Alimentos genéticamente

modificados
Profesora: Bravo Ayala, Marta Margot

Estudiantes:
❏ Flores Velasquez, Gabriela
❏ Oré Maldonado, Kevin Anthony
❏ Valdez Chirinos, Irene
2022 - I
RESUMEN

Con el objetivo de conocer la caracterización, función y ventajas de un transgénico, se

tomó una serie de definiciones y ejemplos que clarifican lo necesario sobre estos
alimentos genéticamente modificados, ya que son encargados de proporcionar
características concretas a los vegetales o frutas que desean emplear. En el caso de las
plantas, por ejemplo, se han alterado sus genes para hacerlas resistentes a
determinadas plagas, para intensificar la producción en el cultivo.En la actualidad, la
soja, algún tipo de patatas y el maíz transgénicos se cultivan en gran parte del mundo y
se han incorporado al consumo general, tanto en su forma natural como en la
producción de alimentos envasados y así como estos se emplean otros más que poseen
un fin determinado dentro de su estudio.

2
INTRODUCCIÓN
La biotecnología moderna o ingeniería genética ha presentado avances
notables desde hace poco más de treinta años. En la actualidad, el mundo es
testigo de los resultados de estas innovaciones a través de la creación de
plantas o animales con características nuevas provenientes de la manipulación
de sus genes, esta cruza barreras naturales entre especies, que no serían
transgredidas siguiendo una evolución natural. De esta manera se crean, en
laboratorios, organismos con nuevas características, denominados
organismos vivos modificados (OVM), conocidos también como transgénicos
[1].

3
ALIMENTOS GENÉTICAMENTE MODIFICADOS
Los alimentos genéticamente modificados (GM) tienen un ADN peculiar usando genes de
otras plantas o animales. Los científicos toman el gen de un rasgo deseado de una planta o
animal e insertan ese gen dentro de una célula de otra planta o animal.
El proceso para crear alimentos GM (transgénicos) es diferente a la cría selectiva. Esta
involucra la selección de plantas o animales con los rasgos deseados y su crianza que con el
tiempo resulta en la descendencia con los rasgos deseados [2].

BENEFICIOS:

● Alimentos más nutritivos y apetitosos

● Plantas resistentes a la sequía y a las enfermedades, que requieren menos recursos
ambientales (como agua y fertilizante)
● Menos uso de pesticidas
● Aumento en el suministro de alimentos a un costo reducido y con una mayor vida útil
● Crecimiento más rápido en plantas y animales
● Alimentos medicinales que se podrían utilizar como vacunas u otros medicamentos

4
ALIMENTOS GENÉTICAMENTE MODIFICADOS
Por su parte, en los OGM de segunda generación se pretende
mejorar la composición del producto o su valor nutritivo, por
medio de la inclusión de vitaminas, atributos medicinales,
eliminación de alérgenos naturales, modificación del contenido de
proteínas, aceites, etc [3].

Algunos ejemplos y para qué fueron tratados químicamente.

● Soja. Modificación en la semilla, para ser más resistente a los

herbicidas.
● Maíz. Genes insertados en el genoma de la planta, para
hacerlo más resistente a insectos.
● Carnes. Aumentar el tamaño y el peso de los animales, y
acelerar la velocidad de su crecimiento.
● Trigo. Hacerlo más resistente antes sequías.

5
APLICACIONES

Las enzimas alimentarias naturales a menudo tienen limitaciones en condiciones de

procesamiento de alimentos sofisticadas y extremas. Las enzimas modificadas
genéticamente están diseñadas para mejorar o alterar las propiedades enzimológicas y/o
aumentar la pureza y el rendimiento de la expresión, cambios que pueden atribuirse a la
secuencia de aminoácidos alterada, resultante de la manipulación de la secuencia del gen.
Este artículo revisa las estrategias básicas para producir enzimas modificadas
genéticamente, como carbohidrasas, proteasas, lipasas, etc., y su uso o uso potencial en el
procesamiento de alimentos. Aunque existen desafíos y preocupaciones de seguridad para el
uso de proteínas recombinantes en el procesamiento de alimentos, las enzimas modificadas
genéticamente son prometedoras debido a los beneficios potenciales para la industria
alimentaria, los consumidores y el medio ambiente.

6
APLICACIONES

Para investigar la importancia de la selectividad del sustrato para la funcionalidad de la xilanasa en la

elaboración de pan, el sitio de unión secundario (SBS) de las xilanasas de Bacillus subtilis (XBS) y
Pseudoalteromonas haloplanktisfue modificado. Esto dio como resultado dos xilanasas con actividad
relativa aumentada hacia arabinoxilano de trigo no extraíble en agua (WU-AX) en comparación con
arabinoxilano de trigo extraíble en agua, es decir, una selectividad de sustrato aumentada, sin cambiar otras
propiedades bioquímicas. La adición de ambas xilanasas modificadas en la elaboración de pan resultó en un
aumento de los volúmenes de pan en comparación con los tipos silvestres cuando se usaba harina débil.
Además, el aumento de volumen máximo se alcanzó con una dosis más baja del mutante en comparación con
el XBS de tipo salvaje. Las xilanasas modificadas pudieron solubilizar más WU-AX y redujeron más el grado
medio de polimerización del arabinoxilano soluble en la masa durante la fermentación. Esto posiblemente
permitió una liberación adicional de agua, lo que podría ser responsable del aumento de los volúmenes de
pan. Funcionalidad SBS alterada y, como resultado,
7
 APLICACIONES
Dentro de la literatura, se ha demostrado que la
fracción de lectina (TBLF) del Frijol Tepary
(Phaseolus acutifolius ) se une específicamente e
induce la muerte celular de diferentes tipos de
células cancerosas. Sin embargo, no se ha
podido desarrollar un fármaco debido a que la
producción es costosa, lenta y presenta bajos
rendimientos.

Con lo cual, el trabajo tiene como objetivo planificar una alternativa para producir una lectina bioactiva por el proceso de
rizosecreción a través de exudados de raíces en plantas genéticamente modificadas. Entonces, se procedió a amplificar
los transcritos de frijol Tepary mediante cebadores degenerados, cuyos productos fueron secuenciados. Entre las
secuencias resultantes, se logró la dilucidación de las lectinas presentes en TBLF.

Su secuencia codificante fue introducida en la planta P. acutifolius utilizando Agrobacterium tumefaciens, para, luego,
hacerlas crecer in vitro. Una vez que las raíces crecieron, las plantas fueron transferidas a condiciones hidropónicas y se
analizaron los exudados de las raíces.

En los resultados se identificó la presencia de una lectina cisgénica glicosilada con actividad biológica, confirmando la
eficacia de la alternativa propuesta para la producción sintética y purificación de esta lectina.

8
 APLICACIONES
Los terpenos, que se encuentran en las uvas, le
otorgan el sabor y aroma característicos a los
vinos, y sirven para identificar la calidad de los
productos vitivinícolas. Los terpenos se derivan a
través de las vías del mevalonato (MVA) y del
fosfato de metileritritol (MEP). En las células
vegetales se pueden manifestar ambas vías; pero
la levadura y otros hongos solo albergan la vía
MVA, mientras que las bacterias producen
terpenos exclusivamente a través de la vía MEP.

En el proceso de fermentación, como la vinificación, muchas cepas de S. cerevisiae pueden hidrolizar enzimáticamente los
terpenos unidos en el mosto de uva y así liberar las contrapartes volátiles libres en el vino. En los últimos años, la
producción de terpenos a través de fábricas de células microbianas se ha convertido en una alternativa prometedora por
manifestar un rápido crecimiento, ser ahorrador de terreno y controlable. Los productos de terpenos de microorganismos
modificados han mostrado un gran potencial en aplicaciones alimentarias que incluyen agentes aromatizantes, agentes
edulcorantes , conservantes e ingredientes alimentarios funcionales.

El trabajo de revisión tiene por objetivo investigar las vías MVA y MEP de la biosíntesis de terpenos y la ingeniería
metabólica de S. cerevisiae para promover la producción de terpenos.

9
CONCLUSIONES
La ingeniería genética promete traer importantes y rápidos avances en la producción de alimentos,
con el objetivo de lograr resolver los diversos problemas en la alimentación, como la desnutrición, el
hambre y la pobreza de una sociedad mundial que se encuentra en continuo crecimiento, mediante el
aumento de la producción, la mejora de la calidad y seguridad de los alimentos. Por lo tanto, es
necesario considerar con mucho cuidado los posibles efectos adversos para garantizar que esta
herramienta tecnológica sea un beneficio para la humanidad. A lo largo de estos años, en la población
han aparecido controversias sobre los alimentos modificados genéticamente, generados por un
escaso conocimiento objetivo del tema; por lo tanto, también se debe tener la intención de disipar los
temores de los consumidores mediante la aplicación de técnicas fiables, proporcionando mayor
evidencia experimental y promoviendo debates y seminarios para aumentar la confianza.

APORTES
Mejoramiento en la calidad, sabor, velocidad de producción, etc de los alimentos. De poder tratar
adecuadamente los efectos secundarios, y mejorando el proceso de los alimentos modificados
genéticamente, se puede solucionar los problemas de plagas y escasez de varios alimentos.

10
REFERENCIAS BIBLIOGRÁFICAS
1. Regulación De Los Transgénicos En El Perú. [citado el 22 de julio de 2022]. Disponible en:
https://www.minam.gob.pe/wp-content/uploads/2015/08/transgenicos_FINALpdf.pdf
2. Reyes S. María Soledad, Rozowski N Jaime. ALIMENTOS TRANSGÉNICOS. Rev. chil. nutr. [Internet]. 2003
Abr [citado 2022 Jul 22] ; 30( 1 ): 21-26. Disponible en:
http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0717-75182003000100003&lng=es.
http://dx.doi.org/10.4067/S0717-75182003000100003.
3. Spendeler Liliane. Organismos modificados genéticamente: una nueva amenaza para la seguridad
alimentaria. Rev. Esp. Salud Publica [Internet]. 2005 Abr [citado 2022 Jul 22] ; 79( 2 ): 271-282. Disponible
en: http://scielo.isciii.es/scielo.php?script=sci_arttext&pid=S1135-57272005000200013&lng=es.
4. MARTÍNEZ-ALARCÓN, Dania, et al. Rhizosecretion of a cisgenic lectin by genetic manipulation of Tepary
bean plants (Phaseolus acutifolius). Journal of Biotechnology. [Internet]. 2019. [Consultado 22 julio 2022].
vol. 306, p. 100013. Disponible en: https://www.sciencedirect.com/science/article/pii/S2590155919300095
5. LIANG, Zijian, et al. Genetic engineering of yeast, filamentous fungi and bacteria for terpene production and
applications in food industry. Food Research International, 2021, vol. 147, p. 110487. Disponible en:
https://www.sciencedirect.com/science/article/pii/S0963996921003860
6. 1. Zhang Y, Geary T, Simpson BK. Genetically modified food enzymes: a review. Curr Opin Food Sci. 2019 Feb
1;25:14–8.
7. 1. Leys S, Pauly A, Delcour JA, Courtin CM. Modification of the secondary binding site of xylanases illustrates
the impact of substrate selectivity on bread making. J Agric Food Chem [Internet]. 2016 Jul 6 [cited 2022 Jul
22];64(26):5400–9. Available from: https://pubs.acs.org/doi/abs/10.1021/acs.jafc.6b01473

11
Accepted Manuscript

Title: Genetically modified food enzymes: A review

Authors: Yi Zhang, Timothy Geary, Benjamin K Simpson

PII: S2214-7993(18)30091-2
DOI: https://doi.org/10.1016/j.cofs.2019.01.002
Reference: COFS 426

To appear in:

Please cite this article as: Zhang Y, Geary T, Simpson BK, Genetically
modified food enzymes: A review, Current Opinion in Food Science (2019),
https://doi.org/10.1016/j.cofs.2019.01.002

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Genetically modified food enzymes: A review

Yi Zhang a, Timothy Geary b, Benjamin K. Simpson a, *

a
Department of Food Science and Agricultural Chemistry, McGill University, Ste-Anne-de-Bellevue,
Quebec H9X 3V9, Canada

PT
b
Institute of Parasitology, McGill University, Ste-Anne-de-Bellevue, Quebec H9X 3V9, Canada

RI
Corresponding author:
Benjamin K. Simpson. E-mail: benjamin.simpson@mcgill.ca

SC
Other authors:
Yi Zhang. E-mail: yi.zhang10@mail.mcgill.ca

U
Timothy Geary. E-mail: timothy.g.geary@mcgill.ca

Highlights
N
A
 Food enzyme modification is developed by advanced genetic techniques
M

 Genetically modified food enzymes can exhibit improved catalytic efficiency, stability, substrate
specificity

D

Most genetically modified food enzymes are designed to be applied in food processing involving
carbohydrates and lipids
TE

Abstract
EP

Natural food enzymes often have limitations in sophisticated and extreme food processing conditions.
Genetically modified enzymes are designed to improve or alter the enzymological properties and/or
increase purity and yield of expression, changes which can be attributed to the altered amino acid sequence,
CC

resulting from gene sequence manipulation. This article reviews the basic strategies for producing
genetically modified enzymes, such as carbohydrases, proteases, lipases, etc., and their use or potential use
A

in food processing. Although there are challenges and safety concerns for using recombinant proteins in
food processing, genetically modified enzymes are promising because of potential benefits for the food
industry, consumers and the environment.

Keywords
Enzyme; food; mutagenesis; genetic modification

Page 1 of 12
1. Introduction
According to the definition of genetically modified organisms (GMOs) by World Health Organization
(WHO), the term “genetically modified enzymes” can be used to refer to situations in which the DNA
encoding the enzyme of interest has been artificially altered. Non-genetically modified enzymes remain the
major source for the food enzyme market, which includes endogenous enzymes produced from animals,
plants and microorganisms. Modern food processing tends to be sophisticated and precise, and sometimes

PT
requires extreme conditions such as high temperature, high pressure, extreme pH, salinity and the use of
solvents. However, native enzymes often have limitations that need to be improved to suit specific food

RI
processing environments. Genetic modification aims to rationally improve enzymes to enhance
characteristics such as purity, yield, specificity, catalytic efficiency, stability, surface property and

SC
multifunctionality, as well as enable cost-effective production and sustainable development for food
processing.
The strategy for producing genetically modified food enzymes is based on genetic engineering and

U
requires comprehensive understanding and integration of bioinformatics, enzymology, biophysical

N
chemistry and molecular biology, etc. Most genetically modified enzymes are expressed in well-established
standard microbial model systems such as E. coli and Pichia pastoris. There are a few strategies for
A
producing genetically modified enzymes depending on information known about the target enzyme and the
M

purpose of its use:

(i) Gene sequence optimization is commonly used for heterologous expression of recombinant
enzymes to optimize codon bias of the host cell, as well as to limit mRNA secondary structure
D

that could inhibit translation [1]. These changes usually increase yield, while the amino acid
TE

sequence remains the same.

(ii) Gene truncation or fusion requires knowledge of the relationship between amino acid sequence
EP

and enzyme function. Gene truncation involves eliminating DNA sequences that are not
functionally important, enabling the enzyme to be expressed in higher yield or with higher
activity; gene fusion constructs a gene with 2 or more enzyme DNA sequences so that the
CC

recombinant enzyme has multiple functions [2].

(iii) Site-directed mutagenesis is a rational design process based on in-depth knowledge of the target
enzyme in terms of structure, catalytic mechanism, active site, etc. DNA sequence(s) encoding
A

specific amino acid(s) that are related to activity or structural stability is (are) replaced, inserted
or deleted to produce enzymes with desired or improved properties [3].
(iv) Directed evolution, a non-rational design process that requires no prior knowledge of the target
enzyme, is an in vitro accelerated process that mimics natural evolution for generating enzymes.
It consists of two key processes: genetic diversity generation for constructing combinatorial

Page 2 of 12
 libraries, and high-throughput screening or selection. The first process can be achieved by
random mutagenesis using error-prone PCR techniques, or gene recombination using
representative DNA-shuffling techniques [4]. The second process for identifying or isolating
improved variants may be quite difficult to perform, so that developing efficient screening
protocols has been a primary goal of innovation [5].
(v) Semi-rational design involves mutations based on sequence, structure or computational models,

PT
followed by small-scale mutagenesis and screening methods [3]. An example of semi-rational
design is site saturation mutagenesis, in which a mutant library is created containing all possible

RI
substitutions at one or more target positions in a gene sequence [6].
This review article summarizes novel food enzymes produced by genetic modification, especially

SC
using strategies (ii)-(v), and discusses their improved properties as well as applications and potential
applications in food processing.

U
2. Genetically modified carbohydrases/glycosidases

N
Carbohydrases are the most widely-used enzymes in the food industry, being employed in baking,
brewing, sweetener production, etc. Some of these enzymes break down polysaccharides into simple sugars.
A
Alpha-amylase (EC 3.2.1.1) is ubiquitous in the starch industry; the commercially available Bacillus
M

licheniformis α-amylase is thermostable but acid-sensitive. Novel Bacillus licheniformis α-amylases

obtained using directed evolution by replacing active site domain His residues with Arg and Asp residues
had higher activity at pH 4.5 than the wild type enzyme [7,8], which enhances starch liquefaction,
D

saccharification and fermentation processes in which the pH of starch slurry is usually < 6.0 [9]. A modified
TE

Rhizopus oryzae α-amylase obtained using site-saturation mutagenesis of His 286 had a higher optimum
temperature (60 oC) and lower optimum pH (4.0–4.5) than the wild-type enzyme [10], properties better
suited for use in high maltose syrup production. A mutant Aspergillus aculeatus β-glucosidase (EC 2.3.1.21)
EP

obtained using site-saturation mutagenesis had improved hydrolytic efficiency especially to cellobiose, and
was used to accelerate the saccharification of alkaline-pretreated bagasse [11]. The substrate specificity of
CC

Thermotoga maritima β-glucosidases after genetic modification was enhanced for quercetin glucosides [12],
suggesting its usage in producing aglycones that have higher pharmaceutical activity compared with the
substrate. Endo-β-1,4-xylanase (EC 3.2.1.8) cleaves the β-1,4 glycosidic linkages in xylans. A variant of
A

this enzyme generated using site-directed mutagenesis exhibited higher specific activity and was utilized in
the production of xylooligosaccharides from wheat straw under thermal and alkaline conditions (50–65 oC,
pH 7–10) [13]. Novel endo-β-1,4-xylanases with modification in secondary binding sites also showed
increased specificity toward water-insoluble, but not water-soluble, wheat arabinoxylan, and its application
in bread making resulted in increased loaf volumes [14]. An engineered β-1,3-1,4-glucanase (EC 3.2.1.73),

Page 3 of 12
hydrolyzing β-1,3-1,4-glucan usually found in avena, rice, oats and barley, had improved catalytic
efficiency, thermostability and halostability following replacement of hydrophobic Lys 48 with Ala and
Leu [15]. Bacillus spp. is the main source of β-1,3-1,4-glucanase used in malt extract production in the
brewing industry, and novel β-1,3-1,4-glucanases were produced by replacing Lys with Ser to form rigid
β-sheet structures. The modified glucanases functioned effectively under industrial operating conditions
with improved optimal temperature and thermostability [16].

PT
Some types of carbohydrases hydrolyze disaccharides into monosaccharides. For example, a novel
Saccharomyces cerevisiae invertase (EC 3.2.26) was engineered to have optimal pH and stability ranges

RI
for the breakdown of sucrose into glucose and fructose; the enzyme was modified by the substitution of
hydrophilic residues in the active site region or peripheral loops with hydrophobic amino acids [17].

SC
Another genetically modified invertase with high transfructosylating activity was used for simple and
efficient production of prebiotic fructooligosaccharides [18]. Commercial β-galactosidase or lactase (EC
3.2.1.23) is commonly applied in lactose-free milk product processing, in which the process temperature is

U
relatively low. A variant of β-galactosidase obtained using directed evolution had high activity in industrial-

N
type conditions for milk processing, i.e., substrate lactose, buffer pH 6.75, and 8 °C [19]. To improve the
transglycosylating activity of this enzyme, a thermoresistant Thermotoga maritima β-galactosidase
A
modified using rational design was obtained with better specificity for the biosynthesis of galacto-
M

oligosaccharides, a prebiotic compound [20].

3. Genetically modified proteases/peptidases

D

The major use of proteases in the food industry is hydrolysis of protein matrices to enhance flavor, texture
TE

or functional properties in diary, meat and fish products. Commonly used food proteases are derived from
animals (such as trypsin, chymotrypsin, chymosin/rennet, pepsin), plants (such as papain, bromelain, ficin),
EP

and microorganism (such as the microbial acid protease Flavourzyme® from Aspergillus oryzae and
proteinase A from S. cerevisiae; bacteria neutral proteases produced by Bacillus circulans BM15 and
Streptomyces nogalator AC 80; and bacteria alkaline proteases produced by Aspergillus niger and Bacillus
CC

subtilis).
Unlike food carbohydrases and lipases, animal/plant-derived food proteases are generally less
commonly found in genetically modified forms, probably because the versatile properties of the native
A

proteases are adequate to suit most required temperature and pH conditions in food processing, and their
catalysis mechanism has been well investigated. It could also be due to the very ample availability of
microbial proteases for industrial uses. Another reason could be the limitations posed for efficient
expression of proteases derived from animal or plant sources in microbial systems.
Recent research on modified food proteases from Aspergillus and Bacillus species indicates that

Page 4 of 12
engineered metalloproteases obtained through site-saturation mutagenesis of His224 had improved affinity
for substrate, making the synthesis of Z-aspartame more cost-effective [21]. An acid protease from a mutant
A. oryzae strain was produced using solid state fermentation with potato pulp powder, with enhanced
glycine releasing activity [22]. A truncated neutral protease from A. oryzae had optimum pH of 8.0 and
optimum temperature of 55 oC, and its enzymological characterization suggested that it was efficient in
antihypertensive peptide production, debittering, and food oil processing [23].

PT
Bacillus spp. are the main producers of alkaline proteases with high thermal and pH stabilities. Subtilisin
nattokinase, a serine protease from B. subtilis var. natto, was modified by site-directed mutagenesis to

RI
obtain higher specific activity and oxidative stability, and may be useful as a potential cardiovascular drug
due to its strong fibrinolytic activity [24]. The cold activity at 10 oC and alkali-resistant properties of a

SC
Bacillus alcalophilus alkaline protease were improved through directed evolution using error-prone PCR
[25] for use in cold-temperature food processing. Similarly, an alkaline serine protease from mesophilic
Bacillus pumilus was engineered to have increased hydrolytic efficiency at 15 oC without compromising

U
thermostability [26]. Modified protases are also applied in the detergent industry. An alkaline protease

N
isolated from an oil-polluted mud flat metagenome by random mutagenesis displayed robust compatibility
with laundry detergents at low temperature (30 °C) and alkaline pH range (pH 8.0–11.0) [27].
A
M

4. Genetically modified lipases

Lipase (EC 3.1.1.3) is the most common biocatalyst for lipid (fats and oils) modification. Bacillus lipases
display activities over wide pH and temperature ranges, and some possess fatty acid specificity. Lipase A
D

from B. subtilis was expressed as a fusion protein with cell wall mannoprotein Pir4 resulting in an
TE

immobilized lipase, and was used as a leavening agent to enhance rheological and aromatic properties of
bread [28]. The secretion of Rhizopus oryzae lipase was improved by rational design of the N-glycosylation
EP

sites, which could meet the industrial demand for Rhizopus lipase, especially for the edible oil and fat
industries [29]. Candida rugosa lipase isozymes modified by site-directed mutagenesis had high catalytic
efficiency for producing fatty acid methyl esters and diglycerides as food emulsifiers, as well as for the
CC

conversion of crude Jatropha curcas seed oil into biodiesel [30,31]. Malassezia globose lipase with specific
activity especially on mono- and diacylglycerol was modified with enhanced thermostability to meet
industrial-scale requirements, making it a potential biocatalyst for synthesis of diacylglycerols in edible oils
A

with health benefits [32]. Engineered thermostable T1 lipases had low activity toward long-chain
triacylglycerols; thus, they are potential biocatalysts to enhance flavor in dairy products by generation of
short-chain fatty acids from milk fats [33]. A Thermomyces lanuginosus lipase modified using semi-rational
design was targeted to enhance methanol tolerance for potential use in biofuel production from waste food
oils and grease [34].

Page 5 of 12
5. Other genetically modified food enzymes
A D-psicose 3-epimerase (EC 5.1.3.31) modified by site-directed mutagenesis displayed high substrate-
binding affinity, catalytic efficiency and thermostability in catalyzing the isomerization of D-fructose to D-
psicose, an ultra-low-calorie sweetener with desirable physiological properties [35]. Similarly, engineered
Caldicellulosiruptor saccharolyticus cellobiose 2-epimerases (EC 5.1.3.11) showed increased specific
activity to directly convert lactose into lactulose without co-production of epilactose, and its application

PT
will boost the commercial availability of lactulose, a non-digestible disaccharide used as a prebiotic food
additive and medicine [36]. A novel trehalose synthase (EC 2.4.1.245) from Picrophilus torridus was

RI
prepared by replacing Pro residues to increase thermostability and catalytic efficiency, and was used to
convert maltose from low-value agricultural products such as sweet potato starch to high-value trehalose

SC
[37]. Yersinia phytases (EC 3.1.3.8) were genetically manipulated using site-directed mutagenesis to
improve pepsin and trypsin resistance and thermostability, enabling potential use for efficient hydrolysis of
phytic acid to inorganic phosphate to improve nutrient uptake from foods [38]. A novel di-D-fructose

U
dianhydride I (DFA I)-forming inulin fructotransferase was engineered using rational design to improve

N
its catalytic behavior, including activity and thermostability, so that it can produce prebiotic DFA I in
industrial processes using inulin as substrate [39]. Nitrilase (EC 3.5.5.1), which converts nitriles into
A
corresponding carboxylic acids, was genetically modified to achieve higher activity for stereospecific
M

production of (R)-(−)-mandelic acid, useful in pharmaceutical production [40]. A mutant Aspergillus niger
α-L-rhamnosidase with improved thermostability and substrate affinity reduced bitterness caused by
naringin in orange juice [41]. 1,4-α-glucan branching enzyme (EC 2.4.1.18) was engineered by Met349
D

mutation to achieve higher activity for the formation of α-1,6-glucosidic linkages, and could form highly
TE

branched potato starch [42]. A novel Bacillus stearothermophilus NO2 cyclodextrin glycosyltransferase
was produced through iterative saturation mutagenesis around the catalytic residues, and was used to
EP

produce 2-O-α-D-glucopyranosyl-L-ascorbic acid, which has great advantages in food processing due to its
stability and facile degradation to L-ascorbic and glucose [43].
CC

6. Safety concerns
Genetically modified food enzymes are currently produced from GMOs. Safety concerns have been
raised regarding potential contamination of food with bacterial toxins or mycotoxins, allergens or
A

uncharacterized extraneous substances as impurities [44,45]. Genetic modification of enzymes may also
change their allergenic properties, posing new potential health risks. For instance, type I sensitisation was
found in a study of 813 exposed industrial workers using genetically modified enzymes [46]. Thus, prior to
marketing, genetically modified food enzymes need approval from various regulatory bodies such as the
US Food and Drug Administration, the Association of Manufacturers and Formulators of Enzyme Products

Page 6 of 12
and the European Food Safety Authority, through processes that vary in different countries. In addition,
ethical and religious concerns have been raised for genetically modified enzymes. For instance, it has been
suggested that raw materials or ingredients used in fermentation of microorganisms to produce enzymes
should be halal [47].

7. Conclusion

PT
Genetically modified food enzymes are typically better suited for specific industrial applications than
their native counterparts, and research on their enzymological properties in comparison to wild-type

RI
enzymes enable us to better understand how to optimize structure-function relationships. Currently, novel
genetically modified food enzymes are mostly used for applications in food processing involving

SC
carbohydrates, followed by lipids. Although there are challenges for safe use, genetic modification
strategies for the development of novel food enzymes are highly promising for the future, and we expect
that more genetically modified enzymes will be introduced in various aspects of food processing.

U
Conflict of Interest
The authors declared no conflict of interest.
N
A
M
D
TE
EP
CC
A

Page 7 of 12
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest

References

PT
1. Rosano GL, Ceccarelli EA: Recombinant protein expression in Escherichia coli: advances and
challenges. Front Microbiol 2014, 5:1-17.

RI
2. Dediu D: An introduction to genetics for language scientists. Cambridge, UK: Cambridge University
Press; 2015.

SC
3. Hua W, El Sheikha AF, Xu J: Molecular techniques for making recombinant enzymes used in food
processing. In Molecular Techniques in Food Biology: Safety, Biotechnology, Authenticity and
Traceability. Edited by El Sheikha A F, Levin R E, Xu J. John Wiley & Sons; 2018:95-112.

U
* This book chapter described in detail the molecular strategies to produce recombinant food enzymes.

N
The two core steps in directed evolution, gene diversification and screening/selection are well discussed
with examples. The procedures and techniques used in rational design are also reviewed.
A
4. Labrou NE: Random mutagenesis methods for in vitro directed enzyme evolution. Curr Protein Pept
M

Sci 2010, 11:91-100.

5. Packer MS, Liu DR: Methods for the directed evolution of proteins. Nat Rev Genet 2015, 16:379.
6. Özgün GP, Ordu EB, Tütüncü HE, Yelboğa E, Sessions RB, Gül Karagüler N: Site saturation
D

mutagenesis applications on Candida methylica formate dehydrogenase. Scientifica 2016: 1-7.

TE

7. Liu Y, Fan S, Liu X, Zhang Z, Wang J, Wang Z, Lu F: A highly active alpha amylase from Bacillus
licheniformis: directed evolution, enzyme characterization and structural analysis. J
EP

Microbiol Biotechnol 2014, 24:898-904.

8. Liu Y, Huang L, Jia L, Gui S, Fu Y, Zheng D, Guo W, Lu F: Improvement of the acid stability of
Bacillus licheniformis alpha amylase by site-directed mutagenesis. Process Biochem 2017,
CC

58:174-180.
9. Cates ES, Dinwiddie JA, Aux G, Batie C, Crabb G: Process for starch liquefaction and fermentation.
U.S. Patent 7,727,726, Washington, DC; 2010.
A

10. Li S, Yang Q, Tang B, Chen A: Improvement of enzymatic properties of Rhizopus oryzae α-amylase
by site-saturation mutagenesis of histidine 286. Enzyme Microb Technol 2018, 117:96-102.
11. Baba Y, Sumitani J-i, Tanaka K, Tani S, Kawaguchi T: Site-saturation mutagenesis for β-glucosidase
1 from Aspergillus aculeatus to accelerate the saccharification of alkaline-pretreated bagasse.
Appl Microbiol Biotechnol 2016, 100:10495-10507.

