumura (1989). Cells were homogenized in ethanol and extracted at
50-60C for 20 min. After an equal volume of n-hexane and one and a half
volumes of water were added to the ethanolic extract and mixed, the n-hexane
layer was obtained by centrifugation. The n-hexane extract was employed for
high performance liquid chromatography. Column systems used were an M&S
packed column (CIS 46 x 150 mm) with methanol-dichloromethane (9: 1, v Iv)
as the mobile phase or a YMC packed column (Sil, 6 x 150 mm) with
n-hexane-dichloromethane-isopropanol (70: 30 : O 2, by volume). The flow
rate was 1 ml/min. Tocopherols were detected at 292 nm.







CH 3

Fig. 1. Chemical structure of vitamin E group.

a- Tocopherol
y- Tocophero 1

0-Tocophero 1

y- Tocotri eno 1


Vitamin E Production


It has been accepted that tocopherols are synthesized through two routes, the

tocopherol route and tocotrienol route. Homogentisic acid, a precursor of

tocopherols, which was derived from the shikimate pathway (the biosynthetic
pathway of aromatic amino acids), incorporates phytyl pyrophosphate or
geranylgeranyl pyrophosphate. When phytyl pyrophosphate is incorporated,
tocopherols are formed and the following additional methylation using
S-adenosyl-L-methionine as the methyl donor occurs to form various types of
tocopherols (tocopherol route). When geranylgeranyl pyrophosphate is incorporated, tocotrienols are formed with a similar methylation. Subsequently,
tocotrienols can be transformed to tocopherols. NADPH is the reducing
coenzyme for the reduction of the side chain (tocotrienol route). In Spinach
chloroplasts, the enzymes involved in the a-tocopherol synthesis were found in
the chloroplast envelope fraction. The prenyltransferase and methyltransferase
activities were found to be tightly bound to the chloroplast envelope
membrane. Thus, it is strongly suggested that the tocopherol biosynthesis is
related to the occurrence of chloroplasts or plastids.


4.1 Screening of Tocopherol-Producing Micro-organisms
The distribution of tocopherol-producing ability was extensively studied with
162 strains from 26 genera of ascosporogenous and asporogenous yeasts, 74
strains from 15 genera representing molds of the Phycomycetes, Ascomycetes
and Fungi Imperfecti, 41 strains from 12 genera of Basidiomycetes, 4 strains of
algae belonging to genera of Euglena, Chlorella and Chlamydomonas, and 5
strains of isolated but unidentified algae. Though occurrences of tocopherol in
yeast (Diplock et al., 1961) and a-tocopherolquinone and its quinol in
eucaryotic and procaryotic micro-organisms (Hughes & Tove, 1982) were
reported, no tocopherol was detected in cells and culture filtrates of yeasts,
molds and Basidiomycetes under the cultured and detection conditions
employed. On the other hand, all algae tested were found to contain
tocopherols, mainly the a-type.
Euglena gracilis Z was the best producer of a-tocopherol among all algae
tested. It appeared on high performance liquid chromatograms that the
tocopherol detected corresponded to the retention time of a-tocopherol and
that it was the dominant component, with only a small amount of fJ(y)- and
6-tocopherols. Shigeoka et al. (1986) demonstrated during a study on the
intracellular localization of tocopherols in E. gracilis Z, that a-tocopherol
constituted more than 97% of total tocopherols.


Y. Tani

4.2 Isolation of a-Tocopherol from EuglelUl gracilis Z CeUs

Cells of E. gracilis Z weighing 21 g as dry matter were obtained from 80 ml of
culture broth. The cell paste was suspended in 100 ml of ethanol and kept at
50C for 30 min. The colorless cell debris were filtered and re-extracted once in
the same manner. The combined extract was partitioned between equal
volume of sodium chloride-saturated water and 150 ml of n-hexane. The
n-hexane phase was dried by anhydrous sodium sulfate and then concentrated
at 50C under reduced pressure. Resultant lipid residues dissolved in 20 ml of
isooctane were passed through Sep-Pak (silica cartridge, Waters Associates)
with a flow rate below 05 ml/ml and then 5 ml of isooctane was passed to
eliminate non-polar carotenoids. After the elution with 15 ml of isooctane-1,4dioxane (98:2, v/v), the eluate was concentrated, dissolved in n-hexane and
then applied on a thin layer chromatography (Kieselgel60F254 ). The thin layer
chromatography was developed with benzene. The band appeared as dark blue
against yellow background under UV light (Rf = 062) was scraped off and
extracted with n-hexane-ethanol (5: 1, v/v). Pale yellow oil, 36 mg, was
obtained in the overall recovery of about 34% by evaporation.
The isolate dissolved in ethanol showed the typical absorption spectrum of
Amax at 292 nm and corresponded to the authentic a-tocopherol. The peak of
the retention time of the isolate on a high performance liquid chromatography,
45 min, corresponded with that of a-tocopherol. Therefore, the oily substance
isolated from E. gracilis Z cells was identified as a-tocopherol.

Production of a-Tocopherol by Euglena gracilis Z

When E. gracilis Z was grown on a conventional algal medium, Koren-Hutner

(KH) medium, the maximum amount of a-tocopherol, 198 mg/liter culture
broth and 026mg/g dry cells, was obtained at 9th day of cultivation. To
improve the productivity and simplify the medium composition, components of
culture medium and their concentrations were investigated using KH medium
as the basal medium. As a result, a medium was derived as an a-tocopherol
production medium consisting of glucose (2%), peptone (12%), inorganic
salts, thiamin and cyanocobalamin, pH 50. The amount of a-tocopherol in the
culture broth increased with increase of the growth and chlorophyll levels in
the cells, and reached a maximum after 10 days of cultivation. The amount of
a-tocopherol was 99 mg/liter culture broth and 11 mg/ g dry cells.
The production of a-tocopherol, however, was much higher under light
conditions than in the dark, indicating that chloroplasts whose differentiation is
induced by light could be related to a-tocopherol productivity.
The effect of additions of a-tocopherol biosynthesis-related compounds on
a-tocopherol production was investigated. L-Tyrosine, L-phenylalanine, isopentenyl alcohol, AMP, 4-hydroxyphenylpyruvate and homogentisate apparently increased the a-tocopherol productivity. Since L-tyrosine and Lphenylalanine share the biosynthetic pathway with tocopherols, their addition

Vitamin E Production


might repress their own biosynthesis and thus increase the biosynthetic flow to
a-tocopherol. Homogentisate is a direct precursor of the tocopherol biosynthesis. Among prenyl alcohols, which might be involved in the tocopherol
biosynthesis in the form of pyrophosphate to be incorporated to the phytyl side
chain, isopentenyl alcohol was slightly effective on a-tocopherol production.
When these compounds were added simultaneously to the a-tocopherol
production medium, a combination of 01 % homogentisate and 001 %
L-tyrosine was the best for a-tocopherol production. Moreover, it was found
that higher productivity can be obtained when this mixture was added to the
culture medium only at 5th day of cultivation rather than at the beginning.
Feeding of ethanol to the medium together with peptone, when glucose was
almost consumed, gave higher production of a-tocopherol than that of
glucose-peptone in spite of the same cell yield.
Supplementation with ethanol and peptone, and homogentisate and Ltyrosine during cultivation was performed finally. As shown in Fig. 2, the
specific production of a-tocopherol increased markedly when the mixture was
fed at 5th, 7th and 9th days of the cultivation. On the other hand, when only
ethanol and peptone were fed, the amount of a-tocopherol in the culture broth
increased according to increase in cell mass but the intracellular content of
a-tocopherol was almost the same as that of the non-supplemented culture.

























15 0

lS 0
Cultivation Time (day)



Fig. 2. a-Tocopherol production in a batch culture. During the cultivation on the

a-tocopherol production medium (A), 1% ethanol and 06% peptone (B), and 1%
ethanol, 06% peptone, 01% homogentisate and 001 % L-tyrosine (C) were added at
periods indicated by arrow. Concentrations indicated are the final concentration (w/v)
in the medium.

Y. Tani

Table 1
Progress of a-Tocopherol Production by E. gracilis Z

Cell yield
(g /liter)


KH medium
a-Tocopherol production medium
Added L-tyrosine
Added homogentisate
Added L-tyrosine, homogentisate
and ethanol
Fed L-tyrosine, homogentisate,
ethanol and peptone

a-Tocopherol produced

(mg/g dry cells)










The amount of a-tocopherol produced in the culture reached 1436 mg/liter

culture broth and 51 mg/g dry cells at 20th day of cultivation.
The increase in intracellular production of a-tocopherol by E. gracilis Z
during the optimization study on culture conditions is summarized in Table 1.
The final productivity in this medium was 718-fold that of the original KH
medium and the intracellular content was increased 196-fold.


As represented in Fig. 1, a-tocopherol is composed of a chroman ring and an

aliphatic side chain. The naturally occurring molecule has three chiral centres



Wittig reaction

Fig. 3. Synthesis of (2R, 4'R, 8'R)-a-tocopherol by coupling optically active chroman
aldehyde with a C I5 phosphonium salt derived from natural phytol.

Vitamin E Production




:C 1 :



Fig. 4. Construction of an optically active C 14 side chain from S-fJ-hydroxyisobutyric


with the (2R, 4'R, 8'R)-configuration. Chemical synthesis of vitamin E-which

is now industrially applied-yields a racemic product. In this respect, several
attempts are made to synthesize the natural optically active compound
(Leuenberger, 1985).
One example of such synthesis involves the coupling of optically active
chroman aldehyde with a CIs-phosphonium salt derived from natural phytol
(Wittig reaction), followed by double bond hydrogenation and hydrolysis of
the protecting group. The natural side chain precursor, phytol, is the source of
the two chiral centres (Fig. 3).
Recently, the total synthesis of the optically active side chain has been
reported. The formation of such side chains starts from small optically active
building blocks (C: or Cn, which are produced via microbial reactions. In this
respect, microbiologically formed S-( + )-fJ-hydroxyisobutyric acid (a C:synthon) has been used to produce a C l4-vitamin E side chain according to the




Fig. S. Microbial reactions affording optically active S- or R-fJ-hydroxyisobutyric acid
(fJ-HIBA). 1. S-fJ-HIBA; Pseudomonas putida or other micro-organisms. 2. S- or
R-fJ-HIBA; depending on micro-organisms used. 3. R-fJ-HIBA; Gluconobacter roseus.
4. R-fJ-HIBA; Candida humicola.


Y. Tani





Fig. 6. Construction of an optically active C I5 side chain.

scheme: C 5 + C; + C i + C; (Fig. 4). Several examples of microbial bioconversion leading to this optically active S- or R-tJ-hydroxyisobutyric acid are given
in Fig. 5.
Another possibility for the synthesis of an optically active C 15 side chain is a
C; + C; + C; condensation (Fig. 6). The optically active C5 -synthon can be
produced from a suitable unsaturated precursor via enantioselective microbial
reduction (Fig. 7). Saccharomyces cerevisiae and Geotrichum candidum are
useful in this respect and formed the C 5 -synthon, (S)-3-methyl-y-butyrolactone, in high yield from suitable precursors.
The optically active chroman ring has also been synthesized using (S)-( + )citramalic acid to introduce chirality; asymmetric hydration of mesaconic acid
into the desired (S)-( + )-citramalic acid has been performed with Clostridium
tetanomorphum (Fig. 8). This synthon has also been used to form the vitamin
D 3 -metabolite, (25S, 26)-dihydroxycholecalciferol.
A more detailed description of the use of microbially obtained optically
active synthons in vitamin E synthesis is given by Leuenberger (1985).



Fig. 7. Bioconversion leading to optically active C 5 -synthons.

Vitamin E Production




Clostridium tetanomorphum













Fig. 8. Synthesis of the optically active chroman moiety from trimethylhydroquinone

and (S)-citramalic acid.

Diplock, A. T., Green, J., Edwin, E. E. & Bunyan, J. (1961). Tocopherol,
ubiquinones and ubichromenols in yeasts and mushrooms. Nature, Lond., 189,
Forbes, M., Zilliken, F., Roberts, G. & Gyorgy, P. (1958). A new antioxidant from
yeast. Isolation and chemical studies, J. Am. Chern. Soc., SO, 385-9.
Green, J., Price, S. A. & Gare, L. (1959). Tocopherols in microorganisms, Nature,
Lond., 184, 1339.
Hughes, P. E. & Tove, S. B. (1982). Occurrence of a-tocopherolquinone and
a-tocopherolquinol in microorganisms. J. Bacteriol., 151, 1397-402.
Leuenberger, H. G. W. (1985). Microbiologically catalyzed reaction steps in the field of
vitamin and carotenoid synthesis. In Biacatalysts in Organic Synthesis, eds. J.
Tramper, M. C. Van der Plas & P. Linko. Elsevier, Amsterdam, pp. 99-118.
Ruggeri, B. A., Gray, R. J. H., Watkins, T. R. & Tomlins, R. I. (1985). Effects of
low-temperature acclimation and oxygen stress on tocopherol production in Euglena
gracilis Z. Appl. Env. Microbial., 50, 1404-8.


Y. Tani

Shigeoka, S., Onishi, T., Nakano, Y. & Kitaoka, S. (1986). The contents and
subcellular distribution of tocopherols in Euglena gracilis. Agric. BioI. Chern., 50,
Taketomi, H., Soda, K. & Katsui, G. (1983). Results of screening test in tocopherols in
microbial realm. Vitamins (Japan), 57, 133-8.
Tani, Y. & Tsumura, H. (1989). Screening of tocopherol-producing microorganisms
and a-tocopherol production by Euglena gracilis Z. Agric. Bioi. Chern., 53, 305-12.

Chapter 7
Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan

The classical investigations performed by Burr & Burr in 1929-1930 clearly
demonstrated that animals cannot survive without an exogenous source of fat.
Subsequent studies have shown that supplementing a fat-free diet with certain
essential fatty acids prevents the occurrence of symptoms of fat deficiency. It is
now well established that both the n-6 and n-3 polyunsaturated fatty acids
(PUFAs) are essential in human nutrition. Current information indicates that
they have important roles in the structure and function of biological membranes. They have recently attracted great interest due to their unique
biological activities, such as lowering of plasma cholesterol level, prevention of
thrombosis, and so on. In addition, they are precursors for the biosynthesis of
a variety of a large family of structurally related C-20 compounds, such as
prostaglandins, prostacyclins, thromboxanes and leukotrienes. Such current
status of PUFAs suggests that they are highly important substances in the
pharmaceutical, medical and nutritional fields. These substances are sometimes
referred to as the vitamin-F-group.
In this chapter, we summarize current progress in the biotechnology of
PUFA production. Emphasis will be placed on possible advantages of
micro-organisms as practical sources of PUFAs. Several reviews summarize
other various aspects of PUFAs and provide excellent background information
(Wagner & Folkers, 1964; Rahm & Holman, 1971; Brenner, 1974; Fluco,
1974; Holloway, 1983; Numa, 1984; Dyerberg, 1986; Needleman et al., 1986;
Horrobin & Huang, 1987).


The n-6 and n-3 families of PUFAs are distinguished by differing positions of
the double bond closest to the methyl terminal group of the fatty acid chain


S. Shimizu & H. Yamada


c:::::::.~ --~ ~~~~ c::::::~~ ~~








c:::::::;:.--~~ ~~- c:::::::;:







-c::;::::; -- c::;:;:;~- c:::;::;:::




20:4 ARA

C:::::::::=H ~C::::::::H - ~~.. ~~. C:::::::




20:5 EPA


22:6 DHA

Fig. 1. Pathways for the biosynthesis of PUFAs of the n-7 (palmitoleic), n-9 (oleic),
n-6 (linoleic) and n-3 (a-linolenic) families. PUFAs include the n-7, n-9, n-6 and n-3

families, which are defined by the position of the double bond closest to the methyl end
of the fatty acid molecule. Desaturation occurs toward the carboxyl end of the molecule
and chain elongation at the carboxyl end, leaving the methyl end unaltered. Usually,
PUFAs of each family are not inconvertible. GLA, r-linolenic acid; DHGLA,
dihomo-r-linolenic acid; ARA, arachidonic acid; EPA, eicosapentaenoic acid; DHA,
docosahexaenoic acid.

(see Fig. 1). Linoleic acid (9,12-cis-octadecadienoic acid, 18: 2 n-6) and
a-linolenic acid (9,12,15-cis-octadecatrienoic acid, 18: 3 n-3) are found abundantly in plants. They are synthesized de novo in higher plants and together
represent nearly all the polyenoic fatty acids present. Linoleic acid comprises
about 50% of the total fatty acids in corn, soybean, sunflower-seed and
cotton-seed oils and over 70% of the fatty acids of safflower oil. a-Linolenic
acid represents over 50% of the total fatty acids of linseed and perilla oils. In
Oenothera biennis (evening primrose) and O. lamarckiana, however, linoleic
acid is further desaturated to form y-linolenic acid (6,9,12-cis-octadecatrienoic
acid, 18: 3 n-6), which comprises about 10% of the total fatty acids in their
seeds. Mosses (Hartmann et al., 1986), algae (Seto et al., 1984) and protozoa
(Erwin & Bloch, 1964), some fungi (Gellerman & Schlenk, 1979; Yamada et
al., 1987, Shimizu et al., 1988a,1989a) and some marine bacteria (Yazawa et
al., 1988) can further add a 2-carbon unit to these C-18 PUFAs to form C-20
Animals cannot synthesize fatty acids with double bonds at either the n-3 or
n-6 positions, and therefore are dependent upon dietary sources for these fatty
acids. Most animals, however, possess enzyme systems for further desaturation
and chain elongation at the carboxyl end of these fatty acid molecules. The
resulting further desaturated or longer-chain derivatives such as y-linolenic
acid, dihomo-y-linolenic acid (8,1l,14-cis-eicosatrienoic acid, 20: 3 n-6),
arachidonic acid (5,8,1l,14-cis-eicosatetraenoic acid, 20: 4 n-6), eicosapentaenoic acid (5,8,1l,14,17-cis-eicosapentaenoic acid, 20:5 n-3, or EPA)
and docosahexaenoic acid (4,7,10,13,16,19-cis-docosahexaenoic acid, 22: 6

Microbial Production of PUFAs


n-3, or DHA), are important as essential fatty acids in human nutrition or as

structural components of cellular membrane phospholipids.
Most aquatic plants can synthesize the longer chain n-3 fatty acids including
EPA and DHA from common metabolites. Fish, shellfish and marine
mammals at the apex of the aquatic food chains contain relatively large
quantities of these n-3 fatty acids, and therefore provide the human diet with a
direct source of preformed EPA or DHA. Terrestrial food chains also contain
some n-3 PUFAs, but they are dominated by the n-6 PUFA family because
most land plants synthesize linoleic acid rather than a-linolenic acid.
A variety of PUFAs other than those of the n-6 and n-3 families have been
detected in living organisms. Some of them are also listed together with the
PUFAs of the n-6 and n-3 families in Table 1.
Table 1
Several Naturally Occurring PUFAs and their Dietary Sources and Biological Functions
Fatty acid families and
their major members
Linoleic acid
9,12-cis-18: 2
y-Linolenic acid (GLA)
6,9,12-cis-18: 3

Dihomo-y-linolenic acid
8, 11,14-cis-20 : 3
Arachidonic acid
5,8,11 , 14-cis-20 : 4
Docosapentaenoic acid
4,7, 10,13, 16-cis-22 : 5
a-Linolenic acid
9,12,15-cis-18: 3
Eicosapentaenoic acid
(EPA) 5,8,11,14,17cis-20:5
Docosahexaenoic acid
(DHA) 4,7,10,13,16,19cis-22: 6
Oleic acid
9-cis-18: 1
Eicosatrienoic acid
5,8,11-cis-20: 3
Palmitoleic acid
7-cis-16: 1

Major dietary

Most vegetable oils

Meat, liver, brain
Some vegetable oils
(soy, linseed), leafy
Fish, shellfish, algae

Fish, shellfish, algae

Animal and vegetable


Biological function,
tissue distribution
and other remarks

Essential fatty acid, minor

component of most tissues
Major fatty acid of evening
primrose seed; used as an additive of feed, cosmetics, etc.
Precursor of 1-group of prostaglandins
Major component of most membrane phospholipids; precursor
of 2-group of prostaglandins
Frequently found in tissues
deficient in n-3 PUF As
Minor component of tissues
Precursor of 3-group of prostaglandins; prevent thrombosis;
used as pharmaceuticals and a
feed additive
Minor component of membrane
phospholipids in retinal photoreceptors, sperm, etc.
Major component of many tissues
Accumulated in total essential
fatty acid deficiency
Prevent apoplexy


S. Shimizu & H. Yamada

The biosynthesis of linoleic acid and a-linolenic acid in higher plants, lower
plants and some micro-organisms takes place via common metabolic C-18 fatty
acids, i.e. stearic acid and oleic acid, from acetate fragments produced during
the catabolism of carbohydrates (Fluco, 1974). Linoleic acid and a-linolenic
acid are usually the terminal products of fatty acid synthesis in higher plants.
These two fatty acids are converted to two distinct families of long-chain and
more unsaturated fatty acids through the n-6 and n-3 routes, respectively, as
shown in Fig. 1. In these routes, the successive chain elongation and
desaturation take place toward the carboxyl end of the fatty acid molecule.
Thus, the members of the n-6 family have six carbons after the last double
bond in the molecule, and the members of the n-3 family have three carbons in
the terminal structure. The first step of the biosynthesis of PUFAs is the
desaturation at the ~6-position of linoleic acid or a-linolenic acid to yield
y-linolenic acid or 6,9,12,15-cis-octadecatetraenoic acid (18: 4 n-3),
respectively (~6-desaturation). The ~6-desaturation is followed by the chain
elongation to the respective C-20 PUFAs, i.e. dihomo-y-linolenic acid and
8,1l,14,17-cis-eicosatetraenoic acid (20:4 n-3). The ~5-position of dihomo-ylinolenic acid in the n-6 route is then desaturated to yield arachidonic acid
(~5-desaturation). Usually, arachidonic acid is the terminal product of the n-6
route. The ~5-desaturation on 20: 4 n-3 fatty acid in the n-3 route results in
the formation of EPA, which is followed by chain elongation and desaturation
to yield DHA. It is usually the end product of the n-3 route.
The n-6 and n-3 routes have been suggested to share the same enzymes or
enzyme systems for the transformations of C-18 PUFAs to C-20 PUFAs
(Brenner, 1974). Therefore, linoleic acid and a-linolenic acid compete with
each other for the same enzymes. In animals, the n-6 family is usually favored
in this competition, forming arachidonic acid. Neither isolation nor detailed
characterization of enzymes responsible for the desaturation and chain
elongation reactions involved in both routes has been performed yet except for
~6 desaturase from rat liver (Okayasu et al., 1981) . Desaturation reactions
require CoA, A TP, Mi+ and NAD(P)H as cofactors, suggesting that
desaturation enzymes are a kind of oxygenase and that only fatty acyl-CoA
derivatives are substrates for the enzymes. Elongation enzymes have also been
suggested to utilize only acyl-CoA derivatives as substrates.
The members of these two families of fatty acids constitute nearly all the
PUFAs in most living organisms. However, small amounts of different families
of fatty acids have been detected. 5,8, ll-cis-eicosatrienoic acid (20: 3 n-9) is
derived from oleic acid (9-cis-octadecenoic acid, 18: 1 n-9) through the n-9
route which involves the following successive reactions: desaturation at the ~6
position of oleic acid to form 6,9-cis-octadecadienoic acid (18: 2 n-9), followed
by its carbon chain elongation to 8,1l-cis-eicosadienoic acid (20:2 n-9) and
desaturation at the ~5-position of the latter to 20: 3 n-9 fatty acid (Fig. 1).
7,10,13-cis-eicosatrienoic acid (20: 3 n-7) is derived from palmitoleic acid


Microbial Production of PUFAs

(9-cis-hexadecenoic acid, 16: 1 n-7) through the n-7 route, as shown in Fig. 1.
These data indicate that there are at least four families of PUFAs in


The available lipid sources relatively rich in C-18 (y-linolenic acid) and C-20
PUFAs are seeds of evening primrose, and animal tissues, algal cells and fish
oils, respectively. For practical purposes, however, these conventional sources
are not satisfactory, both in their lipid contents and in the PUFA contents of
the resultant lipid products (see Table 2).
Micro-organisms are thought to be very promising lipid sources because of
their extremely high growth rates in simple media and the simplicity of their
manipulation. Erwin & Bloch (1964) suggested that lower classes of organisms
including micro-organisms can be classified into several groups based on their
ability to produce PUFAs, as shown in Fig. 2. Show (1965) pointed out that
some fungi belonging to the Mucorales can accumulate relatively large
Table 2
Sources of PUFAs


Mortierella isabellina
Penicillium cyaneum
M. elongata lS-5
M. alpina lS-4
M. alpina lS-4b
M. alpina lS-4
M. alpina 20-17c
M. alpina 20-17d
Chlorella minutissima
Porphyridium cruentum
Euglena gracilis
Seed oile
Porcine liver
Fish oil'




a FA, fatty acid.

b grown with sesame oil extract.
grown at 12e.
d grown with linseed oil at 28e.
e Oenothera biennis.
f Scomber scrombrus.
g not reported.

FAa content




(% in
total FA)






S. Shimizu & H. Yamada



Metazoa; Des; y-lin

----- "I

Green Algae;X;a-lin



Phytomonads;X;a- and y-lin

Yeasts and Fungi; - - - - .

~somonads;Des;a- and y-lin

a-and r-lin

Red Algae;Des;a-lin




Blue-green Algae;Des;a-lin

Photosynthetic Bocteria;An;O

Hypothetical Primitive Organisms

Fig. 2. Presumed phylogenetic relationships between groups of existing protists, and

patterns of fatty acid biosynthesis observed in these groups. An, Anaerobic mechanism
for the synthesis of monoenoic acids; Des, oxidative de saturation for the synthesis of
monoenoic acids; X, unknown ('plant') mechanism for the synthesis of monoenoic
acids; 0, absence of di- and polyunsaturated fatty acids; a-lin, the n-3 route for the
synthesis of PUFAs; y-lin, the n-6 route for the synthesis of PUFAs. For details, see
Erwin & Bloch (1964).

amounts of y-linolenic acid in their mycelia. Based on these early observations, several workers have only since early 1980 started to screen for
micro-organisms capable of accumulating lipids containing these PUFAs in
order to obtain more suitable sources for large-scale preparation of these
Recently, Suzuki and co-workers (Suzuki, 1985, 1988; Suzuki & Yokochi,
1986) found several potent producers of y-linolenic acid through their
extensive screening for filamentous fungi. They reported that a strain of
Mortierella isabellina accumulates about 5 g/liter of y-linolenic acid when
grown in a medium containing glucose as a major carbon source. Based on this
finding, commercial production of y-linolenic acid with a Mortierella fungus
started recently in Japan. Another process is in progress in England.
Yamada and co-workers (Yamada et al., 1987, 1988) found that several
fungal strains produce large amounts of lipids rich in arachidonic acid,
dihomo-y-linolenic acid or EPA, or all of these C-20 PUFAs. They are new
and promising sources for C-20 PUFAs.

"I-Linolenic acid

Suzuki (1985, 1988) reported that several Mortierella strains, belonging to

the subgenus Micromucor produce lipids rich in y-linolenic acid in their

Microbial Production of PUFAs


mycelia when grown in a liquid medium containing glucose or sugar cane

molasses as a major carbon source. They can grow rapidly in a medium
containing a high concentration glucose (200 g/liter) under the conditions of
D.O., 2 ppm; pH, 40; and cultivation temperature, 30C, yielding a mycelial
mass of about 80 g/liter. In the case of M. vinacea, the lipid content of the
resultant mycelia was found to be 48%, by weight, which corresponded to
38 g/liter of the lipids. The purified oil obtained from the mycelia contained
myristic acid (07%, by weight), palmitic acid (272), palmitoleic acid (0,9),
stearic acid (S'7), oleic acid (43,9), linoleic acid (120), ')I-linolenic acid (83),
arachidic acid (0,6), 11-cis-eicosenoic acid (04), behenic acid (01) and erucic
acid (02). The ')I-linolenic acid content in the oil is comparable to that of
Oenothera biennis seed oil. The low content of linoleic acid in the oil may be
advantageous in obtaining highly pure ')I-linolenic acid preparation in high
Recently, M. ramanniana (Hansson & Dostcilek, 1988) and Mucor ambiguus
(Fukuda & Morikawa, 1987) were evaluated for their productivities of
')I-linolenic acid from a viewpoint of biochemical engineering. The latter fungus
immobilized in porous support particles has been shown to excrete lipids
containing ')I-linolenic acid into the culture broth and/or the surface of the cell
wall in the presence of a nonionic surfactant.

Arachidonic acid

Yamada et al. (1987, 1988) assayed productivities of C-20 PUFAs in about 300
fungal isolates from natural sources which are capable of growing on an agar
plate containing stearic acid or elaidic acid as the main carbon source for
growth. They also tested more than 600 stock strains of a wide variety of
micro-organisms, i.e. bacteria (23 genera), actinomycetes (4 genera), yeasts
(20 genera), filamentous fungi (45 genera) and basidiomycetes (11 genera).
Bacteria, actinomycetes, yeasts and basidiomycetes, in general, did not
produce detectable amounts of C-20 PUFAs intracellularly. Most C-20 PUFA
producers were found to be filamentous fungi belonging to the orders of
Mucorales and Entomophthorales. Through this screening, they found that 60
strains of Mortierella and 4S isolates from natural sources produce large
amounts of C-20 PUFAs of the n-6 family (i.e. arachidonic acid and
dihomo-')I-linolenic acid) together with C-18 PUFAs of the same family
(mainly ')I-linolenic acid). Most of the PUFA-producing isolates were found to
belong to the genus Mortierella. It should be noted that all of these Mortierella
strains found as C-20 PUFA producers belong to the subgenus Mortierella.
Neither the stock cultures nor isolates belonging to the subgenus Micromucor
showed any detectable accumulation of C-20 PUFAs, although they produced
n-6 C-18 PUFAs such as ')I-linolenic acid. The arachidonic acid contents of
most of these strains accounted for more than 1S% of the total extractable
fatty acids. These values represent more than SO% of the PUFAs and are
particularly high when compared with those in the case of C-18 PUFAs (Table

hygrophila IFO 5941

zychae CBS 65268
elongata CBS 12171
elongata IS-5 AKU 3999
parvispora 2S-13 AKU 3994
schmuckeri NRRL 2761
alpina IS-4 AKU 3998
alpina 20-17 AKU 3996
alpina CBS 250-53
alpina 1-83 AKU 3995
alpina CBS 21935



























Fatty acid composition (%)C











G For details, see Shimizu et al. (1988b).

b Values are given in mg/ml culture broth or mg/g dry mycelia.
cValues are given in weight %.16:1, palmitic acid; 18:0, stearic acid; 18:1, oleic acid; 18:2, linoleic acid; 18:3, y-linolenic acid; 20:3, dihomo-y-linolenic
acid; 20:4, arachidonic acid; 20:5, EPA. EPA and other PUFAs of the n-3 family were not detected.
d FA = fatty acid.



Table 3

Comparison of Arachidonic Acid Productivities in Mortierella Fungi and their Mycelial Fatty Acid CompositionG

elongata IS-5
alpina IS-4
alpina IS-4
alpina 20-17



(g /liter)


a FA = fatty acid; 20:4, arachidonic acid.







(g /liter)





(% in
total FA)





Table 4
Production of Arachidonic Acid by Selected Mortierella Strains under Optimal Culture Conditions








S. Shimizu & H. Yamada


3). No detectable amount of free fatty acids was found in lipid fractions
extracted with chloroform-methanol, suggesting that the PUFAs produced are
present as triglyceride and/or phospholipid forms . Through this screening,
they obtained three isolates, which were taxonomically identified to be
Mortierella alpina IS-4 , M. alpina 20-17 and M. elongata IS-5, as potent
producers of arachidonic acid.
Subsequently, they investigated the cultural conditions for arachidonic acid
production with the three Mortierella strains (Yamada et al., 1987,1988). These
three fungi utilize not only glucose but also glycerol, maltose, n-hexadecane
and n-octadecane as carbon source for the arachidonic acid production. As
shown in Table 4, all the strains produced more than 1 g/liter of arachidonic
acid under 5-liter bench scale fermentor conditions. Based on the results of
studies on individual and combined factors affecting arachidonic acid production, they selected M . alpina IS-4 as the most promising producer. Using a
2000-liter fermentor, 225 g/liter mycelia (dry weight) containing 440%, by
weight, of the lipids, in which arachidonic acid comprised 31 0% of the total
fatty acids, was produced under the conditions of intermittent feeding of
glucose and 10 days cultivation at 28C (Fig. 3(a . They also reported that the
aradchionic acid in the harvested mycelia can be specifically enriched when
the mycelia are allowed to stand for a further few days at room temperature.
The arachidonic acid content of the resultant mycelia reached nearly 70% of
the total fatty acids (Fig. 3(b. Fractionation of the lipids in the mycelia
demonstrated that triglycerides (701 %) and phospholipids (23,4%) were their



20:~ \'S'3G

l' ::: ~tlMm

10 I i }
9 }{

CultlVOtlon time (cloys)




FA composition (%)

Fig. 3. Production of arachidonic acid by M. alpina IS-4 under the optimal culture
conditions. Mortierella alpina IS-4 was cultivated in a 2000-liter fermentor containing
1400 liters of a medium containing 2% glucose and 1% yeast extract, pH 60 at 28C.
Glucose was added to the medium as shown in (a). Changes in the mycelial fatty acid
composition during growth are shown in (b). 16 : 0, Palmitic acid ; 18 : 0, stearic acid;
18 :2, linoleic acid ; 18 : 3G, y-linolenic acid ; 20:3, dihomo-y-linolenic acid; 20:4 or
Ara , arachidonic acid.

Microbial Production of PUFAs


meyer SOL lar



mil' 1






,I ~';;:I.~I;;:~":"




mil ' 2

Fig. 4. Flow sheet for the production of fungal oil of high arachidonic acid content.

major components, and about 50% of the arachidonic acid was present in the
phospholipid fraction.
The lipids containing arachidonic acid can be obtained as an oil from the
mycelia of M. afpina 1S-4 in a good recovery (80-90%) through the following
successive steps: separation of mycelia by filtration , drying, crushing by ball
mill, extraction of the lipids with n-hexane , removal of insoluble materials by
centrifugation, decolorization and deodorization with active charcoal , and
concentration . The resultant purified oil contains myristic acid (02%, by
weight), palmitic acid (70) , palmitoleic acid (01), stearic acid (28), oleic acid
(66), linoleic acid (59), y-linolenic acid (3 9), 11-cis-eicosenoic acid (09),
11 ,14-cis-eicosadienoic acid (1 3), dihomo-y-linolenic acid (39) and arachidonic acid (674) . The overall process is outlined in Fig. 4. The arachidonic acid
in the purified oil can be isolated as the methyl or ethyl ester in a good
recovery after successive transesterification, liquid-liquid partition chromatography and high-performance liquid chromatography.

Dihomo-y-Linolenic Acid

Most arachidonic acid-producing fungi can accumulate small amounts of

dihomo-y-linolenic acid, when they are cultivated under the conditions for
arachidonic acid production (Yamada et af. , 1988; Shimizu et af. , 1988a).
However, the ratio of dihomo-y-linolenic acid to arachidonic acid in mycelia
is usually only about 02. Yamada et af. (1988) reported that M. afpina
IS-4 accumulated 06 g/Iiter of dihomo-y-linolenic acid. Because dihomo-y-linolenic acid is a precursor of arachidonic acid in the n-6 route, the arachidonic
acid accumulated in the mycelia must be produced via dihomo-y-linolenic


S. Shimizu & H. Yamada







20 :





Time (days)

Fig. 5. Time course of the production of dihomo-y-linolenic acid by M. alpina 1S-4.

Mortierella alpina 1S-4 was cultivated in a 50-liter bench scale fermentor containing 30
liters of a medium containing 2% glucose, 1 % yeast extract and 0022% of sesame oil
extract, pH 60. Glucose was added to the medium at the time indicated by the arrow.
DHGLnA, dihomo-y-linolenic acid; Ara, arachidonic acid.

acid, suggesting that all the arachidonic acid-producing fungi potentially have
the ability to produce large amounts of this fatty acid.
Shimizu et al. (1989a) reported that the mycelial dihomo-y-linolenic acid
content of M. alpina 1S-4 increases, with an accompanying marked decrease in
its arachidonic acid content, on cultivation with sesame oil. They suggested
that this unique phenomenon is due to specific repression of the conversion of
dihomo-y-linolenic acid to arachidonic acid by the oil. The effective factor(s)
causing this phenomenon was found to be present in the non-oil fraction of the
oil, after fractionation of the oil with acetone. In a study on optimization of the
culture conditions for the production of dihomo-y-linolenic acid by M; alpina
1S-4, a medium containing glucose, yeast extract and the non-oil fraction was
found to be suitable for the production. Under the optimal conditions in a
SO-liter bench scale fermentor, the fungus produced 217 g/liter of dihomo-ylinolenic acid (107 mg/ g dry mycelia) (Fig. S). This value accounted for 231 %
of the total mycelial fatty acids. The mycelia were also rich in arachidonic acid
(S3S mg/g dry mycelia, 112%). Other major fatty acids in the lipids were
palmitic acid (241%, by weight), stearic acid (70), oleic acid (201), linoleic
acid (66) and y-linolenic acid (41).

Eicosapentaenoic Acid (EPA)

4.4.1 EPA -Production under Low Temperature Growth conditions

Yamada and co-workers (Shimizu et al., 1988a; Yamada et aI., 1988) investigated the growth conditions causing compositional changes in mycelial fatty

Microbial Production of PUFAs


acids using the arachidonic acid producers obtained through their screening
studies, because they could not find any strains capable of accumulating a
detectable amount of EPA under their screening conditions (see Table 3).
They found that lowering the cultivation temperature caused the additional
accumulation of a PUFA with five double bonds, EPA, by all the arachidonic
acid producers tested, as shown in Table 5. In all cases, cultivation at low
temperature significantly increased the mycelial phospholipid content. For
example, the mycelial lipids extracted from M. alpina 1S-4 grown at 12C
contained 476% (by dry weight) of phospholipids and 357% of triglycerides.
More than 60% of the EPA accumulated in the mycelia was found in the
phospholipid fraction. On the other hand, the lipid from the mycelia grown at
28C contained only 216% of phospholipids, the triglyceride fraction comprising 735% of the total extractable lipids. It should be noted that most of these
arachidonic acid producers grow well at low temperature (6-16C), and yield
enough mycelia in simple growth media.
Subsequently, they studied the mechanism involved in this unique phenomenon. The results of experiments with cell-free extracts of M. alpina 1S-4
demonstrated that the enzyme(s) that catalyzes the formation of EPA is
produced even when the fungus is grown at high temperature, but that the
reaction(s) yielding EPA does not take place at high temperature, as shown in
Table 6. Since all the EPA-producing Mortierella strains accumulate PUFAs of
the n-6 family (i.e. y-linolenic acid, dihomo-y-linolenic acid and arachidonic
acid) in their mycelia and do not accumulate PUFAs of the n-3 family other
than EPA, they suggested that an n-6 PUFA, probably arachidonic acid, may
be a precursor of EPA. If this is the case, an enzyme(s) or enzyme system
catalyzing the methyl-end directed desaturation of arachidonic acid (a17desaturation) may be activated on cold adaptation. The resultant EPA may be
necessary for maintaining a proper membrane fluidity in a low temperature
environment (Shimizu et al., 1988a).
Among various arachidonic acid-producing strains, they selected M. alpina
20-17 as the most promising EPA producer. The fungus was found to
accumulate about O 5 g/liter (266 mgt g dry mycelia) of EPA on cultivation
for 7 days at 12C. This value accounted for 11 % of the total fatty acids in the
extracted lipids. Other major fatty acids in the lipids were palmitic acid (64%,
by weight), stearic acid (48), oleic acid (32), linoleic acid (31), y-linolenic
acid (45) and arachidonic acid (638) (Shimizu et al., 1988b).
4.4.2 Conversion of Linseed Oil to an Oil Containing EPA
Shimizu et al. (1989b) demonstrated that several arachidonic acid-producing
Mortierella strains accumulate detectable amounts of EPA in their mycelia
when grown in media containing a-linolenic acid. This observation suggests
that the n-3 route occurs in the Mortierella fungi as well as animals, although
they are lacking the ability to synthesize a-linolenic acid. This route seems to
be independent of growth temperature, because EPA production takes place
even when they are grown at 28C. The ability of the Mortierella fungi to














Fatty acid composition (%







a For details, See Shimizu et al. (1988a).

b Values are given in mg/g dry mycelia.
C Values are given in weight %. Any fatty acid of the n-3 family other than EPA was not detected. For abbreviations of fatty
acids, see Table 3.

M. hygrophila
IFO 5941
M. elongata
IFO 8570
M. exigua
IFO 8571
M. parvispora
20-24 AKU 3997

M. alpina IS-4
AKU 3998


Table 5
Formation of EPA and Changes in Mycelial Fatty Acid Composition in Mortierella Fungi at Various Growth temperaturea








Microbial Production of PUFAs


Table 6
Enzymatic Formation of EPA by a Cell-Free Extract of M. alpina 1S-4
Grown at 28C'

Omission from
reaction mixture


Cofactors C

Lipid, cofactors


EPA found (nmollml)

on incubation at


o (2)b

o (2)
o (4)
o (3)
o (4)
o (3)




For details, see Shimizu et al. (1988a).

Values obtained with the cell-free extract prepared from the mycelia
grown at 12C are given in parentheses.
ATP, CoA, NADPH and MgCI 2

convert added a-linolenic acid to EPA is very promising from a biotechnological viewpoint because there are various kinds of easily available natural oils
containing a-linolenic acid, and it is expected that they can be converted to
oils rich in EPA on incubation with these fungi. They examined the potential
of such natural oils as precursors of EPA and found that linseed oil, in which
a-linolenic acid amounts to about 60% of the total fatty acids, is the most
suitable for EPA production. Under optimal cultural conditions, M. alpina
20-17 converted 51 % of the a-linolenic acid in the added linseed oil into
EPA, the EPA production reaching 135 g/liter (415 mg/g dry mycelia). This
value is 28-fold higher than that obtained under low temperature growth
conditions. The resultant lipid is rich in either arachidonic acid or EPA,
Another advantage of this conversion is that it can be carried out under normal
growth temperature conditions (20-30C). Under such conditions, the fungal
growth is rapid and dense, and the energy costs for temperature control may
be less than those for cooling.

4.4.3 EPA from Algae

Micro-algae belonging to the genera Skeletonema, Phaeodactylium,
Dunaliella, ... were also found to be high EPA-producers. Algal ponds for
EPA-synthesis and other fine chemicals are already operational in Israel and in
the USA (California, Hawaii) (Borowitzka & Borowitzka, 1988).
Considerable progress has been made in biotechnological production of
PUFAs during the past years. Areas of progress include: (a) the finding of


S. Shimizu & H. Yamada

potent fungal strains rich in y-linolenic acid; (b) determination of optimal

conditions for large-scale submerged culture, which made it possible to operate
industrial-scale production of this fatty acid; (c) the finding that C-20 PUFAs
can be produced by filamentous fungi and algae efficiently; (d) selection of
potent producers of C-20 PUFAs; and (e) determination of culture conditions
for selective accumulation of arachidonic acid, dihomo-y-linolenic acid and
EPA by the selected fungi or algae. These PUFAs may be of considerable
importance in the utilization of these algal cells or fungal mycelia or their oils
as a feed supplement. In addition, they may be used as pharmaceutical
products. Further progress in this field depends on additional technological
development in downstream processes for the large-scale isolation of oils
containing PUFAs and/or pure PUFAs.
In conclusion, it seems to us that the nutritional, pharmacological and
economical potential of micro-organisms as a source of PUFAs are very high.
Borowitzka, M. A. & Borowitzka, L. J. (1988). Micro-algal Biotechnology. Cambridge
University Press, Cambridge, UK.
Brenner, R. R. (1974). The oxidative desaturation of unsaturated fatty acid. Molec.
Cell. Biochem. 3,41-52.
Dyerberg, J. (1986). Linolenate-derived polyunsaturated fatty acids and prevention of
atherosclerosis. Nutr. Rev. 44, 125-34.
Erwin, J. & Bloch, K. (1964). Biosynthesis of unsaturated fatty acids in microorganisms. Science, 143, 1006-12.
FIuco, A. J. (1974). Metabolic alterations of fatty acids. Ann. Rev. Biochem., 43,
Fukuda, H. & Morikawa, H. (1987). Enhancement of ')I-linolenic acid production by
Mucor ambiguus with nonionic surfactant. Appl. Microbiol. Biotechnol. 27, 15-20.
Gellerman, J. L. & Schlenk, H. (1979). Methyl-directed desaturation of arachidonic
acid to eicosapentaenoic acid in the fungus, Saprolegnia parasitica. Biochim.
Biophys. Acta, 573, 23-30.
Hansson, L. & Dostalek, M. (1988). Effect of culture conditions on mycelial growth
and production of ')I-linolenic acid by the fungus Mortierella ramanniana. Appl.
Microbiol. Biotechnol., 28, 240-46.
Hartmann, E., Beutelmann, P., Vandekerkhove, 0., Euler, R. & Kohn, G. (1986).
Moss cell cultures as sources of arachidonic acid and eicosapentaenoic acids. FEBS
Lett., 198,51-5.
Holloway, P. W. (1983). Fatty acid desaturation. In Enzymes, 3rd edn., Vol. XVI, ed.
P. D. Boyer, pp. 63-83. Academic Press, New York.
Horrobin, D. F. & Huang, Y.-S. (1987). The role of linoleic acid and its metabolites in
the lowering of plasma chloesterol and the prevention of cardiovascular disease. Int.
J. Cardiol., 17, 241-55.
Needleman, P., Turk, J., Jakschiik, B. A., Morrison, A. R. & Lefkowith, J. B. (1986).
Arachidonic acid metabolism. Ann. Rev. Biochem., 55,69-102.
Numa, S. (Ed.) (1984). Fatty Acid Metabolism and its Regulation. Elsevier,
Okayasu, T., Nagao, M., Ishibashi, T. & Imai, Y. (1981). Purification and partial
characterization of linoleoyl-CoA desaturates from rat liver microsomes. Arch.
Biochem. Biophys. 206,21-8.

Microbial Production of PUFAs


Rahm, J. J. & Holman, R. T. (1971). Essential fatty acids. In The Vitamins, Chemistry,
Physiology, Pathology, Methods, 2nd edn, Vol. III, ed. W. H. Sebrell Jr & R. T.
Harris, pp. 303-39. Academic Press, New York.
Seto, A., Wang, H. L. & Hesseltine, C. W. (1984). Culture conditions affect
eicosapentaenoic acid content of Chlorella minutissima. J. Am. Oil Chem. Soc., 61,
Shimizu, S., Shinmen, Y., Kawashima, H., Akimoto, K. & Yamada, H. (1988a).
Fungal mycelia as a novel source of eicosapentaenoic acid, Activation of enzyme(s)
involved in eicosapentaenoic acid production at low temperature. Biochem.
Biophys. Res. Commun., 150, 335-41.
Shimizu, S., Kawashima, H., Shinmen, Y., Akimoto, K. & Yamada, H. (1988b).
Production of eicosapentaenoic acid by Mortierella fungi. J. Am. Oil Chem. Soc.,
Shimizu, S., Akimoto, K., Kawashima, H., Shinmen, Y. & Yamada, H. (1989a).
Production of dihomo-y-linolenic acid by Mortierella alpina IS-4, J. Am. Oil Chem.
Soc. 66, 237-41.
Shimizu, S., Kawashima, H., Akimoto, K., Shinmen, Y. & Yamada, H. (1989b).
Microbial conversion of an oil containing a-linolenic acid to an oil containing
eicosapentaenoic acid. J. Am. Oil Chem. Soc. 66, 342-7.
Show, R. (1965). The occurrence of y-linolenic acid in fungi. Biochim. Biophys Acta,
Suzuki, O. (1985). y-Rinorensan no biseibutsu seisan (Production of y-linolenic acid by
micro-organisms). Hakko to Kogyo (Fermentation and Industry), 43, 1024-31.
Suzuki, O. (1988). Production of y-linolenic acid by fungi and its industrialization. In
Proceedings of the World Conference on Biotechnology for the Fats and Oils
Industry, Hamburg. American Oil Chemists' Society, Champaign, Illinois, pp.
Suzuki, O. & Yokochi, T. (1986). Production of y-linolenic acid by fungi. J. Am. Oil
Chem. Soc., 63, 434.
Wagner, A. F. & Folkers, K. (1964). The essential fatty acid group. In Vitamins and
coenzymes. Interscience Publishers, New York, pp. 389-406.
Yamada, H., Shimizu, S. & Shinmen, Y. (1987). Production of arachidonic acid by
Mortierella elongata IS-5. Agric. BioI. Chem. 51, 785-90.
Yamada, H., Shimizu, S., Shinmen, Y., Kawashima, H. & Akimoto, K. (1988).
Production of arachidonic acid and eicosapentaenoic acid by micro-organisms,
Proceedings of the World Conference on Biotechnology for the Fats and Oils
Industry, Hamburg. American Oil Chemists' Society, Champaign, Illinois, pp.
Yazawa, K., Araki, K., Okazaki, N., Watanabe, K., Ishikawa, C., Inoue, A., Numao,
N., & Kondo, K. (1988). Production of eicosapentaenoic acid by marine bacteria. J.
Biochem., 103, 5-7.

Chapter 8


Research Center for Cell and Tissue Culture, Faculty of Agriculture,
Kyoto University, Kyoto 606, Japan


Vitamin K, one of the fat-soluble vitamins, is known as an anti-hemorrhagic

factor. As Bentley & Meganathan (1982) pointed out in their review, vitamin
K was the first discovered growth factor for micro-organisms, though it was
named following the classical approach of animal nutrition in 1935. Dam
discovered a new fat-soluble organic compound which was found to possess
blood coagulation capacity, while establishing a bleeding condition in chicks
fed on a diet formulated for the study of sterol metabolism, and he proposed
the name of vitamin K (K for koagulation).
A major function of vitamin K is to act as a cofactor for the carboxylation of
protein-bound glutamate residues to form y-carboxyglutamates. A simplified
scheme of the blood clotting mechanism is shown in Fig. 1. Of these reactions,
only the synthesis of prothrombin appears to be dependent on vitamin K.
Vitamin K is now widely used for various types of bleeding symptoms:
hemorrhagic disease of new-born infants and hemorrhagic symptoms caused by
administration of antibiotics, salicylic acid or warpharin dosages. Menaquinone (MK, vitamin K 2 ) has much higher therapeutic activity than phylloquinone (vitamin Kl). However, phylloquinone has been used for a long time
as a medicine, because the chemical synthesis of MK is rather difficult and the
natural sources contain only a small amount of MK.
MK was prepared from bacterially putrefied fish meal as the first crystalline
antihemorrhagic vitamin that was clearly different from that of alfalfa,
phylloquinone. Its function in bacteria centers around its major role as an
electron carrier. Although more information is needed on its distribution in
micro-organisms, a correlation has been found between the taxonomical
classification of bacteria based on other criteria and the type of isoprenoid
quinones found in those cells (Collins & Jones, 1981). However, economic
production of MK by a microbiological process has not been demonstrated.

Y. Tani


(1) Precursor of Prothrombin



CO 2 , O 2

Vitamin K (red.)



(2) Prothrombin + Thromboplastin + Ca2 +



(3) Thrombin + Fibrinogen ---~. Fibrin (clot)

Fig. 1. Blood clotting mechanism.


The chemistry of vitamin K was established by the schools of Dam, Doisy and
Karrer, who succeeded in 1939 in isolating two compounds with vitamin K
activity, namely MK-7 (vitamin K 2 (35 from putrefied fish meal and phylloquinone from alfalfa meal (Fig. 2). The structure of phylloquinone was proven
to be 2-methyl-3-phytyl-l,4-naphthoquinone, and that of MK to be 2-methyl-3multiprenyl-l,4-naphthoquinone. The multiprenyl side chain of MK is of
variable length from 5 to 70 carbon atoms. A number of derivatives of the
basic structure also occur in nature; this is reflected in the modification of ring
substitutents at the C-2 or C-3 position or both.
Phylloquinone is a yellow viscous oil and MKs are light yellow microcrystalline plates. Melting points of MKs vary from 35 to 62C on the length of the
multiprenyl side chain. These are soluble in ether, petroleum ether, benzene,
n-hexane and acetone, slightly soluble in methanol and alcohol, and insoluble
in water. These are stable to air and heat, but very unstable for alkali and UV
irradiation. The absorption maxima are 243,248,261,270 and 325 nm with the
molecular extinctions coefficient of 18000-19000 at 248 nm.



Fig. 2. Chemical structure of vitamin K. (a) Menaquinone-n (MK-n); (b) phylloquinone.

Vitamin K2 and Vitamin Kl Production



Intense research based on microorganisms during the last two decades has
clearly proven the MK biosynthetic pathway as summarized by Bentley &
Meganathan (1982).
The biosynthetic pathway of MK in bacteria is now believed to consist of the
biosynthesis of demethyl MK by polyprenylation of 1,4-dihydroxy-2naphthoate, which is formed as the first naphthalenoid intermediate from
shikimate through chorismate(isochorismate) and o-succinylbenzoate, by catalysis of a membrane-associated transferase and its subsequent methylation.
The methyl group is derived from S-adenosylmethionine. Biosyntheses of MK
and aromatic amino acids share a common route until chorismate. The
immediate precursor, shikimate (chorismate), is incorporated to non-carboxyl
carbon atoms of 2-ketoglutarate to form the naphthoquinone nucleus. The
biosynthetic pathway of MK can now be summarized as shown in Fig. 3, to
which recent experimental results are added.
In spite of the increasing need of vitamin K supply, no microbiological process
for MK production has been developed so far. Only recently, work on the
fermentative production of this water-insoluble vitamin by micro-organisms
was initiated by Tani et al. (1984) and is described in this section.

Screening of Menaquinone-producing Micro-organisms

A number of micro-organisms were aerobically grown on media consisting of

glycerol and peptone as carbon and nitrogen sources, respectively. Washed
cells obtained from culture broth were suspended in acetone, and homogenized with a homogenizer. Lipophilic materials in the extract were absorbed on
SEP-PAK C 18 (Waters Associated, Inc.) and eluted with n-hexane. MK in the
n-hexane eluate was detected with a spectrophotometer by its typical UV
absorption spectrum. The amount of MK was calculated from the absorbance
at 248 nm by using the molar extinction coefficient.
For the determination by high performance liquid chromatography, cells
were homogenized in methanol and then extracted at 50-6O"C for 20 min.
Solid materials were removed by centrifugation. The resultant supernatant was
applied on a liquid chromatograph with 46 x 1oo-nm column of Cosmosil 5C18
(Nacalai Tesque Co.). The solvent used was methanol-dichloromethane. MK
was detected with a UV monitor at 248 nm.
Thin-layer and reversed-phase chromatographies were also applied for the
identification of MK. The thin-layer chromatography was performed with
development using a solvent system of benzene-cyclohexane. For the

Y. Tani




H O Y 8hlklmata








H.<f O





..... HO-",




2-8ucclnyH-h clroxr2,4-cyctohexa
- -

2 Z




, _ I.opentenyt-PPI


, _ (n-3)xl.opentenyI-PPI




Fig. 3. Biosynthesis of menaquinone.

reversed-phase chromatography, the thin-layer plate was immersed gently in

5% liquid n-paraffin in n-hexane and stood horizontally in the air to allow
evaporation of n-hexane. The reversed-phase chromatography was performed
with development using a solvent system of acetone-water to which a few
drops of liquid n-paraffin had been added. The MK band was detected under

Vitamin K2 and Vitamin Kl Production


The distribution of isoprenoid quinones was surveyed in about 900 microbial

strains (Tani et al., 1984). The quinone system of molds was tested in 56 strains
of 17 genera representing phycomycetes, ascomycetes and fungi imperfecti. All
of them lacked MK; their quinone system was coenzyme Q (CoQ). The same
result was obtained with 88 strains of yeasts of 32 genera representing
ascosporogenous, asporogenous and ballistosporogenous yeasts, and with 28
strains of Basidiomycetes of 14 genera.
MK was found to be widely distributed in procaryotes. About half of the 112
type strains of bacteria of 27 genera contained MK. The quinone types of the
bacteria tested could be divided into three groups: MK alone, MK and CoQ,
and CoQ alone. MK was found in Gram-positive bacteria, MK and CoQ in
Gram-negative facultative anaerobes, and CoQ in Gram-negative aerobes. The
only exception to this rule were strains of the genus Flavobacterium, which
were Gram-negative facultative anaerobes but had MK mostly or CoQ.
Flavobacterium species showed a higher content of MK than other bacteria.
Therefore, 550 strains of yellow-pigmented bacteria were further isolated.
About 90% of the isolates were MK producers. All actinomycetes tested (40
strains of 6 genera), contained only MK. This coincides well with the
classification of actinomycetes as Gram-positive bacteria.
Actinomycetes, though accounting for a high proportion of potent MK
producers, yield MK with eight or more isoprene units in the C-3 polyprenyl
side chain. In the blood coagulation tests, MK-4, 5 and 6 were shown to have
high biological activity, which decreased with the increase in the number of
isoprenoid units. Consequently, Flavobacterium species was selected to further
study for the improvement of the productivity.

Menaquinone-6 Production by Wild Type and Mutant Strains of

Flavobacterium meningosepticum

Culture conditions for MK-6 production by the wild type strain of F.

meningosepticum were investigated by Tani et al. (1984). The principal cultural
conditions that can be considered as influencing MK production are carbon
source for growth and the oxygen supply to the medium. Among carbohydrates, ribose is closely related to the biosynthesis of MK, sharing the
intermediate of the biosynthesis of aromatic amino acids. In fact, the cellular
content of MK increased when ribose was used in place of glycerol as the
carbon source; but the amount of MK in the culture broth decreased due to
the low cell concentration. When glycerol, glucose or fructose was used as the
carbon source, the growth and the productivity of MK were relatively high.
The cellular content of MK increased and the cell yield decreased with
decrease of aeration.
Prenyl alcohols are involved in MK biosynthesis in the form of pyrophosphates, which are incorporated into the polyprenyl side chain. Isopentenol was
the most effective in increasing the cellular content of MK, followed by
dimethylallyl alcohol.


Y. Tani

Precursors of the naphthoquinone nucleus, shikimate and 0succinylbenzoate, increased the cellular content of MK to some extent.
1-Hydroxy-2-naphthoate (HNA) inhibited the biosynthesis of MK but had
little effect on growth. This suggests that the compound is a specific inhibitor of
the biosynthesis as an analog of 1,4-dihydroxy-2-naphthoate.
The biosynthetic pathways of MK and aromatic amino acids branch at
chorismate after shikimate. The addition of L-tyrosine or its precursor,
p-hydroxyphenylpyruvate, increased the cellular content of MK; however,
L-tryptophan, L-phenylalanine and their precursors, anthranilate and
phenylpyruvate, inhibited the MK production.
Compounds which were found to increase MK production in the experiments described above were examined in various combinations for a possible
synergistic effect. Isopentenol and L-tyrosine formed the best couple. When
isopentenol was fed intermittently to the medium, the amount of MK reached
213 mg per liter of culture broth and 352 mg per g of dry cells after 3 days of
The MK productivity of F. meningosepticum was improved by mutagenization with N-methyl-N'-nitro-N-nitrosoguanidine. A mutant strain resistant to
HNA, which was found to be an inhibitor of MK biosynthesis but not of
growth, produced MK more abundantly: 34 mg per liter of culture broth and
55 mg per g of dry cells (Tani et al., 1986). The mutant strain was less
sensitive to inhibition by HNA on MK biosynthesis than the wild-type strain.
Mutant strains, which displayed KCN resistance, aromatic amino acid
auxotrophy and no-carotenoid productivity, did not show further increase of
MK productivity.
Isolation of MK-6 from cells of the mutant strain was performed as follows:
to a cell paste weighing about 100 g, 1 liter of acetone-ethyl ether was added
and agitated in a mini-jar fermentor for 2 h. Solid materials were filtered and
extracted again in a similar manner. Combined extracts were evaporated under
reduced pressure to less than 10% in volume. Each 10 ml of the concentrate
was partitioned between 500 ml of diethyl ether and 500 ml of sodium
chloride-saturated water. 'the non-polar lipids fraction obtained was dried with
anhydrous sodium sulfate and then evaporated. Resultant oily materials were
dissolved in chloroform and applied on a thin-layer plate which was prepared
by spreading a silica gel suspension (40 g/80 ml H 2 0) on glass plates at a
thickness of 075 mm, followed by activation at 110C for 2 h after drying
overnight at room temperature. Thin-layer chromatography was performed
with development with a solvent system of cyclohexane-benzene. The MK
band was scraped. MK was eluted with chloroform and crystallized twice from
ethanol. Crystalline MK (2204 mg) was obtained from cells of the mutant
strain at a yield of 75%. MK loss occurred in the recrystallization step.

Menaquinone-5 Production by a Mutant Strain of Flavobacterium


The diversity of the types of isoprenoid quinones present in micro-organisms

has been extensively studied as a result of the recent improvement of analytical

Vitamin K2 and Vitamin Kl Production


tools. Different types of MK homo logs were dectected in archaebacteria,

mycoplasmas, and in Gram-negative and Gram-positive bacteria with or
without minor component(s) in relation to taxonomic groupings. However, a
change of the type of MK in the same organism has never been noticed.
It was found that an HNA-resistant mutant, strain HNA 12-D, which was
derived from an MK-6 producer, produced an increased amount of MK, due to
the production of another type of MK in addition to MK-6 (Tani et al., 1985).
The newly formed MK was isolated and identified as MK-5. By combined
feeding of isopentenol and L-tyrosine, the total amount of MK produced by
strain HNA 12-D was 556 mg per liter of culture broth and 919 mg per g of
dry cells in the ratio of MK-6 and MK-5 of 1: 1 7 after 72 h cultivation,
whereas the parent strain produced 253 mg per liter of culture broth and
414 mg per g dry cells of MK-6.
The MK system of the wild type strain of the bacterium was composed only
of MK-6 without any minor components. The quantity of MK-5 produced by
the mutant strain was one to two times as much as that of the MK-6. It was the
first observation of a change in the MK pattern within the same microorganism. HNA is a structural analog of 1,4-dihydroxy-2-naphthoate, an
intermediate of MK biosynthesis. Whether the change in the specificity of the
enzyme, 1,4-dihydroxy-naphthoate: polyprenyl pyrophosphate transferase,
which catalyzes the prenylation of the naphthalenic intermediate using both
solanesyl pyrophosphate and octaprenyl pyrophosphate as substrates, or that
in the supply of polyprenyl pyrophosphates in strain HNA 12-D extended the
MK spectrum by the acquisition of HNA resistance remains to be determined.

Menaquinone-4 Production by a Mutant Strain of Flavobacterium


During further study to develop a new fermentation process for MK

production, a HNA-resistant mutant, strain HNA 250-15, of Flavobacterium
sp. 238-7, which produced MK-6 as the major component of respiratory
quinone with some minor components, showed an increase in MK productivity
due to increased production of a minor component of MK (Tani & Sakurai,
1987). The newly-formed MK was identified as MK-4, which now finds a
clinical use and which is made by a chemical process.
The MK-4 productivity of strain HNA 250-15 was further improved by
making the mutant resistant to usnic acid and menadione. Flavobacterium sp.
238-7 contains MK as the sole quinone in the respiratory chain. Among the
uncouplers and respiratory inhibitors tested, an uncoupler, usnic acid, strongly
inhibited the growth of strain HNA 250-15 at a low concentration. On
subsequent mutagenization of the strain, a mutant, USN r -2, which was
resistant to 4 mg of usnic acid per liter, was found to produce a much larger
amount of MK-4 than strain HNA 250-15, whereas the MK-6 content was
reduced. The productivity of MK-4 was increased more than three times as
compared with that of strain HNA 250-15.

Y. Tani


Growth of Flavobacterium was strictly inhibited by addition of menadione,

an artificial derivative of vitamin K, to the culture medium, and the content of
MK was reduced. Derivation of menadione-resistant mutant strains was then
carried out for enhancement of the MK productivity using strain USN r -2 as
the parent strain. Among the mutant strains obtained, strain K 3 -15, which was
resistant to 20 mg of menadione per liter, showed the highest content of MK-4
and MK-6. The total MK productivity was increased about three times as
compared with that of the wild type strain.
The cultural conditions for MK production were investigated with strain
Kr15. Glycerol and peptone were the most effective carbon and nitrogen
sources, respectively, both for growth and MK production. The amount of
MK-4 increased linearly in proportion with the glycerol concentration,
although the amount of MK-6 hardly changed with increasing glycerol
A variety of natural oils were supplemented to the culture medium at the
final concentration of 005% directly before autoclaving. Among them, cedar
wood oil decreased the growth but significantly increased the MK production.
















8 01)







.. ,.





4 !D



Fig. 4. Improvement of menaquinone production by Flavobacterium sp. 238-7. The left

and right bars indicate the amount of MK in mg/liter of culture broth and mg/g of dry
celis, respectively, the open and shaded bars denoting MK-6 and MK-4, respectively.
The amounts of MK produced by each strain indicated are the values when grown on
the basal medium, except that K3-1sa was grown in the optimized medium and K3-1S b in
the optimized medium supplemented with cedar wood oil.

Vitamin K2 and Vitamin K) Production


MK-4 production increased with increasing cedar wood oil concentration,

whereas the amount of MK-6 decreased. None of the prenyl alcohols increased
the MK production of strain K3-15 at concentrations of 05 and 1 mM, in fact
the amount of MK-4 decreased. Among L-tryptophan, L-phenylalanine,
L-tyrosine, anthranilate and phenylpyruvate, only phenylpyruvate somewhat
increased the MK-4 production in contrast with the case of F. meningosepticum.
The amount of MK produced by strain KT 15 was 1254 mg per liter of
culture broth and 128 mg per g of dry cells, in the ratio of MK-4 and MK-6 of
6: 1, under the optimal culture conditions in the presence of cedar wood oil.
The improvement of MK production by Flavobacterium sp. 238-7 on
mutagenization and optimization of the cultural conditions is summarized in
Fig. 4. The MK productivity of strain KT 15 was almost 10 times that of the
wild type strain. It is noteworthy that the increase of MK was due to an
increase in MK-4 but not MK-6. MK-4 was a minor component of the wild
type strain. The physiological role of MK-4 in the mutant strain is still
uncertain. In the course of cultivation, the cellular content of MK-4 still
increased after the growth had reached the stationary phase.

Extracellular Production of Menaquinone-4 by a Mutant Strain of

Flavobacterium sp. 238-7

It could be considered that the MK productivity of strain K3-15 described in

the previous section, was restricted by the intracellular level causing metabolic
repression of MK biosynthesis. During attempts to further improve the MK
production of the mutant strain, cultivation in a detergent-supplemented
medium resulted in the extracellular production of MK, especially MK-4,
without inhibition of the MK productivity and growth (Tani & Taguchi, 1988).
It was found that a mutant strain, K3-15, of Flavobacterium sp. 238-7
excreted MK into the culture medium supplemented with a detergent. Figure 5
shows the production of MK-4 and MK-6 during the cultivation of strain K3-15
in the basal medium. MK-6, which was the major MK component of the wild
type strain, was produced only in the cells during the cultivation. On the other
hand, MK-4 was also detected in the culture fluid.
A number of detergents (174 in total, i.e. the 110 nonionics, 30 anionics, 21
cationics and 13 amphoterics) were tested as to their induction of MK
excretion into the culture medium. Most of the non-ionic and amphoteric
detergents did not affect the cell growth. On the contrary, many anionic and
cationic detergents inhibited cell growth. A non-ionic detergent, Rikanon VA
5012 (polyoxyethylene oleyl ether), showed highest ability as to induction of
MK excretion into the medium without inhibiting the cell growth. MK-4,
which was newly formed in cells of the mutant strain, was selectively excreted
into the medium, but MK-6, which was the original MK of the wild type strain,
was hardly excreted. In the presence of 005% Rikanon VA 5012, the MK-4

Y. Tani

















CultIvatIon tIme (h)

Fig. 5. Menaquinone production by strain K3-15. 0, growth, .. , intracellular MK-6;

V, extracellular MK-6; e, intracellular MK-4; 0, extracellular MK-4; . , total MK-4.












CultlvaUon time (h)

Fig. 6. Menaquinone production by strain K3-15 in the presence of Rikanon UA 5012.

The symbols are the same as in Fig. 5.

Vitamin K2 and Vitamin Kl Production


productivity increased about 20% compared to that without the detergent, and
the total amount of MK was 39 mg per liter of culture broth.
Figure 6 shows the fermentation course of MK production on cultivation in
the presence of 005% Rikanon VA 5012. The cell mass reached a maximum
level, about 10 g by cell weight per liter of culture broth, after 30 h cultivation,
and then gradually decreased. MK-6 was mainly produced intracellularly and
reached about 12 mg per liter of culture broth after 30 h cultivation. A small
amount of MK-6 was detected in the culture filtrate at the late stage of the
cultivation. On the other hand, the excretion of MK-4 occurred from the
beginning of the cultivation and increased in the later logarithmic phase. The
extracellular production of MK-4 reached its maximum after 48 h cultivation.
The level of intracellular MK-4 was low but increased with increasing
cultivation time. The total amount of MK-4 reached about 25 mg per liter of
culture broth at 48 h cultivation, and the amount of total MK was 39 mg liter.
MK-4 excretion occurred from the onset of the cultivation and the excreted
MK was only MK-4. These facts showed that the excretion of MK-4 was not
due simply to cell lysis. The MK-4leakage might be due to a change in the cell
membrane structure, as a result of exposure to the detergent during growth.
The role of MK-4 in the mutant strain remains unknown, it might be that
MK-4 is not related to the respiratory function, since the growth and growth
rate of strain K3-15 did not decrease when most of the MK-4 had been


Natural vitamin Kh [(7'R,1l'R)-phylloquinone] consists of 2-methyl-1,4naphthoquinone and the optically active phytyl side chain which is attached to
position 3. (3R,7R)-Hexahydrofarnesol, constructed from either of the microbiologically generated optically active lactones [(S)-2-methyl-y-butyrolactone
or (S)-3-methyl-y-butyrolactone] (see chapter 6, p. 102) can be lengthened by
condensation with a trans-Cs-olefin to (7R, llR)-trans-phytol and yields
natural vitamin Kl (Fig. 7).

Fig. 7. Absolute configurations of (7R,1l'R)-phylloquinone.


Y. Tani

Bentley, R. & Meganathan, R. (1982). Biosynthesis of vitamin K (menaquinone) in
bacteria. Microbiol. Rev., 46,241-80.
Collins, M. D. & Jones, D. (1981). Distribution of isoprenoid quinone structural types
in bacteria and their taxonomical implications. Microbiol. Rev., 45, 316-50.
Tani, Y. & Sakurai, N. (1987). Menaquinone-4 production by a mutant of
Flavobacterium sp. 238-7. Agric. BioI. Chem., 51,2409-15.
Tani, Y. & Taguchi, H. (1988). Excretion of menaquinone-4 by a mutant of
Flavobacterium sp. 238-7 in a detergent-supplemented culture. Agric. Bioi. Chem.,
Tani, Y., Asahi, S. & Yamada, H. (1984). Vitamin K2 (menaquinone):screening of
producing microorganisms and production by Flavobacterium meningosepticum. J.
Ferment. Technol., 62,321-7.
Tani, Y., Asahi, S. & Yamada, H. (1985). Production of manaquinone (vitamin K 2 )-5
by a hydroxynaphthoate-resistant mutant derived from Flavobacterium
meningosepticum, a menaquinone-6 producer. Agric. Bioi. Chem., 49, 111-15.
Tani, Y., Asahi, S. & Yamada, H. (1986). Menaquinone (vitamin K 2 )-6 production by
mutants of Flavobacterium meningosepticum, J. Nutr. Sci. Vitaminol., 32, 137-45.


Chapter 9


Department of Biochemistry, Kyoto Prefectural University of Medicine,
Kyoto, Japan

Thiamine was discovered in the course of a search for an agent that would cure
beriberi. The conquest of beriberi began in 1885 when Takaki (1885)
practically eradicated the disease among the Japanese navy by introducing fish,
vegetables, meat and barley into the diet. In 1897 Eijkman (1897) showed that
an experimental polyneuritis in fowl, which closely resembled the polyneuritic
symptoms of beriberi, could be produced by feeding the birds on a diet of
polished rice. When they were fed on unpolished rice they did not develop the
disease. In 1926 Jansen & Donath (1926) isolated a crystalline hydrochloride of
the antineuritic factor from rice bran.
The structure of thiamine, elucidated in the mid-1930s, was established
by Williams (1936) as 3-(4-amino-2-methyl-5-pyridimidinylmethyl)-5-(tJhydroxyethyl}-4-methylthiazolium chloride hydrochloride (Fig. 1). The name
thiamine derives from the chemical nature of the vitamin, in that it has the
thiazole (sulfur-containing) ring attached to a pyrimidine ring with an amine
group. Thiamine was first synthesized in 1936 by Williams & Cline (1936) from
the pyrimidine and thiazole moieties of thiamine. Within a year, Lohmann &
Schuster (1937) isolated thiamine pyrophosphate (TPP) from yeast and showed
that it was the cofactor for the decarboxylation of pyruvic acid. In addition to
thiamine and TPP, thiamine monophosphate (TMP) and thiamine triphosphate
(TIP) have been found in living organisms. These phosphate esters of
thiamine are also shown in Fig. 1.


The double-salt from thiamine with hydrochloric acid (C 12H 17N 4 0SCI HCI;
molecular weight, 33728) consists of monoclinic plates in rosette-like clusters


A. Iwashima


CH2~~ 4' CH,CH,OR










ClH elH




- CHi'

6H 6H

~ C,H,O-P-O-P-OH


Hydroxyethylthlamlne pyrophosphate

Fig. 1. Structures of thiamine and its related compounds.

with slight thiazole odor (Merck Index, 1983). It is soluble in water and
methanol, slightly soluble in ethanol, and practically insoluble in ether,
benzene and chloroform. In aqueous solution thiamine is most stable between
pH 2 and 4, but unstable at alkaline pH (Yurugi et aI., 1979). It is
destroyed by heat if the pH of the solution is above 55: the rate of destruction
increases with the pH. Under dry conditions thiamine is stable to heating at
100C for 24 h. The absorption maximum of thiamine in the UV light is
affected by pH, and it is 243 nm at pH 23 (the molar absorption coefficient:
106 X 103 ), and 235 and 265 nm at pH 74 (Yurugi et al., 1979).
TPP chloride (commercially available in the dried form) is stable when
stored cold in the dark (Yurugi, 1975). In aqueous solution it is stable at
pH 2-6 and OC for 6 months, but it partially decomposes to TMP and
thiamine when allowed to stand for several months at pH 5 and 38C. TIP is
stable under dry conditions, whereas it is unstable in aqueous solution,
especially to heat, and decomposes to TMP and TPP. In alkaline solution TIP
decomposes to TMP and inorganic phosphate. TMP, which is obtained by
hydrolysis of TIP and TPP, is quite stable at acidic and neutral pH. Thiamine
and thiamine phosphate esters are quantitatively converted to thiochrome (Fig.
1) and its phosphate esters which are compounds exhibiting intense blue
fluorescence. This property observed in early studies of thiamine oxidation
serves as the basis for the chemical assay for thiamine (Barger et al; 1935).
Hydroxyethylthiamine pyrophosphate (Fig. 1), a TPP-activated aldehyde
intermediate of the enzyme reactions, it is also known to be present in
micro-organisms and mammalian tissues, especially in muscles (Morita et al.,


Thiamine, as well as other vitamins, is produced by micro-organisms in

extremely small amounts. Usually it is not formed in great excess over the

Microbial Synthesis of Vitamin B 1


organism's own requirement. The production of thiamine by growing cultures

of Escherichia coli ATCC 9637 is 16 and 14 higher than the amount of
thiamine in cells grown in the basal medium (21-23 ng/mg dry weight), when
cysteine and methionine are added at 1 mg per ml of the minimal medium,
respectively (Akagi & Kumaoka, 1963). The addition of some aromatic amino
acids is also stimulatory on thiamine production by growing cultures of the
same organisms: phenylalanine and histidine cause 15- and 12-fold stimulations of thiamine synthesis at 10 and 1 mM, respectively (Iwashima et al.,
1968). Methionine at 1 mM shows an additional effect of 26-fold stimulation of
the thiamine production in the presence of 10 mM phenylalanine. On the other
hand, the addition of purine and pyrimidine bases of nucleic acids to the
growing cultures of E. coli is not effective on thiamine production in the cells.
Thus, the supplement in the growth medium is not very effective to produce
excess thiamine in bacteria and the de novo synthesis of thiamine cannot be
demonstrated in non-growing cells under normal circumstances, because the
control mechanism of thiamine biosynthesis in micro-organisms severely limits
thiamine overproduction in the cells. However, this difficulty can be overcome
by the use of a system that removes the normal control of thiamine over its
biosynthesis, so that the organisms form excess thiamine. Using Salmonella
typhimurium, Newell & Tucker (1966a,b) observed a burst of thiamine
synthesis ensued, increasing the cellular level of thiamine four- to fivefold
(about 200 ng/mg dry weight) when cells were preincubated with adenosine,
washed, and reincubated in minimal medium. Adenosine inhibits the biosynthesis of the pyrimidine moiety of thiamine, and the growth of organisms in the
presence of adenosine therefore lowers the cellular concentration of thiamine.
The normal repressible control by thiamine is thereby lost, and the thiaminesynthesizing enzymes become de repressed to overproduce thiamine. The
thiamine formed occurs almost completely in the form of TPP and is located
Mutants affecting regulation of thiamine biosynthesis from the pyrimidine
and thiazole moieties were isolated from E. coli K12 as resistant strains to
growth inhibition by pyrithiamine, a thiamine antagonist (Kawasaki and Nose,
1969). Two mutants, PT-R1 and PT-R3, have approximately a threefold higher
cellular thiamine content (90-100 ng/mg dry weight) than the parent strain. Yeast
as well as bacteria do not produce large amounts of thiamine under normal
circumstances (Stieglitz et al., 1974). Recently, however, thiamine-excreting
mutants of yeast have been isolated (Silhankova, 1985a). Two mutants of S.
cerevisiae and four mutants of S. carlsbergensis, which excreted up to 920/-tg
thiamine per liter of the culture medium, were obtained by employing several
rounds of UV mutagenesis and selection. The genetic characterization of these
mutants suggested that thiamine excretion phenotype is recessive and under
the control of nuclear genes (Silhankova, 1985b).
The most effective method for the production of thiamine by microorganisms is an enzymatic conjugation of 2-methyl-4-amino-5-hydroxymethylpyrimidine (hydroxymethylpyrimidine) and 4-methyl-5-P-hydroxyethylthiazole
(hydroxyethylthiazole), the preformed pyrimidine and thiazole moieties of



thiamine. Washed cell suspensions of E. coli ATCC 9637 (15-20mg dry cells)
synthesizes 210 ng of thiamine per mg dry weight from 10 IlM each of the two
moieties in the presence of 02% glucose at pH 7 for 1 h at 37C (Iwashima et
al., 1968). With the cells derepressed thiamine biosynthesis by adenine or
thiamine auxotrophs grown in the presence of limited amount of thiamine the
production of thiamine is much more enhanced (Kawasaki et al., 1969). The
cells suspension of baker's yeast also has the capability of joining the
pyrimidine and thiazole moieties of thiamine to form thiamine in the presence
of glucose (Ashida, 1942). In S. cerevisiae, thiamine synthesis from both
moieties added leads to 20-40 times higher contents of intracellular thiamine
in comparison with normal growth conditions (Silhankova, 1985b).


The biosynthesis of thiamine involves the independent formation of the two

ring structures and their subsequent condensation (Fig. 2). The enzymatic
steps involved in the conversion of the pyrimidine and thiazole moieties to
thiamine and TPP were elucidated some years ago (Leder, 1975). Although
the biosynthetic pathway for the formation of hydroxymethylpyrimidine has
not yet been completely elucidated, enough is known to eliminate the
possibility that this pyrimidine and the pyrimidines of nucleic acids do not
share a common biosynthetic pathway (Goldstein & Brown, 1963). An
important contribution to the understanding of the biosynthesis of the
pyrimidine moiety of thiamine in bacteria was the discovery by Newell &
Tucker (1968a,b) of mutants of S. typhimurium with a dual growth requirement for purines as well as for hydroxymethylpyrimidine. The dual growth
requirement was satisfied by 5-aminoimidazole ribonucleotide (AIR) alone and


r I









r- I



6H 6H






Fig. 2. The biosynthesis of thiamine from its pyrimidine and thiazole moieties.

Microbial Synthesis of Vitamin B 1


by compounds that precede AIR on the purine pathway, indicating that AIR is
the last common intermediate on the route to the purines and to the
pyrimidine moiety of thiamine. The biosynthetic pathways of hydroxymethylpyrimidine in micro-organisms differ between bacteria and yeast. In
bacteria, hydroxymethylpyrimidine is synthesized from AIR, and N-1, C-4 and
C-6 of the pyrimidine are derived from the nitrogen, C-1 and C-2 of glycine,
respectively (Estramareix & Lesieu, 1969; Estramareix, 1970; White &
Rudolph, 1979), and the C-2 from formate (Kumaoka & Brown, 1967;
Estramareix & Lesieu, 1969). These results suggest that the imidazole ring is
opened between C-4 and C-S. The remaining three carbon atoms of hydroxymethylpyrimidine (C-S, C-7 and C-8) originate from the ribose part of AIR
which is thus the precursor of all the carbon atoms of this pyrimidine
(Estramareix & Therisod, 1984). Since N-3 and the amino group at C-4 have
been recently shown to be derived from the amide-N atom from glutamine in
E. coli and S. cerevisiae (Tazuya et ai., 1987a), the origin of almost all of the
pyrimidine moiety of thiamine is now known for bacteria, although the
reactions whereby AIR is converted to the pyrimidine remain unknown. In
yeast, on the other hand, C-4 of the pyrimidine originates from formate (David
et ai., 1966) and no significant incorporation of the nitrogen atom and C-2 of
glycine into the pyrimidine is found (Linnett & Walker, 1968; White &
Spenser, 1979). Furthermore, the incorporation of C-3, C-4 and C-S of
pentulose into C-6, C-S and C-7 of the pyrimidine has been recently reported
in S. cerevisiae (Grue-S0rensen et ai., 1986). The origin of C-2, C-8 and N-1 of
hydroxymethylpyrimidine remains to be clarified in the biogenesis of the
pyrimidine moiety in yeast.
Relatively little has been discovered regarding the biosynthetic origins of the
thiazole moiety of thiamine, but some progress has been made in recent years
toward the identification of its precursors. In E. coli and S. typhimurium the
direct utilization of C-2 and the nitrogen atom of tyrosine for the formation of
the thiazole ring was demonstrated (Estramareix & Therisod, 1972; Bellion et
ai., 1976; White & Rudolph, 1978). On the other hand, no incorporation of
C-2 of tyrosine into thiamine is observed in yeast, whereas significant
incorporations of C-2 and the nitrogen atom of glycine into C-2 and N of the
thiazole are shown (Linnett & Walker, 1968, 1969). These observations seem
to indicate that bacteria and yeast use different amino acids as precursors of
the C-2 and nitrogen portion of the thiazole ring, implying that the pathways in
these two micro-organisms are different. Further incorporation studies with
deuterated carbohydrates and related compounds showed that the precursor of
the S-carbon chain (C-4, C-S, C-6, C-7 and C-8) of the thiazole moiety of
thiamine in E. coli is a S-carbon sugar derived from pyruvate and triose
phosphate (White, 1978), which might then react with tyrosine and a sulfur
compound in an undefined series of steps to yield hydroxyethylthiazole. This
has been supported by the incorporation of 1-deoxY-D-threo-pentulose into the
S-carbon chain of the thiazole without carbon-carbon bond cleavage in E. coli
(David et ai., 1981). In yeast, the incorporation pattern leads to the inference



that the 5-carbon chain may not be derived from pyruvate but from
2-pentulose, which is generated from the hexose precursors by oxidative as
well as non-oxidative pentose phosphate pathway (White and Spenser, 1982).
The precursor of the sulfur atom of the thiazole ring is still not known with
certainty, but several experimental observations suggest that either cysteine or
H 2 S which could be produced from cysteine, is the most likely precursor of the
sulfur atom of hydroxyethylthiazole in bacteria and yeast (Bellion & Kirkley,
1977; Tazuya et al.; 1987b).

4.2 Regulation
As described above, the derepression of thiamine biosynthesis by adenosine
was demonstrated for the first time in S. typhimurium (Newell & Tucker,
1966a,b). Adenine and adenosine, in the form of AMP, exert feedback
control over the synthesis of AIR by inhibiting the formation of 5phosphoribosylamine, the first step exclusive to the synthesis of purines and
hydroxymethylpyrimidine. The derepression is unaffected by hydroxyethylthiazole, but does not occur if thiamine or hydroxymethylpyrimidine is
present, or if protein synthesis is prevented during preincubation. Since
derepression by adenosine in a hydroxyethylthiazole auxotroph is prevented by
thiamine, but not by hydroxymethylpyrimidine, it was suggested that hydroxymethylpyrimidine influences the state of repression only by being converted to
thiamine. Derepression in the absence of adenosine also occurs when a
hydroxyethylthiazole auxotroph is grown on a limited supply of thiamine.
These studies were then extended to the regulations of the four enzymes in E.
coli that catalyze the synthesis of TMP from hydroxymethylpyrimidine and
hydroxyethylthiazole: hydroxymethylpyrimidine kinase (EC, phosphomethylpyrimidine kinase (EC, hydroxyethylthiazole kinase (EC and thiamine-phosphate pyrophosphorylase (EC (Kawasaki
et al., 1969). These enzymes are derepressed by adenine and repressed by
thiamine. The derepression is also brought about by preincubating the
growing culture with phenylalanine (Kawasaki et al., 1969). This derepression
is reversed by thiamine or hydroxyethylthiazole, but not by hydroxymethylpyrimidine. Since thiamine synthesis is inhibited by phenylalanine in a thiamine
regulatory mutant (PT-R1) and the inhibition is reversed by hydroxyethylthiazole (Iwashima & Nose, 1970), this inhibition by phenylalanine
appears to account for the derepression in the wild type. However, two-fold
increase in cellular thiamine levels accompanied by the inhibition of the
thiazole synthesis caused by phenylalanine remains unexplained. In a thiamine
regulatory mutant (PT-R1) hydroxyethylthiazole kinase and thiaminephosphate pyrophosphorylase are not subjected to repression by thiamine,
suggesting that it is a constitutive mutant of an operator gene controlling the
synthesis of two enzymes (Kawasaki & Nose, 1969). Several thiamine
regulatory mutants of E. coli affecting other gene loci have been reported
(Kawasaki et al., 1976).

Microbial Synthesis of Vitamin B\


5.1 Microbiological Assays
Assay methods based on use of the thiamine requiring fungus Phycomyces
blakesleeanus and thiamine requiring strains of several bacteria such as
Staphylococcus aureus, E. coli and Lactobacillus species, and S. cerevisiae
have been used (Thomas, 1966).
5.2 Chemical Assays
Chemical assay methods are usually preferred to microbiological methods since
such analyses can be performed rapidly and they are usually more reliable for
routine determination. Although a number of chemical assay methods for
thiamine are known, only the coupling reaction with diazotized paminoacetophenone (Melnick & Field, 1937) and thiochrome reaction have
been developed as precise methods of assay. In recent years, however, the
thiochrome method has been used almost exclusively. As described above,
thiamine in alkaline solution is oxidised quantitatively to thiochrome. As
oxidizing reagents potassium ferricyanide (Hennessy & Cerecedo, 1939) and
cyanogen bromide (Fujiwara & Matsui, 1953) have been routinely used.
Thiochrome formed from non-phosphorylated thiamine is extracted with
isobutanol. Since thiochrome fluoresces intensely under UV illumination, it
can be measured readily in a fluorimeter (excitation 365 nm; emission 430 nm).
Phosphorylated thiamines can be determined after their hydrolysis to thiamine
by phosphatase. Hydroxyethylthiamine gives thiochrome by oxidation with
alkaline ferricyanide, but not with cyanogen bromide. Thiamine gives thiochrome by either oxidizing agent, so that this difference in the oxidation
property is used for the simultaneous determination of thiamine and hydroxyethylthiamine (Morita et al., 1968). For the differential determination of
thiamine and its phosphate esters in biological materials electrophoresis, paper
chromatography, thin-layer chromatography and ion-exchange chromatography can be used to separate thiamine compounds from each other (Yurugi et
al., 1979). In recent years, high-performance liquid chromatography (HPLC)
has been developed for simultaneous determination of thiamine and its
phosphate esters and assay procedures have been established (Kawasaki &
Sanemori, 1985). The oxidation of thiamine compounds in samples can be
carried out either before the chromatography (precolumn derivatization
procedure) or after the chromatography (postcolumn derivatization procedure). Therefore, the precolumn derivatization procedure is essentially the
HPLC of thiochrome and its phosphate esters. A mixing coil with proportioning pump as an additional equipment for thiochrome reaction is required for the
postcolumn derivatization procedure. The intensity of the fluorescent products
is measured with a spectrophotofluorometer. The minimum detection by these
methods has been reported to be 01 pmol or 337 pg as thiamine hydrochloride (Sanemori et al., 1980).




Enzyme Assays

TPP has been estimated manometrically by measuring the evolution of CO2

from pyruvate in the presence of the apoenzyme of yeast pyruvate decarboxylase (Green et al., 1941), and spectrophotometrically in the same reaction
system by measuring acetaldehyde-dependent reduction of NADH2 using
alcohol dehydrogenase (Kajiro, 1957).


6.1 Metabolism
It has been generally accepted that thiamine transport in micro-organisms

occurs via active transport (Iwashima, 1980). In animal cells, evidence has
been accumulated which shows that thiamine is taken up by Na+-dependent
active process (Iwashima, 1980). In both animals and humans thiamine
transport across the small intestine is biphasic: mainly by active transport at
low or physiological concentrations 2 IlM) and predominantly passive at
higher concentrations (Hoyumpa et al., 1975).
Thiamine must be phosphorylated to TPP, a coenzyme form of thiamine, to
exert its physiological functions. The formation of TPP from thiamine is
catalyzed by thiamine pyrophosphokinase (EC in yeast and animal
tissues. In E. coli thiamine is converted to TPP by two successive phosphorylations catalyzed by thiamine kinase (EC and thiamine-monophosphate
kinase (EC TPP is further metabolized to TIP by thiaminediphosphate kinase (EC
Two thiamine cleavage enzymes have been isolated from various natural
sources (Murata, 1982). Thiaminase I (EC is a transferase which
catalyzes an exchange reaction between a base and a thiazole moiety of
thiamine, and thiaminase II (EC mediates in the hydrolysis of

Physiological Functions

The physiological functions of thiamine is exerted in the form of TPP in cells.

TPP serves as the coenzyme for a large numbers of enzyme systems in the
metabolism of carbohydrates and amino acids, including pyruvate
dehydrogenase(cytochrome) (EC, pyruvate oxidase (EC,
pyruvate dehydrogenase(lipoamide) (EC, oxoglutarate dehydrogenase(lipoamide) (EC, 2-oxoisovalerate dehydrogenase(Iipoamide)
(EC, transketolase (EC, pyruvate decarboxylase (EC,
benzoyiformate decarboxylase (EC, oxalylCoA decarboxylase (EC, tartronate-semialdehyde synthase (EC, phosphoketolase (EC, 2-hydroxy-3-oxoadipate synthase (EC, acetolactate synthase
(EC and sulfoacetaldehyde lyase (EC

Microbial Synthesis of Vitamin B)


The physiological function of TIP remains enigmatic, although evidence is

emerging that this form of thiamine is involved in nerve membrane function
(Leder, 1975). It is certain that the neuropathy of beriberi is due to thiamine
deficiency and it is probable that the anorexia and cardiac manifestation of
beriberi are caused by decreased activity of the enzymes for which TPP is
coenzyme. However, much remains unclear about the cellular and molecular
mechanism of thiamine action correlated to thiamine deficiency symptoms.


Thiamine has been synthesized in two basic ways (Matsukawa et al., 1970).
The condensation method, first adapted to the synthesis of thiamine by
Williams & Cline (1936), consists of the condensation of the pyrimidine with
thiazole derivatives. In another method, which was devised by Todd & Bergel
(1937), the pyrimidine moiety is synthesized with an appropriate side chain
conductive to the formation of the remaining structure. Matsukawa (1953)
synthesized thiothiamine as an intermediate for thiamine synthesis. This is
formed by the condensation of 4-amino-5-aminomethyl-2-methylpyrimidine,
CS 2 and 3-acetyl-3-chloro-1-propanol. Thiothiamine is a water-insoluble crystalline compound and is readily converted by oxidation into thiamine in good
Akagi, M. & Kumaoka, H. (1963). Effect of sulfur-containing amino acids on the
production of thiamine by Escherichia coli. J. Vitaminol., 9, 183-7.
Ashida, K. (1942). Synthesis of vitamin B) by microbes. Bulletin of the Agricultural
Chemical Society of Japan., 18, 723-6.
Barger, G., Bergel, F. & Todd, A. R. (1935). Uber das Thiochrom aus Vitamin B)
(Antineurin). Chern. Ber., 68, 2257-62.
Bellion, E. & Kirkley, D. (1977). The origin of the sulfur atom in thiamine. Biochim.
biophys. Acta, 497,323-8.
Bellion, E., Kirkley, D. H. & Faust, J. R. (1976). The biosynthesis of the thiazole
moiety of thiamine in Salmonella typhimurium. Biochim. biophys. Acta, 437,
David, S., Estramareix, B. & Hirshfield, H. (1966). Le formate, precurseur du carbone
4 de la pyrimidine de la thiamine. Bichim. biophys. Acta, 127, 264-5.
David, S., Estramareix, B., Fischer, J-C. & Therisod, M. (1981). 1-Deoxy-o-threo-2pentulose, the precursor of the five-carbon chain of the thiazole of thiamine. J. Am.
Chern. Soc., 103,7341-2.
Eijkman, C. (1897). Eine Beriberiahnliche Krankheit der Huehner. Virchows Arch.
Pathol. Anat., 148, 523.
Estramareix, B. (1970). Biosynthese de la pyrimidine de la thiamine. Origine du
carbone 6 chez Salmonella typhimurium. Biochim. biophys. Acta, 208, 170-7l.
Estramareix, B. & Lesieu, M. (1969). Biosynthese de la pyrimidine de la thiamine.
Origine des carbones 2 et 4 chez Salmonella typhimurium. Biochim. biophys. Acta,



Estramareix, B. & Therisod, M. (1972). La tyrosine, facteur de la biosynthese du

thiazole de la thiamine chez Escherichia coli. Biochim. biophys. Acta, 273,275-82.
Estramareix, B. & Therisod, M. (1984). Biosynthesis of thiamine. 5-Aminoimidazole
ribotide as the precursor of all the carbon atoms of the pyrimidine moiety. J. Am.
Chem. Soc., 106,3857-60.
Fujiwara, M. & Matsui, K. (1953). Determination of thiamine by the thiochrome
reaction. Analyt. Chem., 25, 810-12.
Goldstein, G. A. & Brown, G. M. (1963). The biosynthesis of thiamine. V. Studies
concerning precursors of the pyrimidine moiety. Archs. Biochem. Biophys., 103,
Green, D. E., Herbert, D. & Subrahmanyan, V. (1941). Carboxylase. J. bioI. Chem.,
138, 327-39.
Grue-S(lJrensen, G., White, R. L. & Spenser, I. D. (1986). Thiamin biosynthesis in
Saccharomyces cerevisiae. Origin of the pyrimidine unit. J. Am. Chem. Soc., 108,
Hennessy, D. J. & Cerecedo, I. R. (1939). The determination of free and phosphorylated thiamine by a modified thiochrome assay. J. Am. Chem. Soc., 61, 179-83.
Hoyumpa, A. M., Middleton, H. M. III., Wilson, F. A. & Schenker, S. (1975).
Thiamine transport across the rat intestine. I. Normal characteristics.
Gastroenterology, 68, 1218-227.
Iwashima, A. (1980). Absorption and membrane transport of thiamine. In
Vitaminology, Vol. II, ed. Vitamin Society of Japan, Tokyokagakudojin, Tokyo,
Iwashima, A. & Nose, Y. (1970). Inhibition by phenylalanine of thiazole biosynthesis
in Escherichia coli. J. Bacteriol., 104, 1014-16.
Iwashima, A., Kawasaki, T. Nakamura, M. & Nose, Y. (1968). Effect of amino acids
and purine bases on thiamine synthesis by Escherichia coli. J. Vitaminol., 14,
Jansen, B. C. P. & Donath, W. F. (1926). On the isolation of antiberiberi vitamin.
Proc. Kon. Ned. Akad. Wet., 29, 1390.
Kajiro, Y. (1957). Quantitative determination of thiamine pyrophosphate using
apocarboxylase and alcohol dehydrogenase. J. Biochem., 44, 827-38.
Kawasaki, T. & Nose, Y. (1969). Thiamine regulatory mutants in Escherichia coli. J.
Biochem., 65,417-25.
Kawasaki, T. & Sanemori, H. (1985). Vitamin B 1: Thiamines. In Modern Chromatographic Analysis of the Vitamins, ed. M. G. M. De Ruyter. Marcel Dekker, New
York, pp. 385-411.
Kawasaki, T., Iwashima, A. & Nose, Y. (1969). Regulation of thiamine biosynthesis in
Escherichia coli. J. Biochem., 65,407-16.
Kawasaki, T., Sanemori, H., Egi, Y., Yoshida, S. & Yamada, K. (1976). Biochemical
studies on pyrithiamine-resistant mutants of Escherichia coli K12. J. Biochem., 79,
Kumaoka, H. & Brown, G. M. (1967). Biosynthesis of thiamine. VI. Incorporation of
formate into carbon atom two of the pyrimidine moiety of thiamine. Archs.
Biochem. Biophys., 122,378-84.
Leder, I. G. (1975). Thiamine, biosynthesis and function. In Metabolic Pathways, Vol.
7, ed. D. M. Greenberg, Academic Press, New York, pp. 57-85.
Linnett, P. E. & Walker, J. (1968). Biosynthesis of thiamine. Incorporation experiments with 14C-Iabelled substrates and with [lSN]glycine in Saccharomyces
cerevisiae. Biochem. J., 109, 16l-8.
Linnett, P. E. & Walker, J. (1969). Biosynthesis of thiamine. IV. C-2 of glycine as the
precursor of C-2 of the thiazole moiety in yeast. Biochim. biophys. Acta, 184,

Microbial Synthesis of Vitamin B \


Lohmann, K. & Schuster, Ph. (1937). Untersuchungen tiber die Cocarboxylase.

Biochem. Z., 294, 188-214.
Matsukawa, T. (1953). Chemistry of the thio-thiamine and its related compounds.
Vitamins (Kyoto), 6, 299-310.
Matsukawa, T., Hirano, H. & Yurugi, S. (1970). Preparation of thiamine derivatives
and analogs. Meth. Enzymol., 18A, 141-62.
Melnick, D. & Field, H. (1937). Chemical determination of vitamin B\. 1. Reactions
between thiamine in pure aqueous solution and diazotized p-aminoacetophenone. J.
bioi. Chem., 127,505-14.
Merck Index (1983), 10th edn., ed. M. Windholz. Merck & Co. Rahhway, p. 9134.
Morita, M., Nishibe, Y. & Mineshita, T. (1968). Natural occurrence of 2-(1hydroxyethyl)thiamine in muscular tissues of animals. J. Vitaminol., 14, 230-38.
Murata, K. (1982). Actions of two types of thiaminase on thiamine and its analogues.
Ann. N. Y. Acad. Sci., 378, 146-56.
Newell, P. C. & Tucker, R. G. (1966a). The de-repression of thiamine biosynthesis by
adenosine. A tool for investigating this biosynthetic pathway. Biochem. J., 100,
Newell, P. C. & Tucker, R. G. (1966b). The control mechanism of thiamine
biosynthesis. A model for the study of control of converging pathways. Biochem. J.,
100, 517-24.
Newell, P. C. & Tucker, R. G. (1968a). Precursors of the pyrimidine moiety of
thiamine. Biochem. J., 106,271-77.
Newell, P. C. & Tucker, R. G. (1968b). Biosynthesis of the pyrimidine moiety of
thiamine. A new route of pyrimidine biosynthesis involving purine intermediates.
Biochem. J., 106, 279-87.
Sanemori, H., Ueki, H. & Kawasaki, T. (1980). Reversed-phase high-performance
liquid chromatographic analysis of thiamine phosphate esters at subpicomole levels.
Analyt. Biochem., 102,451-5.
Silhankova, L. (1985a). Yeast mutants excreting vitamin B. and their use in the
production of thiamine rich beers. J. Inst. Brew., 91, 78-81.
Silhankova, L. (1985b). Genetic control of thiamine excretion and of its suppression in
Saccharomyces cerevisiae. J. Inst. Brew., 91,238-41.
Stieglitz, B., Levy, R. & Mateles, R. I. (1974). Thiamine accumulation in yeast. J.
Appl. Chem. Biotechnol., 24, 277-82.
Takaki, K. (1885). On the cause and prevention of Kakke. Transactions of Sei-I-Kawi,
4, Suppl., 29-37.
Tazuya, K., Tanaka, M., Morisaki, M., Yamada, K. & Kumaoka, H. (1987a). Origin
of nitrogen atoms of the pyrimidine moiety of thiamin. Biochem. Int., 14,
Tazuya, K., Yamada, K., Nakamura, K. & Kumaoka, H. (1987b). The origin of the
sulfur atom of thiamin. Biochim. biophys. Acta, 924,210-15.
Thomas, M. H. (1966). Microbiological Assay Technics. In Methods of Vitamin Assay,
ed. Association of Vitamin Chemists. Interscience Publishers, New York, pp.
Todd, A. R. & Bergel. F. (1937). Aneurin. Part VII. A synthesis of aneurin. J. Chem.
Soc., 364-7.
White, R. H. (1978). Stable isotope studies on the biosynthesis of the thiazole moiety
of thiamin in Escherichia coli. Biochemistry, 17, 3833-40.
White, R. H. & Rudolph, F. (1978). The origin of the nitrogen atom in the thiazole
ring of thiamine in Escherichia coli. Biochim. biophys. Acta, 542,340-7.
White, R. H. & Rudolph, F. B. (1979). Biosynthesis of the pyrimidine moiety of
thiamine in Escherichia coli. Incorporations of stable isotope-labeled glycines.
Biochemistry, 18, 2632-6.



White, R. L. & Spenser, I. D. (1979). Thiamine biosynthesis in Saccharomyces

cerevisiae. Origin of carbon-2 of the thiazole moiety. Biochem. J., 179, 315-25.
White, R. L. & Spenser, I. D. (1982). Thiamin biosynthesis in yeast. Origin of the
five-carbon unit of the thiazole moiety. J. Am. Chem. Soc., 104,4934-43.
Williams, R. R. (1936). Structure of vitamin B 1 J. Am. Chem. Soc., 58, 1063-4.
Williams, R. R. & Cline, J. K. (1936). Synthesis of vitamin B 1 J. Am. Chem. Soc., 58,
Yurugi, S., Iwashima, A. & Nose, Y. (1979). Thiamine. In Data Book for
Biochemistry, Vol. 1., ed. Japanese Biochemical Society. Tokyokagakudojin,
Tokyo, pp. 1164-76.
Yurugi, S. (1975). Thiamine (Vitamin Bt). In Experimental Procedures in Biochemistry,
Vol. 13, Vitamins & Coenzymes, ed. Japanese Biochemical Society, Tokyokagakudojin, Tokyo, pp. 55-73.

Chapter 10


Hacettepe University, Chemical Engineering Department, 06532 Beytepe,
Ankara, Turkey

Riboflavin is a water-soluble vitamin which can be produced on a commercial
scale by various fermentation processes using various micro-organisms. Riboflavin is used pure for human nutrition and therapy and in the crude
concentrated form for animal feed supplements. Although it is produced by
both synthetic and fermentation processes, it is believed that the microbial
synthesis should be able to compete quite successfully with the synthetic
chemical processes.
Riboflavin, also called vitamin B2 (lactoflavine, ovoflavine, uroflavine, hepatoflavine, verdoflavine, vitamin G), is a water-soluble vitamin which exhibits a
strong yellowish-green fluorescence.
As early as 1879, A. W. Blyth isolated from whey an impure preparation of
flavine (lactochrome) of a red-orange color. In 1933, P. Gyorgy, R. Kuhn,
and T. Wagner-Jauregg described the isolation of pure, crystallized lactoflavine
and found the vitamin nature of this yellow pigment. The original name of
riboflavin was lactoflavine. Ovoflavine, isolated from eggs; hepatoflavine, from
liver; lactoflavine, from milk; and verdoflavine, from plants, were all identified
as riboflavin. The name riboflavin, indicating that the naturally occurring flavin
is a derivative of o-ribose, was adopted in 1952 by the International
Commission for the Reform of Biochemical Nomenclature (Kirk & Othmer,
1968; Wagner-Jauregg, 1972).
The structure of riboflavin was confirmed in 1934 by synthesis by R. Kuhn
and F. Weygand and P. Karrer and his co-workers. The chemical structure of
vitamin B2 is [6,7-dimethyl-9-(o-1'-ribityl)-isoalloxazine], as shown in Fig. l.

T. Kutsal & M. T. Ozbas


H3 C







~lOh 43NH

Fig. 1. Riboflavin.


Riboflavin (C 17H zoN40 6), crystallizes from 2 N acetic acid, alcohol, water or
pyridine in orange-yellow needles. Its molecular weight is 37636. The
decomposition point is 278-282C, and it darkens at about 240C. The vitamin
is odorless and has a bitter taste. Riboflavin is soluble in water only to the
extent of 10-13 mg in 100 ml at 25-275C, 19 mg in 100 ml at 40C, and
230 mg in 100 ml at 100C. The vitamin dissolves in ethanol and is slightly
soluble in amyl alcohol, cyclohexanol, benzyl alcohol, phenol and amyl
acetate, but is insoluble in ether, chloroform, acetone and benzene. Although
alkali dissolves the vitamin well, these solutions are unstable.
Neutral aqueous solutions of riboflavin have a greenish-yellow color. The
absorption spectrum shows characteristic absorption maxima at 475, 446,
359-375, 268 and 223 nm. The absorption in the visible part of the spectrum is
used for quantitative determination of riboflavin (Wagner-Jauregg, 1972).
Neutral aqueous solutions of riboflavin display intense yellowish-green fluorescence, with a maximum at 565 nm which can be used for quantitative
determination of the vitamin. The fluorescence vanishes on the addition of acid
or alkali, optimum fluorescence occurs at pH 3-8.
Riboflavin is an amphoteric compound. Its dissociation constants are
Ka = 63 X 1O- 1z and KfJ = 05 X 10-5 , the isoelectric point corresponds to a
pH of 60.
The basic factors affecting the stability of riboflavin in food are heat, light
and the reactions which occur in cells during the storage. Neutral solutions of
riboflavin can be sterilized by autoclaving for a short time: only slight
destruction occurs by heating to 120C for 6 h. No appreciable destruction of
the vitamin can be observed during the cooking of food since it is relatively
heat stable, but alkali decomposes riboflavin rapidly (Wagner-Jauregg, 1972;
Heimann, 1980). When milk in bottles is exposed to sunlight, 85% of its
riboflavin content is destroyed within 2 h.
Riboflavin is stable against air and most common oxidizing agents, but when
reduced by reducing agents such as sodium hydrosulfite or hydrogen, riboflavin
readily takes up two hydrogen atoms to form the colorless and non-fluorescent
1,1O-dihydro compound, leucoriboflavin, which can be reconstituted by shaking an aqueous solution with air. This reduction and reoxidation is also useful

Microbial Production of Vitamin B2


in increasing the specificity of riboflavin assay in food and feeds containing

interfering fluorescent substances (Kirk & Othmer, 1968).
Irradiation of riboflavin in solution in the visible light or ultraviolet range
degrades the side chain, yielding a number of compounds. These compounds
are Lumiflavine (6,7 ,9-trimethylisoalloxazine) , C 13H 12N 4 0 2, and Lumichrome
(6,7-dimethylisoalloxazine), C 12H lON 4 0 2. Irradiation with y rays causes loss of
riboflavin in solution but not in dry mixes. Sterilization of foods by irradiation
or treatment with ethylene oxide may also cause destruction of riboflavin (Kirk
& Othmer, 1968).


Riboflavin can be produced in significant quantity by various fermentation

processes using a number of micro-organisms and appropriate media. Vitamin
B2 concentrates for the enrichment of poultry and livestock feeds can be
prepared more cheaply by fermentation processes (Smith & Berry, 1975).
Riboflavin can be synthesized by many various types of micro-organisms.
Bacteria, yeasts, moulds, algae and protozoa are the micro-organisms which
can potentially produce riboflavin.
Different natural sources have been used for the production of riboflavin by
fermenting micro-organisms. Lactose fermenting yeasts are especially
Saccharomyces fragilis,
Clostridium butylicum,
several species of
Lactobacillus, or moulds, e.g., Aspergillus niger, Aspergillus flavus, Penicillium chrysogenum, and species of Fusarium. Whey and other milk byproducts can be employed as lacteal raw materials by these micro-organisms.
Riboflavin was formed in the butyl alcohol-acetone fermentation of cereal
grain mashes or molasses with various strains of Clostridium. Among these
bacteria Clostridium acetobutylicum especially is one of the best producers of
riboflavin (Kirk & Othmer, 1968). A suitable carbohydrate containing mash
cereals, rice, corn, whey and potato starch have been used in these commercial
anaerobic fermentation processes of Clostridium acetobutylicum (WagnerJauregg, 1972).
Most varieties of the yeast species Candida are also able to produce
riboflavin (Goodwin, 1959; Goodwin, 1963a,b). Candida guilliermondia, C.
tropicalis varrhaggi, C. flareri, C. arborea, C. lactis, C. krusei, C. ghoshi, C.
chalmersi, C. lypolytica, C. olea, C. pulcherrima, C. solani and C. melibiosi
were found to synthesize significant amount of riboflavin (Goodwin, 1963a;
Robinson, 1966).
Among these riboflavin producing micro-organisms, two closely related
ascomycetes, Eremothecium ashbyii and Ashbya gossypii (Ainsworth &
Sussman, 1965) are capable of synthesizing very large amounts of this vitamin
under appropriate conditions (Perlman et al., 1952; Hickey, 1954).
A great number of cheap natural materials and industrial wastes have been
used for riboflavin fermentation. Fermentation residues such as grain stillages


T. Kutsal & M. T. Ozbas

from the ethyl alcohol and butyl alcohol-acetone fermentations; sugar

containing substances such as molasses and corn syrup; fish meal, dried yeast,
skim milk, cotton-seed meal, soybean meal, peas, tankage, wheat bran, animal
proteins and malt extract are some of the raw materials employed in industrial
processes for riboflavin production by Eremothecium ashbyii (Hickey, 1954;
Gutcho, 1973).
Ashbya gossypii is also an important ascomycete very similar morphologically to Eremothecium ashbyii. Corn steep liquor, still ages , proteinaceous
agents from animal sources such as tankage, animal stick liquor and meat
scraps, vegetable proteinaceous agents, such as soybean meals, glutens, linseed
and cotton-seed meals have been used as raw materials for commercial
production of riboflavin by Ashbya gossypii (Goodwin, 1959; Robinson, 1966).
Lipids may be of importance for high vitamin B2 production by these two
ascomycetes (Ozbas & Kutsal, 1986a,b, 1987). The lipids studied included corn
oil, soybean oil, cocoanut oil, lard oil, lecithin, menhaden oil, sunflower oil
and others. (Hickey, 1954; Lago & Kaplan, 1981).
It is known that algae synthesize riboflavin but detailed work has not been
reported in literature (Goodwin, 1963b).
Among the protozoa, Chilomonas paramoecium synthesizes riboflavin when
cultured on a medium containing only acetate and inorganic salts (Goodwin,


There are many reports on general factors controlling riboflavin biosynthesis.

Purines stimulate riboflavin over-production in all of the organisms studied.
This holds true not only for complete fermentations, but for cell washed
suspensions as well. Glycine, a known purine precursor, stimulates overproduction in Candida species and in A. gossypii. The stimulatory effects of
purines and glycine in A. gossypii is not due to an effect on growth but is
specific for riboflavin oversynthesis (Demain, 1972). One of the proposed
mechanisms is described as follows.
The biosynthetic pathways for purines, riboflavin, and pteridines are closely
linked. Thus the carbon atoms of glycine are incorporated into these three
classes of compounds in structurally analogous positions. A preformed purine
can be converted directly into riboflavin with loss of only carbon 8 of the
purine ring: other purine carbons and all the ring nitrogens appear finally in
the pyrimidine portion for the riboflavin molecule (Wagner-Jauregg, 1972).
According to some workers who studied the biosynthesis of vitamin,
riboflavin (IV) is formed in E. ashbyii from the purine (I) (9-ribitylxanthin)
t,hrough the unstable 4-ribityl-amino-5-aminouracil (II) and the dioxopteridine
(III) (6,7-dimethyl-8-ribityl-lumazine or DMRL) (Fig. 2). It has been synthesized by condensation of II with diacetyl and the same pathway has been
suggested for the biological synthesis.

Microbial Production of Vitamin B2


R = D-Ribityl



Fig. 2. The biosynthesis of riboflavin.

That lumazine III alone is the precursor of riboflavin (IV) in organisms such
as E. ashbyii and A. ashbyii has been established by biochemical studies. All
the carbon atoms of the o-xylene moiety of riboflavin are derived from the
lumazine III, two molecules of which are converted by riboflavin synthesis into
one molecule of riboflavin (IV) and 4-ribitylamino-5-amino-2,6-dihydroxy
pyrimidine. In this biochemical reaction the lumazine (III) simultaneously
functions as a donor and acceptor of a C 4 unit which consists of the two methyl
groups and the C atoms 6 and 7. There is also a similar mechanism for
riboflavin formation in plants (Wagner-Jauregg, 1972).
Goodwin et al. showed independently that amino acids provide carbon units
and not nitrogen in the biosynthesis of riboflavin. These workers used E.
ashbyii and found that L-threonine, L-serine or L-tyrosine stimulated riboflavin
synthesis whereas L-glutamate, L-aspartate and L-asparagine stimulated both
growth and riboflavin synthesis; L-cystein on the other hand inhibited both
(Goodwin, 1963a,b; Robinson, 1966).


The changes in the culture media and the selection of high riboflavin-producing
mutant cultures cause the most important increases in riboflavin productivity.
In 1950, Pfeifer et al. suggested that some improvement in fermentation
productivity was possible as a result of culture selection. Paralleling the pilot
plant fermentations, a special study was carried out in the laboratory by
Pridham of the NRRL fermentation division to obtain and test natural and
induced variants of Ashbya gossypii NRRL Y -1056 for riboflavin production.
One of the natural variants was chosen as superior to the original strain on the
basis of shaker-flask experiments and was tested in a series of pilot plant
fermentations. This particular strain gave consistently higher yields of

T. Kutsal & M. T. Ozbas


riboflavin than did the standard NRRL Y -1056. The results indicate the
desirability of isolating high-producing strains at regular intervals in order to
maintain riboflavin yields at the highest possible values (Pfeifer et al., 1950).


Commercial fermentation processes for production of riboflavin or riboflavin

concentrates are relatively recent, having been developed since about 1940.
The Ascomycete processes using Eremothecium ashbyii and Ashbya gossypii
are now assuming the most important position in the riboflavin
manufacturing field. Both organisms are capable of synthesizing very large
amounts of riboflavin under appropriate conditions.

7.1 Inoculum
As reported in Section 4, a great number of cheap natural materials and
industrial wastes have been used for riboflavin fermentation by these Ascomycetes (Hickey, 1954; Goodwin, 1959; Robinson, 1966; Gutcho, 1973; Ozbas,
1985). In order to produce very large amounts of riboflavin, an appropriate
combination of these substrates can be chosen and if necessary, several
minerals can be added to the medium.
The generally accepted inoculating procedure is to use a 025-10%
inoculum from an actively growing 24-28 h culture (Goodwin, 1959).

7.2 Medium
In laboratories, cultures of E. ashbyii and A. gossypii are grown on solid agar
media in slant tubes and Petri dishes. The composition of the culture medium
is as follows (g/liter): glucose, 200; peptone, 50; yeast extract, 50; malt
extract, 50; MgS04 7H2 0, 02; K2 HP0 4 , 02. The -pH of the medium is
adjusted to the desired value with 05M H 2 S04 (Ozbas et al., 1984, 1986a,b).
Both liquid and solid media are sterilized at 121C in an autoclave before
inoculation. The initial substrate concentrations can be changed by making the
amount of initial carbon in the medium equivalent to 200 g/liter glucose while
the amounts of other constituents are kept constant.

7.3 Fermentation Course and Parameters

General fermentation conditions that affect micro-organism growth and
productivity are pH, temperature, aeration level, initial substrate concentration such as glucose, and growth factors such as DL-methionine, inositol,
D( +)-biotin and thiamine.

7. 3.1 Effect of initial pH

An initial pH of 60-80 for Eremothecium ashbyii and 55-70 for Ashbya
gossypii were found to be preferable (Hickey, 1954; Goodwin, 1959). The


Microbial Production of Vitamin B2











1.00 ::










Fig. 3. Effect of initial pH on the specific growth (I-') and riboflavin production (v)
rates for E. ashbyii and A. gossypi (30; stirring rate 100 rpm; SGO = 200 g/liter; . , E.
Ashbyii; 0, A. gossypii.

effects of initial pH on specific growth and riboflavin production rates with E.

ashbyii and A. gossypii are shown in Fig. 3. The optimum initial pH value is
65 for both micro-organisms (Ozbas & Kutsal, 1986a,b).
Riboflavin production by these two ascomycetes fits the third group of the
Gaden classification. In these riboflavin fermentations the pH of the medium
decreases with time and drops from its initial value of 65 to approximately 45 .



~Jo\I// "\



o 3 /~_'--_--'-_--'-_ _








0.10 _


0.08 ~

5.0 ::r:













Fig. 4. Changes in medium during fermentation (30; initial pH = 65; stirring rate
750 rpm; aeration rate, 220 cc/min; SGO = 200 g/liter; 0, riboflavin (C R ); . , dry weight
(X); 0, pH).

T. Kutsal & M. T. Ozbas




















T (oC)

Fig. 5. Effect of temperature on the specific growth and riboflavin production rates for
E. ashbyii (stirring rate, 100 rpm; initial pH = 65; 0, SGo = 200 g/liter; . , SSo =
200 g/liter).

At this point the micro-organisms grow rapidly, but no significant amounts of

riboflavin are synthesized. Then both riboflavin production and pH increase
and riboflavin formation terminates at approximately pH 95. Figure 4 shows
characteristic changes in the medium during the fermentation process (Ozbas
et aI., 1984; Ozbas & Kutsal, 1986a, b).
7.3.2. Effect of temperature
The optimum temperature ranges for riboflavin fermentation by Eremothecium
ashbyii and Ashbya gossypii were obtained between 25 and 30C and 26 and
28C respectively (Hickey, 1954; Gutcho, 1973).
Ozbas & Kutsal investigated the effects of temperature by utilizing glucose
and sunflower oil as substrate in the range from 20 to 40C. For E. ashbyii and
A. gossypii maximum specific growth and riboflavin production rates were
obtained at 30C in all medium compositions. At lower temperatures both
growth and riboflavin yield were reduced to values similar to those observed at
higher temperatures (Figs 5 and 6).
The Ascomycete processes for riboflavin production are aerobic and it is
necessary either to aerate or shake the cultures.
Z 3. 3 Effect of initial substrate
For the growth of both micro-organisms the main carbon source is glucose.
Maximum yields are obtained with an initial glucose concentration of
200 g/liter. Glycerol, sunflower oil, whey and several combinations of these
substrates can also be used for riboflavin production. The effects of different
substrates and their initial concentrations on the specific growth and riboflavin

Microbial Production of Vitamin B2








/;or" '\
:/,/" \ \
/0. . . . . 0"'0





T (DC)

Fig. 6. Effect of temperature on the specific growth and riboflavin production rates for

A. gossypii (stirring rate, 100 rpm; initial pH = 65; 0, Sao = 100 g/liter, Sso =
1O0g/liter; e, Sao = 200 g/liter).

production rates and riboflavin yields are given in Table 1 (Ozbas & Kutsal,
7. 3. 4 Fermentor experiments
The effects of agitation and aeration rates on the formation of riboflavin were
observed at the range of 300-1200 rpm and 70-330 cc/min, respectively for
Eremothecium ashbyii. Optimum values were reached as Il = 0176 h- 1 and
v = 445 X 10- 3 g/liter per hat 750-1000 rpm and Il = 0181 h- 1 and v = 369 X
10-3 g/liter per h at 220 cc/min. At these optimum conditions, adding the
sunflower oil to the medium resulting in 50 g/liter glucose and 150 g/liter
sunflower oil showed Il to be 0281 h- 1 and v = 534 X 10- 3 g/liter per h. These
results suggested that in addition to pH, temperature and initial substrate
concentrations, agitation and aeration are also important control parameters
on the riboflavin production (Ozbas et al., 1984).
7. 3. 5 Effect of growth factors
The effect of some growth factors such as oL-methionine, inositol, 0(+ )-biotin
and thiamine for riboflavin production by E. ashbyii and A. gossypii were
investigated in various media.
In order to determine initial concentration effects of these growth factors,
glucose was used as substrate at the optimum growth conditions. While,
o( + )-biotin and thiamine are ineffective in stimulating either growth or
flavinogenesis, maximum specific growth and riboflavin production rates and
riboflavin yields are obtained in the media containing 04 g/liter oL-methionine
and 02 g/liter inositol for E. ashbyii. It is also observed that all of these
four compounds are effective in supporting growth and product formation

Table 1

E. ashbyii
VX 10 3
(gR/gmo per h)


(h -1)




50+ 150
75 + 125
50+ 150
100+ 100

Calculated in the region of exponential growth (Monod kinetic model).

Sunflower oil
Glucose + glycerol
Glucose + glycerol
Glucose + sunflower oil
Glucose + sunflower oil






vX 10 3
(gR/gmo per h)

A. gossypii



(h -1)




Effects of Various Substrates on Specific Growth (It) and Riboflavin Production (v) Rates and Riboflavin Yields (YP/SCa)
for E. ashbyii and A. gossypii (30"C, Stirring Rate: 100 rpm, Initial pH = 65)







Table 2


(h -1)


10 3
(gR/g rno per h)

E. ashbyii





A. gossypii
v X 10 3
(h -1)
(gR/g rno per h)



Initial substrate concentrations were (g/liter):

a Glucose (G), 200; G + sunflower oil (S), 50 + 150; whey (W), 125; DL-methionine (M), 04;
inositol (I), O 2.
b G, 200; G + S, 100 + 100; W, 300; M, 008; I, 05; D( + )-biotin (B), 4 x 10- 7 ; thiamine (T), 004.



Effects of Various Growth Factors on Specific Growth (fJ) and Riboflavin Production (v) Rates and
Riboflavin Yields (YPlScJ for E. ashbyii and A. gossypii (30C, Stirring Rate: 100 rpm, Initial pH = 65)












T. Katsal & M. T. Ozbas


with A. gossypii. Optimum initial growth factor concentrations are found as

008 g/liter oL-methionine, 05 g/liter isositol, 04 Jlg/liter 0(+ )-biotin and
004 g/liter thiamine for this ascomycete.
It was observed that the addition of these growth factors to the medium
containing only whey or glucose as substrate caused riboflavin production by
E. asbhyii and A. gossypii to increase (Table 2). Although higher yields were
obtained in the media containing sunflower oil compared to that of glucose
lower yields were attained with these growth factors in glucose and sunflower oil
medium for both micro-organisms (Ozbas & Kutsal, 1987).

7.4 Flow-Sheet
Descriptions of commercial methods for riboflavin production are primarily
limited to the patent literature.
Commercial methods for cultivation of E. ashbyii may be surface methods
exposed to air, or submerged methods in which air is dispersed throughout the
mash, with or without supplementary agitation (Goodwin, 1959; WagnerJauregg, 1972).
The procedures generally adopted in industry by using A. gossypii utilize
aerated and agitated submerged cultures in liquid media. The air flow rate in
liters/min should be at least 025 times the volume of the medium in liters.
Excessive agitation which interferes with mycelial development reduces the
riboflavin production. Analysis of the experimental results indicates that
satisfactory yields of riboflavin can be obtained if fermentation variables are
closely controlled. In one of these pilot plant experiments, 20% glucose,
18-21 % suitable corn steep liquor, 10% animal stick liquor, and a small
amount of some antifoam agent had been used. After fermentation for
96-120 h at 30C, the riboflavin content of the liquor was 500-600 y / cc. The
fermented liquor had been evaporated to syrup in conventional evaporators
and dried to a concentrate containing 25% riboflavin on steam-heated drum
dryers. In these experiments, inoculum quantity and age, corn steep liquor
from various sources, and use of other nutrients and experimental techniques
are also introduced (Pfeifer et al., 1950; Goodwin, 1959).
The commercial production of riboflavin by A. gossypii is shown in Fig. 7
(Pfeifer et al., 1950).


Fermentation solids containing riboflavin such as residues from butyl alcoholacetone fermentation may be recovered for use as animal feed supplements by
drying and powdering to a crude product. Drying is generally accomplished by
drum or spray drying procedures. These food grade products usually contain
other members of the B-complex, along with minerals and proteinaceous


water in

Air compressor





Fig. 7. Flow sheet for riboflavin production by fermentation (Pfeifer et al., 1950). 1, Glucose; 2, com steep liquor;
3, animal stick liquor; 4, sulphuric acid; 5, soybean oil; 6, caustic soda; 7, inoculum tank; 8, air filter; 9, sterilizer;
10, cooler.

Water out


Laboratory culture
Ashbya gossypii









T. Kutsal & M. T. Ozbas


Pure, crystalline riboflavin is obtained from some liquors containing

sufficiently high concentrations of riboflavin, such as the fermentation broths of
E. ashbyii and A. gossypii.
For the purification of riboflavin from the fermentation broths, it is useful to
remove lipids by extraction with ether, in which the vitamin is insoluble. In
some cases, salts and glycogen can be eliminated from riboflavin concentrates
by fractionate precipitation with alcohol or acetone.
Various methods were developed for recovery of riboflavin from fermented
Two methods were described in 1945 by Hines who found first that by
bacteriological reduction, riboflavin in fermentation liquors of E. ashbyii could
be precipitated as a reddish-brown amorphous product by the action of
Streptococci strains (Hickey, 1954; Wagner-Jauregg, 1972; Periman, 1979).
Streptococcus faecaiis, S. liquefaciens and S. zymoge may be used in these
microbiological methods (Brit. Pat. 621,468).
Some chemical methods can be used also for recovery of riboflavin from
fermented mash. Chemical reducing agents, such as sodium dithionite
(Na2S20 4), stannous and chromous chlorides, and certain other reducing
compounds were employed in these chemical precipitation methods. Precipitation could be obtained more rapidly and almost immediately by the use of
chemical reducing agents. Of these, sodium dithionite has economic and low
toxicity advantages over some of the other useable reducing agents, such as
stannous or chromous chlorides (Hickey, 1954; Wagner-Jauregg, 1972).
The chemical precipitation methods operate primarily on solutions containing at least 20/Jog/ml. In these methods the insoluble matter of mashes is
removed by centrifugation and riboflavin-containing parts are adjusted to
pH 50-55. A reducing agent such as Na2S204 is added at a temperature of
20-30 C and in a ratio of 5 moles of reducing agent to 1 mole of riboflavin
(Brit. Pat. 621,468; US Pat. 2,367,644; US Pat. 2,367,646).
The resulting precipitate is permitted to settle, the supernatant is centrifugated. The residue is filtered. The crude precipitate which is obtained by such
means can be readily converted to crystalline riboflavin. In these processes, the
reddish-brown precipitate is dissolved in a hot polar solvent, such as 75%
aqueous isopropyl alcohol. This mixture is heated to dissolve all the precipitate, except the inert material present, filtered, and the clear filtrate on cooling
gives yellow needlelike crystals of riboflavin which are separated by filtration
or centrifugation. Proportion of solvent to precipitate, and the temperature
employed in obtaining riboflavin are not critical (Brit. Pat. 621 468; US Pat.
2421142; US Pat 2367646).
Riboflavin can be also purified by adsorption methods. Good adsorbents for
riboflavin are Fuller's earth in acid solution, Florisil, Floridin XXF, and
Frankonit in neutral solutions. The vitamin is adsorbed very strongly by
charcoal, but elution is difficult from this adsorbate. A combination of
precipitation and adsorption methods can sometimes be necessary to isolate
pure riboflavin. A general method for the preparation of pure D-riboflavin

Microbial Production of Vitamin B2


from natural sources has been described which is based on adsorption on

Fuller's earth, fractionation with immiscible solvents and acetone, and
crystallization from an aqueous acetone-petroleum ether mixture; aqueous
alcohol solutions have been used for elution of the adsorbates (WagnerJauregg, 1972).