Page 8 of 12
12. Sun H, Xue Y, Lin Y: Enhanced catalytic efficiency in quercetin-4′-glucoside hydrolysis of
Thermotoga maritima β-glucosidase A by site-directed mutagenesis. J Agric Food Chem 2014,
62:6763-6770.
13. Faryar R, Linares-Pastén JA, Immerzeel P, Mamo G, Andersson M, Stålbrand H, Mattiasson B,
Karlsson EN: Production of prebiotic xylooligosaccharides from alkaline extracted wheat
straw using the K80R-variant of a thermostable alkali-tolerant xylanase. Food Bioprod

PT
Process 2015, 93:1-10.
14. Leys S, Pauly A, Delcour JA, Courtin CM: Modification of the secondary binding site of xylanases

RI
illustrates the impact of substrate selectivity on bread making. J Agric Food Chem 2016,
64:5400-5409.

SC
15. Lee JM, Moon SY, Kim Y-R, Kim KW, Lee B-J, Kong I-S: Improvement of thermostability and
halostability of β-1, 3-1, 4-glucanase by substituting hydrophobic residue for Lys48. Int J Biol
Macromol 2017, 94:594-602.

U
16. Niu C, Zhu L, Zhu P, Li Q: Lysine-based site-directed mutagenesis increased rigid β-sheet structure

N
and thermostability of mesophilic 1, 3–1, 4-β-glucanase. J Agric Food Chem 2015, 63:5249-
5256.
A
17. Mohandesi N, Haghbeen K, Ranaei O, Arab SS, Hassani S: Catalytic efficiency and thermostability
M

improvement of Suc2 invertase through rational site-directed mutagenesis. Enzyme Microb

Technol 2017, 96:14-22.
18. Marín-Navarro J, Talens-Perales D, Polaina J: One-pot production of fructooligosaccharides by a
D

Saccharomyces cerevisiae strain expressing an engineered invertase. Appl Microbiol Biotechnol

TE

2015, 99:2549-2555.
19. Rentschler E, Schwarz T, Stressler T, Fischer L: Development and validation of a screening system
for a β-galactosidase with increased specific activity produced by directed evolution. Eur Food
EP

Res Technol 2016, 242:2129-2138.

20. Talens-Perales D, Polaina J, Marín-Navarro J: Structural dissection of the active site of Thermotoga
CC

maritima β-Galactosidase identifies key residues for transglycosylating activity. J Agric Food
Chem 2016, 64:2917-2924.
21. Zhu F, Jiang T, Wu B, He B: Enhancement of Z-aspartame synthesis by rational engineering of
A

metalloprotease. Food Chem 2018, 253:30-36.

** The authors modified a metalloprotease to improve affinity for producing Z-aspartame, a common food
sweetness additive. A computational strategy combining molecular dynamics simulation of the enzyme-
substrate complex with binding free energy calculations was established for engineering the enzyme.
22. Murthy PS, Kusumoto K-I: Acid protease production by Aspergillus oryzae on potato pulp powder

Page 9 of 12
 with emphasis on glycine releasing activity: A benefit to the food industry. Food Bioprod
Process 2015, 96:180-188.
23. Ke Y, Huang W-Q, Li J-z, Xie M-q, Luo X-c: Enzymatic characteristics of a recombinant neutral
protease I (rNpI) from Aspergillus oryzae expressed in Pichia pastoris. J Agric Food Chem 2012,
60:12164-12169.
24. Weng M, Deng X, Bao W, Zhu L, Wu J, Cai Y, Jia Y, Zheng Z, Zou G: Improving the activity of the

PT
subtilisin nattokinase by site-directed mutagenesis and molecular dynamics simulation.
Biochem Biophys Res Commun 2015, 465:580-586.

RI
25. Liu Y, Zhang T, Zhang Z, Sun T, Wang J, Lu F: Improvement of cold adaptation of Bacillus
alcalophilus alkaline protease by directed evolution. J Mol Catal B Enzym 2014, 106:117-123.

SC
26. Zhao H Y, Feng H: Engineering Bacillus pumilus alkaline serine protease to increase its low-
temperature proteolytic activity by directed evolution. BMC Biotechnol 2018, 18:34.
27. Gong B L, Mao R Q, Xiao Y, Jia M L, Zhong X L, Liu Y, Xu P L, Li G: Improvement of enzyme

U
activity and soluble expression of an alkaline protease isolated from oil-polluted mud flat

N
metagenome by random mutagenesis. Enzyme Microb Technol 2017, 106:97-105.
28. Paciello L, Landi C, Orilio P, Di Matteo M, Zueco J, Parascandola P: Bread making with
A
Saccharomyces cerevisiae CEN. PK113‐ 5D expressing lipase A from Bacillus subtilis:
M

leavening characterisation and aroma enhancement. Int J Food Sci Technol 2015, 50:2120-
2128.
29. Yu X W, Yang M, Jiang C, Zhang X, Xu Y: N-Glycosylation engineering to improve the constitutive
D

expression of Rhizopus oryzae lipase in Komagataella phaffii. J Agric Food Chem 2017, 65:6009-
TE

6015.
30. Chang S W, Huang M, Hsieh Y H, Luo Y T, Wu T T, Tsai C W, Chen C S, Shaw J F: Simultaneous
EP

production of fatty acid methyl esters and diglycerides by four recombinant Candida rugosa
lipase’s isozymes. Food Chem 2014, 155:140-145.
31. Kuo T C, Shaw J F, Lee G C: Conversion of crude Jatropha curcas seed oil into biodiesel using
CC

liquid recombinant Candida rugosa lipase isozymes. Bioresour Technol 2015, 192:54-59.
32. Gao C, Lan D, Liu L, Zhang H, Yang B, Wang Y: Site-directed mutagenesis studies of the aromatic
residues at the active site of a lipase from Malassezia globosa. Biochimie 2014, 102:29-36.
A

33. Tang Q, Lan D, Yang B, Khan FI, Wang Y: Site‐ directed mutagenesis studies of hydrophobic
residues in the lid region of T1 lipase. Eur J Lipid Sci Technol 2017, 119:1600107.
34. Tian K, Tai K, Chua BJW, Li Z: Directed evolution of Thermomyces lanuginosus lipase to enhance
methanol tolerance for efficient production of biodiesel from waste grease. Bioresour Technol
2017, 245:1491-1497.

Page 10 of 12
35. Zhang W, Jia M, Yu S, Zhang T, Zhou L, Jiang B, Mu W: Improving the thermostability and catalytic
efficiency of the D-psicose 3-epimerase from clostridium bolteae ATCC BAA-613 using site-
directed mutagenesis. J Agric Food Chem 2016, 64:3386-3393.
36. Shen Q, Zhang Y, Yang R, Pan S, Dong J, Fan Y, Han L: Enhancement of isomerization activity and
lactulose production of cellobiose 2-epimerase from Caldicellulosiruptor saccharolyticus. Food
Chem 2016, 207:60-67.

PT
* This research developed a novel cellobiose 2-epimerase using a random mutagenesis and screening
protocol to enhance lactulose production. The modified enzyme overcame the drawbacks of the wild type

RI
counterpart, such as low enzyme activity and formation of epilactose as by-product.
37. Chou H H, Chang S W, Lee G C, Chen Y S, Yeh T, Akoh CC, Shaw J F: Site-directed mutagenesis

SC
improves the thermostability of a recombinant Picrophilus torridus trehalose synthase and
efficiency for the production of trehalose from sweet potato starch. Food Chem 2010,
119:1017-1022.

U
38. Niu C, Yang P, Luo H, Huang H, Wang Y, Yao B: Engineering of Yersinia phytases to improve

N
pepsin and trypsin resistance and thermostability and application potential in the food and
feed industry. J Agric Food Chem 2017, 65:7337-7344.
A
39. Yu S, Zhang Y, Zhu Y, Zhang T, Jiang B, Mu W: Improving the catalytic behavior of DFA I-forming
M

inulin fructotransferase from Streptomyces davawensis with site-directed mutagenesis. J Agric

Food Chem 2017, 65:7579-7587.
40. Liu Z Q, Zhang X H, Xue Y P, Xu M, Zheng Y G: Improvement of Alcaligenes faecalis nitrilase by
D

gene site saturation mutagenesis and its application in stereospecific biosynthesis of (R)-(−)-
TE

mandelic acid. J Agric Food Chem 2014, 62:4685-4694.

41. Li L J, Wu Z Y, Yu Y, Zhang L J, Zhu Y B, Ni H, Chen F: Development and characterization of an
α-L-rhamnosidase mutant with improved thermostability and a higher efficiency for
EP

debittering orange juice. Food Chem 2018, 245:1070-1078.

** The authors engineered an α-L-rhamnosidase to decrease bitterness in orange juice. The modified
CC

enzyme has improved thermostability, substrate affinity for naringin and better resistance to inhibition by
glucose. This research has practical value in citrus juice processing.
42. Liu Y, Ban X, Li C, Gu Z, Cheng L, Hong Y, Li Z: Met349 mutations enhance the activity of 1, 4-α-
A

glucan branching enzyme from Geobacillus thermoglucosidans STB02. J J Agric Food Chem
2017, 65:5674-5680.
43. Tao X, Wang T, Su L, Wu J: Enhanced 2-O-α-d-glucopyranosyl-L-ascorbic acid synthesis through
iterative saturation mutagenesis of acceptor subsite residues in Bacillus stearothermophilus
NO2 cyclodextrin glycosyltransferase. J Agric Food Chem 2018, 66:9052-9060.

Page 11 of 12
44. De Santis B, Stockhofe N, Wal J-M, Weesendorp E, Lallès J-P, van Dijk J, Kok E, De Giacomo M,
Einspanier R, Onori R: Case studies on genetically modified organisms (GMOs): Potential risk
scenarios and associated health indicators. Food Chem Toxicol 2018, 117:36-65.
45. Srivastava N: Production of food-processing enzymes from recombinant microorganisms. In
Enzymes in Food Biotechnology. Edited by Kuddus M. Academic Press; 2019:739-767.
* This book chapter summarized the technical and scientific involvement of microorganisms in food

PT
enzyme production, and the methodology for identifying genes encoding important enzymes, their cloning
and expression, and purification of recombinant food enzymes.

RI
46. Budnik LT, Scheer E, Burge PS, Baur X: Sensitising effects of genetically modified enzymes used in
flavour, fragrance, detergence and pharmaceutical production: cross-sectional study. Occup

SC
Environ Med 2017, 74:39-45.
47. Ermis E: Halal status of enzymes used in food industry. Trends Food Sci Technol 2017, 64:69-73.

U
N
A
M
D
TE
EP
CC
A

Page 12 of 12
 Article

pubs.acs.org/JAFC

Modification of the Secondary Binding Site of Xylanases Illustrates

the Impact of Substrate Selectivity on Bread Making
Sofie Leys, Anneleen Pauly, Jan A. Delcour, and Christophe M. Courtin*
Laboratory of Food Chemistry and Biochemistry & Leuven Food Science and Nutrition Research Centre (LFoRCe), KU Leuven,
Kasteelpark Arenberg 20, B-3001 Leuven, Belgium

ABSTRACT: To investigate the importance of substrate selectivity for xylanase functionality in bread making, the secondary
binding site (SBS) of xylanases from Bacillus subtilis (XBS) and Pseudoalteromonas haloplanktis was modified. This resulted in two
xylanases with increased relative activity toward water-unextractable wheat arabinoxylan (WU-AX) compared to water-extractable
wheat arabinoxylan, i.e., an increased substrate selectivity, without changing other biochemical properties. Addition of both
modified xylanases in bread making resulted in increased loaf volumes compared to the wild types when using weak flour.
Moreover, maximal volume increase was reached at a lower dosage of the mutant compared to wild-type XBS. The modified
xylanases were able to solubilize more WU-AX and decreased the average degree of polymerization of soluble arabinoxylan in
dough more during fermentation. This possibly allowed for additional water release, which might be responsible for increased
loaf volumes. Altered SBS functionality and, as a result, enhanced substrate selectivity most probably caused these differences.
KEYWORDS: xylanase, arabinoxylan, bread making, substrate selectivity, secondary binding site

■ INTRODUCTION
Endo-β-1,4-xylanases (EC 3.2.1.8), also referred to as xylanases,
are commonly used in bread making processes to enhance
dough manageability and bread quality. Although it is evident
that their functionality is the result of their capacity to
hydrolyze β-(1,4)-linkages in the backbone of arabinoxylan
(AX), their mode of action in bread making is not yet fully
understood, despite much progress over the past decades.1−5
It has been hypothesized that water-unextractable wheat AX
(WU-AX) negatively affects bread volume due to its strong
water-holding capacity, while water-extractable wheat AX (WE-
AX) and solubilized wheat AX (S-AX) with a high molecular
weight have a positive effect since they form highly viscous
aqueous solutions, stabilizing the dough foam structure.6−8
Consequently, xylanases with a high substrate selectivity, i.e.,
the ratio of the capacity to solubilize WU-AX over the capacity
to hydrolyze WE-AX, are desired in bread making. Such
endoxylanases preferentially solubilize WU-AX but hydrolyze
WE-AX and S-AX only to a minimal extent.9
To confirm this hypothesis, extensive research with xylanases
that differed in substrate selectivity was conducted.3,10−12
Disadvantageous for these studies was the use of xylanases of
different microbial origins which differed in more biochemical
properties than only their substrate selectivity. The use of such
xylanases made it difficult to investigate the influence of
substrate selectivity in bread making unambiguously.
Xylanase functionality in bread making is also influenced by
inhibitors. Three different types of xylanase inhibitors have
been identified in wheat flour: Triticum aestivum xylanase
inhibitor (TAXI), xylanase inhibiting protein (XIP), and
thaumatin-like xylanase inhibitor (TL-XI).13 Due to the
formation of enzyme−inhibitor complexes, inhibition sensitive
xylanases are less efficient than inhibition insensitive ones.
These enzymes need higher dosages to ensure optimal
performance.14 Finally, the temperature and pH activity profile
of xylanases determines their efficiency in bread making as
well.15
Recently, the existence of a secondary binding site (SBS)
situated on the surface of the structural unit was discovered in
several single domain enzymes belonging to glycoside hydrolase
families (GH) 8, 10, and 11.16−18 Several functions have been
proposed for these SBSs, including targeting of the enzyme
toward its substrate, guiding substrate into the active site
groove, substrate disruption, enhancing processivity, allosteric
regulation, passing on reaction products, and anchoring the
enzyme to the cell wall of its parent microorganism. These
functions correspond well to the functions which are ascribed
to carbohydrate-binding modules (CBMs) in modular
enzymes.19 Possibly, SBSs compensate for the lack of CBMs
in single domain enzymes.20 Site-directed mutagenesis to
modify this SBS resulted in an increased activity on water-
unextractable substrates compared to water-extractable
ones,21−23 hence effectively changing the substrate selectivity
of the enzyme.
In particular, the GH8 xylanase of Pseudoalteromonas
haloplanktis (XPH) and GH11 xylanase XynA of Bacillus
subtilis (XBS) are both single domain enzymes which contain
an SBS. XBS has a β jelly roll fold structure, which is often
compared to a partially closed right hand, with an SBS located
on the “knuckles” of the enzyme.24 XPH is a psychrophilic
enzyme with a (α/α)6-barrel structure (Figure 1).25 Since both
enzymes show good performance in bread making, they are
ideal tools to investigate the influence of substrate selectivity on
xylanase functionality in bread making.
Received: March 30, 2016
Revised: June 8, 2016
Accepted: June 10, 2016

© XXXX American Chemical Society A DOI: 10.1021/acs.jafc.6b01473

J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry
Figure 1. On the left, the overall structures of the xylanases from Bacillus subtilis (XBS) (PDB 2QZ3) (A) and Pseudoalteromonas haloplanktis (XPH)
(PDB 2B4F) (B) in complex with xylooligosaccharides indicate the presence of a secondary binding site (SBS) on the surface of the enzyme. On the
right, the amino acids responsible for binding of the oligosaccharides to the SBS are indicated for XBS (A) and XPH (B). All figures were drawn
using PyMOL (http://pymol.sourceforge.net/).
The aim of this study is to validate the above-mentioned
hypothesis that substrate selectivity determines, at least in part,
the intrinsic quality of xylanases for bread making. This is done
by selectively modifying the substrate selectivity of XBS and
XPH xylanases through modification of their SBS, and studying
the impact thereof on the bread making process. This way,
nearly identical xylanases are used, which differ only in
substrate selectivity and not in other biochemical properties.
Therefore, XBS and XPH were modified in their SBS and
added to dough and bread followed by evaluating their impact
on the AX population and aqueous extract viscosity of dough
and bread samples. These results were compared to those
obtained with the wild-type enzymes. We here report on the
outcome of this work.
■ MATERIALS AND METHODS
Materials. Escherichia coli cells, transformed with expression
plasmid pEXP5-CT-xyna, were available for heterologous expression
of XBS wild type (XBS WT) (UniProtKB P18429). A stop codon was
introduced in the plasmid after the last nucleotide encoding for the C-
terminal amino acid of the native protein (W185).23 For expression of
XPH, E. coli cells transformed with expression plasmid pEXP5-CT-xph
which contained a de novo synthesized XPH wild-type (XPH WT)
gene (GenBank AJ427921.1) were also available.22 A triple mutant
plasmid of XBS (G56A-T183A-W185A) and a double mutant plasmid
of XPH (W249A-Y315A), hereafter referred to as XBS 3A and XPH
2A, respectively, were constructed using a QuickChange site-directed
mutagenesis kit (Stratagene, La Jolla, CA, USA) as described by
Cuyvers et al.23 and were available in transformed E. coli cells. The
selection of amino acids for modification was based on the crystal
structures of XBS and XPH soaked with oligosaccharides.16,17 Amino
B DOI: 10.1021/acs.jafc.6b01473
Article
acids which were able to interact with the substrate by hydrophobic
stacking interactions or hydrogen bonds were replaced by alanine to
reduce their role in binding of substrate at the SBS (Figure 1).
Xylazyme AX tablets, liquid Azo-wheat AX, and xylooligosaccharides
up to xylohexaose (X6) were obtained from Megazyme (Bray, Ireland).
Xylazyme AX tablets and liquid Azo-wheat AX both contained highly
purified AX from wheat, while the xylooligosaccharides (purity >90%)
were prepared from birchwood xylan. Two European wheat flours, free
from additives, were used in bread making. Crousti flour was a
commercial bread wheat flour obtained from Dossche Mills (Deinze,
Belgium). Soft wheat cultivar Claire from Limagrain (Rilland, The
Netherlands) was conditioned to 16.0% moisture and subsequently
milled with a Bühler MLU-202 laboratory mill (Bühler AG, Uzwil,
Switzerland), yielding three break and three reduction fractions, which
were combined.26 Protein contents [% dry matter (dm), N × 5.7]
were 13.2% and 10.6% for Crousti and Claire flour, respectively,
determined using an adaptation of the AOAC Official Method27 to an
automated Dumas protein analysis system (Vario Max Cube,
Elementar, Hanau, Germany). Ash contents (% dm) were 0.55%
and 0.48%, measured according to Approved Method 08-01.01,28
while the total AX (% dm) and WE-AX contents (% dm) for Crousti
and Claire flour were 1.85% and 0.28% and 1.76% and 0.48%,
respectively. Fresh compressed baker’s yeast was from AB Mauri
(Merelbeke, Belgium). Sodium chloride and sucrose used in the bread
making trials were food grade. All chemicals, solvents, and reagents
were purchased from Sigma-Aldrich (Bornem, Belgium) and were
analytical grade, unless specified otherwise.
Recombinant Expression and Purification. The recombinant
expression of XBS, XPH, and their mutant variants in E. coli
BL21(DE3)pLysS cells was according to Van Craeyveld et al.30
Subsequently, the xylanases were purified by cation exchange
chromatography as described elsewhere.29,30 The enzyme yields of
recombinant XBS and XPH after purification were typically 50−90
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry
mg/L and 100−180 mg/L of culture, respectively. Xylanase purity was
verified by SDS−PAGE and silver staining performed on a
PhastSystem Unit (GE Healthcare, Uppsala, Sweden) according to
GE Healthcare separation technique file 110 and development
technique file 210, respectively.31 Protein concentration of the purified
enzymes was estimated by measuring the extinction at 280 nm in
triplicate with a Nanodrop-1000 spectrophotometer (Thermo Fisher
Scientific, Waltham, MA, USA) using the molar extinction coefficients
calculated with the ProtParam tool (http://expasy.org/tools/
protparam.html). These calculations were based on the amino acid
sequences of the wild-type and mutant xylanases. Errors in the
extinction at 280 nm due to modifications in the amino acid structure
were negligible.
Biochemical Characterization of Xylanases. Specific Activities
toward Different Substrates. The specific activities of the wild-type
and mutant xylanases were determined toward different substrates in
order to assess the relative preference for these substrates. All enzyme
activities were measured at dough pH, which is 5.5,32 in a 100 mM
sodium acetate buffer containing 0.50 mg/mL bovine serum albumin
(BSA). The preincubation and incubation temperature were 30 °C in
all assays, i.e., the dough fermentation temperature.
Specific xylanase activities against WU-AX and WE-AX were
estimated using the polymeric chromophoric substrates Xylazyme
AX and Azo-wheat AX, respectively, as described by Cuyvers et al.23
The bases for these measurements were changes in solubility in water
or in a water−ethanol mixture, respectively, after hydrolysis of the
chromophoric substrate by the xylanase. The color intensity of the
soluble fragments was then proportional to the xylanase activity. For
the determination of the specific activity on Xylazyme AX, an
appropriately diluted enzyme solution (1000 μL) (measurements in
the linear area of enzyme concentration vs extinction) was equilibrated
for 10 min. Then a Xylazyme AX tablet was added. After 60 min of
incubation, the reaction was terminated by addition of 10.0 mL of
1.0% (w/v) tris(hydroxymethyl)aminomethane, vigorous vortex
stirring, and immediate filtration through a MN 615 filter
(Macherey-Nagel, Düren, Germany). The extinction of the filtrate
was determined at 590 nm (E590) [Ultraspec III UV/vis spectropho-
tometer (Pharmacia Biotech, Uppsala, Sweden)] against a control,
prepared by incubating the enzyme solution without substrate. One
Xylazyme unit (Xyl-U) corresponds to the enzyme concentration
required to obtain an E590 of 1.0, under the conditions of the assay.
The specific activity on Azo-wheat AX was determined by first
preincubating an appropriately diluted enzyme solution and the liquid
substrate separately for 10 min. After addition of 500 μL of the
substrate to 500 μL of the enzyme solution, incubation was extended
for 60 min, after which the reaction was terminated by addition of 2.5
mL of ice cold ethanol and vigorous vortex stirring. The samples were
kept on ice for 10 min. After centrifugation for 10 min at 4000g and 4
°C in a Sigma 6-16K centrifuge (Sigma Zentrifugen, Osterode am
Harz, Germany), the E590 of the supernatant was measured against a
control, prepared by incubating the enzyme solution without substrate.
One Azo-wheat unit (Azo-U) was defined as the enzyme
concentration needed to increase E590 by 1.0 under the conditions
of the assay. The substrate selectivity factor (SSF), defined as the
relative preference of xylanases toward WU-AX or WE-AX, was
calculated as the ratio of the specific activity toward Xylazyme AX to
that toward Azo-wheat AX.
Specific xylanase activities against X6 and reaction products of this
hydrolysis were determined by high-performance anion-exchange
chromatography followed by pulsed amperometric detection. Sample
preparation and quantification of the formed hydrolysis products were
done as described by Cuyvers et al.23 One enzyme unit on X6 (X6-U)
was then defined as the concentration of enzyme needed for the
formation of 1.0 μM xylotriose.
Temperature and pH Dependency. Temperature dependency of
xylanase activity was determined by measuring the activity on
Xylazyme AX, under the conditions described above, with incubation
temperatures ranging from 10 to 80 °C with intervals of 10 °C. To
determine pH dependency of xylanase activity, enzyme dilutions were
made in potassium chloride/hydrogen chloride (20 mM, pH 2.0),
C DOI: 10.1021/acs.jafc.6b01473
Article
sodium acetate (100 mM, pH 3.0, 4.0, and 5.0), sodium phosphate (20
mM, pH 6.0, 7.0, 8.0, and 9.0), sodium carbonate (50 mM, pH 10.0
and 11.0), and potassium chloride/sodium hydroxide (20 mM, pH
12.0) buffers, all containing 0.50 mg/mL BSA. The enzyme activities
of these enzyme solutions on Xylazyme AX were measured at 30 °C as
described above.
Inhibition Sensitivity. Wheat flour was suspended in sodium acetate
buffer (100 mM) at pH 5.5 (mass ratio 1:10 flour:buffer). This
suspension was shaken for 30 min at 7 °C (Laboshake, VWR
International, Leuven, Belgium) and centrifuged (10 min, 4000g, 7
°C), after which the supernatant, containing xylanase inhibitors, was
isolated. Xylanase inhibition sensitivity was then measured by
incubating an appropriate xylanase dilution (measurements in linear
area of enzyme concentration vs extinction) (500 μL; 100 mM sodium
acetate buffer pH 5.5) for 30 min at 30 °C with the Crousti wheat
flour extract (500 μL). Subsequently, the xylanase activity on
Xylazyme AX was measured as described above against a control
made from wheat flour extract. Inhibition sensitivity was expressed as a
reduction (%) of xylanase activity.
Bread Making Trials. Dough pieces and bread loaves were
produced in triplicate at 10 g scale, according to the straight-dough
method.33 Flour (10.0 g, 14% moisture base), 0.53 g of fresh yeast, 1.5
g of sodium chloride, 6.0 g of sucrose, deionized water (4.86 mL for
Crousti and 4.20 mL for Claire), and 1.0 mL of enzyme solution
(dialyzed overnight against 100 mM sodium acetate buffer pH 5.5
(volume ratio 1:160 enzyme solution:buffer)) were mixed in a 10 g pin
mixer (National Manufacturing, Lincoln, NE, USA) during 4.0 and 2.5
min for Crousti and Claire flour, respectively. The dough baking
absorption and mixing times were based on Farinograph (Brabender,
Duisburg, Germany) and Mixograph (National Manufacturing,
Lincoln, NE, USA) analyses, respectively.34,35 The xylanase activities
of the dialyzed enzyme solutions were measured against Xylazyme AX
(as described above). For XBS and its mutant, dosages between 100
and 700 Xyl-U/kg of flour were used, while for XPH and its mutant,
dosages between 5 and 35 Xyl-U/kg of flour were used. For control
dough and bread, the enzyme solution was replaced by 1.0 mL of
sodium acetate buffer (100 mM, pH 5.5). The dough was fermented
for 90 min in a fermentation cabinet (National Manufacturing) at 30
°C and 90% relative humidity with intermediate punching at 52 and 77
min. After final punching at 90 min, the dough pieces were molded
and proofed for 36 min (30 °C and 90% relative humidity). Finally,
baking was performed in a rotary oven (National Manufacturing) for
13 min at 232 °C and the bread was immediately weighed after baking.
Loaf volume was measured 120 min after baking by rapeseed
displacement.36 Dough (immediately after mixing and proofing) and
bread samples were frozen in liquid nitrogen. After lyophilization, they
were ground with a laboratory mill (model A10, IKA-Werke GmbH
and Co. KG, Staufen, Germany) and sieved (⌀ = 250 μm) before
further analysis.
Preparation of Aqueous Extracts from Flour, Dough, and
Bread Samples. Aqueous extracts of flour, (fermented) dough, and
bread samples were prepared by suspending the samples (1.0 g) in
10.0 mL of potassium chloride/hydrogen chloride (20 mM) buffer at
pH 3.0. Under these conditions, acid hydrolysis did not occur and
enzymatic breakdown of AX was minimized (results not shown). The
suspensions were shaken for 30 min at 7 °C (Laboshake, VWR
International, Leuven, Belgium), and, after centrifugation (10 min,
4000g, 7 °C), the supernatants were analyzed immediately or frozen
until further analysis.
AX Levels and Compositions. The carbohydrate contents and
compositions obtained after hydrolysis of flour or aqueous extracts of
flour, dough, or bread samples, prepared as described in the previous
section, were determined by gas chromatography. Hydrolysis of flour
(10.0−15.0 mg) was performed in 2.0 M trifluoroacetic acid (TFA)
(5.0 mL), while aqueous extracts (2.5 mL) were hydrolyzed in 4.0 M
TFA (2.5 mL), both for 60 min at 110 °C. Reduction and acetylation
were performed with sodium borohydride and acetic anhydride,
respectively, according the procedure of Englyst and Cummings.37 To
measure reducing end xylose and arabinose contents, reduction was
performed preceding the hydrolysis.38 A combination of reduction and
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry Article

Table 1. Biochemical Characteristics of Wild-Type and Mutant Bacillus subtilis Xylanases (XBS WT and XBS 3A) and
Pseudoalteromonas haloplanktis Xylanases (XPH WT and XPH 2A)a
XBS WT XBS 3A XPH WT XPH 2A
GenBank accession no. P18429 AJ427921.1
modified amino acids G56A-T183A-W185A W249A-Y315A
GH 11 11 8 8
MM (kDa) 20.4 20.2 46.1 45.9
act. on xylohexaose (%)b 100 ± 5 a 96 ± 2 a 100 ± 3 A 115 ± 4 B
act. on Xylazyme AX (%)b 100 ± 2 a 43 ± 1 b 100 ± 6 A 155 ± 4 B
act. on Azo-wheat AX (%)b 100 ± 3 a 34 ± 1 b 100 ± 9 A 128 ± 5 B
SSF 8.3 ± 0.5 a 10.5 ± 0.4 b 10.1 ± 1.1 A 12.2 ± 0.3 B
Topt (°C) 50 a 50 a 30 A 30 A
pHopt 6−7 a 7 a 9 A 9 A
reduction of act. by inhibitors (%) 88 ± 2 a 90 ± 3 a 0±1 A 1±1 A
a
Values with the same small or capital letter for XBS and XPH, respectively, are not significantly different from each other (α = 0.05). bExpressed
relative to the activity of the wild-type enzyme (100%). The activities on xylohexaose (X6), Xylazyme AX, and Azo-wheat AX were for the wild-type
XBS 1 X6-U = 0.17 nM, 1 Xyl-U = 1.03 nM, and 1 Azo-U = 8.58 nM, respectively. For the wild-type XPH, 1 X6-U = 4.73 nM, 1 Xyl-U = 2.71 nM,
and 1 Azo-U = 27.28 nM.