Riboflavin can be assayed by physico-chemical, microbiological, or biological

methods. The most commonly used physico-chemical method involves measurement of the characteristic, yellow-green fluorescence of riboflavin which
has a maximum at 565 nm at pH 6. For complex mixtures such as foods or
feeds, chemical pretreatment of extracts by column chromatography, oxidation, reduction, or extraction is necessary to remove interfering fluorescent
impurities. Riboflavin can be measured by polarography, and also by photometric or fluorometric measurement of lumiflavine formed by the irradiation of
riboflavin in alkaline solutions.
As with most vitamins, the first assays developed for riboflavin depended
upon the measuring of the biological response of animals. Early studies were
carried out in the rat and in the chicken. Animal assays are time-consuming,
expensive, and least accurate, but have the advantage of being based on a
biological animal response, which is important from the nutritional standpoint.
They are particularly useful for assaying riboflavin derivatives since substituent
groups on the riboflavin molecule frequently depress or eliminate the
biological response.
In animal methods the reference standard is crystalline synthetic riboflavin.
For administration in rat growth assays it is useful to prepare a stock solution,
which must be kept in a dark bottle in the refrigerator.
Since riboflavin is widely distributed in plant and animal tissues, test
substances may occur in a variety of forms. Solutions containing high
concentrations of riboflavin can be administered directly in suitable dilutions.
Samples of relatively low potency are mixed with a small portion of the basal
diet. Test materials with a high fat content may increase the riboflavin
requirement of the rat (Pearson, 1967). A human bioassay based on
comparative urinary excretions of riboflavin following test doses of a standard
vs tablets has been used widely for checking the physiological availability of
riboflavin in tablets (Kirk & Othmer, 1968).
One of the microbiological assays of riboflavin is based on the growth of
some micro-organisms. Not many bacteria have a requirement for riboflavin,
and virtually all of those known are lactic acid bacteria. The classic
microbiological assay for riboflavin with Lactobacillus casei which is measured
by titrating the lactic acid formed is in use in the original or modified form. A
more sensitive assay is afforded by the use of Leuconostoc mesenteroides. This
organism is 50 times more sensitive than L. casei. A method for total riboflavin


T. Kutsal & M. T. Ozbas

in body fluids and tissues suitable for clinical and nutritional surveys is based
on the ciliate protozoan Tetrahymena pyriformis (Pearson, 1967; Kirk &
Othmer, 1968).
The principal forms of riboflavin which exist in living cells are riboflavin
5'phosphate (FMN) and flavin adenine dinucleotide (FAD). Both these
phosphates are protein bound and 60-90% of the total riboflavin in natural
products is present in the form of the dinucleotide. Riboflavin performs its
biological functions in a number of different enzyme systems. Two derivatives
of riboflavin, FMN and FAD, serve as prosthetic groups and combine with
specific protein enzymes which catalyze oxidation-reduction reactions in cells
(Pearson, 1967). The good sources of riboflavin are milk, egg, liver, heart,
kidney, green, leafy vegetables, apricot, tomato, beef and poultry meats
(Wagner-Jauregg, 1972).
Riboflavin can thus be produced by chemical synthesis or by fermentation
processes. Although pure crystallized riboflavin for therapeutic purposes is also
made by chemical synthesis, various commercially competitive fermentation
processes have recently been developed for riboflavin production (Ozbas &
Kutsal, 1986a). Other processes start from D-ribose, which is produced by
fermentation (Chapter 11).
In 1891, O. Kiihling synthesized alloxazines by condensation of o-phenylenediamine hydrochloride with alloxan. Using the same principle, in 1934, R.
Kuhn and P. Karrer independently worked out methods for the synthesis of
flavins, based on o-xylene, D-ribose and alloxan as starting materials. Later,
processes of technical importance were developed by employing many refinements in the general synthetic methods of R. Kuhn and P. Karrer (WagnerJauregg, 1972).
Processes employing several chemical procedures have also been patented.
Riboflavin is essential for the growth and normal health of animals as well as of
man. A lack of riboflavin in human or animal diets causes the formation of
certain oral, cutaneous and corneal lesions. Vitamin B2 is widely used in the
food enrichment, pharmaceutical and feed supplement industries (Kirk &
USP riboflavin for therapeutic purposes is administered orally in tablet form
or by injection as a sterile aqueous solution, with niacinamide or other

Microbial Production of Vitamin 8 2


solubilizing agent added. All of the vitamin B complex and most of the
multivitamin preparations also contain riboflavin (Kirk & Othmer, 1968).
Sterile, supersaturated solutions of riboflavin in normal saline have been
employed for intravenous administration (Wagner-Jauregg, 1972).
Some of the cereal products such as flour and bread are enriched with iron,
thiamine and niacin as well as riboflavin to compensate for the loss of these
nutrients that occurs in the processing of wheat (Kirk & Othmer, 1968).
Riboflavin can be synthesized by most higher plants, yeasts, and lower fungi,
and bacteria. The tissues of higher animals are unable to synthesize this
vitamin. For the enrichment of poultry and livestock feeds, riboflavin is usually
added at concentrations of 2-8 g/ton, depending on the species, age and
purpose. The supplement of riboflavin improves health, growth, tissue repair,
and reproduction of the animals and also has an economic importance in
poultry raising and egg production (Kirk & Othmer, 1968; Wagner-Jauregg,
The present world consumption of riboflavin is approximately 125 million
kg for human and animal use combined (Lago & Kaplan, 1981).
Commercial fermentation processes for production of riboflavin or riboflavin
concentrates are relatively recent, having been developed in the past 40 years.
In 1946 processes using the ascomycete A. gossypii were started. Manufacturers using the microbiological process are Merck and Co., Inc. (USA) and
BASF (FRG). Companies manufacturing riboflavin by chemical synthesis
include Hoffmann-LaRoche Inc. (USA and Switzerland), Takeda Chemical
Industries Ltd (Japan), Pfizer, Inc. (USA), and E. Merck (FRG) (Perlman,

Ainsworth, G. C. & Sussman, A. S. (1965). The Fungi, Academic Press, New York,
pp. 37,492.
Brit. Pat. 621,468 (April 11 1949) (to Commercial Solvents Corp.)
Demain, A. L. (1972). Riboflavin oversynthesis, In Annual Reviews of Microbiology
Vol. 26, ed. C. E. Clyton, S. Raffel & M. P. Starr. George Benta Inc., USA, pp.
Goodwin, T. W. (1959). Production and biosynthesis of riboflavin in micro-organisms.
In Progress In Industrial Microbiology, ed. D. J. D. Hockenhull. Heywood &
Company, London, pp. 139-177.
Goodwin, T. W. (1963a). Vitamins. In Biochemistry of Industrial Micro-organisms, ed.
C. Rainbow & A. H. Rose. Academic Press, London, pp. 151-8.
Goodwin, T. W. (1963b). Riboflavin and related compounds. In The Biosynthesis of
Vitamins and Related Compounds, ed. T. W. Goodwin. Academic Press, London,
Gutcho, S. J. (1973). Chemicals by Fermentation, Noyes Data Corporation, Chemical
Technology Review No. 19, New Jersey, pp. 323-25.
Heimann, W. (1980). Fundamentals of Food Chemistry, Ellis Horwood, Chichester, pp.


T. Kutsal & M. T. Ozbas

Hickey, R. J. (1954). Production of riboflavin by fermentation. In Industrial

Fermentations, ed. L. A. Underkofler & R. J. Hickey. Chemical Publishing Co.,
New York, pp. 157-90.
Kirk, R. E. & Othmer, D. F. (1968). Encyclopedia of Chemical Technology, Vol. 17,
John Wiley & Sons, New York, pp. 445-58.
Lago, B. D. & Kaplan, L. (1981). Vitamin fermentations: Bz and B 12 Adv. in Biotech.,
Ozbas, T., Kutsal, T. & CagIar, A. (1984). The production of riboflavin by
Eremothecium ashbyii at various media. In Proceedings of the 3rd European
Congress on Biotechnology, ed. D. Behrens, Verlag Chemie GmbH, Weinheim, pp.
Ozbas, T. (1~~5). Eremotheciu11J ashbyii Ashbya gossypii Mikroorganizmalari lIe
Riboflavin Uretimi. MSc thesis, Hacettepe University, Ankara, Turkey.
Ozbas, T. & Kutsal, T. (1986a). Riboflavin production by Eremothecium ashbyii in a
batch stirred tank fermenter. Biotechnology Letters, 8,441-44.
Ozbas, T. & Kutsal, T. (1986b). Comparative study of riboflavin production from two
micro-organisms: Eremothecium ashbyii and Ashbya gossypii. Enzyme Microb.
Techno!., 8, 593-96.
Ozbas, T. & Kutsal, T. (1987). Effects of inositol, D( + )-biotin and thiamine on
riboflavin production by Ashbya gossypii. In Proceedings of the 4th European
Congress on Biotechnology, ed. O. M. Neijssel, R. R. van der Meer & K. Ch. A.
M. Luyben. Elsevier Science Publishers B.V., Amsterdam, pp. 273-76.
Pearson, W. N. (1967). Riboflavin. In The Vitamins, ed. P. Gyorgy & W. N. Pearson.
Academic Press, New York, pp. 99-101.
Perlman, D. (1979). Microbial process for riboflavin production. In Microbial
Technology, ed. H. J. Peppler & D. Perlman. Academic Press, New York, pp.
Perlman, D., Brown, W. E. & Sylvan, B. L. (1952). Fermentation. Ind. Eng. Chem.,
44, 1996-2001.
Pfeifer, V. F., Tanner, F. W., Vojnovich, C. Jr & Traufler, D. H. (1950). Riboflavin by
fermentation with Ashbya gossypii. Ind. Eng. Chem., 42, 1776-81.
Robinson, F. A. (1966). The Vitamin Co-factors of Enzyme Systems, Pergamon Press,
Braunschweig, pp. 160-63.
Smith, J. E. & Berry, D. R. (1975). The Filamentous Fungi, Vol. 1, Edward Arnold,
London, pp. 67-8, 115-116.
US Pat. 2367644 (January 16 1945), G. E. Hines, Jr (to Commercial Solvents Corp.).
US Pat. 2367646 (January 161945), M. G. W. Millian (to Commercial Solvents Corp.).
US Pat. 2421142 (27 May 1947) J. K. Dale (to Commercial Solvents Corp.).
Wagner-Jauregg, T. (1972). Chemistry. In The Vitamins, ed. W. H. Sebrell Jr & R. S.
Harris. Academic Press, New York, pp. 3-5.

Chapter 11




& M.


Central Research Division, b Corporate Strategy, Takeda Chemical Industries

Ltd, Yodogawa-ku, Osaka 532, Japan

o-Ribose and its derivative, 2-deoxy-o-ribose, are components of RNA and
DNA, respectively. It is of interest that these pentoses are components of such
important biopolymers in the hereditary material of organisms whereas other
ubiquitous pentoses, L-arabinose and o-xylose, are usually present in the plant
cell walls. Why the pentose components of nucleic acids are o-ribose and
2-deoxy-o-ribose is an unresolved problem; o-ribose and its derivatives have
been also found widely in nature.
From the viewpoint of commercial application, o-ribose has long served as
the focus for a starting material for the chemical synthesis of riboflavin (Fig.
1), which has been used in large amounts in pharmaceuticals and livestock feed
additives (see also Chapter 10). Various methods have been investigated and
established for the large-scale production of o-ribose.
Investigations on o-ribose started in 1909 when Levene and Jacobs identified
o-ribose as a component of yeast RNA (Levene & Jacobs, 1909). The
subsequent endeavors of many investigators were devoted to efficiently
preparing o-ribose from yeast RNA, e.g. chemical hydrolysis of yeast RNA
(Levene & Clark, 1921; Levene, 1935) and the enzymatic hydrolysis of yeast
RNA (Bredereck & Rothe, 1938; Bredereck et aI., 1940). Along this line, the
production of o-ribose from 5-amino-4-imidazole-carboxamide riboside by
enzymatic hydrolysis with a bacterial nucleosidase has recently been developed
(Sano et al., 1977a,b).
On the other hand, many investigators tried to establish a method for the
chemical synthesis of o-ribose from such a cheap material as o-glucose. These
studies were stimulated by identification of ribitol, the reduced product of
o-ribose, as a component of vitamin B 2 , riboflavin, and by chemical synthesis
of riboflavin in the middle of the 1930s (Karrer et al., 1935a,b; Kuhn et al.,
1935a,b; von Euler et al., 1935). The chemical process has been used as a


K. Sasajima & M. Yoneda

Fig. 1. Formula of riboflavin.

commercial procedure to produce o-ribose for the synthesis of riboflavin.

Investigations from 1910 to the 1950s on preparing and chemically synthesizing
o-ribose were reviewed succinctly by Jeanlotz & Fletcher (1951) and Overend
& Stacey (1955).
In the early 1950s, the pentose phosphate pathway was established as an
important metabolic pathway other than glycolysis in carbohydrate metabolism
(Wood, 1985). Thus, the biosynthesis of o-ribose was elucidated, which was
very important information for developing a commercial procedure for
o-ribose production using the transketolase (EC mutants (tkt mutants)
of Bacillus species as will be described later.
At the same time, the production of o-ribose by micro-organisms was first
reported by Simonart & Godin (1951). About 10 years later, an unidentified
bacterium was found to excrete both o-ribose and o-ribose 5-phosphate into a
culture medium (Suzuki et al., 1963). A smaller degree of o-ribose accumulation was also described by Saito & Sugiyama (1966).
In the late 1950s, active investigations on the development of nucleotide
flavour enhancers, inosine monophosphate and guanosine monophosphate,
started in Japan. A manufacturing process was established and has been
long used to commercially produce these substances. During the course of
studies on the development of the process, it was found that the mutants of
Bacillus lacking transketolase, a key enzyme of the pentose phosphate pathway, accumulated a large amount of o-ribose (Sasajima & Yoneda, 1971).
Subsequent development research established an economical manufacturing
process for o-ribose production using these mutants. The commercial production of o-ribose for riboflavin synthesis started in the mid-1970s. This process
has now been widely used for 15 years. The characteristic feature of this
process is that mutants of carbohydrate metabolism of Bacillus species are
used. In contrast, wild-type strains of Acetobacter suboxidans, A. xylinum,
etc., are used in L-sorbose fermentation, which is another remarkable
microbial process of sugar production for vitamin C synthesis (Perlman, 1979).
This review focuses on transitions of developments in o-ribose production

Microbial Production of D-Ribose


methods and the related basic investigations. In particular, recent microbial

methods will be described in detail.



The occurrence of o-ribose and its derivatives in nature is summarized in

Table 1.
o-Ribose was first isolated from yeast RNA and identified by Levene &
Jacobs (1909). Thereafter, it was found to be a component of uric acid in the
blood (Davis et al., 1922); of crotonoside in the croton bean (Croton tiglium
L.) (Cherbuliez & Bernhard, 1932a,b); of coenzymes such as NAD and
NADP from yeasts (von Euler & Vestin, 1935; von El,ller & Schlenk, 1937;
von Euler et al., 1942) and from blood cells (Warburg et al., 1935), FAD
(Ochoa & Rossiter, 1939; Klein & Kohn, 1940), UDP-glucose (Caputto et aI.,
1950), and coenzyme A (Baddiley et al., 1953); of vitamin B12 (Beaven et al.,
1949; Brink & Folkers, 1949; Holiday & Petrow, 1949; Brink & Folkers, 1950;
Brink et al., 1950; Buchanan et aI., 1950a,b); and of poly(ADP-ribose) (Doly
& Mandel, 1967). Other coenzymes such as ADP, CDP, and GDP or many
nucleoside antibiotics (Suhadolnik, 1970) also contain o-ribose. It is also
present in the cell walls of Salmonella typhimurium as a component of
lipopolysaccharide (Kauffman et al., 1962) or of Eubacterium saburreum as a
component of polysaccharide (Hoffman et al., 1977; Hofstad & Lygre, 1977);
in the flavonoids of Phlebodium dictyocallis (Mett.) Gomez as a component of
o-glycosidic moieties (Gomez & Wallace, 1986); and in the coccoliths of
Emiliania huxleyi (Lohmann) Hay and Mohler as a component of an acidic
Ca2 + -binding polysaccharide (Borman et al., 1987).
o-Ribose derivatives such as 2-deoxy-o-ribose, 3-deoxy-o-ribose, 3-amino-3deoxy-o-ribose, ribitol, ribonic acid, o-riburonic acid, 5-methylthioribose and
glut amyl o-ribose have also been found as a component of organisms or as a
product of micro-organisms. 2-Deoxy-o-ribose is a component of DNA
(Levene et al., 1930; Chargaff et al., 1949; Chargaff & Lipschitz, 1953) which is
the carrier of genetic information in all organisms except RNA viruses (Avery
et al., 1944; Hershey & Chase, 1952; Watson & Crick, 1953). 3-Deoxy-o-ribose
is a component of an antibiotic, cordycepin, produced by Cordyceps militaris
(Bentley et al., 1951). 3-Amino-3-deoxy-o-ribose is a component of puromycin
produced by Streptomyces alboniger (Baker & Schaub, 1953; Waller et al.,
1953) and an antibiotic, 3'-amino-3'-deoxy-adenosine, produced by
Helminthosporium species (Gerber & Lechevalier, 1962; Guarino & Kredich,
1963). Ribitol (adonitol), the reduced derivative of o-ribose, was first isolated
from Adonis vernalis (Jeanlotz & Fletcher, 1951) and subsequently was found
in the roots of Bupleurum falcatum (Wessely & Wang, 1938), in honeydews
excreted by the genus Ceroplastes (Hackman & Trikojus, 1952), in pneumococci as a component polysaccharide (Rebers & Heidelberger, 1959, 1961;


K. Sasajima & M. Yoneda

Table 1
Occurrence of o-Ribose in Nature

o-Ribose or derivative

Ribonucleic acid
Uric acid

Coenzyme A
Vitamin B12


Poly (ADP-ribose)
Lipopolysaccharide of
Salmonella typhimurium
Polysaccharide of
Eubacterium saburreum
Flavonoids of Phlebodium
dictyocallis (Mett)
Coccolith-associated polysaccharide of Emiliania
huxleyi (Lohmann)
Hay & Mohler
Deoxyribonucleic acid
Adonis vernalis
Roots of Bupreum falcatum
Capsular of pneumococci

Ribitol teichoic acid of

Gram-positive bacteria
Lipopolysaccharide of
Proteus mirabilis
Lipopolysaccharide of Vibrio

Levene & Jacobs, 1909
Davis et aI., 1922
Cherbuliez & Bernhard,
von Euler & Vestin,
1935; Warburg et al.,
1935; von Euler &
Schlenk, 1937; von
Euler et al., 1942
Ochoa & Kossiter, 1939;
Klein & Kohn, 1940
Caputto et al., 1950
Baddiley et al., 1953
Beaven et aI., 1949;
Brink & Folkers, 1949;
Holiday & Petrow, 1949;
Brink & Folkers, 1950;
Brink et aI., 1950;
Buchanan et af., 1950a,b
Doly & Mandel 1967
Kauffmann et al., 1962
Hofstad & Lygre, 1977;
Hoffman et af., 1977
Gomez & Wallace, 1986
Borman et af., 1987

Levene et al., 1930;

Chargaff et af., 1949;
Chargaff & Lipschitz, 1953
Bentley et aI., 1951
Baker & Schaub, 1953;
Waller et al., 1953
Gerber & Lechevalier, 1962;
Guarino & Kredich, 1963
Jeanlotz & Fletcher, 1951
Wessely & Wang, 1938
Rebers & Heidelberger,
1959, 1961; Roberts et
ai., 1963
Ward, 1981
Gmeiner, 1977;
Gmeiner et af., 1977
Miyano et aI., 1983


Microbial Production of D-Ribose

Table l--contd.

D-Ribose or derivative


D-Ribonic acid

Oxidation product of
D-ribose by pseudomonads

D-Riburonic acid

Oxidation product of D-ribose

by Rhizobium meliloti
5' -Methylthioadenosine

Glutamyl D-ribose
D-Ribose 5-phosphate
D-Ribose I-phosphate
D-Ribose 1,5bisphosphate

Brain and kidney of a patient

Foster, 1944; Lockwood
& Nelson, 1946; Weimberg,
Amemura et at., 1981
Mandel & Dunham, 1912;
Suzuki et aI., 1914,
Williams et at., 1986

Pentose phosphate pathway

Precursor of nucleotides
Precursor of nucleotides
Human erythrocyte

Vanderheiden, 1970

Roberts et al., 1963), and in the cell walls of Proteus mirabilis (Gmeiner, 1977;
Gmeiner et al., 1977) and Vibrio parahaemolyticus (Miyano et al., 1983) as a
component of lipopolysaccharides and of various Gram-positive bacteria as a
component of ribitol teichoic acid (Ward, 1981). It is also a component of
riboflavin (Karrer et al., 1935a; Kuhn et al., 1935a; von Euler et aI., 1935).
D-Ribonic acid was the oxidation derivative of D-ribose produced by pseudomonads (Foster, 1944; Lockwood & Nelson, 1946; Weimberg, 1961). Another
oxidation product, D-riburonic acid, was found in the cell wall polysaccharides
of Rhizobium meliloti (Amemura et al., 1981). 5-Methylthioribose is a component of 5'-methylthioadenosine (Mandel & Dunham 1912; Suzuki et at.,
1914, 1924) which has an important role in the metabolism of adenosylthiomethionine (Schlenk & Smith, 1953). Glutamyl D-ribose 5-phosphate was
found as an abnormal storage substance in the brain and kidney of a patient
suffering from renal failure and neurologic deterioration (Williams et al.,
1986). D-Ribose 5-phosphate is an intermediate in the pentose phosphate
pathway (Wood, 1985) which plays an important role in the production of
D-ribose by the tkt mutants of Bacillus species as will be described later.
D-Ribose-l-phosphate and 5-phosphoribosyl I-pyrophosphate take part in
purine and pyrimidine nucleotide biosynthesis (Hartmann, 1970). o-Ribose
1,5-bisphosphate was found in human erythrocytes (Vanderheiden, 1970).


The chemical and physical properties of o-ribose and its derivatives have
already been described in detail by Jeanlotz & Fletcher (1951), Overend &


K. Sasajima & M. Yoneda


" " 0"

" "
Fig. 2. Formula of o-ribose.

Stacey (1955) and Windholz et al. (1983). The formulae of o-ribose is shown in
Fig. 2. The chemical and physical properties of o-ribose and its derivatives are
as follows.
o-Ribose C5 H lO0 5 ; mol. wt 15013, C 4000%, H 671; 05329%. Melting
point of crystalline o-ribose is 87C (Levene & Jacobs, 1909). Crystals are
plates or needles. o-Ribose shows mutarotation, final []b - 23'7 (4,5% in
water) (Phelps et al., 1934). Soluble in water, slightly insoluble in alcohol.
Anhydroribose C5 H s0 4; m.p. 229-230C (Bredereck et al., 1940).
-Aniline-N-o-ribopyranoside; m.p. 125-127C, [m + 634~ + 486
(C = 10% in pyridine) (Berger & Lee, 1944; Berger et al., 1944).
Benzylphenylhydrazone ClsH22N204; chrysanthemum-like crystals, m.p.
128C (Sasajima & Yoneda, 1971).
p-Bromophenylhydrazone; m.p. 166-167C (van Ekenstein & Blanksma,
1913; Steiger, 1936)
Methyl-o-riboside; crystals from ethyl acetate, m.p. 83-84 C, []~ - 1136
(p = 3) (Windholz et al., 1983).
Phenylosazone C17H2oN'403; yellow needles from pyridine + water, m.p.
The infrared absorption of o-ribose from 8 to 15 microns was measured by
Kuhn (1950). A polarographic study of o-ribose was described by Cantor &
Peniston (1940); and crystallographic properties were reported by Keenan
As already mentioned above, o-ribose is a component of biologically important substances such as RNA and coenzymes. After the classical elucidation of
the glycolytic pathway, the pentose phosphate pathway was established as
another important carbohydrate pathway in the early 1950s. The pentose
phosphate pathway and related metabolic pathways are shown in Fig. 3. It has
the following roles; (1) supplement of the o-ribose frame for nucleic acid
biosynthesis (Hartman, 1970); (2) the biosynthesis of o-erythrose 4-phosphate,
one of the essential precursors of aromatic amino acids and vitamins
(Greenberg, 1969); (3) the biosynthesis of C-7 and C-8 carbohydrate components of lipopolysaccharide in the cell walls of Gram-negative bacteria
(Eidels & Osborn, 1971; Woisetschlager & Hogenauer, 1986), and (4) the

TCA cycle

-- --"-- -- ---.- ---


o-Xylulose 5-phosphate


o-Fructose 6-phosphate

Glyceraldehyde 3-phosphate



Shikimic acid

o-Erythrose 4-phosphate

Sedoheptulose 7-phosphate


Purine nucleotide
Pyrimidine nucleotide



O-Ribose 1-PhosPhate]




Ribitol teichoic acid

o-Ribose 5-phosphate

~ Isom'

o-Ribulose 5-phosphate





Vitamin K
Folic acid
Fig. 3. Pentose phosphate pathway and related pathwayso



r--_ _ _ _

o-Glucose 6-phosphate









K. Sasajima & M. Yoneda

biosynthesis of ribitol, a component of ribitol teichoic acid in the cell walls of

Gram-positive bacteria (Ward, 1981).
The active investigations of two American research groups headed by
Bernard Horecker and Ebraim Racker in the early 1950s, following the
observation of the oxidation of D-glucose 6-phosphate to 6-phospho-Dgluconate by Warburg & Christian (1936, 1937) and the identification of
D-ribose 5-phosphate as one of the oxidation products by Dickens (1938a,b),
established the pathway scheme. Thus, the biosynthetic pathway from Dglucose to D-ribose has been elucidated. The roles of the pathway mentioned
above were established through subsequent active investigations. Recently, all
aspects of the pathway were reviewed by Wood (1985). The tkt mutants of
Bacillus species used for commercially producing D-ribose could be isolated on
the basis of the knowledge accumulated about the pentose phosphate pathway.
The catabolism of D-ribose in Escherichia coli has also been studied
extensively. It was elucidated that D-ribose is used as a carbon and energy
source through the pentose phosphate pathway (David & Wiesmeyer, 1970).
D-Ribose catabolism is initiated by transportation into the cells through two
distinct D-ribose transport systems (David & Wiesmeyer, 1970). Intracellular
D-ribose is phosphorylated by ribokinase (EC (David & Wiesmeyer,
1970) and catabolized further through the pentose phosphate pathway. The
high-affinity D-ribose transport system consists of a binding protein (Curtis,
1974; Willis & Furlong, 1974; Galloway & Furlong, 1977) and membranebound components (Iida et aI., 1984; Lopilato et ai., 1984). The D-ribose
binding protein serves not only as a component of the transport system but
also as the attractant receptor for the chemotactic response to D-ribose (Adler,
1969; Hazelbauer & Adler, 1971; Galloway & Furlong, 1977). The gene
encoding the D-ribose binding protein was recently cloned and sequenced
(Groarke et a/., 1983). The D-ribose binding protein of Salmonella typhimurium has also been described (Aksamit & Koshland, 1972, 1974).
The existence of the D-ribose operon (rbs) encoding for the genes of the
D-ribose binding protein, membrane-bound components, and the ribokinase of
E. coli K-12 was first elucidated by Anderson & Cooper (1970). However, the
rbs operon in E. coli B/r was determined to be located at 2 min and expressed
constitutively (Abou-Sabe & Richman, 1973) while, in E. coli K-12, it is
located at 735 min (now 83 min) and expressed inducibly (Anderson &
Cooper, 1970). The duplicate rbs operon at 2 min represents a transposition of
the 83-min region to 2 min in strain E. coli B/r (Abou-Sabe et al., 1982). A
silent copy of the 83-min rbs operon is also present in E. coli B/r (Abou-Sabe
et ai., 1982). Recently, the rbs operon was investigated by Lopilato et al.
The catabolism of D-ribose in some other bacteria has also been studied.
D-Ribose might be catabolized into the pentose phosphate pathway after being
transported and cleaved into C2 and C3 compounds by phosphoketolase (EC in Lactobacillus species, though the initial reaction was not elucidated
(Bernstein, 1953; Heath et ai., 1958). In Rhizobium meliloti (Duncan, 1981)

Microbial Production of D-Ribose


and R. leguminosarum (Dilworth et al., 1986), o-ribose is metabolized into the

pentose phosphate pathway after phosphorylation by ribokinase. Pseudomonads oxidized o-ribose into o-ribonic acid (Foster, 1944; Lockwood & Nelson,
1946; Weimberg, 1961). o-Ribose is converted to acetate and propionate in
Bacteroides ruminicola (Caldwell & Newman, 1986) and is converted to
o-ribulose by o-ribose isomerase (EC of Mycobacterium smegmatis
(Izumori et aI., 1975). Recently, the ribokinase gene of Staphylococcus hyicus
subsp. hyicus was cloned (Keller et al., 1984).
In yeasts, there are two metabolic pathways; one initiated by phosphorylation of o-ribose through ribokinase (Sable, 1950, 1952) and another by
reduction into ribitol (Barnett, 1976), which is oxidized into o-ribulose and,
after phosphorylation, metabolized into the pentose phosphate pathway.
o-Ribose is reduced by aldose reductase (EC from Pachysolen
tannophilus (Morimoto et al., 1987).
5.1 Wild-Type Micro-organisms
The first o-ribose accumulation was found in the culture medium of Penicillium
brevi-compactum (Simonart & Godin, 1951) during studies on the biosynthesis
of the components of benzene derivatives (Godin, 1953a,b). Since o-ribose
was detected by paper chromatography, the amount of accumulated o-ribose
was unknown. About 10 years later, an unidentified bacterium was reported to
accumulate both o-ribose and D-ribose 5-phosphate (Suzuki et al., 1963).
Pseudomonas reptilivola was also reported to accumulate o-ribose in a culture
medium at the rate of 012 mg of o-ribose per ml (Saito & Sugiyama, 1966).
All the micro-organisms used in these studies were wild-type and not much
o-ribose was accumulated. However, the amount of o-ribose produced by the
tkt mutants of Bacillus species was remarkable (Sasajima & Yoneda, 1971) and
the procedure, with subsequent improvement, has been used to commercially
produce o-ribose for riboflavin synthesis for 15 years. A detailed description of
the isolation and improvement of o-ribose-producing tkt mutants of Bacillus
species is given in the following section.
5.2 Transketolase Mutants (tkt Mutants) of Bacillus Species
5. 2.1 Isolation of tkt Mutants of Bacillus Species
Investigations on the production of the nucleotide flavor enhancers, inosine
monophosphate and guanosine monophosphate, by enzymatic degradation of
yeast RNA started in Japan in the late 1950s (Ogata, 1976). The process was
established and has been used for their commercial production ever since.
During this time and subsequently, the production of inosine monophosphate
and guanosine monophosphate or their precursors inosine and guanosine by


K. Sasajima & M. Yoneda

fermentation became a research subject in many academic and company

laboratories in Japan. Various types of mutants of purine nucleotide biosynthesis were isolated (Ogata et al., 1976) and have been used on the industrial
scale to produce these nucleotides. During the course of studies on improving
inosine-producing mutants of Bacillus species (Nogami et al., 1968), mutants
defective in the pentose phosphate pathway were isolated because it was
assumed that more o-ribose frame would be supplied for purine nucleotide
biosynthesis in such mutants (Sasajima et al., 1970). Contrary to expectation,
inosine productivity was not improved in such mutants. However, it was found
that a large amount of o-ribose was accumulated in the culture medium by the
tkt and o-ribulose 5-phosphate 3-epimerase (EC mutants of Bacillus
species (Sasajima & Yoneda, 1971, 1974b).
Although the reactions catalyzed by transketolase had been elucidated, it
was not simple to isolate tkt mutants. They had never been isolated at that
time because the pentose phosphate pathway is very complicated and,
moreover, plays a role in supplying the precursor of aromatic amino acids such
as L-tryptophan, L-tyrosine and L-phenylalanine and of aromatic vitamins such
as coenzyme Q, vitamin K, and folic acid (Fig. 3). In the beginning, mutants
which could not utilize L-arabinose or o-gluconate that are metabolized
through the pentose phosphate pathway were isolated. No tkt mutant could be
found among them, though aromatic amino acids were complemented in the
culture medium used for mutant isolation. However, a o-ribulose 5-phosphate
3-epimerase mutant was isolated during these studies (Sasajima et al., 1970;
Sasajima & Yoneda, 1974b).
Next, attempts were made to first isolate mutants requiring shikimic acid,
which is an intermediate of aromatic amino acids and vitamins (Fig. 3), and to
secondly select non-utilizers of o-gluconate as a carbon source among them.
Ten shikimic acid-requiring mutants were thus isolated and two of them were
found not to utilize o-gluconate as a carbon source (Sasajima et al., 1970).
These mutant strains were proved to be transketolase-deficient mutants by
enzymatic determinations (Sasajima & Yoneda, 1974a,b). The other shikimic
acid-requiring mutants were assumed to be those deficient in enzymes of
aromatic biosynthesis before shikimic acid. The ratio of 2 to 10 coincided
exactly with that anticipated. These tkt mutants accumulated about 40 mg of
o-ribose per ml in the culture medium (Sasajima & Yoneda, 1971). The
o-ribulose 5-phosphate 3-epimerase mutant produced a large amount of both
o-ribose and o-ribulose at the same time in a two to one ratio.
Josephson & Fraenkel (1969) independently isolated tkt mutants of
Escherichia coli, which did not accumulate any o-ribose (Fraenkel, D.G.,
1973, pers. comm.). Eidels & Osborn (1971) also isolated tkt mutants of
Salmonella typhimurium to elucidate the synthetic pathway of L-glycero-omanno-heptose, a component of the lipopolysaccharide in the outer membrane. However, it was not described whether o-ribose was produced by the
tkt mutants or not.
The tkt and o-ribulose 5-phosphate 3-epimerase mutants could not utilize

Microbial Production of D-Ribose


o-gluconate, o-xylose, and L-arabinose which are catabolized through the

pentose phosphate pathway (Sasajima & Yoneda, 1974b; Sasajima et al.,
1977). The tkt mutants among them, moreover, required aromatic amino acids
such as L-tryptophan, L-tyrosine and L-phenylalanine but, unexpectedly, no
aromatic vitamins for their growth (Sasajima & Yoneda, 1974b; Sasajima et
al., 1977). Aromatic vitamins might be synthesized from these complemented
aromatic amino acids.
As the revertants derived from the tkt mutants did not accumulate o-ribose,
it was concluded that their o-ribose accumulation depended on transketolase
mutations (Sasajima & Yoneda, 1971). The phenomenon was the first example
of the production of carbohydrates by carbohydrate metabolism mutant
The investigation on the isolation of the tkt mutants of other bacterial strains
such as Bacillus subtilis, B. pumilus, Brevibacterium thiogenitalis, B.
ammoniagenes, Arthrobacter globiformis, Aerobacter aerogenes and
Micrococcus denitrificans revealed that the tkt mutants of only Bacillus species
among them could accumulate o-ribose (Sasajima et aI., 1985b).
A difference in the productivity of pentoses was observed between strains of
Bacillus species. For example, the tkt mutants of a strain of Bacillus subtilis
IFO 3026 accumulated both o-ribose and o-ribulose at the same time (Sasajima
et aI., 1972; Sasajima, 1976; Sasajima & Yoneda, 1984; Sasajima et aI., 1985b).
The tkt mutants of another strain of B. subtilis IFO 12114 accumulated only
o-ribose (Sasajima & Yoneda, 1984). The difference in the kind of pentoses
accumulated might be due to the strain difference in kind or substrate
specificity of the phosphatases which would hydrolize intracellulariy accumulated o-ribose 5-phosphate and o-ribulose 5-phosphate to o-ribose and
o-ribulose during excretion.
Transketolase activity is functional when thiaminepyrophosphate and Mg2+
ion are coexistent (Wood, 1985). It was, therefore, considered that thiaminerequiring mutants would also accumulate o-ribose. This was shown to be so by
isolating the thiamine-requiring mutants of Bacillus subtilis IFO 12114 but the
amount of o-ribose accumulated was much less than that in the case of the tkt
mutants (Sasajima & Yoneda, 1984; Sasajima et al., 1985b). This is probably
because it is essential to add a small amount of thiamine to the culture medium
for enzymes other than trans keto lase which require it to function. Thus, it is
impossible to make transketolase activity complete!y zero by this mutation,
and consequently they produce a smaller amount of o-ribose than do the tkt

5.2.2 Improvement of rrribose-producing tkt mutants

The original tkt mutants accumulated o-ribose at the rate of about 40 mg/ml
(Sasajima & Yoneda, 1971). Improvement of these mutants by subsequent
mutations resulted in isolating high-level o-ribose-producing mutants which
produced about 70 mg of o-ribose per ml in the culture medium (Sasajima et
al., 1972, 1985b; Sasajima, 1976; Sasajima & Yoneda, 1984). Two mutation

K. Sasajima & M. Yoneda


steps at which the yield of o-ribose production was remarkably improved were
the following; (1) a mutation of the regulation of D-glucose dehydrogenase
(EC synthesis (Yokota et al., 1979; Yokota & Sasajima, 1981) and
(2) an asporogenous mutation (Sasajima et al., 1972, 1985b; Sasajima, 1976;
Sasajima & Yoneda, 1984)
D-Glucose dehydrogenase in Bacillus species, is known to be synthesized at
stage III of sporulation in parallel with morphological changes during spore
formation (Warren, 1968), exists only in spores (Fujita et al., 1977) and has the
role of supplying energy during germination (Otani et al., 1986). The initiation
of sporulation is also known to occur only after the carbon source is depleted.
In the improved mutant strain, the enzyme was found in vegetative cells while
no morphological changes concerning sporulation occurred (Yokota et al.,
1979; Yokota & Sasajima, 1981). It was presumed that o-glucose was
converted to o-ribose in the mutant strain through two metabolic pathways,
the non-phosphorylated pathway (from o-glucose to the pentose phosphates
via D-gluconate) and the phosphorylated pathway (the oxidative branch of the
pentose phosphate pathway) (Fig. 4). The gluconate pathway, probably,
contributed to the high D-ribose productivity of the improved strain. However,
there was also the possibility of o-gluconate as well as o-ribose accumulation.
The ratio of o-ribose and o-gluconate production depended on the partial
pressure of oxygen during fermentations, i.e. oxygen deficiency led to
D-gluconate rather than o-ribose accumulation (Sasajima & Yoneda, 1984).
D-Ribose accumulation by the mutant strain was about 60 mg/ml (Sasajima et
al., 1972, 1985b; Sasajima, 1976; Sasajima & Yoneda, 1984).
The second-step asporogenous mutation improved the capability of o-ribose
production up to 70 mg/ml (Sasajima et al., 1972, 1985b; Sasajima, 1976;
Sasajima & Yoneda, 1984). In this case, the mechanism for the improvement
was not clear. It might be a mutation by which carbohydrate metabolism is
changed in the direction of higher-level o-ribose accumulation. The high-yield

D-Glucose - - - - - - - o-Gluconate

o-Glucose-6-phosphate -


0- Ribulose-5-phosphate

0- Ribose-5-phosphate

Fig. 4. Pathways of o-ribose formation from o-glucose in high-yield mutant strains.

Microbial Production of D-Ribose


mutant strain has been used for 15 years to produce o-ribose for riboflavin
synthesis on an industrial scale.

5.2.3 Pleiotropic properties of D-ribose-producing tkt mutants

As already described, the tkt mutants not only could never utilize carbon
sources which are catabolized through the pentose phosphate pathway but also
required complementation of aromatic amino acids in the culture medium for
their growth. Moreover, it was revealed that the tkt mutants also had another
interesting property, i. e. functions concerning the cell surface changed
pleiotropically. The pleiotropic change is described below.
The first unexpected phenomenon was that the tkt mutants could not utilize
o-glucose as a sole carbon source (Sasajima & Yoneda, 1974b; Sasajima et at.,
1977) though o-glucose is the best carbon source for o-ribose production
(Sasajima & Yoneda, 1984; Sasajima et at., 1985b). The mutant strain seemed
preferentially to utilize other substances in the medium as a carbon and energy
source for growth, and later to become able to metabolize o-glucose for
o-ribose production. Investigation on the phenomenon revealed that the
phosphoenolpyruvate-dependent phosphotransferase system was defective
(Sasajima & Kumada, 1979); especially, the transport function of the Enz n glc
(phosphohistidinoprotein-hexose phosphotransferase; EC of the
PEP-dependent o-glucose phosphotransferase system (Roseman, 1969; Postma
& Roseman, 1976; Saier, 1977; Dills et at., 1980; Postma & Lengeler, 1985),
was defective in the tkt mutant.
Another biochemical change was found in the catabolite repression (Magasanik, 1961). The utilization of o-mannitol, o-fructose and glycerol was
inhibited by o-glucose while utilization of sorbitol was not (Sasajima &
Kumada, 1981a). The synthesis of the enzymes of o-mannitol catabolism, such
as o-mannitol-1-phosphate dehydrogenase (EC, and of the 0mannitol transport system was found to be hypersensitive to o-glucose
repression. Meanwhile, the synthesis of enzymes of sorbitol catabolism, such
as sorbitol dehydrogenase, and of the sorbitol transport system was resistant to
o-glucose repression. D-Gluconate, o-xylose, and L-arabinose were also found
to repress the synthesis of the enzymes of D-mannitol catabolism. These
changes in the regulation of enzyme synthesis might be caused by a defect in
the membrane.
Subsequent investigations on various functions concerning the cell surface
revealed; (1) the tkt mutant formed chains of cells during the exponential
growth phase; (2) the sensitivity to bacteriophages SPlO and SP01 was altered;
(3) the cell walls of the mutant strain autolyzed during incubation more slowly
than those of the parental strain, and (4) the sporulation frequency of the
mutant strain remarkably decreased (Sasajima & Kumada, 1981b, 1983a). The
mutant strain was non-motile because of a deficiency in flagellation (Sasajima
& Kumada, 1983b). The isolation and characterization of a true revertant
(transketolase reversion) confirmed that these pleiotropic properties were


K. Sasajima & M. Yoneda

generated by a single mutation in transketolase (Sasajima & Kumada, 1979,

The investigation on chemical changes in the membrane composition
revealed that the SDS-polyacrylamide gel electrophoresis pattern of the
membrane proteins and the thin layer chromatography pattern of the
membrane lipids were different in the parental and mutant strains (Sasajima et
al., 1985a).
The tkt mutants of Gram-negative bacteria such as Escherichia coli and
Salmonella typhimurium could not synthesize L-glycero-o-manno-heptose
(Eidels & Osborn, 1971), which is a component of the outer membrane
lipopolysaccharides and essential for a functional cell surface (Ames et al.,
1974; Koplow & Goldfine, 1974; Bayer et al., 1975; Irvin et al., 1975; Smit et
al., 1975; van Alphen et al., 1976). It is of interest that the tkt mutants of both
Gram-negative and Gram-positive bacteria are defective in cell surface
functions, though the mechanisms were probably different. It can be said that
transketolase plays an important role not only in carbohydrate metabolism and
aromatic biosynthesis but also in the cell surface functions of bacteria.
5. 2.4 Production of l-deoxy-ketoses by tkt mutants of Bacillus pumilus
During studies on the relation between o-ribose productivity and transketolase
mutation in various bacteria, it was found that the tkt mutants of Bacillus
pumilus produced a new monosaccharide, 1-deoxy-o-altro-heptulose (Yokota
et al., 1978). Further investigations revealed that 1-deoxy-o-altro-heptulose
7-phosphate was synthesized from o-ribose-5-phosphate and oL-acetoin
(Yokota & Sasajima, 1983). The accumulation of o-ribose 5-phosphate in the
cells of the tkt mutant strains presumably stimulated formation of 1-deoxy-oaltro-heptulose, which might accumulate in the culture medium after dephosphorylation during excretion (Yokota & Sasajima, 1983). Subsequent investigations revealed that cell-free extracts of various micro-organisms catalyzed
the formation of 1-deoxy-o-fructose, 1-deoxy-o-sorbose, 1-deoxy-o-tagatose,
1-deoxy-o-threo-pentulose, 1-deoxY-L-threo-pentulose and 1-deoxyerythrulose
(Yokota & Sasajima, 1984a, b). It is of much interest that, among them,
1-deoxy-o-threo-pentulose was described as a precursor of the thiazole ring of
thiamine (Therisod et al., 1981). This substance was also described as being
produced by Streptomyces hygroscopicus (Hoeksema & Baczynskyj, 1976;
Slechta & Johnson, 1976). It was elucidated that the formation of 1-deoxyketoses was catalyzed by pyruvate dehydrogenase (lipoamide) (EC
and acetoin dehydrogenase (Yokota & Sasajima, 1986). It was also found that
another similar enzyme, 2-ketoglutarate dehydrogenase (EC produced
2,3-dideoxy-hex-4-ulosonic acid (Yokota & Sasajima, 1985).

6.1 Inoculation


In the case of Penicillium brevi-compactum (Simonart & Godin, 1951), an

unidentified bacterium (Suzuki et al., 1963), and Pseudomonas reptilivora

Microbial Production of D-Ribose


Table 2
Inoculum Medium of Bacillus Species

Soluble starch
Corn steep liquor
K2 HP0 4
KH2 P04

Concentration (%)



(Saito & Sugiyama, 1966), the cells grown on bouillon agar slants were directly
inoculated into the fermentation medium. In the case of the tkt mutants of
Bacillus species (Sasajima & Yoneda, 1971), the inoculum culture incubated at
37C was used. The inoculum medium composition is shown in Table 2.
Adenosine and guanosine can be excluded from the medium in the case of
high-producing mutant strains whose requirements were compensated by
further reversion mutations.


Fermentation Media

The composition of the fermentation media used for o-ribose production are
shown in Tables 3, 4, 5 and 6 (Simonart & Godin, 1951; Suzuki et ai., 1963;
Saito & Sugiyama, 1966; Sasajima and Yoneda, 1971).
Suzuki et ai. (1963) described that the Mn2+ ion stimulated o-ribose
accumulation while o-ribose 5-phosphate was accumulated more in the absence
of Mn2+ ion in the case of an unidentified bacterium. Fe2+ and Zn 2+ ions are
also effective in stimulating o-ribose accumulation.
In the case of Pseudomonas reptilivora, o-ribose accumulation from various
carbon sources other than o-glucose was observed (Saito & Sugiyama, 1966).
o-Mannose, o-fructose and o-gluconic acid were found to be as good substrates
as o-glucose.

Table 3
Medium of Penicillium

NaN0 3
KH2 P04


Concentration (%)




K. Sasajima & M. Yoneda



Table 4
Medium of an

K zHP04
MgS0 47HzO
CaClz2H2 0
Yeast extract


Concentration (%)

In the case of Bacillus species, it was also found that various carbon sources
were useful for D-ribose accumulation as shown in Table 7 (Sasajima &
Yoneda, 1971). D-Glucose, D-mannose, sorbitol, D-mannitol, maltose, and
lactose were good substrates. Various kinds of natural nitrogen sources as well
as dried yeast are useful (Table 8) (Sasajima & Yoneda, 1971). Corn steep
liquor-based medium was used for a large-scale production (Asai et ai., 1978).
It was necessary to complement aromatic amino acids to the medium in the
case where corn steep liquor was used as nitrogen source (Asai et ai., 1978).
Table 5
Fermentation Medium of Pseudomonas
KH2 P0 4
Sodium citrate
Corn steep liquor
Yeast extract

Concentration (%)

Table 6
Fermentation Medium of Bacillus Species
Ca3 (P04 )z
Dried yeast

Concentration (%)
125; 150
(pH 70)


Microbial Production of D-Ribose

Table 7
D-Ribose Formation from Various Carbon Sources

Carbon source

Soluble starch

D-Ribose formed (mg/ml)

Strain Shi5"

Strain Shi7"






"Strains SHi5 and Shi7 are tkt mutants of a Bacillus

species (Sasajima & Yoneda, 1971).
Table 8
Efficiency of Various Nitrogen Sources for D-Ribose

Nitrogen source

Dried yeast
Meat extract
Com steep liquor
Yeast extract
Cotton seed meal
Cotton seed flour
Com gluten meal
Soy bean meal

D-Ribose formed (mg/ml)







Strain S27, a transformant from strain Shi7 which compensated for both adenine and xanthine requirements
(Sasajima & Yoneda, 1971). D-Glucose was used as the
carbon source.

6.3 Fennentation Course

In the case of Bacillus species tkt mutants, D-ribose accumulation started at 1
or 2 days in proportion to D-glucose assimilation and ceased at 4 days
(Sasajima & Yoneda, 1971). The cell population reached about 1010 cells/ml in
2 days. Thus, D-ribose was accumulated during the stationary phase. This
implied that only the aged cells excreted D-ribose into the medium.
A large-scale fermentation course using a high-level producing strain is


K. Sasajima & M. Yoneda







;f. 10

~ QI
::> '"



:;: I


.- ::>


I .;;;


L -_ _ _ _L...'-_ ..




Time (hr)

Fig. 5. Time course of n-ribose fermentation.

shown in Fig. 5 (Sasajima et al., 1972, 1985b; Sasajima, 1976; Sasajima &
Yoneda, 1984).
As already described in the section on 'Producing micro-organisms', the
improved strains have the ability to produce n-gluconate as well as n-ribose.
n-Ribose is made from n-glucose through two pathways: one is a nonphosphorylated pathway via n-gluconate and the other is a phosphorylated
pathway via n-glucose 6-phosphate (Fig. 4). Both n-gluconate and n-glucose
6-phosphate are converted to 6-phospho-n-gluconate, the common intermediate of both pathways, and finally converted to n-ribose.
Appropriate culture conditions are required for the smooth conversion of
n-gluconate to 6-phospho-n-gluconate. Otherwise, n-gluconate accumulates in
the culture medium resulting in a reduced production of n-ribose (Asai et al.,
1978; Sasajima & Yoneda, 1984). Effective factors are the pH of media, an
oxygen concentration in the medium enough for three oxidative reactions in
the pathway from n-glucose to n-ribose (Fig. 4) (Sasajima & Yoneda, 1984),
and the temperature of the culture (Asai et al., 1978). The maximum yield of
n-ribose was noted at 366C when exponential growth cells were used as the
inoculum (Asai et al., 1978).
In the course of preparing n-ribose by chemical or enzymatic hydrolysis of
RNA, n-ribose was first converted to the p-bromophenylhydrazone (van

Microbial Production of D-Ribose


Ekenstein & Blanksma, 1913; Steiger, 1936) or the arylamide riboside (Berger
& Lee, 1944; Berger et al., 1944). After subsequent purification, o-ribose was
recovered by hydrolysis (Jeanlotz & Fletcher, 1951; Overend & Stacey, 1955).
In the case of preparing o-ribose by enzymatic hydrolysis of 5-amino-4imidazole-carboxamide-riboside with a nucleosidase, o-ribose was separated
and purified from the other product, 5-amino-4-imidazole-carboxamide, by
column chromatography with ion-exchange resin Dowex 50 (H+) (Sano et al.,
1977b). Finally, o-ribose was obtained as the lyophilized form.
Saito & Sugiyama (1966) purified o-ribose from the culture medium of
Pseudomonas reptilivora by first separating microbial cells by centrifugation
and subsequent column chromatographies with chromatopile and strong-base
anion resin (borate form). Finally, they obtained o-ribose in crystalline form
from the syrup.
Sasajima & Yoneda (1971) purified and isolated o-ribose from the culture
medium of Bacillus species tkt mutant by removing microbial cells, column
chromatography with carbon, yeast treatment to assimilate the residual
substrate o-glucose and excluding metal and anionic ions with ion exchange
resins, amberlite IR120 and amberlite IRA400. o-Ribose was crystallized by
adding ethanol to the syrupy solution.
Hough et al. (1948, 1949) described the separation of o-ribose from
coexisting sugars by partition chromatography with a powdered cellulose


o-Ribose can be determined by reagents for pentoses, e.g. orcinol reagent

(Bial, 1902, 1903; Mejbaum, 1939; Miller et al., 1951; Dische, 1962b), carbazol
reagent (Dische, 1962a) and phloroglucinol reagent (Dische, 1962b ). By
orcinol reagent o-ribose in the culture medium is overestimated because of
coexisting substances in the culture medium (Sasajima & Yoneda, 1971).
o-Ribose can be also assayed by densitometry after paper partition chromatography (Sasajima & Yoneda, 1971), and can be determined by gas-liquid
chromatography (Laine & Sweeley, 1973; Petersson, 1974).


The preparative method of o-ribose by chemical synthesis was first described

by van Ekenstein & Blanksma (1913). In 1936, Steiger repeated this process
(Fig. 6). The starting material was prepared from calcium o-gluconate
according to the method of Hockett & Hudson (1934). The process was later
modified by Berezovskii and Rodionova (1953).
Another method of o-ribose synthesis was described by Gehrke & Aichner
(1927). This process was improved by Austin & Humoller (1934) in L-ribose


K. Sasajima & M. Yoneda


boiling aqueous
electrolyticaloxidation; HCOH _..<:.pyn<.:;"=di::.:.:ne,--+) HCO H __

tH2 0H


6H 20H

o-arabonic acid

o-ribonic acid



amalgam ..


Fig. 6. Chemical synthesis of D-ribose according to the method of van Ekenstein &
B1anksma (1913).


Zn dust

acetic aCid)



I H 0








CH 20H

Fig. 7. Chemical synthesis of D-ribose according to the method of Gehrke & Aichner

Microbial Production of D-Ribose


CH2 0H




6H s
H 2 C-O



3,5-benzylidene-l-deoxy-l- I-nitro-l-deoxyD-ribose
Fig. 8. Chemical synthesis of D-ribose according to the method of Sowden (1950).

H 2 C-O"

/CH 3
"---CH 3


H 2 C-O,
/,CH 3
"CH 3
CH 3S0 2 0CH



HC-O....... y,,/CH 3
H 2 C-O/
"CH 3
H 2 C-O
"---CH 3
1,2: 5,6-diisopropylidene-o-glucose 3-0-mesyl-l,2: 5,6-diisopropylidene-o-glucose
CH 3S0 2 0CH


CH 3S02 0CH





CH 20H



CH 2 0H

CH 20H
3-0-mesyl-o-glucose 2-0-mesyl-4-0-formyl-D-arabinose o-ribose
Fig. 9. Chemical synthesis of D-ribose according to the method of Smith (1955).


K. Sasajima & M. Yoneda

preparation and was used to prepare o-ribose for the chemical synthesis of
riboflavin (Karrer et al., 1935b; Kuhn et al., 1935b). The process includes
arabinal as an intermediate as shown in Fig. 7.
The formation of o-ribose from o-glucose was described by Sowden (1950)
and Smith (1955) as shown in Figs 8 and 9.
Stroh et al. (1964) compared the above four methods and concluded that the
best yield was obtained by the method of Smith (1955). The overall yield was
o-Ribose has long been prepared for the chemical synthesis of riboflavin from
o-glucose by one of the chemical processes described above. However, a
microbial production process was developed in the early 1970s (Sasajima et al.,
1972, 1985b; Sasajima, 1976; Sasajima & Yoneda, 1984). The procedure, using
the tkt mutant of Bacillus species is economical and prevents environmental
pollution by avoiding the use of mercury, which is essential for the electrolytic
reduction required in the chemical process (Harris et al., 1972). The industrial
microbial production of o-ribose for the chemical synthesis of riboflavin started
in 1973; it has given substantial results for about 15 years. Some companies in
the world have been using this process to produce o-ribose for the synthesis of
riboflavin, which is used not only for pharmaceuticals but also for animal feed
additives worldwide in large quantities.
o-Ribose itself has been also used to treat some cases of heart diseases. It
was suggested that administered o-ribose bypasses the pentose phosphate
pathway and increases the size of the phosphoribosylpyrophosphate pool
resulting in stimulation of nucleotide synthesis, maintenance of adenine
nucleotide levels, and protection of the myocardium in various heart diseases
(Zimmer et al., 1984). Exercise-induced muscle pain and stiffness due to
myoadenylate deaminase deficiency was also successfully treated with o-ribose
(ZOllner et al., 1986).
D-Ribose is a natural component of biologically important substances such as
nucleic acids and coenzymes. It has been also found in many nucleoside
antibiotics. Nucleoside analogs have been developed as antiviral agents. It is
hoped that o-ribose can be used in the chemical synthesis of drugs to treat viral
and bacterial infectious diseases in the future.
Abou-Sabe, M. & Richman, J. (1973). On the regulation of o-ribose metabolism in E.
coli B/r. II. Chromosomal location and fine structure analysis of the o-ribose
permease and o-ribokinase structural genes by PI transduction. Molec. Gen. Genet.,
122, 303-12.

Microbial Production of D-Ribose


Abou-Sabe, M., Pilla, J., Hazuda, D. & Ninfa, A. (1982). Evolution of the D-ribose
operon of Escherichia coli B/r. J. Bacteriol., 150, 762-9.
Adler, J. (1969). Chemoreceptors in bacteria. Science, 166, 1588-97.
Aksamit, R. & Koshland, D. Jr (1972). A ribose binding protein of Salmonella
typhimurium. Biochem. biophys. Res. Commun., 48, 1348-53.
Aksamit, R. R. & Koshland, D. E. Jr (1974). Identification of the ribose binding
protein as the receptor for ribose chemotaxis in Salmonella typhimurium.
Biochemistry, 13, 4473-8.
Alphen, W. van, Lugtenberg, B. & Berendsen, W. (1976). Heptose-deficient mutants
of Escherichia coli K12 deficient in up to three major outer membrane proteins.
Molec. Gen. Genet., 147,263-9.
Amemura, A., Hisamatsu, M., Ghai, S. K. & Harada, T. (1981). Structural studies on
a new polysaccharide, containing D-riburonic acid, from Rhizobium me/iloti IFO
13336. Carbohydr. Res., 91,59-65.
Ames, G. F.-L., Spudich, E. N. & Nikaido, H. (1974). Protein composition of the
outer membrane of Salmonella typhimurium: effect of lipopolysaccharide mutations.
J. Bacteriol., 117,406-16.
Anderson, A. & Cooper, R. A. (1970). Biochemical and genetical studies on ribose
catabolism in Escherichia coli. J. Gen. Microbiol., 62,335-9.
Asai, T., Doi, M., Kono, T. & Fukuda, H. (1978). Kinetic study on the production of
D-ribose by Bacillus sp. J. Ferment. Technol., 56,91-5.
Austin, W. C. & Humoller, F. L. (1934). The preparation of I-ribose. J. Am. Chem.
Soc., 56, 1152-3.
Avery, O. T., MacLeod, C. M. & McCarty, M. (1944). Studies on the chemical nature
of the substance inducing transformation of pneumococcal types. J. expo Med., 79,
Baddiley, J., Thain, E. M., Novelli, G. D. & Lipmann, F. (1953). Structure of
coenzyme A. Nature, Lond., 171, 76.
Baker, B. R. & Schaub, R. E. (1953). Achromycin. Synthetic studies. III. Synthesis of
3-amino-D-ribose, a hydrolytic product. J. Am. Chem. Soc., 75, 3864-5.
Barnett, J. A. (1976). The utilization of sugars by yeasts. Adv. Carbohydr. Chem.
Biochem., 32, 125-234.
Bayer, M. E., Koplow, J. & Goldfine, H. (1975). Alterations in envelope structure of
heptose-deficient mutants of Escherichia coli as revealed by freeze-etching. Proc.
natn. Acad. Sci. U.S.A., 72,5145-9.
Beaven, G. R., Holiday, E. R., Johnson, E. A., Ellis, B., Mamalis, P., Petrow, V. &
Sturgeon, B. (1949). The chemistry of anti-pernicious anaemia factors. Part III.
5: 6-Disubstituted benziminazoles as products of acid hydrolysis of vitamin B 12 J.
Pharm. Pharmac., 1,957-70.
Bentley, H. R., Cunningham, K. G. & Spring, F. S. (1951). Cordycepin, a metabolic
product from cultures of Cordyceps militaris (Linn.) Link. Part II. The structure of
cordycepin. J. Chem. Soc., 1951,2301-5.
Berezovskii, V. M. & Rodionova, E. P. (1953). Transformations and synthesis of
carbohydrates. VII. Synthesis of D-ribose. Sbornik Statel Obshchel Khim., 2,
939-43; Chem. Abstr., 49,6838 (1955).
Berger, L. & Lee, J. (1944). Arylamine-N-glycosides. Part II. Arylarnine-N-pentosides
and complex salt formation studies. J. Org. Chem., 11, 84-90.
Berger, L., Solmssen, U. V., Leonard, F., Wenis, E. & Lee, J. (1944). Arylamine-Nglycosides. Part III. Hydrolysis of arylamine-N-pentosides and the preparation of
crystalline D-ribose. J. Org. Chern., 11,91-4.
Bernstein, I. A. (1953). Fermentation of ribose-C4 by Lactobacillus pentosus. J. BioI.
Chem., 205, 309-16.
Bial, M. (1902). Die Diagnose der Pentosurie. Deut. med. Wochschr., 28,


K. Sasajimo & M. Yoneda

Bial, M. (1903). Ueber die Diagnose der Pentosurie mit dem von mir angegebenen
Reagens. Deut. med. Wochschr., 29,477-8.
Borman, A. H., Jong, E. W. de, Thierry, R., Westbroek, P., Bosch, L., Gruter, M. &
Kamerling, J. P. (1987). Coccolith-associated polysaccharides from cells of
Emiliania huxleyi (Haptophyceae). J. Phycol., 23, 118-23.
Bredereck, H. & Rothe, G. (1938). Nucleinsauren, VI. Mitteil: Darstellung der
Nucleoside durch enzymatische Hydrolyse der Hefenucleinsaure; zugleich ein
Beitrag zur Darstellung der d-Ribose. Ber., 71,408-11.
Bredereck, H., Kothnig, M. & Berger, E. (1940). Uber die d-Ribose (Darstellung
einer krystallisierten Anhydroribose). Ber., 73,956-62.
Brink, N. G. & Folkers, K. (1949). Vitamin B 12 . VI. 5 ,6-Dimethylbenzimidazole , a
degradation product of vitamin B12 . J. Am. Chem. Soc., 71, 295l.
Brink, N. G. & Folkers, K. (1950). Vitamin B 12 . X. 5 ,6-Dimethylbenzimidazole , a
degradation product of vitamin B 12 . J. Am. Chem. Soc., 72, 4442-3.
Brink, N. G., Holly, F. W., Shunk, C. H., Peel, E. W., Cahill, J. J. & Folkers, K.
(1950). Vitamin B 12 . IX. 1-a-Ribofuranosido-5,6-dimethylbenzimidazole, a degradation product of vitamin B 12 . J. Am. Chem. Soc., 72, 1866.
Buchanan, J. G., Johnson, A. W., Mills, J. A. & Todd, A. R (1950a). The isolation of
a phosphorus-containing degradation product from vitamin B 12c Chem. Ind., 1950,
Buchanan, J. G., Johnson, A. W., Mills, J. A. & Todd, A. R. (1950b). Chemistry of
the vitamin BI2 group. Part I. Acid hydrolysis studies. Isolation of a phosphoruscontaining degradation product. J. Chem. Soc., 1950, 2845-55.
Caldwell, D. R & Newman, K. (1986). Pentose metabolism by Bacteroides ruminicola
subsp. brevis strain BA. Curro Microbio/., 14, 149-55.
Cantor, S. M. & Peniston, Q. P. (1940). The reduction of aldoses at the dropping
mercury cathode: estimation of the aldehyde structure in aqueous solutions. J. Am.
Chem. Soc. 62, 2113-2l.
Caputto, R, Leloir, L. F., Cardini, C. E. & Paladini, A. C. (1950). Isolation of the
coenzyme of the galactose phosphate-glucose phosphate transformation. J. Bioi.
Chem., 184, 333-50.
Chargaff, E. & Lipshitz, R (1953). Composition of mammalian desoxyribonucleic
acids. J. Am. Chem. Soc., 75, 3658-6l.
Chargaff, E., Vischer, E., Doniger, R, Green, C. & Misani, F. (1949). The
composition of the desoxypentose nucleic acids of thymus and spleen. J. Bioi.
Chem., 177,405-16.
Cherbuliez, E. & Bernhard, K. (1932a). Recherches sur la graine de croton. I. Sur Ie
crotonoside (2-oxy-6-amino-purine-D-riboside). Helv. Chim. Acta, 1S,464-7l.
Cherbuliez, E. & Bernhard, K. (1932b). Remarques sur Ie crotonoside. Helv. Chim.
Acta, IS, 978-80.
Curtis, S. J. (1974). Mechanism of energy coupling for transport of D-ribose in
Escherichia coli. J. Bacterio/., 120,295-303.
David, J. & Wiesmeyer, H. (1970). Regulation of ribose metabolism in Escherichia
coli. I. The ribose catabolic pathway. Biochim. biophys. Acta, 208,45-55.
Davis, A. R, Newton, E. B. & Benedict, S. R (1922). The combined uric acid in beef
blood. J. Bioi. Chem., 54,595-9.
Dickens, F. (1938a). Oxidation of phosphohexonate and pentose phosphoric acids by
yeast enzymes. I. Oxidation of phosphohexonate. II. Oxidation of pentose
phosphoric acids. Biochem. J., 32, 1626-44.
Dickens, F. (1938b). Yeast fermentation of pentose phosphoric acids. Biochem. J., 32,
Dills, S. S., Apperson, A., Schmidt, M. R & Saier, M. H. Jr (1980). Carbohydrate
transport in bacteria. Microbiol. Rev., 44,385-418.

Microbial Production of D-Ribose


Dilworth, M. J., Arwas, R., McKay, I. A., Saroso, S. & Glenn, A. R. (1986). Pentose
metabolism in Rhizobium leguninosarum MNF300 and in cowpea Rhizobium
NGR234. J. Gen. Microbiol., 132,2733-42.
Dische, Z. (1962a). Color reaction of ketoses with carbazole and sulfuric acid. In
Methods in Carbohydrate Chemistry, Vol. I, ed. R. L. Whistler, M. L. Wolfrom, J.
N. BeMiller & F. Shafizadeh. Academic Press, New York, London, pp. 481-2.
Dische, Z. (1962b). Color reactions of pentoses. In Methods in Carbohydrate
Chemistry, Vol. I, ed. R. L. Whistler, M. L. Wolfrom, J. N. BeMiller & F.
Shafizadeh. Academic Press, New York, London, pp. 484-8.
Doly, J. & Mandel, P. (1967). Mise en evidence de la biosyntbese in vivo d'un
polymere compose, Ie polyadenosine diphosphoribose dans les noyaux de foie de
poulet. C. R. hebd. Slanc. Acad. Sci., Paris, Ser. D., 264,2687-90.
Duncan, M. J. (1981). Properties of tn5-induced carbohydrate mutants in Rhizobium
meliloti. J. Gen. Microbiol., 122,62-7.
Eidels, L. & Osborn, M. J. (1971). Lipopolysaccharide and aldoheptose biosynthesis in
transketolase mutants of Salmonella typhimurium. Proc. natn. Acad. Sci. U.S.A.,
Ekenstein, W. A. van & Blanksma, J. J. (1913). Over o-ribose. Chem. Weekblad., 10,
Euler, H. von & Vestin, R. (1935). Zur Kenntnis der Wirkungen der Co-Zymase. Z.
physiol. Chem., 237, 1-5.
Euler, H. von & Schlenk, F. (1937). Co-Zymase. Z. physiol. Chem., 246,64-82.
Euler, H. von, Karrer, R., Malmberg, M., Schopp, K., Benz, F., Becker, B. & Frei, P.
(1935). Synthese des Lactoflavins (Vitamin B2 ) und anderer Flavine. Helv. Chim.
Acta, 18,522-35.
Euler, H. von, Karrer, P. & Usteri, E. (1942). Der Zucker der Cozymase. Helv. Chim.
Acta, 25, 323-5.
Foster, J. W. (1944). Microbiological aspects of riboflavin. III. Oxidation studies with
Pseudomonas riboflavina. J. Bacteriol., 48,97-111.
Fujita, Y., Ramaley, R. & Freese, E. (1977). Location and properties of glucose
dehydrogenase in sporulating cells and spores of Bacillus subtilis. J. Bacteriol., 132,
Galloway, D. R. & Furlong, C. E. (1977). The role of ribose-binding protein in
transport and chemotaxis in Escherichia coli K12. Archs. Biochem. Biophys., 184,
Gehrke, M. & Aichner, F. X. (1927). Uber das Arabinal. Ber., 60,918-22.
Gerber, N. N. & Lechevalier, H. A. (1962). 3'-Amino-3'-deoxyadenosine, an
antitumor agent from Helminthosporium sp. J. Org. Chem., 27, 1731-2.
Gmeiner, J. (1977). The ribitol-phosphate-containing lipopolysaccharide from Proteus
mirabilis, strain D52. Investigations on the structure of O-specific chains. Eur. J.
Biochem., 74, 171-BO.
Gmeiner, J., Mayer, H., Fromme, I., Kotelko, K. & Zych, K. (1977). Ribitolcontaining lipopolysaccharides from Proteus mirabilis and their serological relationship. Eur. J. Bioc;hem., 72, 35-40.
Godin, P. (1953a). Etude du metabolisme ternaire de Penicillium brevi-compactum. I.
Analyse chromatographique du milieu de culture. Biochim. biophys. Acta, 11,
Godin, P. (1953b). Etude du metabolisme ternaire de Penicillium brevi-compactum. II.
Analyse chromatographique de la solution glucosee substitutee au milieu de culture
initial. Biochim. biophys. Acta, 11, 119-21.
Gomez, L. D. & Wallace, J. W. (1986). Flavonoids of Phlebodium. Biochem. Syst.
Ecol., 14,407-8.
Greenberg, D. M. (1969). Biosynthesis of amino acids and related compounds. In


K. Sasajima & M. Yoneda

Metabolic Pathways, Vol. III, ed. D. M. Greenberg. Academic Press, New York,
London, pp. 237-315.
Groarke, J. M., Mahoney, W. C., Hope, J. N., Furlong, C. E., Robb, F. T., Zalkin,
H. & Hermodson, M. A. (1983). The amino acid sequence of o-ribose-binding
protein from Escherichia coli K12. I. Bioi. Chem., 258, 12952-6.
Guarino, A. J. & Kredich, N. M. (1963). Isolation and identification of 3'-amino-3'deoxyadenosine from Cordyceps militaris. Biochim. biophys. Acta, 68,317-19.
Hackman, R. H. & Trikojus, V. M. (1952). The composition of the honeydew excreted
by Australian coccids of the genus Ceroplastes. Biochem. I., 51, 653-6.
Harris, R. S., Wagner-Jauregg, T., Horwitt, M. K. & Witting, L. A. (1972).
Riboflavin. In The Vitamins, Vol. V (2nd edn), ed. W. H. Sebrell & R. S. Harris.
Academic Press, New York, London, pp. 1-96.
Hartman, S. C. (1970). Purines and pyrimidines. In Metabolic Pathways, Vol. IV, ed.
D. M. Greenberg. Academic Press, New York, London, pp. 1-68.
Hazelbauer, G. L. & Adler, J.' (1971). Role of the galactose binding protein in
chemotaxis of Escherichia coli toward galactose. Nature New Bioi., 230, 101-4.
Heath, E. C., Hurwitz, J., Horecker, B. L. & Ginsburg, A. (1958). Pentose
fermentation by Lactobacillus plantarum. I. The cleavage of xylulose 5-phosphate
by phosphoketolase. I. Bioi. Chem., 231, 1009-29.
Hershey, A. D. & Chase, M. (1952). Independent functions of viral protein and nucleic
acid in growth of bacteriophage. I. Gen. Physiol., 36,39-56.
Hockett, R. C. & Hudson, C. S. (1934). Improvements in the preparation of
o-arabinose from calcium gluconate. I. Am. Chem. Soc., 56, 1632-3.
Hoeksema, H. & Baczynskyj, L. (1976). A new metabolite from Streptomyces
hygroscopicus, II. Identification as 1-deoxy-o-threo-pentulose. I. Antibiot., XXIX,
Hoffman, J., Lindberg, B., Hofstad, T. & Lygre, H. (1977). Structural studies of the
polysaccharide antigen of Eubacterium saburreum, strain L.452. Carbohydr. Res.,
Hofstad, T. & Lygre, H. (1977). Composition and antigenic properties of a surface
polysaccharide isolated from Eubacterium saburreum, strain L452. Acta Pathol.
Microbiol. Immunol. Scand. Section B, 85, 14-17.
Holiday, E. R. & Petrow, V. (1949). Vitamin B12 as a 5:6-dimethylbenziminazole
derivative. I. Pharm. Pharmac., 1,734-5.
Hough, L., Jones, J. K. N. & Wadman, W. H. (1948). Application of paper partition
chromatography to the separation of the sugars and their methylated derivatives on
a column of powdered cellulose. Nature, Lond., 162, 448.
Hough, L., Jones, J. K. N. & Wadman, W. H. (1949). Quantitative analysis of
mixtures of sugars by the method of partition chromatography. Part IV. The
separation of the sugars and their methylated derivatives on columns of cellulose. I.
Chem. Soc., 1949,2511-16.
!ida, A., Harayama, S., lino, T. & Hazelbauer, G. L. (1984). Molecular cloning and
characterization of genes required for ribose transport and utilization in Escherichia
coli K-12. I. Bacteriol., 158,674-82.
Irvin, R. T., Chatterjee, A. K., Sanderson, K. E. & Costerton, J. W. (1975).
Composition of the cell envelope structure of a lipopolysaccharide-defective
(heptose-deficient) strain and a smooth strain of Salmonella typhimurium. I.
Bacteriol., 124,930-41.
Izumori, K., Rees, A. W. & Elbein, A. D. (1975). Purification, crystallization, and
properties of o-ribose isomerase from Mycobacterium smegmatis. I. Bioi. Chem.,
250, 8085-7.
Jeanlotz, R. W. & Fletcher, H. G. Jr (1951). The chemistry of ribose. Adv. Carbohydr.
Chem. Biochem., 6, 135-74.

Microbial Production of D-Ribose


Josephson, B. L. & Fraenkel, D. G. (1969). Transketolase mutants of Escherichia coli.

1. Bacteriol., 100, 1289-95.
Karrer, P., Schopp, K. & Benz, F. (1935a). Synthesen von Flavin IV. Helv. Chim.
Acta, 18, 426-9.
Karrer, P., Becker, B., Benz, F., Frei, P., Salomon, H. & Schopp, K. (1935b). Zur
Synthese des Lactoflavins. Helv. Chim. Acta. 18,1435-48.
Kauffmann, F., Jann, B., KrUger, L., Liideritz, D. & Westphal, O. (1962). Zur
Immunochemie der O-Antigen von Enterobacteriaceae. VIII. Analyse der Zuckerbausteine von Polysacchariden weiterer Salmonella- und Arizona-O-Gruppen.
Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Abt. 1: Orig., 186, 509-16.
Keenan, G. T. (1926). The optical properties of sugars. 1. Wash. Acad. Sci., 16,
Keller, G., Schleifer, K. H. & Gotz, F. (1984). Cloning of the ribokinase gene of
Staphylococcus hyicus subsp. hyicus in Staphylococcus carnosus. Archs Microbiol.,
140, 218-24.
Klein, J. R. & Kohn, H. I. (1940). The synthesis of flavin-adenine dinucleotide from
riboflavin by human blood cells in vitro and in vivo. 1. Bioi. Chem., 136, 177-89.
Koplow, J. & Goldfine, H. (1974). Alterations in the outer membrane of the cell
envelope of heptose-deficient mutants of Escherichia coli. 1. Bacteriol., 117,527-43.
Kuhn, L. P. (1950). Infrared spectra of carbohydrates. Analyt. Chem., 22,276-83.
Kuhn, R., Reinemund, K., Kaltschmitt, H., Strobele, R. & Trischmann, H. (1935a).
Synthetisches 6,7-Dimethyl-9-d-ribo-flavin. Naturwissenschaften, ~, 260.
Kuhn, R., Reinemund, K., Weygand, F. & Strobele, R. (1935b). Uber die Synthese
des Lactoflavins (Vitamin B2). Ber., 68, 1765-74.
Laine, R. A. & Sweeley, C. C. (1973). O-Methyl oximes of sugars. Analysis as
O-trimethylsilyl derivatives by gas-liquid chromatography and mass spectrometry.
Carbohydr. Res., 27, 199-213.
Levene, P. A. (1935). Note on the preparation of crystalline o-mannose and of
crystalline o-ribose. 1. Bioi. Chem., 108,419-20.
Levene, P. A. & Clark, E. P. (1921). d-Ribohexosaminic acids. 1. BioI. Chem., 46,
Levene, P. A. & Jacobs, W. A. (1909). Uber Inosinsiiure. Ber., 42, 1198-203.
Levene, P. A., Mikeska, L. A. & Mori, T. (1930). On the carbohydrate of
thymonucleic acid. 1. Bioi. Chem., 85, 785-7.
Lockwood, L. B. & Nelson, G. E. (1946). The oxidation of pentoses by Pseudomonas.
1. Bacteriol., 52, 581-6.
Lopilato, J. E., Garwin, J. L., Emr, S. D., Silhavy, T. J. & Beckwith, J. R. (1984).
o-Ribose metabolism in Escherichia coli K-12: genetics, regulation and transport. 1.
Bacteriol., 158, 665-73.
Magasanik, B. (1961). Catabolite repression. Cold Spring Harbor Symp. Quant. Bioi.,
Mandel, J. A. & Dunham, E. K. (1912). Preliminary note on a purine-hexose
compound. 1. Bioi. 9hem., 11, 85-6.
Mejbaum, W. (1939). Uber die Bestimmung kleiner Pentosemengen, insbesondere in
Derivaten der Adenylsiiure. Z. Physiol. Chem., 258, 117-20.
Miller, G. L., Golder, R. H. & Miller, E. E. (1951). Determination of pentoses.
Analyt. Chem., 23,903-5.
Miyano, K., Ishibashi, M., Kunita, N., Takeda, Y. & Miwatani, T. (1983). Demonstration of the presence of alditols and galacturonic acid in Vibrio parahaemolyticus
010 lipopolysaccharide. FEMS Microbiol. Lett., 20,225-8.
Morimoto, S., Tawaratani, T., Azuma, K., Oshima, T. & Sinskey, A. J. (1987).
Purification and properties of aldose reductase from Pachysolen tannophilus. 1.
Ferment. Technol., 65, 17-21.


K. Sasajima & M. Yoneda

Nogami, I., Kida, M., Iijima, T. & Yoneda, M. (1968). Studies on the fermentative
production of purine derivatives. Part I. Derivation of guanosine and inosineproducing mutants of a Bacillus strain. Agric. Bioi. Chem., 32, 144-52.
Ochoa, S. & Rossiter, R. J. (1939). Flavin-adenine-dinucleotide in rat tissues.
Biochem. J., 33,2008-16.
Ogata, K. (1976). History and prospects. In Microbial Production of Nucleic AcidRelated Substances, ed. K. Ogata, S. Kinoshita, T. Tsunoda & K. Aida. Kodansha
Ltd, Tokyo and John Wiley & Sons, New York, London, Sydney, Toronto, pp.
Ogata, K., Kinoshita, S., Tsunoda, T. & Aida, K. (1976). Microbial Production of
Nucleic Acid-Related Substances. Kodansha Ltd., Tokyo and John Wiley & Sons,
New York, London, Sydney, Toronto.
Otani, M., Ihara, N., Umezawa, C. & Sano, K. (1986). Predominance of gluconate
formation from glucose during germination of Bacillus megaterium OM B1551
spores. J. Bacteriol., 167, 148-52.
Overend, W. G. & Stacey, M. (1955). Chemistry of ribose and deoxyribose. In The
Nucleic Acids, Vol. I, ed. E. Chargaff & J. N. Davidson. Academic Press, New
York, pp. 9-80.
Perlman, D. (1979). Ketogenic fermentation processes. In Microbial Technology, Vol.
II, ed. H. J. Peppler & D. Perlman. Academic Press, New York, San Francisco,
London, pp. 173-7.
Petersson, G. (1974). Gas-chromatographic analysis of sugars and related hydroxy
acids as acyclic oxime and ester trimethylsilyl derivatives. Carbohydr. Res., 33,
Phelps, F. P., Isbel, H. S. & Pigman, W. (1934). Mutarotation of beta-d-ribose and
beta-I-ribose. J. Am. Chern. Soc., 56, 747-8.
Postma, P. W. & Lengeler, J. W. (1985). Phosphoenolpyruvate: carbohydrate
phosphotransferase system of bacteria. Microbiol. Rev., 49,232-69.
Postma, P. W. & Roseman, S. (1976). The bacterial phosphoenolpyruvate: sugar
phosphotransferase system. Biochim. biophys. Acta, 457,213-57.
Rebers, P. A. & Heidelberger, M. (1959). The specific polysaccharide of type VI
Pneumococcus. I. Preparation, properties and reactions. J. Am. Chem. Soc., 81,
Rebers, P. A. & Heidelberger, M. (1961). The specific polysaccharide of type VI
Pneumococcus, II. The repeating unit. J. Am. Chem. Soc., 83,3056-9.
Roberts, W. K., Buchanan, J. G. & Baddiley, J. (1963). The specific substance from
Pneumococcus type 34(41). The structure of a phosphorus-free repeating unit.
Biochem. J., 88, 1-7.
Roseman, S. (1969). The transport of carbohydrates by a bacterial phosphotransferase
system. J. Gen. Physiol., 54, 138s-180s.
Sable, H. Z. (1950). Phosphorylation of ribose and adenosine in yeast extracts. Proc.
Soc. exptl. Bioi. Med., 75,215-19.
Sable, H. Z. (1952). Pentose metabolism in extracts of yeast and mammalian tissues.
Biochim. biophys. Acta, 8,687-97.
Saier, M. H. Jr (1977). Bacterial phosphoenolpyruvate:sugar phosphotransferase
systems: structural, functional, and evolutionary interrelationships. Bacteriol. Rev.,
Saito, N. & Sugiyama, S. (1966). D-Ribose formation by Pseudomonas reptilivora.
Agric. BioI. Chem., 30,841-6.
Sano, K., Yokozeki, K. & Mitsugi, K. (1977a). Enzymatic production of 5-amino-4imidazole-carboxamide from 5-amino-4-imidazole-carboxamide-riboside by Bacillus
thiaminolyticus. Optimal conditions for the enzyme formation and the enzymatic
reaction. Agric. Bioi. Chern., 41,2331-4.
Sano, K., Yokozeki, K. & Mitsugi, K. (1977b). Screening of micro-organisms

Microbial Production of D-Ribose


producing 5-amino-4-imidazole-carboxamide and o-ribose from 5-amino-4imidazole-carboxamide-riboside. Agric. Bioi. Chem., 41,2463-4.
Sasajima, K. (1976). o-Ribose. In Microbial Production of Nucleic Acid-Related
Substances, ed. K. Ogata, S. Kinoshita, T. Tsunoda & K. Aida. Kodansha Ltd,
Tokyo and John Wiley & Sons, New York, London, Sydney, Toronto, pp. 199-204.
Sasajima, K. & Kumada, T. (1979). Deficiency of o-glucose transport in transketolase
mutant of Bacillus subtilis. Inst. Ferment. Res. Commun., 9, 17-26.
Sasajima, K. & Kumada, T. (1981a). Change in the regulation of enzyme synthesis
under catabolite repression in Bacillus subtilis pleiotropic mutant lacking transketolase. Agric. Bioi. Chem., 45,2005-12.
Sasajima, K. & Kumada, T. (1981b). Cell surface change of Bacillus subtilis
pleiotropic mutant lacking transketolase. Inst. Ferment. Res. Commun., 10,3-9.
Sasajima, K. & Kumada, T. (1983a). Cell surface change of Bacillus subtilis pleiotropic
mutant lacking transketolase: properties of various revertant strains. Inst. Ferment.
Res. Commun., 11,3-9.
Sasajima, K. & Kumada, T. (1983b). Deficiency of flagellation in a Bacillus subtilis
pleiotropic mutant lacking transketolase. Agric. Bioi. Chem., 47, 1375-6.
Sasajima, K. & Yoneda, M. (1971). Carbohydrate metabolism mutants of a Bacillus
species. Part II. o-Ribose accumulation by pentose phosphate pathway mutants.
Agric. Bioi. Chem., 35, 509-17.
Sasajima, K. & Yoneda, M. (1974a). Simple procedures for the preparation of o-ribose
5-phosphate ketol-isomerase, o-ribulose 5-phosphate 3-epimerase and 0sedoheptulose 7-phosphate: o-glyceraldehyde 3-phosphate glycolaldehydetransferase. Agric. Bioi. Chem., 38, 1297-1303.
Sasajima, K. & Yoneda, M. (1974b). o-Sedoheptulose 7-phosphate: o-glyceraldehyde
3-phosphate glycolaldehydetransferase and o-ribulose 5-phosphate 3-epimerase
mutants of a Bacillus species. Agric. Bioi. Chem., 38, 1305-10.
Sasajima, K. & Yoneda, M. (1984). Production of pentoses by micro-organisms.
Biotechnol. Genet. Engineer. Rev., 2, 175-213.
Sasajima, K., Nogami, I. & Yoneda, M. (1970). Carbohydrate metabolism mutants of a
Bacillus species. Part I. Isolation of mutants and their inosine formation. Agric.
Bioi. Chem., 34,381-9.
Sasajima, K., Fukuhara, T., Matsukura, A., Nakanishi, I. & Yoneda, M. (1972).
o-Ribose production by pentose phosphate pathway mutants of Bacillus. Paper
presented at the Fourth International Fermentation Symposium, Kyoto, 19-25
Sasajima, K., Kumada, T. & Yokota, A. (1977). A pleiotropy in carbohydrate
metabolism of Bacillus subtilis mutant lacking transketolase. Inst. Ferment. Res.
Commun., 8, 69-77.
Sasajima, K., Yokota, A. & Kumada, T. (1985a). Alteration of the membrane
composition of Bacillus subtilis pleiotropic mutant lacking transketolase. Inst.
Ferment. Res. Commun., 12,5-18.
Sasajima, K., Yokota, A. & Yoneda, M. (1985b). o-Ribose production by bacteria.
Kagaku to Seibutsu, 23,240-8 (in Japanese).
Schlenk, F. & Smith, R. L. (1953). The mechanism of adenine thiomethylriboside
formation. J. Bioi. Chem., 204, 27-34.
Simonart, P. & Godin, P. (1951). Production de pentoses, d'acide 2-ceto gluconique et
d'acide glucuronique par Penicillium brevi-compactum. Bull. Soc. Chim. Belg., 60,
Slechta, L. & Johnson, L. E. (1976). A new metabolite from Streptomyces hygroscopicus. I. Fermentation and isolation. J. Antibiot., XXIX, 685-7.
Smit, J., Kamio, Y. & Nikaido, H. (1975). Outer membranes of Salmonella
typhimurium: chemical analysis and freeze-fracture studies with lipopolysaccharide
mutants. J. Bacteriol., 124, 942-58.


K. Sasajima & M. Yoneda

Smith, D. C. C. (1955). The preparation of o-ribose and 2-deoxy-o-ribose from

glucose. Chem. Ind., 1955,92-3.
Sowden, J. C. (1950). The condensation of nitromethane with o-erythrose and
2,4-benzylidene-o-erythrose. I. Am. Chem. Soc., 72,808-11.
Steiger, M. (1936). Zur Herstellung von o-Ribose. Helv. Chim. Acta, 19, 189-95.
Stroh, H.-H. von, Dargel, D. & Haussler, R. (1964). Vergleichende Untersuchungen
zur Synthese von o-Ribose aus o-Glucose. I. Prakt. Chem., 23,309-17.
Suhadolnik, R. J. (1970). Nucleoside Antibiotics. Wiley-Interscience, New York,
London, Sydney, Toronto.
Suzuki, T., Tanaka, N., Tomita, F., Mizuhara, K. & Kinoshita, S. (1963). Bacterial
accumulation of ribose and ribose 5-phosphate. I. Gen. Appl. Microbiol., 9, 457-8.
Suzuki, U., Shimamura, T. & Matsunaga, S. (1914). Studies on oryzanin VI. I. Tokyo
Chem. Soc., 34, 1123-75 (in Japanese).
Suzuki, U., Odaka, S. & Mori, T. (1924). Uber einen neuen schwefelhaltigen
Bestandteil der Hefe. Biochem. Z., 154,278-89.
Therisod, M., Fischer, J. C. & Estramareix, B. (1981). The origin of the carbon chain
in the thiazole moiety of thiamine in Escherichia coli: incorporation of deuterated
1-deoxy-o-threo-2-pentulose. Biochem. biophys. Res. Commun., 98,374-9.
Vanderheiden, B. S. (1970). Phosphate esters in human erythrocytes. VII. Further
evidence for ribose 1,5-diphosphate as a natural metabolite. Biochim. Biophys.
Acta, 215, 242-8.
Waller, C. W., Eryth, P. W., Hutchings, B. L. & Williams, J. H. (1953). Achromycin.
The structure of the antibiotic puromycin I. I. Am. Chem. Soc., 75,2025.
Warburg, O. & Christian, W. (1936). Verbrennung von Robison-Ester durch
Triphospho-Pyridin-Nucleotid. Biochem. Z., m, 44O-l.
Warburg, O. & Christian, W. (1937). Abbau von Robisonester durch TriphosphoPyridine-Nucleotid. Biochem. Z., 292,287-95.
Warburg, 0., Christian, W. & Griese, A. (1935). Wasserstoffiibertragendes CoFerment, seine Zusammensetzung und Wirkungsweise. Biochem. Z., 282, 157-205.
Ward, J. B. (1981). Teichoic and teichuronic acids: biosynthesis, assembly, and
location. Microbiol. Rev., 45, 211-43.
Warren, S. C. (1968). Sporulation in Bacillus subtilis. Biochemical changes. Biochem.
I., 109, 811-18.
Watson, J. D. & Crick, F. H. C. (1953). Molecular structure of nucleic acids. A
structure for deoxyribose nucleic acid. Nature, Lond., 171, 737-8.
Weimberg, R. (1961). Pentose oxidation by Pseudomonas fragi. I. Bioi. Chem., 236,
Wessely, F. & Wang, S. (1938). Uber ein Vorkommen von Adonit. Monatsh. Chem.,
72, 168.
Williams, J. C., Verani, R., Alcala, H., Butler, I. J. & Rosenberg, H. S. (1986).
Glutamyl ribose-5-phosphate storage disease: nephrotic syndrome and cerebral
atrophy. Pediatr. Pathol., 5, 277-94.
Willis, R. C. & Furlong, C. E. (1974). Purification and properties of a ribose binding
protein from Escherichia coli. I. Bioi. Chem., 249,6926-9.
Windholz, M., Budavari, S., Blumetti, R. F. & Otterbein, E. S. (1983). The Merck
Index. Merck & Co., Rahway, p. 1185.
Woisetschlager, M. & Hogenauer, G. (1986). Cloning and characterization of the gene
encoding 3-deoxy-o-manno-octulosonate 8-phosphate synthetase from Escherichia
coli. I. Bacteriol., 168,437-9.
Wood, T. (1985). The Pentose Phosphate Pathway. Academic Press, Orlando, San
Diego, New York, Austin, London, Montreal, Sydney, Tokyo, Toronto.
Yokota, A. & Sasajima, K. (1981). Derepressed syntheses of sporulation marker
enzymes in a Bacillus species mutant. Agric. Bioi. Chem., 45,2417-23.
Yokota, A. & Sasajima, K. (1983). Enzymatic formation of a new monosaccharide,

Microbial Production of D-Ribose


I-deoxy-o-altro-heptulose phosphate, from oL-acetoin and o-ribose 5-phosphate by

a transketolase mutant of Bacillus purnilus. Agric. Bioi. Chern., 47, 1545-53.
Yokota, A. & Sasajima, K. (1984a). Formation of I-deoxy-o-threo-pentulose and
I-deoxY-L-threo-pentulose by cell-free extracts of micro-organisms. Agric. Bioi.
Chern., 48, 149-58.
Yokota, A. & Sasajima, K. (1984b). Formation of I-deoxy-o-fructose, I-deoxy-osorbose, I-deoxy-o-tagatose and I-deoxy-erythrulose by cell-free extracts of bacteria
and Actinomycetes. Agric. BioI. Chern., 48, 1643-5.
Yokota, A. & Sasajima, K. (1985). Formation of 2,3-dideoxy-hex-4-ulosonic acid by
carboligase activity of 2-ketoglutarate dehydrogenase in micro-organisms. [nst.
Ferrnent. Res. Cornrnun., 12, 19-33.
Yokota, A. & Sasajima, K. (1986). Formation of I-deoxy-ketoses by pyruvate
dehydrogenase and acetoin dehydrogenase. Agric. BioI. Chern., 50,2517-24.
Yokota, A., Sasajima, K. & Horii, S. (1978). Accumulation of a new monosaccharide,
I-deoxy-o-altro-heptulose (l-deoxy-sedoheptulose) by transketolase mutants of
Bacillus purnilus. Agric. Bioi. Chern., 42,2245-52.
Yokota, A., Sasajima, K. & Yoneda, M. (1979). Reactivation of inactivated o-glucose
dehydrogenase of a Bacillus species by pyridine and adenine nucleotides. Agric.
Bioi. Chern., 43, 271-8.
Zimmer, H.-G., Ibel, H., Suchner, U. & Schad, H. (1984). Ribose intervention in the
cardiac pentose phosphate pathway is not species-specific. Science, 223, 712-4.
Zollner, N., Reiter, S., Gross, M., Pongratz, D., Reimers, C. D. Gerbitz, K., Paetzke,
I., Deufel, T. & Hiibner, G. (1986). Myoadenylate deaminase deficiency: successful
symptomatic therapy by high dose oral administration of ribose. Klin. Wochenschr.,
64, 1281-90.

Chapter 12


Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan

Pantothenic acid (R- or D-( + )-N-(2,4-dihydroxy-3,3-dimethyl-1-oxobutyl)-~
alanine; for chemical formula, see Table 1) was first isolated in the 1930s from
liver and found to be an essential growth factor for yeasts (Williams and
co-workers, 1943). During this same period, it was also identified independently with the chick anti-dermatitis factor, the filtrate factor, the chick
anti-pellagra factor and an essential growth factor for lactic acid bacteria. Later
these activities were shown to be identical with pantothenic acid. Since the
factor could be obtained from a variety of plants and animal tissues, it was
designated pantothenic acid, meaning 'from everywhere' by Williams. The
compound is also referred to as vitamin B 5 It was independently synthesized
by two groups in 1940.
Pantothenic acid occurs in all types of living organisms in a free form and in
conjugated forms such as coenzyme A, pantetheine (Lactobacillus bulgaricus
factor) and 4'-phosphopantetheine (Acetobacter suboxydans factor) (see Table
1). The coenzyme form of the vitamin is coenzyme A. It was discovered as an
essential cofactor for the acetylation of sulfonamide in the liver and the
acetylation of choline in the brain by Lipmann and co-workers (1953). Since
that time, it has been identified with 'active acetate' and has been found to be
essential for a variety of biochemical transacylation reactions. 4'Phosphopantetheine is also a coenzyme form of pantothenic acid. It functions
as a prosthetic group of the acyl carrier protein of fatty acid synthetase, citrate
cleaving enzyme and enzymes involved in the synthesis of peptide antibiotics.
These early studies have been reviewed by several authors (Williams, 1943;
Lipmann, 1953; Wagner & Folkers, 1964; Vagelos 1973; Vandamme, 1981;
Kleinkauf & von Dohren, 1983).

Pantothenic acid can be obtained as a colorless, viscous oil by drying under
high vacuum in P 2 0 5 . It is an acid and has a marked tendency to absorb water

MW: 47653

MW: 24121

CH 3

4H 16NO HP


MW: 31327



4' -Phosphopantothenic acid (Ba salt)

CH 3 0H

CH 3
4H 16NaN0 5



Sodium o-pantothenate I

CH l


Calcium o-pantothenate

4H17NOs MW: 21923


o-Pantothenic acid

Table 1

Soluble in water; insoluble in ethanol. Unstable to bases. Free acid is

unstable. [a J~ + 90 (c = 3 3). King & Strong (1951); Okada et al.

White, hygroscopic crystals. Decomposed by acids and bases. Solutions

are stable between pH 5 and 7. mp 122-124C; [aJ~ + 271 (c = 2).
For solubility, see calcium pantothenate

White needles. Moderately hygroscopic. Soluble in water, glycerol;

slightly soluble in alcohol, acetone; insoluble in ether, benzene,
chloroform. Decomposed by bases. Solutions are stable between pH 5
and 7. mp 195-196C (dec); [aJ~ + 282 (c = 5).

Unstable, viscous oil. Extremely hydroscopic, easily decomposed by

acids, bases, heat. Soluble in water, ethyl acetate, dioxane, glacial
acetic acid; moderately soluble in ether, amyl alcohol; insoluble in
benzene, chloroform. Solutions are stable between pH 5 and 7.
[aHi' + 375. Stiller et al. (1940).

Pantothenic Acid and its Naturally-Occurring Derivatives


MW: 32238

I .I





MW: 41642

MW: 27837

MW: 55472

I .I



CH 3

MW: 35835



4' -Phosphopantetheine (Sa salt)

CH 30H

CH 3



C 1H 22N 20 4S

CH 3


CH 3 0H




4' -Phosphopantothenoyl-L-cysteine (Sa salt)




Pantothenoyl-L-cysteine (Sa salt)



Soluble in water; slightly soluble in ethanol; insoluble in ether. Unstable

to acids and bases. Easily oxidized in air. [ll']~ + 133 (c = 225);
Oxidized form, [ll']~ + 122. Acetobacter suboxidans factor. Baddiley
& Thain (1953); Moffatt & Khorana (1961); Nagase (1967).

Disulfide form of pantetheine. Glassy, colorless to light yellow

substance. Unstable to acids. [ll']~ + 171 (c = 32). For solubility and
ref., see pantetheine.

Syrup or glass. Soluble in water; slightly soluble in alcohol, insoluble in

ether, benzene, chloroform, ethyl acetate. Unstable to acids and
bases. Easily oxidized in air. [ll']~ + 122 (c = 345). Lactobacillus
bulgaricus factor. Shimizu et al. (1965).

Soluble in water; slightly soluble in alcohol. Unstable to acids and bases.

Easily oxidized in air. [ll']~ 0 (c = 2). Baddiley & Mathias (1954);
Nagase (1967).

Soluble in water, methanol, moderately soluble in ethanol; insoluble in

ether. Unstable to acids and bases. [ll']D - 14 (c = 2). Ohta et al.

I . I


(Li salt)





MW: 68756







Dephospho-coenzyme A

Soluble in water, methanol; insoluble in acetone. Unstable to acids and

bases. For ref., see coenzyme A.


I .I


CH .1


. 2

MW: 76755






MW: 43841



4' -Phosphopantetheine-S-sulfonate (Ca salt)



O-CH 2





Coenzyme A


Soluble in water. [a]~ + 72 (c = 195). Bifidus factor. Yoshioka &

Tamura (1971).

Soluble in water; insoluble in ethanol, ether, acetone. Decomposed to

panthetheine-2' ,4'-cyclic phosphate and 3' ,5'-ADP in 1 N NaOH
(100C, 2 min). Decomposed to pantetheine-4'-phosphate and adenine
in 1 N HCI (1()()OC, 5 min). Easily oxidized in air. Moffatt & Khorana
(1961); Shimizu (1970); Shimizu et al. (1979a).








MW: 45754

CH 3




OA-6129A (Na salt)

MW: 79168





CH 20 H


4' -O-(!3-glucopyranosyl)-D-pantothenic acid

= 10). Growth factor of Leuconostoc.

Unstable to acids and bases. [c:r]ii + 11.60 (c = 10). Antibiotic produced

by Streptomyces sp. OA-6129. Yoshioka et al. (1983).

Soluble in water. [c:r]ii - 18 20 (c

Amachi et at. (1971).

Table l-contd.

Vitamin B s , Coenzyme A and Related Compounds


4 '-~tetheine
: - -pantothenic aci.d---\
:-pantoic acid ~
: '









C~HZ~ --~-~~--f- i- S-al.anine-J,..cysteamineJ,


c;> H

Ho-~ ~~~



Fig. 1. The structure of coenzyme A.

from the air. It breaks down into fJ-alanine and pantoic acid by alkaline
hydrolysis. The latter forms a lactone, i.e. D-( - )-pantoyl lactone, readily in
acid solution or upon heating. Acid hydrolysis of pantothenic acid gives
fJ-alanine and pantoyl lactone. Pantothenic acid is soluble in water, ethyl
acetate, dioxane, acetic acid, ether and amylalcohol, but is insoluble in
benzene and chloroform.
The structure of pantothenic acid contains a single asymmetric center, so
that it is optically active; only the natural D-( + )-isomer has vitamin activity.
The absolute configuration of the vitamin has been defined as R (Hill & Chan,
1970). The conformation of the vitamin has also been reported (Fritz & LOwe,
The calcium salt of pantothenic acid can be obtained as needle crystals from
methanol. It is moderately hygroscopic and is rather more stable to heat, air
and light than the free acid. It is soluble in water and glycerol and slightly
soluble in alcohol and acetone. Reviews by Baddiley (1955) and Wagner &
Folkers (1964) summarize early studies on the chemistry of pantothenic acid.
Table 1 lists several naturally-occurring derivatives of pantothenic acid. They
can be grouped into three types based on their chemical structures, i.e. simple
pantothenate derivatives, pantetheine derivatives in which cysteamine (or its
analogs) attaches by an amide linkage and coenzyme A derivatives in which
the pantetheine is adenosylylated (see Fig. 1).
Pantothenyl alcohol, an alcohol analog of pantothenic acid, is also a
pharmaceutically important unnatural derivative. Unnatural analogs of pantothenic acid and coenzyme A were reviewed by Shimizu (1970).

Although pantothenic acid cannot be synthesized by animals, micro-organisms
and plants are generally able to produce it from the precursors pantoic acid

S. Shimizu & H. Yamada

pyruvic acid


~ CH"'r-COOH

a-aoetolactic acid


tlUD-pantothenic acid



- 1





Fig. 2. The pathway for the biosynthesis of pantothenic acid.

and p-alanine, through catalysis of the enzyme pantothenate synthetase (EC Animals, plants and micro-organisms can convert pantothenic acid to
4' -phosphopantetheine and coenzyme A, the metabolically active forms of the
vitamin. The pathway for the biosynthesis of pantothenic acid and coenzyme A
has been elucidated by several workers since the early 1950s and has been
reviewed (Brown & Reynolds, 1963; Plaut et ai., 1974; Abiko, 1975; Brown &
Williamson, 1982). The pathway leading to the vitamin and the coenzymes
from common precursors can be summarized as shown in Figs 2 and 3.

3.1 /I-Alanine
Three routes to p-alanine have been reported. Several micro-organisms have
been reported to form p-alanine by a--decarboxylation of L-aspartic acid (Fig.
2, reaction 1). Confirmatory evidence for this conversion was provided by
Williamson & Brown (1979), who purified (to apparent homogeneity) from
extracts of Escherichia coli an enzyme that catalyzes the a--decarboxylation of
L-aspartic acid to yield p-alanine and CO2 They also reported that the enzyme
is missing in a mutant of E. coli that requires either p-alanine or pantothenate
as a nutritional factor, but is present in the wild-type strain and in a revertant
strain of the mutant. It has also been suggested that p-alanine is produced by
decarboxylation of N-carbamoyl-p-alanine formed from uracil on the basis of
the observation that mutants of Salmonella typhimurium lacking the ability to
degrade uracil require N-carbamoyl-p-alanine, p-alanine or pantothenate as a
nutritional factor (Fig. 2, reactions 2, 3 and 4) (West et al., 1985). p-Alanine
may also be produced by transamination of malonylsemialdehyde produced
from propionic acid (Fig. 2, reaction 5 or 6), because enzyme activity
catalyzing this conversion was detected in several micro-organisms. However,
there have been no further studies concerning this reaction.

Vitamin B 5 , Coenzyme A and Related Compounds



Pantoic Acid

The route to pantoic acid from pyruvate as shown in Fig. 2 has been
elucidated mainly in E. coli and Neurospora crassa. Two enzymes catalyzing
the conversion of pyruvate to a-ketoisovalerate (Fig. 2, reactions 7 and 8) in
this route are shared by the route for the biosynthesis of the branched chain
amino acids. In E. coli, two enzyme activities have been detected for the
conversion of a-ketoisovalerate to ketopantoic acid (Fig. 2, reaction 9); one is
dependent on tetrahydrofolate and the other is not. The physiological
significance of the tetrahydrofolate-independent activity seemed to be questionable because of its high Km values for formaldehyde (10 mM) and
a-ketoisovalerate (100 mM). The fact that a mutant mlssmg the
tetrahydrofolate-dependent activity requires pantothenate for growth, whereas
the tetrahydrofolate-independent activity is found in the same amounts in the
same mutant also give concrete evidence supporting the theory that the
tetrahydrofolate-dependent enzyme is the one responsible for the ketopantoate
needed for the biosynthesis of pantothenate. The tetrahydrofolate-dependent
enzyme, i.e. ketopantoate hydroxymethyltransferase (EC, has been
purified and characterized in some detail. The observation that pantoate,
pantothenate and coenzyme A are all allosteric inhibitors of this enzyme also
supports this conclusion (Powers & Snell, 1976; Teller et al., 1976).
The reduction of ketopantoic acid to D-pantoic acid (Fig. 2, reaction 10) is
catalyzed by an NADPH-dependent enzyme, ketopanoic acid reductase (EC The enzyme activity has been detected in Saccharomyces cerevisiae
and E. coli. The same reduction is also catalyzed by a-acetohydroxyacid
isomeroreductase (EC which is the responsible enzyme for the
conversion of a-acetolactate to a-ketoisovalerate (Fig. 2, reaction 8) (Primerano & Burns, 1983). Recently, Shimizu et al. (1988a) isolated ketopantoic acid
reductase in a crystalline form from Pseudomonas maltophilia and characterized it in some detail. They also demonstrated that this enzyme is due the
enzyme for D-pantoic acid formation, necessary for the biosynthesis of
pantothenic acid, due to the observation that mutants lacking this enzyme
require either D-pantoic acid or pantothenate for growth and the revertants
regain this activity.

Coenzyme A

The pathway for the biosynthesis of coenzyme A from pantothenic acid,

L-cysteine and A TP as shown in Fig. 3 was first demonstrated by Brown
(1959a,b) based on his studies with Proteus morganii and the early observations in the 1950s with pig liver and other organisms. Later, the validity of this
pathway was confirmed by Abiko and co-workers who separated and characterized the enzymes involved in this pathway in rat liver (Abiko, 1975).
The first step in this pathway is the phosphorylation of pantothenic acid (Fig.
3, reaction 1) by pantothenate kinase (EC The enzyme has been
purified and characterized from rat liver (Abiko et al., 1972) and

S. Shimizu & H. Yamada


r---" X
AlP ~







Pantothenic acid

P-Pontothenic acid







AlP "




!... __________________ CoA _________ 1I

Fig. 3. The pathway for the biosynthesis of coenzyme A from pantothenic acid,
L-cysteine and ATP. For chemical structures of each compound, see Table l.
Abbreviation used: CoA, coenzyme A.

Brevibacterium ammoniagenes (Shimizu et al., 1973). Both enzymes receive

allosteric inhibition by coenzyme A, the end product of the pathway. Since
such feedback inhibition by coenzyme A is observed only at this step and no
other regulation mechanism has been known, this inhibition seems to be the
most important mechanism for controlling the cellular level of coenzyme A.
Pantetheine is also phosphorylated by the same enzyme to yield 4'phosphopantetheine (Fig. 3, reaction 7), which can be converted to coenzyme
The condensation of 4' -phosphopantothenic acid with L-cysteine to yield
4' -phosphopantothenoyl-L-cysteine (Fig. 3, reaction 2) is catalyzed by 4'phosphopantothenoyl-L-cysteine synthetase (EC The mammalian
enzyme requires A TP as energy for the condensation whereas the bacterial
enzyme preferably utilizes CTP (Brown, 1959a). 4'-Phosphopantothenoyl-Lcysteine is then decarboxylated to yield 4' -phosphopantetheine (Fig. 3,
reaction 3) by 4' -phosphopantothenoyl-L-cysteine decarboxylase (EC
The enzyme has been shown to be independent of pyridoxal 5'-phosphate.
This step is the only one which does not require A TP among the five steps in
the biosynthesis of this coenzyme.
The pyrophosphate linkage formation between 4' -phosphopantetheine and
ATP to yield 3' -dephospho-coenzyme A (Fig. 3, reaction 4) and the
phosphorylation of it (Fig. 3, reaction 5) are respectively catalyzed by
dephospho-coenzyme A pyrophosphorylase (EC and dephosphocoenzyme A kinase (EC In rat liver, both enzymes are present as a
complex or a bifunctional enzyme (Suzuki et al., 1967). The reversibility of the
former reaction may be important for controlling cellular levels of coenzyme A
and 4' -phosphopantetheine.
The presence of an alternate route to yield 4' -phosphopantetheine via
pantetheine (Fig. 3, reactions 6 and 7) has been suggested in Acetobacter

Vitamin B 5 Coenzyme A and Related Compounds


suboxydans, Lactobacillus helveticus and others (Brown, 1959b), but confirmatory evidence for this is not available.

3.4 Control Mechanisms for the Biosynthesis

The allosteric inhibition of ketopantoic acid hydroxymethyltransferase of E.
coli by o-pantoic acid, pantothenic acid or coenzyme A may be involved as a
control mechanism in pantothenate biosynthesis (Powers & Snell, 1976). On
the other hand, such inhibition was not observed in the case of ketopantoic
acid reductase of P. maltophilia (Shimizu et al., 1988a).
In the pathway to coenzyme A from pantothenic acid, the feedback
inhibition of pantothenate kinase by coenzyme A and 4'-phosphopantetheine
has been demonstrated to be involved as a control mechanism in the
biosynthesis (Abiko et al., 1972; Shimizu et al., 1973; Vallari et at., 1987).
Since this inhibition was generally observed regardless of species and the other
four steps following this reaction are not inhibited by coenzyme A or
4' -phosphopantetheine significantly, this may be one of the most important
mechanisms to control cellular levels of coenzyme A. No other mechanism
such as repression has been observed in either pantothenate or coenzyme A
Pantetheinase, which specifically degrades pantetheine to pantothenic acid
and cysteamine, may also be an important enzyme, because coenzyme A can
be degraded to pantetheine enzymatically and pantetheine can be reused as a
precursor of coenzyme A after phosphorylation by pantothenate kinase.
Cellular coenzyme A levels may be influenced by competition between
pantetheinase and pantothenate kinase towards their substrate, pantetheine
(Wittwer et al., 1983).


4.1 Pantothenic Add

At present, commercial production of pantothenate depends exclusively on
chemical synthesis. As outlined in Fig. 4, the conventional chemical process
involves reactions yielding racemic pantoyl lactone from isobutyraldehyde,
formaldehyde and cyanide, optical resolution of the racemic pantoyl lactone to
0-( - )-pantoyl lactone with quinine, quinidine, cinchonidine, brucine and so
on and condensation of 0-( - )-pantoyllactone with p-alanine. This is followed
by isolation as the calcium salt and drying to obtain the final product. A
problem of this chemical process apart from the use of poisonous cyanide is the
troublesome resolution of the racemic pantoyl lactone and the reracemization
of the remaining L-( + )-isomer. Therefore, most of the recent studies in this
area have been concentrated on development of an efficient method to obtain
0-( - )-pantoyllactone.


S. Shimizu & H. Yamada


Fig. 4. Outline of the chemical synthesis of D-pantothenic acid.

To skip this resolution-reracemization step, several microbial or enzymatic

methods have been proposed. They roughly fall into two types based on the
starting substrate used (Yamada & Shimizu, 1988).
Recently, an efficient combined chemi-enzymatic method, which involves an
efficient one-pot synthesis of ketopantoyl lactone as a starting substrate,
followed by stereospecific reduction of it to 0-( - )-pantoyl lactone using
microbial cells as a catalyst was reported (Hata et al., 1987; Shimizu et al.,
1984a, 1987a). As shown in Fig. 5, ketopantoyl lactone is synthesized from
isobutyraldehyde, sodium methoxide, diethyl oxalate and formalin. The
reaction is performed in one step at room temperature with a yield of 810%.
Stereospecific reduction of ketopantoyl lactone to 0-( - )-pantoyl lactone is
carried out with washed cells of Rhodotorula minuta or Candia parapsilosis as
a catalyst and glucose as energy for the reduction. About 50 or 90 g/liter of
optically pure 0-( - )-pantoyl lactone can be produced with a molar yield of
nearly 100% by R. minuta or C. parapsilosis, respectively. The enzyme
catalyzing this conversion has been isolated as a crystalline form from C.
parapsilosis cells and characterized. It is a novel carbonyl reductase which
specifically catalyzes the reduction of conjugated polyketone compounds (Hata
et al.,1989a,b; Shimizu et al., 1988b).
Racemic pantoyllactone can also be used as a starting substrate. Shimizu et
ale (1987b) reported that Nocardia asteroides cells specifically oxidize the
L-( + )-isomer in a racemic mixture of pantoyllactone to ketopantoyllactone,
which is then converted to 0-( + )-pantoyl lactone by the reduction with C.
parapsilosis cells as described above. In these two enzymatic steps, the
coexisting 0-( - )-isomer remains without any modification (Fig. 6, reactions 1
and 2). Under suitable conditions, 72 g/liter of 0-( - )-pantoyl lactone was
obtained from 80 g/liter of oL-pantoyl lactone. Similar specific oxidation and
reduction reactions can also be carried out with a single micro-organism as
catalyst. On incubation with washed cells of Rhodococcus erythropolis,

Vitamin B s , Coenzyme A and Related Compounds




~ ~CHg~-~~CHO ~TOOMe

2 x



















~~/ ___________

~ -alanine

(R)-(+)-pantothenic acid

Fig. S. The reaction pathway for the chemicoenzymatic synthesis of 0-( - )-pantoyl

D-( - )-pantoyl lactone in the reaction mixture reached 182 g/liter with a
molar yield of 905% (optical purity, 944% e.e.). This unique conversion
proceeds through the successive reactions as follows: (1) the enzymatic
oxidation of L-( + )-pantoyl lactone to ketopantoyl lactone (the same enzyme
as that in N. asteroides has been suggested to be the responsible enzyme for
this oxidation); (2) the rapid and spontaneous hydrolysis of ketopantoyl
lactone to ketopantoic acid, and (3) the enzymatic reduction of the ketopantoic
acid to D-pantoic acid. The enzyme catalyzing this reduction seemed to be
ketopantoic acid reductase, because R. erythropolis cells could not utilize



(L-PL forming}


(D-PL forming)

(HC1)1 (5)


(3)t<PH 7)





Fig. 6. Reactions involved in the enzymatic transformation to 0-( - )-pantoyl lactone.
Abbreviations used: L-PL, L-( + )-pantoyllactone; D-PL, 0-( - )-pantoyllactone; KPL,
ketopantoyllactone; KPA, ketopantoic acid; D-PA, o-pantoic acid.


S. Shimizu & H. Yamada

ketopantoyl lactone as substrate, different from C. parapsilosis (see Fig. 6,

reactions 1, 3 and 4) (Shimizu et at., 1987c).
These three enzymatic methods are simple and require no reracemization
step, which is necessary for the conventional chemical resolution. A chemical
reduction of ketopantoyl lactone which uses rhodium complex as a catalyst has
also been reported to yield 0-( - )-pantoyllactone stereospecifically (Ojima et
at., 1978).

4.2 Coenzyme A
The production methods for coenzyme A roughly fall into chemical and
microbial ones. The chemical methods have been reviewed by Shimizu (1970)
and Mautner (1970). They are not practical due to their complexity.
Therefore, commercial production is carried out by microbiological methods.
Extraction of coenzyme A from yeast cells has been performed since the early
1950s. Usually, cells of baker's or brewer's yeasts, which are relatively rich in
coenzyme A, are used as the source. Later, an efficient enzymatic method
using Brevibacterium ammoniagenes cells as the catalyst was developed.
These microbial methods were reviewed by Shimizu & Yamada (1986).
A successful enzymatic method using the biosynthetic route of coenzyme A
from pantothenic acid, L-cysteine and ATP was first reported by Ogata et al.
(1970), who found that B. ammoniagenes has all five enzymes necessary for the
biosynthesis of coenzyme A in high activities. The above three substrates,
when added to a reaction mixture containing the bacterial cells, were
converted to coenzyme A with a satisfactory yield (2-3 g/liter). They also
found that the same organism can accumulate coenzyme A directly in the
culture medium on addition of pantothenic acid, L-cysteine and AMP,
adenosine or adenine in the presence of a surfactant, cetylpyridinium chloride,
and high levels of glucose (usually 10%), K2 HP0 4 and MgS0 4 7H2 0. Under
optimal conditions, the amount obtained was 55 g/liter. Most coenzyme A in
the medium was present in the disulfide form due to the vigorous shaking
during the reaction. After treatment of the culture filtrate with Duolite S-30,
charcoal and Dowex 1 (Cl-), and the reduction of the disulfide, the very pure
thiol form was obtained in a high yield. The mechanism of this coenzyme A
production has been suggested to be that shown in Fig. 7 (for details, see
Shimizu et at., 1979a,b).
In order to improve the product yield further, the regulation mechanism of
the biosynthesis was investigated. As described in the previous section, the
biosynthesis of coenzyme A in B. ammoniagenes is mainly controlled by the
feedback inhibition of pantothenate kinase by coenzyme A. This was the main
problem in the practical production, because the over-produced coenzyme A
itself stopped the biosynthesis. Two methods to abolish this feedback
inhibition have been reported.
A synthetic scheme in which the reaction is initiated by the condensation of
4' -phosphopantothenic acid and L-cysteine or the transadenosylylation of

Villlmin B s , Coenzyme A and Related Compounds



~ Cysteine







Fig. 7. Reaction sequences of coenzyme A production with Brevibacterium ammoniagenes under A TP-generating conditions. Abbreviations used: PRPP, 5phosphoribosyl-l-pyrophosphate; CoA, coenzyme A.

4' -phosphopantetheine was investigated, because these routes do not involve

phosphorylation of pantothenic acid or pantetheine by pantothenate kinase.
Replacement of the enzymatic phosphorylation of pantothenate or pantetheine
with chemical phosphorylation followed by the enzymatic reaction increased
the yield of coenzyme A 10-20 times. Yields from 4' -phosphopantothenic acid
and 4'-phosphopantetheine were 33 g/liter (molar yield based on ATP, 645%)
and 115 g/liter (100%), respectively (Shimizu et a/., 1983). This method is
applicable to coenzyme A production under the A TP-generating conditions
(see Fig. 7). 4'-Phosphopantothenic acid (25 g/liter), L-cysteine (15 g/liter),
and AMP (33 g/liter) added to the culture broth of B. ammoniagenes were
converted to coenzyme A with a yield of 23 g/liter (Shimizu & Yamada, 1984).
Another way to improve the yield is to use mutants derepressed for the
feedback inhibition or those showing elevated pantothenate kinase activity. A
mutant of B. ammoniagenes that is resistant to oxypantetheine (the cor-

Fig. 8. Time course of coenzyme A production by an oxypantetheine-resistant mutant

of Brevibacterium ammoniagenes under ATP-generating conditions. (A) Production
from pantothenic acid, L-cysteine and AMP. (B) Production from pantetheine and
AMP. For details, see Shimizu et al. (1984b). 0, Coenzyme A (CoA); . , pantothenic
acid (PaA); D, ATP; .; ADP; D., AMP; .6, pH; x, cell growth.


S. Shimizu & H. Yamada

responding oxygen analog of pantetheine) was found to have an elevated

activity of pantothenate kinase. Under the ATP-generating conditions, the
yields of coenzyme A from pantothenic acid (3-6 g/liter), L-cysteine
(1-8 g/liter) and AMP (6 g/liter) or from pantetheine (5 g/liter) and AMP
(6g/liter) were 93 or U5g/liter, respectively. These values were about
three-fold higher than those obtained with the parent strain, and 70-100% of
the added AMP was converted to coenzyme A (Shimizu etal., 1984b) (Fig. 8).
Continuous production of coenzyme A through five enzymatic steps using
gel-entrapped cells of B. ammoniagenes has also been reported (Shimizu et al.,
1979b; Yamada et al., 1980).
Continuous production of coenzyme A through five enzymatic steps using
gel-entrapped cells of B. ammoniagenes has also been reported (Shimizu et al.,
1979b; Yamada et al., 1980).


4'-Phosphopantetheine and Other Intermediates in Coenzyme A


4' -Phosphopantetheine together with other intermediates in coenzyme A

biosynthesis can be effectively synthesized by using B. ammoniagenes cells as
the catalyst and by modifying the reaction conditions (Shimizu et al., 1979a):
4' -phosphopantothenic acid on omission of L-cysteine from the reaction
mixture, 4'-phosphopantetheine on addition of CTP or CTP and GTP in place
of A TP, because pantothenic acid is phosphorylated in the presence of CTP or
GTP as well as A TP and the condensation of 4' -phosphopantetheine with
nucleoside triphosphate is specific for A TP, and 3' -dephospho-coenzyme A on
treatment of the reaction mixture containing coenzyme A with 3'-nucleotidase.
The amounts of these intermediates obtained by this method are summarized
in Table 2.

Test micro-organisms normally used for the microbiological assay of pantothenic acid are Lactobacillus plantarum ATCC 8014 (L. arabinosus 17-5), L.
casei ATCC 7469 and Saccharomyces uvarum ATCC 9080 (S. carlsbergensis).
Lactobacillus plantarum is suitable for determining unconjugated pantothenate
in samples. It should be noted that pantetheine, when simultaneously present
in a molar ratio to pantothenate of more than O 5, gives positive errors in the
determination. Saccharomyces uvarum also shows almost specific growth
response to free pantothenate, but fJ-alanine stimulates its growth. Hence, an
assay procedure employing this organism is also the one chosen for determining the pantothenic acid that occurs in natural products together with other
pantothenate forms. Lactobacillus casei responds not only to pantothenate but
also several conjugated forms of pantothenate. Lactobacillus helveticus ATCC


Pantothenic acid
Pantothenic acid
4' -Phosphopantothenic acid and
Pantothenic acid
and L-cysteine
Pantothenic acid
and L-cysteine
Pantothenic acid
and L-cysteine


4' -Phosphopantothenic acid

4' -Phosphopantothenic acid
4' -Phosphopantetheine



ITP and CfP
























Productivity (mg/ml)
enzyme source

For details, see Shimizu et al., 1979a,b.

Numbers correspond to those given in Fig. 3.
C In this case, the nucleoside monophosphates added are presumed to be converted to the corresponding nucleoside triphosphates and used for
the reactions in a similar manner as shown in Fig. 7.

4' -Phosphopantetheine
4' -Phosphopantetheine
3' -Dephosphocoenzyme A

4' -Phosphopantetheine

4' -Phosphopantetheine

Table 2

Production of the Intermediates in Coenzyme A Biosynthesis by Brevibacterium ammonillgenes

S. Shimizu & H. Yamada


12046 and L. bulgaricus Bl are recommended for determination of pantetheine (or pantethine), because both these organisms require more than 100
times as much pantothenic acid as pantethine to give the same response.
Treatment of samples and details of assay procedures have been described
(Bird & Thompson, 1967).
A sensitive enzymatic assay method using pantothenase has been reported
(Airas, 1986), but the enzyme is not commercially available. Chemical and
physical methods have also been reported. They are often used in determining
pantothenic acid in pharmaceutical products. They are not suitable for
determination of natural samples because of their low sensitivity.


The current world capacity of calcium pantothenate production and its demand
are presumed to be about 4000 and 3600-4000 tons/year, respectively. It is
mainly used as an additive of animal feed (about 3000 tons/year) and as a
pharmaceutical product (about 600 tons/year). Pantothenyl alcohol is used as a
source of panthothenate activity for pharmaceutical vitamin products. Pantothenyl alcohol itself has no pantothenate activity; in fact, it is a competitive
growth inhibitor of several pantothenate-requiring lactic acid bacteria. However, it has been demonstrated to be quantitatively converted to pantothenic
acid in the animal body, and to be equivalent to pantothenic acid in man.
Pantethine, the disulfide of pantetheine, and coenzyme A are also used as
pharmaceutical products in several countries. They have been suggested to be
effective in reducing cholesterol level, curing fatty liver, and related diseases.
Some sulfonate derivatives of pantetheine or coenzyme A (Bifidus factors),
such as 4'-phosphopantetheine-S-sulfonate, which were originally isolated
from carrot roots have been shown to be growth factors of Bifidobacterium
(Yoshioka & Tamura, 1971). Addition of the bifidus factors to dried milk for
infants has been suggested to be useful in improving the quality of the milk. A
carbapenem antibiotic, OA-6129A, (Table 1) produced by Streptomyces sp.
OA-6129, may be an interesting example suggesting a new use of the vitamin
as building block for its synthesis (Yoshioka et al., 1983).

Research from the authors laboratory was supported, in part, by a Grant in aid
of Scietific Research from the Ministry of Education, Science and Culture of

Vitamin B s , Coenzyme A and Related Compounds


Abiko, Y. (1975). Metabolism of coenzyme A. In Metabolic Pathways, 3rd edn, Vol. 7,
Metabolism of Sulfur Compounds, ed. D. M. Greenberg. Academic Press, New
York, pp. 1-25.
Abiko, Y., Ashida, S. & Shimizu, M. (1972). Purification and properties of Dpantothenate kinase from rat liver. Biochim. biophys. Acta, 268,364-72.
Airas, R. K. (1986). Pantothenase-based assay of pantothenic acid. Meth. Enzymol.,
Amachi, T., Iwamoto, S. & Yoshizumi, H. (1971). A growth factor for malo-lactic
fermentation bacteria. Part II. Structure and synthesis of a novel pantothenic acid
derivative isolated from tomato juice. Agric. Bioi. Chem., 35, 1222-30.
Baddiley, J. (1955). The structure of coenzyme A. Adv. Enzymol. 16, 1-22.
Baddiley, J. & Mathias, A. P. (1954). Coenzyme A. Part IX. The synthesis of
pantothenoylcysteine, its 4'-phosphate, and related compounds as possible precursors of the coenzyme. J. Chem. Soc., 2803-12.
Baddiley, J. & Thain, E. M. (1953). Coenzyme A. Part VIII. The synthesis of
pantetheine 4'-phosphate (Acetobactor stimulatory factor), a degradation product of
the coenzyme. J. Chem. Soc., 1610-15.
Bird, O. D. & Thompson, R. Q. (1967). Pantothenic acid. In The Vitamins, Chemistry,
Physiology, Pathology, Methods, 2nd edn, Vol. VII, ed. P. Gyorgy & W. N.
Pearson. Academic Press, New York, pp. 209-41.
Brown, G. M. (1959a). The metabolism of pantothenic acid. J. Bioi. Chem., 234,
Brown, G. M. (1959b). Assay and distribution of bound form of pantothenic acid. J.
Bioi. Chem., 234, 379-82.
Brown, G. M. & Reynolds, J. J. (1963). Biogenesis of the water-soluble vitamins. Ann.
Rev. Biochem., 32,419-62.
Brown, G. M. & Williamson, J. M. (1982). Biosynthesis of riboflavin, folic acid,
thiamine, and pantothenic acid. Adv. Enzymol., 53,345-81.
Fritz, H. & LOwe, W. (1962). Konformationsbestimmung der Pantothensaure und des
Pantoylamides durch Protonenresonanz-spektroskopie. Angew. Chem., 74,751-3.
Hata, H., Shimizu, S. & Yamada, H. (1987). Enzymatic production of D-( - )-pantoyl
lactone from ketopantoyl lactone. Agric. Bioi. Chem., 51, 3011-16.
Hata, H., Shimizu, S., Hattori, S. & Yamada, H. (1989a). Ketopantoyl lactone
reductase from Candida parapsilosis, Purification and characterization as a conjugated polyketone reductase. Biochim. Biophys. Acta, 990, 175-81.
Hata, H., Shimizu, S., Hattori, S., & Yamada, H. (1989b). Ketopantoyl lactone
reductase is a conjugated polyketone reductase. FEMS Microbiol. Lett., 58, 87-90.
Hill, R. K. & Chan, T. H. (1970). The absolute configuration of pantothenic acid.
Biochem. Biophys. Res. Commun., 38, 181-3.
King, T. E. & Strong, F. M. (1951). Synthesis and properties of pantothenic acid
monophosphates. J. Bioi. Chem. 191, 515-21.
Kleinkauf, H. & Dohren, H. von (1983). Non-ribosomal peptide formation on
multifunctional proteins. Trends. Biochem. Sci. 8,281-3.
Lipmann, F. (1953). Symposium on chemistry and function of coenzyme A. Fedn.
Proc., n,673-715.
Mautner, H. G. (1970). Synthesis of coenzyme A analogs, Meth. Enzymol., 18A,
Moffatt, J. G. & Khorana, H. G. (1961). Nucleoside polyphosphates. XII. The total
synthesis of coenzyme A. J. Am. Chem. Soc., 83, 663-75.
Nagase, O. (1967). Investigations on pantothenic acid and its related compounds. IV.


S. Shimizu & H. Yamada

Chemical studies. (3). Syntheses of D-pantetheine 4'-phosphate and N-Dpantothenoyl-L-cysteine 4'-phosphate. Chem. Pharm. Bull., 15,648-54.
Ogata, K., Shimizu, S. & Tani, Y. (1970). A new preparation method of coenzyme A
with micro-organisms. Agric. Bioi. Chem., 34, 1757-9.
Ohta, G., Nagase, 0., Hosokawa, Y., Tagawa, H. & Shimizu, M. (1967). Investigations on pantothenic acid and its related compounds. III. Chemical studies. (2).
Synthesis of di-D-pantothenoyl-L-cystine. Chem. Pharm. Bull., 15,644-7.
Ojima, I., Kogure, T., Terasaki, T. & Achiwa, K. (1978). Effective biomimetic route
to D-( + )-pantothenic acid using asymmetric hydrogenation catalyzed by a chiral
rhodium complex in the key step. I. Org. Chem., 43,3444-6.
Okada, S., Nagase, O. & Shimizu, M. (1967). Investigations on pantothenic acid and
its related compounds. VII. Chemical studies. (5). Synthesis of D-pantothenic acid
4'-phosphate. Chem. Pharm. Bull., 15,713-5.
Plaut, G. W., Smith, C. M. & Alworth, W. L. (1974). Biosynthesis of water-soluble
vitamins. Ann. Rev. Biochem., 43, 899-922.
Powers, S. G . & Snell, E. E. (1976). Ketopantoate hydroxymethyltransferase, II.
Physical, catalytic, and regulatory properties. I. Bioi. Chem. 251, 3786-93.
Primerano, D. A. & Burns, R. O. (1983). Role of acetohydroxy acid isomeroreductase
in biosynthesis of pantothenic acid in Salmonella typhimurium. I. Bacteriol., 153,
Shimizu, M. (1970). Pantothenic acid and coenzyme A: Preparation of CoA analogs.
Meth. Enzymol., 18A, 322-8.
Shimizu, S. & Yamada, H. (1984). Enzymatic process for the high yield production of
coenzyme A. In Third European Congress on Biotechnology, Vol. I. Verlag
Chemie, Weinheim, pp. 401-5.
Shimizu, S. & Yamada, H. (1986). Coenzymes. In "Biotechnology", Vol. 4, ed. H.-J.
Rehm & G. Reed. VCH Verlagsgesellschaft, Weinheim, pp. 159-84.
Shimizu, M., Ohta, G., Nagase, 0., Okada, S. & Hosokawa, Y. (1965). Investigations
on pantothenic acid and its related compounds. I. Chemical studies. (1). A novel
synthesis of pantethine. Chem. Pharm. Bull., 13, 180-8.
Shimizu, S., Kubo, K., Tani, Y. & Ogata, K. (1973). Purification and properties of
pantothenate kinase from Brevibacterium ammoniagenes IFO 12071. Agric. Bioi.
Chem., 37,2863-70.
Shimizu, S., Tani, Y. & Ogata, K. (1979a). Synthesis of coenzyme A and its
biosynthetic intermediates by microbial processes. Meth. Enzymol., 62, 236-45.
Shimizu, S., Tani, Y. & Yamada, H. (1979b). Synthesis of coenzyme A by immobilized
bacterial cells. In Immobilized Microbial Cells, ed. K. Venkatsubramanian.
American Chemical Society, Washington, DC, pp. 87-100.
Shimizu, S., Komaki, R., Tani, Y. & Yamada, H. (1983). A high yield method for the
preparative synthesis of coenzyme A by combination of chemical and enzymic
reactions. FEBS Lett., 151, 303-6.
Shimizu, S., Hata, H. & Yamada, H. (1984a). Reduction of ketopantoyl lactone to
D-( - )-pantoyl lactone by micro-organisms. Agric. Bioi. Chem., 48, 2285-9l.
Shimizu, S., Esumi, A., Komaki, R. & Yamada, H. (1984b). Production of coenzyme
A by a mutant of Brevibacterium ammoniagenes resistant to oxypantetheine. Appl.
Environ. Microbiol., 48, 1118-22.
Shimizu, S., Yamada, H., Hata, H., Morishita, T., Akutsu, S. & Kawamura, M.
(1987a). Novel chemoenzymatic synthesis of 0-( - )-pantoyl lactone. Agric. Bioi.
Chem., 51, 289-90.
Shimizu, S., Hattori, S., Hata, H. & Yamada, H. (1987b). Stereoselective enzymatic
oxidation and reduction system for the production of D-( - )-pantoyl lactone from a
racemic mixture of pantoyllactone. Enzyme Microb. Technol., 9,411-16.
Shimizu, S., Hattori, S., Hata, H. & Yamada, H. (1987c). One-step microbial

Vitamin B 5 , Coenzyme A and Reloted Compounds


conversion of a racemic mixture of pantoyllactone to optically active D-( - )-pantoyl

lactone. Appl. Environ. Microbiol., 53, 519-22.
Shimizu, S., Kataoka, M., Chung, M. C. M. & Yamada, H. (1988a). Ketopantoic acid
reductase of Pseudomonas maltophilia 845, Purification, characterization and role in
pantothenate biosynthesis. J. BioI. Chem., 263, 12077-84.
Shimizu, S., Hattori, S., Bata, H. & Yamada, H. (1988b). A novel fungal enzyme,
NADPH-dependent carbonyl reductase, showing high specificity to conjugated
polyketones, purification and characterization. Eur. J. Biochem. 174, 37-44.
Stiller, E. T., Harris, S. T., Finkelstein, J., Keresztesy, J. C. & Folkers, K. (1940).
Pantothenic acid. VIII. The total synthesis of pure pantothenic acid. J. Am. Chem.
Soc., 62, 1785-90.
Suzuki, T., Abiko, Y. & Shimizu, M. (1967). Investigation of pantothenic acid and its
related compounds, XII. Biochemical studies (7). Dephospho-CoA pyrophosphorylase and dephospho-CoA kinase as a possible bifunctional enzyme complex. J.
Biochem., 62, 642-9.
Teller, J. H., Powers, S. G. & Snell, E. E. (1976). Ketopantoate hydroxymethyltransferase, I. Purification and role in pantothenate biosynthesis. J. BioI. Chem., 251,
Vagelos, P. R. (1973). Acyl group transfer (Acyl carrier protein). In The Enzymes, 3rd
edn, Vol. 8, ed. P. D. Boyer. Academic Press, New York, pp. 155-99.
Vallari, D. S., Jackowski, S. & Rock, C. O. (1987). Regulation of pantothenate kinase
by coenzyme A and its thioesters. J. BioI. Chem., 262, 2468-71.
Vandamme, E. J. (1981). Properties, biogenesis and fermentation of the cyclic
decapeptide antibiotic, Gramicidin S. In Topics in Enzyme and Fermentation
Biotechnology, Vol. 5, ed. A. Wiseman. Horwood-Wiley. New York, London, pp.
Wagner, A. F. & Folkers, K. (1964). Pantothenic acid and coenzyme A. In Vitamins
and Coenzymes. Interscience Publishers, New York, pp. 93-137.
West, T. P., Traut, T. W., Shanley, M. S. & O'Donovan, G. A. (1985). A Salmonella
typhimurium strain defective in uracil catabolism and fJ-alanine synthesis. J. Gen.
Microbiol., 131, 1083-90.
Williams, R. J. (1943). The chemistry and biochemistry of pantothenic acid. Adv.
Enzymol., 3, 253-87.
Williamson, J. M. & Brown, J. M. (1979). Purification and properties of L-aspartate-Il'decarboxylase, an enzyme that catalyzes the formation of fJ-alanine in Escherichia
coli. J. Bioi. Chem., 254, 8074-82.
Wittwer, C. T., Burkhard, D., Ririe, K., Rasmussent, R., Brown, J., Wyse, B. W. &
Hansen, R. G. (1983). Purification and properties of a pantetheine-hydrolyzing
enzyme from pig kidney. J. Bioi. Chem., 258, 9733-8.
Yamada, H. & Shimizu, S. (1988). Microbial and enzymatic processes for the
production of biologically and chemically useful compounds. Angew. Chem. Intern.
Eng. Ed., 27,622-42.
Yamada, H., Shimizu, S. & Tani, Y. (1980). Synthesis of coenzymes by immobilized
cell systems. In Enzyme Engineering, Vol. 5, ed. H. H. Weetall and G. Royer.
Plenum Publishing, New York, pp. 405-1l.
Yoshioka, M. & Tamura, Z. (1971). Bifidus factors in carrot. II. The structure of the
factor in fraction IV. Chem. Pharm. Bull., 19, 178-85.
Yoshioka, T., Kojima, I., Isshiki, K., Watanabe, A., Shimauchi, Y., Okabe, M. &
Fukagawa, Y. (1983). Structure of OA-6129A, Bh B2 and C, new carbapenem
antibiotics produced by Streptomyces sp. OA-6129. J. Antibiot., 36, 1473-82.