Figure 2. Specific loaf volumes of the breads in which XBS (A, B) and XPH (C, D) wild types (XBS WT and XPH WT) and mutants (XBS 3A and
XPH 2A) were incorporated as a function of enzyme dosage. The specific loaf volumes are expressed relative to a control bread without xylanase
(fixed at 100%). Enzyme dosages are expressed in Xylazyme units per kg flour. Two types of wheat flour were used: Crousti (A, C) and Claire (B,
D). All values are averages of triplicate experiments, and error bars show the standard deviations. Values with the same small letter at different
enzyme dosages of one xylanase are not significantly different from each other (α = 0.05). Wild-type xylanases which significantly differ from the
mutant xylanase at the same dosage are indicated with an asterisk.

acetylation, without prior hydrolysis, was performed to determine free
xylose and arabinose. The formed alditol acetates (1.0 μL) were
separated on a Supelco SP-2380 polar column (30 m × 0.32 mm inner
diameter; 0.2 μm film thickness) (Supelco, Bellefonte, PA, USA) in an
Agilent chromatograph (Agilent 6890 series, Wilmington, DE, USA)
equipped with autosampler, splitter injection port (split ratio 1:20),
and flame ionization detector. The carrier gas was helium. Separation
was at 225 °C while injection and detection were at 270 °C. The total
AX and WE-AX content of flour and the AX content in dough and
bread extracts was 0.88 times the sum of the xylose and arabinose
contents. The total arabinose content was corrected for free arabinose
as well as arabinose originating from arabinogalactan-peptides
(AGP).39 The total xylose content was corrected for the presence of
free xylose. All contents were expressed on a dm base. The average
degree of polymerization (avDP) of the soluble AX was calculated as
the sum of the xylose and arabinose content divided by the reducing
end xylose content. The arabinose content was corrected for free

D DOI: 10.1021/acs.jafc.6b01473
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry Article

Figure 3. Levels of solubilized WU-AX in dough and bread samples supplemented with xylanases of Bacillus subtilis [XBS wild type and XBS mutant
(XBS 3A)] (300 Xylazyme units/kg flour) (A, B) and Pseudoalteromonas haloplanktis [XPH wild type and XPH mutant (XPH 2A)] (25 Xylazyme
units/kg flour) (C, D) after mixing, fermentation, and baking. Two types of flour were used: Crousti (A, C) and Claire (B, D) flour. No xylanase was
incorporated in control dough or bread. According to a Tukey test (α < 0.05), values with the same small letter at different bread making stages of
control bread or with the same xylanase are not significantly different. Values that are not significantly different for control bread or different
xylanases at one particular baking stage are indicated with the same capital letter.

arabinose and arabinose present in AGP, while the xylose and reducing
xylose content was corrected for free xylose.
Viscosity. The viscosity of extracts of flour, dough, and bread
samples (500 μL), was determined with a Brookfield DV II+
viscometer (Brookfield Engineering Laboratories Inc., Stoughton,
MA, USA) at 30 °C with a CP40 cone and a constant shear rate of 60
s−1. The specific viscosity was defined as the relative viscosity, i.e., the
ratio of the viscosity of the extract and that of a potassium chloride/
hydrogen chloride (25 mM; pH 3.0) buffer, minus one.3
Statistical Analysis. Specific activities toward Xylazyme AX and
Azo-wheat AX were determined in 5-fold, while all other analyses were
done in triplicate. All data were analyzed using statistical software JMP
Pro 11 (SAS Institute, Cary, NC, USA) to verify whether mean values
were significantly different at a difference level (α) of 0.05 using the
two-way ANOVA procedure.
an ideal substrate to measure differences in inherent activity.23
For XBS, mutations in the SBS had no effect on the activity on
X6. In contrast, the activity on X6 increased by 15% by mutating
the SBS of XPH WT (Table 1). The hydrolysis pattern of X6
was identical for all mutant xylanases and their wild-type
counterparts (results not shown).
On both Xylazyme AX and Azo-wheat AX, significant
differences in activity were revealed after mutating the SBS.
XBS 3A showed only 34% of the activity of XBS WT for Azo-
wheat AX, while for Xylazyme AX, XBS 3A activity was 43% of
that of the wild-type enzyme. In contrast to XBS, the activities
on both chromophoric substrates increased for XPH 2A
compared to its wild type. Especially the activity on Xylazyme

■
RESULTS
Biochemical Characterization of Xylanases. Biochem-
ical properties of XBS and XPH enzymes modified in their SBS
were compared to those of their corresponding wild-type
enzymes to investigate whether modifying the SBS of a xylanase
influences the SSF without affecting other biochemical
properties.
Specific Activity toward Different Substrates. Since X6 is
too small to reach the active site and SBS simultaneously, it is
E DOI: 10.1021/acs.jafc.6b01473
AX (155%) increased, while the activity increase on Azo-wheat
AX (128%) was less pronounced. Since the activity toward
water-unextractable Xylazyme AX was higher for both mutant
xylanases than toward water-extractable Azo-wheat AX, the SSF
increased by 26% and 21% for XBS and XPH, respectively
(Table 1).
Temperature and pH Dependency of Xylanase Activity.
The wild-type and modified XBS showed maximal xylanase
activity at 50 °C, while the optimal temperature was 30 °C for
the psychrophilic wild-type XPH and its mutant (Table 1).
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry Article

Figure 4. Degree of polymerization (avDP) of the soluble AX fraction in dough and bread samples enriched with xylanases of Bacillus subtilis [XBS
wild type and XBS mutant (3A)] (300 Xylazyme units/kg flour) (A, B) and Pseudoalteromonas haloplanktis [XPH wild type and XPH mutant (2A)]
(25 Xylazyme units/kg flour) (C, D) after mixing, fermentation, and baking. Two types of flour were used: Crousti (A, C) and Claire (B, D) flour.
To the control dough or bread, no xylanase was added. According to a Tukey test (α < 0.05), values with the same small letter at different bread
making stages of control bread or with the same xylanase are not significantly different. Values that are not significantly different for control bread or
different xylanases at one particular baking stage are indicated with the same capital letter.

Maximal activity was at pH 6.0 to 7.0 for XBS and at pH 9.0
for XPH, for both wild-type enzymes as well as their mutant
counterparts (Table 1).
Inhibition Sensitivity. The addition of a wheat flour extract
reduced the xylanase activity of the XBS wild type and mutant
88% and 90% respectively, compared to the activity measure-
ment in the absence of wheat flour extract. Neither XPH wild
type nor mutant was affected by inhibitors.
Functionality of XBS and XPH in Bread Making.
Changes in Dough and Bread Properties. The addition of
wild-type and mutant xylanases to the bread making process
influenced the manageability of the dough. After mixing, dough
supplemented with XPH enzymes felt drier compared to dough
supplemented with XBS enzymes. After fermentation, dough
supplemented with XPH enzymes became more sticky, while
the dough manageability of dough supplemented with XBS
enzymes improved. Dough manageability after 90 min
fermentation limited the enzyme dosages that can be used to
700 Xyl-U/kg flour for XBS enzymes and 35 Xyl-U/kg flour for
XPH enzymes, but no differences were observed between the
wild-type and mutant xylanases.
Addition of xylanases significantly increased specific loaf
volumes, for both flour types used (Figure 2). Only at 700 Xyl-
U/kg Crousti flour, the specific loaf volume obtained with XBS
F DOI: 10.1021/acs.jafc.6b01473
3A was significantly lower than with XBS WT. In bread making
with Claire flour, the maximal increase in specific loaf volume
was reached at a lower dosage for XBS 3A (300 Xyl-U/kg flour)
compared to XBS WT (500 Xyl-U/kg flour). Moreover, the
volume increase at optimal dosages was higher when XBS 3A
was used. For Crousti flour, no differences in specific loaf
volume were observed when using the wild-type or mutant
XPH, regardless of the dosage used (Figure 2C). Maximal
volume increases of 20.3% and 19.1% were obtained with XPH
WT and XPH 2A, respectively, at an enzyme dosage of 25 Xyl-
U/kg flour. Specific volumes of the Claire loaves supplemented
with XPH 2A were higher than those supplemented with XPH
WT, for all dosages tested (Figure 2D). Maximal volume
increases of 23.3% and 31.9% were obtained with XPH WT and
XPH 2A, respectively.
AX Properties during Bread Making. To evaluate changes in
solubilized WU-AX and avDP, the AX population was
monitored during different steps of the bread making process
(Figures 3 and 4). For this, enzyme dosages of 300 Xyl-U/kg
flour for XBS wild type and mutant and 25 Xyl-U/kg flour for
XPH wild type and mutant were used since these dosages
resulted in significant differences in volume of Claire bread
between wild-type and mutant xylanases (cf. previous section).
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry Article

Table 2. Solubilization of Water-Unextractable Arabinoxylan (WU-AX) and Decrease of the Average Degree of Polymerization
(avDP) during the Fermentation Stage in Crousti and Claire Bread Making Processesa
solubilization of WU-AX during fermentation (% of total WU-AX) decrease in avDP during fermentation
Crousti flour Claire flour Crousti flour Claire flour
control 14 ± 1 c 3±0 c 6±0 b 6±1 c
XBS WT 19 ± 3 b 7±4 b 4±5 b 12 ± 2 b
XBS 3A 29 ± 3 a 17 ± 4 a 11 ± 2 a 24 ± 2 a
control 14 ± 1 c 3±0 c 6±0 b 6±1 c
XPH WT 16 ± 3 b 12 ± 3 b 15 ± 4 a 31 ± 1 b
XPH 2A 26 ± 7 a 29 ± 3 a 20 ± 8 a 46 ± 1 a
a
Wild types and mutants of the xylanases of Bacillus subtilis (XBS) and Pseudoalteromonas haloplanktis (XPH) were added to both flour types. A
control dough without xylanase addition was also investigated. The results of the statistical analysis are shown per enzyme. Values with the same
small letter are not significantly different.

Figure 5. Specific viscosity of aqueous extracts of dough and bread samples relative to a potassium chloride/hydrogen chloride buffer (20 mM; pH
3.0). A wild-type and mutant xylanase of Bacillus subtilis [(XBS wild type and XBS mutant (3A)] (300 Xylazyme units/kg flour) (A, B) and
Pseudoalteromonas haloplanktis [XPH wild type and XPH mutant (2A)] (25 Xylazyme units/kg flour) (C, D) were incorporated in dough and bread
samples prepared with Crousti (A, C) and Claire (B, D) flour. To the control samples, no xylanase was added. According to a Tukey test (α < 0.05),
values with the same small letter at different bread making stages of control bread or with the same xylanase are not significantly different. Values that
are not significantly different for control bread or different xylanases at one particular baking stage are indicated with the same capital letter.

Solubilization of WU-AX. The level of WU-AX solubilization
in dough after mixing and fermentation was significantly higher
when xylanases were used (Figure 3). For Crousti as well as
Claire bread making, no differences in solubilized WU-AX
content after mixing were observed between incorporation of
the wild-type and mutant XBS. During the fermentation stage,
XBS 3A solubilized more WU-AX than its wild-type counter-
part (Table 2). Immediately after mixing, solubilization of WU-
AX in Claire flour had progressed less for XPH 2A compared to
G DOI: 10.1021/acs.jafc.6b01473
the wild-type counterpart. Again, the mutant XPH was able to
solubilize more WU-AX during fermentation compared to the
wild type. After baking, the amount of solubilized WU-AX was
decreased for all samples tested.
avDP of the Soluble AX Fraction. During mixing and
fermentation, the avDP of the soluble AX fraction of xylanase
supplemented doughs was significantly lower compared to
control dough samples (Figure 4). When XBS enzymes were
incorporated in Crousti bread, no differences between the wild-
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry
type and mutant xylanase were observed after mixing. During
fermentation, XBS WT was not able to further lower the avDP
of the soluble AX fraction and this in contrast to XBS 3A
(Table 2). When XBS was added to bread making with Claire
flour, the avDP of the soluble AX fraction was already lower
after the mixing stage when XBS WT was compared to XBS 3A.
During fermentation, the avDP of the soluble AX fraction
further decreased. Since this decrease was more outspoken for
XBS 3A, there was no significant difference anymore between
the wild-type and mutant XBS after fermentation. Furthermore,
the avDP of the soluble AX population increased after baking,
except when XBS WT was added to Crousti bread. For the
control breads, in contrast, avDP decreased after baking.
The results obtained with XPH were similar to those of XBS
(Figure 4). After mixing and fermentation in Crousti bread
making, no differences were observed between the wild-type
and mutant XPH. However, the avDP of the soluble AX
decreased more pronouncedly during fermentation when using
the mutant XPH (Table 2). After baking, the avDP of the
soluble AX fraction was significantly lower with XPH 2A than
with XPH WT. For Claire bread making, the avDP of the
soluble AX fraction after mixing was lower for the wild-type
XPH compared to the mutant. During fermentation, the
decrease in avDP was more outspoken for XPH 2A. The avDP
of the soluble AX population increased again after baking.
Extract Viscosity at Different Phases of Bread Making.
While XBS enzymes had no impact on extract viscosity after
mixing, Crousti dough supplemented with XBS enzymes
showed a significantly lower extract viscosity after fermentation
than control dough (Figure 5). No differences were observed
between XBS WT and XBS 3A. When XBS enzymes were used
in dough and bread samples made with Claire flour, no
differences were observed compared to the control.
Supplementation of XPH had no impact on extract viscosity
after mixing. Dough and bread supplemented with XPH
enzymes showed a significantly lower extract viscosity after
fermentation and baking than control dough and bread. No
differences in specific viscosity were noticed between addition
of the wild-type and mutant XPH.
■ DISCUSSION
Modifying the SBS of XBS and XPH by replacing aromatic
amino acids responsible for binding of the substrate by an inert
alanine residue resulted in an increased SSF and, hence, an
increased preference for WU-AX (Table 1). Other biochemical
properties were not influenced by these mutations. As shown
for XPH 2A, modifying of the SBS did not automatically
decrease the activity on polymeric substrates as the activity of
the mutant XPH on Xylazyme AX and Azo-wheat AX was
increased by 55% and 28%, respectively, compared to XPH
WT. With the wild-type XPH, the substrate was probably
bound too tightly to the SBS for optimal activity. Modifying
this position could therefore allow a better positioning or faster
throughput of the substrate with respect to the active site. As
mentioned by Cuyvers et al.,23 the increased activity of XPH 2A
toward X6 was probably the result of subtle positional change of
residues located in the active site after modification of the SBS.
Since modifying the SBS provided xylanases that only differed
in SSF, they were ideal tools to investigate the specific influence
of substrate selectivity in bread making.
Equivalent units (Xyl-U) of wild-type and mutant xylanases
were added to bread making processes based on Crousti and
Claire flour. Differences in dough manageability were observed
H DOI: 10.1021/acs.jafc.6b01473
Article
after addition of XBS and XPH. Dough supplemented with
XPH was dry after mixing but became more sticky during
fermentation, while the inverse was observed for XBS. This can
be explained by the inhibition sensitivity of the xylanases: XPH
is not sensitive to inhibitors and keeps working during the
entire process, while the activity of XBS, which is inhibited by
TAXI,40 most probably decreases rapidly during bread making.
Differences in inhibition sensitivity also explain why XPH was
added in lower dosages than XBS. Modifying the SBS had no
impact on inhibition sensitivity.
While little if any significant difference in specific loaf volume
was observed between wild-type and mutant xylanases in bread
making with Crousti (Figure 2), an additional volume increase
of 7.1% and 8.6% was obtained after supplementation of the
mutant XBS and mutant XPH, respectively, at their optimal
dosages in Claire bread making. The stronger gluten network of
dough made with Crousti flour probably masked possible
additional improvements of the mutant xylanases. In contrast,
when using the weak Claire flour, the effects of incorporation of
xylanases with different SSF were more pronounced.
Since equivalent units (Xyl-U) of wild-type and mutant
xylanases were used in bread making, it was expected that also
equivalent amounts of WU-AX were solubilized during the
different bread making stages. The amount of solubilized WU-
AX after mixing was indeed similar for wild-type and mutant
xylanases (Figure 3). The increase in solubilized WU-AX was
accompanied by a decrease in the avDP of the soluble AX
population (Figure 4). It is possible that the fragments derived
from solubilizing WU-AX had a lower avDP than the native
WE-AX present in flour. Additionally, native WE-AX and S-AX
might have been hydrolyzed into smaller fragments, which
would also result in a lower avDP. These fragments would have
a minimal chain length of five xylose molecules since XPH and
XBS are not able to hydrolyze substrates with a degree of
polymerization lower than six.41 However, since XBS 3A and
XPH 2A had a lower activity toward WE-AX and S-AX than the
wild-type enzymes (Table 1), we had expected that
incorporation of the mutant xylanases in bread making would
result in differences in avDP of the soluble AX fraction. This
was not the case, and, based on this, it can be concluded that
the lower avDP of solubilized WU-AX was probably responsible
for the decrease in overall avDP.
During fermentation, an additional amount of WU-AX was
solubilized (Figure 3 and Table 2). Especially for Crousti bread
making, this was not least caused by the xylanase activity of the
flour itself, as shown for the control dough. For xylanase
supplemented dough samples, the amount of additionally
solubilized WU-AX was lower compared to what was
solubilized earlier during the mixing stage. Both mutant
xylanases solubilized WU-AX to a larger extent and decreased
the avDP of soluble AX more strongly compared to their
respective wild types. This is remarkable since the same activity
on water-unextractable substrate of wild-type and mutant
xylanases was added. In the past, several functions have already
been described for SBSs.20 Possibly, a number of these
functions were slightly enhanced after modification and ensured
the xylanase to become more active. This resulted in a more
convenient hydrolysis for the mutant xylanases compared to the
wild type.
XPH degraded S-AX more severely than XBS as deduced
from the avDP of soluble AX after fermentation (Figure 4). A
reason for the lower hydrolyzing capacity of XBS is probably its
inhibition sensitivity toward TAXI, as this inhibitor was present
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry
in excess in wheat flour. It has already been shown that
inhibition-insensitive xylanases show a stronger degradation of
S-AX and WE-AX than inhibition-sensitive ones.11,14
WE-AX with a high molecular weight have a high viscosity
forming capacity and are able to stabilize the liquid films
surrounding the gas cells in bread dough.42 Despite differences
in solubilization degree between wild-type and mutant
xylanases, no differences in specific viscosity were observed
(Figure 5). Liquid films in bread dough are indeed more
concentrated than the extracts used for measuring specific
viscosity. Alternatively, it is possible that differences in bread
loaf volume are only determined by WU-AX solubilization
degree and accompanying water release and not by the bulk
viscosity of aqueous phase. Roels and co-workers43 showed that
increasing the baking absorption up to 10% above the
Farinograph baking absorption results in higher loaf volumes.
This could also explain the increased specific loaf volume when
mutant xylanases were added to bread making. Since Claire
flour is a weak flour type, the impact of additional water release
due to degradation of WU-AX might be more pronounced.
During the early baking phase, it was expected that XPH
would still hydrolyze AX, which would then result in an
additional decrease of the avDP of the soluble AX fraction. A
decrease in solubilized WU-AX level was, however, observed
(Figure 3). This indicates that part of the previously solubilized
AX turned unextractable again, probably due to cross-linking of
AX molecules with other AX molecules or flour components
and/or due to physical inclusion in the gelatinized starch matrix
since no amylase was added during the extraction.3,44,45 For
Claire bread making, it appears that AX molecules with a low
molecular weight preferentially turned unextractable again since
the avDP was increased after baking (Figure 4). On the one
hand, smaller AX molecules are more mobile and less sensitive
to steric hindrance, which makes them able to cross-link more
easily with other constituents. On the other hand, the fact that
small molecules are more mobile renders them also easier to
extract.
In conclusion, this study shows that increasing the preference
of a xylanase for water-unextractable substrates compared to
water-extractable ones, i.e., increasing its substrate selectivity,
can enhance its functionality in bread making. Due to increased
WU-AX solubilization during fermentation, these xylanases
ensure that more of the WU-AX that is detrimental in bread
making is converted into S-AX. This possibly allows for
additional water release, responsible for an increase in specific
loaf volume and/or a lower optimal dosage when weak flour
was used. With the experimental approach followed here,
viscosity differences could not be pinpointed as driving
mechanism for differences between wild-type and mutant
enzymes. These conclusions, of course, have to be seen in the
right perspective. The functionality of xylanases in bread
making is determined by more biochemical properties than
only their substrate selectivity.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: +32 (0) 16 32 19 17. Fax: +32 (0) 16 32 19 97. E-
mail: christophe.courtin@biw.kuleuven.be.
Funding
S.L. acknowledges the Agency for Innovation through Science
and Technology in Flanders (IWT-Vlaanderen, Brussels,
Belgium) for financial support. A.P. wishes to acknowledge
the Research FoundationFlanders (FWO-Vlaanderen, Brus-
I DOI: 10.1021/acs.jafc.6b01473
Article
sels, Belgium) for a position as postdoctoral researcher. This
research is also part of the Methusalem program “Food for the
future” (2007−2014) at the KU Leuven.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
Dr. Emmie Dornez (Mauri Research, The Netherlands) is
gratefully thanked for fruitful discussions.
■
ABBREVIATIONS USED
α, difference level; AGP, arabinogalactan-peptides; avDP,
average degree of polymerization; AX, arabinoxylan; Azo-U,
Azo-wheat unit; BSA, bovine serum albumin; CBMs,
carbohydrate-binding modules; dm, dry matter; E590, extinction
at 590 nm; GH, glycoside hydrolase family; S-AX, solubilized
AX; SBS, secondary binding site; SSF, substrate selectivity
factor; TAXI, Triticum aestivum xylanase inhibitor; TFA,
trifluoroacetic acid; TL-XI, thaumatin-like xylanase inhibitor;
WE-AX, water-extractable AX; WT, wild type; WU-AX, water-
unextractable AX; X6, xylohexaose; X6-U, xylohexaose unit;
XBS, xylanase XynA of Bacillus subtilis; XBS 3A, G56A-T183A-
W185A mutant of xylanase XynA of Bacillus subtilis; XIP,
xylanase inhibiting protein; XPH, xylanase of Pseudoalteromonas
haloplanktis; XPH 2A, W249A-Y315A mutant of xylanase of
Pseudoalteromonas haloplanktis; Xyl-U, Xylazyme unit
■
REFERENCES
(1) Rouau, X.; El-Hayek, M. L.; Moreau, D. Effect of an enzyme
preparation containing pentosanases on the bread-making quality of
flours in relation to changes in pentosan properties. J. Cereal Sci. 1994,
19, 259−272.
(2) Courtin, C. M.; Delcour, J. A. Arabinoxylans and endoxylanases
in wheat flour bread-making. J. Cereal Sci. 2002, 35, 225−243.
(3) Courtin, C. M.; Gelders, G. G.; Delcour, J. A. Use of two
endoxylanases with different substrate selectivity for understanding
arabinoxylan functionality in wheat flour breadmaking. Cereal Chem.
2001, 78, 564−571.
(4) Butt, M. S.; Tahir-Nadeem, M.; Ahmad, Z.; Sultan, M. T.
Xylanases and their applications in baking industry. Food Technol.
Biotechnol. 2008, 46, 22−31.
(5) Jaekel, L. Z.; da Silva, C. B.; Steel, C. J.; Chang, Y. K. Influence of
xylanase addition on the characteristics of loaf bread prepared with
white flour or whole grain wheat flour. Ciênc. Technol. Aliment. 2012,
32, 844−849.
(6) Izydorczyk, M. S.; Biliaderis, C. G. Cereal arabinoxylans: advances
in structure and physicochemical properties. Carbohydr. Polym. 1995,
28, 33−48.
(7) Meuser, F.; Suckow, P. Non-starch polysaccharides. In Chemistry
and Physics of Baking; Blanchard, J. M. V., Frazier, P. J., Galliard, T.,
Eds.; The Royal Society of Chemistry: London, U.K., 1986; pp 42−61.
(8) Courtin, C. M.; Roelants, A.; Delcour, J. A. Fractionation−
reconstitution experiments provide insight into the role of
endoxylanases in bread-making. J. Agric. Food Chem. 1999, 47,
1870−1877.
(9) Moers, K.; Celus, I.; Brijs, K.; Courtin, C. M.; Delcour, J. A.
Endoxylanase substrate selectivity determines degradation of wheat
water-extractable and water-unextractable arabinoxylan. Carbohydr.
Res. 2005, 340, 1319−1327.
(10) Moers, K.; Courtin, C. M.; Brijs, K.; Delcour, J. A. A screening
method for endo-β-1,4-xylanase substrate selectivity. Anal. Biochem.
2003, 319, 73−77.
(11) Dornez, E.; Verjans, P.; Arnaut, F.; Delcour, J. A.; Courtin, C.
M. Use of psychrophilic xylanases provides insight into the xylanase
functionality in bread making. J. Agric. Food Chem. 2011, 59, 9553−
9562.
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry
(12) Verjans, P.; Dornez, E.; Delcour, J. A.; Courtin, C. M. Selectivity
for water-unextractable arabinoxylan and inhibition sensitivity govern
the strong bread improving potential of an acidophilic GH11
Aureobasidium pullulans xylanase. Food Chem. 2010, 123, 331−337.
(13) Fierens, E.; Gebruers, K.; Courtin, C. M.; Delcour, J. A.
Xylanase inhibitors bind to nonstarch polysaccharides. J. Agric. Food
Chem. 2008, 56, 564−570.
(14) Trogh, I.; Sørensen, J. F.; Courtin, C. M.; Delcour, J. A. Impact
of inhibition sensitivity on endoxylanase functionality in wheat flour
breadmaking. J. Agric. Food Chem. 2004, 52, 4296−4302.
(15) Verjans, P. Functionality of extremophilic xylanases in bread
making. Ph.D. dissertation; K.U. Leuven: Leuven, Belgium, 2010; 185
pp.
(16) De Vos, D.; Collins, T.; Nerinckx, W.; Savvides, S. N.;
Claeyssens, M.; Gerday, C.; Feller, G.; Van Beeumen, J.
Oligosaccharide binding in family 8 glycosidases: Crystal structures
of active-site mutants of the β-1,4-xylanase pXyl from Pseudoaltermonas
haloplanktis TAH3a in complex with substrate and product.
Biochemistry 2006, 45, 4797−4807.
(17) Vandermarliere, E.; Bourgois, T. M.; Rombouts, S.; Van
campenhout, S.; Volckaert, G.; Strelkov, S. V.; Delcour, J. A.; Rabijns,
A.; Courtin, C. M. Crystallographic analysis shows substrate binding at
the − 3 to + 1 active-site subsites and at the surface of glycoside
hydrolase family 11 endo-1,4-β-xylanases. Biochem. J. 2008, 410, 71−
79.
(18) Ludwiczek, M. L.; Heller, M.; Kantner, T.; McIntosh, L. P. A
secondary xylan-binding site enhances the catalytic activity of a single-
domain family 11 glycoside hydrolase. J. Mol. Biol. 2007, 373, 337−
354.
(19) Guillén, D.; Sánchez, S.; Rodríguez-Sanoja, R. Carbohydrate-
binding domains: multiplicity of biological roles. Appl. Microbiol.
Biotechnol. 2010, 85, 1241−1249.
(20) Cuyvers, S.; Dornez, E.; Delcour, J. A.; Courtin, C. M.
Occurrence and functional significance of secondary carbohydrate
binding sites in glycoside hydrolases. Crit. Rev. Biotechnol. 2012, 32,
93−107.
(21) Moers, K.; Bourgois, T.; Rombouts, S.; Beliën, T.; Van
Campenhout, S.; Volckaert, G.; Robben, J.; Brijs, K.; Delcour, J. A.;
Courtin, C. M. Alteration of Bacillus subtilis XynA endoxylanase
substrate selectivity by site-directed mutagenesis. Enzyme Microb.
Technol. 2007, 41, 85−91.
(22) Cuyvers, S.; Dornez, E.; Delcour, J. A.; Courtin, C. M. The
secondary substrate binding site of the Pseudoalteromonas haloplanktis
GH8 xylanase is relevant for activity on insoluble but not soluble
substrates. Appl. Microbiol. Biotechnol. 2011, 92, 539−549.
(23) Cuyvers, S.; Dornez, E.; Rezaei, M. N.; Pollet, A.; Delcour, J. A.;
Courtin, C. M. Secondary substrate binding strongly affects activity
and binding affinity of Bacillus subtilis and Aspergillus niger GH11
xylanases. FEBS J. 2011, 278, 1098−1111.
(24) Collins, T.; Gerday, C.; Feller, G. Xylanases, xylanase families
and extremophilic xylanases. FEMS Microbiol. Rev. 2005, 29, 3−23.
(25) Van Petegem, F.; Collins, T.; Meuwis, M.; Gerday, C.; Feller,
G.; Van Beeumen, J. The structure of a cold-adapted family 8 xylanase
at 1.3 Å resolution: Structural adaptations to cold and investigation of
the active site. J. Biol. Chem. 2003, 278, 7531−7539.
(26) Delcour, J. A.; Vanhamel, S.; De Geest, C. Physico-Chemical
and Functional Properties of Rye Nonstarch Polysaccharides. I.
Colorimetric Analysis of Pentosans and Their Relative Monosacchar-
ide Compositions in Fractionated (Milled) Rye Products. Cereal Chem.
1989, 66, 107−111.
(27) AOAC. Official Methods of Analysis, 16th ed.; Association of
Official Analytical Chemists: Washington, DC, 1995.
(28) AACC International. Approved Methods of Analysis, 11th ed.;
Method 08-01.01. AshBasic Method. Approved Nov 3, 1999;
AACC International: St. Paul, MN, 2000.
(29) Pollet, A.; Vandermarliere, E.; Lammertyn, J.; Strelkov, S. V.;
Delcour, J. A.; Courtin, C. M. Crystallographic and activity-based
evidence for thumb flexibility and its relevance in glycoside hydrolase
family 11 xylanases. Proteins: Struct., Funct., Genet. 2009, 77, 395−403.
J DOI: 10.1021/acs.jafc.6b01473
Article
(30) Van Craeyveld, V.; Dornez, E.; Holopainen, U.; Selinheimo, E.;
Poutanen, K.; Delcour, J. A.; Courtin, C. M. Wheat bran AX properties
and choice of xylanase affect enzymic production of wheat bran-
derived arabinoxylan-oligosaccharides. Cereal Chem. 2010, 87, 283−
291.
(31) Laemmli, U. K. Cleavage of Structural Proteins during the
Assembly of the Head of Bacteriophage T4. Nature 1970, 227, 680−
685.
(32) Jayaram, V. B.; Cuyvers, S.; Lagrain, B.; Verstrepen, K. J.;
Delcour, J. A.; Courtin, C. M. Mapping of Saccharomyces cerevisiae
metabolites in fermenting wheat straight-dough reveals succinic acid as
pH-determining factor. Food Chem. 2013, 136, 301−308.
(33) Shogren, M. D.; Finney, K. F. Bread-making test for 10 grams of
flour. Cereal Chem. 1984, 61, 418−423.
(34) AACC International. Approved Methods of Analysis, 11th ed.;
Method 54-21.02. Rheological Behaviour of Flour by Farinograph:
Constant Flour Weight Procedure. Approved Nov 3, 1999; AACC
International: St. Paul, MN, 2000.
(35) AACC International. Approved Methods of Analysis, 11th ed.;
Method 54-40.02. Mixograph Method. Approved Nov 3, 1999; AACC
International: St. Paul, MN, 2000.
(36) Vanhamel, S.; Vandenende, L.; Darius, P. L.; Delcour, J. A. A
volumeter for breads prepared from 10-grams of flour. Cereal Chem.
1991, 68, 170−172.
(37) Englyst, H. N.; Cummings, J. H. Simplified method for the
measurement of total non-starch polysaccharides by gas-liquid
chromatography of constituent sugars as alditol acetates. Analyst
1984, 109, 937−942.
(38) Courtin, C. M.; Van den Broeck, H.; Delcour, J. A.
Determination of reducing end sugar residues in oligo- and
polysaccharides by gas−liquid chromatography. J. Chromatogr. A
2000, 866, 97−104.
(39) Loosveld, A.-M. A.; Grobet, P. J.; Delcour, J. A. Contents and
Structural Features of Water-Extractable Arabinogalactan in Wheat
Flour Fractions. J. Agric. Food Chem. 1997, 45, 1998−2002.
(40) Gebruers, K.; Brijs, K.; Courtin, C. M.; Fierens, K.; Goesaert,
H.; Rabijns, A.; Raedschelders, G.; Robben, J.; Sansen, S.; Sørensen, J.
F.; Van Campenhout, S.; Delcour, J. A. Properties of TAXI-type
endoxylanase inhibitors. Biochim. Biophys. Acta, Proteins Proteomics
2004, 1696, 213−221.
(41) Pollet, A. Functional and structural analysis of glycoside
hydrolase family 8, 10 and 11 xylanases with focus on Bacillus subtilis
xylanase A. Ph.D. thesis; KU Leuven: Belgium, 2010; 207 pp.
(42) Gan, Z.; Ellis, P. R.; Schofield, J. D. Gas cell stabilisation and gas
retention in wheat bread dough. J. Cereal Sci. 1995, 21, 215−230.
(43) Roels, S. P.; Cleemput, G.; Vandewalle, X.; Nys, M.; Delcour, J.
A. Bread Volume Potential of Variable-Quality Flours with Constant
Protein Level As Determined by Factors Governing Mixing Time and
Baking Absorption Levels. Cereal Chem. 1993, 70, 318−323.
(44) Wang, M.; Hamer, R. J.; van Vliet, T.; Oudgenoeg, G.
Interaction of water extractable pentosans with gluten protein: effect
on dough properties and gluten quality. J. Cereal Sci. 2002, 36, 25−37.
(45) Wang, M.; van Vliet, T.; Hamer, R. J. How gluten properties are
affected by pentosans. J. Cereal Sci. 2004, 39, 395−402.
J. Agric. Food Chem. XXXX, XXX, XXX−XXX
 Journal of Biotechnology 306S (2019) 100013

Contents lists available at ScienceDirect

Journal of Biotechnology
journal homepage: www.elsevier.com/locate/jbiotec

Rhizosecretion of a cisgenic lectin by genetic manipulation of Tepary bean

$
plants (Phaseolus acutifolius)
Dania Martínez-Alarcón a,1, * , Alejandra Mora-Avilés b,1 , Arantxa Espinoza-Núñez b ,
Luz M Serrano Jamaica b , Andrés Cruz-Hernández a , Angelina Rodríguez-Torres a ,
José L Castro-Guillen c, Alejandro Blanco-Labra c , Teresa García-Gasca a, *
a
Facultad de Ciencias Naturales, Universidad Autónoma de Querétaro, Santiago de Querétaro, 76230, Querétaro, Mexico
b
Departamento de Biotecnología de Plantas, Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP) Campo Experimental Bajío, Mexico
c
Departamento de Bioquímica y Biotecnología de Plantas, CINVESTAV Unidad Irapuato, Irapuato 36821, Guanajuato, Mexico

A R T I C L E I N F O A B S T R A C T

Keywords: Tepary bean (Phaseolus acutifolius) lectin fraction (TBLF) has been shown to specifically bind and induce cell death of
Genetic modification different types of cancer cells and also has exhibited an effect on early colon tumorigenesis. However, the
Phaseolus acutifolius development of a pharmaceutical formula is not possible yet because the production process is expensive and slow
Cisgenic lectin
and provides low yields. Therefore, the purpose of the present work was to develop a strategy to produce one
Rhizosecretion
Tepary bean
bioactive lectin by rhizosecretion through root exudates on genetically modified plants. Amplification of Tepary
bean transcripts was performed using degenerate primers, and the products obtained were sequenced. Multiple
alignments of sequences led to elucidating one of the lectins present in TBLF. Its coding sequence was flanked by an
N-terminal secretion signal peptide and a 6xHis-tail. This construction was introduced into P. acutifolius plants using
Agrobacterium tumefaciens to subsequently carry out the in vitro growth of the plants. When roots grew, plants were
transferred to hydroponic conditions and root exudates were analyzed. Results showed the presence of a
glycosylated cisgenic lectin with biological activity, confirming that the strategy followed provides an alternative for
the synthetic production and purification of this lectin.