Chapter 13


Research Center for Cell and Tissue Culture, Faculty of Agriculture, Kyoto
University, Kyoto 606, Japan

In 1934, vitamin B6 (or pyridoxine) was discovered by Gyorgy as a rat pellagra
preventive factor. Soon after the determination of its chemical nature the total
synthesis of the vitamin was accomplished. Moreover, clinical investigations
have clarified that the vitamin has therapeutic and prophylactic uses for many
diseases. Since microbiological assays pointed towards several forms of the
vitamin B 6-active compound the enzymatic conversion of various forms of
vitamin B6 into pyridoxal 5 '-phosphate (pyridoxal-P), the active form of the
vitamin in vivo, has been intensively investigated. Pyridoxine, pyridoxamine,
pyridoxal and their 5 -phosphate esters are now recognized as principal vitamin
B6 compounds (Fig. 1).
Since Gunsalus & Bellamy first reported in 1944 that pyridoxal-P is the
coenzyme of L-tyrosine decarboxylase, it has become evident that pyridoxal-P
or pyridoxamine 5' -phosphate (pyridoxamine-P) is a coenzyme for various
enzyme systems associated with the metabolism of amino acids, amines,
saccharides, fatty acids, etc. A large number of vitamin B6 enzymes has now
been found, and the mechanisms of vitamin B6 dependent reactions have been
extensively investigated.
Several biotechnological methods for synthesis of vitamin B6 compounds
have been investigated and are described in this chapter. However, the
industrial production is still limited to chemical synthesis processes for
pyridoxine and pyridoxal-Po


Pyridoxine (3-hydroxy-4,5-dihydroxymethyl-2-methylpyridine) exhibits the

properties of a stable hydroxylated weak nitrogen base. Hydrolytic agents such

Y. Tani




CH2 0R2

,9'4 5


CH 3

1 61

Pyridoxine: R\ = CH 20H, R2 = H
Pyridoxal: R\ = CHO, R2 = H
Pyridoxamine: R\ = CH2NH2, R2 = H
Pyridoxine 5'-phosphate: R\ = CH20H, R2 = P0 3 H 2
Pyridoxal 5'-phosphate: Rl = CHO, R z = P03HZ
Pyridoxamine 5'-phosphate: Rl = CHzNHz, Rz = P03H2
Fig. 1. Chemical structure of vitamin B 6

as mineral acids or aqueous alkali, hot or cold, do not affect the vitamin. With
ferric chloride, pyridoxine reacts as a phenolic substance giving a reddish
brown coloration. In alkaline solution, pyridoxine on treatment with 2,6dichloroquinone chlorimide gives an immediate blue color fading to reddishbrown.
Pyridoxine hydrochloride occurs as white platelets, melting point 204-206C
with decomposition. The free base melts at 160C. The compound is optically
inactive. Both base and hydrochloride readily sublime without decomposition.
The hydrochloride is freely soluble in water but sparingly in alcohol and
acetone. The base is soluble in methanol. Rapid destruction of pyridoxine by
light occurs in neutral and alkaline solutions. In 01 M hydrochloric acid there is
very little destruction.
The tautomeric properties of pyridoxine are well illustrated by the changes
in its UV absorption produced by varying the hydrogen ion concentration. The
single maximum at 2925 nm at pH 2 diminishes in intensity at pH 45, and
concomitantly a new maximum appears at 3275 nm. This latter band increases
in intensity when the pH is changed to 675, and the 2925 nm maximum
disappears but a new band appears at 2560 nm. When the pH is further raised
to 102, both bands increase in intensity and shift to shorter wavelengths.
The existence of other forms of pyridoxine was recognized as a result of the
comparison of microbiological assays on extracts of natural materials with the
values based on chemical and animal assays.
When pyridoxine was treated with mild-oxidizing agents, a marked increase
in l?io-activity towards the micro-organism, Lactobacillus casei, was observed.
Autoclaving of pyridoxine in the presence of the assay medium or amino acids,
greatly increased the activity of pyridoxine towards the test organism,
Streptococcus faecalis R. The products formed by treating pyridoxine with
animating agents and mild-oxidizing agents were, respectively, the amino and
aldehyde derivatives of pyridoxine. The compounds are pyridoxamine (2methyl-3-hydroxy-4-aminomethyl-5-hydroxymethylpyridine) and pyridoxal (2methyl-3-hydroxy-4-formyl-5-hydroxymethylpyridine) .
The availability of pyridoxal in synthetic form permitted identification of

Microbial Production of Vitamin B6


pyridoxal-P, the coenzyme discovered as an essential cofactor for the enzymatic decarboxylation of amino acids. Pyridoxamine-P was subsequently
discovered as a naturally occurring compound by virtue of its differential
activity in promoting growth of certain lactic acid bacteria. Finally, pyridoxine
5'-phosphate (pyridoxine-P) (although it plays no known essential role in
metabolism) can be formed from pyridoxine by tissue enzymes and further
transformed to pyridoxal-P by other tissue enzymes. It appears therefore to
also be a naturally occurring compound. Direct analysis of many tissue extracts
also indicates its natural occurrence.
A number of microbiological procedures are established for the estimation
of vitamin B 6. The method with Saccharomyces cerevisiae (S. carlsbergensis, S.
uvarum) is widely used for the assay of total vitamin B6 in non-esterified form.
The phosphates of vitamin B6 should be converted to the free form by acid or
enzymatic hydrolysis before applying the bioassay. A differential assay
employs Lactobacillus casei which responds to pyridoxal and Streptococcus
faecalis R which responds to pyridoxal and pyridoxamine. Chemical reactions
with 2,6-dichloroquinone chlorimide and diazotized p-aminoacetophenone are
less sensitive and non-specific but are useful for the detection on paper
chromatograms and paper electrophoregrams. Recent progress in high performance liquid chromatography is applied for quantitative and simultaneous
detection of all vitamin B6 compounds by determining UV absorption and


The current knowledge of the biosynthesis and metabolism of vitamin B6 is
summarized in Fig. 2.
Although the de novo biosynthesis of vitamin B6 has been investigated since
the 1940s using micro-organisms, the complete biosynthetic pathway still
remains obscure. Clear evidence of a direct precursor in the vitamin B6
biosynthesis was demonstrated for the first time by Tani & Dempsey (1973). In
a vitamin B6 requiring WG3 strain mutant derived from Escherichia coli B, the
vitamin B6 requirement could be satisfied by glycolaldehyde, and 14C_
glycolaldehyde was incorporated into vitamin B6 without significant dilution.
Subsequently, the biosynthetic route of glycolaldehyde in E. coli (Tani et al.,
1977) and the incorporated position of glycolaldehyde in vitamin B6 (Hill et al.,
1975) were elucidated. Based on these results, a hypothetical sequence of
vitamin B6 biosynthesis, is presented in Fig. 3 in which the condensation of
glycolaldehyde with a 'C4 compound' forming a 'C6 compound' such as
a,p-dihydroxy-p-hydroxymethyl valerate, as well as the following introduction
of a 'C3 compound' are involved. Another route, in which all carbon atoms of
pyridoxine are derived from trioses, is also proposed (Vella et aI., 1980).
As to the biodegradation of vitamin B6, research was performed with

Y. Tani



D-Glutamate 2-Ketoglutarate



'-. '-. A

\. .J.

Pyn'doxme. P

~~i~hOOPhat" ATPt~o'Pha". ~~i~PhO'Pha",


""" Pyridoxine
Pyridoxine glUCOSi1

Pyridoxine glucuronide

5-Pyridoxic acid lactone


5-Pyridoxic acid
a-Hydroxymethyl-a' -(N-acetylaminomethylene )succinate

a-Hydroxymethylsuccinate semi aldehyde + acetate + NH3 + cO 2




4-Pyridoxic acid lactone

2-Methyl-3-hydroxy-5-formyl4-Pyridoxic acid




a-(N -acetylaminomethylene)succinate

Succinate semialdehyde + acetate

+NH 3 +C02

Fig. 2. Metabolism of vitamin B 6

pseudomonads, to which pyridoxine and pyridoxamine were fed as carbon

and/or nitrogen source (Nyns et al., 1969).
The biosynthesis of pyridoxal-P from pyridoxine is considered to occur
through pyridoxine-+ pyridoxine-P-+ pyridoxal-P rather than pyridoxine-+
pyridoxal-+ pyridoxal-P, based on experimental results so far obtained.
Pyridoxine is transformed into pyridoxal by an oxidase or a dehydrogenase.
Pyridoxine oxidase activity was found in an enzyme preparation extracted from
rabbit liver, and the oxidase reaction of pyridoxine to pyridoxal proceeded
only in the presence of a coupling reaction with an aldehyde oxidase which
oxidized pyridoxal to 4-pyridoxic acid. Pyridoxine dehydrogenase requires
NADP as a coenzyme. Both oxidase and dehydrogenase reactions are

Microbial Production of Vitamin B6




~~~;;~~'::ed) H}r-'-CH'OH C, COrU~d


**CH2 0H

.. a, f;I-dihydroxy

Fig. 3. Hypothetical biosynthetic sequence of pyridoxine.

reversible, though the reaction equilibria favour the formation of pyridoxine

from pyridoxal.
The phosphorylation of free vitamin B6 is catalyzed by pyridoxal kinase,
which phosphorylates pyridoxine and pyridoxamine as well as pyridoxal in the
presence of A TP to form the corresponding phosphate esters. The enzyme
activity occurs widely in mammalian tissues, bacteria, molds, and yeasts.
Another phosphotransferring reaction using various organic phosphorus
compounds was found in many micro-organisms such as molds, yeasts, and
bacteria, and found to be catalyzed by the phosphotransferring ability of
phosphatase (Tani & Ogata, 1972).
Pyridoxine-P oxidase catalyzes the oxidative deamination of pyridoxamine-P
as well as the oxidation of pyridoxine-Po The coenzyme of the enzyme is FMN.
The enzyme activity is widely present in bacteria and it is strongly inhibited by
compounds phosphorylated at the 5-position of the pyridine ring, but weakly
inhibited by compounds having an aldehyde residue at the 4-position of the
pyridine ring (Yamamoto et al., 1965). The fact that the enzyme is strongly
inhibited by pyridoxal-P (the reaction product) might be attributed to
metabolic regulation of the in vivo biosynthesis of pyridoxal-P. It seems likely
that pyridoxine-P or pyridoxamine-P stored in cells is oxidized to pyridoxal-P
when necessary.

Y. Tani


Anaerobic bacteria, clostridia, lack the pyridoxine-P oxidase activity but

have the pyridoxamine-P-2-ketoglutarate transaminase activity which is absent
in aerobic bacteria. The transaminase purified from cell-free extract of
Clostridium kainantoi uses 2-ketoglutarate as the amino group acceptor and
D-glutamic acid as the amino group donor (Tani et al., 1972b).


Production of Vitamin 8 6 by Fermentation

Extensive screening for vitamin B6 producers among micro-organisms has been

pedormed. However, the productivity was quite insufficient to establish an
industrial microbiological process of vitamin B6 production. Almost all
micro-organisms synthesize vitamin B6 intracellularly 005-03 mg per g of dry
cells, and excrete it in amounts of 005-05 mg per liter of culture medium.
Table 1 is a list of relatively high producers. The extracellular vitamin B6 is the
free form.
The biosynthesis of vitamin B6 in E. coli is known to be repressed by a
feedback control system. However, in a high producer, Flavobacterium sp.
238-7, the amount of vitamin B6 excreted was increased even when cultivated
with the addition of each form of vitamin B6 (Tani et al., 1972a). An enzyme
which reduces glycolate to glycolaldehyde as the first step of the biosynthesis of
vitamin B6 is present in the particulate fraction of Flavobacterium sp. 238~7.
The activity and synthesis of the enzyme was not affected by added vitamin B6
(Tani et al., 1984). From these results, Flavobacterium sp. 238-7 can be
considered to lack this feedback control system for the vitamin B6 biosynthesis,
resulting in the formation of large amounts of vitamin B 6
A typical time course of vitamin B6 production by Flavobacterium sp. 238-7
in a glycerol-peptone medium is shown in Fig. 4.
Table 1
Extracellular Production of Vitamin B6 by Micro-organisms

Klebsiella sp.
Achromobacter cycloclastes
Flavobacterium sp. 238-7
Bacillus subtilis
Candida albicans (n-paraffin-grown)
Saccharomyces marxianus
Pichia guilliermondii (n-paraffin-grown)

Vitamin B6 produced (mg/liter)




Microbial Production of Vitamin B6





0.6 '"...








0.4 ....01

....<I 4




0.2 .c...:-




Cultivation time (h)

Fig. 4. Production of vitamin B6 by Flavobacterium sp. 238-7.


Production of Pyridoxal-P from Pyridoxine-P and Pyridoxine

As pyridoxine and pyridoxine-P can be chemically synthesized with ease, the

industrial production of pyridoxal-P from pyridoxine and pyridoxine-P by
pyridoxine-P oxidase catalysis with or without enzymatic phosphorylation is
considered to be worthy of study.
Pyridoxine-P is oxidized by a flavoprotein, pyridoxine-P oxidase. It was
found that amino acids accelerate the enzyme reaction due to relief of the
aldehyde inhibition by pyridoxal-P of the enzyme activity, in which these
amino acids combined non-enzymatically with an aldehyde residue of
pyridoxal-P to form a Schiff base. In view of applying this enzyme reaction for
industrial production of pyridoxal-P, a suitable industrial medium for
Alcaligenes faecalis was constructed with corn steep liquor, ammonium sulfate
and glycerol. The cells obtained were autolyzed by addition of toluene or ethyl
acetate and used as the enzyme source. The reaction proceeded adequately on
addition of aspartic acid or glutamic acid, which are relatively cheap, and
resulted in a complete conversion of pyridoxine-P to pyridoxal-P.
An acid phosphatase of Escherichia freundii Kl can phosphorylate pyridoxine as the most effective phosphoryl acceptor among free vitamin B6 using
various kinds of organic phosphates as the phosphoryl donor. As E. freundii
had pyridoxine-P oxidase and pyridoxine dehydrogenase activities together
with the phosphotransferase activity, a considerable amount of pyridoxal-P was
formed as well as pyridoxine-P in the phosphotransferring reaction of


Y. Tan;

pyridoxine using cells or cell-free extract of the organism (Tani et al., 1969). In
particular, an appreciable amount of pyridoxal-P was formed by use of
phenylphosphate as the phosphoryl donor under alkaline conditions. Under
the optimal reaction conditions, 04 g of pyridoxal-P was formed from 2 g of
pyridoxine incubated with phenylphosphate and cells for 24 h.

Production of Pyridoxal from Pyridoxine

Oxidation of pyridoxine to form pyridoxal is catalyzed by pyridoxine dehydrogenase. A biochemical process for production of pyridoxal from pyridoxine was studied with the pyridoxine dehydrogenase of dried cells of baker's
yeast (Ogata et al., 1968). The reaction equilibrium of the enzyme lies to the
reduction of pyridoxal to pyridoxine. However, a large amount of pyridoxal
(pyridoxal-carbonyl complex) was accumulated on addition of various carbonyl reagents to the reaction mixture. For example, the formed pyridoxal was
non-enzymatically converted into pyridoxal-semicarbazone on addition of
semicarbazide, and the maximum yield of pyridoxal reached 80% based on the
amount of pyridoxine added. Pyridoxal-semicarbazone is highly insoluble in
water and can easily be isolated by crystallization from the reaction mixture.

Production of Vitamin B6 Derivatives

A number of vitamin B6 derivatives such as vitamin B6 fatty acid esters and

vitamin B6 amino acid compounds are synthesized chemically. On the other
hand, few naturally occurring derivatives have been found: glucuronide of
pyridoxine is found in human urine, sulfates (3-position) of pyridoxine or
pyridoxal are formed by a homogenate of rat liver, and an indole-pyridoxal
complex is found in milk.
It was also found that a new derivative of vitamin B6 was formed in the
culture filtrates of some strains belonging to the genera Sarcina and
Micrococcus, cultivated in medium containing pyridoxine and sucrose, maltose
of phenyl-a-glucoside as carbon source. the derivative isolated from the
culture broth consisted of two components. These were separated by Dowex
1 x 2 (borate form) column chromatography and identified as pyridoxine
4'-a-D-glucoside and pyridoxine 5'-a-D-glucoside (Fig. 5) by comparison with
chemically synthesized samples (Ogata et al., 1969). Under optimal conditions
using Micrococcus sp. No. 431, pyridoxine glucoside, 35-37 g/liter, was
produced in the culture broth of the organism from 2 g of pyridoxine per liter.
The pyridoxine glucoside-synthesizing enzyme was purified and characterized to be as a-glucosidase having the ability of transglucosidation. The
bioactivity of pyridoxine glucoside using S. cerevisiae as the test organism was
about 20% of that of its molar equivalent of pyridoxine, and it increased with
incubation time, suggesting the gradual hydrolysis of pyridoxine glucoside into

Microbial Production of Vitamin B6



H3 C


Pyridoxine 4' -a-D-glucoside

Fig. 5. Chemical structure of pyridoxine glucoside.

Pyridoxine glucoside is relatively more stable than pyridoxine toward heat

and UV irradiation. Pyridoxine glucoside could be transported into rabbit
erythrocytes. The transport of pyridoxine glucoside was about 10-20% of that
of its molar equivalent of pyridoxine (Kawai et al., 1972).

Gunsalus, I. C. & Bellamy, W. D. (1944). A function of pyridoxal. 1. BioI. Chern.,
Hill, R. E., Horsewood, P., Spenser, I. D. & Tani, Y. (1975). Biosynthesis of vitamin
B 6 Incorporation of glycolaldehyde into pyridoxal. 1. Chern. Soc. Perkin, I,
Kawai, F., Yamada, H. & Ogata, K. (1972). Studies on transglycosidation to vitamin
B6 by micro-organisms. VII. Transport of pyridoxine glucoside into rabbit erythrocytes. 1. Vitarninol., 18, 183-8.
Nyns, E. J., Zach, D. & Snell, E. E. (1969). The bacterial oxidation of vitamin B6
VIII. Enzymatic breakdown of a-(N-acetylaminomethylene) succinic acid. 1. Bio!.
Chern., 244,2601-5.
Ogata, K., Tochikura, T., Kawata, S., Yamamoto, S. & Kawai, H. (1968). Pyridoxal
derivative formation from pyridoxine by baker's yeast, Agric. BioI. Chern., 32,
Ogata, K., Uchida, Y., Kurihara, N., Tani, Y. & Tochikura, T. (1969). Studies on
transglycosidation to vitamin B6 by micro-organisms. II. Chemical structure of
pyridoxine glucoside. 1. Vitarninol., 15, 160-66.
Gyorgy, P. (1934). Vitamin B6 and pellegra-like dermatitis in rats. Nature, 133,
Tani, Y. & Dempsey, W. B. (1973). Glycolaldehyde is a precursor of pyridoxal
phosphate in Escherichia coli B. 1. Bacterio!., 116,341-5.


y. Tani

Tani, Y. & Ogata, K. (1972). Studies on vitamin B6 metabolism in micro-organisms.

Part IX. Microbial phosphorylation of vitamin B6 through a new phosphotransferring reaction (5). Phosphorylation of pyridoxine by several phosphatases. Agric.
Bioi. Chem., 36, 173-80.
Tani, Y., Kawata, S., Tochikura, T. & Ogata, K. (1969). Studies on vitamin B6
metabolism in micro-organisms. Part VIII. Microbial phosphorylation of vitamin B6
through a new phosphotransferring reaction (4). Formation of pyridoxal 5'phosphatae from pyridoxine. Agric. BioI. Chem., 33,314-20.
Tani, Y., Nakamatsu, T., Izumi, Y. & Ogata, K. (1972a). Studies on vitamin B6
metabolism in micro-organisms. Part XI. Extracellular formation of vitamin B6 by
marine and terrestrial micro-organisms and its control. Agric. Bioi. Chem., 36,
Tani, Y., Ukita, M. & Ogata, K. (1972b). Studies on vitamin B6 metabolism in
micro-organisms. Part X. Further purification and characterization of pyridoxamine
5'-phosphate-a-ketoglutarate transaminase from Clostridium kainantoi. Agric. BioI.
Chem., 36, 181-8.
Tani, Y., Morita, H. & Ogata, K. (1977). Glycolaldehyde synthesizing pathway
involved in vitamin B6 biosynthesis in Escherichia coli B. Agric. BioI. Chem., 41,
Tani, Y., Nishise, H., Morita, H. & Yamada, H. (1984). Vitamin B6 biosynthesis and
glycolate reductase in Flavobacterium sp. 238-7. J. Nutr. Sci. Vitaminol., 30,
Vella, G. J., Hill, R. E., Mootoo, B. S. & Spenser, I. D. (1980). The status of
glycolaldehyde in the biosynthesis of vitamin B6 J. BioI. Chem., 255, 3042-8.
Yamamoto, S., Tochikura, T. & Ogata, K. (1965). Studies on vitamin B6 metabolism in
micro-organisms. Part III. Pyridoxine phosphate and pyridoxamine phosphate
oxidation (3). Product inhibition and reactivation with amino compounds. Agric.
Bioi. Chem., 29,597-604.

Chapter 14


Department of Agricultural Chemistry, Kyoto University, Kyoto 606, Japan

In 1927, Boas described a skin injury produced in rats by feeding raw egg
white, as well as the occurrence in various foodstuffs of a 'protective factor X'
that prevented and cured this injury (Boas, 1927). Gyorgy et al. (1939) tracked
down this protective factor and called it vitamin H after the German Haut for
skin. Kogi & Tonnis (1936) isolated a yeast growth factor, biotin vitamin B8
from egg yolk in the form of its crystalline methyl ester and Kogi (1937)
determined its empirical formula. Biotin was later shown to be identical to
vitamin H, the protective factor X, and to coenzyme R. Melville et al. (1942)
determined the correct structure of biotin, which was later confirmed by total
chemical synthesis in the Merck laboratories (Harris et al., 1943, 1944a,b,
1945) and verified by X-ray crystallography (Traub, 1956; Trotter & Hamilton,
Studies on the biosynthesis of biotin started in 1937 with studies on the
nutritional requirements of micro-organisms (Mueller, 1937a, b; du Vigneaud
et al., 1942; Eakin & Eakin, 1942; Dittmer & du Vigneaud, 1944; Dittmer &
du Vigneaud, 1944; Lilly & Leonian, 1944; Stokes & Gunness, 1945).
Subsequently, through radiochemical and genetic-biochemical studies (Tatum,
1945; Pontecorvo, 1953; Ryan, 1956), part of the biosynthetic pathway was
proposed. Thereafter, a hypothetical pathway was established by Okumura et
al. (1962a,b) based on their investigations of the microbial production of
glutamic acid using biotin-requiring bacteria. Ogata et al. (1965a,b) and
Iwahara et al. (1966a,b,c) made extensive investigations of biotin biosynthesis
and verified the main pathway described by Okumura et al. (1962a,b) using
growing and resting cells: pimelic acid - 7-keto-8-aminopelargonic acid
(KAPA)-7,8-diaminopelargonic acid (DAPA)-dethiobiotin (DTB)biotin (Izumi & Ogata, 1977). Through the extensive studies of three groups,
Eisenberg (1973), Pai (1971) and Izumi (Izumi & Ogata, 1977), the enzymic


Y. Izumi & H. Yamada

system involved in the biosynthetic pathway from pimelic acid to DTB had
been established by 1975.
In the 1950s it became clear that biotin was involved as a cofactor in many
biochemical processes, especially a number of carboxyl transfer reactions.
Investigations on the function of biotin in the reaction mechanisms were
obstructed by the fact that biotin is active only when covalently attached to an
e-amino group of a lysyl residue of the enzyme (Kosow et al., 1962), as was
suggested earlier by the isolation of biocytin (e-N-biotinyllysine) from yeast
extract (Wolf et al., 1952).
Wakil et al. (1958) and Wakil & Gibson (1960) showed that highly purified
acetyl-CoA carboxylase, isolated from avian liver extracts, was enriched with
respect to biotin and sensitive to inhibition by avidin, suggesting the tight
binding of the vitamin to the carboxylase and its probable role in acetyl-CoA
carboxylation. Lynen et al. (1959) made the interesting observation that the
A TP- and Mi+ -dependent carboxylation of free biotin was catalyzed by a
bacterial p-methylcrotonyl-CoA carboxylase. This model reaction led to the
formation of an unstable carboxybiotin derivative which was identified as
1'-N-carboxybiotin (Knappe et al., 1961; Lynen et al., 1961). Subsequently, it
was demonstrated with several biotin enzymes that the site of carboxylation of
the enzyme-bound prosthetic group was the same as that of free biotin (Moss
& Lane, 1971).
The existence of separate subsites for the catalysis of each partial reaction
has been unequivocally demonstrated with acetyl-CoA carboxylase of
Escherichia coli (Fall & Vagelos, 1972; Polakis et al., 1974). These investigations have shown that there are three dissimilar subunits involved in catalysis:
(a) the biotin carboxyl carrier protein (CCP), (b) the biotin carboxylase (BC),
which catalyzes the carboxylation of the CCP biotin, and (c) the carboxyl
transferase (Cf), which catalyzes the carboxyl transfer from the CCP biotin to
acetyl-CoA. Wood's group (Wood & Barden, 1977) has unequivocally
demonstrated with transcarboxylase that each partial reaction is catalyzed by a
separate subunit and that a distinct CCP subunit is present.
Major portions of the amino acid sequence of the CCP from transcarboxylase of Propionibacterium shermanii and acetyl-CoA carboxylase from
Escherichia coli have been determined (Wood & Barden, 1977). Up to now,
the sequences of the portions around biotin of the CCP or the corresponding
domains have been elucidated with almost all the biotin enzymes including
chicken liver acetyl-CoA carboxylase (Takai et al., 1987), human pyruvate
carboxylase (Lamhonwah et al., 1987) and human propionyl-CoA carboxylase
(Lamhonwah et al., 1987). Recently, Takai et al. (1988) have deduced for the
first time the complete amino acid sequence of a biotin enzyme, chicken liver
acetyl-CoA carboxylase.
The structure of biotin is shown in Fig. 1. The naturally occurring biotin,
(+ )-cis-hexahydro-2-oxo-1H-thieno-(3,4-d)-imidazole-4-n-valeric acid, is ex-

Microbial Production of Biotin


* asymmetric carbon

Fig. 1. Chemical structure (a) and absolute configuration (b) of ( + )-biotin.

pressed in many ways: d-biotin, d-( + )-biotin, D-biotin and D-( + )-biotin.
Here, we describe it as (+ )-biotin, and unless otherwise stated, the term
'biotin' indicates ( + )-biotin. The structure of crystalline biotin has also been
investigated by X-ray diffraction techniques and is now known precisely
(Traub, 1956; Trotter & Hamilton, 1966; Stallings & de Titta, 1985). X-ray
crystallographic analysis of (+ )-biotin and of the carboxybiotin derivative
described previously revealed that the bicyclic ring system has a boat-like
configuration (Fig. 1(B. The planar ureido ring projects upward at an angle
of 620 with respect to the plane (plane A) formed by the four carbon atoms
of the tetrahydrothiophene ring. Another plane comprising S-1, C-2, and C-5
tilts upward at an angle of 37 6 with respect to plane A (Bonnemere et al.,
1965). Because of the cis orientation of the ureido ring with respect to the
aliphatic side chain, C 6 resides approximately 28 A from the 3' -N of the
ureido ring (Traub, 1956). Repulsion between the 3'-N and C-6 positions is
thought to cause the greater-than-anticipated C 3-Cz-C6 angle of 119
(Traub, 1956).
There are seven other forms which can be chemically synthesized: (-)biotin, (+)- and (- )-epibiotin (cis-form); (+)- and (- )-allobiotin, and
( + )- and ( - )-epiallobiotin (trans-form).
The crystals of ( + )-biotin are colorless, fine long needles: m.p. 232-3C,
[{1']22=91 (c=1, 01NNaOH). The isoelectric point is pH35. The pH of
001 % aqueous solution is 45. Solubility in water is about 22 mg/100 ml at
25C and it is more soluble in hot water or in dilute alkali. Solubility in 95%
ethanol is about 80 mg/100 ml at 25C. It is insoluble in other common organic
solvents. It is stable in air, in a wide temperature range, and up to about pH 9.
It is also stable after autoclaving in 6 N H 2S04 at 120C for 1 h. Accordingly,
most of the naturally occurring biotin bound to protein can be hydrolyzed to
free forms by such autoclave treatments without any problem.
Biotin combines with avidin, a protein in raw egg white, and streptavidin, a
protein produced by an actinomycete to become inactive.
Various chemical reactions with biotin and their products are shown in
Fig. 2.

Y. Izumi & H. Yamada







ammo a c l d /

Os -"2'.
(t'U \


0'" "0




JL /




biotin azide



thienyl"aleric acid

, ,

~ H2N

3 HN



t(CH3J2S0 4
NH2 (HNO,)

-- H


~ H, (Rane, Ni)

bIotln~ methyl ester





biotin sulfone


biotin acid '-- HN NH

sOO--- I--l


biotin amide

Qlotln sulfoxide
















adipic acid






gonic acid




biotin hydrazide

Fig. 2. General chemical reactions of biotin and their products.


Using about 1000 strains of molds, bacteria and actinomycetes, Ogata et al.
(1965a,b) and Iwahara et al. (1966a,b) investigated the accumulation of
biotin-vitamers when pimelic acid was added to media as a precursor. They
found a promoting effect of pimelic acid in a large number of strains. On
addition of pimelic acid to the medium of Bacillus sphaericus IFO 3525,
20-200 Ilg/ml of 'total biotin', which is the amount of biotin-vitamer given by
the bioassay using Saccharomyces cerevisiae and includes DTB, KAP A and
DAPA as well as biotin, was accumulated. This was several hundred times
more than has hitherto been reported in micro-organisms. The main component of the biotin-vitamers formed from pimelic acid was (+ )-DTB.
Afterwards, Yamada et al. (1983) found this strain also produced relatively
large amounts (08-11Ilg/ml) of biotin from pimelic acid, as well as from
DTB, under optimized conditions.
Ogino et al. (1974a,b) demonstrated a production method using an
n-paraffin-utilizing bacterium from a new compound, oL-cis-tetrahydro-2-oxo4-n-pentylthieno-(3,4-d)-imidazoline (oL-TOPTI), which is a biotin analog
having a methyl group instead of a carboxyl group of the biotin molecule.
TOPTI was chemically synthesized from Nt ,N3 -dibenzyl-oL-cis-tetrahydrothieno-(3,4-d)-imidazoline-2,4-dione (compound VII in Fig. 6) via a
Grignard reaction with n-pentyl magnesium bromide, dehydration in the

Microbial Production of Biotin


presence of an acidic catalyst, catalytic hydrogenation, and debenzylation with

concentrated HBr. In this production method, n-paraffin was used as a carbon
and energy source for cell growth with concurrent transformation (co-oxidation
of TOPTI to biotin). First, n-paraffin-utilizing micro-organisms that cooxidized TOPTI were selected from natural sources. Three strains, identified
as Corynebacterium sp., were the most excellent producers of biotinol and
biotin from TOPTI. Therefore the conversion of TOPTI to biotin was assumed
to occur via p-oxidation. In a medium containing 2% n-paraffin and 02%
urea, with the addition of 50 mg DL-TOPTI/100 ml after 24 h of cultivation, a
maximum conversion of about 60% (032 mg ( )-biotin/ml) was obtained
96 h after the addition of TOPTI. However, selective degradation of (+)biotin occurred after prolonged incubation, leaving the ( - )-isomer. Thus, to
avoid such degradation, mutants that were incapable of assimilating n-paraffin
and of degrading biotin, but capable of utilizing acetate, were derived. One of
the mutants produced 065 mg/ml of ( )-biotin from 08 mg/ml of DL-TOPTI
(805% conversion).




Biotin can be synthesized by a great number of micro-organisms, while some

organisms which cannot synthesize biotin require it for growth. Biotin is also
known to be synthesized by plants (Watanabe et al., 1982). Almost all the
studies on the biosynthesis of biotin have used micro-organisms. The pathway
of biotin biosynthesis has been established, as shown in Fig. 3. All the enzymes
which are involved in the reaction from pimelic acid to DTB have been
elucidated and were found to be novel types. Table 1 summarizes the
properties of pimelyl-CoA synthetase, KAPA synthase, DAPA aminotransferase and DTB synthetase (Izumi et al., 1980; Izumi, 1984).
The final step from DTB to biotin through a sulfur introduction has not been
enzymically resolved yet. However, there have been some reports on the
biosynthesis using growing cells and resting cells. Yamada et al. (1983) showed
that resting cells of B. sphaericus IFO 3525 exhibited high activity of biotin
synthesis from DTB. Izumi et at. (1973) also showed that resting cells of
Rhodotorula glutinis which form appreciable amounts of biotin from DTB,
formed biotin from DTB only in the presence of methionine, particularly the
L-form. The isotopic experiment revealed that sulfur contained in one molecule
of L-methionine was incorporated into one molecule of biotin.
Parry & Kunitani (1976), Guillerm et al. (1977), Parry & Naidu (1980) and
Trainor et al. (1980) have developed a stereospecific synthesis of DTB and
examined the mechanism of the conversion of DTB to biotin by using
specifically labelled 3H-DTB and growing cells of Aspergillus niger and E. coli,
respectively. They concluded that the introduction of sulfur at C-1 and C-4


Y. Izumi & H. Yamada

Pimelic acid

Pimelyl CoA

7-Keto-B'aminopelargonic acid

78-0iaminopelargonic acid



Fig. 3. Biosynthetic pathway of biotin. CD Pimelyl-CoA synthetase, (6) KAPA synthase,

OAPA aminotransferase, @) OTB synthetase.

positions of DTB takes place with the loss of two hydrogen atoms at

C-1 and C-4 and without the loss of hydrogen atoms at C-2 or C-3. These
results suggest that unsaturation does not occur at C-2 or C-3.
Table 2 (Izumi et al., 1981) shows that among the strains of bacteria and
yeasts tested for activities of the four biotin biosynthetic enzymes, only B.
sphaericus IFO 3525, a DTB producer as described previously, showed
significant activities for all four enzymes.

Microbial Production of Biotin


Table 1
Properties of Biotin Biosynthetic Enzymes
(a) Pimelyl-CoA synthetase of Bacillus megaterium
Optimum temperature
Optimum pH
Km value: pimelic acid
Substrate specificity
Nucleotide requirement
Metal ion requirement

27 x 10- 4 M
55 x 10- 4 M
15 x to- 3 M
15 X to- 3 M
Pimelic acid
(Other dicarboxylic
acids were inert)
Mi+, Mn2+
EDTA, a, a' -dipyridyl,

(b) KAPA synthase of Bacillus sphaericus

Optimum temperature
Optimum pH
Thermostability (10 min)
Amino acid specificity
Coenzyme requirement

0-6O"C (100%)
Phenylhydrazine, semicarbazide,
oL-penicillamine, isoniazid,
o-phenanthroline, citrate,
L-cysteine, glycine, o-alanine
L-serine, o-,L-histidine

(c) DAPA Aminotransferase of Brevibacterium divaricatum

Molecular weight
Optimum temperature
Optimum pH
Thermostability (10 min)
pH Stability (30 min)
Amino donor specificity
Amino acceptor specificity
Coenzyme requirement
Km values: KAPA
pyridoxal 5' -phosphate
pyridoxamine 5' -phosphate

280, 320, 410 nm
O-60C (100%)
70-100 (100%)
KAPA (100)
7-Amino-8-ketopelargonate (1)
Pyridoxal 5' -phosphate
Pyridoxamine 5' -phosphate
069 x to- 4 M
055 x 10-3 M
083 X to- 6 M
12 X to- 6 M
Phenylhydrazine, semicarbazide
hydroxylamine, o-cycloserine,
oL-penicillamine, isoniazid,
p-chloromercuribenzoate, HgCI
iodoacetate; Co 2+, Ass+, KCN

Y. Izumi & H. Yamada


Table l---contd.
(d) DTB Synthetase of Pseudomonas graveolens
Optimum temperature
Optimum pH
7 0-80
0-45C (100%)
DAPA (100)
Substrate specificity


Nucleotide requirement
Metal ion requirement
Km values: DAPA
HC03 -




Biotin diaminoCOOH carboxylic acid (10)

ATP (100), CTP (20), UTP (10)

GTP (20), ITP (10)
Mg2+ Mn2+ Fe 2+
1 X l(j-4 M '
1 X 1O- 2 M
5 X 10- 5 M
3 X 10- 3 M
EDTA, a, a' -dipyridyl, o-phenanthroline

Regulation of Biosynthesis

The biosynthetic regulation mechanism has also been elucidated at the enzyme
level. Eisenberg & Star (1968) have reported that KAPA synthase of E. coli is
almost completely repressed by the addition of 5 ng/ml of biotin to the
medium. Eisenberg & Krell (1969) and Pai (1969) observed similar repression
by biotin of DTB synthetase. Izumi & Ogata (1977) found that KAPA
synthase of B. sphaericus and DAPA aminotransferase of Brevibacterium
divaricatum were repressed by the addition of Olllg/ml of biotin to the
medium. Moreover, they found that, in contrast to the complete repression of
DTB synthetase in B. megaterium by O25Ilg/ml of biotin, pimelyl-CoA
synthetase was not repressed by even 1llg/ml of biotin. The biotin synthesizing reaction by resting cells of B. sphaericus was repressed by 1llg/ml of
biotin. In this way, a strong repressive action of biotin has been demonstrated
on all the enzymes between pimelyl-CoA and DTB and in the biosynthetic
step(s) between DTB and biotin. The corepressor has been found to be
biotinyl-AMP in E. coli (Prakash & Eisenberg, 1979). The repressor protein
(Eisenberg et al., 1982) and biotin operon (Szybalski & Szybalski, 1982) have
also been elucidated.



Biotin is known to be degraded by molds, yeasts and bacteria via f3-oxidation

of the side chain of the molecule (Izumi & Ogata, 1977; Tanaka et aI., 1988).
The biotin-degrading bacterium, Mycoplana sp. No. 166, formed bisnorbiotin,
a compound having two less carbon atoms in its side chain than biotin, from

Table 2


Escherichia coli AKU 007

Klebsiella pneumoniae IFO 12059
Enterobacter aerogenes IFO 12010
Alcaligenes faecalis IFO 3160
Bacillus megaterium NI 8100
Bacillus roseus lAM 1257
Bacillus sphaericus IFO 3525
Brevibacterium divaricatum NRRL 2311
Pseudomonas graveolens IFO 3460









Specific activityc (units/mg protein)

The amounts of biotin-vitamers produced by organisms grown with pimelic acid.

The amount of biotin-vitamer given by the bioassay using Lactobacillus plantarum, which includes biotin and biotin
C 1, pimelyl-CoA synthetase; 2, KAPA synthetase; 3, DAPA aminotransferase; 4, DTB synthetase.





Total biotin

True biotin b

Saccharomyces kloeckerianus IFO 0016

Lipomyces starkeyi IFO 0678
Sporobolomyces salmonicolar IFO 0374
Sporobolomyces salmonicolor IFO 1038
Sporobolomyces coprophilus IFO 1442
Rhodotorula glutinis IFO 0415



Biotin-vitamer Producing Abilitiesa and Biotin Biosynthetic Enzyme Activities of Various Yeasts and Bacteria








Y. Izumi & H. Yamada


biotin, as well as j3-hydroxybisnordethiobotin and tetranordethiobiotin from

DTB in a resting cell system (Osakai et ai., 1986a). Biotinyl-CoA synthetase,
the first enzyme involved in biotin degradation, was purified to homogeneity
from Mycopiana sp. No. 166 (Yamada et ai., 1984; Tanaka et aI., 1986, 1988).
The enzyme was a monomer with a molecular weight of 55 000. The enzyme
catalyzes the stoichiometric conversion of biotin, ATP and CoA into biotinylCoA, AMP and inorganic pyrophosphate. DTB is also effective as a substrate.


As described previously, a strong feedback repression by biotin seems to be

one of the main problems that makes microbial production difficult. This
control must be overcome in order to microbially produce large quantities of
biotin and its vitamers. Some potent biotin antimetabolites have already been
found (Fig. 4) (Izumi & Ogata, 1977). Therefore, it has become possible to use
these antimetabolites for the selection of regulatory mutants producing biotin
and its vitamers.

Actithiazic acid






5-(2-Thienylhlvalerie acid



Fig. 4. Various biotin antimetabolites.

Microbial Production of Biotin


Pai (1975) and Eisenberg et al. (1975) isolated a-dehydrobiotin-resistant

mutants of E. coli, which showed enhanced excretion levels of biotin,
derepressed levels of the biotin biosynthetic enzymes, and resistance to
repression by biotin. However, the level of their accumulation was still quite
low (less than 100 ng/mI). Since actithiazic acid, or acidomycin (ACM) (Izumi
et al., 1981) and 5-(2-thienyl)-n-valeric acid (TVA) (Izumi et aI., 1978) were
found to be biotin antagonists which inhibited the biosynthetic step of biotin
from DTB and the DAPA aminotransferase reaction, respectively, Yamada et
al. (1983) and Tanaka et al. (1988) induced ACM- and/or TV A-resistant
mutants from B. sphaericus IFO 3525. ACM- and TV A-resistant mutant,
AB12, excreted 71 times more biotin (535Ilg/ml) than the wild type strain.
The TVA-resistant mutant, A16, excreted 116 times more total biotin
(400 Ilg/ml). Strain AB12 accumulated 9'5Ilg/ml of biotin under the optimum
reaction conditions using 1 mg/ml of pimelic acid. Strain AB12 was considered
to have enhanced biotin synthesizing enzyme activity, resulting in survival even
in the presence of ACM, whereas the repression of the enzymes by biotin was
still not diminished.
Recently, the production of biotin by bacteria which were transformed by
recombinant DNA has been reported. Hirono et al. (1986) and Ifuku et al.
(1986, 1987) have derived an a-dehydrobiotin-resistant mutant from E. coli.
They transformed E. coli with the biotin operon coding for the biotin
biosynthetic enzymes. The maximum production of biotin by the transformant
was 121lg/ml under the optimized conditions (Hirono et al., 1986). Osawa et
al. (1987, 1989) obtained an ACM- and TVA-resistant mutant of B. sphaericus
IFO 3525, followed by transforming of E. coli strain with the plasmid carrying
under appropriate culture conditions, the transformant excreted 161lg/ml of
biotin from DL-DTB in a 48-h cultivated medium (Osawa et al., 1989). They
have recently determined the nucleotide sequence of the bio B gene of B.
sphaericus and compared the amino acid sequence of its gene product, biotin
synthetase, with that of E. coli, suggesting conservation in the catalytic
mechanism for both enzymes (Osawa et aI., 1989).
Since the acetate produced in the culture medium was found to inhibit the
growth, Ifuku & Yanagi (1988) derived a tluoroacetate-resistant mutant from
the E. coli transformant, which had a level of acetate production 10 times
lower than the parent strain. When the mutant was cultivated in a glucose-fed
system with a controlled dissolved oxygen level in a jar fermentor, the growth
and the accumulation of biotin reached 55 mg (dry cell weight)/ml and
105 Ilg/ml, respectively.


Dethiobiotin Production by B. sphaericus IFO 3525 (Ogata, 1970)

Medium: 10 g peptone, 5 g Casamino acids (Difco), 100 g soybean meal, 20 g

glycerol, 1 g K 2HP0 4, 05 g KCI, 05 g MgS0 47H20, 10 mg FeS047H20,


Y. Izumi & H. Yamada

10 mg MnSOc6HzO, 200 Ilg thiamine HCI and 1 g pimelic acid in 1 liter of

tap-water, pH 72.
Inoculum and fermentation course: The cells of the bacterium, B.
sphaericus, from a slant culture are incubated in 30 ml of the medium in a
300-ml shaking flask for 4-5 days at 28C on a reciprocal shaker
(140 strokes/min).
6.2 Biotin Production by an ACM- and TVA-resistant Mutant ABU of B.
sphaericus IFO 3525 (Yamada et al., 1983)
Medium: 20g glycerol, 50g Proteose peptone (Difco), 5g Casamino acids
(Difco, vitamin-free), 1 g KHzP04, 05 g KCI, 05 g MgS047HzO,
10 mg FeS04'7HzO, 10 mg MnS044-6HzO and 20llg thiamineHCI in 1 liter
of tap-water, pH 70.
Inoculum and fermentation course: Cells of bacterium AB12 were inoculated into 3 ml of the medium in a test tube and incubated at 28C with
shaking. After 1 day of cultivation, pimelic acid was added (1 mg/ml) to the
medium and cultivation was continued at 37C (optimum temperature for
biotin biosynthesis from DTB) for 3 more days. The addition of the precursor
after 1 day of cultivation is important for biotin production in order to avoid
repression by the produced biotin itself of the biotin biosynthetic enzymes (see
Section 4.2).
Biotin and biotin-vitamers accumulated in the culture media are purified by
active carbon treatment and ion-exchange column chromatography (Ogata,
1970). After centrifugation of the culture medium, the pH of the supernatant is
adjusted to about 2-3 with concentrated HCI-15 g of active carbon is added
to 1 liter of the culture medium and stirred mechanically for 4 h at room
temperature. The active carbon is collected on a Buchner funnel and washed
with about 200 ml of water. The active carbon cake is suspended in about
200 ml of 50% ethanol-28% ammonia water (18: 1, v/v) mixture and stirred
for 4 h at room temperature. The active carbon is filtered off onto a Buchner
funnel, and the active carbon cake is washed with ethanol-ammonia mixture
three to four times. The filtrates are concentrated to dryness in vacuo at room
temperature. The concentrate obtained is dissolved in about 40 ml of 90%
ethanol, and insoluble materials are removed by centrifugation; the clear
supernatant solution is again concentrated to a volume of about 5 ml in vacuo
at room temperature. This concentrated solution is subjected to ion-exchange
chromatography. A few milliliters of the solution of biotin-vitamers (containing about 20-500 Ilg of total biotin, pH 70) is quantitatively taken up onto a
Dowex 1 X2-formate column (06 cmz x 19 cm). The column is washed with
100 ml of deionized water, which removes the unadsorbed biotin-vitamers such

Microbial Production of Biotin


as DAPA, then the biotin-vitamers (e.g., biotin, biotin sulfoxide and DTB)
are eluted with 0012 M formic acid, and lO-ml fractions are collected. Biotin
concentrations are quantitatively determined by microbiological assays with
Saccharomyces cerevisiae and Lactobacillus plantarum.


There are three methods for the determination of biotin and its vitamers:
microbiological, chemical and enzymatic assay methods. Table 3 summarizes
the characteristics of the three methods. The microbiological assay has been
most commonly used among the three methods. This assay uses the microorganisms Lactobacillus plantarum ATCC 8014 (Scheiner, 1985),
Saccharomyces cerevisiae ATCC 7754 (Gyorgy, 1967) and Bacillus subtilis
AKU 236 (Iwahara et al., 1966c). In addition, a biotin-requiring mutant of E.
coli, C162, (bio B-, His-) has also been used as a test of the growthpromoting activities of biotin-vitamers (Salib et al., 1979; Osakai et al., 1986b).
These micro-organisms have biotin-vitamer specifi.cities for growth, which
enables us to assay them differentially. Table 4 shows the microbiological
activities of various biotin-vitamers when assayed with L. plantarum, S.
cerevisiae and B. subtilis. The microbiological assay also has the advantage of
high sensitivity. However, one of the main disadvantages of the method is that
it requires the time-consuming cultivation of assay organisms and the laborious
preparation of washed cell suspensions. Tanaka et al. (1987a) have developed
improved assay methods for biotin using lyophilized cells of L. plantarum and
E. coli C162 and glycerol-suspended cells of S. cerevisiae. Figure 5 shows the
standard curves by the paper disk plate method and turbidimetric method
using such cells. Those lyophilized or glycerol-suspended cells, which are
preserved at -20C, can be used for the assay for more than 1 year or half a
year, respectively.


Biotin Deficiencies in Animals

Feeding a diet either low in biotin or, more frequently, using a diet amended with
with raw egg white, produced biotin deficiencies in the rat, chicken, poultry, pig,
monkey, man and other species (Gyorgy & Larger, 1968). Chicken and poultry,
however, require an outside source of biotin, even when egg white is replaced
by other proteins. The effectiveness of egg white in producing biotin deficiency
is due to the presence of the glycoprotein avidin, which forms a complex with
biotin that renders the biotin unavailable to the host. Avidin also forms tight
complexes when biotin is the prosthetic group of an enzyme, resulting in the
loss of enzyme activity. The biotin-binding capacity is present not only in the





1. Ureido ring-containing
biotin-related compounds
with no microbiological
activity can be assayed.
2. Fast assay
3. Sample should be dried
before assay
1. Avidin-combinable
(ureido ring compounds)
can be assayed.
2. Fast assay

Avidin is reacted with the dye, 4hydroxyazobenzene-2' -carboxylic

acid (HABA) to form the avidinHABA complex (A. max = 500 om),
followed by the addition of biotin.
Since the HABA in the complex is
substituted by biotin, the added
biotin is assayed from the
colorimetrical determination of the
remaining avidin-HABA complex.

1. High sensitivity
2. High specificity
to various biotinvitamers
3. Even crude samples can
be assayed.
4. Time-consuming (6-27 h)


Coloring reaction of the ureido ring

with p-dimethylcinnamaldehyde
(DACA) under anhydrous
conditions (reddish orange to pink
A. max = 533 nm)

Growth of L. plantarum,
S. cerevisiae, B. subtilis
and E. coli


Table 3
Assay range

(05Ilg biotin/ml
gives M = 0069)


Turbidimetric method:

Paper disc (agar

diftUsion) method:
01-50 llg/ml
(as (+ )-biotin)

Assay Methods of Biotin and Its Related Compounds

Green (1970)

McCormick &

Tanaka et al.


PC method


Avidin3HC 4C)biotin

10-200 nmol/ml

5-2000 pg/ml

1. High sensitivity
2. Fast Assay (1 h)
3. The yeast cells should
be prepared just before

The enzymatic binding of biotin in situ

to the pyruvate carboxylase (PC)
apoprotein of biotin-deficient bakers'
yeast and the subsequent estimation
of the PC activity by a 14C02 -fixation

>01 ng Biotin
(Variable according
to the used amounts
of avidin:
available avidin binds
1-13llg biotin)

1. High sensitivity
2. Fast assay (2 h)
3. BCS is stable

1. Avidin-combinable
compounds (ureido ring
compounds) can
be assayed.
2. Fast assay

The H 20 2 formed through the coupled

reactions of biotinyl-CoA synthetase
(BCS) and acyl-CoA oxidase (AOD)
is assayed by the peroxidase (POD)
reaction, which gives a fluorescent

Given amount of avidin, of which

biotin combinability (a) is known, is
reacted with a sample containing
biotin(x), followed by the further
addition of excess 3H-biotin. x is
obtained from the equation x = a b, where b is the combined 3H_


Tanaka et al.

Hood (1979)
Dakshinamurti &
Allan (1979)

Y. Izumi & H. Yamada


Table 4
Microbiological Activities of Various Biotin-Vitamers

Activitya toward

L. plantarum

Intermediates of
biotin biosynthesis
(+ )-biotin
( + )-dethiobiotin
( )-dethiobiotin
7,8-diaminopelargonic acid
pimelic acid
Intermediates of
biotin biodegradation
( + )-bisnorbiotin
( + )-bisnorbiotin sulfoxide
( + )-bisnordethiobotin
Naturally occurring
biotin-related compounds
( + )-biotin amide
dethiobiotin amide
( + )-biotin D-sulfoxide
( + )-biotin L-sulfoxide
Chemically synthesized
biotin-related compounds
( )-oxibiotin
( + )-selenobiotin
( )-carbobiotin
( + )-biotin methyl ester



B. subtilis

















Relative activity of equimolar concentration of each compound.

egg whites, but also in the egg yolks of avian species and the turtle (Moss &
Lane, 1971). Since avidin is heat-labile, prolonged heating of egg white
denatures the avidin and destroys its biotin-binding capacity. Another biotinbinding protein similar to avidin, streptavidin, is produced by a Streptomyces
strain (Chaiet & Wolf, 1964).
Biotin deficiency produces dermatitis and perosis in chicken and poultry
(Gyorgy, 1968); alopecia, seborrheic skin changes, spasticity of the hind legs
and cracks in the feet of pigs. The activities of the biotin-dependent enzymes
are also decreased (Whitehead, 1981). These enzymes are involved in
carboxylation, transcarboxylation, and decarboxylation reactions, and function
in the vitally important metabolic processes of glucose and fat synthesis (Moss
& Lane, 1971; Wood & Barden, 1977; Whitehead, 1981). Among the most
important enzymes are acetyl-CoA carboxylase, pyruvate carboxylase and
propionyl-CoA carboxylase.
It was generally believed that the combination of biotin in the feed

Microbial Production of Biotin










1X10 3
Biotin (ng/ml X 71l1/disk)


5X10 3

20X10 3





Biotin (ng/ml x 100 III/2m!)


Biotin (ng/ml X 100 III/2m!)


Fig. 5. Standard curves of microbiological assays of biotin. (A) Paper disc plate (agar
diffusion) method using lyophilized cells of L. plantarum (x) and E. coli (e) and
glycerol-suspended cell of S. cerevisiae (0). (B) Turbidimetric method using lyophilized
cells of L. plantarum. (C) Turbidimetric method using glycerol-suspended cells of S.
cerevisiae (0) and lyophilized cells of E. coli (e). From Tanaka et al. (1987a).

ingredients plus the biotin produced in the intestine by bacteria supplied

sufficient biotin to meet the poultry's requirement. However, since 1966, a
number of reports on biotin deficiency in commercial flocks have appeared
(Scheiner & DeRitter, 1975). Apparent biotin deficiencies in swine under
commercial conditions were also reported (Scheiner & DeRitter, 1975).
In humans, infant seborrheic dermatitis and the related Leiner's disease are


Y. Izumi & H. Yamada

biotin-responsive (Svejcar & Homolka, 1950). Biotin deficiency has been

reported in individuals during prolonged total parenteral nutrition (Bozian et
al., 1981; Mock et al., 1981). Biotin administration has successfully controlled
multiple carboxylase deficiency even in unborn infants (Baumgartener et al.,
1981; Munnich et al., 1981; Thoene et al., 1981; Roth et al., 1982). It has been
suggested that diseases related to biotin metabolism may be more common
than previously thought (Tanaka, 1981).
9.2 GrowthPromoting Activity
Biotin deficiency also causes a marked decrease in the activities of several
glycolytic enzymes in the liver, e.g. glucokinase, phosphofructokinase and
pyruvate kinase. The activities of these enzymes increased rapidly after the
administration of biotin (Dakshinamurti & Cheah-Tan, 1968, 1970; Dakshinamurti & Hong, 1970; Dakshinamurti et al., 1970), while other glycolytic
enzymes such as hexose phosphate isomerase were not affected by biotin
administration. Boeckx and Dakshinamurti (1974) showed that biotin administration to biotin-deficient rats resulted in increased stimulation, by more
than twofold, of amino acid incorporation into protein, both in vivo and in
vitro in rat liver, pancreas, intestinal mucosa and skin. They also found that
the synthesis of some proteins such as serum albumin, a major product of the
liver protein-synthetic machinery, was stimulated more than twofold, but
others were not stimulated at all. The effect of biotin on protein synthesis was
preceded by stimulation in the incorporation of orotic acid into nuclear and
ribosomal RNA.
Vesely (1982) and Vesely et al. (1984) found that biotin and its analogs at
01-1 IJM enhanced soluble guanylate cyclase activity two- to threefold in rat
liver, kidney, colon, cerebellum, and heart. Since cyclic GMP, a product of
guanylate cyclase reaction, is known to increase the growth of fibroblasts and
thymocytes and also to increase RNA and protein synthesis, these results
suggest that the growth-promoting effect of biotin might be mediated by cyclic
GMP. Spence & Koudelka (1984) also found that addition of biotin in the
presence of insulin elicited an increase in the intracellular content of cyclic
GMP, followed by an increase in glucokinase in cultured rat hepatocytes.

10.1 Method of Goldberg et 01.

The industrial synthesis of biotin which is presently carried out is based on a
method developed by Hoffmann-La Roche, Inc. (Goldberg & Sternbach, 1949;
Gyorgy & Larger, 1968). As illustrated in Fig. 6, the synthesis is characterized
by the use of a meso-diaminosuccinic acid derivative as a starting material,
which contains two groups in the same spatial arrangement as the two amino


Microbial Production of Biotin



7R (e)


~R (d)
CH-CH - - CH-CH - - HC--CH __ Ht-CH -


~r (a) ~HR ~HR (b) R~

taaH taoH

taaH tOOH










~R (e)


Ac=CH 3Ca







of (+)-IX



Fig. 6. Flow sheet of one chemical process for the industrial preparation of ( + )-biotin.
(a) Benzylamine; (b) phosgene; (c) acetic anhydride; (d) Zn, a mixture of acetic acid
and acetic anhydride; (e) H 2S, HCI; (f) C2H50(CH2hMgBr; (g) H 2, Raney nickel; (h)
HBr, acetic acid; (i) silver o-camphorsulfonate; (j) isopropanol; (k) sodium diethylmalonate; (I) HBr. From Goldberg & Sternbach (1949).

groups present (in substituted form) in the biotin molecule, i.e. the mesoconfiguration in diaminosuccinic acid derivatives corresponding to the cisstructure of the two amino groups in a ring compound such as biotin.
Moreover, since resolution is carried out at the intermediate stage (XII-XIII)
it permits the direct production of the optically and biologically active biotin,
( + )-biotin.
10.2 Synthesis from Sugars
Efficient stereospecific total syntheses of (+ )-biotin from the sugars, 0mannose, o-glucose and o-glucosamine, were achieved by Ohrui & Emoto
(1975), Ogawa et al. (1977) and Ohrui et al. (1978), respectively. The synthesis
from o-mannose is shown in Fig. 7. o-Mannose is first converted to the
aldehyde (II) through isopropyridenylation, benzoylation, selective isopropydenylation, and periodate oxidation. The Wittig reaction and hydrogenation of

Y. Izumi & H. Yamada


<" s~.





















Fig. 7. Stereospecific synthesis of ( + )-biotin from D-mannose. From Ohrui & Emoto

the aldehyde yield compound (IV) which has a side chain-like biotin. The
treatment of (IV) with NaOCH3 in methanol, followed by the reduction of the
resulting aldehyde (V) with NaBH4 produces (VII). Treatment of (VII) with
NaS affords a tetrahydrothiophene derivative (VIII) which is converted via
(XI) to (X). Treatment of (X) with NaN3 gives a diazido compound (XI).
Catalytic reduction of the azido groups of (XI) in a mixture of methanol and
acetic anhydride gives a diacetoamido derivative (XII). Treatment of (XII)
with Ba(OHb followed by the treatment with phosgene affords ( + )-biotin.
Thus, (+ )-biotin is synthesized from D-mannose in a good yield, because a
five-membered ring consisting of isopropyridene, which protects the hydroxyl
groups of the sugar, is used to fix the molecular conformation during the
intramolecular substitution reactions.


The most practical uses of biotin are as a pharmaceutical and as a supplement

of culture media for amino acid production, where the biotin-requiring
bacteria such as Corynebacterium glutamicum and Brevibacterium ftavum for
glutamic acid and lysine production are used (Aida, 1986). Recently, the
vitamin has attracted increasing interest as a food and feedstuff supplement
(Pearson et al., 1976; Whitehead, 1981; Kornegay, 1985) as can also be seen
from the studies of its growth-promoting effect as described in Section 9.2.
There may be a marked increase in the demand and supply of the vitamin as a
feedstuff supplement if its price becomes lower. At present, the industrial
preparation of biotin is carried out through a chemical process. The present
price of biotin as a biochemical reagent is US $17000/10 g (price list of Sigma,
1988). Therefore, the price of bulk supplies can be supposed to be much lower
than the reagent price. In order to economically compete with the present
chemical process, the present production levels via fermentation processes
should be greatly improved (the maximum level so far reported is 105 mg/liter

Microbial Production of Biotin


as described in Section 5). If a fermentation process uses a biotin precursor

such as pimelic acid or azelaic acid, the cost of such compounds should also be
taken into consideration.
Aida, K. (1986). An overview of the microbial production of amino acids. In Progress
in Industrial Microbiology, Vol. 24 (Biotechnology of Amino Acid Production), ed.
K. Aida, I. Chibata, K. Nakayama, K. Takinami & H. Yamada. KodanshaElsevier, Tokyo, pp. xxi-xxv.
Baumgartener, R., Suormala, T., Wick, H., Bachmann, C. & Jaggi, K. H. (1981).
Biotin dependency causing multiple carboxylase deficiency in vivo. Pediatr. Res.,
15, 1189.
Boas, M. A. (1927). The effect of desiccation upon the nutritive properties of
egg-white. Biochem. J., 21, 712-24.
Boeckx, R. L. & Dakshinamurti, K. (1974). Biotin mediated protein biosynthesis.
Biochem. J., 140, 549-56.
Bonnemere, c., Hamilton, J. A., Steinrauf, L. K. & Knappe, J. (1965). Structure of
the bis-p-bromoanilide of carbon dioxide biotin. Biochemistry, 4, 240-45.
Bozian, R. c., Moussavian, N. & Piepmeyer, J. L. (1981). Biotin deficiency during
prolonged home total parenteral nutrition (TPN). Clin. Res., 29, 622A
Chaiet, L. & Wolf, F. J. (1964). The properties of streptavidin, a biotin-binding protein
produced by Streptomycetes. Archs Biochem. Biophys., 106, 1-5.
Dakshinamurti, K. & Allan, L. (1979). Isotopic dilution assay for biotin: Use of
[3H]biotin. In Methods in Enzymology, Vol. 62, ed. D. B. McCormick & L. D.
Wright. Academic Press, New York, pp. 284-9.
Dakshinamurti, K. & Cheah-Tan, C. (1968). Liver glucokinase of the biotin deficient
rat. Can. J. Biochem., 46, 75-80.
Dakshinamurti, K. & Cheah-Tan, C. (1970). Biotin-mediated synthesis of hepatic
glucokinase in the rat. ArdIS Biochem. Biophys., 127, 17-21.
Dakshinamurti, K. & Hong, H. C. (1970). Regulation of key hepatic glycolytic
enzymes. Enzymol. BioI. Clin., 11, 423-8.
Dakshinamurti, K., Tarrago-Litvak, L. & Hong, H. C. (1970). Biotin and glucose
metabolism. Can. J. Biochem., 48, 493-500.
Dittmer, K. & du Vigneaud, V. (1944). Antibiotins. Science, 100, 129-31.
Eakin, R. E. & Eakin, E. E. (1942). A biosynthesis of biotin. Science, 96, 187-8.
Eisenberg, M. A. (1973). Biotin: biogenesis, transport and their regulation. Adv.
Enzymol., 38,317-72.
Eisenberg, M. A & Krell, K. (1969). Dethiobiotin synthesis from 7,8diaminopelargonic acid in cell-free extracts of a biotin auxotroph of Escherichia coli
K-12. J. Bioi. Chem., 244,5503-9.
Eisenberg, M. A & Star, C. (1968). Synthesis of 7-keto-8-aminopelargonic acid, a
biotin vitamer, in cell-free extracts of Escherichia coli biotin auxotrophs. J.
Bacteriol., 96, 1291-7.
Eisenberg, M. A, Mea, B., Prakash, O. & Eisenberg, M. R. (1975). Properties of
a-dehydrobiotin-resistant mutants of Escherichia coli K-12. J. Bacteriol., 122,
Eisenberg, M. A., Prakash, o. & Hsiung, S. C. (1982). Purification and properties of
the biotin repressor-a bifunctional protein. J. Bioi. Chem., 257, 15167-73.
Fall, R. R. & Vagelos, P. R. (1972). Acetyl coenzyme A carboxylase. Molecular forms
and subunit composition of biotin carboxyl carrier protein. J. Bioi. Chem., 247,


Y. Izumi & H. Yamada

Goldberg, M. W. & Sternbach, L. H. (1949). US Patents 2,489,232-2,489,238.

Green, N. M. (1970). Spectrophometric determination of avidin and biotin. In Methods
in Enzymology, Vol. 18A, ed. D. B. McCormick & L. D. Wright. Academic Press,
New York, pp. 418-24.
Guillerm, G., Frappier, F., Gaudry, M. & Marquet, A. (1977). On the mechanism of
conversion of dethiobiotin to biotin in Escherichia coli. Biochimie, 59, 119-26.
Gyorgy, P. (1967). Biotin. In The Vitamins, 2nd edn, Vol. 7, ed. W. H. Sebrell & R. S.
Harris. Academic Press, New York, pp. 303-10.
Gyorgy, P. & Larger, B. W. (1968). Biotin. In The Vitamins, 2nd edn, Vol. 2, ed. W.
H. Sebrell & R. S. Harris. Academic Press, New York, pp. 262-357.
Gyorgy, P., Kuhn, R. & Lederer, E. (1939). Attempts to isolate the factor (vitamin H)
curative of egg white injury. J. Bioi. Chem., 131,745-59.
Harris, S. A., Wolf, D. E., Mozingo, R. & Folkers, K. (1943). Synthetic biotin.
Science, fYI, 447-8.
Harris, S. A., Wolf, D. E., Mozingo, R., Anderson, R. C., Arth, G. E., Easton, N.
R., Heyl, D., Wilson, A. N. & Folkers, K. (1944a). Biotin. II. Synthesis of biotin.
J. Am. Chem. Soc., 66, 1756-7.
Harris, S. A., Easton, N. R., Heyl, D., Wilson, A. N. & Folkers, K. (1944b). Biotin.
IV. Synthesis of 4-benzamido-3-ketotetrahydrothiophene. J. Am. Chem. Soc., 66,
Harris, S. A., Wolf, D. E., Mozingo, R., Arth, G. E., Anderson, R. C., Easton, N. R.
& Folkers, K. (1945). Biotin. V. Synthesis of dl-biotin, dl-allobiotin and dl-epiallobiotin. J. Am. Chem. Soc., 67,2096-100.
Hirono, Y., Kojima, T. & Kimura, H. (1986). Japan Patent 14,901.
Hood, R. L. (1979). Isotopic dilution assay for biotin: Use of [14C]biotin. In Methods in
Enzymology, Vol. 62, ed. D. B. McCormick & L. D. Wright. Academic Press, New
York, pp. 279-83.
Ifuku, O. & Yanagi, M. (1988). Cloning of biotin operon of Escherichia coli and
breeding of biotin producers. Hakka to Kogyo (in Japanese), 46, 102-11.
Ifuku, 0., Haze, S., Kishimoto, J., Sakamoto, T. & Yanagi, M. (1986). Japan Patent
Ifuku, 0., Haze, S., Sakamoto, T. & Kishimoto, J. (1987). Japan Patent 155,018.
Iwahara, S., Emoto, Y., Tochikura, T. & Ogata, K. (19600). Studies on biosynthesis of
biotin by micro-organisms. Part III. The isolation of crystalline desthiobiotin from
culture filtrate of Bacillus sphaericus. Agric. Bioi. Chem., 30,64-7.
Iwahara, S., Takasawa, S., Tochikura, T. & Ogata, K. (1966b). Studies on biosynthesis
of biotin by micro-organisms. Part IV. Conversion of desthiobiotin to biotin by
various kinds of micro-organisms. Agric. Bioi. Chem., 30, 385-92.
Iwahara, S., Takasawa, S., Tochikura, T. & Ogata, K. (1966c). Studies on biosynthesis
of biotin by micro-organisms. Part V. degradation of desthiobiotin by molds. Agric.
Bioi. Chem., 30, 1069-75.
Izumi, Y. (1984). Studies on the metabolism of biotin and its regulation in microorganisms. Nippon Nogei Kagaku Kaishi, 58,891-900.
Izumi, Y. & Ogata, K. (1977). Some aspects of the microbial production of biotin.
Adv. Appl. Microbio!., 22, 145-76.
Izumi, Y., Sugisaki, K. & Ogata, K. (1973). Incorporation of the sulfur of Lp5SJmethionine into the biotin molecule by intact cells of Rhodoturula glutinis.
Biochim. biophys. Acta, 304,887-90.
Izumi, Y., Fukuda, H., Tani, Y. & Ogata, K. (1978). The mode of action of
5-(2-thienyl)-valeric acid on biotin biosynthesis. Agric. Bioi. Chem., 42, 579-84.
Izumi, Y., Tani, Y. & Yanada, H. (1980). Biotin biosynthesis in micro-organisms.
Bull. Inst. Chem. Res. Kyoto Univ., 58,434-47.
Izumi, Y., Kano, Y., Inagaki, K., Kawase, N., Tani, Y. & Yamada, H. (1981).
Characterization of biotin biosynthetic enzymes of Bacillus sphaericus, a dethiobiotin producing bacterium. Agric. Bioi. Chem., 45, 1983-9.

Microbial Production of Biotin


Kogl, F. (1937). Wirkstoffprinzip und Pflanzenwachstum. Naturwissenschaften, 25,

Kogl, F. & Tonnis, B. (1936). Uber des Bios-Problem. Darstellung von krystallisierten
Biotin aus Eigelb. Z. Physiol. Chem., 242, 43-73.
Knappe, J., Ringelmann, E. & Lynen, F. (1961). Zur biochemischen Funktion des
Biotins. III. Die dchemische Konstitution kdes enzymatisch gebildeten Carboxybiotins. Biochem. Z., 335, 168-76.
Kornegay, E. T. (1985). Biotin in swine nutrition. Ann. N. Y. A cad. Sci., 447, 112-21.
Kosow, D. P., Huang, S. C. & Lane, M. D. (1962). Propionyl coenzyme A
holocarboxylase synthesis. I. Preparation and properties of the enzyme system. J.
BioI. Chem., 237, 3633-9.
Lamhonwah, A. M., Quan, F. & Gravel, R. A. (1987). Sequence homology around in
the biotin-binding site of human propionyl-CoA carboxylase and pyruvate carboxylase. Archs Biochem. Biophys., 254,631-6.
Lilly, V. G. & Leonian, L. H. (1944). Antibiotin effect of desthiobiotin. Science, 99,
Lynen, F., Knappe, J., Lorch, E., Juttig, G. & Ringelmann, E. (1959). Die
biochemischen Funktion des Biotin. Angew. Chem., 71,481-91.
Lynen, F., Knappe, J., Lorch, E., Juttig, G., Ringelmann, E. & Lachance, J. P.
(1961). Zur biochemischen Funktion des Biotins. II. Reinigung und Wirkungsweise
der ,8-Methyl-crotonyl-Carboxylase. Biochem. Z., 335, 123-67.
McCormick, D. B. & Roth, J. A. (1970). Colorimetric determination of biotin and
analogs. In Methods in Enzymology, Vol. 18A, ed. D. B. McCormick & L. D.
Wright. Academic Press, New York, pp. 383-5.
Melville, D. B., Moyer, A. W., Hofmann, K. & du Vigneaud, V. (1942). The structure
of biotin: The formation of thiophenevaleric acid from biotin. J. BioI. Chem., 146,
Mock, D. M., de Lorimer, A. A., Liebman, W. M., Sweetman, L. & Baker, H.
(1981). Biotin deficiency: An unusual complication of parenteral alimentation. New
Engl. J. Med., 304,820-23.
Moss, J. & Lane, D. (1971). The biotin-dependent enzymes. Adv. Enzymol., 35,
Mueller, J. H. (1937a). Pimelic acid as a growth accessory for a strain of the diphtheria
bacillus. Science, 85, 502-3.
Mueller, J. H. (1937b). Pimelic acid as a growth accessory for the diphtheria bacillus. J.
Bioi. Chem., 119,121-9.
Munnich, A., Saudubray, J. M., Carre, G., Coude, F. X., Ogier, H., Charpentier, C.
& Frezal, J. (1981). Defective biotin absorption in multiple carboxylase deficiency.
Lancet, 2, 263.
Ogata, K. (1970). Microbial synthesis of dethiobiotin and biotin. In Methods in
Enzymology, Vol. 17A, ed. D. B. McCormick & L. D. Wright. Academic Press,
New York, pp. 390-94.
Ogata, K., Tochikura, T., Iwahara, S., Takasawa, S., Ikushima, K., Nishimura, A. &
Kikuchi, M. (1965a). Studies on biosynthesis of biotin by micro-organisms. Part I.
Accumulation of biotin-vitamers by various micro-organisms. Agric. Bioi. Chem.,
Ogata, K., Tochikura, T., Iwahara, S., Ikushima, K., Takasawa, S., Kikuchi, M. &
Nishimura, A. (1965b). Studies on biosynthesis of biotin by micro-organisms. Part
II. Identification of biotin-vitamers accumulated by various micro-organisms. Agric.
Bioi. Chem., 29, 895-901.
Ogawa, T., Kawano, T. & Matsui, M. (1977). A biomimetic synthesis of (+ )-biotin
from D-glucose. Carbohydrate Res., 57, c31-5.
Ogino, S., Fujimoto, S. & Aoki, Y. (1974a). Cooxidation of dl-cis-tetrahydro-2-oxo-4n-pentyl-thieno-(3,4-d)-imidazoline (dl-TOPTI) by soil isolates of the genus
Corynebacterium. Agric. Bioi. Chem., 38, 275-8.


Y. Izumi & H. Yamada

Ogino, S., Fujimoto, S. & Aoki, Y. (1974b). Production of biotin by microbial

transformation of dl-cis-tetrahydro-2-oxo-4-n-pentyl-thieno-(3 ,4-d)-imidazoline (dlTOPTI). Agric. Bioi. Chern., 38,707-12.
Ohrui, H. & Emoto, S. (1975). Stereospecific synthesis of ( + )-biotin. Tetrahedr. Lett.,
Ohrui, H., Sueda, N. & Emoto, S. (1978). An alternative synthesis of (+ )-biotin from
carbohydrate, a formal synthesis of ( + )-biotin from D-glucosamine. Agric. Bioi.
Chem., 42,865-8.
Okumura, S., Tsugawa, R., Tsunoda T. & Motozaki, S. (1962a). Studies on the
L-glutamic acid fermentation. Part III. On the biotin synthetic path in bacteria
producing L-glutamic acid. Nippon Nogei Kagaku Kaishi, 36,599-605.
Okumura, S., Tsugawa, R., Tsunoda, T. & Motozaki, S. (1%2b). Studies on the
glutamic acid fermentation. Part IV. The universality of biotin biosynthetic
pathway. Nippon Nogei Kagaku Kaishi, 36,605-13.
Osakai, M., Izumi, Y., Nakamura, K. & Yamada, H. (1986a). Bacterial degradation of
biotin and dethiobiotin: Isolation and identification of {:J-hydroxy and keto
metabolites. Agric. Bioi. Chem., SO,311-6.
Osakai, M., Tamura, K., Izumi, Y. & Yamada, H. (1986b). Reinvestigation of
compound X, a suspected biotin intermediate: Identification of N-formyl derivatives of biotin and dethiobiotin. J. Nutr. Sci. Vitaminol., 32, 279-86.
Osawa, I., Fuji, N. & Kamogawa, K. (1987). Japan Patent 275,684.
Osawa, I., Speck, D., Kisou, T., Hayakawa, K., Zinsius, M., Gloeckler, R., Lemoine,
Y. & Kamogawa, K. (1989). Cloning and expression in Escherichia coli and Bacillus
subtilis of the biotin synthetase gene from Bacillus sphaericus. Gene, in press.
Pai, C. H. (1969). Biosynthesis of desthiobiotin in cell-free extracts of Escherichia coli.
J. Bacteriol., 99, 696-701.
Pai, C. H. (1971). Biosynthesis of biotin: Synthesis of 7,8-diaminopelargonic acid in
cell-free extracts of Escherichia coli. J. Bacteriol., 105, 793-800.
Pai, C. H. (1975). Biochemical and genetic characterization of dehydrobiotin resistant
mutants of Escherichia coli. Molec. Gen. Genet., 134,345-57.
Parry, R. J. & Kunitani, M. G. (1976). Biotin biosynthesis. 1. The incorporation of
specifically tritiated dethiobiotin into biotin. J. Am. Chem. Soc., 98, 4024-26.
Parry, R. J. & Naidu, M. V. (1980). Biotin biosynthesis. Incorporation of 5(RS)-3Hdethiobiotin into biotin. Tetrahedr. Lett., 21, 4783-6.
Pearson, J. A., Johnson, A. R., Hood, R. L. & Fogerty, A. C. (1976). Fatty liver and
Kidney syndrome in chicks. I. Effect of biotin in diet. Aust. J. BioI. Sci., 29,
Polakis, S. E., Guchhait, R. B., Zwergel, E. E. & Lane, M. D. (1974). Acetyl
coenzyme A carboxylase system of Escherichia coli: Studies on the mechanism of
the biotin carboxylase- and carboxyl transferase-catalyzed reactions. J. Bioi. Chem.,
Pontecorvo, G. (1953). The genetics of Aspergillus nidulans. Adv. Genet., 5,208-38.
Prakash, O. & Eisenberg, M. A. (1979). BiotinyI5'-adenylate: Corepressor role in the
regulation of the biotin genes of Escherichia coli K-12. Proc. Natn. Acad. Sci.
U.S.A., 76, 5593-602.
Roth, K. S., Yang, W., Allan, L., Saunders, M., Gravel, R. A. & Dakshinamurti, K.
(1982). Prenatal administration of biotin in biotin responsive multiple carboxylase
deficiency. Pediatr. Res., 16, 126.
Ryan, F. J. (1956). Contributions from microbiology to the concept of the gene and
allelism. Japan J. Genet., 31,265-73.
Salib, A. G., Frappier, F., Guillerm, G. & Marquet, A. (1979). On the mechanism of
conversion of dethiobiotin to biotin in Escherichia coli. III. Isolation of an
intermediate in the biosynthesis of biotin from dethiobiotin, Biochem. Biophys.
Res. Commun., 88,312-9.

Microbial Production of Biotin


Scheiner, J. (1985). Biotin. In Methods of Vitamin Assay, 4th edn, ed. J. Augustin, B.
P. Klein, D. A. Becker & P. B. Venugopal. John Wiley & Sons, New York, pp.
Scheiner, J. & DeRitter, E. (1975). Biotin content of feedstuffs. J. Agric. Food Chem.,
23, 1157-62.
Spence, J. T. & Koudelka, A. P. (1984). Effects of biotin upon the intracellular level of
cGMP and the activity of glucokinase in cultured rat hepatocytes. J. Bioi. Chem.,
Stallings, W. & Titta, G. T. de (1985). Crystallographic investigations of biotin and
carboxybiotin derivatives. Ann. N. Y. Acad. Sci., 447, 152-68.
Stokes, J. L. & Gunness, J. (1945). Microbial activity of synthetic biotin, its optical
isomers, and related compounds. J. BioI. Chem., 157, 121-6.
Svejcar, J. & Homolka, J. (1950). Experimental experiences with biotin in babies. Ann.
Paediat., 174, 175-93.
Szybalski, E. H. & Szybalski, W. (1982). A physical map of the Escherichia coli bio
operon. Gene, 19,93-103.
Takai, T., Wada, K. & Tanabe, T. (1987). Primary structure of the biotin-binding site
of chicken liver acetyl-CoA carboxylase. FEBS Lett., 2U, 98-102.
Takai, T., Yokoyama, C., Wada, K. & Tanabe, T. (1988). Primary structure of chicken
liver acetyl-CoA carboxylase deduced from cDNA sequence. J. Bioi. Chem., 263,
Tanaka, K. (1981). New light on biotin deficiency. N. Engl. J. Med., 304,839-40.
Tanaka, M., Yamamoto, H., Izumi, Y. & Yamada, H. (1986). Purification and
properties of biotinyl-CoA synthetase from Mycoplana sp. No. 166. Archs Biochem.
Biophys., 251,479-86.
Tanaka, M., Izumi, Y. & Yamada, H. (1987a). Biotin assay using lyophilized and
glycerol-suspended cultures. J. Microbiol. Meth., 6,237-46.
Tanaka, M., Izumi, Y. & Yamada, H. (1987b). Enzymatic assay for biotin using
biotinyl-CoA synthetase. Agric. BioI. Chem., 51, 2585-6.
Tanaka, M., Izumi, Y. & Yamada, H. (1988). Microbial metabolism and production of
biotin. Vitamin (Japan), 62,305-15.
Tatum, E. L. (1945). Desthiobiotin in the biosynthesis of biotin. J. Bioi. Chem., 160,
Thoene, J., Baker, H., Yoshino, M. & Sweetmen, L. (1981). Biotin-responsive
carboxylase deficiency associated with subnormal plasma and urinary biotin. N.
Engl. J. Med., 304,817-20.
Trainor, D., Parry, R. J. & Gitterman, A. (1980). Biotin biosynthesis. 2. Stereochemistry of sulfur introduction at C-4 of dethiobiotin. J. Am. Chem. Soc., 102, 1467-8.
Traub, W. (1956). Crystal structure of biotin. Nature, Lond., 178,649-50.
Trotter, J. & Hamilton, J. A. (1966). The absolute configuration of biotin.
Biochemistry, 5,713-4.
Vesely, D. L. (1982). Biotin enhances guanylate cyclase activity. Science, 216,1329-30.
Vesely, D. L., Wormser, H. C. & Abramson, H. N. (1984). Biotin analogs activate
guanylate cyclase. Molec. Cell. Biochem., 60, 109-14.
Vigneaud, V. du, Dittmer, K., Hauge, E. & Long, B. (1942). The growth-stimulating
effect of biotin for the diphtheria bacillus in the absence of pimelic acid. Science, 96,
Wakil, S. J. & Gibson, D. M. (1960). Studies on the mechanism of fatty acid synthesis.
VIII. The participation of protein-bound biotin in the biosynthesis of fatty acids.
Biochim. Biophys. Acta, 41, 122-9.
Wakil, S. J., Titchener, E. B. & Gibson, D. M. (1958). Evidence for the participation
of biotin in the enzymic synthesis of fatty acids. Biochim. Biophys. Acta, 29, 225-6.
Watanabe, K., Yano, S. & Yamada, Y. (1982). The selection of cultured plant cell lines
producing high levels of biotin. Phytochemistry, 21,513-6.


Y. Izumi & H. Yamada

Whitehead, C. C. (1981). The assessment of biotin status in man and animals. Proc.
Nutr. Soc., 40, 165-72.
Wolf, D. E., Valiant, J., Peck, R. L. & Folkers, K. (1952). Synthesis of biocytin. J.
Arn. Chern. Soc., 74, 2002-3.
Wood, H. G. & Barden, R. E. (1977). Biotin enzymes. Ann. Rev. Biochern., 46,
Yamada, H., Osakai, M., Tani, Y. & Izumi, Y. (1983). Biotin overproduction by biotin
analog-resistant mutants of Bacillus sphaericus. Agric. Bioi. Chern., 47, 1011-6.
Yamada, H., Osakai, M. & Izumi, Y. (1984). Formation of biotinyl-CoA synthetase,
the first enzyme involved in microbial biotin degradation. Agric. BioI. Chern., 48,

Chapter 15


Farmitalia Carlo Erba, Via dei Gracchi 35, 20146, Milano, Italy


From the chemical point of view, the term vitamin B12 is synonymous with
cyanocobalamin, which is by far the most important component of the large
family of the cobalt corrinoids. In this review, however, as in most publications, the word vitamin B12 is given a very broad meaning so as to include all
the cobalt corrinoids of the cobalamin group. Other cobalt corrinoids which
can be considered analogous or precursors of cobalamins are also included.
Discovered in 1948 and identified as the anti-pernicious anemia factor,
vitamin B12 has been thoroughly investigated from several points of view,
including biochemical significance, biosynthesis in micro-organisms, production
in commercial amounts and applications. The B12 biosynthetic pathway in
micro-organisms is now largely known and good production processes are
available and will be described in this review.
Due to its implications in extremely important biochemical reactions,
vitamin B12 originated wide interest in large and useful applications in human
and animal health care. Many analogues and derivatives were obtained and
studied to find new activities with particular emphasis towards B12 molecules
endowed with anti-cancer activity. Unfortunately, none of these substances
were of particular interest and the application of vitamin B12 remained limited
to anti-pernicious anemia activity, polyvitaminic specialties and as a feed
supplement in husbandry. The interest in the development of vitamin B12
reached a peak in the 1960s and then decreased slowly.
At present, a well consolidated but steady market and good production
processes are in the hands of a very limited number of companies. Like other
vitamins, B12 found its niche and from the economic point of view it can now
be considered as a 'mature product'.
Vitamin B 12 , a natural product endowed with important biological properties,
was isolated in 1948, almost contemporaneously but independently and from


C. Spalla et aI.

different sources, by three industrial research groups, i.e. by Glaxo Laboratories (Smith, 1948a,b) in the UK and by Merck & Co. Laboratories (Rickes et
al., 1948a, b) and Lederle Laboratories (Stokstad et al., 1948) in the USA.
The two first groups found this compound in liver extracts and identified it as
the anti-pernicious anemia factor responsible for the effective control of the
disease. These experimental programs were related to a continuous study over
a 20-year period of the therapeutic effect of whole beef liver in pernicious
In their communication Rickes et al. also reported the presence of significant
amounts of vitamin B12 in media fermented by a grisein-producing strain of
Streptomyces griseus as well as in culture broths from fermentations by
Mycobacterium smegmatis, Lactobacillus arabinosus, Bacillus subtilis, Streptomyces roseochromogenes and Streptomyces antibioticus. The vitamin B12
produced was measured by the microbiological assay based on the growth
response of Lactobacillus lactis (Domer strain) devised by Shorb (1947, 1948).
It was also at about the same time that Stokstad et al. (1948) reported the
production by Flavobacterium solare of a growth promotor, active in the
so-called 'animal protein factor' assay in chicks and effective in the treatment
of pernicious anemia in the human body which was identified as vitamin B 12 .
Microbial synthesis was also implicated in the early studies on the presence of
'animal protein factor' in manures and feces (Cary et al., 1946; Hartman and
Cary, 1946; Rubin & Bird, 1946; Lillie et al., 1948). Subsequent investigations,
which showed that the potency of incubated feces and manures was higher
than the freshly voided material, confirmed the earlier hypothesis of the
importance of micro-organisms in this synthesis (McGinnis et al., 1947; Sahashi
et al., 1953). With this background of information, the finding of vitamin B12 in
sewage was not unexpected (Hoover et al., 1951), and some efforts have been
made to exploit this source (Bernhauer & Friedrich, 1954; Friedrich &
Bernhauer 1958).
Although vitamin B12 is present in small amounts in almost every animal
tissue, it originates from micro-organisms. Depending on the nature of their
nutritional habits and digestive physiology, animals obtain the vitamin from
their own intestinal flora or from other animals through their meat diet. An
exogenous supply is mandatory for man. The importance of microbial synthesis
of this group of vitamins has been summarized as follows by Smith
It seems probable that the only primary source of vitamin B12 in nature is
the metabolic activity of micro-organisms; there is no convincing evidence
for its elaboration in tissues of higher plants or animals. It is synthesized by a
wide range of bacteria and actinomycetes, though apparently not to any
extent by yeasts or fungi.
Robbins et al. (1950) have concurred and stated:
It appears probable that the synthetic activity of micro-organisms, especially

bacteria and actinomycetes, is the original source of vitamin B12 in nature.

Microbial Production of Vitamin B12


The elucidation of the chemical structure of vitamin B12 is one of the most
outstanding examples of the successful collaboration of the chemist, the X-ray
crystallographer, and the biologist. The chemistry of this complex molecule has
been reviewed by those actively connected with the research and their
interpretation of the X-ray crystallography (Brink et al., 1954; Hodgkin et al.,
1955, 1957) and the degradative studies leading to the confirmation of the
structural formula, proposed by Dr Hodgkin and associates, should be
consulted for a guide to the literature (Folkers & Wolf, 1954; Wolf & Folkers,
1954; Bonnett et al., 1955; Folkers et al., 1957).
Ten years were spent from the first efforts of Woodward and Eschenmoser
until a full chemical synthesis was achieved (Krieger, 1973; Maugh, 1973). It
turned out to be a very difficult and it involved a group in Zurich, and a group
in Cambridge (UK) with more than 100 people. The synthesis requires about
70 steps and is of no value for industrial purposes. Today, vitamin B12 is
exclusively obtained by fermentation processes utilizing high-producing microorganisms or, less frequently, sewage.


Vitamin B12 (or cyanocobalamin) belongs to the large family of the cobalt
corrinoids exhibiting the general formula presented in Fig. 1. The molecule of
cobalamins, which are the most interesting cobalt corrinoids, is formed by the
following entities: a macrocycle in planar position, constituted by 4 reduced
pyrrol rings linked through their N -atoms to a cobalt atom in central position
(corrin) and showing 6 side chains (3 acetamide and 3 propionamide residues).
Almost perpendicular to the macrocycle a nucleotide, formed by phosphate,
ribose and 5,6-dimethylbenzimidazole is linked, through a coordination bond,
to the cobalt atom as shown in Fig. 2. The group of the cobalamins includes
cyanocobalamin or true vitamin B 12 , hydroxocobalmin or vitamin B 12 a,
methylcobalamin or mecobalamin and 5'-desoxyadenosilcobalamin or cob ammide or coenzyme B12 of Barker (Barker et al., 1958), characterized by a
cyano, hydrozyl, methyl or 5'-desoxyadenosyl radical respectively (Fig. 2).
Other cobaltcorrinoids with different heterocyclic bases (like substituted
benzimidazoles or purines), have been found in various micro-organisms,
either spontaneously or after the supply of the corresponding base to the
fermentation medium. Pseudovitamin B12 and factor III, in which the base is
adenosine and 5-hydroxybenzimidazole respectively, are the most frequently
encountered. They can be considered analogues of vitamin B 12 . A list of these
compounds is reported by Perlman (1959) and by Marvyn & Smith (1964).
Natural analogues designated as 'incomplete' are also known, for instance
factor B which is devoid of the ribose-phosphate moiety. All the known
analogues showed low activity as growth factors for vertebrates and no
therapeutic effect but they are very potent growth factors for the microorganisms. Cyanocobalamin, hydroxycobalamin, methylcobalamin and vitamin

C. Spalla et al.







'~ ____ ,_


H,C '0><' . . . . . . . . 0-






Trivial name




Vitamin Bua




C obamam ide or
coenzyme Bu:

Fig. 1. General structure of cobaltocorrinoids with references to cobalamins (Florent &

Ninet, 1979).

B12 coenzyme (5,6-dimethylbenzimidazol-5'-deoxyadenosylcobalamin) are the

only compounds of industrial interest.
About the nomenclature, as shown in Fig. 1, the corrinoid molecule without
the nucleotide is named cobinamide or factor B. When only the base is absent,
the compound is called cobamide and finally the complete form, i.e. with the
5,6 dimethylbenzimidazole attached, is named cobalamin.
The cobalamins are water-soluble compounds which crystallize in red
needles and are present in nature in their coenzyme form. Vitamin B12
coenzyme is very unstable to light and only stable at low temperature (-lOOC)
in aqueous solution. At room temperature it is easily transformed to
hydroxycobalamin which has been firstly isolated from Streptomyces aureofaci-

Microbial Production of Vitamin

Vitamin B12



V1tamin B12 Coenzyme

Fig. 2. Structures of vitamin BI2 and BI2 coenzyme (Perlman, 1967).

ens. (Pierce et al., 1949, 1950). This compound is also unstable to light and
temperature and in the presence of cyanide originates the more stable, well
known cyanocobalamin. Methylcobalamin is the active form by which vitamin
B12 is involved in methionine biosynthesis in Escherichia coli from which it was
first isolated (Guest et al., 1962). As the other native forms mentioned above,
it is also unstable to light.
The B12 corrinoids show the best stability at pH ranging from 4 to 6; they are
sensitive to ascorbic acid, to thiols, to gamma rays. They are endocellular
products and are extracted from the cells by treatment with water, neutral
buffer solution, lower alcohols or aqueous acetone. When natural corrinoids
like vitamin B12 coenzyme, hydroxycobalamin or methylcobalamin are to be
obtained, the extraction must be performed in the dark. Such a procedure is
not needed when cyanocorrinoids are wanted. In fact, the addition of cyanide
to the cell material transforms the unstable coenzyme forms to the stable
cyano-form. This procedure is usually applied in industry. A comprehensive
review of the chemical and physical properties of vitamin B12 corrinoids has
been written by Friedrich (1975).


As the research on vitamin B12 progressed, it soon became clear that its best
source in nature is represented by micro-organisms. Actually, it is synthesized


C. Spalla et al.

by a wide variety of bacteria and actinomycetes, while its production by yeasts,