1. Introduction
Lectins are proteins that interact with carbohydrates, either free or
bounded to cell membranes (Sharon and Lis, 2002). In plants, these
proteins are found mainly in seeds and storage tissues, where they reach
up to 10% of the total proteins, whereas in leaves, roots, and stems, their
concentration is lower (Liener, 1976; Etzler, 1985; Chrispeels and
Raikhel, 1991). Lectins from some species, mainly legumes, have shown
an ability to selectively adhere to tumor-cell oligosaccharides and inhibit
their growth (Díaz et al., 1999; De Mejía and Prisecaru, 2005; Ferriz
Martínez et al., 2010).
Tepary bean (Phaseolus acutifolius), which grows in the north of
Mexico and in the south of the United States and its seeds contain high
concentration of lectins. By different chromatographic and electropho-
retic methods it was possible to obtain the Tepary bean lectin fraction
(TBLF), which has a cytotoxic effect on colon cancer cells (García-Gasca
et al., 2012) and, under oral administration, displays low toxicity and
inhibits early colon tumorigenesis in rats by apoptosis (Ferriz-Martínez
et al., 2015; Moreno-Celis et al., 2017). The partial sequences of the two
main lectins of TBLF (Lectin_A and Lectin_B) were identified by Torres-
Arteaga et al. (2016) through mass spectrometry sequencing. Also
identified in these lectins was the presence of N-glycans, which were
shown to be necessary for their function.
Extraction and purification of the Tepary bean lectins has been carried
out through the use of a wide range of strategies. Unfortunately, these
strategies commonly provide low yields (García-Gasca et al., 2012;
Torres-Arteaga et al., 2016). Therefore, it is necessary to develop a system
to obtain higher quantities of these lectins at low cost, without
compromising its functionality. Transformation facilities and modern
plant tissue culture techniques allow genetically modified plants to

Abbreviations: IRM, induction and regeneration medium; PGIP, Phaseolus vulgaris polygalacturonase; PBS, phosphate buffer solution; cisTBL1, cisgenic Tepary bean
lectin; TBL, Tepary bean lectin; TBLF, Tepary bean lectin fraction.
$
This article was originally published in Journal of Biotechnology: X. Journal of Biotechnology: X is now discontinued and the article is republished here for the reader’s
convenience. For citation purposes, please use the publication details of this article; Journal of Biotechnology, 306S.
* Corresponding authors.
E-mail addresses: dania_667@hotmail.com (D. Martínez-Alarcón), mora_alejandra@yahoo.com (A. Mora-Avilés), alejandroblancolabra@gmail.com
(A. Blanco-Labra), tggasca@uaq.edu.mx (T. García-Gasca).
1
These authors share as first authors.

http://doi.org/10.1016/j.btecx.2019.100013
Received 26 April 2019; Received in revised form 5 August 2019; Accepted 4 September 2019
Available online 25 September 2019
0168-1656/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
D. Martínez-Alarcón et al.
function as suitable bioreactors for the production of recombinant
proteins and/or industrially relevant metabolites (Giddings et al., 2000).
Plants provide a suitable environment for protein folding and disulfide
bond formation, multimeric protein assembly, and post-translational
modifications (Borisjuk et al., 1999). However, the biggest disadvantage
of producing proteins in genetically modified plants is that their
purification involves tissue maceration which destroys the production
system and the release of proteins of interest at all stages of production
(including partially folded or partially processed proteins), decreasing the
efficiency of the process (Gaume et al., 2003; Cabanes-Macheteau et al.,
1999). To overcome these drawbacks, in 1999 Borisjuk et al. (Borisjuk
et al., 1999) proposed, for the first time, a root-secretion system. This
technique couples signaling peptides for the flux of functional proteins to
a hydroponic medium through the roots of genetically modified plants,
providing higher yields than conventional extraction methods. As the
root-secretion system allows the expression of proteins from the original
species, the correct folding and post-translational processing of complex
proteins is assured, providing an alternative manufacturing platform that
increases performance and simplifies the purification procedure (Drake
et al., 2009). Therefore, this could also represent an excellent alternative
method of preserving the native N-glycans of leguminous lectins. As we
use one of the lectin genes from the same plant, the resulting product will
be a cisgenic protein (Telem et al., 2013). Therefore, the objective of this
study was to transform Tepary bean plants in order to obtain a cisgenic
lectin by rhizosecretion and to evaluate the functionality of the mature
protein by hemagglutination analysis.
2. Materials and methods
2.1. Tepary bean lectin fraction
Tepary bean lectin fraction was obtained by purification from Tepary
bean seeds, using Tris-HCl pH 8.0 as extraction solvent at 4
C, followed by
sequentialprecipitationwithammoniumsulfateat40%and60%saturation.
Finally, the protein extract was separated by molecular weight exclusion
chromatography using G-75 Sephadex (Torres-Arteaga et al., 2016).
2.2. Phaseolus acutifolius specimens
Plants were germinated from Tepary bean seeds (P. acutifolius)
purchased in the local market of Hermosillo, Sonora, Mexico. One specimen
was deposited at the “Dr. J. Rzedowski” herbarium of the Natural Sciences
Faculty of the Autonomous University of Querétaro for its identification.
2.3. Preparation of the sample
Tepary bean seeds were previously washed with water, cultivated for
3 weeks and watered every third day. Subsequently, stems and leaves
were pulverized in a mortar with liquid nitrogen. Total RNA was extracted
using Trizol1 and incubated in the presence of DNase. The cDNA was
generated by using the RevertAidTM H Minus First Strand Kit (# K1631
Fermentas, Ontario, Canada), according to the manufacturer's instruc-
tions. Subsequently, a polymerase chain reaction (PCR) was performed
with the degenerate primers F1 (50 - ATGGCTTCCTCCAACTTCTCCA-30 ),
R1 (50 -TAGAGGATTTGGTTGAGGACGA-30 ), F2 (5-CCTCTTCCTTCCGC
TTCTCAC-3) y R2 (5-AAGAGAGGAGGTCGTT CGTTTC-3). Those primers
were designed based on conserved sequences of 36 leguminous lectins
from multiple origins. Amplification conditions were: 5 min at 95
C for
one cycle; 1 min at 94
C, 1 min at 52
C, 2 min at 72
C for 35 cycles; and
7 min at 72
C for final extension.
2.4. Sequencing
The PCR product was cloned into a pCR2.1-TOPO vector (Invitrogen
Life Technologies, Carlsbad, CA, USA). Then, it was inserted into
2
Journal of Biotechnology 306S (2019) 100013
Escherichia coli Top10 chemocompetent cells by heat shock (Sambrook
et al., 1989). Transformed colonies were inoculated in LB medium with
100 mg L1 of ampicillin as the selection agent and incubated at 37
C for
24 h. Finally, plasmids were purified by miniprep method (Birnboim and
Doly, 1979) and sequenced at the National Laboratory of Agricultural,
Medical and Environmental Biotechnology (LANBAMA, San Luis Potosi,
Mexico).
2.5. Cloning
A construction coupling the P. vulgaris polygalacturonase (PGIP)
signal peptide at the 5-end, and 6xHis-tail at the 3-end of the Tepary bean
lectin gene was synthesized by GenScript USA Inc. Once the construction
was obtained, it was released from the commercial vector pUC57 by
digestion carried out with the BamHI restriction enzyme (Fermentas,
Ontario, Canada), and incubated at 37
C from 3 to 5 h. The sequence was
introduced in a previously digested vector pBI121, by a ligation carried
out with the T4 ligase (Fermentas, Ontario, Canada), incubating at 16
C
for 16 h. Subsequently, the result of the previous ligation was cloned into
chemocompetent cells of E. coli Top 10 by heat shock (Sambrook et al.,
1989). The transformed colonies were recovered for growth on plates of
LB medium with 100 mg/mL of ampicillin as selection agent, incubated at
37
C for 24 h, and finally the plasmids were extracted by the miniprep
method (Birnboim and Doly, 1979).
2.6. Modeling
To visualize the lectin molecular modeling, the program RASTOP
(Molecular Graphics Visualization Software) was used (Sayle and
Bernstein, 2007). The comparative modeling of the mRNA-deduced
protein sequence of the Tepary bean lectin was made using the following
five homology-based comparative modeling websites:
1 FUGUE-Homstrad: Sequence-structure homology recognition (Wil-
liams et al., 2001).
2 3D-JIGSAW—Comparative modeling (Bates et al., 2001).
3 CPHmodels 3.2 Server (Nielsen et al., 2010).
4 ESyPred3D Web Server 1.0 (Lambert et al., 2002).
5 Geno3D: Automatic comparative molecular modelling of protein
(Combet et al., 2002).
2.7. Agrobacterium tumefaciens transformation
Competent cells of A. tumefaciens strain GV2260 were transformed
with pBI121 (Promega, Madison, WI, USA) by the heat shock method
(Sambrook et al., 1989). Transformed cells were cultivated on solid LB
medium with 100 mg L1 of kanamycin as selection agent and incubated
at 28
C for 24–48 h. A single Agrobacterium colony was selected and
grown in a liquid culture to exponential growth (OD600 = 0.8).
2.8. Phaseolus acutifolius transformation
Tepary bean seeds were disinfected by the gas-chlorine method
composed by hydrochloric acid and commercial bleach in a ratio of 5:0.16
(v/v) for 8–10 h. Subsequently, bean seeds were soaked in sterile distilled
water for 17 h under aseptic conditions. After this time, embryos were
removed from the seeds and cultivated on petri dishes with induction and
regeneration medium (IRM), which consisted of solid Gamborg B5
medium (Gamborg et al., 1968), amended with 2% sucrose, 100 mg L1 of
myo-inositol, 1 mg L1 of pyridoxine, mgL1 of thiamine, and 10 mg L1
of BAP for 5 days at 25  2
C and a photoperiod of 16 h light and 8 h
darkness. Hypocotyls were obtained by dissecting apical and radicular
meristems from five-day-old embryos. Liquid IRM was combined with A.
tumefaciens (O.D.600 = 0.8) in a 5:1 (v/v) ratio. Hypocotyls were
incubated at room temperature for 10 min in this solution with gentle
movements. Liquid was discarded and explants were placed in co-
D. Martínez-Alarcón et al. Journal of Biotechnology 306S (2019) 100013

Fig. 1. Multiple alignment of codifying sequences for leguminous lectins and primers design. A) Alignment of the 50 end of 36 leguminous lectins; green boxes indicate F1/
F2 primers and the location where they hybridize the consensus sequence. B) Alignment of the 30 end of 36 leguminous lectins; orange boxes indicate the complementary
sequences of R1/R2 primers and the location where they hybridize the consensus sequence. Conserved residues are indicated in blue, and the accession number of each
sequence is on the left.

cultivation medium, which consisted of IRM amended with 200 mM of
acetosyringone remaining in this medium for 5 days in the same
environmental conditions previously mentioned. Hypocotyls were
transferred to IRM solid medium amended with 300 mg L1 of timentin
for Agrobacterium elimination for 6 days under the same conditions.
Subsequently, hypocotyls were transferred to a selection medium which
consisted of IRM medium with 300 mg L1 of timentin and 50 mg L1 of
kanamycin added to it (Sigma Chemical Co., St. Louis, MO, USA).
Explants were incubated for bud formation and shoot differentiation,
with successive transfers to fresh medium every 6 days. Independent
shoots were transferred to tall petri dishes with selection medium, where
they remained until reaching 1 cm height. Later, they were transferred to
Magenta1 boxes with selection medium without growth regulators to
promote elongation and root formation. Once the kanamycin-resistant
plants fully formed their rooting system, they were transferred to liquid
selection medium to establish the hydroponic conditions for rhizosecre-
tion. After seven days of incubation in liquid selection medium, an aliquot
was collected and lyophilized.
2.9. Periodic acid Schiff staining
Protein concentration was determined by the Bradford method
(Bradford, 1976). To verify the presence of glycosylations of the cisgenic
lectin, 20 mL and 10 mL of crude root-exudate were run by SDS-PAGE,
adding 40 mg native lectin as a positive control. Subsequently, they were
transferred to a nitrocellulose membrane, applying a current of 0.2 A for
60 min, and the membrane was stained using periodic acid–Schiff reagent
(Sigma Chemical Co., St. Louis, MO, USA) according to the manufacturer's
instructions.

3
D. Martínez-Alarcón et al. Journal of Biotechnology 306S (2019) 100013

Fig. 2. Identification of lectins in transcripts of Phaseolus acutifolius. A) Amplification of sequences with different combinations of degenerate primers; molecular size
marker (1). F1/R1 (2), F2/R2 (3), F1/R2 (4), F2/R2 (5), amplicon purified from band excision of agarose electrophoresis (6). B) Polymerase chain reaction (PCR) of
transformed colonies, molecular size marker (1 and 16), PCR screening of E. coli colonies after transformation (2–15 and 17–21). C) Multiple alignment of the amplified
sequence from P. acutifolius (cisTBL1), the deduced sequence for P. acutifolius lectin (accession number: AAA82181.1) and the peptides obtained by mass spectrometry for
Lectin-A (LA_SequencedPeptides) and Lectin-B (LB_SequencedPeptides). Matches are indicated in red and mismatches are indicated in blue and black.

2.10. Immunoblotting
To verify the presence of cisgenic lectin in root exudates of the
genetically modified plants, 25 mL of each sample was placed in
nitrocellulose membranes previously activated with 1X TBS. The
membranes were left to dry for 15 min and then blocked for 15 min
with anti-histidine primary monoclonal antibodies coupled to alkaline
phosphatase (Novus Biological Europe, Science Park, CAM, UK), in a
1:1000 dilution. Finally, the membranes were revealed using 3,3-
diaminobenzidine (Sigma Chemical Co, St. Louis, MO, USA).
2.11. Purification of the cisgenic protein from the culture media
The supernatant was diluted with phosphate buffer 4X pH: 7.4 in a 1:3
ratio and flowed through 5 mL nickel columns (Invitrogen Carlsbad, CA,
USA). Subsequently, the columns were washed with an ascending
imidazole gradient until the desired product was recovered.
2.12. Hemagglutination assay
Glutaraldehyde fixed type A + erythrocytes (0.5% in PBS) (Turner
and Liener, 1975) were used for testing the hemagglutination activity of
rTBL1 (García-Gasca et al., 2012; Jaffé, 1980), protein concentration was
determined by Bradford method (Bradford, 1976). Then, TBLF (8.2 mL,
324.7 mg P/mL) was used as a positive control and PBS (100 mL) was used
as a negative control. The tested samples were 100 mL of non-transformed
radicular exudate and 100 mL of genetically modified radicular exudate
(26.76 mg P/mL). Agglutination was determined using an inverted
microscope after incubation at 37
C for 2 h.

4
D. Martínez-Alarcón et al. Journal of Biotechnology 306S (2019) 100013

Fig. 3. Modeling the structure of the deduced lectin. Models generated by EsyPred3D (A), CPH-models (B), Geno3D (C), 3D-JIGSAW (D). X: Front view. Y: Top view. The
order of the polypeptide sequence is indicated by a gradual change of color from blue (N-terminal end) to red (C-terminal end).

Fig. 4. Matched model of the predicted protein structure of lectins phylogenetically related to TBL1. A) Matched model without glycans. The level of conservation is
indicated in color scale from blue to red, blue being the most conserved regions and red the less conserved ones. B) Matched model with glycans: calcium ions (yellow),
manganese ion (green), and glycosylation (red). X. Front view. Y: Top view.

3. Results were sequenced independently and, based on multiple alignments, it was

3.1. Analysis of lectin transcripts from P. acutifolius
mRNA amplification of a Tepary bean lectin was performed using
degenerated primers designed to bind conserved sequences of leguminous
lectins (Fig. 1). The amplification of lectin transcripts was specific, and the
size ranged from 740 to 894 bp. The largest product was purified and
introduced in a TOPO vector for subsequent transformation of E. coli
(Fig. 2A). Nineteen colonies resulted positive from transformation, and
plasmid purification was performed for three of them (Fig. 2B). Vectors
determined a sequence of 831 nucleotides corresponded to a polypeptide
chain of 277 amino acids, later referred to as Tepary bean lectin 1 (TBL1).
This sequence displayed high identity with La and LB proteins obtained by
Torres-Arteaga et al. (2016), and with a P. acutifolius lectin previously
deduced by Mirkov et al. (1994) (accession number: AAA82181.1). The
coverage percentage of identity was 97.83%, with 6 amino acid changes in
position 17 P changed to A, in position 139 R changed to K, in position 180
GQ changed to VN, in position 228 R changed to S, and in position 239 T
changed to S (Fig. 2C). In contrast with the amplified sequence, none of the
sequences reported by Torres-Arteaga et al. (2016) displayed the N-

5
D. Martínez-Alarcón et al. Journal of Biotechnology 306S (2019) 100013

Fig. 5. Phaseolus acutifolius transformation. A) Synthesized nucleotide construction scheme. B–F) Development stages of P. acutifolius regeneration. A) Four-day-old
embryo germinated in induction and regeneration medium. B) Dissection of embryo to obtain the hypocotyl. C) Development of de novo organogenic buds. D)
Differentiation of kanamycin resistant shoots. E) Development of kanamycin resistant complete plants.

terminal peptide, indicating that this signal peptide is cleaved during
protein processing. No additional lectin transcripts were identified in the
transcripts; however, the possibility that they are present is not ruled out.
3.2. TBL1 structure
Conformational modeling of the deduced polypeptide sequence
showed that TBL1 has a large proportion of beta sheets interconnected
by small regions of helices (Fig. 3). These data are consistent with
previous reports of legume lectins and provide a good explanation of the
high resistance that TBLF lectins display against degradation in the
gastrointestinal tract (Ferriz-Martínez et al., 2015).
Multiple alignment of polypeptide lectin sequences from different
legumes revealed that the glycosylation sites are highly conserved
between species and are distant from the C-terminus and N-terminus,
none of these sites coincide with differential residues between the TBL1
sequence and the peptides reported by Torres-Arteaga et al. (2016) for
TBLF lectins (Fig. 4), suggesting that they are conserved. The
carbohydrate-binding pocket in all models was found adjacent to calcium
and manganese divalent cations, which are also necessary for carbohy-
drate binding activity. Those structures also revealed that, in the
monomeric conformation, the C- and N- terminals are exposed,
suggesting that their possible modification should not compromise its
quaternary structure.
3.3. Genetic construction design
Based on the structural studies of these models, a genetic construction
was designed for the synthetic expression of TBL1. The signal peptide of
PGIP was coupled to the 50 end of the TBL nucleotide sequence and a
6xHis-tail was coupled to the 30 end followed by a stop codon. The
construction was flanked by cutting sites for BamHI and codon
optimization was done in order to avoid non-specific internal cleavage
by restriction enzymes. Insert and vector were digested with BamHI, and
the construction was introduced into pBI121 under regulation of the
CaMV 35 s promoter and NOS terminator (Fig. 5A).

6
D. Martínez-Alarcón et al. Journal of Biotechnology 306S (2019) 100013

Fig. 6. Characterization and purification of the cisgenic lectin in root exudates. A) SDS-PAGE stained by periodic acid–Schiff (PAS) reaction; (1) Positive control (10 mg of
TBL), (2 and 3) Root exudates from transformed Tepary bean plants (20 and 10 ml, respectively), (4) Negative control, radicular exudate of non-transformed Tepary bean
plants. B) Identification of cisgenic proteins by immunoblotting. On the left: sample (25 mL of root exudate from transgenic plants); on the right: negative control (25 mL of
root exudate from non-transformed plants). C) Purification of cisTBL1 from root exudates. (1) Elution with 200 mM phosphate buffer of Imidazole (10 mL, 1.2 mg) after
desalination, (2) Root exudate from a transformed plant prior to purification (20 mL), (3) Positive control (10 mg TBLF). D) Biological activity of root exudates. The first
row shows inverted microscopy images with a 10X magnification objective, whereas the second row shows these same regions with 20X magnifications. Column 1) PBS
negative control, Column 2) Root exudates from non-transformed plants, Column 3) Root exudates of genetically modified plants, Column 4) Positive control of native
lectin (TBLF).

3.4. Genetic transformation of Phaseolus acutifolius
Tepary bean germination rate was between 90% and 95%. Five days
after germination, embryos were dissected and cultivated in germination
medium, and apical and radical meristems were cut off to obtain the
hypocotyl explant (Fig. 5B and C). After 10 days, the first direct
organogenic buds were formed on the hypocotyl in selection medium
(Fig. 5D). Kanamycin-resistant differentiated shoots were separated from
the original explant and transferred to an elongation medium for further
growth and root formation (Fig. 5E and F). Once primary roots developed,
the plants were transferred to a liquid medium.
3.5. Characterization and purification of cisTBL1
Seven days after transferring the plants to hydroponic conditions, a
sample of crude culture medium was analyzed by periodic acid Schiff
staining on SDS-PAGE revealing the presence of a unique glycoprotein in
root exudates of genetically modified Tepary bean plants, whereas no
evidence of a similar band in root exudates from wild-type plants was
found (Fig. 6A). Subsequently, the presence of cisTBL1 in the root
exudates of genetically modified plants was confirmed by a dot-blot assay
with anti-histidine antibodies, but no presence of this protein was
detected in root exudates of non-transformed plants (Fig. 6B). After
purification by nickel affinity chromatography, a single protein was
obtained, which corresponded in size to the shorter lectin present in the
TBLF (Fig. 6C). These data also evidenced the absence of phase shifting
during transcription, as the His-tail introduced at the C-end of the protein
was successfully translated. The cisTBL1 yield ranged between 12.21 and
20.97 mg of cisgenic protein per gram of dry root.
3.6. Agglutination activity of root exudates of genetically modified plants
The agglutination test using human A + erythrocytes revealed that
the root exudates of transgenic plants presented similar biological activity
with respect to TBLF (11,958.15 and 12,018.59 AU/mg protein), whereas
no agglutination in root exudates from non-transformed plants was
observed (Fig. 6D).
4. Discussion
High identity between the amplified transcript with LA and LB
peptides indicates that TBL1 is one of the main components of TBLF. This
information was later confirmed by the characterization and purification
of the cisgenic product, where it was found that TBL1 corresponds to the
lowest molecular weight lectin observable in the electrophoretic profile of
TBLF. Although it is true that cisTBL1 has a slightly larger size than its
native counterpart, the alteration of the electrophoretic migration pattern
could be attributed to the addition of 6xHis-tail at its C-terminal end.
Structural models allow us to propose a three-dimensional structure of
TBL1 and its prosthetic groups, also allowing the identification of