fungi and tissues of higher plants or animals has not been reported. At first,
vitamin B12 used for human therapy and as a supplement for animal feeds was
obtained as a by-product of antibiotic-producing fermentations. These included
processes for the production of streptomycin by S. griseus (Schindler &
Reichstein, 1952; Smith & Ball, 1953), neomycin with S. fradiae (Jackson et
al., 1951), chlorotetracyclin by S. aureofaciens (Pierce et al., 1950), and
others. As the demand for the vitamin grew, it became profitable to employ
specific production processes in which vitamin B12 was the only product. A
number of processes were tested and some of them operated on a commercial
scale (Periman, 1959). The products of some of these fermentations were used
mainly for the enrichment of feeds poor in animal protein factor, a use
promoted by Ansbacher et al. (1949), while the pure material was isolated to
be used in human therapy.
As already mentioned, fermented feces and manures showed higher vitamin
B12 content than the unfermented material. Hence, sources for isolation of
vitamin B 12-producing micro-organisms became, among others, hen feces
(Stokstad et al., 1948), poultry litter, intestinal contents of animals, like bovine
rumen (Di Marco & Spalla, 1957) as well as sewage sludge. Among the
organisms shown to produce vitamin B12 there are bacteria from the following
genera; Aerobacter, Agrobacterium, Alcaligenes, Azotobacter, Bacillus,
Clostridium, Corynebacterium, Escherichia, Flavobacterium, Methanobacillus,
Mycobacterium, Propionibacterium, Proteus, Pseudomonas, Rhizobium,
Serratia, Streptococcus, Xanthomonas (Periman, 1959). Significant quantities
of vitamin B12 have shown to be produced by the following species of
actinomycetes: Nocardia rugosa, N. gardneri, Streptomyces albidoftavus, S.
antibioticus, S. aureofaciens, S. aureus,S. farinosus, S. fradiae, S. griseus, S.
olivaceus, S. roseochromogenes and S. vinaceus (Periman, 1959).
The isolation of vitamin B12 formed by the above mentioned microorganisms was carried out, and, in many instances, examination of the isolated
materials showed that several types of vitamin B12 and of its analogues were
present. It is not unlikely that there are still some undiscovered vitamin
B 12-related compounds present in some cultures. So, for instance, Friedrich &
Bernhauer (1958) reported to have found in a sample of sewage sludge,
besides the identified c5-benzimidazolecobamidecyanide, some 17 other cobamides in their extracts.
Numerous micro-organisms have been considered as a potential source of
vitamin B12 for industrial production (Table 1). Because of their rapid growth
and high productivity, Propionibacterium shermanii and Pseudomonas denitrificans were finally selected for industrial purposes.
The usefulness of an industrial process with methane-utilizing bacteria has
still to be proven, while the one based on the use of hydrocarbons does not
seem attractive (Florent & Ninet, 1979).
Improvement of productivity is achieved by screening for spontaneous or
induced mutants obtained with mutagenic treatments. Whatever the strains

Microbial Production of Vitamin





Table 1
Production by Different Strains

Micromonospora sp.
Nocardia rugosa
Propionibacterium freudenreichii
Propionibacterium shermanii
Methanobacillus omelianskii
Protaminobacter ruber
Coryneformbacterium strain XF
Corynebacterium and Rhodopseudomonas
Nocardia gardneri


Wagman et al. (1969)
Farmaceutici Italia (1971a)
Uelaf (1960)
Speedie & Hull (1960)
Pantskhava & Bykhovsky (1966)
Kojima et al. (1976)
Dumenil et al. (1979; 1981)
Nakao et al. (1974)
Kyowa Hakko Kogyo Co. Ltd

and the culture conditions may be, it is necessary to add to the medium some
elements essential for vitamin biosynthesis, like cobalt ions and frequently
5,6-dimethylbenzimidazole. The addition of potential precursors such as
glycine, threonine, c>-aminolevulinic acid and aminopropanol proved sometimes to be beneficial.


Investigations on the biosynthesis of the vitamin B12 molecule had mainly the
objective of elucidating the biosynthetic pathway leading to the tetrapyrrole
macrocycle and that leading to the incorporation of the nucleotide to form
vitamin B 12 .
With few exceptions all authors agree with the general pathway shown in
Fig. 3. Two molecules of c>-aminolevulinic acid condense to form porphobilinogen which is the basic unit of the macrocycle. Latest studies identify
in uroporphyrinogen III the point where the biosynthesis of vitamin B12 and
that of porphyrins diverge. One of the most striking structural differences
between the corrin and the porphyrin ring system is the lack in the former of
a methene bridge between one of the pairs of pyrroline-type rings.
Even if the sequence of methylations in the course of the biosynthesis of the
corrin ring remains unknown, early work from Shemin's laboratory showed
that all of the methyl groups are derived from methionine (Bray & Shemin,
1963; Shemin & Bray, 1964).
Substances endowed with a strong stimulatory effect on the production of
vitamin B12 are betaine and choline. Ansbacher et al. (1949) reported that out
of all the microbial processes available in 1949, the best ones were characterized by the disappearance of choline from the medium.
Miller & Putter, quoted by Demain et al. (1968), discovered that betaine

C. Spalla et al.


.--Su-C-Cn-y-l--C-oA----,~lle~~%In:id I-I_X_2__ '-I-----,~_'


CH, units

Cobyrinic acid


5' -deaxyadenosine



Fig. 3. General pathway for the biosynthesis of cobalamins (Florent & Ninet, 1979).

and choline stimulate vitamin B12 production by Pseudomonas denitrificans;

this has been confirmed by Miller & Rosenblum (1960) and by McDaniel
(1961). Stern & Friedman (1960) found these two compounds to stimulate by
as much as five-to-sixfold the formation of vitamin B12 by Rhizobium trifolii,
R. meliloti, Agrobacterium sp. and Bacillus megaterium.
The stimulatory effect of betaine and choline on vitamin B12 production by
Pseudomonas denitrificans grown in a chemical defined medium is shown in
Fig. 4 (Demain et al., 1968). The effect of betaine and choline on vitamin B12
production by P. denitrificans in an industrial medium was studied by
Garofano & Merli of these laboratories (unpublished data). The complex
medium currently utilized for industrial production contains high concentration
of sugar beet molasses which, being particularly rich in betaine, is a very good

Microbial Production of Vitamin



- 12







:;: 4

Additive (mg Iml)

Fig. 4. EtIect of betaine and choline on vitamin BJ2 production by Pseudomonas

denitrificans (Demain et aI., 1968).

raw material for vitamin Bt2 production. Even in this case, however, addition
to the medium of an extra amount of betaine exerts a stimulatory effect on the
produced vitamin Bt2 which increases from 90 to 140 Ilg/ml. Under the same
conditions betaine can be substituted by choline with similar results provided
the medium be suitably buffered with CaC03
As far as the biochemical mechanism of the stimulation of vitamin B12
production by betaine and choline is concerned, it has been demonstrated that
it is not due to the methyl groups' donation to the corrin ring of vitamin B t2 , as
methionine only is the precursor of these 'extra' methyl groups (Demain &
White, 1971). According to Demain et al. (1968) and White & Demain (1971),
betaine is an absolute requirement also for porphyrin over-production and it is
needed by P. denitrijicans for over-production of both corrins and porphyrins.
Although the site of betaine action in the biosynthetic pathway is unknown,
these data point out that betaine may be a positive effector for some early
common step of corrin and porphyrin synthesis.
The industrial production of vitamin B t2 , likewise that of many other microbial
metabolites, is carried out utilizing strains resulting from long lasting and

c. Spalla et al.


intense programs of genetic modification directed to improve their characteristics, mainly, but not only, by increasing their productivity (Kyowa Hakko
Kogyo Co. Ltd, 1970; Nakao et at., 1974; Pliva Co., 1964; Pierrel S.p.A.,
These programs consist essentially of the treatment of the producing
micro-organisms with a mutagenic agent and of the selection of the strains
bringing mutations leading to some practical advantage (higher productivity,
good genetic stability, resistance to higher concentrations of substances present
in the medium, higher rate of growth, etc.). As this kind of mutant is
extremely rare (of the order of 10-5), their selection obtained by direct
examination of all the strains deriving from the mutagenic treatments is a very
tedious and long-lasting task. As soon as knowledge on the biosynthetic
pathway progresses, this essentially random technique is implemented and
more rational and productive techniques are adopted.
Since an inverse correlation had been found between vitamin B12 and
porphyrin productions (Di Marco et at., 1961), porphyrin-less mutants of
vitamin B l2-producing strains were searched. This was done in a very simple
way with Nocardia rugosa, taking advantage of the fact that porphyrins are
excreted in the medium surrounding the colonies on the Petri dishes forming a
reddish halo. Mutants without halo were easily detected by direct examination
of the isolation plates. Out of 180 haloless isolated colonies, about 30 proved
to be really unable to produce porphyrins and two of them showed a significant
increase in vitamin B12 production (Marnati & Spalla, unpublished data).
A more sophisticated technique leading to the same results in
Propionibacterium shermanii has been described by Barrere et at. (1981). The
catalase is an enzyme with a porphyrin moiety. Hence, mutants unable to
synthesize porphyrins cannot produce catalase and, on the other hand, among
mutants lacking catalase activity there are good chances of finding strains
unable to synthesize porphyrins. Catalase-less mutants were easily identified
on Petri dishes because they were unable to develop oxygen bubbles from a
drop of oxygen peroxide put onto the colony surface.
Other techniques consist of looking for mutants resistant to high concentrations of the vitamin BI2 precursor 5,6-dimethylbenzimidazole or to the
antimetabolite ethionine and others. The increase of vitamin B12 productivity,
like that of other metabolites, has been a stepwise process. A significant
example of strain improvement referred to P. denitrificans is shown in Fig. 5.


During the last 30 years, several reviews dealing with research on vitamin B 12
production have been published (Perlman, 1959, 1967; Prescott & Dunn, 1959;
Marvyn & Smith, 1964; Friedmann & Cagen, 1970).

Microbial Production of Vitamin B12


Fig. 5. Mutation and selection program for the improvement of vitamin B12 production
by Pseudomonas denitrificans. The mutagens employed are the following ones: 1,
ultraviolet; 2, ethyleneimine; 3, nitrosomethyluretane; 4-7, N-methyl-N-nitro-Nnitrosoguanidine; 8,10, ultraviolet-bromouracil; 9,11, psoralene near UV light (Garofano & Merli, unpublished data).

The main objectives of the research work according to Florent & Ninet
(1979) have been:
to acquire a better knowledge of the biosynthetic pathway in order to
support biochemical and genetic studies;
to improve strains which turned out to be the most useful ones (propionibacteria and Pseudomonas);
to replace traditional sugars by more economical nutrients in the media;
to improve extraction and purification techniques.

Production by Propionibacteria

Several microaerophilic propionibacteria produce cobaltocorrinoids in conventional carbohydrate media supplemented with cobalt and without aeration. For
this process Propionibacterium freudenreichii A Tee 6207, P. shermanii A Tee
13673 and their mutants which can synthesize their own dimethyl benzimidazole



Spalla et aI.

Lyophilized with skim milk

Test tube with medium (a) incubated 4 days at 30C

2-liter Erlenmeyer flask with 04 liter of medium (b) incubated 2 days at
30C without agitation

30-liter stainless steel fermentor with 10 liters medium (c), incubated 24 h
at 30C without aeration and with frequent adjustment of pH to 65 with
aqueous NH 4 0H solution

SOO-liter stainless steel fermentor with 340 liters of medium (d) sterilized
40 min at 120C, inoculated with 7 liters of second-stage seed culture and
incubated at 30C, during the first 80 h under slight nitrogen pressure (no
aeration) and with slow agitation, then during the next 88 h with agitation
and aeration (2 m3 h): pH adjusted to 70 by aqueous NH 4 0H addition
along the whole fermentation
Fig. 6. Vitamin B12 from Propionibacterium shermanii. Semi-pilot plant scale fe~IlI:en
tation process. Medium (a): tryptone, 10 g; yeast extract, 10 g; filtered tomato JUice,
200 ml; agar, 15 g; tap-water to 1 liter; pH adjusted to 72. Medium (b): medium (a)
without agar. Medium (c): corn-steep liquor, 20 g; dextrose, 90 g; tap-water to 1 liter;
pH adjusted to 65. Medium (d): corn-steep liquor, 40 g; dextrose (sterilized separately), 100 g; cobalt chloride, 20 mg; tap-water to 1 liter; pH adjusted to 70. (Adapted
from Florent & Ninet, 1979).

Microbial Production of Vitamin B 12


(DBI) are the most useful ones. Since aeration favours DBI formation, the use
of a two-stage culture is advisable in order to obtain the highest yields. In the
first stage, anaerobic culture is run to almost total depletion of sugar which
promotes the growth of the bacteria and cob amide biosynthesis. Then, in the
second stage, an aeration shift leads to DBI biosynthesis and conversion of
cobamide to cobalamin. The two stages can be carried out batchwise in the
same tank or continuously in two connected fermentors (Speedie & Hull,
1960). Almost no information is available on strain preservation and culture
maintenance conditions.
The fermentation media consist mainly of glucose or inverted molasses
(50-100 g/liter) with small amounts of ferrous, manganous and magnesium
salts in addition to cobalt salts (60-100 mg/liter), of buffering or neutralizing
agents and nitrogenous compounds. Among these, yeast preparations, casein
hydrolysates and other traditional sources are currently used. Corn steep
liquor (50-70 g/liter) is frequently preferred since it supplies some lactic and
pantothenic acids which enhance the growth of the bacteria. Generally, the
temperature of the culture is 30C and the pH is controlled at around 65-70.
Reports indicate yields of 25-40 mg/liter of vitamin B12 by propionibacteria.
Figure 6 gives an example of a current process using P. shermanii.


Production by Pseudomonas

Several Pseudomonas species are vitamin B12 producers but currently the most
used species is P. denitrificans (Miller & Rosenblum, 1960). In contrast to the
propionibacteria fermentation, that of Pseudomonas is characterized by the
fact that its growth parallels cobalamin biosynthesis under aerobic conditions
throughout the whole fermentation process.
Accordingly, fermentation is conducted with aeration and agitation in a
single tank, batchwise or continuously. The strain requires traditional nutrients, such as yeast extract, sucrose and several mineral salts in the growth
medium. In order to enhance vitamin B12 production, the medium has to be
supplemented at the beginning of the culture with 10-25 mg/liter of DBI and
40-200 mg/liter of cobaltous nitrate (McDaniel, 1961; Daniels, 1970).
Likewise, betaine and to some extent choline, have favorable effects in
activating some biosynthetic steps as mentioned before. Glutamic acid
stimulates the growth of bacteria (Daniels, 1966; Koike & Hattori, 1975).
Owing to its low costs and high betaine and glutamic acid content, beet
molasses are preferentially used in industrial fermentations at a 60-120 g/liter
concentration in conjunction with 2-5 g/liter of ammonium phosphate and
some oligoelements. The optimum temperature is 28C, with a pH around 70.
By mutation and selection of proper strains, vitamin B12 production rose
from 5 to 120-140 mg/liter within the last 20 years (Long, 1962; Lago &
Demain, 1969; Ferni & Pennella, unpublished data). The process is outlined in
Fig. 7. In Table 2 the composition of typical media for vitamin B12 production
with selected mutants of P. denitrificans is reported.


C. Spalla et al.

Lyophilized with skim milk

Test tube with medium A incubated 4 days at 28C

2-liter round-bottomed flask with 06 liter of medium E inoculated with a
working culture slants and incubated 48 hr at 28C on rotating shaker at
120 r.p.m. with 90 mm excursion.

40-80 liter stainless steel fermentor with 25-50 liter medium E inoculated
with 1-12% seed culture first stage, incubated 25-30h at 32C, 200
r.p.m. agitation 02 bar pressure, 05 v/vm' aeration.

500-liter stainless steel fermentor with 300-liter medium P inoculated with
5% seed culture second stage, incubated 140-160 hr at 32C, 02 bar
pressure. Impeller speed regulated in order to assure sufficient oxygen
during the first phase of growth (20-30 hr) and then to maintain constant
the oxygen consumption at the level reached during the first phase. The
fermentation lasts 140-150 h and gives a production of about 120130,ug/ml
Fig. 7. Vitamin B12 from Pseudomonas denitrificans. Pilot plant scale production. For
media composition see Table 2. (Femi & Pennella, unpublished data.).

Microbial Production of Vitamin

Table 2

Media for the Production of Vitamin




with Pseudomonas denitrificans

Media (g/liter)

Sugar beet molasses

Ammonium phosphate
Ammonium sulfate
Magnesium sulfate
Zinc sulfate
5-6 dimethylbenzimidazole
Cobalt chloride
Difeo agar
Tap water










An important parameter to follow during the production phase is the level of

dissolved .oxygen in the culture. At the beginning of the process the
concentration of the oxygen must be sufficiently high in order to avoid
limitation in the ratio of growth (p02 > 0). During the production phase, on
the contrary, oxygen becomes the growth limiting factor and this somehow
induces the onset of vitamin B12 production. It is therefore necessary to supply
oxygen (air) in limited amounts so as to maintain the p02levei at a value near
to zero.

8.1 Hydroxocobalamin
This compound is produced by cultures of numerous bacteria and streptomycetes and by P. shermanii (Ilieva, 1971). During the extraction, successive
transformations of native cobalamins in their sulfate, nitrate and chloro
derivatives, are required before a final hydrolytic treatment by Amberlite IRA
400 (OH-), to generate hydroxocobalamin (Kaczka et al., 1956; Pierrel
S.p.A., 1963a). The chemical synthesis from cyanocobalamin can also be
performed through sulfite and nitrite derivatives (Pierrel S.p.A., 1963b; Smith,

8.2 Deoxyadenosylcobalamin (Coenzyme Bu)

Coenzyme B12 is directly extracted from P. shermanii (Chinoin-Gyogyszer,
1971; Ilieva & Popova, 1974), from P. freudenreichii (Sifa, 1964) and from

C. Spalla et al.


Nocardia rugosa (Farmaceutici Italia, 1971a), where it can represent up to

80% of the total native cobalamins.
Coenzyme B12 is always endocellular; after harvest of the cells by centrifugation, extraction is performed in the cold by an acetone-water mixture, or at
80-100 C during a short time by a 2% phenol-aqueous solution or an
ethanol-water mixture. All operations have to be conducted in the cold, under
subdued light and avoiding any drastic pH condition, in the presence of
cyanide in order to reach a global yield of 80-85% in the extraction process
(Kaken Kagaku, 1965; Chinoin-Gyogyszer, 1971).


During the last 30 years the extraction processes have been improved many
times; they usually include the following main steps: solubilization of cobalamins and conversion to cyanocobalamin and isolation of a crude product, 80%
pure (utilizable directly for animal feeding), followed by purification to a
95-98% level (for medical use). Usually the whole broth or an aqueous
suspension of harvested cells is heated at 80-120 C for 10-30 min at
pH 65-85 in order to extract the vitamin B 12 . The conversion to cyanocobalamin is obtained by treating the heated broth or cell suspension with cyanide
or thiocyanate.
The extraction operation units are the classical ones and they are combined
to achieve an efficient and inexpensive process. An extraction process
presently used is briefly outlined in Fig. 8. For further details see Florent &
Ninet (1979).

The usefulness and value of microbiological assays in studying biosynthesis of
vitamin B12 and in its detection in natural materials cannot be overemphasized.
Indeed, considering it in retrospect, the vitamin B12 isolation by the American
group of researchers at Merck in comparison to the English group at Glaxo
was due to the possibility of the former one to make use, instead of clinical
tests, of the microbiological method. This was based on the growth response of
Lactobacillus lactis (Dorner strain) to the vitamin as first described by Shorb
(1947, 1948).
Although the sensitivity of this micro-organism to vitamin B12 is high, the
fact that it responds to ribosides and desoxyribosides and that its growth
response is influenced by changes in aeration of the culture as well as by
genetic variability of the culture (Shorb & Briggs, 1948), led to its substitution
with Lactobacillus leichmanii strains ATCC 4797 and A TCC 7830 which gave
more reproducible results (Skeggs et al., 1948). This turbidimetric assay found
general acceptance for the analysis of fermentation samples and other

Microbial Production of Vitamin B 12


Heating 30 min. at 120C Cooling, adjustment of pH to 85

Addition of KCN, agitation 16 h at 25C
Addition of zinc chloride (200 g)
Adjustment of pH to 80
Addition of filter-aid (200 g), agitation,

P. denitrificans

Extraction 3 times with 350 ml of a cresol
and carbon tetrachloride mixture (1: 2


Addition of 500 ml n-butanol
Extraction 3 times with 100 ml water

Organic extract I

Extraction 3 times with 30 ml of a cresol
and carbon tetrachloride mixture (1: 2

Aqueous extract

Organic extract II

Addition of 200 ml acetone and 100 ml


Crude vitamin B12

Dissolution in 10 ml methanol
Chromatography on activated aluminia
(4 g). Elution with 2% acetic acid in methanol. Precipitation with ether

Pure vitamin
Fig. 8. Crystalline vitamin



isolation process (adapted from Florent & Ninet, 1979).


C. Spalla et aI.

materials, although a number of problems were encountered. Further investigations showed that all of the vitamin B I2-requiring lactobacilli respond, at
varying degrees, to all of the cobamides but not to cobinamide. These
differences in response have been used as the basis for differential-type assays
but unsatisfactory results have been obtained when samples contained more
than two types of cobamides.
Davis & Mingioli (1950) reported that certain mutants derived by treatment
of E. coli (ATCC 9637) with UV light required either vitamin B12 or
methionine for growth. One of these (strain 113-3) has been, and still is,
widely used as a test organism in bioassay for the determination of the
cobamide content of fermentation materials and other natural products. As
this strain responds to all of the cobamides as well as to cobinamide (Ford &
Hunter, 1955) it is very useful for the detection of all of these factors in
bioautographs of paper chromatograms of such solutions (Ford & Holdsworth,
1952). This strain is cultivated in a very simple medium and can be used in the
agar-diffusion assay as well as in the turbidimetric assay method which is more
The usefulness of the growth response of certain protozoa as a method of
measuring the vitamin BI2 content of samples of natural products was first
reported by Hutner et al. (1949). They found that Euglena gracilis is vitamin
BIZ- and thiamin-requiring. In this case chlorophyll production under light
serves as a measure of the vitamin BI2 content in the samples. All of the
cobamides stimulate the growth of this organism, but some problems in
assaying fermentation samples are encountered as a result of the strain
sensitivity to antibiotics which may be present in them (Robbins et al., 1953).
This test is recommended for vitamin BI2 determination in serum.
Another protozoal species useful in bioassay for cobamides is Ochromonas
malhamensis (Ford, 1953; Hutner et al., 1953) which has been, and still is,
widely used since the finding that only 5,6-dimethylbenzimidazolecobamide
stimulates its growth. Because of this benzimidazole-cob amide specificity, O.
malhamensis is used in substitution of animal tests like chicks and others
where, as in man, only these compounds are active.
The possibility of using chemical or physical methods for the estimation of
the cobamides content in fermentation samples or other crude biological
materials has been attempted. In these cases it is mandatory to perform
preliminary separations of the cobamides from other substances contained in
the samples by thin-layer or column chromatography (Tortolani et al.,
1970a,b; Vogelmann & Wagner, 1973; Tortolani & Mantovani, 1974), or
electrophoresis (Tortolani & Ferri, 1974). Afterwards, several classical methods are available to assay isolated fractions or pure products; spectrophotometry in the visible spectrum (Farmaceutici Italia, 1971b; Celletti et al., 1976)
or in the IR spectrum (Goldstein et al., 1974), atomic absorption (Whitlock et
al., 1976), potentiometry (Goldstein & Duca, 1976) and enzymatic analysis
(Hermann & Mueller, 1976; Schneider, 1979).
The radioisotope dilution method is widely used (Cooper et al., 1979; Begley

Microbial Production of Vitamin



& Hall, 1979; Beck, 1979). Also the application of HPLC for rapid corrinoid
determination has been reported (Jacobsen et al., 1979).



The biological properties of vitamin B12 are many and this has never been
found for any other natural product; interesting biochemical, haematological,
neurological and oncological aspects in man have been reported. This
physiological versatility can be explained by the fact that vitamin B12 is
involved in the functioning of numerous enzymes, which have vital roles in
assuring a normal, undisturbed evolution of the metabolism.

11.1 Biochemical mechanisms

Historically, among vitamin B12-dependent enzyme reactions, the isomerization of L-glutamic acid to /3-methylaspartic acid is the most important one
because its study led to the isolation of vitamin B12 coenzyme (Barker et al.,
1958). This enzyme, found in Clostridium tetanomorphum, in Acetobacter
suboxydans (Kato et aI., 1968), in Rhodopseudomonas sphaeroides and
Rhodospirillum rubrum (Ohmori et al., 1971) was also investigated in the
animal system (plasma of the sheep) where it was found to be vitamin
B12-independent (Marston et al., 1961). On the other hand, among biochemical disorders caused by vitamin B l2-deficient diet in man, a decreased
polyglutamate biosynthesis has been observed (Reed et al., 1979).
The degradation of propionic acid in higher organisms and its formation in
bacteria is performed by methyl-malonyl-CoA which is present in a catalytic
equilibrium with succinyl-CoA and regulated by a specific isomerase. Smith &
Monty (1959) showed that in liver homogenates of vitamin B 1rdeficient rats
the methyl-malonyl-CoA-mutase activity was strongly reduced. Successive
experiments performed on mitochondria from rat liver (Gurnani et al., 1960),
beef liver (Stern & Friedman, 1960), sheep kidney (Lengyel et al., 1960) and
extracts from Propionibacterium shermanii demonstrated the necessary presence of adenosylcobalamin for the functioning of methyl-malonyl-CoAmutase which has been isolated from Propionibacterium shermanii
(Kellermeyer et al., 1964), from liver (Cannata et al., 1965) and more recently
also reported for Rhodospirillum rub rum and Rhodopseudomonas sphaeroides
(Friedrich, 1975).
The presence of methyl-malonyl-CoA-mutase in animal tissues has been
reported by Cardinale et al. (1969) while its determination in leucocytes has
been described by Whitaker & Giorgio (1973). High levels of this enzyme were
observed in leucocytes of vitamin B 12-deficient patients as in rat tissues affected
by the same disorder. Degradation of methyl-malonyl-CoA through succinylCoA is of vital importance because of the toxicity of methylmalonic acid


C. Spa//o et aI.

(methylmalonylacidemia). Similarly a defective function of methylmalonylCoA racemase due to vitamin Bt2 deficiency causes accumulation of propionic
acid which is also toxic (propionicacidemia). The ribonucleotide reductase, an
enzyme catalyzing the reduction of ribonucleotides to desoxyribonucleotides
has been demonstrated as vitamin B t2-dependent in Lactobacillus leichmannii
and in Euglena gracilis (Beck, 1968). Its vitamin B12 dependency in animals is
still under discussion (Hopper, 1979).
In many micro-organisms and animals, the methionine synthetase which
necessitates as methyl donor the 5-methylhydrofolic acid, is vitamin B t2dependent.
acidhomocysteinmethyltransferase) from Escherichia coli and from liver, are very
similar and have identical coenzyme requirements. Some micro-organisms can
perform this synthesis even through a vitamin Bt2-independent pathway.
Similarly, animals may have an alternative way of methionine synthesis: rats
deficient in methionine, folic acid and vitamin Bt2 respond to betaine and
homocystein utilizing the enzyme betain-homocysteine methyltransferase
which catalyses the CH3 transfer from betain to homocysteine under formation
of methionine (Weissbach & Tylor, 1970).
Today, the mechanism generally accepted which explains the anti-anemic
activity of vitamin Bt2 is based on the interrelationship between the vitamin,
folic acid and methionine (Weissbach & Tylor, 1970). DNA synthesis
necessitates thymidilic acid which is produced by thymidilatesynthetase and
catalyzed by tetrahydrofolic acid derived from CHr-FH4 which in turn is
originated by the vitamin B t2-dependent methioninesynthetase.
In the absence of vitamin B t2 , folic acid is accumulated as CH3-FH4 and the
FH4 amount falls under the permissible level needed for the normal functioning of the entire cycle. In this case, vitamin Bt2 plays, in contrast to folic acid,
a secondary role as anti-pernicious anemia factor. The mechanism here
described is known as the 'folate trap'.
Vitamin Bt2 is also involved in the nucleic acid synthesis. In fact,
experimental data from systems where this synthesis is particularly active like
neoplastic mouse cells (Rotherham et al., 1971) and lymphocytes of vitamin
B t2 -deficient patients (Haurani, 1973), showed the existence of the interrelationship thymidine-vitamin B t2-folic acid. Vitamin Bt2 participation in the
functioning of methanesynthetase has been reviewed by Stadtman (1967) and
is important insofar as there are production processes of vitamin Bt2 employing
methanobacteria, currently under investigation (Florent & Ninet, 1979;
Dumenil et al., 1979, 1981).
11.2 Absorption, transport, metabolism and elimination
The physiological pathway of vitamin Bt2 transport is rather well known:
intrinsic factor (IF) secreted by the gastric mucosa picks up the minute
amounts of vitamin Bt2 out of the huge bulk of food ingested and mixed
thoroughly in the stomach. The firmly-bound vitamin Bt2 is well protected

Microbilll Production of Vitamin B12


against both the greediness of intestinal micro-organisms and loss in the stools
and is safely transported down to the gastrointestinal tract until it reaches a
section of the distal ileum, where it is reabsorbed by the intestinal mucosa.
This process is mediated by specific receptors of the mucosa cells for the
intrinsic-extrinsic-factor complex. Within the intestinal mucosa cells the
intrinsic-extrinsic-factor complex is split, and vitamin Bt2 is now transported
across the cell for reabsorption from its vascular contact side.
The IF (intrinsic factor of Castle; Castle et al. 1930) is a glycoprotein of
55 000-60 000 mol. wt showing variable sugar moieties according to the animal
species bearing it. The IF of pigs and man are very similar and this fact helped
much the elucidation of the absorption mechanism it plays. It is assumed that
IF possesses two receptor sites, one for the vitamin and the other for the
microvilli of the epithelial cells of the ileum. The absorption takes place at a
neutral pH and in presence of Ca or Mg ions (Shinton, 1972). As for the
formation of the IF-vitamin B12 complex, the presence of 5,6-benzimidazole
seems essential (Friedrich, 1975). There are other vitamin Bt2 binder proteins
isolated from the blood serum, named transcobalamins (TC). Among them TC
I and TC II are the most important ones. The biological turnover of vitamin
Bt2 in human liver is much lower than for other soluble vitamins, being its
half-life over 12 months.
Vitamin Bt2 deficiency causes pernicious anaemia, a disease which in the past
usually led to death within 1 month after detection and inevitably within 3 years.
As mentioned before, classical observations and experiments revealed the
cause of the disease in the lack of intrinsic factor produced in the stomach.
This factor was necessary for the reabsorption of vitamin B 12 . The anaemia is
hyperchromic, i.e. there is more haemoglobin available than red cells.
Maturation and division of erythroblasts are disturbed and the resulting cells
are too few and too big, i.e. macrocytic and megalocytic. Similar deficiencies
are seen in the neutrophyls which are hypersegmented and diminished in
number. These anomalies can be reversed by parenteral injection of extrinsic
factor (vitamin B 12) which immediately releases the inhibition of cell maturation and corrects the anaemia in a few days. Patients affected by pernicious
anaemia show a series of disorders like bone marrow depression, transverse
myelitis (myelopathy), peripheral neuropathy and subacute combined degeneration (Reynolds, 1979). A report on clinical diseases related to deficiencies of vitamin B12 transport proteins is given by Hitzig et al. (1979).
Vitamin B12 deficiency is revealed best by its serum level, since experiments on
the presence and determination of this compound showed that plasma and
serum are primarily indicative of this disorder, other tissues or organs being
affected only later on (Sullivan, 1970). The normal Bt2 serum level is
200-900 pg/ml and its total binding capacity is about 500-1100 pg/ml so that


C. Spalla et al.

the serum is saturated by only about 60% with vitamin B 12 . A serum B12 level
of about 200 pg/ml is considered critical. Clinical deficiency shows serum
concentrations of 150-80 pg/ml. Patients affected by leukemia show significant
higher serum levels due to a TC I increase. The presence of several cobalamins
in the plasma and the prevalence among them of methylcobalamin was
demonstrated first by Lindstrand & Stahlberg (1963) and Lindstrand (1964).
The optimal requirement of a healthy person for vitamin B12 is 05-1/-lg daily
according to Herbert (1968). The normal daily turnover of vitamin B12 is
25/-lg. In order to have a reabsorbtion of l/-lg of vitamin from an oral dose of
20/-lg, an amount of IF factor is required capable of binding 5-10 /-lg.
A vitamin B12 nutritional-dependent deficiency is rare and observable only
after about 10-20 years; it usually occurs in persons who live on a pure
vegetable diet. Oral treatment of vitamin B12 deficiency caused by the absence
or a defective function of IF allows a low level of reabsorbtion varying from 1
to 3% of a dosage of 100000 /-lg. Therefore, a maintenance dose of about
300 /-lg daily of cyanocobalamin at intervals of 3-4 days is suggested.
According to Berlin et al. (1968) even higher doses (5OO-1OO0/-lg daily) are
suggested. Tablets of 10 and 100 mg are available. In case of parenteral
treatments in presence of minor vitamin B12 deficiencies a dosage of
500-1000 /-lg of hydroxycobalamin injected 3-4 times at regular intervals
within a year is suggested (Heinrich, 1970). Neuropathies caused by vitamin
B12 deficiency need massive dosages of the order of 1000/-lg daily for the first
week followed by 1000 /-lg twice a week for several months administered by
injection (Friedrich, 1975). The assumption that vitamin B12 lacks side-effects
even at high doses has to be revised.
Vitamin B12 is available in numerous forms and grades. As the US market is
still using almost exclusively cyanocobalamin, the major portion of B12
material in the world keeps being distributed in that form. However, in
Europe hydroxycobalamin is preferred since it shows a better uptake in the
liver, less urinary excretion and a more sustained serum level than cyanocobalamin (Heinrich, 1970) and permits a less frequent dosage regimen than needed
for the cyano form. Hydroxycobalamin is often combined with cob amide
(vitamin B12 coenzymes).
The pharmaceutical use of vitamin B12 varies considerably. In the United
States injections of vitamin B12 used as a general boost against tiredness as well
as a remedy for miscellaneous aches and pains are virtually unknown; in many
other countries, particularly in France, Italy and Spain the above-mentioned
indications are the predominant ones. On the other hand, the US is by far the
single largest market on a per capita basis, for vitamin B12 used in dietary
supplement formulations; straight vitamin B12 tablets are available in US stores
in strengths of 100 /-lg. Vitamin B12 finds a fairly large use as feed supplement.
Generally it is dosed into all animal feeds in Europe and the USA with the
exception of ruminants. The dosage levels are of 10-30 mg/ton of feed for
poultry, pigs and for calves as milk replacer.
The overall world market of vitamin B 12 , including both pharmaceutical and

Microbial Production of Vitamin



animal feed use, is around 3 tons/year which is not expected to increase in the
near future. The bulk price varies from 3 to 4 US $/g. The main producers of
this vitamin are Rhone-Poulenc and Roussel-Uelaf (France), Glaxo (UK),
Merck (USA) and Medimpex (Yugoslavia).

The help of Anna Tomassoni in the translation and typing of the text is

Ansbacher, S., Hill, W. H., Tieman, J. W., Downing, J. F. & Caldwell, J. H., (1949).
Microbially synthesized APF (animal protein factor). Fedn Proc., 8, 180.
Barker, H. A., Weisbach, H. & Smith, R. D. (1958). A coenzyme containing
pseudovitamin B 12 . Proc. natn. Acad. Sci. U.S.A., 44, 1093-7.
Barrere, G., Geneste, B. & Satatier, A. (1981). Fabrication de la vitamine B 12 ,
l'amelioration d'un procede. Pour La Science, Novembre 1981, 56-60.
Beck, R. A., (1979). Essential prerequisites for the analysis of cyanocobalamin in
biochemical complex samples using radiometric competitive binding assays. In
Vitamin B 12 , ed. B. Zagalak & W. Friedrich. W. de Gruyter, Berlin, New York,
Beck, W. S. (1968). Desoxyribonucleotide synthesis and the role of B12 in erythropoiesis. In Vitamins and Hormones, Vol. 26, ed. R. S. Harris, G. G Wool, J. A.
Loraine & K. V. Thimann. Academic Press, New York, London, pp. 413-42.
Begley, J. A. & Hall, C. A. (1979). Effect of residual extract products and the type of
binders (R or IF) on serum vitamin B12 levels by radioisotope dilution assay. In
Vitamin B 12 , ed. B. Zagalak & W. Friedrich. W. de Gruyter, Berlin, New York,
Berlin, H., Berlin, R. & Brante, G. (1968). Oral treatment of pernicious anaemia with
high doses of vitamin B12 without intrinsic factor. Acta Med. Scand., 184,247.
Bemhauer, K. & Friedrich, W. (1954). Uber die vitamine der B13 gruppe. Angew.
Chem., 66,776-80.
Bonnett, R., Cannon, J. R., Johnson, A. W., Sutherland, L., Todd, A. R. & Smith, E.
L. (1955). Structure of vitamin B 12 . Nature, Lond., 176,325-8.
Bray, R. C. & Shemin, D. (1963). On the biosynthesis of vitamin B 12 . J. Bioi. Chem.,
238, 1501-8.
Brink, c., Hodgkin, D. C., Lindsey, J., Pickworth, J., Robertson, J. H. & White, J.
G. (1954). Structure of vitamin B 12 . Nature, Lond. 174, 1169-7l.
Cannata, J. J. B., Focesi, A., Mazumder, R. & Ochoa, S. (1965). Metabolism of
propionic acid in animal tissues. XII properties of mammalian methylmalonylcoenzyme A mutase. J. BioI. Chem., 240, 3249-57.
Cardinale, G. J., Dreyfus, P. M., Auld, P. & Abeles, R. H. (1969). Experimental
vitamin B12 deficiency: its effect on tissue vitamin B 12-coenzyme levels and on the
metabolism of Methylmalonyl-CoA. Archs Biochem. Biophys., 131,92-9.
Cary, C. A., Hartman, A. M., Dryden, L. P. & Likely, G. D. (1946). An unidentified
factor essential for rat growth. Fedn Proc., 5, 128.


C. Spalla et aI.

Castle, W. B., Townsend, W. C. & Heath, C. W. (1930). Observations on etiologic

relationship of achylia gastrica to pernicious anemia; nature of reaction between
normal human gastric juice and beef muscle leading to clinical improvement and
increased blood formation similar to etIect of liver feeding. Am. J. Med. Sci., 180,
Celletti, P., Moretti, G. P. & Petrangeli, B. (1976). Determinazione spettrofotometrica
ditIerenziale nel visibile di alcune cobalamine in presenza di sostanze interferenti.
Farmaco, ed. prat., 31, 413-19.
Chinoin-Gyogyszer (1971). Crystalline coenzyme B12 by fermentation. French patent
Cooper, J. P., Jonas, E. & Whithead, V. M. (1979). Vitamin B12 assay: an evolution
of radiodilution assay using cobinamide to increase specificity. In Vitamin B 12 , ed.
B. Zagalak & W. Friedrich, W. de Gruyter, Berlin, New York, pp. 655-6.
Daniels, J. H. (1966). Inhibition of growth of Pseudomonas denitrificans by aminoacids.
Can. J. Microbiol., 12, 1095-8.
Daniels, J. H. (1970). Some factors influencing vitamin B12 production by Pseudomonas
denitrificans. Can. J. Microbiol., 16,809-15.
Davis, B. D. & Mingioli, E. S. (1950). Mutants of Escherichia coli requiring
methionine or vitamin B 12 . J. Bacteriol., 60, 17-28.
Demain, A. L. & White, R. F. (1971). Porphyrin overproduction by Pseudomonas
denitrificans: essentiality of betaine and stimulation of ethionine. J. Bacteriol., 107,
Demain, A. L., Daniels, H. J., Schnable, L. & White, R. F. (1968). Specificity of the
stimulatory etIect of betaine on the vitamin B12 fermentation. Nature, Lond. 220,
Di Marco, A. & Spalla, C. (1957). La produzione di cobalamine da fermentazioni con
una nuova specie di Nocardia: Nocardia rugosa. Giorn. Microbiol., 4,24-30.
Di Marco, A., Boretti, G. & Spalla, C. (1961). Biosynthesis of cyanocobalamin in
Nocardia rugosa. Sci. Repts. 1st. Sup. Sanita (Rome), 1, 355-67.
Dumenil, G., Cremieux, A., Couderc, R., Chevalier, J., Guiraud, H. & Ballarini, D.
(1979). Production of Vitamin B12 by gram-variable methanol-utilizing bacteria.
Biotechnol. Lett., 1, 371-6.
Dumenil, G., Cremieux, A., Chevalier, J. & Guiraud, H. (1981). Vitamin B12
formation by a gram-variable methanol-utilizing bacterium. Biotechnol. Lett., 3,
Farmaceutici Italia (1971a). Cobamide preparation by fermentation of Nocardia rugosa
7892 Fi. French Patent 2,078,641.
Farmaceutici Italia (1971b). Cobalamin analysis of fermentation brews by automatic
solvent extraction and photometric assay. French Patent, 2,083,992.
Florent, J. & Ninet, L. (1979). Vitamin B 12 . Microbial Technology, ed. H. J. Peppler &
D. Perlman. Academic Press, New York, pp. 497-519.
Folkers, K. A. & Wolf, D. E. (1954). Chemistry of vitamin B 12. In Vitamins and
Hormones, Vol. 12, ed. R. S. Harris, G. F. Marriam & K. V. Thimann. Academic
Press, New York, pp. 1-51.
Folkers, K. A., Robinson, F. M. & Shunk, C. H. (1957). Vitamin B12 und Intrinsic
Factor, ed. H. C. Heinrich. Enke, Stuttgart, pp. 3-31.
Ford, J. E. (1953). Microbiological tests. Nature, Lond., 171, 149-50.
Ford, J. E. & Holdsworth, E. S. (1952). An improved bioautographic technique.
Biochem. J., 53, XXII-XXIII.
Ford, J. E. & Hunter, S. H. (1955). Role of vitamin B12 in the metabolism of
micro-organisms. In Vitamins and Hormones, Vol. 13, ed. R. S. Harris, G. F.
Marriam & K. V. Thimann. Academic Press, New York, pp. 101-36.
Friedman, H. C. & Cagen, L. M. (1970). Microbial biosynthesis of Vitamin B12-like
compounds. Ann. Rev. Microbiol., 24, 160-208.

Microbial Production of Vitamin



Friedrich, W. (1975). Vitamin B12 und verwandte Corrinoide. In Fermente, Hormone,

Vitamine, Vol. 111/2, ed. R. Amann & W. Dirscherl. Georg Thieme, Stuttgart, pp.
Friedrich, W. & Bernhauer, K. (1958). Die isolierung der Benzimidazol-und 5-Methylbenzimidazol-cobalamin-Analoga aus Faulschlamm. Chem. Ber., 91,2061-5.
Goldstein, S. & Duca, A. (1976). Determination of cyanobalamin by thermal
decomposition of the cyano group using an ion-selective electrode. J. Pharm. Sci.,
65, 1831-3.
Goldstein, S., Ciupitoiu, A., Vasilescu, V. & Duca, A. (1974). Determination of
Vitamin B12 by I.R. absorption spectrometry of the cyano-group in benzyl alcohol.
Can. J. Pharm. Sci., 9, 96-8.
Guest, J. R., Friedman, S., Woods, D. D. & Smith, E. L. (1962). A methyl analogue
of cobamide coenzyme in relation to methionine synthesis by bacteria. Nature,
Lond., 195,340-2.
Gurnani, S., Mistry, S. P. & Johnson, B. C. (1960). Function of vitamin B12 in
methylmalonyl metabolism I. Effect of a cofactor form of B12 on the activity of
metbylmalonyl-CoA isomerase. Biochim. biophys. Acta, 38, 187-8.
Hartman, A. M. & Cary, C. A. (1946). Occurrence in foods, of an unidentified factor
essential for rat growth. Fedn Proc., 5, 137.
Haurani, F. I. (1973). Vitamin B12 and the megaloblastic development. Science, IS2,
Heinrich, H. C. (1970). Renale Vitamin B 12-Ausscheidung. Dtsch. Med. Wschr., 95,
Herbert, V. (1968). Nutritional requirements for vitamin B12 and folic acid. Am. J. din.
Nutr., 21,743-52.
Hermann, R. & Mueller, O. (1976). Enzymatische Vitamin B I2 -Bestimmung. Hoppe
Seyler's, Z. physiol. Chem., 357,1695-8.
Hitzig, W. H., Frater-Schroeder, M. & Seger, R. (1979). Clinical disease related to
deficiencies of vitamin B12 transport proteins. In Vitamin B 12 , ed. B. Zagalak & W.
Freidrich. W. De Gruyter, Berlin, New York, pp. 1009-16.
Hodgkin, D. C., Pickworth, J. & Robertson, J. (1955). Structure of vitamin B12 .
Nature, Lond., 176, 325-8.
Hodgkin, D. C., Kamper, J. Lindsey, J., Mackay, M., Pickworth, J., Robertson, J. H.,
Shoemaker, C. B., White, J. G., Prosen, R. J. & Trubelood, K. N. (1957).
Structure of vitamin B 12. Proc. R. Soc., 242, 228-63.
Hoover, S. R., Jasewicz, L. B. & Porges, N. (1951). Vitamin B12 in activated sewage
sludge. Science, 114,213.
Hopper, S. (1979). An investigation of a possible role for coenzyme B12 in ribonucleotide reduction in rabbit bone marrow. In Vitamin B 12 , ed. B. Zagalak & W.
Friedrich. W. de Gruyter, Berlin, New York, pp. 1025-8.
Hutner, S. H., Provasoli, L., Stokstad, E. L. R., Hoffmann, C. E., Belt, M., Franklin,
A. L. & Juckes, T. H. (1949). Assay of anti-pernicious anemia factor with Englena.
Proc. Soc. exptl. Bioi. Med., 70, 118-20.
Hutner, S. R., Provasoli, L. & Filfus, J. (1953). Nutrition of some phagotrophic
fresh-water chrysomonads. Ann. N. Y. Acad. Sci., 56, 852-62.
Ilieva, M. (1971). Nauch. Tr., Vissh Inst. Khranit Vkusova Prom-st., Plovdiv., 18,
Ilieva, M. & Popova, Y. (1974). Nauch Tr., Vissh. Inst. Khranit Vkusova Prom-st.,
Plovdiv., 21, 105-12.
Jackson, W. G., Whitefield, G. B. & De Vries, W. H. (1951). The isolation of Vitamin
B12 from neomycin fermentations. J. Am. Chem. Soc., 73,337-41.
Jacobsen, D., Green, R. & Savage, M. (1979). Rapid determination of corrinoids by
high performance liquid chromatography. In Vitamin B 12 , ed. B. Zagalak & W.
Friedrich. W. de Gruyter, Berlin, New York, pp. 663-4.



Spalla et al.

Kaczka, E. A., Wolf, D. E. & Folkers, K. (1956). Vitamin B12 analogs and processes
for preparing the same. US Patent 2,738,302, Merck and Co., Inc.
Kaken Kagaku, K. K. (1965). Japanese Patent 65/4,440.
Kato, R, Hayashi, M. & Kamikubo, T. (1968). Isolation of 5,6Dimethylbenzimidazolylcobamide coenzyme as a cofactor for glutamate formation
from Acetobacter suboxydans. Biochim. biophys. Acta, 165,233-7.
Kellermeyer, R W., Allen, S. H. G., Stjernholm, R & Wood, H. G. (1964).
Methylmalonyl isomerase. IV. Purification and properties of the enzyme from
propionibacteria. J. Bioi. Chem., 239,2562-9.
Koike, 1. & Hattori, A. (1975). Growth yield of a denitrifying bacterium Pseudomonas
denitrificans under aerobic and denitrifying conditions. J. Gen. Microbiol., 88, 1-10.
Kojima, 1., Salo, H., Maruhashi, K. and Fujiwara, Y. (1976). Abstr. Int. Ferment.
Symp. 5th, 1976, No. 7-04, p. 134.
Krieger, J. H. (1973). Vitamin B 12 : the struggle toward synthesis. Chem & Eng. News,
51, 16-19, 25, 27, 28.
Kyowa Hakko Kogyo Co. Ltd (1970). Vitamin B12 by fermentation. Japanese Patent
Lago, B. D. & Demain, A. L. (1969). Alternate requirement for vitamin B12 or
methionine in mutants of Pseudomonas denitrificans, a vitamin B12 producing
bacterium. J. Bacteriol., 99,347-9.
Lengyel, Rajarshi & Ochoa, S. (1960). Metabolism of propionic acid in animal tissues.
VI Mammalian methylmalonyl isomerase and vitamin B12 coenzymes. Proc. natn.
Acad. Sci. U.S.A., 46, 1312-19.
Lillie, R J., Denton, C. A. & Bird, H. R. (1948). Relation of vitamin B12 to the
growth factor present in cow manure. J. BioI. Chem., 176, 1477-8.
Lindstrand, K. (1964). Isolation of Methylcobalamin from natural source material.
Nature, Lond., 204, 188-9.
Lindstrand, K. & Stahlberg, K. G. (1963). On vitamin B12 forms in Human Plasma.
Acta Med. Scand., 174,665.
Long, R A. (1962). Production of vitamin B 12 . U.S. Patent 3,018,225, Merck and Co.,
McDaniel, L. E. (1961). Production of cobalamins. US Patent 3,000,793: Merck and
Co., Inc.
McGinnis, J., Stevens, J. M. & Groves, K. (1947). The in vitro synthesis of a chick
growth promoting factor in hen feces. Poultry Sci., 26,432-3.
Marston, H. R, Allen, S. H. & Smith, R M. (1961). Primary metabolic defect
supervening on Vitamin B\2 deficiency in sheep. Nature, Lond., 190, 1085-91.
Marvyn, L. & Smith, E. L. (1964). The biochemistry of vitamin BI2 fermentation.
Progr. Industr. Microbiol., 5, 152-201.
Maugh, T. H. (1973). Vitamin B 12 : after 25 years, the first synthesis. Science, 170,
Miller, 1. M. & Rosenblum, C. (1960). Production of vitamin B\2 using deltaaminolevulinic acid. US Patent 2,939,822: Merck and Co., Inc.
Nakao, Y., Hisano, K., Kanemaru, T. & Yamano, T. (1974). Vitamin B12 production
by cultivation of Rhodopseudo monas and Corynebacterium bacteria in hydrocarbon
medium. Japanese Patent 74/15796: Takeda Chemical Ind. Ltd.
Ohmori, H., Ishitani, H., Sato, K., Shimizuki, S. & Fukui, S. (1971). Vitamin BI2
dependent glutamate mutase activity in photosynthetic bacteria. Biochem. biophys.
Res. Commun., 43, 156-62.
Pantskhava, E. S., & Bykhovsky, V. Y. (1966). Prod. Mikrobnogo Sin., Akad. Nauk
Latv. SSR Inst. Mikrobiol. 81-94: Chem. Abstr., 67 (1967) 10335.
Perlman, D. (1959). Microbial synthesis of cobamides. Adv. Appl. Microbiol., 1,
Perlman, D. (1967). Microbial production of therapeutic compounds. In Microbial

Microbial Production of Vitamin



Technology, ed. H. J. Peppler. Van Nostrand-Reinhold, New Jersey,

Pierce, J. V., Page, A. C. Jr, Stokstad, E. L. R. & Jukes, T. H. (1949). Crystallization
of vitamin B 12 b. J. Am. Chem. Soc., 71, 2952.
Pierce, J. V., Page, A. C. Jr, Stokstad, E. L. R & Jukes, T. H. (1950). Studies of
some characteristics of vitamin B 12 b. J. Am. Chem. Soc., 72, 2615-16.
Pierrel, S. p. A. (1963a). Hydroxocobalamine from cyanocobalamine by ion exchange.
French patent 1,325,304.
Pierrel, S. p. A. (1963b). Pure hydroxocobalamine from natural materials by ion
exchange. French Patent 1,325,308.
Pierrel, S. p. A. (1965). Preparation of hydroxycobalamin and cyanocobalamin (II) by
cultivating Propionibacterium Freundenreichii or Shermanii in a nutrient medium to
produce (II) and its analogues. British Patent 1,007,972.
Pliva, T. F. K. P. (1964). Production of vitamin B12 by a bacterial culture of
Propionibacterium vanielli petko vic grown on a nitrogen containing medium.
Belgian Patent 651,153.
Prescott, S. C. & Dunn, C. G. (1959). The production of Vitamin B 12 . In Industrial
Microbiology, 3rd edn. McGraw-Hill, New York, pp. 482-96.
Reed, B., Dinn, J., McCannon, S., Wilson, P., O'Sullivan, H., Wein, D. G. & Scott,
J. M. (1979). Alterations in mammalian cells induced by inactivation of vitamin B12
with nitrous oxide. In Vitamin B 12 , ed. B. Zagalak & W. Friedrich, W. de Gruyter,
Berlin, New York, pp. 1061-4.
Reynolds, E. H. (1979). The neurology of vitamin B12 deficiency. In Vitamin B 12 , ed.
B. Zagalak & W. Friedrich. W. de Gruyter, Berlin, New York, pp. 1001-8.
Rickes, E. L., Brink, N. G., Koniuszy, E. R., Wood, T. R & Folkers, K. (1948a).
Crystallin Vitamin B 12 . Science, 107, 396.
Rickes, E. L., Brink, N. G., Koniuszy, F. R, Wood, T. R. & Folkers, K. (1948b).
Vitamin B12 cobalt complex. Science, 108, 134.
Robbins, W. J., Hervey, A. & Stebbins, M. E. (1950). Studies on Euglena and vitamin
B 12 . Science, 112, 455.
Robbins, W. J., Hervey, A. & Stebbins, M. E. (1953). Euglena and vitamin B 12 . Ann.
N. y. Acad. Sci., 56, 818-30.
Rotherham, J., Price, F. M., Otani, T. T. & Evans, V. J. (1971). Some aspects of the
role of vitamin B12 and folic acid in DNA-thymidine synthesis in a neoplastic C3H
mouse cell strain. J. natn. Cancer Inst., 47, 277.
Rubin, M. & Bird, H. R. (1946). A chick growth factor in cow manure-I. Its
non-identity with chick growth factors previously described. J. BioI. Chem., 24,
Sahashi, Y., Iwamoto, K. & Hayashi, J. (1953). Biosynthesis of vitamin B12 in various
organisms. J. Biochem., 40, 227-44.
Schindler, O. and Reichstein, T. (1952). Isolierung von Vitamin B12 aus Leberextrakt
sowie aus Ruckstanden der Streptomycinifabrikation. He/v. Chim. Acta, 35,
Schneider, Z. (1979). Enzymatic estimation of Vitamin B 12 . In Vitamin B 12 , ed. B.
Zagalak & W. Friedrich. W. de Gruyter, Berlin, New York, pp. 673-4.
Shemin, D. & Bray, R. C. (1964). The biosynthesis of corrin structure of vitamin B 12 .
Ann. N. Y. Acad. Sci., 112, 615-21.
Shinton, N. Y. (1972). Vitamin B12 and folate metabolism. Br. Med. J., 1,556-9.
Shorb, M. S. (1947). Unidentified essential growth factors for Lactobacillus Lactis
found in refined liver extracts and in certain natural materials. J. Bacteriol., 53, 669.
Shorb, M. S. (1948). Activity of Vitamin B12 for the growth of Lactobacillus lactis.
Science, 107,397-8.
Shorb, R S. & Briggs, G. M. (1948). The effect of dissociation in Lactobacillus lactis
cultures on the requirement for vitamin B 12 J. Bioi. Chem., 176, 1463-4.


C. Spalla et aI.

Sifa (1964). 5,6-dimethylbenzimidazole-cobamide-coenzme. French Patent 1,368,892.

Skeggs, H. R., Huff, J. W., Wright, L. D. & Bosshardt, D. K. (1948). The use of
Lactobacillus leichmanii in the microbiological assay of the animal protein factor. J.
Bioi. Chem., 176,1459-60.
Smith, E. L. (1948a). Purification of anti-pernicious anaemia factors from liver. Nature,
Lond., 161,638-9.
Smith, E. L. (1948b). Presence of cobalt in the anti-pernicious anaemia factor. Nature,
Lond., 162, 144-5.
Smith, E. L. (1950-51). Nutrition. Abstract and review, 20, 795-809.
Smith, E. L. (1965). Hydroxocobalamin. US Patent 3,167,539: Glaxo Lab., Ltd.
Smith, E. L. & Ball, S. (1953). Nitrous acid treatment for separating impurities from
antipernicious anaemia active material. US Patent 2,630,401.
Smith, R. M. & Monty, K. (1959). Vitamin B12 and propionate metabolism. Biochem.
biophys. Res. Commun., 1,105-9.
Speedie, J. D. & Hull, G. W. (1960). Cobalamin producing fermentation process. US
Patent 2,951,017.
Stadtman, T. C. (1967). Methane fermentation. Ann. Rev. Microbiol., 21, 122-40.
Stern, J. R. & Friedman, D. C. (1960). Vitamin B12 and methylmalonyl CoA
isomerase. Biochim. biophys. Res. Commun., 2,82-7.
Stockstad, E. L. R., Page, A. C. Jr, Pierce, J., Franklin, A. L., Jukes, T. H., Heinle,
R. W., Epstein, M. & Welch, A. D. (1948). Activity of microbial animal protein
factor concentrates in pernicious anaemia. J. Lab. Clin. Med., 33, 860-64.
Sullivan, L. W. (1970). Differential diagnosis and management of the patient with
megaloblastic anemia. Am. J. Med., 48,209.
Tortolani, G. & Ferri, P. G. (1974). Electrophoretic separation of vitamin B12
derivatives. J. Chromatogr., 88,430-33.
Tortolani, G. & Mantovani, V. (1974). SP-Sephadex column chromatography of
cobalamins contained in liver extracts. J. Chromatogr., 92,201-6.
Tortolani, G., Bianchini, P. & Mantovani, V. (1970a). Separation and determination of
cobalamins on an SP-Sephadex column. J. Chromatogr., 53,577-9.
Tortolani, G., Bianchini, P. & Mantovani, V. (1970b). Separazione mediante cromatografia su colonna di varie cobalamine. Farmaco, ed. prat., 25, 772-5.
Uelaf (1960). French Patent 1,264,016.
Vogelmann, H. & Wagner, F. (1973). Elutionschromatographische Trennungen von
Corrinoiden an dem Unpolaren Adsorbenz Amberlite XAD-2. J. Chromatogr., 76,
Wagman, G. H., Gannon, R. D. & Weinstein, M. J. (1969). Production of vitamin B12
by Micromonospora. Appl. Microbiol., 17,648-9.
Weissbach, H. & Tylor, R. T. (1970). Roles of vitamin B12 and folic acid in Methionine
synthesis. In Vitamins and Hormones, Vol. 28, ed. R. S. Harris, P. L. Munson & E.
Diczfalusy. Academic Press, New York, London, pp. 415-40.
Whitaker, T. R. & Giorgio, A. J. (1973). A direct radioassay of methylmalonylcoenzymeA mutase using enzymatically synthesized DL-3 14C methylmalonyl-CoA.
Analyt. Biochem., 53, 522-32.
White, R. F. & Demain, A. L. (1971). Catabolism of betaine and its relationship to
cobalamin overproduction. Biochim. biophys. Acta, 237, 112-19.
Whitlock, L. L., Melton, J. R. & Billings, T. J. (1976). Determination of vitamin B12 in
dry feeds by atomic absorption spectrophotometry. J. Assoc. Analyt. Chem., 59,
Wolf, D. E. & Folkers, K. A. (1954). Vitamin B12-II. Chemistry. In The Vitamins,
Vol. 1, ed. W. H. Serbell & R. S. Harris. Academic Press, New York, pp. 397-417.

Chapter 16


Kyowa Hakko Kogyo Co, Ltd, Tokyo Research Laboratories 3-6-6, Asahimachi, Machida-shi, Tokyo, 194, Japan

ATC, aspartate transcarbamylase; AU, 6-azauracil; CPS, carbamyl phosphate
synthetase; CTP; cytidine-trisphosphate. DHO, dihydroorotase; DHOdeh,
dihydrooroatate dehydrogenase; OA, orotic acid; OMP, oritidine 5'-monophosphate; OMPdec, OMP decarboxylase; OPRT, orotate phosphoribosyltransferase; OR, orotidine; PRPP, phosphoribosylpyrophosphate; pyr, pyrimidine; UMP, uridine-monophosphate; ura, uracil


Orotic acid was first discovered in cow's milk in 1905 and was later found to
accumulate as an intermediate in the biosynthetic pathway of pyrimidine
nucleotides in a variety of mutants of micro-organisms. On the other hand, a
similar substance-which was a growth factor for rats, chickens and lactic acid
bacteria-had been isolated from distillers' dried solubles and had been known
as vitamin B 13 . It was proved in 1953 that orotic acid and vitamin B13 are
identical. Orotic acid is generally formed in mammals, so it is not a vitamin in
a strict sense.
Several reviews have been published on the production of orotic acid
(Furuya, 1976; Enei, 1984; Kuninaka, 1986) and the biosynthesis of pyrimidine
compounds (O'Donovan & Neuhard, 1970; Shiio, 1972).
In this chapter, recent aspects of the microbial production of orotic acid by
the de novo pathway will be described.
Orotic acid has a complex history. It was isolated from the whey of cow's milk
(Biscaro & Belloni, 1905) and subsequently found to be present in the milk of

* Present address: Central Research Laboratory, Shikishima Baking Co. Ltd, 3-5,
Shirakabe, Higashi-ku, Nagoya, 461 Japan.

K. Takayama & A. Furuya


various mammals. It was not until 1930 that the chemical structure of orotic
acid was perfectly known (Bachstez, 1930). Further investigations on the
compound were carried out late in the 1940s. Mitchell and co-workers (1947,
1948) observed that one of the pyrimidine-requiring mutants of Neurospora
utilized orotic acid instead of the pyrimidines, while the others, which did not
utilize orotic acid, accumulated large quantities of orotic acid in a culture
medium. Afterwards, isotopic experiments led to the conclusion that orotic
acid was the important intermediate in the biosynthetic pathway of pyrimidine
On the other hand, a growth factor for rats and chickens, was discovered
from distillers' dried solubles (DDS) and named vitamin B13 (Novak & Hauge,
1948). Moreover, Wright et al. (1950) observed that DDS was also effective in
promoting the growth of Lactobacillus bulgaricus and orotic acid could
substitute for DDS. Subsequently, evidence has been given that vitamin B13
was identical to orotic acid (Manna & Hauge, 1953).
The accumulation of orotic acid by mutants of several species of bacteria was
reported, but the amounts accumulated were very low. In 1961, the accumulation of orotic acid by a uracil-requiring mutant of Micrococcus glutamicus
(synon. Corynebacterium glutamicum) (a glutamic acid-producing microorganism), was reported (Tanaka et al., 1961; Kinoshita & Tanaka, 1963;
Konishita et al., 1963). This was the first report aimed at the industrial
production of orotic acid by fermentation procedure. Studies on the microbial
production of orotic acid have mainly been reported by Japanese researchers.
The industrial production of orotic acid by fermentation started in the
middle of the 1970s in Japan.




Chemical structure and properties of orotic acid (1,2,3,6-tetrahydro-2,6-dioxo4-pyrimidinecarboxylic acid; uracil-6-carboxylic acid) and oritidine (3-{J-oribofuranosylorotic acid, 6-carboxyuridine) are shown in Fig. 1 and Table 1.
Data are quoted from the Merck Index (10th edn, 1983). Potassium, sodium,








Fig. 1. Chemical structure of orotic acid (a) and orotidine (b).


Microbial Production of Vitamin B13

Table 1
Properties of Orotic Acid and Oritidine

Orotic acid

Molecular weight
Melting point


Absorption max.

282 nm
7680 (pH 44-72)
Soluble in water
about 17 mg/ml



Turns brown near 200C
but failed to melt at 400C
9570 (01 N HCl)
Soluble in hot water,
lower aliphatic alcohols

magnesium, calcium and ammonium salts of orotic acid are scarcely soluble in

3.2 Assay
Aqueous solutions of orotic acid and orotidine can be easily assayed
spectrophotometrically using the UV absorption property (282 and 268 nm,
respectively). In the case of a culture broth or a mixture solution, UV
detection combined with paper chromatography. TLC or HPLC is a very
useful method for determination. With regard to HPLC, ODS column (reverse
phase) is preferably applicable. Orotic acid can also be assayed selectively by a
microbiological method using Lactobacillus buigaricus 09 (Wright et ai., 1950)
and via a colorimetric method using dimethylaminobenzaldehyde (Adachi,


Micro-organisms which accumulate orotic acid and/or orotidine through the de

novo pathway are summarized in Table 2. Accumulation of orotic acid was first
reported using a pyrimidine auxotrophic mutant of Neurospora and a maximal
accumulation of 1 3 mg of orotic acid per ml was obtained (Mitchell et ai.,
1948). Thereafter, accumulation of orotic acid by Aerobacter aerogenes
(Brooke et ai., 1954), Escherichia coli (Yates & Pardee, 1956) and Serratia
marinorubra (Belser, 1961) were reported, but these studies focussed on the
biosynthetic pathways of pyrimidine derivatives and the accumulated amounts
were very low.
Later on, in 1961, it was reported that a uracil-requiring mutant of
Corynebacterium giutamicum accumulated 14 mg/ml in a culture medium
containing 10% glucose (Tanaka et ai., 1961). This strain was subsequently
found to be deleted in orotate phosphoribosyltransferase. Nakayama et ai.
(1965) examined the accumulation of orotic acid and orotidine by uracilrequiring mutants derived from the genera Corynebacterium, Brevibacterium,

pyr[NaF added]

Neurospora sp.
Penicillum commune
Candida tropicalis
Candida lipolytica

ARs , adenosine sensitive

pyrpyr[AUR added]
[AUR added]
uraura- (OPRT)
uraura- + adeuraurauraamino acidura- (OPRT)
urawild (OMP dec)
uraura- (OMP dec)
ura- (OMP dec)

Genetic character

Aerobacter aerogenes
Escherichia coli
Escherichia coli
Escherichia coli
Serratia marinorubra
Corynebacterium glutamicum
Bacillus subtilis
Corynebacterium glutamicum
Corynebacterium rathayi
Brevibacterium ammoniagenes
Bacillus subtilis
Escherichia coli
Brevibacterium ammoniagenes
Corynebacterium sp.
Escherichia coli
Arthrobacter paraffineus
Streptomyces showdoensis
Escherichia coli
Bacillus subtilis





Table 2
Orotic Acid-Producing Micro-organisms

Mitchell et al. (1948)

Sugimoto et af. (1962)
Watanabe et al. (1968)
Takayama et af. (1971)

Brooke et al. (1954)

Yates & Pardee (1956)
Skoda & Sorm (1958)
Handschumacher (1958)
Belser (1961)
Tanaka et al. (1961)
Konishi et af. (1963)
Nakayama et af. (1965)
Nakayama et af. (1965)
Nakayama et af. (1965)
Nakayama et al. (1965)
~akayama et al. (1965)
Skodova et al. (1969a,b)
Skodova et al. (1969b)
Machida et af. (1969,1970)
Kawamoto et al. (1970)
Ozaki et al. (1972)
Shimosaka et al. (1984)
Doi et al. (1988)








Microbial Production of Vitamin B13


Bacillus and Escherichia. Consequently, they found that C. glutamicum KY

9824 (ura- + ade-) and B. ammoniagenes KY 7349 (ura-) accumulated large
amounts of orotic acid; the maximal amount was 60 mg/ml by the latter.
Skodova & Skoda (1969) and Skodova et al. (1969) also reported the
accumulation of orotic acid using a uracil-requiring mutant of B. ammoniagenes CCEB 394, obtaining 58 mg/ml in a medium containing 5%
glucose, and also found a deletion of orotate phosphoribosyltransferase in the
mutant. Watanabe et al. (1968) observed the accumulation of orotic acid by an
adenine or hypoxanthine-requiring mutant of Candida tropicalis and obtained
a maximal accumulation of 7 mg/ml in a medium. Aspartic acid was recognized
to be effective as a precursor. Ozaki et al. (1972) found that a uracil-leaky
mutant of Streptomyces showdoensis (a showdomycin producer) accumulated
orotic acid and orotidine, obtaining 12 and 08 mg/ml respectively, and
confirmed the considerably decreased activity of OMP decarboxylase.
Research on the accumulation of orotic acid from n-paraffin has also been
carried out. Kawamoto et al. (1970) observed the accumulation of orotic acid
and orotidine, obtaining 67 mg/ml and 81 mg/ml, respectively, using a
uracil-requiring mutant induced from Arthrobacter paraffineus, a hydrocarbonutilizing bacterium. Takayama et al. (1971) also observed the accumulation of
orotic acid and orotidine, obtaining 76 mg/ml and 10 mg/ml, respectively,
using a uracil-requiring mutant induced from Candida lipolytica, a
hydrocarbon-utilizing yeast.
Machida & Kuninaka (1969) and Machida et al. (1970) reported that
wild-type strains of E. coli K-12 produced orotic acid, but other strains of E.
coli did not. Later, Womack & O'Donovan (1978) investigated this observation and concluded that this was caused by a mutation in the pyrF gene
encoding OMP decarboxylase. Shimosaka et al. (1984) reported the excretion
of orotic acid by an adenosine-sensitive (20 mM) strain induced from E. coli
C-600, derived from the K-12 strain. They also found that this growth
inhibitory effect was reversed by co-addition of pyrimidines to the medium; the
mutant displayed a lower level (7%) of activity for orotate phosphoribosyltransferase than a parent strain.
Recently, Doi et al. (1988) reported the accumulation of orotic acid and
orotidine by a uracil-requiring mutant F-100 induced from Bacillus subtilis IFO
14386; the accumulation was 55 and 27 mM, respectively. Successively, they
derived a uridine-producing mutant from the parent strain IFO 14386 which
was assumed to display a potent pyrimidine de novo pathway.



Biosynthesis of pyrimidine nucleotides

The de novo biosynthesis of pyrimidine nucleotides has been studied in detail in

bacteria, especially Escherichia coli and Salmonella typhimurium, in fungi and





CO 2



Cit.rull ine




Di hydrooro tate


























Fig. 2. Synthesis of pyrimidine nucleotide.







O.. C.... N/CH-COOH

~H, 9H,




Microbial Production of Vitamin B13


in mammals. Several reviews have been previously presented (O'Donovan &

Neuhard, 1970; Shiio, 1972; Enei, 1984). The pathway appears to be universal
in all organisms. The de novo pathway of pyrimidine biosynthesis is shown in
Fig. 2.
The atoms of the pyrimidine nucleus are derived from three simple
precursors, CO 2 , NH3 (from ammonia or glutamine) and aspartic acid.
Carbamyl phosphate is an intermediate of pyrimidine and arginine biosynthesis. Orotic acid is formed via N-carbamyl aspartic acid and dihydroorotic acid.
Orotic acid reacts with PRPP to OMP and successively decarboxylated to yield
UMP, which is the starting substance for synthesis of cytidine and thymidine
Orotic aciduria is a genetic disorder of pyrimidine biosynthesis in man in
which orotic acid accumulates in the blood and is excreted in the urine. The
disorder is alleviated by the feeding of uridine or cytidine.

Regulation of the biosynthetic pathway of UMP

In Salmonella typhimurium and Escherichia coli, biosynthesis de novo of UMP

is catalyzed by six enzymes encoded by six unlinked genes and operons. The
first enzyme, carbamyl phosphate synthetase (CPS), is essential for both
pyrimidine and arginine biosynthesis. Synthesis of this enzyme is regulated by
cumulative repression exerted by a pyrimidine nucleotide and arginine.
(ATC) ,
phoribosyltransferase (OPRT) and OMP decarboxylase (OMPdec) are found
to be repressed by uri dine nucleotides, whereas synthesis of dihydroorotase
(OHO) and dihydroorotate dehydrogenase (OHOdeh) are found to be
repressed primarily by cytidine nucleotides.
Carbamyl phosphate synthetase (CPS) is subjected to feedback inhibition by
uridine nucleotides, especially UMP, and this inhibition is reversed by
ornithine. Aspartate transcarbamylase (ATC) is sensitive to feedback inhibition by CTP and to activation by A TP.
5-Fluorouracil is a powerful antimetabolite of pyrimidine nucleotides. A
mutant of S. typhimurium resistant to a combination of 5-tluorouracil and
5-tluorouridine isolated by Jensen et al. (1982) possessed fourfold elevated
pools of the pyrimidine nucleotide triphosphate. Furthermore, the specific
activities of ATC and OPRT in the mutant were 40-fold and 7-fold higher than
in the parent strain when grown in minimal media and the synthesis of these
two enzymes was not repressed by pyrimidine compounds.
In Bacillus subtilis, the regulatory mechanism appeared to be considerably
different from that of Salmonella (Doi et al., 1989). Synthesis of all six
enzymes, catalyzing UMP biosynthesis, was strongly repressed by uracil
compounds and CPS was inhibited by UMP. In a mutant strain resistant to
uracil analogues such as 2-thiouracil or 6-azauracil, activities of the six enzymes
increased to 16-30-fold higher than in the parent strain. Moreover, the


K. Takayama & A. Furuya

repression by uracil compounds to the six enzymes and the feedback inhibition
by UMP to CPS were remarkably lost.
It has been observed that great differences exist for the regulation of the
pathway as a whole, especially for ATC, even though a common pathway is
found in all organisms (O'Donovan & Neuhard, 1970).


6.1 Fermentation
Most orotic acid-accumulating strains require uracil for growth, hence the
concentration of uracil in media severely affects the accumulation of orotic acid
as well as cell growth. Figure 3(a) and (b) shows the remarkable effects of
uracil on the accumulation of orotic acid by C. glutamicum and B.
ammoniagenes, respectively (Nakayama et al., 1965). Maximal accumulations
were obtained at concentrations of uracil which allowed the micro-organisms
to grow to about half-maximal levels. Similar results were obtained with
uracil-requiring mutants of other micro-organisms (Brooke et aI., 1954;
Tanaka et al., 1961; Konishi et al., 1963; Kawamoto et aI., 1970; Ozaki et al.,
1972; Doi et al., 1988).
For the accumulation of orotic acid by a mutant of Candida tropicalis,
addition of aspartic acid to the medium is effective. The amino acid has been
shown to be a metabolic precursor of orotic acid, and the accumulation of

( a)


_ 2.5

_ 2.4




-., 2.0




~ 1.6





&... __ -cr--- ___ 00

~ 1.0 \

\ jP',



E 0.5 )1









Uracil (pg/ml)


~ 1.2

0.22; 0 0 . 4



0.4:; ~0.8
i .~





0.4 ~







Fig. 3. Effect of uracil on orotic acid accumulation. (a) C. glutamicum KY 9824, (b) B.
ammoniagenes KY 7349 . , Orotic acid, /':,; orotidine; 0, Growth.

Microbial Production of Vitamin B13


orotic acid proceeds in parallel with aspartic acid consumption (Watanabe et

al., 1968). However, such effects of aspartic acid were not observed in the case
of C. glutamicum (Tanaka et al., 1961).
Recently, a mutant resistant to 5-ftuorouracil was derived from C. glutamicum ATCC 14275 (a uracil-requiring strain) by Takayama & Matsunaga
(1987). The activities of both carbamyl phosphate synthetase (CPS) and
aspartate transcarbamylase (ATC) in the mutant were recognized to increase,
comparing to the parent strain ATCC 14275 to some extent. By using the
improved strain, about 40-50 mg orotic acid per ml was accumulated in a
culture medium containing 17% glucose or other carbohydrates employing the
technique of submersed culture with stirring and aeration for 3-4 days.
Accumulation of orotic acid and/or orotidine by uracil-requiring mutants
can be ascribed to two genetic blocks, that is, orotate phosphoribosyltransferase (OPRT) and OMP decarboxylase. Since nucleotides
permeate poorly through cell membranes, it is assumed that OMP is excreted
only after dephosphorylation to orotidine or further degradation to orotic acid.
Therefore, when accumulation of orotidine in addition to orotic acid is
observed, it is reasonable to assume a deletion of OMP decarboxylase.
However, when accumulation of orotic acid alone is observed, it should be
confirmed which one of the two enzymes is deleted.
6-Azauracil, a pyrimidine analogue, is a powerful inhibitor of OMP
decarboxylase in micro-organisms in vivo. It exerts its effect only after being

Fig. 4. Bacterial cells and crystals of orotate in the orotic acid fermentation broth.
(scanning electron micrograph).

K. Takayama & A. Furuya





H,C" NH,


I + I


/ + I '--+ EtO,CHC







Fig. 5. Chemical synthesis of orotic acid.

converted to 6-aza-UMP, inhibits OMP decarboxylase, and causes the accumulation of OMP, which subsequently inhibits OPRT. As a result, orotic
acid is excreted into the medium.

Isolation and purification

Orotic acid in the fermentation broth usually exists as crystals of ammonium,

potassium and/or sodium salt due to its limited solubility (Fig. 4). Crude
orotate can be easily separated from the fermentation broth with repeated
decantation of supernatant liquid. After this operation, the resulting precipitates of crude orotate are, for example, dissolved in a solution of hydrochloric
acid or sulfuric acid at an elevated temperature and then the solution is filtered
to remove impurities and an excess of acid. For further purification, treatments
based on active carbon and on ion exchange resin are generally used.


Synthesis of orotic acid involving the condensation of diethyl oxalacetate with

S-methylthiourea is believed to be the most practical method in the laboratory
(Fig. 5(a. Hydantoin sometimes arises in reactions designed to make
pyrimidines. Diethyl oxaloacetate and urea can be converted into orotic acid
under suitable conditions (Fig. 5(b (Brown, 1984).


Orotic acid has been used as a hepatic drug. However, it may be contrarily
said that large doses of orotic acid result in liver disturbance, because of the
imbalance between purine and pyrimidine compounds.

Microbial Production of Vitamin B 13


Considerable quantities of orotic acid are applied in the field of medicine

and in health food in the form of free orotic acid, or its calcium, magnesium,
potassium or iron salt. The demand for orotate is estimated worldwide to be
about 100 tons/year.
Orotic acid is an intermediate in pyrimidine biosynthesis, hence, expansion
of the utilization of orotic acid for the synthesis of various types of pyrimidine
compounds is expected in the near future.

Adachi, S. (1963). A colorimetric determination of orotic acid. (I) Improvement of
p-dimethylaminobenzaldehyde method. Vitamin (Japanese), 27, 433-7.
Bachstez, M. (1930). Constitution of orotic acid. Ber. Dtsch. Chem. Ges., 863,
Belser, W. L. (1961). Uracil biosynthesis in Serratia marinorubra. Biochem. Biophys.
Res. Commun., 4,56-60.
Biscaro, G. & Belloni, E. (1905). Sur un nouveau constituant du lait. Monit. Sci., 19,
Brooke, M. S., Ushiba, D. & Magasanik, B. (1954). Some factors affecting the
excretion of orotic acid by mutants of Aerobacter aerogenes. J. Bacteriol., 68,
Brown, D. J. (1984). Pyrimidines and their benzo derivatives. In Comprehensive
Heterocyclic Chemistry, Vol. 3, ed. A. R. Katritzky & C. W. Rees. Pergamon
Press, Oxford, pp. 57-155.
Doi, M., Tsunemi, Y., Asahi, S., Akiyama, S. & Nakao, Y. (1988). Bacillus subtilis
mutants producing uridine in high yields. Agric. Bioi. Chem., 52, 1479-84.
Doi, M., Asahi, S., Tsunemi, Y. & Akiyama, S. (1989). Mechanism of uridine
production by Bacillus subtilis mutants. Appl. Microbiol. Biotechnol. (in press).
Enei, H. (1984). Fermentative production of pyrimidines, pyrimidine nucleosides and
pyrimidine nucleotides. Hakko to Kogyo (Japanese), 42, 562-9.
Furuya, A. (1976). Production of other nucleic acid-related substances. In Microbial
Production of Nucleic Acid-related Substances, ed. K. Ogata, S. Kinoshita, K. Aida
& T. Tsunoda, Kodansha, Tokyo, pp. 184-91.
Handschumacher, R. E. (1958). Bacterial preparation of orotidine-5'-phosphate and
uridine-5'-phosphate. Nature, Lond., 18'1., 1090-91.
Jensen, K. F., Neuhard, J. & Schack, L. (1982). RNA polymerase involvement in the
regulation of expression of Salmonella typhimurium pyr genes. Isolation and
characterization of fluorouracil-resistant mutant with high, constitutive expression of
the pyrB and pyrE genes due to mutation in rpoBe. The EMBO Journal, 1, 69-74.
Kawamoto, I., Nara, T., Misawa, M. & Kinoshita, S. (1970). Fermentative production
of orotic acid and orotidine from hydrocarbon. Agric. Bioi. Chem., 34, 1142-9.
Kinoshita, S., Tanaka, K. (1963). Production of orotic acid by fermentation method.
US Patent, 3 086 917.
Kinoshita, S., Tanaka, K. & Kimura, K. (1963). Production of orotic acid by
fermentation method. Japanese Published Examined Patent Application, S38-9950.
Konishi, S., Shiro, T. & Takahashi, M. (1963). Accumulation of orotidine by