7
D. Martínez-Alarcón et al.
common motifs between lectins, and to identify possible interferences
where modifications are proposed. On the other hand, immunoblot
analysis indicates that cisTBL1 was successfully secreted through the root
exudates of genetically modified plants, and the Schiff staining assay
confirms that this protein was glycosylated, indicating that PGIP signal
peptide successfully redirected the protein through the secretory
pathway.
Finally, the agglutination analysis confirmed that the excreted
product maintained the carbohydrate-binding properties of the native
lectins of Tepary bean. Native lectins of TBLF have the ability to
agglutinate A + erythrocytes. As lectins that are not properly folded lose
their carbohydrate-binding activity, cisTBL1 agglutination suggest a
proper protein folding. Therefore, it is concluded that the proposed
method allows the successful production and secretion of the bioactive
lectin towards root exudates of genetically modified plants, providing a
continuous production platform that avoids the inconveniences associat-
ed with the purification of lectins from seeds. The fact that the
agglutination activity was conserved suggests that there is no change
in the lectin molecular structure; however, this needs to be confirmed.
Further studies will focus on optimizing this procedure, as well as on
characterizing the cisgenic protein produced and its effects on cancer
cells.
Author contributions
Dania Martínez-Alarcón, Arantxa Espinoza-Núñez, Luz M Serrano
Jamaica performed the experiments, Alejandra Mora-Avilés advised
plant culture experiments, Andrés Cruz-Hernández and Angelina
Rodríguez-Torres advised genetic transformation experiments, José L
Castro-Guillen advised the molecular modeling, Alejandro Blanco-Labra
and Teresa García-Gasca directed and advised the complete work and
managed the resources.
Declaration of Competing Interest
The authors declare no financial or commercial conflict of interest.
Acknowledgments
We would like to thank F. Josué López Martínez and Elizabeth
Mendiola-Olaya for their technical assistance, and Consejo Nacional de
Ciencia y tecnología (CONACYT) Programa de Fortalecimiento de la
Calidad Educativa (PFCE) Ciencia Básica (CB-2014-01-241181) and
PFCE program for their financial support.
References
Bates, P.A., Kelley, L.A., MacCallum, R.M., Sternberg, M.J.E., 2001. Enhancement of
protein modelling by human intervention in applying the automatic programs 3D-
JIGSAW and 3D-PSSM. Proteins 5, 39.
Birnboim, H.C., Doly, J., 1979. A rapid alkaline extraction procedure for screening
recombinant plasmid DNA. Nucl. Acids Res. 7, 1513.
Borisjuk, N.V., Borisjuk, L.C., Logendra, S., Petersen, F., Gleba, Y., Raskin, I., 1999.
Production of recombinant proteins in plant root exudates. Nat. Biotechnol. 17, 466–
469.
Bradford, M.A., 1976. Rapid and sensitive method for the quantification of microgram
quantities of protein utilizing the principle of protein-dye-binding. Anal. Biochem. 72,
248–252.
8
Journal of Biotechnology 306S (2019) 100013
Cabanes-Macheteau, M., Fitchette-Lainé, A.C., Loutelier-Bourhis, C., Lange, C., Vine, N.D.,
Ma, P., Lerouge, J.K.C., Faye, L., 1999. N-Glycosylation of a mouse IgG expressed in
transgenic tobacco plants. Glycobiol 9, 365.
Chrispeels, M.J., Raikhel, N.V., 1991. Lectins, lectin genes, and their role in plant defense.
Plant Cell 3, 1.
Combet, C., Jambon, M., Deléage, G., Geourjon, C., 2002. Geno3D: Automatic comparative
molecular modelling of protein. Bioinformatics 18, 213.
De Mejía, E.G., Prisecaru, V.I., 2005. Lectins as bioactive plant proteins: a potential in
cancer treatment. Crit. Rev. Food Sci. Nutr. 45, 425.
Díaz, P.H., González, O.M., De Pablos Vélez, Y.R., Ganem Báez, C., 1999. Aplicaciones de
las lectinas. Rev. Cuba. Hematol. Inmunol. Hemoter. 15, 91.
Drake, P.M., Barbi, T., Sexton, A., McGowan, E., Stadlmann, J., Navarre, C., Paul, M.J.,
Ma, J.K., 2009. Development of rhizosecretion as a production system for recombinant
proteins from hydroponic cultivated tobacco. FASEB J. 10, 3581.
Etzler, M.E., 1985. Plant lectins: molecular and biological aspects. Ann. Rev. Plant Physiol.
36, 209.
Ferriz Martínez, R.A., Torres-Arteaga, I.C., Blanco-Labra, A., García-Gasca, T., 2010. The
role of plant lectins in cancer treatment. In: Mejia-Vazquez, C., Navarro-Fos, S. (Eds.),
New Approaches in the Treatment of Cancer, Vol. 1. Nova Science Publishers, Mexico,
pp. 71–89 Ch. 4.
Ferriz-Martínez, R.A., García-García, K., Torres-Arteaga, I., Rodríguez-Méndez, A.J.,
Guerrero-Carrillo, M.J., Moreno-Celis, U., Ángeles-Zaragoza, M.V., Blanco-Labra, A.,
Gallegos-Corona, M.A., Robles-Álvarez, J.P., Mendiola-Olaya, E., Andrade-
Montemayor, H.M., García, O.P., García-Gasca, T., 2015. Tolerability assessment of a
lectin fraction from Tepary bean seeds (Phaseolus acutifolius) orally administered to rats.
Toxicol. Reports. 2, 63.
Gamborg, O.L., Miller, R.A., Ojima, E., 1968. Nutrient requirements of suspension cultures
of soybean root cells. Cell Res. 50, 151.
García-Gasca, T., Hernández-Rivera, E., López-Martínez, J., Castañeda, A.L., Yllescas-
Gasca, L., Rodríguez-Méndez, A.J., Mendiola-Olaya, E., Castro-Guillén, J.L., Blanco-
Labra, A., 2012. Effects of Tepary bean (Phaseolus acutifolius) protease inhibitor and
semipure lectin fractions on cancer cells. Nutr. Cancer 64, 1269.
Gaume, A., Komarnytsky, S., Borisjuk, N., Raskin, I., 2003. Rhizosecretion of recombinant
proteins from plant hairy roots. Plant Cell Rep. 21, 1188.
Giddings, G., Allison, G., Brooks, D., Carter, A., 2000. Transgenic plants as factories for
biopharmaceuticals. Nat. Biotechnol. 18, 1151.
Jaffé, W.G., 1980. Hemagglutinins. Toxic Constituents of Plant Foodstuffs. Academic Press,
New York, NY, USA, pp. 73–102.
Lambert, C., Leonard, N., De Bolle, X., Depiereux, E., 2002. ESyPred3D: Prediction of
proteins 3D structures. Bioinformatics 18, 1250.
Liener, I., 1976. Phytohemagglutinins. Ann. Rev. Plant Physiol. 27, 291.
Mirkov, T.E., Wahlstrom, J.M., Hagiwara, K., Finardi-Filho, F., Kjemtrup, S., Chrispeels,
M.J., 1994. Evolutionary relationships among proteins in the phytohemagglutinin
arcelin-alpha amylase inhibitor family of the common bean and its relatives. Plant Mol.
Biol. 26, 1103.
Moreno-Celis, U., López-Martínez, J., Blanco-Labra, A., Cervantes-Jiménez, R., Estrada-
Martínez, L.E., García-Pascalin, A.E., Guerrero-Carrillo, M.J., Rodríguez-Méndez, A.J.,
Mejía, C., Ferríz-Martínez, R.A., García-Gasca, T., 2017. Phaseolus acutifolius lectin
fractions exhibit apoptotic effects on colon cancer: preclinical studies using
dimethilhydrazine or azoxi-methane as cancer induction agents. Molecules 22, 1670.
Nielsen, M., Lundegaard, C., Lund, O., Petersen, T.N., 2010. CPHmodels-3.0 - Remote
homology modeling using structure guided sequence profiles. Nucl. Acids Res. 576.
Sambrook, J., Fritsh, E., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual. Cold
Spring Harbor Laboratory Press, NY.
Sayle, R., Bernstein, H.J., 2007. ValadonP. Molecular Graphics Visualisation Software.
Version 2.1. .
Sharon, N., Lis, H., 2002. How proteins bind carbohydrates: lessons from legume lectins. J.
Agric. Food Chem. 50, 6586–6591.
Telem, R.S., Wani, S.H., Singh, N.B., Nandini, R., Sadhukhan, R., Bhattacharya, S.,
Mandal, N., 2013. Cisgenics - a sustainable approach for crop improvement. Curr.
Genomics 14 (7), 468–476.
Torres-Arteaga, I., Mendiola-Olaya, E., García-Gasca, T., Ángeles-Zaragoza, M.V., García-
Santoyo, V., Castillo, J.A., Aguirre, C., Phinney, B., Blanco-Labra, A., 2016.
Characterization of two non-fetuin-binding lectins from Tepary bean (Phaseolus
acutifolius) seeds with differential cytotoxicity on colon cancer cells. J. Glycobiol. 5, 1.
Turner, R.H., Liener, I.E., 1975. The use of glutaraldehyde-treated erythrocytes for assaying
the agglutinating activity of lectins. Anal. Biochem. 68 (2), 651–653.
Williams, M.G., Shirai, H., Shi, J., Nagendra, H.G., Mueller, J., Mizuguchi, K., Miguel, R.
N., Lovell, S.C., Innis, C.A., Deane, C.M., Chen, L., Campillo, N., Burke, D.F., Blundell,
T.L., de Bakker, P.I., 2001. Sequence-structure homology recognition by iterative
alignment refinement and comparative modeling. Proteins 5, 92.
 Food Research International 147 (2021) 110487

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Review

Genetic engineering of yeast, filamentous fungi and bacteria for terpene

production and applications in food industry
Zijian Liang, Hang Zhi, Zhongxiang Fang, Pangzhen Zhang *
School of Agriculture and Food, Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Parkville, VIC 3010, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Terpenes are a major class of natural aromatic compounds in grapes and wines to offer the characteristic flavor
Terpenes and aroma, serving as important quality traits of wine products. Saccharomyces cerevisiae represents an excellent
Saccharomyces cerevisiae cell factory platform for large-scale bio-based terpene production. This review describes the biosynthetic path­
Escherichia coli
ways of terpenes in different organisms. The metabolic engineering of S. cerevisiae for promoting terpene
Fungi
biosynthesis and the alternative microbial engineering platforms including filamentous fungi and Escherichia coli
Metabolic engineering
Food applications are also elaborated. Additionally, the potential applications of the terpene products from engineered microor­
ganisms in food and beverage industries are also discussed. This review provides comprehensive information for
an innovative supply way of terpene via microbial cell factory, which could facilitate the development and
application of this technique at the industrial scale.

terpenes exclusively through the MEP pathway. In the fermentation

1. Introduction
Terpenes and their derivative terpenoids are a major class of natural
aromatic compounds present in both free and glycosidically bound
forms in many edible plants and their fruits, to offer a strong odor which
may protect the plants by deterring herbivores and by attracting pred­
ators and parasites of herbivores (Liang, Fang, et al., 2020). They are
structurally related to isoprene (C5H8) and classified as monoterpenes
(C10), norisoprenoids (C13), sesquiterpenes (C15), diterpenes (C20),
triterpenes (C30) and tetraterpenes (C40) based on the number of
isoprene units (Li, Howell, Fang, & Zhang, 2020). Terpenoids are
modified from terpenes and usually contain oxygen atoms. This article
uses the term terpenes to represent terpenes and terpenoids for easier
discussion. Terpenes derived from grapes, especially monoterpenes, are
responsible for various sensory characteristics of wine products, such as
citrus, fruity and floral flavors (Li et al., 2020). Additionally, terpenes
are engaged in the pharmaceutical industry due to their multiple bio­
activities, including antioxidative, antimicrobial and anti-inflammatory
properties (Pardo, Rico, Gil, & Orejas, 2015). From the chemical aspect,
terpenes are mainly derived from isopentenyl pyrophosphate (IPP) and
its isomer dimethylallyl pyrophosphate (DMAPP) via mevalonate (MVA)
and methylerythritol-phosphate (MEP) pathways (Zhang et al., 2016).
Both MVA and MEP pathways exist in plant cells. However, yeast and
other fungi only harbor the MVA pathway, while bacteria produce
process such as winemaking, many S. cerevisiae strains can enzymati­
cally hydrolyze the bound terpenes in grape must and thus release the
free volatile counterparts in wine (Liang, Fang, et al., 2020). S. cerevisiae
strains can also utilize the MVA pathway to yield various monoterpenes
but the amount is low due to the deficiency of monoterpene synthases
(MTSs) (Pardo et al., 2015).
Solvent extraction from plant materials is the conventional process
for commercial terpene production, although with multiple demerits
such as time-consuming and labor-intensive process, low yield and
environment unfriendliness (Liang, Zhang, & Fang, 2020). In recent
years, terpene production via microbial cell factories has emerged as a
promising alternative to the traditional method (Paramasivan & Mut­
turi, 2017). Microorganism serves as an efficient terpene-producing
platform for being fast-growing, land-saving and controllable. Yeast
especially S. cerevisiae with a clear genetic background, is the well-
established host to express heterologous terpene synthase (TPS) genes.
Similar to plant cells, S. cerevisiae contains an inherent endomembrane
system for microsomal enzymes to stand on (Sun, Zhao, & Li, 2019). A
rational strain improvement by metabolic engineering can enhance
terpene production in S. cerevisiae. Overexpression of HMG-CoA (3-hy­
droxy-3-methylglutaryl-CoA) reductase (HMGR) variant, ERG20 over­
expression, ERG9 downregulation, integration of extra IDI1 gene and
deletion of ROX1 constitute the main engineering strategies to modify
the MVA pathway in S. cerevisiae (Paramasivan & Mutturi, 2017). In

* Corresponding author.
E-mail address: pangzhen.zhang@unimelb.edu.au (P. Zhang).

https://doi.org/10.1016/j.foodres.2021.110487
Received 26 March 2021; Received in revised form 21 May 2021; Accepted 23 May 2021
Available online 31 May 2021
0963-9969/© 2021 Elsevier Ltd. All rights reserved.
Z. Liang et al. Food Research International 147 (2021) 110487

Nomenclature CyaA adenylate cyclase

cAMP cyclic adenosine monophosphate
IPP isopentenyl pyrophosphate CRP cAMP receptor protein
DMAPP dimethylallyl pyrophosphate CPPS ent-copalyl diphosphate synthase
MVA mevalonate KS ent-kaurene synthase
MEP methylerythritol-phosphate AA-CoA acetoacetyl-CoA
TPS terpene synthase MVAP mevalonate-5-phosphate
MTS monoterpene synthase MVAPP mevalonate-5-pyrophosphate
STS sesquiterpene synthase IP isopentenyl phosphate
DTS diterpene synthase DXP 1-deoxy-D-xylulose-5-phosphate
HMG-CoA 3-hydroxy-3-methylglutaryl-CoA CD-ME 4-(cytidine 5′ -diphospho)-2-C-methyl-D-erythritol
HMGR HMG-CoA reductase MEC 2C-methyl-D-erythritol-2,4-cyclodiphosphate
HMGS HMG-CoA synthase AACT acetoacetyl-CoA thiolase
GPP geranyl diphosphate MVK mevalonate kinase
GPPS GPP synthase PMK MVAP kinase
FPP farnesyl diphosphate MDD MVAPP decarboxylase
FPPS FPP synthase MPD MVAP decarboxylase
GGPP geranylgeranyl diphosphate IPK IP kinase
GGPPS GGPP synthase DXS DXP synthase
IDS isoprenyl diphosphate synthase DXR DXP reductoisomerase
G3P glyceraldehydes-3-phosphate MCT CD-ME cytidylyltransferase
IDI isopentenyl diphosphate isomerase CMK CD-ME kinase
MCT 2-C-methyl-D-erythritol 4-phosphate cytidylyltransferase MDS MEC synthase
CD-ME 4-(cytidine 5′ -diphospho)-2-C-methyl-D-erythritol HDS HMBPP synthase
HMBPP 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate HDR HMBPP reductase
NADH nicotinamide adenine dinucleotide GPPS GPP synthase
NADPH nicotinamide adenine dinucleotide phosphate FPPS FPP synthase
ADA acetaldehyde dehydrogenase (acylating) GGPPS GGPP synthase
xPK xylulose-5-phosphate specific phosphoketolase SS squalene synthase
PTA phosphotransacetylase KO-KAH kaurene oxidase-kaurenoic acid hydroxylase
UAS upstream activating cis-element CPR cytochrome P450 reductase
SUE sterol uptake enhancement UDP uridine diphosphate
NPP neryl diphosphate UGT UDP glucose-dependent glucosyltransferase
NPPS NPP synthase PKS polyketide synthase
ABC ATP-binding cassette PT prenyltransferase
MIC minimum inhibitory concentration FAD flavin adenine dinucleotide
GRAS generally recognized as safe FMO FAD-dependent monooxygenase
CRISPR clustered regularly interspaced short palindromic repeat AT acetyltransferase
sgRNA single guide RNA MT methyltransferase
AMA1 autonomous maintenance in Aspergillus SDR short-chain dehydrogenase/reductase
HA hemagglutinin TS taxadiene synthase
ATMT Agrobacterium tumefaciens-mediated transformation GMC glucose-methanol-choline
LB left border GRAS generally recognized as safe
RB right border

addition, construction of orthogonal pathway also exhibits great po­
tential for efficient monoterpene production in S. cerevisiae (Ignea et al.,
2019). Nowadays, nearly 100 TPS genes have been identified in fungi
(Zhang et al., 2020) and many of them are potential heterologous genes
for the functional TPSs expression in S. cerevisiae, indicating fungi as a
reservoir of heterologous TPS genes. Filamentous fungi especially
Aspergillus oryzae can act as an alternative platform for terpene biosyn­
thesis. Likewise, E. coli is also a common terpene-producing host, which
is highly efficient and more tolerant of monoterpene toxicity (Dugar &
Stephanopoulos, 2011; Ingy & Wim, 2017).
So far, terpene products from engineered microorganisms are mainly
from the S. cerevisiae and E. coli platforms (Lyu, Lee, & Chen, 2019).
These products have displayed great potential in food applications
including flavoring agents, sweetening agents, preservatives and func­
tional food ingredients, indicating the latent commercial value of the
microbial technology. This review aims to investigate the MVA and MEP
pathways of terpene biosynthesis and the metabolic engineering of
S. cerevisiae for promoting terpene production. The article also
highlights the future prospect of fungi as a reservoir of heterologous TPS
genes, and filamentous fungi and E. coli as the alternative hosts to
S. cerevisiae in terpene synthesis, and further discusses the applications
of the microbial engineering derived terpene products in food industry.
This review provides guidance for future researches and industrial ap­
plications in terpene bioproduction from genetically engineered
microorganisms.
2. Terpene biosynthesis
Monoterpenes, sesquiterpenes and diterpenes as major terpene sub­
classes are respectively produced from the corresponding intermediates
geranyl diphosphate (GPP), farnesyl diphosphate (FPP) and ger­
anylgeranyl diphosphate (GGPP) through MVA and MEP pathways
(Fig. 1A) (Zhuang, 2013). These intermediates are formed by a contin­
uous condensation of IPP and DMAPP from head to tail, then subjected
to cyclization and post modifications (e.g., acetylation, oxidation) to
generate terpene products (Wang, Quan, & Xiao, 2019).

2
Z. Liang et al. Food Research International 147 (2021) 110487

Fig. 1. Terpene biosynthesis (A) via mevalonate (MVA) and/or methylerythritol-phosphate (MEP) pathways in plant (B), Saccharomyces cerevisiae (C) and Escherichia
coli (D). Dashed arrow lines represent multiple steps. TCA cycle, tricarboxylic acid cycle; AA-CoA, acetoacetyl-CoA; HMG-CoA, 3-hydroxy-3-methylglutaryl-CoA;
MVA, mevalonate; MVAP, mevalonate-5-phosphate; MVAPP, mevalonate-5-pyrophosphate; IP, isopentenyl phosphate; IPP, isopentenyl pyrophosphate; DMAPP,
dimethylallyl pyrophosphate; G3P, glyceraldehyde 3-phosphate; DXP, 1-deoxy-D-xylulose-5-phosphate; MEP, methylerythritol-phosphate; CD-ME, 4-(cytidine 5′ -
diphospho)-2-C-methyl-D-erythritol; MEC, 2C-methyl-D-erythritol-2,4-cyclodiphosphate; HMBPP, 4-hydroxy-3-methylbut-2-enyl diphosphate; GPP, geranyl pyro­
phosphate; FPP, farnesyl diphosphate;; GGPP, geranylgeranyl pyrophosphate; AACT, acetoacetyl-CoA thiolase; HMGS, HMG-CoA synthase; HMGR, HMG-CoA
reductase; MVK, mevalonate kinase; PMK, MVAP kinase; MDD, MVAPP decarboxylase; MPD, MVAP decarboxylase; IPK, IP kinase; IDI, isopentenyl diphosphate
isomerase; DXS, DXP synthase; DXR, DXP reductoisomerase; MCT, CD-ME cytidylyltransferase; CMK, CD-ME kinase; MDS, MEC synthase; HDS, HMBPP synthase;
HDR, HMBPP reductase; GPPS, GPP synthase; FPPS, FPP synthase; GGPPS, GGPP synthase.

The structural diversity of terpenes is determined by three enzymes
isoprenyl diphosphate synthase (IDS), TPS and cytochrome P450. IDS is
responsible for controlling the branching points of terpene biosynthesis
by catalyzing the sequential condensation of IPP and DMAPP to form the
intermediates with different chain lengths (Wang & Ohnuma, 1999).
TPSs can catalyze carbocation-driven cyclization, rearrangement, and
elimination reactions on acyclic prenyl diphosphate precursors, leading
to the formation of a vast variety of acyclic and cyclic terpene scaffolds
(Tholl, 2006). Cytochrome P450 is another modifying enzyme to add
diversity to terpene structure through oxidation, hydroxylation, as well
as ring rearrangement and closure (Zhou & Pichersky, 2020).
2.1. Terpene biosynthesis in plant
In plant cells, MVA and MEP pathways are primarily employed to
meet the biosynthetic need for diverse terpenes (Fig. 1B). The MVA
pathway starts with the successive condensation of three molecules of
acetyl-CoA into HMG-CoA via acetoacetyl-CoA thiolase encoded by
ERG10 and HMG-CoA synthase (HMGS) encoded by ERG13. Afterward,
HMG-CoA is catalyzed by HMGR to form mevalonate. Mevalonate is
further subjected to a two-step phosphorylation and decarboxylation to
produce IPP that can be isomerized to DMAPP by isopentenyl diphos­
phate isomerase (IDI) encoded by IDI1. One DMAPP and two IPP mol­
ecules are further acted upon by FPP synthase (FPPS) encoded by ERG20
to form FPP, a precursor of sesquiterpene. Two molecules of FPP can be
condensed by squalene synthase (Erg9p) to yield squalene acting as a
pivotal precursor of triterpene and sterol (Wang, Quan, et al., 2019).
In MEP pathway, glyceraldehydes-3-phosphate (G3P) and pyruvate
are successively subjected to condensation and reduction to yield MEP,
then catalyzed by 4-(cytidine 5′ -diphospho)-2-C-methyl-D-erythritol
(CD-ME) cytidylyltransferase (MCT) to form CD-ME. CD-ME is further
converted to 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate (HMBPP)
by phosphorylation, cyclization and ring opening. Then, one DMAPP
and five IPP molecules are simultaneously generated from HMBPP under
the catalysis of HMBPP reductase (HDR). DMAPP and IPP can be
condensed by corresponding catalytic enzymes to form GPP, FPP and
GGPP, thus contributing to the production of different terpenes. The
biosynthesis of sesquiterpenes has been detailed in our previous review
(Li et al., 2020). It is worth mentioning that GGPP also contributes to the
biosynthesis of norisoprenoids (Wang, Quan, et al., 2019).
2.2. Terpene biosynthesis in yeast, other fungi and bacteria
S. cerevisiae is the most widespread yeast species responsible for
fermentations in food and beverage production. Similar to plant cells, S
cerevisiae also serves as a production host for diverse terpenes but only
harbors the MVA pathway within the cytoplasm (Zhuang, 2013)
(Fig. 1C). With acetyl-CoA as the primary precursor, GPP, FPP, GGPP
and squalene are natively synthesized in the yeast as the precursors of
terpenes, norisoprenoids and sterols (Paramasivan & Mutturi, 2017).

3
Z. Liang et al. Food Research International 147 (2021) 110487

Likewise, other fungi only possess the MVA pathway for terpene pro­
duction (Zhang, Nielsen, & Liu, 2017). Fungi contain a large quantity of
TPS genes that can encode crucial enzymes for terpene biosynthesis
(Zhang et al., 2020). Sesquiterpenes, diterpenes and triterpenes are the
major classes of terpenes found in fungi (Quin, Flynn, & Schmidt-
Dannert, 2014). Most bacteria especially E. coli produce terpenes
exclusively through the MEP pathway (Fig. 1D), which can be sub­
divided into two modules as that of plant cell: the feeding module where
G3P and pyruvate are generated from sugar substrates and the MEP
module which produces terpenes as the final product (Liu, Sun, et al.,
2013).
3. Engineering of Saccharomyces cerevisiae
The metabolic engineering of S. cerevisiae is based on the modifica­
tion in the MVA and orthogonal pathways employing genetic tools and
techniques, which is mainly ascribed to two actions: increasing pre­
cursor supply and downregulating competing pathways (Paramasivan &
Mutturi, 2017). Vector is the most common genetic tool responsible for
delivering and manipulating foreign DNA inside a host cell. Plasmid is
recommended as a viable vector due to its small size and easy manip­
ulation for controlled gene expression. The recombinant expression by
plasmid is maintained at gene dosage levels similar to that of the
chromosome while avoiding the interruption of native genes (Jones,
Kim, & Keasling, 2000). Additionally, the plasmid with multiple copy
numbers can increase the expression of target proteins (Chen, Partow,
Scalcinati, Siewers, & Nielsen, 2012). Therefore, the optimization of
plasmid is beneficial to the efficiency of terpene synthesis in engineered
microorganism. So far, overexpression of HMGR variant, ERG20 over­
expression, ERG9 downregulation, integration of extra IDI1 gene, dele­
tion of ROX1 and construction of orthogonal pathway have been
adopted as the effective engineering strategies for enhancing terpene
production in S. cerevisiae (Fig. 2). Here, we summarize terpene products
including their yield from different S. cerevisiae strains engineered by
these strategies (Table 1).
3.1. Overexpression of HMG-CoA reductase (HMGR) variant
HMGR was identified as a key rate-limiting enzyme in the MVA
pathway in the 1980s, which is responsible for the conversion of HMG-
CoA into mevalonate in yeast (Bach, 1986). Hmg1p and Hmg2p are two
reductase isozymes separately encoded by HMG1 and HMG2, thereinto,
Hmg1p plays a bigger role by expressing about 83% of HMGR activity
and presents in a more stable form (Wang, Quan, et al., 2019). HMGR is
composed of two terminal domains. The hydrophobic N-terminus an­
chors the protein to endoplasmic reticulum membrane, while the hy­
drophilic C-terminus possesses all the enzyme activity and is oriented for
projection into the cytoplasm (Liscum et al., 1985). Overexpression of
HMGR is known to promote the MVA pathway, despite it is still nega­
tively affected by feedback regulation (DeBose-Boyd, 2008). The trans­
lation of Hmg1p may be regulated by the upstream intermediates such
as mevalonate, although the actual mechanism remains unknown. In
contrast, the downstream intermediates such as sterol and nonsterol
isoprenoids impose feedback inhibition on Hmg2p, especially at high
pathway flux. Nonsterol signal GGPP can change Hmg2p folding, lead­
ing to Hmg2p degradation that is further stimulated by oxysterol (Burg
& Espenshade, 2011). In addition, FPP-derived farnesol also contributes
to the alteration of the Hmg2p conformation into a less folded state,
thereby increasing the susceptibility of Hmg2p to degradation (Kuranda,
François, & Palamarczyk, 2009). To overcome the limitation, the N-
terminal truncated and soluble form of Hmg1p (tHmg1p) which is free
from feedback regulation, is available to increase the accumulation of
intermediates and thus terpene precursors, thereby boosting the terpene
yield (Farhi et al., 2011; Wang, Quan, et al., 2019). For monoterpene,
limonene can be produced in S. cerevisiae by expressing tHMG1 and
Citrus limon limonene synthase at the yield level of 1.48 mg/L (Beh­
rendorff, Vickers, Chrysanthopoulos, & Nielsen, 2013). For sesquiter­
pene, in the study of Farhi et al. (2011), the overexpression of tHMG1,
mitochondrion-targeted heterologous FPPS (mtFDPS) and a
mitochondrion-targeted sesquiterpene synthase (mtCsTPS) can produce
1.5 and 20 mg/L of valencene and amorphadiene respectively, which
accounted for 8- and 20-fold enhancements in the production. When the

Fig. 2. Engineering strategies for enhancing terpene production in Saccharomyces cerevisiae. Dashed arrow lines represent multiple steps. HMG-CoA, 3-hydroxy-3-
methylglutaryl-CoA; MVA, mevalonate; IPP, isopentenyl pyrophosphate; DMAPP, dimethylallyl pyrophosphate; GPP, geranyl pyrophosphate; FPP, farnesyl
diphosphate; GGPP, geranylgeranyl pyrophosphate; NPP, neryl diphosphate; AACT, acetoacetyl-CoA thiolase; HMGS, HMG-CoA synthase; HMGR, HMG-CoA
reductase; MVK, mevalonate kinase; PMK, mevalonate-5-phosphate (MVAP) kinase; MDD, mevalonate-5-pyrophosphate (MVAPP) decarboxylase; IDI, isopentenyl
diphosphate isomerase; SS, squalene synthase; FPPS, FPP synthase; GGPPS, GGPP synthase; NPPS, NPP synthase; MTS, monoterpene synthase; STS, sesquiterpene
synthase; DTS, diterpene synthase.

4
Z. Liang et al. Food Research International 147 (2021) 110487

Table 1
Selective studies of terpenes/terpenoids from engineered Saccharomyces cerevisiae strains.
Terpene/ Maximum Strain Engineering strategy Terpene synthase Reference
Terpenoid yield (mg/L)

Monoterpene
(+)-Limonene 1.48 EPY210C Overexpression of tHMG1 Limonene synthases from Citrus (Behrendorff
limon and Mentha spicata et al., 2013)
0.028 AE9 K197G Expression of an ERG20 mutant (K197G) (+)-Limonene synthase from (Jongedijk et al.,
Citrus limon 2015)
166 AM94 a. Construction of orthogonal pathway using substrate NPP for NPP-specific variant limonene (Ignea et al.,
monoterpene synthase synthase from Citrus limon 2019)
b. Introduction of ergosterol-repressed ERG1 promoter upstream of (ClLimS)
the ERG20 gene
917.7 yJGZ1 a. Construction of orthogonal pathway using substrate NPP for N-terminal truncated limonene (Cheng et al.,
monoterpene synthase of terpenes/terpenoids from engineered b. synthase from Citrus limon 2019)
Dynamic downregulation of ERG20 by replacing the ERG20 promoter
with the glucose-sensing HXT1 promoter
9.8 AY12α a. Construction of orthogonal pathway using substrate NPP for N-terminal truncated limonene (Hu et al., 2020)
monoterpene synthase of terpenes/terpenoids from engineered b. synthase from Citrus limon
Expression of N-terminal truncated NPPS encoded by tNDPS1
c. Expression of tLS–tNDPS1 fusion enzyme
(-)-Limonene 0.060 AE9 K197G Expression of an ERG20 mutant (K197G) (-)-Limonene synthase from (Jongedijk et al.,
Perilla frutescens 2015)
Sabinene 17.5 AM94 a. Expression of a degradation-stabilized variant of HMG2 (bearing a Sabinene synthase from Salvia (Ignea et al.,
K6 to R mutation) pomifera 2014)
b. Expression of ERG20 mutants (A99C, A99L, A99F, A99W, F96W,
N127W, F96W-N127W)
c. Monoallelic deletion of ERG9
113 AM94 a. Construction of orthogonal pathway using substrate NPP for NPP-specific variant sabinene (Ignea et al.,
monoterpene synthase synthase from Salvia pomifera 2019)
b. Introduction of ergosterol-repressed ERG1 promoter upstream of (SpSabS)
the ERG20 gene
Cineole 485 BY4741 a. Integration of a K6R variant of HMG2 into the genome Cineole synthase from Salvia (Ignea et al.,
b. Integration of a promoter-IDI1-ts cassette fruticosa (CinS1) 2011)

Monoterpenoid
Geraniol 5 AE9 K197K Expression of ERG20 mutants (K197G, E, S) Geraniol synthase from Ocimum (Fischer et al.,
basilicum 2011)
1690 YZG13-GE1 a. Dynamic control of ERG20 by replacing the ERG20 promoter with Geraniol synthase from (Zhao et al.,
the glucose-sensitive HXT1 promoter Valeriana officinalis 2017)
b. Optimization of carbon source feeding strategies (glucose and
ethanol ratio)
c. Deletion of OYE2 or ATF1
d. LEU2 auxotrophic complementation
6.77 CEN.PK2-1C a. Overexpression of tHMG1 N/A (Liu, Zhang,
b. Integration of IDI1 gene et al., 2013)
Linalool 0.3 AE9 K197K Expression of ERG20 mutants (K197G, E, S) N/A (Fischer et al.,
2011)
Citronellol 0.3 AE9 K197K Expression of ERG20 mutants (K197G, E, S) N/A (Fischer et al.,
2011)

Sesquiterpene
Farnesene 13.8 CEN.PK113-5D a. Expression of acetoacetyl-CoA synthase (nphT7) Farnesene synthase from Citrus (Tippmann
b. Overexpression of tHMG1 junos (FarnSyn) et al., 2017)
c. Overexpression of mutated acetyl-CoA carboxylase (ACC1**)
2240 CEN.PK2-1C, CEN.PK2- a. Substituting ALD6, ACS1 and ACS2 (‘PDH-bypass’) reactions with N/A (Meadows et al.,
1D or CEN.PK113-7D acetaldehyde dehydrogenase (acylating) (ADA) 2016)
b. Replacing the native NADPH-dependent HMGR with an NADH-
specific version (NADH-HMGR) from Silicibacter pomeroyi
c. Overexpression of xylulose-5-phosphate specific phosphoketolase
(xPK) and phosphotransacetylase (PTA)
Bisabolene 994 EPY300 a. Overexpression of UPC2-1, tHMG1, ERG20 Abies grandis α-bisabolene (Peralta-Yahya
b. Downregulation of ERG9 synthase (AgBIS) et al., 2011)
400 BY4741, CEN.PK2 a. Deletion of ROX1 Abies grandis α-bisabolene (Özaydın et al.,
synthase (AgBIS) 2013)
800 BY4741, CEN.PK2 a. Deletion of ROX1 Abies grandis α-bisabolene (Özaydın et al.,
b. Deletion of YJL064W and YPL062W synthase (AgBIS) 2013)
Caryophyllene 125 BY4741 a. Tandem heterologous deletion of ubiquitin ligase Ubc7p, Ssm4p Caryophyllene synthase from (Ignea et al.,
and the endoplasmic reticulum resident protein Pho86p Salvia fructicosa 2012)
b. Overexpression of a degradation stabilized variant of Hmg2p (K6R)
δ-Cadinene 40 BY4741 a. Overexpression of mutated HMG2 (K6R) and ERG20 Sesquiterpene synthase (P330) (Ignea et al.,
b. Downregulation of ERG9 from Salvia pomifera 2011)
trans- 10 BY4741 a. Integration of a K6R variant of HMG2 into the genome Sesquiterpene synthase (P330) (Ignea et al.,
β-Caryophyllene b. Upregulation of ERG20 from Salvia pomifera 2011)
c. Haploinsufficiency for ERG9
Valencene 3.0 CEN.PK113-5D Downregulation of ERG9 Valencene synthase (GFTpsD) (Asadollahi
from grapefruit et al., 2008)
1.5 BDXe
(continued on next page)