auxotrophic mutant of Bacillus subtilis. Ann. Mtg. Agr. Chem. Soc. Japan, Abstract
(Japanese), p. SO.
Kuninaka, A. (1986). Nucleic acids, nucleotides and related compounds. In
Biotechnology, Vol. 4, ed. H. J. Rehm & G. Reed. VCH, Weinhein, pp. 108-9.


K. Takayama & A. Furuya

Machida, H. & Kuninaka, A. (1969). Studies on the accumulation of orotic acid by

Escherichia coli K12. Agric. Bioi. Chem., 33, 868-75.
Machida, H., Kuninaka, A. & Yoshino, H. (1970). Studies on the accumulation of
orotic acid by Escherichia coli K12. Part 2. Mechanism of the accumulation. Agric.
Bioi. Chem., 34, 1129-35.
Manna, L. & Hauge, S. M. (1953). A possible relationship of Vitamin B13 to orotic
acid. J. Bioi. Chem., 202,91-6.
Merck Index, 10th edn (1983). Merck & Co., Rahway, NJ, USA.
Mitchell, H. K. & Houlahan, M. B. (1947). Investigation on the biosynthesis of
pyrimidine nucleotides in Neurospora. Fedn Proc., 6,506-9.
Mitchell, H. K., Houlahan, M. B. & Nyc, J. F. (1948). The accumulation of orotic acid
by a pyrimidineless mutant of Neurospora. J. Bioi. Chem., 172,525-9.
Nakayama, S., Sato, Z., Tanaka, H. & Kinoshita, S. (1965). Accumulation of orotidine
and orotic acid by pyrimidine-auxotrophic mutant of micro-organisms. J. Agric.
Chem. Soc., Japan (Japanese), 39,118-22.
Novak, A. F. & Hauge, S. M. (1948). Isolation of the unidentified growth factor
(Vitamin, B 13) in distillers' dried solubles. J. BioI. Chem., 174,647-51.
O'Donovan, G. A. & Neuhard, J. (1970). Pyrimidine metabolism in micro-organisms.
Bacteriol. Rev., 34,278-343.
Ozaki, M. & Kimura, T. (1972). Mechanism of the accumulation of orotic acid and
orotidine by a mutant of Streptomyces showdoensis. Amino Acid Nucleic Acid, 26,
Ozaki, M., Tagawa, S. & Kimura, T. (1972). The accumulation of orotic acid and
orotidine by a mutant of Streptomyces showdoensis. Amino Acid. Nucleic Acid, 26,
Shiio, I. (1972). Regulation of nucleotide biosynthesis. Seikagaku (Japanese), 44, 7-21.
Shimosaka, M., Fukuda, Y., Murata, K. & Kimura, A. (1984). Growth inhibition by
purine derivatives and its reversal by pyrimidine derivatives in a mutant of
Escherichia coli K12. Agric. Bioi. Chem., 48, 1303-10.
Skoda, J. & Sorm, F. (1958). Accumulation of nucleic acid metabolites in Escherichia
coli exposed to the action of 6-azauracil. Biochim. biophys. Acta, 28, 659-60.
Skodov3, H. & Skoda, J. (1969). Mechanism of overproduction of orotic acid by a
mutant of Brevibacterium ammoniagenes. Appl. Microbiol., 17, 188-9.
Skodov3, H., Solinov3, H., Skoda, J. & Dyr, J. (1969). Orotic acid formation by
pyrimidine-deficient mutants of Brevibacterium ammoniagenes. Folia Microbiol.,
Sugimoto, H., Iwata, T. & Ishiyama, J. (1962). Studies on phosphodiesterases
produced by micro-organisms. (2) The influence of fluoride on the excretion of
nucleic acid derivatives by fungi. J. Agric. Chem. Soc., Japan (Japanese), 36,
Takayama, K. & Matsunaga, T. (1987). Production of orotic acid by fermentation
method. Japanese Patent Application, S62-261715.
Takayama, K., Kawabata, T.& Abe, S. (1971). Production of orotic acid by
fermentation method. Japanese Published Examined Patent Application, S516756.
Tanaka, K., Nakajima, Y. & Kinoshita, S. (1961). Accumulation of orotic acid by a
mutant strain of Micrococcus glutamicus. Ann. Mtg. Agr. Chem. Soc. Japan,
Abstract (Japanese), p. 14.
Watanabe, A., Tani, K. & Sasaki, Y. (1968). Accumulation of orotic acid by
auxotrophic mutant of Candida tropicalis. Amino Acid. Nucleic Acid (Japanese),
Womack, J. E. & O'Donovan, G. A. (1978). Orotic acid excretion in some wild-type
strains of Escherichia coli K-12. J. Bacteriol., 136,825-7.

Microbial Production of Vitamin B13


Wright, L. G., Huff, J. W., Skeggs, H. R., Valentik, K. A. & Bosshardt, D. K.

(1950). Orotic acid, a growth factor for Lactobacillus bulgaricus. 1. Am. Chern.
Soc., 72,2312-13.
Yates, R. A. & Pardee, A. B. (1956). Pyrimidine biosynthesis in Escherichia coli. 1.
BioI. Chern., 221, 743-55.

Chapter 17



PLIVA Pharmaceutical, Chemical, Food And Cosmetic Industry, Research

Institute, Zagreb, Yugoslavia


Microbial conversion of several chemical substances deserves special attention,

mainly due to its economical and ecological advantages as compared to
conventional chemical routes. Industrial production of vitamin C is an example
of sophisticated chemical manufacturing involving highly complex and expensive chemical and microbial technology. In this process only one step is
mediated by micro-organisms. This review is to provide a complete discussion
of results, methods and current ideas of L-ascorbic acid (vitamin C) biosynthesis by micro-organisms, particularly since this field has only rudimentally been
covered in the literature (Razumovskaya, 1962; Kulhanek, 1970; Crawford &
Crawford, 1980).
Many excellent articles have been published on L-ascorbic acid chemical
synthesis, biosynthesis in plants and animals and its biological role in living
organisms (Sebrell & Harris, 1967; Jaffe, 1984). All these interesting issues will
only briefly be covered in this review, the main objective being to focus on
micro-organisms involved in the biosynthetic steps and conditions to L-ascorbic
acid as a final product.


L-Ascorbic acid was first isolated by the famous Hungarian biochemist A.

Szent-Gyorgyi in 1928, under the name of 'crystalline hexuronic acid'. He used
natural material such as ox adrenal cortex, orange juice and cabbage juice for
the isolation of a new compound. In the process of conducting this work,
Szent-Gyorgyi and his co-workers established the identity of 'hexuronic acid',
e.g. the presence of an acidic group and the easy reversible nature of its
oxidation. In addition they found that green paprika was a particularly rich


V. Delic, D. Sunic & D. VlaSic

Table 1
Some Historical, Production and Economical Data on L-Ascorbic Acid Development
Time period










First isolation of 'hexuronic
Description of the structure.
Reichstein's first and second
chemical synthesis. First productions of vitamins by
chemical synthesis; microbial
oxidation as a step in
chemical synthesis.
L-Ascorbic acid synthesis on
industrial scale.
L-Ascorbic acid from AspergilIus niger; never reproduced.
Increasing of production
capacity in USA.
Preparation of intermediates
by biosynthesis
Many chemical and technical
modifications; e.g. electrochemical oxidation
Production in USA
Increasing production
c. 14% year. Total
production in the world.
Biosynthesis of L-ascorbic
acid intermediates by


Szent-Gyorgyi, 1928




Reichstein & Grtissner, 1934

Jaffe, 1984
Geiger-Hiiber & Galli, 1945

c. 17


Jaffe, 1984.
Yamazaki & Miki, 1953
Yamazaki, 1953a, b, c
Crawford & Crawford, 1980

c. 3000

Inflation; increasing in the

raw material and energy
Total world production.
34 000-36 000
Biosynthesis of 2-keto-Lgulonic acid by two-stage
Biosynthesis of 2-KLG on a
commercial scale in People's
Republic of China.
One step biosynthesis of 2-KLG
by genetically modified
Erwinia herbicola.
Total world production.
Predictions: (a) growth of nominal
capacity of main world
producers;" (b) intensive efforts on
rDNA techniques; (c) use of biocatalysts from modified enzymes
and/or cells and (d) research on
various bioreactors, e.g.
membrane reactors.

" See listing of world producers in Table 4.



Jaffe, 1984
Jaffe, 1984
Ullmanns, 1983

325-425 Isono et al.. 1968

Okazaki et al.. 1968
Okazaki et al.. 1969
Kanzaki & Okazaki, 1970
7,15-830 Jaffe, 1984
Ullmanns, 1983

Ullmanns, 1983
Sonoyama et al.. 1982
Kieslich, 1984

Anderson et al.. 1985


Microbial Production of Vitamin C


source of vitamin C (Svirbely & Szent-Gyorgyi, 1933). Detailed descriptions of

'hexuronic acid' led to the discovery of the correct structure of vitamin C, later
on named ascorbic acid by Szent-Gyorgyi and Haworth (Crawford & Crawford, 1980).
These data, together with those derived from the study of L-ascorbic acid
oxidation products, led Hirst et al. (1933) to propose that this compound was
3-keto-L-sorbosone. The early results of L-ascorbic acid structure elucidation are
summarized in several reviews (Haworth & Hirst, 1933; Hirst, 1939; Crawford
& Crawford, 1980).
The importance of L-ascorbic acid in human health care was established soon
after its discovery and caused tremendous interest in all aspects of its scientific
and industrial development. Hence, it is not surprising that up to now
L-ascorbic acid has been and still is one of the major products of interest to the
pharmaceutical, chemical and food industries. The main stages of the history
of L-ascorbic acid over the past 60 years are listed in Table 1.
Its medical importance and the role of L-ascorbic acid in human health have
been very well documented in over 22 000 publications that have appeared in
the literature from 1966 to 1984 (Jaffe, 1984). Since humans and some animals
lack L-gulono-y-lactone oxidase, the key oxidizing enzyme in liver for
L-ascorbic acid biosynthesis, they have to consume vitamin C from exogenous
sources. The main symptoms of L-ascorbic acid deficiency, mainly due to a
dietary lack of fresh fruit and vegetables, are general weakness, fatigue and
listlessness followed by a shortage of breath and aching in the bones. As the
illness-known from ancient times as scurvy-progresses, the skin becomes dry
and rough followed by a swelling of the gums, easy bleeding and a loss of
teeth. Obviously the lack of L-ascorbic acid affects many different metabolic
processes in living cells such as collagen synthesis, amino acid synthesis,
immune response, etc., although the precise mechanisms of its actions are not
clear. L-Ascorbic acid, an important reducing agent functioning as an electron carrier, plays an important role in the respiration and oxidative reactions
in the cells.




Names previously used for L-ascorbic acid were: cevitamic acid, hexuronic
acid, redoxon, scorbutamin, vitamin C and L-xylo-ascorbic acid. Throughout
this review, the name L-ascorbic acid will be used.
In 1965 the trivial name ascorbic acid or L-ascorbic acid was recommended
by the IUPAC-IUB Commission on Biochemical Nomenclature as a name for
vitamin C. The systematic name for L-ascorbic acid is L-threo-hex-2-enonic acid
y-lactone or L-threo-hex-2-enomo l,4-lactone (IUPAC-IUB, 1965).

V. Delic, D. Sunic & D. VlaJic



H5 0







'-- H 0
H"'" H"'"
_ _ =0

Fig. 1. Chemical constitution of L-ascorbic acid.

3.2 Properties
L-Ascorbic acid is white, crystalline solid melting at 192C. Specific rotation in
water at the sodium D line is +215C. In solution, L-ascorbic acid has: pKl of
417, pK2 of 1157. The most acidic proton is that of the 3C-hydroxyl group.
Crystals of sodium and calcium L-ascorbate have the metals associated with
0-3. Reversible oxidation takes place at + 127 m V.
Infrared, ultraviolet, IH-nuclear magnetic resonance (nmr) and 13C-nmr
spectra have all been reported (see: Schauenstein et al., 1948; Billman et al.,
1972; The Merck Index, 1983). Ogawa et al. (1977), studied the conformation
of L-ascorbic acid in deuterium oxide by 13C-nmr spectroscopy. They concluded that in aqueous solution the most favoured rotamer of L-ascorbic acid is
in the form (b) (Fig. 1). The rotamer is not present in the crystalline state of
L-ascorbic acid. The approximate conformation in the crystal is shown by
structure form (a) (Fig. 1).
Elucidation of L-ascorbic acid stereochemistry played an important role in
the determination of its stereomolecular structure and in the development of
chemical synthesis for industrial production. The configuration of L-ascorbic
acid was assignment to the L-series and confirmed by synthesis from o-glucose
and L-xylose. Chiral centers at C-2 and C-3 of these sugars were in the correct
position to become C-5 and C-4 of L-ascorbic acid, respectively (Reichstein &
Griissner, 1934). Other carbohydrates, e.g. L-gulose, L-idose, L-galactose and
L-talose have the correct chirality at C-4 and C-5 to be potential raw materials
for the synthesis, but their high cost makes them impractical.
Since fruit and vegetables are the main natural sources of L-ascorbic acid,
biosynthetic pathways in plants were subjected to extensive studies. It has been
shown that L-ascorbic acid is a product of hexose phosphate metabolism
including conversion of D-glucose or o-galactose to L-ascorbic acid (Fig. 2).

Microbial Production of Vitamin C






D-Glucuronic acid

o-Galacturonic acid

o-Glucuronic acid

L-Gulonic acid

L-Galactonic acid

L-Gulonic acid

UDP-Glucuronic acid





L-2-Keto-gulono- y-lactone


L-Ascorbic acid

L-Ascorbic acid

Fig. 2. Proposed routes of L-ascorbic acid biosynthesis in plants (a) and animals (b),
involving inversion of configuration; UDP-uridine diphosphate.
Studies by using radioactive o-glucose C4 C-1 of 14C_6, respectively) have
shown that the main six-carbon chain is conserved and radioactivity found at
C-1 or C-6 atoms of L-ascorbic acid. In this retention of configuration,
oxidation occurs at C-1, followed by lactonization, then again oxidation at C-2
and C-3, and finally epimerization at C-5 to give L-ascorbic acid (Loewus,
1980). Alternative biosynthetic pathways involve configurational inversion in
which the main precursor is turned to L-galactono-y-Iactone, more active than
L-gulono-y-Iactone, which is oxidized at C-2 to give L-ascorbic acid (in
o-glucose series epimerization at C-3 occurred prior to the last oxidation).
Biosynthesis of L-ascorbic acid in animals is somewhat simpler (Fig. 2). In
mammals, L-ascorbic acid is synthesized in the liver via the glucuronic acid


V. Delif, D. Sunif & D. Vlalic

pathway, likewise in plant biosynthesis. Studies with radioactive labelled

D-glucose have shown that 14C_1 (or 14C_6) became the 14C_6 (or 14C_1) atom of
L-ascorbic acid until the D-glucose chain remained intact (Bums, 1967).
After 1896, when Bertrand for the first time described microbial oxidation of
sorbitol to sorbose by the micro-organism Bacterium xylinum (subsequently
classified as Acetobacter xylinum) (Bertrand, 1896), more than 30 years passed
until this microbial reaction was used as a biooxidation step in L-ascorbic acid
synthesis. Afterwards, research in the field of microbial biosynthesis of
L-ascorbic acid intermediates became more intensive and further knowledge
accumulated quickly.
During the past 50 years there have been a lot of attempts to find ways of
microbial conversion of intermediates in L-ascorbic acid synthesis which could
compete with the Reichstein-Griissner synthesis from an economical point of
view. Different micro-organisms were screened for their ability to carry out
reactions on carbohydrates as well as methods for new biosynthetic steps in
those routes. These reactions, micro-organisms and media are listed in Table
2. L-Ascorbic acid intermediates biosynthetic routes, including micro-organisms,
are shown in Fig. 3.
In this review we present names of micro-organisms as reported in the
original articles. Since their first presentation in the literature, the taxonomy of
these micro-organisms has undergone many changes due to their morphological, cultural and biochemical properties, e.g. Acetobacter suboxydans has been
renamed Gluconobacter oxydans subsp. suboxydans. t For actual names of
micro-organisms see Bergey's Manual of Systematic Bacteriology (1984). Most
of the micro-organisms mentioned here are mutants with blocked side
reactions in 2-keto-L-gulonic acid biosynthetic routes from carbohydrates,
which could utilize alternative carbon sources.

5.1 Biooxidation of L-sorbose to 2-keto-L-gulonic acid

In the classical Reichstein-Griissner synthesis (see p.323), L-sorbose is converted to 2-keto-L-gulonic acid (2-KLG) by chemical reactions. In this part we
are describing the possibilities of L-sorbose conversion to 2-KLG by microorganisms (Fig. 3, routes 2, 3).
The first biooxidation of L-sorbose to 2-KLG by micro-organisms was
described by Tengerdy (1961a, b) who used UV-irradiated mutant strains of
Pseudomonas. After 4 days of cultivation in the medium containing 20 g/liter

t 'Names included in the Approved List of Bacterial Names are the only names which
are nomenclaturally valid as at the 1st January, 1980. All other names which have
appeared in the literature prior to 1st January, 1980 are nomenclaturally invalid'
(Skerman et af., 1980).

IFO 14464

oxydans U-13
Pseudogluconobacter saccharoketogenes

N1197 A


IFO 3293




L-Sorbose~ 2-KLG



L-Sorbose~ 2-KLG

L-Sorbose ~ 2-KLG






D-Sorbitol ~ L-sorbose
Sorbitol ~ 2-KLG


Pseudomonas sp.

IFO 3243



L-Sorbose 7; urea 05-08; K2HP04 0007; KH2P0 4

0003; glycerol 02; MgS04 x 7H20 001; CaC03 05
and com steep liquor 05%. pH 60-65. Addition of
urea or (NH4)zHP04 during fermentation
L-Sorbose 100; glycerol 05; yeast ext. 150 g/liter; salts.
Cultivation on rotary shaker, 30C, 4 days
1000 liters of the ferm. medium containing L-sorbose 15;
CaC0 3 5; com steep liquor 2; dried yeast 03; Actcol
003; (NH4)2S04 003; Na2S20 3 x 5H20 05 and
FeS04 x 7H20 01 %.

30 liters of the medium containing sorbitol 5; glucose 05;

yeast ext. 05 and CaC0 3 2%. 28-29C, 150 h,
aeration 15-24 liters/min
Glycerol 1; L-sorbose 20; com steep liquor 10; NZ amine
B 5; K2HP04 X 3H20 07; KH2P04 03; NaCl 05;
MgS04 x 7H20 01 and FeS04 x 7H20 01 g/liter. 10
liters of medium in 14-liter glass fermentors was
inoculated with 2% cell suspension from inoculum
medium. 28C; pH 7 0-72; dow silicone oil
(0 25 g/liter) was used as antifoam; 96 h; aeration 1 v/v
per min; stirred 350 rpm
KH2P04 004; MgS04 x 7H 20 002; NZ-amine B 025;
yeast ext. 05 and sorbose 2%, 28C, aeration 1 v/v per
min; pH 8,1-4 days
L-Sorbose 7; glycerol 05; yeast ext. 05; MgS04 x 7H2O
05 and CaC0 3 1%. 25C, pH65, 220 rpm, 96h

Media, conditions

Table 2
Micro-organisms and Media Used in Biosynthesis of L-Ascorbic Acid Intermediates


Fujiwara et ai.,
Nogami et ai., 1987

Yan et al., 1981

Yin et al., 1980

Tsukada &

Huang, 1962

Tengerdy, 1961a, b

Obata et aI., 1975













melanogen us

aeruginosa T-S1






Ca-L-idonate~ 2-KLG



2-ketoCa-n-gluconate ~ hexonic
(1: 1)
Ca-idonate~ 2-KLG


L-Sorbosone ~ 2-KLG


IFO 3293

L-Sorbose ~ L-sorbosone




Ca-idonate and Ca-gluconate (7: 3)

50; ~OAc 12; KH 2P0 4 05; MgS04 x 7H20 03;
FeS04 x 7H20 OS g/liter. Aeration, 6 days
Ca-L-idonate 7; Ca-n-gluconate 3; yeast ext. 25;
(NH4)2HP04 02; KH2P0 4 01 and MgS04 0025%.
28C, 225 rpm, 3 days
Glucose 05; corn steep liquor 10; KHZP0 4 01;
MgS04 x 7H20 005; Ca-5-keto-gluconate
hydrogenation product 34% (containing idonic acid
147 and gluconic acid 72%); 1 liter. 2SoC, shake
cultured, 76 h, 40 g 2-KLG formed

Glycerol 25; Na-citrate 50 g/liter; mineral salts; 2SoC,

overnight; fresh medium (95 ml) containing. Lsorbosone 10 g/liter was inoculated with 05 ml of the
old medium, 28C
Ca-L-idonate; maltose (or glucose) 3; com steep liquor 3;
KH2P04 03 and MgS04 x 7H20 01 part; 50 parts of
P. mildenbergii (grown on 5% glucose and 05%
yeast). pH 55-60
Idonic acid 250; gluconic acid 250; NH4 0Ac 12; K3P04
5; MgS04 1 g; 10 liters H 20. pH 65; 2-3 days; AcOH
added to pH 55, kept 7-S days

Sorbose 5; com steep liquor 25; MgS04 x 7H20 001;

KH2P04 003; K2HP04 007; NaCl 005 and glycerol
005%. Aeration, 2SoC, 120h
L-Sorbose 7; glycerol 05; yeast ext. 05; MgS04 x 7H2O
05; CaC03 1%. pH 68, 30C, 250 rpm, 7 days

IFO 1210S


Media, conditions
30C, 110 rpm, aeration 900 liters/min, pressure
05 kg/cm2, 4 days, 1231 mg/mI2-KLG formed




Table 2-contd.


Mochizuki et al.,

Takeda et al. ,

1963a, b

Fujisawa et al.,

Yamazaki, 1955

Gray, 1947a

Makover & Pruess,


Kitamura &
Perlman, 1975


Mochizuki et al. ,












Erwinia sp.
SHS 2629001
der. from
(ATCC 31626)

IFO 3263

Acetobacter sp.

o-glucose- Ca-2,5-DKG


Glucose- 5-ketogluconic
Glucose- 5-KDG
o-Glucose- 2,5-DKG

5-KDG- L-idonic acid

L-gulonate- 2-keto-Lgulonate

ATCC 10768

FERM-P 1905
(ATCC 21914)

Ca-L-gulonate- Ca-2-KLG


2 Liters of medium (pH 60) containing glucose 110; corn

steep liquor 05; (NH 4)zHP0 4 058; KH 2P0 4 15;
MgS04 x 7H 20 005; urea 05 g/liter; CuS0 4 x 5H2O
1; nicotinic acid 03 mg/liter. 1; nicotinic acid
o 3 mg/liter. 28C, 36 h, 1700 rpm, aeration o 75 v/v x
min. An additional 55 g glucose/liter was added at 20 h
and the pH was maintained at 55
o-Glucose 58; com steep liquor 1-13; (NH4)zHP04 O'56;
CaC03 184; p-2000 antifoam 002%; pH 68.
186 m3 /10 m3 ; 160 rpm; 36 m3 air/min; 28C; 26 h.
During fermentation, 2304 kg 50% (wt/wt) o-glucose
was added

Glucose 10-15; yeast 075-10%;

26C, 7 days
Glucose 10; yeast ext. 05; glycerol 05; MgS04 x 7H2O
05; CaC03 1%; pH70. 25-26C, 300 rpm, 3 days

Glucose 10%, 33 h

Ca-L-gulonate 120; com steep liquor 5; octadecyl ale.

03; maltose or sorbitol 5 and H 20 1000 parts; pH
60-61. 50 parts of inoculum; 25C; aerated; 8 days;
90 parts Ca-2-KLG formed
4 Liter stirred fermenter containing 2 liters of production
medium containing corn steep liquor 8; meat digest 6;
(NH4)zHP0 4 1; glucose 10; CaC0 3 3 and Ca-Lgulonate 80 g/liter. After ferm. at 28-30C for 48 h
with stirring and aeration, 40 g/liter Na-L-gulonate
ferm. broth was added followed by an addnl. 30 g/liter
at -72h
Cells of log phase were suspended in 400 ml of 01 M
phosphate buffer (pH 686) mixed with 200 ml of 6%
K-salt of 5-KDG and held at 30C for 48 h


Sonoyama et al.,

Kita & Hall,


Teramoto et al.,
Stroshane &
Perlman, 1977

Stubbs et al.,

Sonoyama et at. ,

Kita, 1979

Gray, 1947b











500 ml of medium contg. glucose 10; o-mannitol 20; com

steep liquor 10; yeast ext. 15; casamino acids 10;
K2HP04 4; KH2P04 1; ~Cl1; CaCI2 001 and
K2S04 26 g/liter with three further additions of
10 g/liter o-glucose. 28C; 800 rpm; over 60 h
Galactonic acid or galactono-y-Iactone 05; com steep
liquor 025; ~CI 01; glycine 07; MgS04 x 7H2O
005; Na-glutamate 02; EtOH 15%; trace metals;
pH 4.2. 300C, stirring, aeration, 48 h. 043 g/liter Lascorbic acid formed

o-glucose- 2-KLG

o-glucose- 2-KLG








Galactonic- L-ascorbic


o-Glucose 2; com steep liquor 3; NaN0 3 0345; KH2P04

0067; p-2000 antifoam 000167%; ZnS04 X 7H20 49;
MnCl2 x 4H20 08; thiamine hydrochloride 022; Cao-pantothenate 017 mg/liter; pH 69; 28C; 160 rpm;
118 air/min. NaN03 , o-glucose and Ca-2,5-DKG
added during fermentation
100 ml medium contg. cerelose 2; (~)2HP04 1;
KH2P04 1; MgS04 X 7H20 05; beet molasses 2 and
glycine 02 g/liter; pH 67. After 22 h 15 ml of an
Acetobacter cerinus ferm. broth contg. 15-20% 2,5DKG was added 28C; addnl. 52 h; pH 65
Glucose 3; glycerol 20; yeast ext. 5; peptone 5; CaC03
75 g/liter; pH 70. 1 g/liter 2-KLG formed

Ca-5-DKG- Ca-2-KLG

SHS 752001
der. from
SHS 0007
(ATCC 31090)


Media, conditions



Table 2-contd.

Roland et al. ,

Anderson et al. ,
Estell et al., 1985
Hardy et al. ,


Kita & Hall,

Sonoyama et ai,








Microbial Production of Vitamin C






G. m.,.nog''''










l-Gulonic aCid


f f


HO~C~H_ HOfC~O=-~~~C01













l H O ~C:O~H


2-Keto- Lgulonlc .. cid








gulonlc acid





Po mildenber9li
p tluorescens
G mel.ilnogenus



L - Idose






L - Sorbosone



L - Idonic

B '.tosocedu'tum





ILao_'d_-=:=:'n='~=~'=S~=-~= ;'='~=:__


Acetobacter cerlnus
AcetomonCiis albosesamae






2,5-0rKeto-Ogluconrc acid
Erwima cltreus
Erwinla herbrcola

Fig. 3. Biosynthesis of intermediates in L-ascorbic acid routes by different microorganisms. 1, Reichstein-Griissner route; 2, L-idose route; 3, L-sorbosone route; 4,
5-keto-o-gluconic and L-idonic acid route; 5, 2,5-diketo-o-gluconic acid route.

L-sorbose, 13 g/liter of substrate was utilized and 3 g/liter of 2-KLG formed.

Tengerdy (1961a, b) proposed the one-step reaction involving oxidation of the
C-1 of L-sorbose to 2-KLG. Another Pseudomonas mutant (P. viridiflava
NRRL B-94), which oxidized L-sorbose to 2-KLG with 50% yield, was isolated
by Huang (1962).
Different genera of eubacteria were screened for their ability to oxidize
sorbitol and L-sorbose. Isono et al. (1968) found that micro-organisms
producing keto-sugar acid from sorbitol were restricted to the genera
Acetobacter, Gluconobacter and Pseudomonas whereas those producing acids


V. Delic, D. Sunic & D. VlaIic

from L-sorbose were distri1?uted widely in the 12 genera: Acetobacter,

Alcaligenes, Aerobacter, Azotobacter, Bacillus, Escherichia, Gluconobacter,
Klebsiella, Micrococcus, Pseudomonas, Serratia and Xanthomonas. As shown
by Mochizuki et al. (1969a), P. aeruginosa was able to oxidize sorbose only to
L-idonic acid. Kanzaki & Okazaki (1970) suggested a multi-step conversion of
L-sorbose to 2-KLG by Pseudomonas aeruginosa in which L-idose and L-idonic
acid were intermediates. Pseudomonas aeruginosa IFO 3898 converted Lsorbose via L-idose and L-idonic acid to 2-KLG efficiently (Fig. 3, route 2).
There are only a few genera of micro-organisms capable of oxidizing sorbitol
to 2-KLG. Some unidentified Acetobacter and Pseudomonas strains were able
to produce 2-KLG from sorbitol (Takeda Chemical Industries, 1964). It was
reported that these strains, after 150 h of fermentation in the medium
containing 5% sorbitol, produced 2-KLG with an overall yield of 8-9%.
Acetobacter sp. IFO 3243 cultivated in 30 liters of fermentation medium
converted 4% sorbitol to 2-KLG (Obata et al., 1975).
Okazaki et al. (1968, 1969) studied the conversion of sorbitol to 2-KLG by
means of G. melanogenus IFO 3292. After 7 days of cultivation in a medium
containing 10% sorbitol, the fermentation broth was found to contain neutral
and acidic compounds which were identified' as: L-sorbose, 5-keto-o-mannonic
acid, 2-keto-o-gluconic acid, 2-KLG, L-idonic acid, o-mannonic acid, 0fructose and 5-keto-o-fructose. As a result of the accumulation of these
compounds the authors proposed the following metabolic pathway in G.
melanogenus: sorbitol--+ L-sorbose--+ L-idose--+ L-idonic acid--+ 2-KLG (Fig. 3,
route 2).
In the early 1970s, a number of publications appeared describing the
biooxidation of L-sorbose to 2-KLG by Gluconobacter meianogenus IFO 3293.
Tsukada & Perlman (1972a, b, c) found conditions for the conversion of 2% of
metabolized L-sorbose to 2-KLG and proposed the direct conversion of
L-sorbose to 2-KLG. Later on, the 'sorbosone pathway' was found to be
involved. Employing this route, L-sorbose was converted to 2-KLG via
L-sorbosone as an intermediate (Fig. 3, route 3). The first step of this reaction
was reversible and mediated, in both directions, by an enzyme(s) fraction of
Pseudomonas putida ATCC 21812 (Makover et al., 1975).
Growing cultures or cell-free extracts of G. melanogenus IFO 3293 were also
able to carry out this reaction (Kitamura & Perlman, 1975). The limiting factor
of this route was the reversibility of the first step as well as the poor absorption
of L-sorbose into the cells (Makover et al., 1975). Oxidation of L-sorbose to
L-sorbosone was efficiently improved 2-3 times by the addition of organic
solvents or detergents, due to the increased cell membrane permeability
(Martin & Perlman, 1975).
Yin et al. (1980) and Yan et al. (1981) established the optimal fermentation
medium for producing 2-KLG from L-sorbose by the bacterial strain
Gluconobacter oxydans N 1197 A. In the medium containing 7 or 10% of
L-sorbose and 05-07% of urea concentration, the yield of 2-KLG was about
30 or 37 g/liter, respectively.

Microbial Production of Vitamin C


It seems likely that researchers in the People's Republic of China are able to
produce 2-KLG from L-sorbose (10 g/liter) over L-sorbosone on a commercial
scale with a 90% yield using bacteria of the genera Gluconobacter and Bacillus
(Kieslich, 1984; Lu et al., 1985; Ning et al., 1988).
Fujiwara et al. (1987) have used the G. oxydans strain, which has a high
activity of L-sorbose dehydrogenase, to produce 2-KLG. After a 4-day
cultivation of G. oxydans U-13 in the L-sorbose medium 2-KLG was formed
with a 64% yield. Also, G. oxydans U-13 was able to produce 2-KLG during
cultivation in o-sorbitol medium.
Recently, Nogami et al. (1987) reported the capability of Pseudogluconobacter saccharoketogenes IFO 14464 (novel genus, isolated from soil
origin) and a concomitant bacterium Bacillus megaterium IFO 12108 to
produce 2-KLG in a high yield (76%) during cultivation in a 2 m 3 fermenter
containing 1000 liters of L-sorbose medium. Bacillus megaterium, which was
not directly involved in the oxidation process, promoted the growth of the
oxidative strain (P. saccharoketogenes) and increased the yield of 2-KLG. (Fig.
3). Compared with a pure culture of the oxidative strain the mixed culture was
able to oxidize higher concentrations of L-sorbose to 2-KLG in a shorter period
of time. The bacteria of some other genera could be successfully used instead
of B. megaterium, e.g. Pseudomonas, Proteus, Citrobacter, Enterobacter,
Erwinia, Xanthomonas and Flavobacterium.


Biosynthesis of 2-keto-L-gulonic acid from 5-keto-D-gluconic and Lidonic acid

Chemical reduction of 5-keto-o-gluconic acid (5-KDG) resulted in a mixture of

L-idonic and o-gluconic acid that could be converted to 2-KLG by various
micro-organisms. The first microbial biooxidation of L-idonic acid to 2-KLG
was reported by Gray (1947a), who found that Pseudomonas mildenbergii was
capable of performing this reaction (Fig. 3, route 4). In this process L-idonic
acid or its salts were prepared along with o-gluconic acid by a chemical
reduction of the corresponding salts of 5-keto-gluconic acid. L-Idonic acid was
subsequently biooxidized by P. mildenbergii to 2-KLG, while o-gluconic acid
was oxidized back to 5-keto-gluconic acid by A. suboxydans. 5-Keto-gluconic
acid was reused in the process.
Pseudomonas fluorescens oxidized well both calcium-L-idonate and calciumo-gluconate (Yamazaki & Miki, 1953; Yamazaki, 1953a, b, c). After 7-8 days
of cultivation in the medium containing the mixture of L-idonic and o-gluconic
acid (ratio 1: 1) the producing strains of genus Pseudomonas were able to
convert both of the added compounds to 2-keto-hexonic acids with a 40-50%
yield. The main problem in this process was the isolation of 2-KLG from the
mixture of 2-keto-hexonic acids obtained at the end of fermentation. 2-Ketoguion ate was isolated in a 50% yield (Yamazaki, 1953c, 1955). 5-Keto-ogluconate, used in this procedure, was obtained by the biooxidation of
o-glucose with A. suboxydans (Yamazaki, 1953a).


V. Delic, D. Sunic & D. VlaJic

With a different starting mixture of Ca-idonate and Ca-gluconate (7: 3,

respectively) and after a 6-day cultivation of P. f/uorescens, Fujisawa et al.
(1963a, b) obtained a 77% yield of 2-KLG based on L-idonic acid. Similar
results were obtained by Alieva (1966) using the same micro-organism and by
Takeda et al. (1964) using P. aeruginosa T-81 strain.
It seems likely that the ability for oxidation of L-idonic acid to 2-KLG is
widely spread in genus Pseudomonas, e.g. P. aeruginosa, P. mildenbergii, P.
chlororaphis and P. pseudomallei (Nakanishi et al., 1964a, b). Apart from
genus Pseudomonas, some other micro-organisms have also been reported to
oxidize L-idonic acid to 2-KLG (Shoemaker, 1956; Chas. Pfizer & Co., 1958:
unidentified bacteria ATCC 11867 and ATCC 11868; Mochizuki et al., 1969b:
Acetobacter melanogenus and A. suboxydans).
A new approach to L-ascorbic acid intermediate synthesis was applied by
Sonoyama et al. (1974b). Instead of chemical agents, they used the cells or a
crude enzyme preparation of Brevibacterium ketosoreductum sp. nov. FERMP 1905 derived from ATCC 21914 to reduce 5-KDG to L-idonic acid.

Biooxidation of L-gulonic to 2-keto-L-gulonic acid

A combination of chemical and biochemical reactions for the preparation of

2-KLG from L-idonic acid was described by Gray (1947b). In this combination
the first step was the chemical conversion of L-idonic to L-gulonic acid, while
the second step was the biooxidation of L-gulonic acid to 2-KLG by A.
suboxydans. After 8 days of cultivation 75% of L-gulonic acid was oxidized to
2-KLG. As shown by Perlman (1959), P. aeruginosa was able to oxidize even
L-gulono-Iactone to 2-KLG in a 44% yield. Some strains of genus
Xanthomonas were also able to oxidize L-gulonic acid to 2-KLG. For example,
Xanthomonas translucens A TCC 10768 was cultivated in the medium containing 80 g/liter calcium-L-gulonate (added twice during cultivation). In this
instance the overall yield of 2-KLG was 85% (Kita, 1979).
5.4 Biosynthesis of 2-keto-L-gulonic acid from o-glucose
Studies on o-glucose biooxidation to keto-gluconic acids began 50 years ago.
Bernhauer & Knobloch (1938, 1940) reported biosynthesis of 2-keto-gluconic
and 5-keto-gluconic acids from glucose or gluconate by micro-organisms of
genus Acetobacter (Fig. 3, route 4, 5). Stubbs et al. (1940, 1943) obtained a
90% yield of 5-keto-gluconic acid from 10% glucose in the medium with A.
suboxydans and 82% with an unidentified strain. Teramoto et al. (1946)
obtained identical yield of 5-keto-gluconic acid using an unidentified
Acetobacter strain. As shown by Razumovskaya & Vasil'eva (1956), A.
suboxydans was able to oxidize o-glucose not only to 5-keto-gluconic but also
to diketo-gluconic acid.
Advancement along this route was achieved when Katznelson et al. (1953)
established oxidation of o-glucose to 2,5-diketo-o-gluconic acid (2,5-DKG) by

Microbial Production of Vitamin C


Acetobacter melanogenus MA 6.2 without prior splitting of the o-glucose

molecule. The route they proposed was glucose- gluconic acid- 2-ketogluconic acid-2,5-DKG. The same route was studied by Wakisaka (1964a)
who reported the microbial oxidation of glucose to 2,5-DKG by Pseudomonas
sesami. Soon afterwards a new species of Pseudomonas genus (P.
albosesamae) capable of a high production of 2,5-DKG from o-glucose was
isolated (Wakisaka, 1964b, c). Acetobacter fragum (Oga et al., 1972) and A.
cerinus (Kita & Hall, 1981b) were also able to produce 2,5-DKG under
aerobic conditions in the medium containing o-glucose. As much as 95% of the
starting o-glucose was converted to 2,5-DKG by A. cerinus. Stroshane &
Perlman (1977) obtained 2,5-DKG in a 93% yield from o-glucose by A.
melanogenus ATCC 9937.
The authors proposed a metabolic pathway which included o-gluconic acid
and 5-KDG as intermediates. In this biosynthesis the main problem was the
instability of 2,5-DKG and decomposition of up to 80% during the isolation
The procedure for 2-KLG biosynthesis from o-glucose was improved by
Sonoyama et al. (1982, 1983b) who used a two-stage fermentation system (Fig.
4). They reported a process which included biosynthesis in 10 m 3 conventional
fermenters for the production of calcium-2-KLG from o-glucose over calcium2,5-DKG by two different micro-organisms. The first step in this process was
o-glucose oxidation to Ca-2,5-DKG by the mutant strain SHS 2629001 of
Erwinia sp. derived from ATCC 31626 and, subsequently, the stereospecific
reduction (to eliminate 2-keto-o-gluconic acid formation) of Ca-2,5-DKG to
Ca-2-KLG by the mutant strain SHS 752001 of Corynebacterium sp. derived
from ATCC 31090. At the end of cultivation, Erwinia sp. was treated with
sodium dodecyl sulfate to decrease the number of viable cells so the broth with
accumulated Ca-2,5-DKG could be used directly for the next reduction,
mediated by Corynebacterium sp. Thus, in this process, the isolation of
Ca-2,5-DKG was avoided. The total fermentation time for this two-stage
process was about 100 hand Ca-2-KLG accumulated up to 1063 mg/ml
(846% yield from o-glucose). Media used in this process are listed in Table 3.
Compared to some other processes previously described in this review, a
two-stage fermentation procedure for 2-KLG production has some advantages
that make it a possible candidate for industrial application. These are: (a) high
producing strains; (b) elimination of 2,5-DKG isolation from the broth; (c)
commercially feasible media for both biosyntheses; (d) reliability and reproducibility and (e) a simplified process for 2-KLG isolation (absence of
compounds like o-glucose, Ca-L-idonate, Ca-o-gluconate, Ca-5-KDG, Ca-2KDG).
Results obtained by Sonoyama and his co-workers on this subject have been
very well documented in a number of publications on the mixed cultures of
Acetomonas albosesamae ATCC 21998 and Brevibacterium sp. ATCC 31083,
as well as on the screening, isolation and description of micro-organisms
capable of the oxidation of glucose to 2,5-DKG (Erwinia species) and

V. Delic, D. Sunic & D. VlaJic


11~ ~~';7ri."IP'
(ATCC 31(90)

l28"'C. .ah


:.:.:: :::.::::

.....:.. :

28'C. lOh. 275 rpm

10.600 ml 0/ pn:seed medium in a


28"C. 2' h. 215 rpm

.20 ,<> of the

seed tncdium in
l ~ml rermentor

28"C. 10 h. z.w rpm;

0,(2 ml .ir/min.

1860 tilfa of 1M

ferment. medium in
100mJ {ermen.or
inoculated with the
4DO I:llrCl oIlhc:
fennent. medium in
IO-m 3 fennenlor
inoculated with the
entire content of
tl'M: JCCd fennenu~:t

cnlirc conterU of
the seed fcrmc:nlc:r

28"C' 26 h. ,60 rpm;


3.6 m3 Iit/m.!n. (&Ilk

prcuure 1.5 .... /an 2;

tdd.ition of D1lucosc
28"C. 18h . '60 rpm;
1-18 m1 lir/min:
NINO) odQ"j


4200 Ijll'd ollhc:


as a hydrogen

fermen ted broth:


Add.itJon of SOS. after 6 I'll


the IQlal viabk; cell

populldon docrc.&$Cd
!Torn 5 109 t. 9.

7430 litra of tbe
fermented broth;

U16 m& Ca-2-KLGlmi

Fig. 4. Flow-sheet for the production of 2-keto-L-gulonic acid by two-stage fermentation; SDS, sodium dodecylsulfate; Ca-2,5-DKG, calcium-2,5-diketo-o-gluconate; Ca-2KLG, calcium-2-keto-L-gulonate.

Microbial Production of Vitamin C


Table 3
Media Used in the Biosynthesis of Calcium 2,S-Diketo-o-Gluconate and
Calcium 2-Keto-L-Gulonate by Erwinia sp. SHS 2629001 and
Corynebacterium sp. SHS 725001


Erwinia sp.

Agar slant
yeast extract
KH2 P04
MgS04 x 7H2 O
Preseed medium
yeast extract
com steep liquor
NaN0 3
KH2 P0 4
MgS04 x7H2 O
p-200 antifoam
Seed medium
com steep liquor
NaN0 3
KH2 P04
MgS04 x7H 2 O
p-2000 antifoam
Fermentation medium
com steep liquor
NaN0 3
KH2 P0 4
p-2000 antifoam
ZnS04x 7H2 O
MnCl2 x 4H2 O
thiamine hydrochloride
Ca-0- Pantothenate


Corynebacterium sp.
49 mg/liter
08 mg/liter
022 mg/liter
017 mg/liter


V. De/if, D.

Sunil &

D. Vlalif

reduction of 2,5-DKG to 2-KLG, like Brevibacterium ketosoreductum, Bacillus

megaterium, Arthrobacter simplex, Staphylococcus aureus, Micrococcus,
Pseudomonas, etc. (Sonoyama et al., 1974a, 1976, 1983a, 1986).
Mutants of Corynebacterium sp. ATCC 31088, 31089 and 31090 with
improved bioconversion abilities have been selected after N-methyl-N'-nitroN-nitrosoguanidine mutagenic treatment. These mutants were blocked in
degradative metabolic pathways of 2,5-DKG, reducing 2,5-DKG only to
2-KLG in a high molar yield with the addition of glucose as a hydrogen donor
(Sonoyama et al., 1987a, b).
2-KLG is the end product of all described biosynthetic routes starting from
D-glucose, L-sorbose or L-gulonic acid, respectively. This is the ultimate
intermediate in L-ascorbic acid biosynthesis by micro-organisms (or chemicals).
Subsequently, 2-KLG could easily be converted by chemical methylation and
cyclization to L-ascorbic acid.

5.5 Direct Biosynthesis of L-Ascorbic Add

Direct biosynthesis by micro-organisms is the most challenging means of
L-ascorbic acid synthesis. Until now, there have been only few articles
describing micro-organisms and their ability for direct biosynthesis of Lascorbic acid from different carbohydrates. Various groups of organisms were
found to be able to produce L-ascorbic acid from different sugar substrates but
concentrations of substrates and yields of L-ascorbic acid are still very low.
Geiger-Huber & Galli (1945) found that Aspergillus niger was able to
synthesize L-ascorbic acid (15 mg/lOO ml) after 6-8 days of cultivation in the
medium containing 15% sucrose. Sucrose was also a suitable substrate for
L-ascorbic acid biosynthesis when Aspergillus flavus was used (Ramakrishnan
& Desai, 1956). Sastry & Sarma (1957) cultivated A. niger in glucose medium.
They proposed the following metabolic pathway as follows: D-glucoseD-glucuronic acid- L-gulono-Iactone- 2-keto-L-gulonolactone- L-ascorbic
Recently, Roland et al. (1985, 1986) reported that strains of the yeast
Candida (e.g. C. norvegensis KCC MF 42) were able to produce L-ascorbic
acid during cultivation in the medium containing L-galactonic substrates, such
as L-galactonic acid, L-galactonic acid esters or L-galactono-y-Iactone. According to these results it seems likely that biosynthetic pathways in some yeasts
and moulds are similar to those proposed for plants (see Fig. 2).
Some green algae also displayed the ability to produce L-ascorbic acid.
Chlorella pyrenoidosa cultivated on a glucose medium for 101 h produced
15 g/liter of L-ascorbic acid (Skatrud & Huss, 1987).

5.6 Biocatalysts Involved in Biosynthesis of L-Ascorbic Add Intennediates

Recently, new methods applying biological catalysts-biocatalysts have been
developed for conversion of different raw materials to desired end-products.

Microbial Production of Vitamin C


These new technologies and processes significantly reduce the cost of biosynthetic production (Aiba et al., 1973; Rehm & Reed, 1981; Karube et al.,
1984; Kulbe & Knopki, 1986).
These methods rely on preparation of biocatalysts such as immobilized
enzyme(s), parts of cells or whole cells (Klein & Wagner, 1978; Scott, 1987;
Brodelius and Vandamme, 1987). The biocatalysts can be used for the
production of various substances in a single-step enzyme reaction (6aminopenicillanic acid from penicillin G; fructose from glucose, etc.). The
application of biocatalysts for multi-step enzyme biosynthesis (for instance,
antibiotics) is rudimentally described in the literature (Morikawa et al., 1979,
1980; Vandamme, 1984). This interesting area represents a new type of
biosynthesis, and as such this approach to L-ascorbic acid synthesis should not
be neglected in the future.
Martin & Perlman (1976) described conversion of L-sorbose to L-sorbosone
by biocatalysts. They used immobilized cells of native Gluconobacter melanogenus IFO 3293 entrapped in polyacrylamide gel. Authors concluded that
conversion depended on the concentration of monomer used. The higher the
monomer concentration, the more L-sorbosone accumulated. Different results
were found with immobilized lyophilized cells. In this case a very active
preparation of biocatalysts was obtained with low monomer concentration due
to the impaired permeability of lyophilized cells, allowing diffusion of the toxic
monomers into the cytoplasm. Fujiwara et al. (1987a, b) described isolation,
purification and characterization of two enzymes (L-sorbose dehydrogenase
and L-sorbosone dehydrogenase) from Gluconobacter oxydans UV-10 involved
in conversion of L-sorbose to 2-KLG via L-sorbosone.
The cells of Gluconobacter oxydans strain No.6 VNIVI capable of oxidizing
o-sorbitol to L-sorbose were entrapped in polyacrylamide gel. The highest
oxidative activity of such a biocatalyst, reused five times, was 33 g of L-sorbose
per liter per h (Pomortseva & Krasil'nikova, 1983).
The production of L-sorbose by cells of Gluconobacter suboxydans (ATCC
621) immobilized in polyacrylamide gel was carried out in a continuous
process. The entrapped cells of this strain were capable of almost completely
converting o-sorbitol into L-sorbose at a rate of about 7 kg/m3 per h during a
long period of time (Stefanova et al., 1987).
Bailey et al. (1985) developed a mathematical model on K-carrageenanimmobilized cells of Acetobacter suboxydans oxidizing ethyl alcohol to acetic
acid. This model, which predicted the cells, substrate and product concentration in reaction with the biocatalyst, could possibly be used for biooxidative
studies in the L-ascorbic acid pathway.
Studies on enzyme participation in the pathways of o-sorbitol oxidation were
done by Shinaga-,a et al. (1982). They described the solubilization, purification
and characteristics of o-sorbitol dehydrogenase from the membrane fraction of
Gluconobacter suboxydans var. a IFO 3254.
Recently, Sonoyama & Kobayashi (1987) reported two 2,5-DKG reductases
(reductase I and II) isolated from Corynebacterium sp. SHS 752001 mutant


V. De/ie, D. Sunie & D. V/aIiC

derived from ATCC 31090. Reductase I differed from reductase II in its

molecular weight, its higher specific activity and its affinity for substrate
2,5-DKG. Reductase II is essentially the same enzyme described by Anderson
et aZ. (1985) and Miller et aZ. (1987). Both reductases had the ability for the
stereospecific reduction of the keto group of 2,5-DKG at the C-5 position and
were NADPH-dependent.
Molecular biology techniques developed in the last 15 years and the possibility
of gene manipulation in vitro (recombinant DNA technologies-rDNA),
particularly in microbial cells, resulted in the creation of hybrid features and
the non-conventional production of several substances in bioreactors. Gene
cloning was extended to many different micro-organisms, showing diversity of
application in a wide range of products like hormones, vaccines, enzymes,
amino acids, vitamins, antibiotics, etc. (Old & Primrose, 1985; Primrose,
1987). The classical example is cloning of genes for human insulin in the
bacterium Escherichia coli (Goeddel et aZ., 1979) and the production of this
hormone by fermentation (Frank & Chance, 1983). These techniques were
also usefully applied in the construction of improved microbial strains for
producing penicillin G-acylase, an enzyme involved in the hydrolysis of
penicillin G to 6-aminopenicillanic acid (Mayer et al., 1979; Vandamme, 1984;
Garcia & Buesa, 1986).

Cloning of 2,S-Diketo-D-Gluconic Acid Reductase Gene in Genus


As described in previous parts of this review, many micro-organisms have the

ability to catalyse a reaction on a sugar molecule and to carry out a conversion
on the L-ascorbic acid pathway. The shortest route for o-glucose conversion to
2-keto-L-gulonic acid (2-KLG) is the two-step reaction via 2,5-diketo-ogluconic acid (2,5-DKG) (Fig. 3, route 5). In this route, o-glucose is oxidized
by microbial dehydrogenases to give 2,5-DKG. On the other hand, a group
of-for example-coryneforming bacteria was able to reduce enzymatically
(2,5-DKG reductase) 2,5-DKG to 2-KLG, the last intermediate in L-ascorbic
acid synthesis that could be easily chemically converted into L-ascorbic acid.
Sonoyama et al. (1982) demonstrated a two-stage fermentation using mutants
of bacteria Erwinia sp. and Corynebacterium sp. to convert o-glucose to
Based on this data, the idea of Anderson et al. (1985) was to combine the
traits of these two micro-organisms in a single cell by gene manipulation
techniques and to simplify the conversion of o-glucose to 2-KLG. The strategy
was to clone the 2,5-DKG reductase gene from Corynebacterium sp. ATCC

Microbial Production of Vitamin C


31090 in Erwinia herbicola ATCC 21988 (also known as Acetomonas

albosesamae) which was able to oxidize D-glucose into 2,S-DKG. Anderson et
isolated and characterized 2,S-DKG reductase from
Corynebacterium sp. and through Edman degradation determined a sequence
of 40 amino acids from the NH2 -terminus of the enzyme. Two synthetic DNA
probes, 43 nucleotides long and corresponding to the NH2 -terminal amino acid
sequence were used to detect by hybridization the 2,S-DKG reductase gene
between fragments obtained by Bam HI digestion of total Corynebacterium sp.
DNA. Fragments ranging in size from 20 to 2S kb were isolated and cloned in
a Bam HI site of plasmid pBR 322. A partial genomic library was obtained in
E. coli and the transformed colonies were checked for the presence of
2,S-DKG reductase gene by hybridization. A fragment was subcloned in
bacteriophage M13mp9 cut by a combination of two restriction enzymes Pst I
and Sma I for further vector construction. A simplified route for suitable
vector construction is given in Fig. S. For good 2,S-DKG reductase gene
expression in bacterium Erwinia herbicola, deletion was constructed upstream
of the ATG start codon. A 'strong' tryptophan (trp) promoter linked to the
Shine-Dalgarno sequence (Shine & Dalgarno, 1974) from plasmid pHGH2071ptrpL\RIS' (De Boer et al., 1983) was attached to the 2,S-DKG reductase
coding region instead of the deleted seque~ce. This was accomplished by using
deletion primer and Alu I fragments of M13mp9 treated with Klenow fragment
of DNA polymerase I, T4 DNA ligase and deoxynucleotide triphosphates
(dNTPs) to give heteroduplex mit12 RF (replicative form) molecule used to
transform E. coli JM101. Deletion of the upstream region between Nco I
restriction site and A TG starting codon was selected by plaque hybridization
(A). A replicative form of one isolate mit12L\ having desired deletion was used
to excise 2,S-DKG reductase gene by Eco RI and Hind III digestion. A new
hybrid vector was constructed from three different parts (B). 2,S-DKG
reductase gene (lS kb) from mit12L\ was mixed with a small fragment (10 kb)
of plasmid pHGH207-1ptrpL\R1S' digested with Pst I and Eco RI. This small
fragment contained tryptophan promoter region essential for 2,S-DKG reductase gene expression. The third part of the hybrid vector was a larger fragment
(36 kb) of plasmid pBR322 obtained after digestion with Hind III and Pst I.
Three parts were mixed and ligated and this mixture was used to transform E.
coli MM294. Transformants were selected for ampicillin resistance. Hybrid
plasmid pmit12L\ trp1 AmpR could not be used for the selection of transformed
cells of Erwinia herbicola since this micro-organism is naturally resistant to
ampicillin. Thus, a functional gene for tetracycline resistance was reconstructed, since after pBR322 digestion with Hind III-Pst I loss of tetracycline resistance occurred. For this purpose plasmid pBR322 was digested with
Eco RI and Sph I to generate the fragment of 0S6 kb containing the
tetracycline promoter region. This fragment was ligated with a S4 kb fragment
from plasmid pmit 12L\/trp1 AmpR isolated from E. coli MM294 and digested
with Hind III and Sph I (C). The resulting 2,S-DKG reductase expression
vector ptrpl-3S was used to transform Erwinia herbicola and transformants


V. Delic, D. Sunic & D. VlaIic

2.5- DKG red. gene

Annealing of deletion primer

and Alul f ra9m~nt$, DNA
pOlymerase I (Klenow), dNTPs.
T4 DNA ligase


Transfectron ot E. colt JMIOI
w ith heterodupr;;-;;~ ' 2RF

mil 12 6



Screening for phage

containing deletion






pBR 322




3 6kb

-----J,I ~


select ion for AmpR (olonies



5 .6kb~


/ o.56kb


Transfo,.mat ion of "",,!!,!Ill>1!....!lS=~. selection for Tet R cOlOnies

Erw ini. herblcola w ith

2, 5- DKG reductase gene on

pl.smid ptrpl- 35

Fig. 5. Construction of 2,5-diketo-D-gluconic acid reductase cloning vector and

transformation of Erwinia herbicola. Only relevant restrictipn sites are given; AmpR,
ampicillin resistance; Tet, R tetracycline resistance; trp, tryptophan promoter; kb, kilo
base; dNTPs, deoxynucleotide tryphosphates.

were selected on tetracycline. Isolated transformants, checked for enzyme

activity, showed 54 times higher 2,5-DKG reductase activity in crude extracts
in comparison with E. herbicola containing plasmid pBR322. Activity was
50-100 times higher in E. herbicola than in E. coli MM294 when transformed
with the same ptrpl-35 plasmid.

Microbial Production of Vitamin C


For the production of 2-KLG, E. herbicola with plasmid ptrp1-35 was

cultivated in a o-glucose-containing medium and subsequently on the medium
containing 2% glycerol. After additional incubation, 1 g/liter 2-KLG was
found and analysed by HPLC. Identity of 2-KLG was confirmed by the usual
analytical methods (Lazarus & Seymour, 1986).
A similar approach has recently been applied by Hardy et al. (1987) and
Grindley et al. (1988) for the construction of 2,5-DKG reductase expression
vector. Fragments containing 2,5-DKG reductase sequence were taken from
Corynebacterium sp. SHS 7520001 after Sau 3A digestion and inserted into
plasmid pUCS digested by Bam HI restriction enzyme. Recombinant plasmid
was used for the transformation of Escherichia coli JM83 and colonies of the
resulting library were screened by hybridization with the 17-mer DNA
synthetic probe to locate the 2,5-DKG reductase gene. Two recombinant
plasmids, pCBR10 and pCBR13, were found to contain 2,5-DKG reductase
gene approximately 30 kb long. These plasmids were used for further vector
construction using different promoters (lac, trp, tac, the operator and
promoter regions of phage A, the control region of fd coat protein, the
promoters of yeast, etc.). Plasmid pPLred332 containing APL promoter region
upstream of 2,5-DKG reductase region from pCBR13 was used to transform
the mutant strain Erwinia citreus ER 1026 obtained from E. citreus SHS 2003
(ATCC 31623). The newly-formed E. citreus ER 1116 strain was able to
convert o-glucose in a single biosynthetic step to 2-KLG after 65 h of
cultivation. o-Glucose was added three times during biosynthesis and at the
end 198 g/liter of 2-KLG was found (overall yield was 494%).
6.2 Molecular Biology of Other Micro-organisms Involved in 2-Keto-L-Gulonic
Add Biosynthesis
Members of the genus Acetobacter and Gluconobacter are industrially important because of their ability for high yield conversion of carbohydrates to
various oxidation products. In recent years more data on the molecular biology
and genetics of these micro-organisms have been accumulated.
Fukaya et al. (1985a) showed that as much as 818% of Acetobacter strains
harboured mainly multicopy plasmids of low molecular weight while 638% of
Gluconobacter strains harboured plasmids ranging from 10 to 17 Md.
The transformation method for Acetobacter strains was improved by
polyethylene glycol and treatment of cells by divalent cations (Ca2 +, S~+,
Ba2+, Mn2 +). Efficiency of transformation was HP transformants per I-'g DNA
(Fukaya et al., 1985b).
Construction of shuttle vectors for Gluconobacter suboxydans and
Escherichia coli was a further step in the application of recombinant DNA
technology for this group of micro-organisms, although efficiency of transformation was low (1ij2 per I-'g DNA) (Fukaya et al., 1985c).
Two phages isolated from 54 different strains of Gluconobacter (Robakis et
al., 1985a) were characterized and discovered to be unusually large, containing


V. Delit, D. Sunil & D. Vlaiil

double stranded DNA. Construction of a restriction map of bacteriophage A-I

genome revealed circular DNA containing cohesive ends (Robakis et al.,
1985b). Palleroni (1986) showed that Gluconobacter oxydans IFO 3293 could
accept plasmid RSF 1010 both by transformation and conjugation but the
yields of transformants were low, and obviously could be improved.
A new approach was used by Kahn and Manning (1988) to create
Gluconobacter oxydans strains capable of higher production of 2-KLG from
L-sorbose. They carried out additional mutation in G. oxydans UV-lO mutant
strain by using insertional transposon mutagenesis (PI:: Tn5). Mutant strain
M23-15 produced more 2-KLG (333 g/liter) than the parent strain
(196 g/liter) due to reduced 2-KLG-reductase activity. As a consequence of
this reduction, decreased flux of 2-KLG to L-idonic acid occurred. Besides this,
new recombinant strains of G. oxydans IFO 3293 were obtained by using
plasmid vectors constructed of broad host range plasmid RSF 1010 and a part
of chromosomal DNA of the wild type strain containing L-sorbose dehydrogenase DNA sequence. One of these recombinant strains, obtained by
conjugative transfer with constructed plasmid pURU010 into the parent strain
(18 glliter of 2-KLG) showed much higher conversion of L-sorbose to 2-KLG
(28 g/liter).
Some species of the genus Pseudomonas, particularly P. aeruginosa and P.
putida, are well characterized genetically and biochemically. This includes the
development of host-vector systems for gene cloning by using restrictionnegative host strains of P. aeruginosa and P. putida, and the construction of
suitable plasmids or bacteriophage DNA as vectors (Bagdasarian & Timmis,
1982; Holloway, 1986).
Broad host range plasmid RSF 1010 is one of the possible candidates for a
cloning vector in Pseudomonas cells. Some derivatives of this plasmid (pKT
230; pKT 231) with stable maintenance and propagation in host strains, have
several unique cleavage sites, and insertional inactivation inside antibiotic
resistance determinants is possible. Recently, new cloning vectors were
constructed by inserting the entire sequence of plasmid pUC 13 into
derivatives of the RSF 1010 plasmid. These new vectors, named pWS, were
stable both in E. coli and Pseudomonas putida (Werneke et al., 1985). With
regard to the cloning genes of Pseudomonas, P. putida possesses a potent gene
engineering system, even more promising than E. coli.
These findings in those industrially important micro-organisms, which have
the capacity for oxidation products from carbohydrates and polyalcohols, have
opened up further possibilities for the creation of new routes for L-ascorbic
acid biosynthesis by recombinant DNA technology.
L-Ascorbic acid was the first vitamin to be produced in commercial quantities
(Table 1). It is produced in a very large, integrated and automated process

Microbial Production of Vitamin C


involving both batch and continuous operations. Modern industrial production

of L-ascorbic acid is based on Reichstein-Griissner synthesis established more
than 50 years ago. According to available information, it seems that most
L-ascorbic acid producers use this procedure for industrial production.


The Reichstein-Griissner synthesis

The first chemical synthesis was reported by Reichstein et al. (1933; 1934).
L-Ascorbic acid was synthetized from L-xylose over L-xylosone but this route
was never commercialized. It is interesting that when Reichstein and coworkers published the first synthesis of L-ascorbic acid, the correct structure
was not yet known.
The most important synthesis of L-ascorbic acid for the manufacturing
processes is that reported by Reichstein & Griissner (1934), well known in
literature as Reichstein's second synthesis (Fig. 3, route 1).
This synthesis from o-glucose included the inversion of the glucose chain,
which means the rearrangement of the molecule where C-1 of o-glucose
became C-6 of L-ascorbic acid. The main steps of the Reichstein-Griissner
synthesis were: reduction at C-1 (o-glucose- o-sorbitol) followed by microbial
biooxidation at C-5 (o-sorbitol- L-sorbose), acetone treatment (L-sorboseL-sorbose diacetone), oxidation at C-6 (L-sorbose diacetone- 2-keto-L-gulonic
acid (2-KLG) diacetone), hydrolysis (2-KLG diacetone- 2-KLG), esterification (2-KLG- 2-KLG methyl ester) and lactonization (2-KLG methyl
ester- L-ascorbic acid). o-Sorbitol was oxidized to L-sorbose with Acetobacter
xylinum in a 60% yield. Later on A. suboxydans became the micro-organism
of choice due to its high effective oxidative ability. The modified ReichsteinGriissner synthesis is a very efficient process for L-ascorbic acid synthesis
and served as the basis for modern industrial production.
From an industrial point of view, the Reichstein-Griissner synthesis has two
crucial features. The first is low cost of raw material, e.g. o-glucose which is a
starting compound, and second, fully protected intermediate L-sorbose diacetone for the second oxidation which precludes other side reactions.
Many chemical and technological modifications have improved the
efficiencies of each step. This multi-step chemical synthesis, which includes
micro-organisms for oxidation in one step, remains till now the principal and
the most economical process for industrial production of L-ascorbic acid. The
process steps including microbial oxydation are outlined in Fig. 6.
o-Sorbitol obtained by the hydrogenation of o-glucose was oxidized by
Acetobacter suboxydans into the key intermediate L-sorbose. L-Sorbose is then
condensed with acetone to form sorbose diacetone which is oxidized to a
diacetone derivative of 2-keto-L-gulonic acid (2-KLG). After hydrolysis and
esterification 2-KLG diacetone gives 2-KLG methyl ester which is converted to
L-ascorbic acid by cyclization. The procedure requires about 17 kg of
L-sorbose per kg of L-ascorbic acid with an overall yield of 66% (Jaffe, 1984).


V. Delil, D. Sunil & D. V[alil




Acetob~cter Suboxy'd~ns

Filtered air





Fig. 6. Synthesis of L-ascorbic acid including biooxidation of D-sorbitol to L-sorbose by

Acetobacter suboxydans.

7.2 Product recovery and purification

Some steps of industrial processing include generally sophisticated product
purification and recovery.
L-Sorbose isolation from cultivation broth is related with demands for the
separation of A. suboxydans biomass accumulated in the course of n-sorbitol
fermentation and elimination of other accompanying impurities. The product
purification and recovery of L-sorbose as a solid is a continuous process which
includes not only filtration, evaporation and crystallization steps but also
purification in anion and cation ion exchange columns.
The acetone treatment of L-sorbose results in a mixture of mono- and
diacetone-L-sorbose. Because benzene readily extracts diacetone-L-sorbose,
diacetone-L-sorbose can be separated in a benzene/water system, while the
monoacetone-L-sorbose in the aqueous layer is recovered and recycled.
2-Keto-L-gulonic methyl ester is recovered as a crystalline product. After the

Microbial Production of Vitamin C


formation of 2-KLG methyl ester crystals, mother liquors and methanol

washes are accumulated for a second recovery of ester crystals.
At the end of the Reichstein-Griissner process crude L-ascorbic acid is
filtered and purified by recrystallization. The pure product is isolated and
dried. Mother liquors are accumulated and recycled for L-ascorbic acid
recovery (Jaffe, 1984).


Since L-ascorbic acid is widely distributed in different plant and animal tissues,
as well as in crystalline form added to a variety of products (e.g. in food,
drinks, etc.), the determination and quantification of its content deserve a
special analytical approach.
Apart from these difficulties L-ascorbic acid is easily subjected to oxidation
by moderate oxidants, thus yielding dehydroascorbic acid which frequently
interferes with L-ascorbic acid determination.
As analytical details on L-ascorbic acid determination are not within the scope
of this review, readers interested in this field of L-ascorbic acid are recommended
to a very pertinent review by Pelletier (1985). The main analytical methods will
be mentioned here only briefly.
Bioassay as a method was not frequently used for L-ascorbic acid determination. This method based on guinea-pig protection against scurvy, was further
applied to determine if certain L-ascorbic acid derivatives or other compounds
exhibited anti-scurvy activity.
Most determinations are based on colorimetric methods using blue dye
2,6-dichloroindophenol (Hughes, 1983). In the presence of L-ascorbic acid, dye
is reduced, resulting in colourless or pink colour solution. Direct titration of
L-ascorbic acid with 2,6-dichloroindophenol is at present the most frequently
used oxido-reduction method applicable to diverse types of samples yielding
reproducible results.
Murty & Rao (1979) used iodine salts for the determination of L-ascorbic
acid. In this case different dye indicators could be used (naphthol blue black,
Amaranth, Brilliant Ponceau 5R) in titration mixture, liberating iodine which
in the presence of starch yielded a blue colour. Reduction of ferric, cupric and
mercuric ions by L-ascorbic acid has been the basis of simple and sensitive
methods for measuring L-ascorbic acid. Corresponding ferrous, cuprous or
mercurous coloured complexes with different substances were used as a
measure of L-ascorbic acid.
Pelletier & Brassard (1977) developed a method for the determination of the
total content of L-ascorbic acid, dehydroascorbic acid and diketogluconic acid
in the samples on the basis of coupling with 2,4-dinitrophenyl hydrazine.
Gas-liquid chromatography of trimethylsilyl L-ascorbic acid derivatives was
also adapted for quantitative determination of L-ascorbic acid in different
samples (Schlack, 1974).

V. De/ie, D. Sunie & D. V/aIie


High performance liquid chromatography (HPLC) seems to be the most

precise method for the determination of L-ascorbic acid in samples of different
origins (Geigert et al., 1981; Rose & Nahrwold, 1981) although it remains to
be established which method is the most accurate and precise for general
Recently developed peroxidase-based colorimetric determination of Lascorbic acid has certain advantages over previously mentioned analytical
methods (Thompson, 1987). Determination is based on the interference of
L-ascorbic acid with the horseradish enzyme peroxidase which causes a
decrease in the rate of oxidative coupling of chromogenic reagents like
4-aminoantipyrine or 3,5-dichloro-2-hydroxy-phenylsulphonate. The resulting
change in absorbance at 510 nm is related to the concentration of L-ascorbic


Pharmaceutical applications represent the largest segment of the L-ascorbic

acid market. L-Ascorbic acid is not used only for the prevention and treatment
of scurvy, which is of course its primary application, but also for the
prevention and treatment of other pathological conditions and maintenance of
good health in general. It can be used for the prevention and treatment of
certain types of anaemia, some cardiovascular diseases, wound repair and
normal healing processes. Due to its stimulative role toward the immune
system, L-ascorbic acid is used in the prevention and treatment of various
infections such as common colds and others.
L-Ascorbic acid has gained extensive application in the food industry due to
Table 4
Main World Producers of L-Ascorbic Acida


Hoffmann-La Roche
Takeda Chemicals Ltd

City, country

Grenzach, FRG
Belvedere, New Jersey, USA
DaIry, UK
Osaka, Japan
Darmstadt, FRG
Grenaa, Denmark
Croton, Connecticut, USA
Zagreb, Yugoslavia


a Total world production (including Eastern countries, China, India and

Brasil) is assumed to be 65 000-70 000 t/year.
b According to available information Pfizer's plant in Croton has been

Microbial Production of Vitamin C


its antioxidant properties. It is used as an antioxidant in the commercial

preparation of oils, fats, meats, beer, soft .drinks, milk products and canned
and frozen food. It is also used as a dough and flour conditioner in the flour
industry. L-Ascorbic acid is applicable in photography as a developing agent,
and in metallurgy as a reducing agent (Jaffe, 1984).
At the present time, L-ascorbic acid is one of the most used and the most
widely produced vitamins of importance in the pharmaceutical, food, cosmetic
and other industries. The largest world producers are listed in Table 4
(Ullmans, 1983; Kieslich, 1984).
Although no alternative chemical route has reached the economical importance of Reichstein's second synthesis, current developments of chemical
synthesis and biosynthesis by micro-organisms may provide the important basis
in the future of L-ascorbic acid production. Many future views are focused on
the direct biosynthesis of 2-KLG from o-sorbitol or o-glucose by geneticallymanipulated organisms.
Nevertheless, the main effort in the future could be designated as the
substitution of the chemical pathway for L-ascorbic acid synthesis with a
stereospecific bioconversion in one single step and continuous operation.
The authors gratefully acknowledge the expert secretarial assistance of Miss
Jasenka Korajlija during the preparation of the manuscript, and thank Dr
Ivanka Pavusek for suggestions and reading of the manuscript.
Aiba, S., Humprey, A. E. & Milles, N. F. (1973). Biochemical Engineering. Academic
Press, New York and London.
Alieva, R. M. (1966). Formation of 2-keto-L-gulonic acid by resting cells of
Pseudomonas fluorescens. Tr. Petergofsk. Bioi. Inst., Leningr. Gos. Univ., 19,
Anderson, S., Marks, C. B., Lazarus, R., Miller, J., Stafford, K., Seymour, J., Light,
D., Rastetter, W. & Estell, D. (1985). Production of 2-keto-L-gulonate an
intermediate in L-ascorbate synthesis, by a genetically modified Erwinia herbicola.
Science, 230, 144-9.
Bagdasarian, M. & Timmis, K. N. (1982). Host: vector systems for gene cloning in
Pseudomonas. In Current Topics in Microbiology and Immunology, Vol. 96, ed. P.
H. Hofschneider & W. Goebel. Springer, Berlin, Heidelberg, New York, pp.


V. Delit, D. Sunit & D. VlaJit

Bailey, K. M., Venkatasubramanian, K. & Karkare, S. B. (1985). Immobilized live cell

reactor dynamics following dilution rate shift to growth conditions: cell synchrony
effects. Biotechnol. Bioeng., 27, 1208-13.
Bergey's ManuaL of Systematic BacterioLogy (1984). Vol. 1, ed. N. R. Krieg & J. G.
Holt. The Williams & Wilkins Co., Baltimore.
Bernhauer, K. & Knobloch, H. (1938). Decomposition of glucose by Acetobacter
suboxydans. Naturwissenschaften, 26,819.
Bernhauer, K. & Knobloch, H. (1940). Oxidations with acetic acid bacteria. VI
Comparative study on the formation of reducing-sugar carboxylic acids and the
preparation of 2-keto-gluconic acid. Biochem. z., 303, 308-15.
Bertrand, G. (1896). Preparation biochimique du sorbose. C.R. Hebd. Stanco Acad.
Sci., Paris, 122, 900-3.
Billman, J. H., Sojka, S. A. & Taylor, P. R. (1972). Investigations of keto-enol
tautomerism by carbon-13 nuclear magnetic resonance spectroscopy. 1. Chem. Soc.,
Perkin Trans., 2, pp. 2034-5.
Brodelius, P. & Vandamme, E. J. (1987). Immobilized cell systems. In BiotechnoLogy,
Vol. 7A (Enzyme TechnoLogy), 8, ed. J. F. Kennedy. VCH, Weinheim, pp.
Burns, J. J. (1967). Ascorbic acid. In Metabolic Pathways, Vol. 1, ed. D. M.
Greenberg. Academic Press, New York, pp. 394-411.
Chas. Pfizer & Co. (1958). 2-0xo-L-gulonic acid. British Patent No. 800 634.
Crawford, T. C. & Crawford, S. A (1980). Synthesis of L-ascorbic acid. Adv.
Carbohydr. Chem. Biochem., 37,79-155.
De Boer, H. A., Comstock, L. J. & Vasser, M. (1983). The tac promoter: a functional
hybrid derived from the trp and lac promoters. Proc. Natl. Acad. Sci. U.S.A., 80,
Estell, D. A, Light, D. R., Rastetter, W. H., Lazarus, R. A. & Miller, J. V. (1985).
Biosynthetic 2,5-diketogluconic acid reductase recombinant cells and expression
vectors for its production, and its use in preparing 2-keto-L-gulonic acid. European
Patent No. 132308.
Frank, B. H. & Chance, R. E. (1983). Two routes for producing human insulin utilizing
recombinant DNA technology. Much. med. Wschr., 125 (Suppl. 1), 14-20.
Fujisawa, T., Yamazaki, M., Nishiyama, K., Shimizu, K. & Sawai, H. (1963a).
2-Ketogulonic acid. Japan Patent No. 637725.
Fujisawa, T., Yamazaki, M., Nishiyama, K., Shimizu, K. & Sawai, H. (1963b).
Purification of 2-ketogulonic acid. Japan Patent No. 638714.
Fujiwara, A., Hoshino, T. & Sugisawa, T. (1987a). Enzym und Verfahren zu seiner
Herstellung. European Patent No. 248 400.
Fujiwara, A, Hoshino, T. & Sugisawa, T. (1987b). Purification and characterization of
L-sorbosone dehydrogenase of GLuconobacter. European Patent No. 248 401.
Fujiwara, A, Sugisawa, T., Shinjoh, M., Setoguchi, Y. & Hoshino, T. (1987c).
Process for the manufacture of ketogulonic acid. European Patent No. 213 591.
Fukaya, M., Iwata, T., Entani, E., Masai, H., Uozumi, T. & Beppu, T. (1985a).
Distribution and characterization of plasmids in acetic acid bacteria. Agric. BioI.
Chem., 49, 1349-55.
Fukaya, M., Takayama, K., Okumura, H., Masai, H., Uozumi, T. & Beppu, T.
(1985b). Improved transformation method for Acetobacter with plasmid DNA.
Agric. BioL. Chem., 49, 2091-7.
Fukaya, M., Okumura, H., Masai, H., Uozumi, T. & Beppu, T. (1985c). Development
of a host-vector system for GLuconobacter suboxydans. Agric. BioL. Chem., 49,
Garcia, J. L. & Buesa, J. M. (1986). An improved method to clone penicillin acylase
genes: Cloning and expression in Escherichia coLi of penicillin G acylase from
KLuyvera citrophila. 1. Biotechnol., 3, 187-95.

Microbial Production of Vitamin C


Geiger-Huber, M. & Galli, H. (1945). Uber den Nachweis der L-Ascorbinsaure ais
Stoffwechsel-produkt von Aspergillus niger. Helv. Chim. Acta, 28, 248-50.
Geigert, J., Hirano, D. S. & Neidleman, S. L. (1981). High performance liquid
chromatographic method for the determination of L-ascorbic acid and D-isoascorbic
acid. J. Chromatogr., 206, 396.
Goeddel, D. V., Kleid, D. G., Bolivar, F., Heyneker, H. L., Yansura, D. G., Crea,
R., Hirose, T., Kraszewski, A., Itakura, K. & Riggs, A. D. (1979). Expression in
Escherichia coli of chemically synthesized genes for human insulin. Proc. Natl.
A cad. Sci. U.S.A., 76, 106-10.
Gray, E. B. (1947a). 2-Keto-gulonic acid and its salts. US Patent No.2 421611.
Gray, E. B. (1947b). 2-Keto-gulonic acid and its salts. US Patent No.2 421612.
Grindley, J. F., Payton, M. A., Van de Pol, H. & Hardy, K. G. (1988). Conversion of
glucose to 2-keto-L-gulonate, an intermediate in L-ascorbate synthesis, by a
recombinant strain of Erwinia citreus. Appl. Environ. Microbiol., 54, 1770-5.
Hardy, K., Van de Pol, H., Grindley, J. & Payton, M. A. (1987). Production of a
vitamin C precursor using genetically modified organisms. WO 87/00863.
Haworth, W. N. & Hirst, E. L. (1933). Ascorbic acid. Chem. Ind. (London), 52,
Hirst, E. L. (1939). The structure and synthesis of vitamin C and its analogs. Fortschr.
Chem. Org. Naturst., 2, 132-59.
Hirst, E. L., Percival, E. G. V. & Smith, F. (1933). Ascorbic acid. Nature, Lond., 131,
Holloway, B. W. (1986). Organization of the Pseudomonas genome. In Proceedings of
the 5th International Symposium on the Genetics of Industrial Micro-organisms, ed.
M. Alacevic, D. Hranueli & Z. Toman. Pliva, Zagreb, pp. 393-401.
Huang, H. T. (1962). 2-Keto-L-gulonic acid. US Patent No.3 043 749.
Hughes, D. E. (1983). Titrimetric determination of ascorbic acid in commercial liquid
diets with 2,6-dichloroindophenol. J. Pharmac. Sci., 72, 126-9.
Isono, M., Nakanishi, K., Sasajama, K., Motizuki, K., Okazaki, H. & Yoshino, H.
(1968). 2-Keto-L-gulonic acid fermentation. I Paper chromatographic characterization of metabolic products from sorbitol and L-sorbose by various bacteria. Agric.
Bioi. Chem., 32,424-31.
IUPAC-IUB (1965). Commission on Biochemical Nomenclature. Biochim. Biophys.
Acta, 107,4.
Jaffe, G. M. (1984). Ascorbic acid. In Encyclopedia of Chemical Technology, Vol. 24,
ed. R. E. Kirk & D. F. Othmer. John Wiley & Sons, New York, pp. 8-40.
Kahn, M. S. & Manning, R. F. (1988). Process for the preparation of keto-gulonic acid.
European Patent No. 276 832.
Kanzaki, T. & Okazaki, H. (1970). 2-Keto-L-gulonic acid fermentation. IV L-Sorbose
metabolism in Pseudomonas aeruginosa. Agric. BioI. Chem., 34, 432-6.
Karube, I., Suzuki, S. & Vandamme, E. J. (1984). Antibiotic production with
immobilized living cells. In Biotechnology of Industrial Antibiotics, ed. E. J.
Vandamme. Marcel Dekker, New York, Basel, pp. 762-80.
Katznelson, H., Tanebaum, S. W. & Tatum, E. L. (1953). Glucose, gluconate and
2-ketogluconate oxidation by Acetobacter melanogenus. J. Bioi. Chem., 204,43-59.
Kieslich, K. (1984). Present state of biotechnological productions of pharmaceuticals.
In Proceedings of the 3rd European Congress on Biotechnology, Vol. IV. VCH,
Weinheim, pp. 39-72.
Kita, D. A. (1979). Fermentation process for converting L-gulonic acid to 2-keto-Lgulonic acid. US patent No.4 115 812.
Kita, D. A. & Hall, K. E. (1981a). Preparation of 2-keto-L-gulonic acid. US Patent No.
Kita, D. A. & Hall, K. E. (1981b). Process for producing 2,5-diketo-gluconic acid. US
Patent No.4 263 402.


V. Delii, D. Sunii & D. VlaIiC

Kitamura, I. & Perlman, D. (1975). Conversion of L-sorbose to L-sorbosone by

Gluconobacter melanogenus. Biotechnol. Bioeng., 17,349-59.
Klein, J. & Wagner, F. (1978). Immobilized whole cells. Biotechnology, Dechema, 82.
Verlage Chemie, Weinheim, New York, pp. 142-64.
Kulbe, K. D. & Knopki, G. (1986). Process for the intrasequential cofactor
regeneration in enzymatic synthesis, particularly when producing vitamine C. WO
Kulhanek, M. (1970). Fermentation processes employed in vitamin C synthesis. In
Advances in Applied Microbiology, Vol. 12, ed. D. Perlman. Academic Press, New
York, London, pp. 11-33.
Lazarus, R. A. & Seymour, J. L. (1986). Determination of 2-keto-L-gulonic and other
ketoaldonic and aldonic acids produced by ketogenic bacterial fermentation. Analyt.
Biochem., 157,360-6.
Loewus, F. A. (1980). L-Ascorbic acid: metabolism, biosynthesis, function. In
Biochemistry of Plants, Vol. 3, ed. J. Preiss. Academic Press, New York, pp.
Lu, S., Huang, S. & Yu, J. (1985). Rules of the supply and demand of oxygen and scale
up for vitamin C fermentation. Huaxue Fanying Gongcheng Yu Gongyi, 1, 72-82.
Makover, S. & Pruess, D. L. (1976). 2-Keto-L-gulonic acid. Swiss Patent No. 58016l.
Makover, S., Ramsey, G. B., Vane, F. M. & Witt, C. G. (1975). New mechanisms for
the biosynthesis and metabolism of 2-keto-L-gulonic acid in bacteria. Biotechnol.
Bioeng., 17, 1485-514.
Martin, C. K. A. & Perlman, D. (1975). Stimulation by organic solvents and detergents
of conversion of L-sorbose to L-sorbosone by Gluconobacter melanogenus IFO 3293.
Biotechnol. Bioeng., 17, 1473-83.
Martin, C. K. A. & Perlman, D. (1976). Conversion of L-sorbose to L-sorbosone by
immobilized cells of Gluconobacter melanogenus IFO 3293. Biotechnol. Bioeng., 18,
Mayer, H., Collins, J. & Wagner, F. (1979). Cloning of the penicillin G-acylase gene of
Escherichia coli ATCC 11105 on multicopy plasmids. In Plasmids of Medical,
Environmental and Commercial Importance, ed. K. M. Timmis & A. Puhler.
Elsevier/North-Holland Biomedical Press, Amsterdam, pp. 187-94.
Merck Index, The, 10th edn. (1983). Merck & Co., Rahway, New Jersey, p. 844.
Miller, J. V., Estell, D. A. & Lazarus, R. A. (1987). Purification and characterization
of 2,5-diketo-o-gluconate reductase from Corynebacterium sp. J. Bioi. Chem., 262,
Mochizuki, K., Kanzaki, T., Kusunoki, T. & Okazaki, H. (1969a). Idonic acid by
oxidation of sorbose with Pseudomonas. Japan Patent No. 6908073.
Mochizuki, K., Kanzaki, T., Okazaki, H. & Doi, M. (1969b). 2-Keto-L-gulonic acid by
Acetobacter fermentation of L-idonic acid. Japan Patent No. 69 08074.
Morikawa, Y., Karube, I. & Suzuki, S. (1979). Penicillin G production by immobilized
whole cells of Penicillium chrysogenum. Biotechnol. Bioeng., 21,261-70.
Morikawa, Y., Karube, I. & Suzuki, S. (1980). Continuous production of bacitracin by
immobilized living whole cells of Bacillus sp. Biotechnol. Bioeng., 22, 1015-23.
Murty, N. K. & Rao, K. R. (1979). Vitamine C (ascorbic acid). In Methods in
Enzymology, Vol. 62, ed. D. B. McCormick &am