5
Z. Liang et al. Food Research International 147 (2021) 110487

Table 1 (continued )
Terpene/ Maximum Strain Engineering strategy Terpene synthase Reference
Terpenoid yield (mg/L)

a. Overexpression of tHMG1 Valencene synthase from Citrus (Farhi et al.,

b. Expression of mtFDPS sinensi (CsTPS1) 2011)
c. Expression of mtCsTPS
539.3 BJ5464 a. Deletion of ROX1 Valencene synthase from (Chen et al.,
b. Overexpression of ERG10, ERG13, tHMG1, ERG12, ERG8, ERG19, Callitropsis nootkatensis 2019)
ERG20 and IDI1
c. Downregulation of ERG9
d. Deletion of BTS1, DPP1, LPP1
Amorphadiene 20 BDXe a. Overexpression of tHMG1 Amorpha-4,11-diene synthase (Farhi et al.,
b. Expression of mtFDPS from Artemisia annua 2011)
c. Expression of mtCsTPS
α-Copaene 12 BY4741 a. Integration of a K6R variant of HMG2 into the genome Sesquiterpene synthase (P330) (Ignea et al.,
b. Upregulation of ERG20 from Salvia pomifera 2011)
c. Haploinsufficiency for ERG9
α-Cubebene 7 BY4741 a. Integration of a K6R variant of HMG2 into the genome Sesquiterpene synthase (P330) (Ignea et al.,
b. Upregulation of ERG20 from Salvia pomifera 2011)
c. Haploinsufficiency for ERG9

Sesquiterpenoid
trans-Nerolidol 64.4 BEJY12 a. Delection of ERG9 N/A (Jackson, 2005)
b. Overexpression of tHMG1
Farnesol 20.71 BEJY12 a. Delection of ERG9 N/A (Jackson, 2005)
b. Overexpression of tHMG1
Cubebol 1.6 CEN.PK113-5D Downregulation of ERG9 Cubebol synthase (GFTpsC) (Asadollahi
from grapefruit et al., 2008)
9.9 CEN.PK113-5D a. Overexpression of tHMG1 Cubebol synthase from Citrus (Asadollahi
b. Downregulation of ERG9 paradisi et al., 2010)
Patchoulol 16.9 CEN.PK113-5D Downregulation of ERG9 Patchoulol synthase (Asadollahi
(PatTps177) from patchouli et al., 2008)

Diterpene &
Diterpenoid
Casbene 108.5 NCYC 3608 a. Expression of a truncated GGPPS from Phomopsis amygdali Casbene synthase from Ricinus (Callari et al.,
b. Dynamic control of ERG20 and ERG9 by replacing their native communis 2018)
promoters with glucose-sensitive HXT1 promoter and ergosterol-
sensitive ERG1 promoter

Steviol glycosides N/A Haploid strain (MATα a. Optimization of the kaurene oxidase-kaurenoic acid hydroxylase Kaurene synthase (Gold et al.,
HOΔura3Δhis3-Δleu2Δ) (KO-KAH) activities 2018)
b. Expression of cytochrome P450 reductase (CPR)
c. Expression of UDP (uridine diphosphate) glucose-dependent
glucosyltransferases (UGTs)
Sclareol 750 BY4741 a. Deletion of ROX1 Sclareol synthase from Salvia (Trikka et al.,
b. Deletion of YGR259C, YNR063W, VBA5, YER134C and DOS2 sclarea 2015)

Triterpene &
Triterpenoid
β-Amyrin 36.5 INVSc1 a. Integration of IDI1 gene from E. coli β-Amyrin synthase from (Zhang et al.,
b. Overexpression of ERG20 and ERG9 from S. cerevisiae Glycyrrhiza glabra 2015)
Protopanaxadiol 1189 BY4742 a. Overexpression of tHMG1, ERG20, ERG9 Protopanaxadiol synthase from (Dai et al., 2013)
b. Overexpression of squalene epoxidase (ERG1) Panax ginseng
c. Increasing protopanaxadiol synthase activity through codon
optimization

Tetraterpene
β-Carotene 47.18 BY4741, BY4742 a. Introduction of the positive GGPP synthase mutant (crtE03M) N/A (Zhou et al.,
b. Overexpression of tHMG1, crtI and crtYB 2017)
c. Increasing β-carotene ketolase activity via directed evolution
d. Balancing the metabolic flux by adjusting copy numbers of the rate-
limiting enzymes

Norisoprenoid
β-Ionone 5 SCIGS22 a. Integration of the tHMG1 gene N/A (López et al.,
b. Overexpression of ERG20 2015)
c. Downregulation of ERG9
d. Deletion of LPP1 and DPP1 genes
e. Overexpression of BTS1, crtYB and crtI genes
f. Expression of CCD1 gene

N/A, not available; S. cerevisiae, Saccharomyces cerevisiae; E. coli, Escherichia coli; NPP, neryl diphosphate; NPPS, NPP synthase; ADA, acetaldehyde dehydrogenase
(acylating); NADPH, nicotinamide adenine dinucleotide phosphate; NADH, nicotinamide adenine dinucleotide; HMGR, HMG-CoA (3-hydroxy-3-methylglutaryl-CoA)
reductase; xPK, xylulose-5-phosphate specific phosphoketolase; PTA, phosphotransacetylase; GGPP, geranylgeranyl diphosphate; GGPPS, GGPP synthase; KO-KAH,
kaurene oxidase-kaurenoic acid hydroxylase; CPR, cytochrome P450 reductase; UDP, uridine diphosphate; UGTs, UDP glucose-dependent glucosyltransferases.

6
Z. Liang et al.
tHMG1 engineering combined with the expression of acetoacetyl-CoA
synthase (nphT7), mutated acetyl-CoA carboxylase (ACC1**) and far­
nesene synthase from Citrus junos (FarnSyn), the final farnesene yield can
reach 13.8 ± 4.1 mg/L (Tippmann, Ferreira, Siewers, Nielsen, & Chen,
2017). During the biosynthesis, excess nicotinamide adenine dinucleo­
tide (NADH) is readily oxidized by O2, leading to the inefficient capture
of free energy from glucose in farnesene (Dugar & Stephanopoulos,
2011). Therefore, another study replaced the native nicotinamide
adenine dinucleotide phosphate (NADPH)-specific HMGR with an
NADH-dependent one in the strain, together with the substitution of
acetaldehyde dehydrogenase (ALD6) and acetyl-CoA synthases (ACS1
and ACS2) (‘PDH-bypass’) reactions with acetaldehyde dehydrogenase
(acylating) (ADA) and overexpression of xylulose-5-phosphate specific
phosphoketolase (xPK) and phosphotransacetylase (PTA), which suc­
cessfully boosted the farnesene production to 2240 mg/L (Meadows
et al., 2016). It is worth mentioning that cubebol production can be
increased by 30% when overexpressing tHMG1 from an episomal high
copy plasmid by the integration of strong inducible GAL10 promoter
rather than genomic integration (Asadollahi, Maury, Schalk, Clark, &
Nielsen, 2010). In addition, the overexpression of tHMG1 can also aid in
tetraterpene biosynthesis. Zhou et al. (2017) overexpressed tHMG1,
phytoene desaturase (crtI), phytoene synthase (crtYB) and GGPP syn­
thase (GGPPS) mutant (crtE03M) in an engineered diploid strain of
S. cerevisiae with increased β-carotene ketolase activity and balanced
metabolic flux, resulting in a 1.8-fold increase to 47.18 mg/L in astax­
anthin production. Engineering of Hmg2p also aims for improved
terpene production while inhibiting protein degradation. For instance, a
single point mutation replacing lysine 6 with an arginine (K6R) in HMG2
led to nearly three times higher production of cineole, α-copaene and
α-cubebene in S. cerevisiae while trans-β-caryophyllene and δ-cadinene
also increased by two-fold (Ignea et al., 2011). The stability of Hmg1p
and Hmg2p are negatively affected by their positive interactors,
including ubiquitin ligase Ubc7p, Ssm4p and the endoplasmic reticulum
resident protein Pho86p. These proteins participate in the degradation
of endoplasmic reticulum transmembrane proteins, and the deletion of
them contributes to the improved stabilization of Hmg1p and Hmg2p
(Paramasivan & Mutturi, 2017). For instance, the caryophyllene yield
using S. cerevisiae expressing caryophyllene synthase from Salvia fructi­
cosa can be boosted to 125 mg/L by deleting the target genes of all three
proteins and overexpressing the stabilized variant of Hmg2p, which is
11-fold higher than that of the wild strain (Ignea et al., 2012).
3.2. Upregulation of ERG20
Upregulation of the ERG20 gene is a specific strategy to stimulate
monoterpene biosynthesis in S. cerevisiae by increasing the intracellular
GPP pool in the MVA pathway (Liu, Zhang, Du, Chen, & Zhou, 2013).
GPP, formed by the condensation of IPP and DMAPP, is known as the
principal precursor for monoterpene biosynthesis (Wang, Quan, et al.,
2019). Unlike plants, yeast does not harbor a specific GPP synthase
(GPPS), resulting in no or limited monoterpene production (Zhuang,
2013). In S. cerevisiae, Erg20p, the dual-specificity enzyme encoded by
ERG20, has both GPPS and FPPS activities, catalyzing the sequential
formation of GPP and thus FPP. However, GPP only functions as an in­
termediate of FPP synthesis since it is tightly bound to the catalytic site
of Erg20p and oriented to the conversion to FPP (Zhao, Bao, Li, Shen, &
Hou, 2016). To increase the GPP flux in S. cerevisiae, a series of ERG20
mutants are constructed for the endogenous shift from FPP to GPP
accumulation (Zhuang, 2013). In the study of Fischer, Meyer, Claudel,
Bergdoll, and Karst (2011), S. cerevisiae ERG20 gene was subjected to a
sequence of amino acid mutations at position K197, where K197 G, E
and S successfully boosted the production of geraniol, linalool and
citronellol by 10- to 20-fold. It is worth mentioning that the sterol
content, growth rate, and monoterpenol production abilities varied in
different mutated strains. Similarly, an ERG20 mutant strain (AE9
K197G) was also deployed to simultaneously generate both enantiomers
7
Food Research International 147 (2021) 110487
of limonene by expressing (+)-limonene synthase from C. limon, and
(-)-limonene synthase from Perilla frutescens (Jongedijk et al., 2015). In
addition, a double mutation of F96 to W and N127 to W was also
introduced into the ERG20 gene to enhance the production of sabinene
by up to 10.4-fold to 0.53 mg/L in diploid S. cerevisiae strain. The fusion
of the mutated Erg20p (F96W-N127W) with sabinene synthase from
Salvia pomifera (SpSabS1) can boost sabinene yield by 3.5-fold to 1.87
mg/L. A 7-fold increase in production can be achieved when the fusion
occurred in the erg20/ERG20 heterozygous deletion strain, while the
introduction of a second plasmid vector generating the same fusion can
further boost the production to 17.5 mg/L (Ignea, Pontini, Maffei,
Makris, & Kampranis, 2014). Ignea et al. (2011) integrated galactose
promoter in one of the ERG20 alleles in diploid S. cerevisiae strain,
leading to 25%, 30% and 50% increases in δ-cadinene, cineole and trans-
β-caryophyllene production respectively, whereas the contents of
α-copaene and α-cubebene grew slightly. Another study reported the
dynamic control of ERG20 by replacing the ERG20 promoter with the
glucose-sensing HXT1 promoter, which can successfully redistribute
GPP flux and thus attain a 3.8-fold increase to 867.7 mg/L in geraniol
production after optimizing carbon source feeding. The geraniol yield
can be further elevated to 1.69 g/L by involving OYE2 (encoding an
NADPH oxidoreductase) deletion and LEU2 (auxotrophic marker)
complementation in addition to ERG20 dynamic control (Zhao et al.,
2017). It is well worth mentioning that an auxin-inducible degradation
of Erg20p can successfully redirect metabolic flux towards monoterpene
production, which is an advanced metabolic engineering tool
demanding further study to determine its feasibility (Lu, Peng, Ebert,
Dumsday, & Vickers, 2021).
3.3. Downregulation of ERG9
Another possible strategy to increase the availability of terpene
precursors is to lower their flux toward sterol synthesis (Paramasivan &
Mutturi, 2017). Sterol is a cell membrane component derived from
squalene, being responsible for maintaining the permeability, fluidity,
as well as the structural integrity of cell membranes and presenting as
signal compounds when cells communicate with one another (Zhuang,
2013). ERG9, which encodes squalene synthase (Erg9p), is targeted for
further modification of the MVA pathway in addition to ERG20. Erg9p
functions in the transformation of FPP into squalene while acting
downstream of Erg20p (Ignea et al., 2011). The downregulation of ERG9
can be attained via the deletion of the gene. In the study of Chen et al.
(2019), deleting 45-bp in the range of upstream activating cis-element
(UAS) in ERG9 promoter can promote the production of valencene in the
S. cerevisiae strain with valencene synthase from Callitropsis nootkatensis.
Monoallelic deletion of ERG9 together with the expression of HMG2
(bearing a K6 to R mutation) can yield 0.05 mg/L sabinene (Ignea et al.,
2014). Further, ERG9 downregulation can also be coupled with the
overexpression of tHMG1, ERG20, global transcription regulator of ste­
rol pathway UPC2-1, as well as Abies grandis α-bisabolene synthase
(AgBIS), reaching 994 ± 241 mg/L in bisabolene production (Peralta-
Yahya et al., 2011). The combination of ERG9 deletion and tHMG1
overexpression was also reported to produce 20.71 mg/L farnesol and
64.40 mg/L trans-nerolidol which were not identified in the wild type
yeast strains (Jackson, 2005). Nevertheless, deletion of ERG9 gene can
compromise cellular activity in haploid cells, while no detrimental effect
but decreased mRNA levels of the gene is reported when deleting one
allele in diploid cells. In the study of Ignea et al. (2011), a loxP-his5-loxP
cassette was used to delete one of the two ERG9 alleles in the S. cerevisiae
strain, resulting in a significant decline in monoterpene production (e.g.
cineole) but a drastic increase in sesquiterpene production (e.g. trans-
β-caryophyllene, α-copaene, δ-cadinene). This may be ascribed to FPP
accumulation, which might lead to feedback inhibition on Erg20p with
GPPS activity and thus reduce the GPP availability (Ignea et al., 2015).
During aerobic growth conditions, yeast cells can not actively assimilate
exogeneous ergosterol that is vital for yeast growth, so the complete
Z. Liang et al.
deletion of ERG9 gene responsible for endogenous ergosterol cannot be
attained (Asadollahi et al., 2010). To complement the gene deletion,
Zhang, Jennings, Robinson, and Poulter (1993) adopted a SUE (sterol
uptake enhancement) mutation which is a yeast line that enables the
aerobic uptake and utilization of exogenous sterol, while the endoge­
nous ergosterol biosynthetic pathway becoming dispensable. In
response, a large quantity of pathway intermediates (DMAPP, IPP and
FPP) are diverted to the biosynthesis of other non-essential terpene
products.
Replacement of the ERG9 promoter with a repressive one also serves
as a promising alternative to ERG9 deletion. Asadollahi et al. (2008)
replaced the endogenous ERG9 promoter with the methionine repressive
promoter MET3 in the S. cerevisiae cells, resulting in a decreased
ergosterol content and increased levels of FPP derived compounds like
sesquiterpenes. The yield of valencene and patchoulol separately
increased by 50% and 70%, while the cubebol content remained nearly
unchanged. The addition of higher amounts of methionine to the me­
dium can further increase the patchoulol content by 47% to 16.9 mg/L.
Another study reported the same replacement strategy, which increased
the cubebol production more than threefold from 3.2 mg/L to 9.9 mg/L.
However, the simultaneous overexpression of tHMG1 in this background
decreased cubebol production by 10% (Asadollahi et al., 2010). Glucose-
regulated promoters such as HXT1 promoter can also be used to replace
the ERG9 promoter. Under this condition, integration of the tHMG1
gene, overexpression of ERG20, GGPPS gene (BTS1), carotenogenic
genes (crtYB and crtI) and carotenoid cleavage dioxygenase gene
(CCD1), deletion of lipid phosphate phosphatase gene (LPP1) and
diacylglycerol pyrophosphate phosphatase gene (DPP1) increased the
β-ionone production by 23-fold to 5 mg/L in the engineered S. cerevisiae
strain (López et al., 2015). Another study replaced the ERG9 and ERG20
promoters with ergosterol-sensitive ERG1 promoter and glucose-
sensitive HXT1 promoter respectively, together with the expression of
truncated GGPPS from Phomopsis amygdali and casbene synthase from
Ricinus communis, which successfully boosted the casbene production by
3.6-fold to 108.5 mg/L (Callari, Meier, Ravasio, & Heider, 2018). It is
reported that ERG9 can also be overexpressed to enrich the squalene
pool for triterpene biosynthesis. In the study of Dai et al. (2013), a 262-
fold increase in protopanaxadiol production can be observed in
S. cerevisiae with enhanced protopanaxadiol synthase from Panax
ginseng, when combining the overexpression of ERG9, tHMG1, ERG20
and squalene epoxidase (ERG1).
3.4. Integration of extra IDI1 gene
The gene IDI1 catalyzes the isomerization of IPP to DMAPP by
encoding for an essential isomerase IDI, contributing to the subsequent
generation of GPP and monoterpene. Therefore, the introduction of IDI1
gene into the yeast genome is well accepted as a promising strategy to
improve the monoterpene production which depends on the copy
number of IDI1 (Ignea et al., 2011). According to Liu, Zhang, et al.
(2013), the integration of an extra copy of IDI1 into the genome of
S. cerevisiae can lead to a 50% increase in geraniol yield, while the IDI1
overexpression using a multi-copy plasmid boosted geraniol production
by 1.45-fold. A 6-fold increase can also be observed when combining the
overexpression of IDI1 and tHMG1. In the study of Ignea et al. (2011),
the high copy number plasmid pYES2-myc was used to integrate the IDI1
gene under the control of the galactose promoter in the genome of
S. cerevisiae with cineole synthase from Salvia fruticosa. For the modified
strain, the overexpression of IDI1 boosted the cineole production to 250
mg/L which was 5-fold higher than that of the unmodified one. Under
the HMG2 expression (K6R), the same study also cloned the IDI1 gene
into the COD4 cassette to obtain the COD40 construct which was then
integrated into the yeast chromosome, producing cineole at the yield
level of 200 mg/L. A further increase to 485 mg/L can be achieved when
a plasmid expressing IDI1 was introduced, confirming that the integra­
tion of additional copies can significantly promote yields. It is worth
8
Food Research International 147 (2021) 110487
mentioning that the expression of IDI1 may exert no enhancing effect on
sesquiterpene yield, but contribute to the expansion of squalene pool for
triterpene overproduction. For instance, the co-expression of IDI1,
ERG20 and ERG9 in S. cerevisiae with β-amyrin synthase from Glycyr­
rhiza glabra can enhance the accumulation of squalene and thus
β-amyrin, resulting in 30% and 46% increases in β-amyrin production in
plasmid-based and genome-integrated yeast strains respectively (Zhang
et al., 2015).
3.5. Deletion of ROX1
ROX1 is a transcriptional factor inhibiting the expression of hypoxia-
induced genes in MVA pathway and ergosterol biosynthesis (Henry,
Nickels, & Edlind, 2002; Montañés, Pascual-Ahuir, & Proft, 2011),
which can be knocked out to enhance the mevalonate levels in cells and
thus terpene production. In the study of Jakočiūnas et al. (2015), single
deletion of ROX1 in S. cerevisiae can significantly boost valencene yield
compared to other single gene alterations of ERG9, BTS1, YPL062W and
YJL064W. Under ROX1 deletion, the valencene production volume
reached 539.3 mg/L which was a 160-fold increase when additional
engineering strategies (overexpression of ERG10, ERG13, tHMG1,
ERG12, ERG8, ERG19, ERG20 and IDI1, downregulation of ERG9,
deletion of BTS1, DPP1, LPP1) were simultaneously employed in the
S. cerevisiae strain with valencene synthase from C. nootkatensis (Chen
et al., 2019). Similarly, a two-fold increase in bisabolene production was
reported in S. cerevisiae by expressing bisabolene synthase from
A. grandis and deleting ROX1 gene. In addition, the deletion of YJL064W
and YPL062W genes can further increase the flux through the MVA
pathway, thereby boosting the bisabolene yield from 400 mg/L to 800
mg/L (Özaydın, Burd, Lee, & Keasling, 2013). According to Trikka et al.
(2015), the combination of ROX1 deletion and the deletion of other five
genes (YGR259C, YNR063W, VBA5, YER134C and DOS2) which limit
the productivity of carotenoid and potentially diterpene, induced a 40-
fold increase in carotenoid, and consequently a 12-fold increase in
sclareol yield to 750 mg/L in S. cerevisiae with sclareol synthase from
Salvia sclarea. It is worth mentioning that ROX1 deletion in S. cerevisiae
can facilitate the expression of multiple ERG genes participating in
ergosterol biosynthesis, leading to 2.5- to 16-fold-lower susceptibilities
to two sterol biosynthesis inhibitors, azoles and terbinafine (Henry et al.,
2002).
3.6. Recent advances on engineering strategies
Construction of orthogonal pathway is an emerging strategy to
enhance monoterpene production in yeast by using neryl diphosphate
(NPP) as substrate for MTSs (Ignea et al., 2019). NPP is formed by the
condensation of IPP and DMAPP in the cis-configuration under the
catalysis of NPP synthase (NPPS) encoded by SlNDPS1 gene (Schilmiller
et al., 2009). Unlike GPP, NPP is not consumed by Erg20p for FPP
conversion. MTSs can be engineered to specifically accept NPP as the
substrate for yielding different monoterpene products. The orthogonal
route branches out from the MVA pathway at the point following the
synthesis of IPP and DMAPP. The branch point functions as a metabolic
valve which can be controlled dynamically to regulate fluxes between
the canonical and orthogonal branches (Ignea et al., 2019). For instance,
Ignea et al. (2019) successfully installed dynamic control by introducing
ergosterol-repressed ERG1 promoter upstream of the ERG20 gene in
S. cerevisiae with NPP-specific variants S. pomifera sabinene synthase
(SpSabS) and C. limon limonene synthase (ClLimS), which directed
fluxes to the orthogonal pathway due to the increased sterol levels
(Leber et al., 2001), and thus led to 7.1- and 6.8-fold increases in sabi­
nene and limonene production respectively. Similarly, Cheng et al.
(2019) established an orthogonal biosynthetic pathway involving NPPS
from Solanum lycopersicum and N-terminal truncated C. limon limonene
synthase. Under this condition, dynamic downregulation of the ERG20
gene by replacing the ERG20 promoter with the glucose-sensing HXT1
Z. Liang et al.
promoter can boost the limonene yield by 6-fold to 917.7 mg/L in the
engineered S. cerevisiae strain. In the study of Hu et al. (2020), the N-
terminal truncated genes of (+)-limonene synthase (tLS) and NPPS
(tNDPS1) were codon-optimized and heterologously expressed to
reconstruct an orthogonal pathway in S. cerevisiae, resulting in a 17-fold
increase to 6.97 mg/L in (+)-limonene production. The yield can be
further enhanced by 40% to 9.8 mg/L by adopting a tLS–tNDPS1 fusion
enzyme.
3.7. Terpene product toxicity
Indeed, the engineering strategies can significantly promote mono­
terpene synthesis in the MVA and orthogonal pathways in S. cerevisiae,
yet the increased amounts of intermediates and monoterpenes are toxic
and hazardous to the strain (Moser & Pichler, 2019). Monoterpenes can
readily compromise the cell membrane integrity and function, as well as
intracellular energy metabolism (Uribe, Ramirez, & Peña, 1985).
Therefore, irreversible cell damage can be induced by high monoterpene
levels. For instance, complete growth inhibition and substantial viability
loss were detected in both mutant and wild type S. cerevisiae strains in
the presence of 60 mg/L limonene (Brennan, Turner, Krömer, & Nielsen,
2012). To improve S. cerevisiae tolerance toward diverse monoterpenes,
heterologous expression of efflux pumps like ATP-binding cassette
(ABC) transporter can achieve expulsion and isolation of monoterpenes.
For instance, GcABC-G1 is a specialized ABC transporter promoting the
survival rate of S. cerevisiae cells by 3, 7 and 30 times in the presence of
(-)-β-pinene, (+)-limonene and (+)-3-carene, respectively. However, the
transformed yeast cells showed no increased tolerance to the racemic
α-pinene (Wang et al., 2013). Monoterpenes are sparingly water-soluble
oils that predominantly induce phase toxicity rather than molecular
toxicity. Therefore, the upregulation of ABC transporters can only
address low levels of molecular toxicity but no phase toxicity. For
instance, ABC transporters were proved to fail in limonene toxicity
alleviation since limonene interferes with cell wall integrity rather than
membranes in S. cerevisiae (Brennan, Krömer, & Nielsen, 2013; Hu et al.,
2012). Truncation of tricalbin protein (tTcb3p1-989) was first reported as
a successful engineering of phase tolerance in S. cerevisiae (Brennan
et al., 2015). tTcb3p1-989 can improve tolerance toward limonene,
β-pinene and myrcene by 9-, 11- and 8-fold. Under limonene stress,
tTcb3p1-989 served to maintain cell wall integrity and enhance cell wall
resilience. It also displayed negligible perturbations to strain growth
since fewer genes involved in cell wall maintenance and biogenesis were
affected in the reconstructed strain. Interestingly, most monoterpenes
have no inhibitory effect on strain growth at their aqueous saturation
points that are well below their minimum inhibitory concentrations
(MIC) (Brennan et al., 2012). Therefore, the toxic level of monoterpene
may be hard to achieve. In contrast to monoterpene, the toxic effect of
sesquiterpene product on yeast cells is negligible due to the low pro­
duction level and weak toxicity (Schempp, Drummond, Buchhaupt, &
Schrader, 2018).
4. Fungal terpene synthase genes as TPSs reservoir
Fungi are well-known for their production of diverse terpenes. Spe­
cifically, the identity and number of terpene products depend on the
fungal TPSs active sites in which the substrate is activated by a metal-
binding motif to generate a carbocation intermediate and guided
through a series of cyclization reactions (Shaw et al., 2015). So far,
about 100 fungal TPSs had been functionally characterized and most of
them are sesquiterpene synthases (STSs). STSs from plants and fungi
conserve metal-binding motifs composed of an NSE/DTE and an
asparagine-rich motif. The NSE motif is highly preserved in fungal STSs,
while the asparagine-rich motif in fungal STSs [D(D/E/N)XX(D/E)] is
different to some extent from that found in plants [DDXX(D/E)]. Outside
the motif, the TPSs sequences of plants and microorganisms have limited
similarity, indicating the difficulty to predict the functions of fungal
9
Food Research International 147 (2021) 110487
STSs through investigating plant STSs (Zhang et al., 2020). In compar­
ison with plants, fungi possess a higher TPSs diversity which out­
numbers by at least 10 times (Blackwell, 2011). For instance, each
fungus in the phylum Basidiomycota was reported with an average of
10–20 putative TPS homologues (Wawrzyn, Quin, Choudhary, López-
Gallego, & Schmidt-Dannert, 2012). Therefore, the fungal TPSs and
their terpene products are abundant natural resources.
Various fungal TPS genes have been identified through the heterol­
ogous expression in E. coli, but many of them remain largely untapped
(Table 2). From Agrocybe aegerita, Zhang et al. (2020) identified 11 STSs
that were cloned and characterized in the engineered terpene-
overproducing E. coli chassis strain. Nine of the STSs produced at least
one sesquiterpene except for Agr10 and Agr11, indicating that they were
not functional probably due to their incomplete metal-binding regions
(NSE motif). δ-Cadinene and β-myrcene were generated by expressing
Agr4, which is a bifunctional enzyme responsible for the biosynthesis of
both sesquiterpenes and monoterpenes. The expression of Agr1 also
synthesized δ-cadinene, which accounted for 60% of the total terpenes
produced. Four major sesquiterpenes in similar proportions were pro­
duced by Agr3-expressing E. coli, including α-muurolene (32%),
γ-muurolene (22%), δ-cadinol (21%) and δ-cadinene (24%). Agr3 is
closely related to CpSTS3 from Clitopilus pseudopinsitus and Cop3 from
Coprinopsis Cinerea, both of which mainly produce α-muurolene. E. coli
expressing the Agr8 gene mainly produces γ-muurolene and β-cadinene.
In other studies, the heterologous expression of the Copu3 and CpSTS3
genes from the Basidiomycota Coniophora puteana mainly yielded
cubebol and α-cadinol (Mischko, Hirte, Fuchs, Mehlmer, & Brück, 2018;
Nagamine et al., 2019), while the expression of ShSTS12 from Stereum
hirsutum primarily produced α-cubebene and β-cubebene (Nagamine
et al., 2019; Quin, Flynn, Wawrzyn, Choudhary, & Schmidt- Dannert,
2013). Additionally, Nagamine et al. (2019) found that the E. coli
expression of CpSTS18, CpSTS5 and CpSTS11 from Clitopilus pseudo-
pinsitus generated γ-cadinene, γ-murolene and alloaromadendrene,
respectively. α-Gurjunene, β-gurjunene and β-copaene are the major
products from the expression of PpSTS06, Galma_104215 and Copu2
genes that respectively derive from Postia placenta, Galerina marginata
and C. puteana (Ichinose & Kitaoka, 2018; Mischko et al., 2018; Zhang
et al., 2020). For the brown-rot basidiomycete P. placenta, myrcene and
linalool are the main monoterpene products generated by the expression
of PpSTS25, while the expression of PpSTS06 mainly produced α-gur­
junene (Ichinose & Kitaoka, 2018). From Daldinia eschscholzii and
Hypoxylon sp., Wu et al. (2016) separately identified EC12-PGS and
CI4A-CPS genes by the heterologous expression in E. coli, producing the
corresponding sesquiterpenes including α-guaiene, β-pinene and β-cha­
migrene. Similar researches reported that α-farnesene, nerolidol, and
germacrene D were derived from C. pseudo-pinsitus, Hypoxylon sp., and
S. hirsutum separately containing three STS genes encoding CpSTS14,
Hyp1 and ShSTS10 (Nagamine et al., 2019; Quin et al., 2013; Shaw
et al., 2015). It should be noted that the functions of fungi are different
between ascomycetes and basidiomycetes, while basidiomycetes are
more dominant in the total number and variety of terpene products
(Zhang et al., 2020).
Rational engineering of fungal TPSs is an effective strategy to
enhance enzyme specificity and activity due to the plasticity of TPSs.
The TPSs plasticity refers to the capacity of TPSs to modify their func­
tions with a small number of amino acid substitutions (Yoshikuni, Fer­
rin, & Keasling, 2006). Site-directed mutagenesis and contact mapping
mutagenesis are the recent methods to identify plasticity residues which
control the profiles of STSs that are then transformed into a variety of
more specific synthases (Greenhagen, O’Maille, Noel, & Chappell, 2006;
Yoshikuni et al., 2006). For instance, site-directed mutagenesis can
successfully identify plasticity residues in trichodiene synthase from
Fusarium sporotrichioides and generate a different distribution of
sesquiterpene products including some newly identified sesquiterpenes
(Vedula, Jiang, Zakharian, Cane, & Christianson, 2008). The study of
Shaw et al. (2015) also showed that the mutation of one amino acid in
Z. Liang et al. Food Research International 147 (2021) 110487

Table 2
Selective studies of fungal terpene synthase (TPS) genes.
Terpene/Terpenoid Fungus TPS Plasmid/Vector Identification approach Reference

Monoterpene
β-Pinene Hypoxylon sp. CI4A-CPS pJBEI-3122, pBbE1a, pBbE2k Heterologous expression in E. coli (Wu et al., 2016)
Myrcene Postia placenta PpSTS25 pGYRG Heterologous expression in E. coli (Ichinose & Kitaoka, 2018)

Monoterpenoid
Linalool Postia placenta PpSTS25 pGYRG Heterologous expression in E. coli (Ichinose & Kitaoka, 2018)
1,8-Cineole Hypoxylon sp. Hyp3 pTrc99 Heterologous expression in E. coli (Shaw et al., 2015)

Sesquiterpene
α-Farnesene Clitopilus pseudo-pinsitus CpSTS14 pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Nagamine et al., 2019)
Germacrene D Stereum hirsutum ShSTS10 pUCBB, pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Quin et al., 2013)
(Nagamine et al., 2019)
α-Selinene Agrocybe aegerita Agr8 pET Heterologous expression in E. coli (Zhang et al., 2020)
β-Selinene Agrocybe aegerita Agr8 pET Heterologous expression in E. coli (Zhang et al., 2020)
δ-Cadinene Agrocybe aegerita Agr1/Agr4 pET Heterologous expression in E. coli (Zhang et al., 2020)
β-Cadinene Agrocybe aegerita Agr8 pET Heterologous expression in E. coli (Zhang et al., 2020)
γ-Cadinene Clitopilus pseudo-pinsitus CpSTS18 pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Nagamine et al., 2019)
γ-Muurolene Clitopilus pseudo-pinsitus CpSTS5 pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Nagamine et al., 2019)
α-Muurolene Agrocybe aegerita Agr3 pET Heterologous expression in E. coli (Zhang et al., 2020)
α-Guaiene Daldinia eschscholzii EC12-PGS pJBEI-3122, pBbE1a, pBbE2k Heterologous expression in E. coli (Wu et al., 2016)
β-Chamigrene Hypoxylon sp. CI4A-CPS pJBEI-3122, pBbE1a, pBbE2k Heterologous expression in E. coli (Wu et al., 2016)
Alloaromadendrene Clitopilus pseudo-pinsitus CpSTS11 pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Nagamine et al., 2019)
α-Gurjunene Postia placenta PpSTS06 pGYRG Heterologous expression in E. coli (Ichinose & Kitaoka, 2018)
β-Gurjunene Galerina marginata Galma_104215 pET Heterologous expression in E. coli (Zhang et al., 2020)
β-Copaene Coniophora puteana Copu2 pACYC Heterologous expression in E. coli (Mischko et al., 2018)
α-Cubebene Stereum hirsutum ShSTS12 pUCBB, pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Quin et al., 2013)
(Nagamine et al., 2019)
β-Cubebene Stereum hirsutum ShSTS12 pUCBB, pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Quin et al., 2013)
(Nagamine et al., 2019)

Sesquiterpenoid
Cubebol Coniophora puteana Copu3 pACYC Heterologous expression in E. coli (Mischko et al., 2018)
Nerolidol Hypoxylon sp. Hyp1 pTrc99 Heterologous expression in E. coli (Shaw et al., 2015)
α-Cadinol Clitopilus pseudo-pinsitus CpSTS3 pC9SC103, pDP1031, pDP1032 Heterologous expression in E. coli (Nagamine et al., 2019)

E. coli, Escherichia coli.

Hypoxylon sp. can lead to the conversation of Hyp3 from one 1,8-cineole
synthase to one limonene, myrcene or farnesene synthase. Currently, the
heterologous genes for the functional expression of TPSs in S. cerevisiae
are mainly derived from plant sources, or synthetically generated
(Paramasivan & Mutturi, 2017). Therefore, fungi might serve as a po­
tential source of heterologous TPS genes with an enormous diversity,
especially for sesquiterpenes.
5. Alternative microbial engineering platforms
5.1. Filamentous fungi
Similar to S. cerevisiae, filamentous fungi can function as a viable host
for heterologous production of terpenes (Table 3), although relevant
research is still at the early stage. It should be noted that a filamentous
fungus of A. oryzae, which has been defined as a Generally Recognized as
Safe (GRAS) organism by US FDA (Food and Drug Administration,
2018), has been adopted to reconstitute biosynthetic machinery, usually
by introducing heterologous terpene biosynthetic genes using genome-
editing method, to yield various terpene products, including aphidico­
lin (Fujii et al., 2011), pyripyropene (Itoh et al., 2010) and andrastin
(Matsuda, Awakawa, & Abe, 2013).
Generally, the terpene biosynthetic genes are amplified from the
genomic DNA and cloned into the expression vectors to obtain expres­
sion plasmids before introducing into fungal host strains for over­
expression (Nagamine et al., 2019). A. oryzae usually serves as a
genetically tractable and efficient ascomycete host for basidiomycete
terpene gene expression via plasmid vectors. For instance, the heterol­
ogous expression of the seven pleuromutilin biosynthesis genes from
Clitopilus passeckerianu (Pl-sdr, Pl-p450-2, Pl-p450-1, Pl-ggs, Pl-cyc, Pl-atf,
Pl-p450-3) in A. oryzae, can boost the production of diterpene product,
pleuromutilin by up to 20-fold compared to C. passeckerianu (Bailey
et al., 2016). However, this bottom-up synthetic method is limited by a
long time required for the insertion of multiple genes. In response, a
rapid reconstitution strategy was developed to introduce heterologous
aflatrem biosynthetic genes into A. oryzae for the biosynthesis of indole
diterpenes aflatrem and β-aflatrem (Tagami et al., 2014). This strategy
requires two rounds of tandem transformation which use two plasmids
carrying the same selectable marker and promoter/terminator but
different target genes. Furthermore, it is possible to simultaneously
prepare several types of transformants with various genes in a single
transformation, and introduce multiple genes with a single selectable
marker gene. The use of multiple plasmids with a small number of target
genes in a tandem transformation has contributed to the rapid and
simple reconstitution of fungal biosynthetic machinery. Fungal clus­
tered regularly interspaced short palindromic repeat (CRISPR)/Cas9
system is a recently developed genome editing technology for heterol­
ogous production of terpenes in A. oryzae. The CRISPR/Cas9 system
comprises only two components, the Cas9 nuclease of Streptococcus
species and a single guide RNA (sgRNA). In the recipient genome, the
sgRNA can recognize the target sequence, which is subsequently cleaved
by the Cas9 protein, creating a double-stranded break. When the break is
repaired by the cell via homologous recombination, gene integration can
be introduced at the target locus (Katayama et al., 2016; Shi et al.,
2019). In the study of Nagamine et al. (2019), A. oryzae can successfully
express 29 sesquiterpene synthase genes and diterpene biosynthetic
genes from C. pseudo-pinsitus and S. hirsutum by introducing CRISPR/
Cas9 plasmid and donor DNA plasmid with target genes. In addition,
CRISPR/Cas9 system has a synergetic effect with forced recycling system
which involves an autonomous maintenance in Aspergillus (AMA1)-
based replicating plasmid containing a drug resistance marker ptrA.
Conditional expression of Aoace2 gene from AMA1-based plasmid is
detrimental to fungal growth, which allows for forced recycling of the
plasmid and thus repeatable genetic engineering (Katayama et al.,
2019). For instance, Liu et al. (2019) reconstituted a Basidiomycota
diterpene erinacine biosynthetic gene cluster in A. oryzae via a

10
Z. Liang et al. Food Research International 147 (2021) 110487

Table 3
Selective studies of terpenes/terpenoids from different engineered filamentous fungi.
Fungi Terpene/ Yield (mg/ Plasmid/Vector Biosynthetic gene Fold Reference
Terpenoid L) increase

Aspergillus oryzae Aphidicolin N/A pTAex3, pPTRI, pUSA, GGPP synthase (PbGGS), diterpene synthase (PbACS), N/A (Fujii et al.,
pAdeA two monooxygenases (PbP450-1 and PbP450-2) genes 2011)
Pyripyropene N/A pTAex3, pPTRI Polyketide synthase (PKS), prenyltransferase (PT), CoA N/A (Itoh et al.,
ligase, flavin adenine dinucleotide (FAD)-dependent 2010)
monooxygenase (FMO), cytochrome P450 (P450),
acetyltransferase (AT) genes
Andrastin N/A pTAex3, pPTRI, pAdeA, PKS, PT, FMO, methyltransferase (MT), terpene cyclase N/A (Matsuda
pUSA (AdrI) genes et al., 2013)
Aflatrem N/A pUARA2, pUSA2 GGPP synthase (PaxG), prenyltransferase (PaxC), N/A (Tagami
epoxidase (PaxM), cyclase (PaxB), cytochrome P450 et al., 2014)
(AtmP/Q), reverse-mode prenyltransferase (AtmD)
genes
Pleuromutilin 84.24 pYES2 Cytochrome P450, AT, terpene cyclase, GGPP 20 (Bailey et al.,
synthetase (GGS), P450 monooxygenase, short-chain 2016)
dehydrogenase/reductase (SDR) genes
N/A pC9SC103, pDP1031, Sesquiterpene synthase, glucose-methanol-choline N/A (Nagamine
pDP1032 (GMC) oxidase, oxidation/reduction enzyme, et al., 2019)
cytochrome P450, flavoprotein, GGPP synthase genes
Erinacine Erinacine pRGE-g801/g201, pC9pA- Cytochrome P450, short-chain dehydrogenase, N/A (Liu et al.,
Q: 4.7 g801, pC9pGg401/g601, polyprenyl synthase, UbiA prenyltransferase, enoyl- 2019)
Erinacine pDP201a, pDP401a, (acyl carrier protein) reductase, UDP-glucoronosyl and
Q2: 1.1 pDP601a UDP-glucosyl transferase, membrane bound O-acyl
transferase, GMC oxidase genes

Aspergillus nidulans Gamma- N/A pHHO5 Fusarium fujikuroi ent-kaurene synthase (cps/ks), N- N/A (Bromann
terpinene, ent- terminally truncated and hemagglutinin (HA)-tagged et al., 2016)
kaurene gamma-terpinene synthase from Citrus unshiu (gTerpS),
cluster-specific transcription factor (pbcR), ent-pimara-8
(14),15-diene synthase, GGPP synthase, cytochrome
P450 genes

Paraconiothyrium Taxol N/A pET32, pPK2, pBlueScript Taxus canadensis GGPP synthase gene 3 (Soliman
SSM001 spp. KS+ et al., 2017)
Alternaria alternata Taxadiene 0.0619 ± pGB123, pGB125, pGB127 IDI, tHMG1, taxadiene synthase (TS) genes N/A (Bian et al.,
0.0063 2017)

N/A, not available; GGPP, geranylgeranyl diphosphate; PKS, polyketide synthase; PT, prenyltransferase; FAD, flavin adenine dinucleotide; FMO, FAD-dependent
monooxygenase; AT, acetyltransferase; MT, methyltransferase; SDR, short-chain dehydrogenase/reductase; GMC, glucose-methanol-choline; UDP, uridine diphos­
phate; HA, hemagglutinin; TS, taxadiene synthase.

combination of CRISPR/Cas9 system and forced recycling system, pro­
ducing erinacine Q and Q2 at the yield levels of 4.7 and 1.1 mg/L,
respectively. This combined strategy can eliminate the lengthy
screening step and use unlimited biosynthetic gene introduction because
of marker-free genome editing.
Transgenic technology can also apply to other filamentous fungi to
improve terpene production. In the study of Bromann, Toivari, Viljanen,
Ruohonen, and Nakari-Setälä (2016), Aspergillus nidulans was engi­
neered to heterologously produce gamma-terpinene and ent-kaurene.
Fusarium fujikuroi ent-kaurene synthase (cps/ks), and the N-terminally
truncated and hemagglutinin (HA)-tagged gamma-terpinene synthase
from Citrus unshiu (gTerpS), were overexpressed in A. nidulans strain by
using a plasmid that incorporated the synthase genes under the control
of gpdA promoter. A plant GGPPS gene was also proved to be feasible for
integration into the Taxol-producing Paraconiothyrium SSM001 genome
via Agrobacterium tumefaciens-mediated transformation (ATMT) to in­
crease the GGPPS copy number, which enhanced the production level of
a diterpene anticancer drug, Taxol (paclitaxel) (Soliman, Mosa, El-
Keblawy, & Husseiny, 2017). ATMT was carried out by delivering the
genetic components that are bordered by the left border (LB) and right
border (RB) of its helper plasmid derived from three different
A. tumefaciens Ti plasmids (octopine, nopaline and L,L-succinamopine
plasmids) (Hood, Gelvin, Melchers, & Hoekema, 1993). In addition,
feeding the GGPPS-transformed fungus with exogenous GGPP can
further improve the synthesis efficiency of both terpenes and Taxol. The
introduction of new copies of GGPPS requires more terpene precursors
such as GPP or FPP which are consumed by the original fungal GGPPS to
produce native GGPP for the synthesis of essential diterpenes. Therefore,
the addition of extra GGPP can be adopted by the original terpene
pathway, which ensures FPP and/or GPP supply to activate GGPPS to
improve its transcription and Taxol production (Soliman et al., 2017).
Bian et al. (2017) also developed an effective alternative to the semi-
chemical synthesis of Taxol using endophytic fungus Alternaria alter­
nate TPF6. Specifically, a heterologous MVA pathway was transformed
into A. alternata TPF6, then engineered by introducing the plant
taxadiene-forming gene via ATMT method, resulting in the production
of 61.9 ± 6.3 μg/L taxadiene, which can be converted into Taxol by 19
enzymatic steps including hydroxylation and other oxygenation re­
actions of the taxadiene skeleton (Hezari & Croteau, 1997). Up to date,
the potential of filamentous fungi as a terpene-producing platform has
not been fully exploited, particularly in the engineering and optimiza­
tion of endogenous MVA pathway.
5.2. Escherichia coli
With the advancement of metabolic engineering and transgenic
technology, yeast is viewed as a convenient and efficient host for large-
scale terpene production (Paramasivan & Mutturi, 2017). However,
some engineered yeasts are incapable of enhancing the yield of certain
terpene products such as cineole and cubebol (Asadollahi et al., 2008;
Asadollahi et al., 2010; Ignea et al., 2011). In response, E. coli serves as a
promising alternative host (Moser & Pichler, 2019). E. coli has a similar
genetic background to yeast, indicating its great potential in terpene
synthesis using microbial-directed heterologous pathway (Pitera, Pad­
don, Newman, & Keasling, 2007).
The advantages of E. coli in terpene synthesis are mainly ascribed to
two aspects. First, E. coli is a highly efficient host for terpene biosyn­
thesis. E. coli can synthesize the basic structure units of terpenes via its

11
Z. Liang et al.
inherent MEP pathway which surpasses the MVA pathway in stoichi­
ometry and accumulates fewer byproducts (Dugar & Stephanopoulos,
2011). Compared to S. cerevisiae, E. coli consumes less time when pro­
ducing the same terpene due to its faster growth rate (Ingy & Wim,
2017). Secondly, E. coli is more tolerant of monoterpene products such
as limonene and α-pinene (Himejima, Hobson, Otsuka, Wood, & Kubo,
1992). According to Chubukov et al. (2015), the non-oxidized limonene
exhibited minimal toxicity to wild-type E. coli, while a point mutation in
alkyl hydroperoxidase (ahpC) can further alleviate the acute limonene
toxicity caused by limonene hydroperoxide. E. coli also displayed
enhanced tolerance to cyclohexane, pinene and limonene by expressing
the AcrAB-TolC efflux proteins (Aono, Tsukagoshi, & Yamamoto, 1998;
Dunlop et al., 2011). However, E. coli is also reported with multiple
drawbacks as a microbial host. So far, E. coli has not been classified as a
GRAS organism by US FDA due to the possible contamination of the final
product by endotoxins (Ingy & Wim, 2017). Also, the MEP pathway in
E. coli suffers from the natural cryptic regulation via various mechanisms
including cellular compartmentalization, gene duplication, bifunctional
fusion proteins, transcriptional control and redox regulation. Introduc­
tion of the heterologous MVA pathway to E. coli might offer a bypass to
the highly regulated MEP pathway (Davies, Jinkerson, & Posewitz,
2015). For instance, E. coli heterologously expressing a functional MVA
pathway, along with the overexpression of FPPS and integral membrane
phosphatase PgpB, can produce up to 526.1 mg/L of farnesol that is
higher than the reported yield (145.7 mg/L) in S. cerevisiae (Wang, Park,
Choi, & Kim, 2016). In an in silico comparison between the MEP and
MVA pathways with different hosts, E. coli is superior to S. cerevisiae in
the IPP production when using the heterologous MVA pathway, the
native MEP pathway, and both pathways combined, respectively (Meng,
Wang, Hua, Zhang, & Wang, 2011). Further, the MVA and MEP path­
ways exist a synergic effect for terpene synthesis. In the study of Wang,
Quan, et al. (2019), the flux of the MVA and MEP pathway in the dual
pathway overexpressing E. coli strain were 1.5- and 4.8-fold higher than
that of the strains overexpressing MEP and MVA pathway alone,
respectively. In addition, cytochrome P450 expression in E. coli is poor
compared with S. cerevisiae since most P450s are membrane-anchored to
the endoplasmic reticulum which are inserted by P450 reductases for co-
expression (Moser & Pichler, 2019). Indeed, E. coli is a viable terpene-
producing host, but the efficiency of E. coli and S. cerevisiae in terpene
production need to be further investigated and compared for selecting
the most suitable microbial host.
Recent advances in the metabolic engineering of E. coli for terpene
production is supported by research efforts (Table 4). In the study of
Jeong et al. (2020), E. coli HP variant obtained from parental E. coli
DH5α strain by adaptive evolution, displayed an increased cell fitness in
terpene synthesis under adenylate cyclase (CyaA) mutation, as evi­
denced by a five-fold increase to 193.2 mg/L in taxadiene production.
Specifically, an insertion of mobile element IS10 in CyaA successfully
reduced the intracellular concentration of cyclic adenosine mono­
phosphate (cAMP), which can regulate its receptor protein (CRP) to
rewire cell metabolism and thus improve cell fitness for terpene pro­
duction. In addition, Zong et al. (2019) reported a 29.51-fold increase in
nerol production in E. coli through a heterologous MVA pathway, when
combining the overexpression of acetyl-CoA acetyltransferase (ERG10)
amplified from S. cerevisiae genomic DNA. An attempt of replacing
Erg20p with A. grandis geranyl diphosphate synthase (GPPS2) was also
observed in promoting linalool production in E. coli with Actinidia arguta
linalool synthase by increasing the availability of GPP (Kong, Fu, Li, Pan,
& Guo, 2020).
5.3. Trends in microbial terpene production
Currently, plant-derived terpene products are unable to meet the
gradually growing usage needs for commercial applications, rendering
the microorganism as a feasible alternative terpene-producing platform.
However, microbial cell factories still demand further improvement for
12
Food Research International 147 (2021) 110487
the commercial production of terpenes, especially in precursor supply,
pathway optimization and microbial tolerance to terpene products. In
the future, more efforts should be invested toward identifying the bot­
tlenecks of MVA or MEP pathways, so as to maximize the flux and in­
crease the precursor supply. Engineering of terpene synthase enzymes
and introducing a heterologous pathway can be performed in parallel to
test potential combinatorial effects on terpene production. Engineering
and optimization of orthogonal pathway could be a main direction of
future research on S. cerevisiae as a terpene-producing platform, which
can be attained by reducing the flux of competing pathway and chan­
neling the metabolic flux of MVA pathway toward NPP. Filamentous
fungi as a potential microbial cell factory also expect the development of
engineering strategies to modify the endogenous MVA pathway for
enhancing terpene synthesis. In addition, the synergic effect of MVA and
MEP pathways overexpressing in E. coli on production of specific ter­
penes is worth further investigation. Future research should also
consider the potential strategies in increasing microbial tolerance or
attenuate terpene toxicity in the presence of high levels of terpene
products. The effect of the fine-tuning of multi-gene expression on the
synthesis of other isoprenoid-derived compounds also awaits further
exploration.
While E. coli is the mostly used model bacterial platform, other
bacteria such as Bacillus subtilis has also been engineered for terpene
production. B. subtilis is regarded as an attractive host for terpene bio­
production due to its high isoprene productivity (Pramastya, Song,
Elfahmi, Sukrasno, & Quax, 2020) and its generally recognized as safe
(GRAS) status (Food and Drug Administration, 2018). In recent years,
B. subtilis has been reported effective in heterologous production of
terpenes and their intermediate via different engineering strategies
including CRISPR/Cas9 system, upregulation of MEP pathway genes
(dxs, ispD, ispF, ispH, ispE, ispC, ispG, idi) and overexpression of GGPPS
and FPPS (Pramastya et al., 2020; Song et al., 2020; Song et al., 2021).
Recent research also confirmed that portable bacterial CRISPR tran­
scriptional activation method can be applied to the metabolic engi­
neering in Pseudomonas putida, which makes it a possible host for
terpene production (Kiattisewee et al., 2021). Apart from bacteria,
another major prokaryotic host is cyanobacteria. Cyanobacteria are
considered a potential microbial platform for sustainable bioproduction
because these photosynthetic prokaryotes require only CO2 and light for
growth. Enhanced terpene production in these microorganisms can be
attained using a combinatorial metabolic engineering approach of mu­
tation in GGPPS gene, modulation of GPPS expression, and optimization
of the MEP pathway such as overexpression of IDI and 1-deoxy-D-xylu­
lose-5-phosphate synthase (DXS) (Lin, Zhang, & Pakrasi, 2021; Rodri­
gues & Lindberg, 2021). Further, other emerging bacterial platforms
such as Rhodobacter capsulatus, Pseudomonas putida and Corynebacterium
glutamicum have been developed. Valencene, patchoulol and many other
terpenes have been successfully produced by these heterogeneous pro­
duction platforms (Kogure & Inui, 2018; Loeschcke & Thies, 2015;
Troost et al., 2019).
In addition to Saccharomyces as the model yeast platform, attempts
have also been made to other yeast species, such as Yarrowia lipolytica
and Lipomyces starkeyi which can be used to produce (-)-α-bisabolol,
linalool, α-farnesene, α-santalene, α‑zingiberene by overexpressing the
rate-limiting enzymes (e.g. HMGR, FPPS, IDI) of the MVA pathway and
downregulating the competing squalene synthesis pathway (Cao, Wei,
Lin, & Hua, 2017; Dai et al., 2021; Jia et al., 2019; Ma et al., 2021; Yang,
Nambou, Wei, & Hua, 2016). In addition to yeast and filamentous fungi,
researchers have also explored other eukaryotic hosts such as Chlamy­
domonas reinhardtii and Phaeodactylum tricornutum as new microbial
platforms, which can produce betulin, 13R(+) manoyl oxide, (E)-
α-bisabolene (D’Adamo et al., 2019; Lauersen et al., 2018; Wichmann,
Baier, Wentnagel, Lauersen, & Kruse, 2018). Overall, studies on these
emerging platforms are still limited compared to S. cerevisiae, E. coli and
Aspergillus, thus it may have a long way to go to achieve their com­
mercial applications.
Z. Liang et al. Food Research International 147 (2021) 110487

Table 4
Selective studies of recent advanced engineering strategies for terpene/terpenoid production in Escherichia coli.
Terpene/ Maximum yield Strain Engineering strategy Terpene synthase Reference
Terpenoid (mg/L)

Farnesol 526.1 DH5α a. Introduction of a heterologous MVA pathway N/A (Wang et al.,
b. Overexpression of FPPS and integral membrane 2016)
phosphatases PgpB and YbjG
Taxadiene 193.2 Adaptive HP variant a. Introduction of a heterologous MVA pathway Synthetic taxadiene (Jeong et al.,
of DH5α b. CyaA mutation synthase 2020)
Nerol 1.564 ± 0.102 K-12 MG1655 (DE3) a. Introduction of a heterologous MVA pathway Nerol synthase from Glycine (Zong et al.,
b. Overexpression of IDI1, MVD1, ERG8, ERG12, tHMG1 and max 2019)
ERG13
c. Overexpression of acetyl-CoA acetyltransferase (ERG10)
Linalool 63 ± 5.6 MG165 a. Introduction of a heterologous MVA pathway Linalool synthase from (Kong et al.,
b. Expression of Abies grandis geranyl diphosphate synthase Actinidia arguta 2020)
(GPPS2) instead of Erg20p
c. Overexpression of IDI1

N/A, not available; MVA, mevalonate; FPPS, farnesyl diphosphate (FPP) synthase; CyaA, adenylate cyclase.

6. Potential food applications of terpene products from
engineered microorganisms
Up to date, metabolic engineering of S. cerevisiae and E. coli are
gradually applied in food and beverage industries for the production of
various terpene products potentially acting as flavoring agents, sweet­
ening agents, preservatives and functional food ingredients (Table 5).
S. cerevisiae is more attractive for food applications due to its safety
which promotes approval of new food products (Lyu et al., 2019). In
contrast, filamentous fungi as a terpene-producing platform have not
been fully explored for food application yet, thus a promising new area
for food industry. Geraniol is one of the most in-demand
monoterpenoids with a pleasant rose-like aroma, which can be
massively produced in an engineered S. cerevisiae strain by deploying
amino acid mutations in ERG20, replacing the ERG20 promoter with
HXT1 promoter, or integrating extra IDI1 gene into the yeast genome
(Fischer et al., 2011; Liu, Zhang, et al., 2013; Zhao et al., 2017). It is a
flavoring additive applied in chewing gum, candy, ice cream, beverages,
and many other products (Kutyna & Borneman, 2018). During the
alcoholic fermentation of non-aromatic Parellada grape must, the
expression of the Ocimum basilicum (sweet basil) geraniol synthase gene
in a S. cerevisiae wine strain can successfully yield 750 μg/L of geraniol,
which is higher than its olfactory sensory threshold and its quantities in
wines obtained from the highly aromatic Muscat grapes. Further,

Table 5
Bioactivities and food applications of terpene products from engineered microorganisms.
Food category Terpene product Engineered Bioactivity Food application Reference
microorganism

Flavoring agent Geraniol S. cerevisiae Natural aroma/flavor a. Aroma/flavor additive in (Kutyna &
chewing gum, candy, ice Borneman,
cream, beverages 2018)
b. Aroma and monoterpene (Pardo et al.,
enhancement in wine 2015)
c. Replacing flavor hop (Denby et al.,
addition in classical 2018)
brewing
Linalool S. cerevisiae Natural aroma/flavor Replacing flavor hop (Denby et al.,
addition in classical 2018)
brewing
Limonene S. cerevisiae, E. coli Natural aroma/flavor Aroma/flavor additive in (Duetz et al.,
candy and soft drinks 2003)
Nerolidol S. cerevisiae Natural aroma/flavor N/A (Kutyna &
Borneman,
2018)
Valencene S. cerevisiae Natural aroma/flavor Aroma/flavor additive in (Beekwilder
food and drinks et al., 2014)

Sweetening Steviol glycosides S. cerevisiae, E. coli, Sweet taste Alternative and low-caloric (Prakash et al.,
agent A. nidulans sweetener 2014)
Mogroside V S. cerevisiae Sweet taste Sweetener in low-calorie (Zhao & Li,
sweet beverages 2018)

Preservative Monoterpenoids (geraniol, linalool, S. cerevisiae, E. coli Antimicrobial activity N/A (Lyu et al.,
limonene, etc.), sesquiterpenoids 2019)
(farnesene, cubebol, farnesol, etc.)

Functional food Limonene S. cerevisiae, E. coli Gastroprotective, antiviral, antihyperalgesic, Imparting health benefits to (Vieira et al.,
ingredients antidiabetic, anticancer, antinociceptive, Chinese Baijiu 2018)
antioxidant and anti-inflammatory activities (Hu et al.,
2020)
Steviol glycosides S. cerevisiae, E. coli, Anti-hyperglycemic, antihypertensive, anti- N/A (Yılmaz et al.,
A. nidulans inflammatory, anti-carcinogenic and 2020)
antioxidant activities
Geraniol, linalool, nerol S. cerevisiae Antioxidant activity Formulation of antioxidant (Wang, Chen
rich functional food et al., 2019)

N/A, not available; S. cerevisiae, Saccharomyces cerevisiae; E. coli, Escherichia coli; A. nidulans, Aspergillus nidulans.

13
Z. Liang et al.
geraniol can be metabolized by yeast enzymes to excess monoterpenes,
leading to a significant increase in total monoterpene concentration in
wine. Therefore, the geraniol-engineered yeast can positively affect the
terpene profile of wine, resulting in a monoterpene enhanced wine with
fruity and flowery aromas (Pardo et al., 2015). In specific types of beer,
the dry-hopped flavor and aroma are mostly ascribed to the presence of
geraniol and linalool, which originate from the hop flowers added to the
wort during fermentation in classical brewing. In the study of Denby
et al. (2018), S. cerevisiae as a brewing strain can be metabolically
engineered by overexpressing tHMGR and mutated FPPS (FPPS*), so as
to increase GPP accumulation. This strain was also equipped with a full-
length geraniol synthase from O. basilicum (ObGES) and truncated
linalool synthase from Mentha citrata (t67-McLIS) to produce linalool
and geraniol, thereby replacing flavor hop addition. Limonene is
another monoterpene frequently used as a flavoring agent in candy and
soft drinks (Duetz, Bouwmeester, van Beilen, & Witholt, 2003), which
can be yielded in large amount in engineered S. cerevisiae by expressing
tHMG1 mutant (Behrendorff et al., 2013) or ERG20 mutant (Jongedijk
et al., 2015), or by establishing an NPP-based orthogonal pathway
(Cheng et al., 2019; Hu et al., 2020; Ignea et al., 2019), or in engineered
E. coli with a heterologous MVA pathway (Alonso-Gutierrez et al., 2013).
Nerolidol is a typical sesquiterpene officially permitted as food flavoring
additive by US FDA (Food and Drug Administration, 2020), giving a
wood- and fresh bark-like scent (Kutyna & Borneman, 2018). The het­
erologous production of this sesquiterpene by S. cerevisiae can be largely
enhanced in the background of ERG9 downregulation and/or tHMG1
overexpression (Jackson, 2005). Another sesquiterpene valencene is
also commercially used as an additive to impart or reinforce fruity and
woody flavor in food and drinks with a market volume of about 10,000
kg per year (Beekwilder et al., 2014). This bicyclic sesquiterpene can be
massively produced in engineered S. cerevisiae via different engineering
strategies including tHMG1 overexpression (Farhi et al., 2011), ERG9
downregulation (Asadollahi et al., 2008; Chen et al., 2019) and ROX1
deletion (Chen et al., 2019; Jakočiūnas et al., 2015).
At present, there is a growing consumer demand for natural sweet­
ening agents, which can offer the similar sweetness of common sugar
and restrict the high calorie intake. Therefore, S. cerevisiae is recom­
mended as a microbial cell factory to manufacture alternative and low-
caloric sweetening agents. Steviol glycosides are typical sweeteners that
are a group of diterpene glycosides (rubusoside, steviolbioside, dulco­
side A, stevioside, rebaudioside-A, B, C, D, E, F, and M). They can be de
novo synthesized from GGPP with five heterologous enzymes in
S. cerevisiae, conferring characteristic sweet tastes that are about 30 to
250 times more potent than sucrose (Gold et al., 2018; Prakash, Mar­
kosyan, & Bunders, 2014). E. coli is also a microbial chassis for the high-
level production of steviol glycosides. The ent-copalyl diphosphate
synthase (CPPS) and ent-kaurene synthase (KS) from Stevia rebaudiana
can be heterologously expressed in E. coli to produce ent-kaurene, which
is a dedicated precursor responsible for steviol glycoside biosynthesis
(Kong, Kang, Kim, Oh, & Lee, 2015). It is worth mentioning that ent-
kaurene can also be produced in engineered A. nidulans by over­
expressing F. fujikuroi ent-kaurene synthase (Bromann et al., 2016),
indicating the vast potential of filamentous fungi as an engineering
platform for desired terpene products in food industry. Similarly, mog­
roside V also serves as a sweetener in low-calorie sweet beverages since
it is nearly 300 times sweeter than sucrose. It is a cucurbitane-type tri­
terpenoid saponin derived from the introduction of Arabidopsis thaliana
cytochrome P450 reductase 1 (AtCPR1) and various genes encoding key
enzymes in the biosynthesis pathway in S. cerevisiae (Zhao & Li, 2018).
Multiple monoterpenoids (geraniol, linalool, limonene, etc.) and
sesquiterpenoids (farnesene, cubebol, farnesol, etc.) generated through
metabolic engineering of S. cerevisiae and E. coli, are potential food
preservatives due to their high antimicrobial properties. However, the
microbial technology has not been engaged in the industrial production
of food preservatives, which is mainly due to the low yield, poor toxicity
resistance of yeast to terpene products, as well as the high production
14
Food Research International 147 (2021) 110487
costs from feedstock supply (Lyu et al., 2019). It is worth mentioning
that the food additives produced with genetically engineered microbes
usually get low consumer acceptance. Therefore, sustained information
and education are required in parallel with the application of the mi­
crobial technology to overcome the negative public perception
(Schempp et al., 2018).
Several microbial engineering derived terpene products can poten­
tially serve as bioactive components of functional food due to their
multiple health benefits. For instance, limonene is reported as being
gastroprotective, antiviral, antihyperalgesic, antidiabetic, anticancer,
antinociceptive, antioxidant and anti-inflammatory, indicating its po­
tential application in functional food and high-value food products
(Vieira, Beserra, Souza, Totti, & Rozza, 2018). During Chinese Baijiu
fermentation, Hu et al. (2020) successfully established an orthogonal
pathway in S. cerevisiae for elevating the heterologous production of
(+)-limonene which can impart potential health benefits to Chinese
Baijiu. Steviol glycosides are diterpenoid sweeteners derived from
engineered microorganisms, which could be another bioactive compo­
nent of functional food by possessing anti-hyperglycemic, antihyper­
tensive, anti-inflammatory, anti-carcinogenic and antioxidant activities
(Yılmaz, Görgüç, Uygun, & Bircan, 2020). Likewise, geraniol, linalool
and nerol with effective antioxidant activities, which are monoterpenoid
products produced from engineered microorganisms, may also
contribute to the formulation of antioxidant rich functional food (Wang,
Chen, & Hou, 2019).
More research efforts need to be invested in terpene production using
microbial engineering platforms in food industry. Future research could
explore the potential of filamentous fungi in the production of desired
terpenes for food application. For instance, A. oryzae is a promising
terpene-producing platform for this purpose due to its GRAS status. It is
also worthy of exploring the production of valuable terpenes by engi­
neering S. cerevisiae during wine fermentation, which can modify or
improve the aroma/flavor of wine. In addition, the technical obstacles in
food application of microbial engineering derived terpene products, the
safety concern regarding to the microbial engineering in food process­
ing, and the influence of the processing parameters on the terpene
product quality also deserve further investigation.
7. Conclusion
Terpene production via microbial cell factories is of great signifi­
cance to substitute the traditional supply way. This review introduces
the biosynthetic pathways of terpenes in different organisms. Both MVA
and MEP pathways exist in plant cells, while S. cerevisiae and other fungi
only harbor the MVA pathway. Bacteria such as E. coli only have
inherent MEP pathway and can be introduced a heterologous MVA
pathway through engineering. Engineering of S. cerevisiae for promoting
terpene synthesis are described comprehensively, where overexpression
of HMGR variant, ERG20 upregulation, ERG9 downregulation, IDI1 gene
integration, ROX1 deletion and orthogonal pathway construction
constitute the main engineering strategies to enhance terpene produc­
tion. In recent years, numerous TPS genes have been identified in fungi,
serving as potential heterologous TPS genes for the functional TPSs
expression in S. cerevisiae. Apart from S. cerevisiae, filamentous fungi and
E. coli also serve as efficient hosts for terpene synthesis. Nowadays, the
diverse terpene products from engineered microorganisms have been
gradually applied in food and beverage industries as flavoring agents,
sweetening agents, preservatives and functional food ingredients.
Future studies are required to investigate the heterologous production of
fungi-derived terpenes in S. cerevisiae and to improve the efficiency of
S. cerevisiae, filamentous fungi and E. coli in terpene production, so as to
expand their production scale and promote industrial applications.
Declaration of Competing Interest
The authors declare that they have no known competing financial
Z. Liang et al. Food Research International 147 (2021) 110487

interests or personal relationships that could have appeared to influence Cao, X., Wei, L.-J., Lin, J.-Y., & Hua, Q. (2017). Enhancing linalool production by
engineering oleaginous yeast Yarrowia lipolytica. Bioresource Technology, 245,
the work reported in this paper.
1641–1644. https://doi.org/10.1016/j.biortech.2017.06.105.
Chen, H., Zhu, C., Zhu, M., Xiong, J., Ma, H., Zhuo, M., & Li, S. (2019). High production
of valencene in Saccharomyces cerevisiae through metabolic engineering. Microbial
Acknowledgement Cell Factories, 18(1), 1–14. https://doi.org/10.1186/s12934-019-1246-2.
Chen, Y., Partow, S., Scalcinati, G., Siewers, V., & Nielsen, J. (2012). Enhancing the copy
number of episomal plasmids in Saccharomyces cerevisiae for improved protein
This review research was supported by the food science research production. FEMS Yeast Research, 12(5), 598–607. https://doi.org/10.1111/j.1567-
group of the Faculty of Veterinary and Agricultural Sciences, University 1364.2012.00809.x.
Cheng, S., Liu, X., Jiang, G., Wu, J., Zhang, J.-L., Lei, D., … Zhao, G.-R. (2019).
of Melbourne. Orthogonal engineering of biosynthetic pathway for efficient production of limonene
in Saccharomyces cerevisiae. ACS Synthetic Biology, 8(5), 968–975. https://doi.org/
10.1021/acssynbio.9b00135.
Chubukov, V., Mingardon, F., Schackwitz, W., Baidoo, E. E., Alonso-Gutierrez, J., Hu, Q.,
Author Contribution
… Mukhopadhyay, A. (2015). Acute limonene toxicity in Escherichia coli is caused by
limonene hydroperoxide and alleviated by a point mutation in alkyl
Zijian Liang and Hang Zhi drafted the original manuscript. Critical hydroperoxidase AhpC. Applied Environmental Microbiology, 81(14), 4690–4696.
inputs and corrections were successively provided by Zhongxiang Fang, https://doi.org/10.1128/AEM.01102-15.
D’Adamo, S., Schiano di Visconte, G., Lowe, G., Szaub-Newton, J., Beacham, T.,
and Pangzhen Zhang during the preparation process. Pangzhen Zhang Landels, A., … Matthijs, M. (2019). Engineering the unicellular alga Phaeodactylum
helped in the conception and structure design of the manuscript and tricornutum for high-value plant triterpenoid production. Plant Biotechnology Journal,
final proofing of the manuscript for submission. 17(1), 75–87. https://doi.org/10.1111/pbi.12948.
Dai, Z., Liu, Y., Zhang, X., Shi, M., Wang, B., Wang, D., … Zhang, X. (2013). Metabolic
engineering of Saccharomyces cerevisiae for production of ginsenosides. Metabolic
References Engineering, 20, 146–156. https://doi.org/10.1016/j.ymben.2013.10.004.
Dai, Z., Pomraning, K. R., Panisko, E. A., Hofstad, B. A., Campbell, K. B., Kim, J., …
Magnuson, J. K. (2021). Genetically engineered oleaginous yeast Lipomyces starkeyi
Alonso-Gutierrez, J., Chan, R., Batth, T. S., Adams, P. D., Keasling, J. D., Petzold, C. J., &
for sesquiterpene α-zingiberene production. ACS Synthetic Biology. https://doi.org/
Lee, T. S. (2013). Metabolic engineering of Escherichia coli for limonene and perillyl
10.1021/acssynbio.0c00503.
alcohol production. Metabolic Engineering, 19, 33–41. https://doi.org/10.1016/j.
Davies, F. K., Jinkerson, R. E., & Posewitz, M. C. (2015). Toward a photosynthetic
ymben.2013.05.004.
microbial platform for terpenoid engineering. Photosynthesis Research, 123(3),
Aono, R., Tsukagoshi, N., & Yamamoto, M. (1998). Involvement of outer membrane
265–284. https://doi.org/10.1007/s11120-014-9979-6.
protein TolC, a possible member of the mar-sox regulon, in maintenance and
DeBose-Boyd, R. A. (2008). Feedback regulation of cholesterol synthesis: Sterol-
improvement of organic solvent tolerance of Escherichia coli K-12. Journal of
accelerated ubiquitination and degradation of HMG CoA reductase. Cell Research, 18
Bacteriology, 180(4), 938–944. https://doi.org/10.1128/JB.180.4.938-944.1998.
(6), 609–621. https://doi.org/10.1038/cr.2008.61.
Asadollahi, M. A., Maury, J., Møller, K., Nielsen, K. F., Schalk, M., Clark, A., & Nielsen, J.
Denby, C. M., Li, R. A., Vu, V. T., Costello, Z., Lin, W., Chan, L. J. G., … Keasling, J. D.
(2008). Production of plant sesquiterpenes in Saccharomyces cerevisiae: Effect of
(2018). Industrial brewing yeast engineered for the production of primary flavor
ERG9 repression on sesquiterpene biosynthesis. Biotechnology and Bioengineering, 99
determinants in hopped beer. Nature Communications, 9(1), 965. https://doi.org/
(3), 666–677. https://doi.org/10.1002/bit.21581.
10.1038/s41467-018-03293-x.
Asadollahi, M. A., Maury, J., Schalk, M., Clark, A., & Nielsen, J. (2010). Enhancement of
Duetz, W. A., Bouwmeester, H., van Beilen, J. B., & Witholt, B. (2003). Biotransformation
farnesyl diphosphate pool as direct precursor of sesquiterpenes through metabolic
of limonene by bacteria, fungi, yeasts, and plants. Applied Microbiology and
engineering of the mevalonate pathway in Saccharomyces cerevisiae. Biotechnology
Biotechnology, 61(4), 269–277. https://doi.org/10.1007/s00253-003-1221-y.
and Bioengineering, 106(1), 86–96. https://doi.org/10.1002/bit.22668.
Dugar, D., & Stephanopoulos, G. (2011). Relative potential of biosynthetic pathways for
Bach, T. (1986). Hydroxymethylglutaryl-CoA reductase, a key enzyme in phytosterol
biofuels and bio-based products. Nature Biotechnology, 29(12), 1074–1078. https://
synthesis? Lipids, 21(1), 82–88. https://doi.org/10.1007/BF02534307.
doi.org/10.1038/nbt.2055.
Bailey, A. M., Alberti, F., Kilaru, S., Collins, C. M., de Mattos-Shipley, K., Hartley, A. J., …
Dunlop, M. J., Dossani, Z. Y., Szmidt, H. L., Chu, H. C., Lee, T. S., Keasling, J. D., …
Foster, G. D. (2016). Identification and manipulation of the pleuromutilin gene
Mukhopadhyay, A. (2011). Engineering microbial biofuel tolerance and export using
cluster from Clitopilus passeckerianus for increased rapid antibiotic production.
efflux pumps. Molecular Systems Biology, 7(1), 487. https://doi.org/10.1038/
Scientific Reports, 6(1), 25202. https://doi.org/10.1038/srep25202.
msb.2011.21.
Beekwilder, J., van Houwelingen, A., Cankar, K., van Dijk, A. D., de Jong, R. M.,
Farhi, M., Marhevka, E., Masci, T., Marcos, E., Eyal, Y., Ovadis, M., … Vainstein, A.
Stoopen, G., … Bosch, D. (2014). Valencene synthase from the heartwood of Nootka
(2011). Harnessing yeast subcellular compartments for the production of plant
cypress (Callitropsis nootkatensis) for biotechnological production of valencene. Plant
terpenoids. Metabolic Engineering, 13(5), 474–481. https://doi.org/10.1016/j.
Biotechnology Journal, 12(2), 174–182. https://doi.org/10.1111/pbi.12124.
ymben.2011.05.001.
Behrendorff, J. B., Vickers, C. E., Chrysanthopoulos, P., & Nielsen, L. K. (2013). 2, 2-
Fischer, M. J., Meyer, S., Claudel, P., Bergdoll, M., & Karst, F. (2011). Metabolic
Diphenyl-1-picrylhydrazyl as a screening tool for recombinant monoterpene
engineering of monoterpene synthesis in yeast. Biotechnology Bioengineering, 108(8),
biosynthesis. Microbial Cell Factories, 12(1), 1–12. https://doi.org/10.1186/1475-
1883–1892. https://doi.org/10.1002/bit.23129.
2859-12-76.
Food and Drug Administration (2018). Microorganisms & microbial-derived ingredients
Bian, G., Yuan, Y., Tao, H., Shi, X., Zhong, X., Han, Y., & Liu, T. (2017). Production of
used in food (partial list). Retrieved from https://www.fda.gov/food/generally-rec
taxadiene by engineering of mevalonate pathway in Escherichia coli and endophytic
ognized-safe-gras/microorganisms-microbial-derived-ingredients-used-food-partia
fungus Alternaria alternata TPF6. Biotechnology Journal, 12(4), 1600697. https://doi.
l-list. Accessed January 3, 2020.
org/10.1002/biot.201600697.
Food and Drug Administration (2020). Code of federal regulations (CFR). Title 21: Food
Blackwell, M. (2011). The Fungi: 1, 2, 3 … 5.1 million species? American Journal of
and drugs. Chapter I - food and drug administration, department of health and
Botany, 98(3), 426-438. https://doi.org/10.3732/ajb.1000298.
human services, subchapter B - food for human consumption (continued), Part 172 -
Brennan, T. C., Williams, T. C., Schulz, B. L., Palfreyman, R. W., Krömer, J. O., &
food additives permitted for direct addition to food for human consumption, subpart
Nielsen, L. K. (2015). Evolutionary engineering improves tolerance for replacement
f - flavoring agents and related substances, subpart 172.510 - natural flavoring
jet fuels in Saccharomyces cerevisiae. Applied Environmental Microbiology, 81(10),
substances and natural substances used in conjunction with flavors. Retrieved from
3316–3325. https://doi.org/10.1128/AEM.04144-14.
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?
Brennan, T. C. R., Krömer, J. O., & Nielsen, L. K. (2013). Physiological and
fr=172.510. Accessed September 10, 2020.
transcriptional responses of Saccharomyces cerevisiae to d-limonene show changes to
Fujii, R., Minami, A., Tsukagoshi, T., Sato, N., Sahara, T., Ohgiya, S., … Oikawa, H.
the cell wall but not to the plasma membrane. Applied and Environmental
(2011). Total biosynthesis of diterpene aphidicolin, a specific inhibitor of DNA
Microbiology, 79(12), 3590–3600. https://doi.org/10.1128/AEM.00463-13.
polymerase α: Heterologous expression of four biosynthetic genes in Aspergillus
Brennan, T. C. R., Turner, C. D., Krömer, J. O., & Nielsen, L. K. (2012). Alleviating
oryzae. Bioscience, Biotechnology, and Biochemistry, 75(9), 1813–1817. https://doi.
monoterpene toxicity using a two-phase extractive fermentation for the
org/10.1271/bbb.110366.
bioproduction of jet fuel mixtures in Saccharomyces cerevisiae. Biotechnology and
Gold, N. D., Fossati, E., Hansen, C. C., DiFalco, M., Douchin, V., & Martin, V. J. J. (2018).
Bioengineering, 109(10), 2513–2522. https://doi.org/10.1002/bit.24536.
A combinatorial approach to study cytochrome P450 enzymes for de novo
Bromann, K., Toivari, M., Viljanen, K., Ruohonen, L., & Nakari-Setälä, T. (2016).
production of steviol glucosides in baker’s yeast. ACS Synthetic Biology, 7(12),
Engineering Aspergillus nidulans for heterologous ent-kaurene and gamma-terpinene
2918–2929. https://doi.org/10.1021/acssynbio.8b00470.
production. Applied Microbiology and Biotechnology, 100(14), 6345–6359. https://
Greenhagen, B. T., O’Maille, P. E., Noel, J. P., & Chappell, J. (2006). Identifying and
doi.org/10.1007/s00253-016-7517-5.
manipulating structural determinates linking catalytic specificities in terpene
Burg, J. S., & Espenshade, P. J. (2011). Regulation of HMG-CoA reductase in mammals
synthases. Proceedings of the National Academy of Sciences, 103(26), 9826. https://
and yeast. Progress in Lipid Research, 50(4), 403–410. https://doi.org/10.1016/j.
doi.org/10.1073/pnas.0601605103.
plipres.2011.07.002.
Henry, K. W., Nickels, J. T., & Edlind, T. D. (2002). ROX1 and ERG regulation in
Callari, R., Meier, Y., Ravasio, D., & Heider, H. (2018). Dynamic control of ERG20 and
Saccharomyces cerevisiae: Implications for antifungal susceptibility. Eukaryotic Cell, 1
ERG9 expression for improved casbene production in Saccharomyces cerevisiae.
(6), 1041–1044. https://doi.org/10.1128/EC.1.6.1041-1044.2002.
Frontiers in Bioengineering and Biotechnology, 6, 160. https://doi.org/10.3389/
fbioe.2018.00160.

15
Z. Liang et al.
Hezari, M., & Croteau, R. (1997). Taxol biosynthesis: An update. Planta Medica, 63(04),
291–295. https://doi.org/10.1055/s-2006-957684.
Himejima, M., Hobson, K. R., Otsuka, T., Wood, D. L., & Kubo, I. (1992). Antimicrobial
terpenes from oleoresin of ponderosa pine tree Pinus ponderosa: A defense
mechanism against microbial invasion. Journal of Chemical Ecology, 18(10),
1809–1818. https://doi.org/10.1007/BF02751105.
Hood, E. E., Gelvin, S. B., Melchers, L. S., & Hoekema, A. (1993). New Agrobacterium
helper plasmids for gene transfer to plants. Transgenic Research, 2(4), 208–218.
https://doi.org/10.1007/BF01977351.
Hu, F., Liu, J., Du, G., Hua, Z., Zhou, J., & Chen, J. (2012). Key cytomembrane ABC
transporters of Saccharomyces cerevisiae fail to improve the tolerance to d-limonene.
Biotechnology Letters, 34(8), 1505–1509. https://doi.org/10.1007/s10529-012-0931-
6.
Hu, Z., Lin, L., Li, H., Li, P., Weng, Y., Zhang, C., … Xiao, D. (2020). Engineering
Saccharomyces cerevisiae for production of the valuable monoterpene d-limonene
during Chinese Baijiu fermentation. Journal of Industrial Microbiology and
Biotechnology, 47(6–7), 511–523. https://doi.org/10.1007/s10295-020-02284-6.
Ichinose, H., & Kitaoka, T. (2018). Insight into metabolic diversity of the brown-rot
basidiomycete Postia placenta responsible for sesquiterpene biosynthesis: Semi-
comprehensive screening of cytochrome P450 monooxygenase involved in
protoilludene metabolism. Microbial Biotechnology, 11(5), 952–965. https://doi.org/
10.1111/1751-7915.13304.
Ignea, C., Cvetkovic, I., Loupassaki, S., Kefalas, P., Johnson, C. B., Kampranis, S. C., &
Makris, A. M. (2011). Improving yeast strains using recyclable integration cassettes,
for the production of plant terpenoids. Microbial Cell Factories, 10(1), 4. https://doi.
org/10.1186/1475-2859-10-4.
Ignea, C., Pontini, M., Maffei, M. E., Makris, A. M., & Kampranis, S. C. (2014).
Engineering monoterpene production in yeast using a synthetic dominant negative
geranyl diphosphate synthase. ACS Synthetic Biology, 3(5), 298–306. https://doi.org/
10.1021/sb400115e.
Ignea, C., Raadam, M. H., Motawia, M. S., Makris, A. M., Vickers, C. E., &
Kampranis, S. C. (2019). Orthogonal monoterpenoid biosynthesis in yeast
constructed on an isomeric substrate. Nature Communications, 10(1), 3799. https://
doi.org/10.1038/s41467-019-11290-x.
Ignea, C., Trikka, F. A., Kourtzelis, I., Argiriou, A., Kanellis, A. K., Kampranis, S. C., &
Makris, A. M. (2012). Positive genetic interactors of HMG2 identify a new set of
genetic perturbations for improving sesquiterpene production in Saccharomyces
cerevisiae. Microbial Cell Factories, 11(1), 162. https://doi.org/10.1186/1475-2859-
11-162.
Ignea, C., Trikka, F. A., Nikolaidis, A. K., Georgantea, P., Ioannou, E., Loupassaki, S., …
Kampranis, S. C. (2015). Efficient diterpene production in yeast by engineering
Erg20p into a geranylgeranyl diphosphate synthase. Metabolic Engineering, 27,
65–75. https://doi.org/10.1016/j.ymben.2014.10.008.
Ingy, I. A., & Wim, J. Q. (2017). A glimpse into the biosynthesis of terpenoids. KnE Life
Sciences, 3(5), 81–98. https://doi.org/10.18502/kls.v3i5.981.
Itoh, T., Tokunaga, K., Matsuda, Y., Fujii, I., Abe, I., Ebizuka, Y., & Kushiro, T. (2010).
Reconstitution of a fungal meroterpenoid biosynthesis reveals the involvement of a
novel family of terpene cyclases. Nature Chemistry, 2(10), 858–864. https://doi.org/
10.1038/nchem.764.
Jackson, B. E. (2005). Yeast as a host for sesquiterpene production ((Ph.D)). Houston, Texas,
USA: Rice University.
Jakočiūnas, T., Bonde, I., Herrgård, M., Harrison, S. J., Kristensen, M., Pedersen, L. E., …
Keasling, J. D. (2015). Multiplex metabolic pathway engineering using CRISPR/Cas9
in Saccharomyces cerevisiae. Metabolic Engineering, 28, 213–222. https://doi.org/
10.1016/j.ymben.2015.01.008.
Jeong, S.-H., Park, J.-B., Wang, Y., Kim, G.-H., Zhang, G., Wei, G., … Kim, S.-W. (2020).
Regulatory molecule cAMP changes cell fitness of the engineered Escherichia coli for
terpenoids production. Metabolic Engineering, 65, 178–184. https://doi.org/10.1016/
j.ymben.2020.11.009.
Jia, D., Xu, S., Sun, J., Zhang, C., Li, D., & Lu, W. (2019). Yarrowia lipolytica construction
for heterologous synthesis of α-santalene and fermentation optimization. Applied
Microbiology and Biotechnology, 103(8), 3511–3520. https://doi.org/10.1007/
s00253-019-09735-w.
Jones, K. L., Kim, S.-W., & Keasling, J. (2000). Low-copy plasmids can perform as well as
or better than high-copy plasmids for metabolic engineering of bacteria. Metabolic
Engineering, 2(4), 328–338. https://doi.org/10.1006/mben.2000.0161.
Jongedijk, E., Cankar, K., Ranzijn, J., van der Krol, S., Bouwmeester, H., & Beekwilder, J.
(2015). Capturing of the monoterpene olefin limonene produced in Saccharomyces
cerevisiae. Yeast, 32(1), 159–171. https://doi.org/10.1002/yea.3038.
Katayama, T., Nakamura, H., Zhang, Y., Pascal, A., Fujii, W., & Maruyama, J.-I. (2019).
Forced recycling of an AMA1-based genome-editing plasmid allows for efficient
multiple gene deletion/integration in the industrial filamentous fungus Aspergillus
oryzae. Applied and Environmental Microbiology, 85(3), e01896–01818. https://doi.
org/10.1128/AEM.01896-18.
Katayama, T., Tanaka, Y., Okabe, T., Nakamura, H., Fujii, W., Kitamoto, K., &
Maruyama, J.-I. (2016). Development of a genome editing technique using the
CRISPR/Cas9 system in the industrial filamentous fungus Aspergillus oryzae.
Biotechnology Letters, 38(4), 637–642. https://doi.org/10.1007/s10529-015-2015-x.
Kiattisewee, C., Dong, C., Fontana, J., Sugianto, W., Peralta-Yahya, P., Carothers, J. M., &
Zalatan, J. G. (2021). Portable bacterial CRISPR transcriptional activation enables
metabolic engineering in Pseudomonas putida. Metabolic Engineering. https://doi.org/
10.1016/j.ymben.2021.04.002.
Kogure, T., & Inui, M. (2018). Recent advances in metabolic engineering of
Corynebacterium glutamicum for bioproduction of value-added aromatic chemicals
and natural products. Applied Microbiology and Biotechnology, 102(20), 8685–8705.
https://doi.org/10.1007/s00253-018-9289-6.
16
Food Research International 147 (2021) 110487
Kong, M. K., Kang, H.-J., Kim, J. H., Oh, S. H., & Lee, P. C. (2015). Metabolic engineering
of the Stevia rebaudiana ent-kaurene biosynthetic pathway in recombinant Escherichia
coli. Journal of Biotechnology, 214, 95–102. https://doi.org/10.1016/j.
jbiotec.2015.09.016.
Kong, S., Fu, X., Li, X., Pan, H., & Guo, D. (2020). De novo biosynthesis of linalool from
glucose in engineered Escherichia coli. Enzyme and Microbial Technology, 140,
109614. https://doi.org/10.1016/j.enzmictec.2020.109614.
Kuranda, K., François, J., & Palamarczyk, G. (2009). The isoprenoid pathway and
transcriptional response to its inhibitors in the yeast Saccharomyces cerevisiae. FEMS
Yeast Research, 10(1), 14–27. https://doi.org/10.1111/j.1567-1364.2009.00560.x.
Kutyna, D. R., & Borneman, A. (2018). Heterologous production of flavour and aroma
compounds in Saccharomyces cerevisiae. Genes, 9(7), 326. https://doi.org/10.3390/
genes9070326.
Lauersen, K. J., Wichmann, J., Baier, T., Kampranis, S. C., Pateraki, I., Møller, B. L., &
Kruse, O. (2018). Phototrophic production of heterologous diterpenoids and a
hydroxy-functionalized derivative from Chlamydomonas reinhardtii. Metabolic
Engineering, 49, 116–127. https://doi.org/10.1016/j.ymben.2018.07.005.
Leber, R., Zenz, R., Schröttner, K., Fuchsbichler, S., Pühringer, B., & Turnowsky, F.
(2001). A novel sequence element is involved in the transcriptional regulation of
expression of the ERG1 (squalene epoxidase) gene in Saccharomyces cerevisiae.
European Journal of Biochemistry, 268(4), 914–924. https://doi.org/10.1046/j.1432-
1327.2001.01940.x.
Li, Z., Howell, K., Fang, Z., & Zhang, P. (2020). Sesquiterpenes in grapes and wines:
Occurrence, biosynthesis, functionality, and influence of winemaking processes.
Comprehensive Reviews in Food Science and Food Safety, 19(1), 247–281. https://doi.
org/10.1111/1541-4337.12516.
Liang, Z., Fang, Z., Pai, A., Luo, J., Gan, R., Gao, Y., … Zhang, P. (2020). Glycosidically
bound aroma precursors in fruits: A comprehensive review. Critical Reviews in Food
Science and Nutrition, 1–29. https://doi.org/10.1080/10408398.2020.1813684.
Liang, Z., Zhang, P., & Fang, Z. (2020). Modern technologies for extraction of aroma
compounds from fruit peels: A review. Critical Reviews in Food Science and Nutrition,
1–24. https://doi.org/10.1080/10408398.2020.1840333.
Lin, P.-C., Zhang, F., & Pakrasi, H. B. (2021). Enhanced limonene production in a fast-
growing cyanobacterium through combinatorial metabolic engineering. Metabolic
Engineering Communications, 12, e00164. https://doi.org/10.1016/j.mec.2021.
e00164.
Liscum, L., Finer-Moore, J., Stroud, R., Luskey, K., Brown, M. S., & Goldstein, J. L.
(1985). Domain structure of 3-hydroxy-3-methylglutaryl coenzyme A reductase, a
glycoprotein of the endoplasmic reticulum. Journal of Biological Chemistry, 260(1),
522–530.
Liu, C., Minami, A., Ozaki, T., Wu, J., Kawagishi, H., Maruyama, J.-I., & Oikawa, H.
(2019). Efficient reconstitution of Basidiomycota diterpene erinacine gene cluster in
Ascomycota host Aspergillus oryzae based on genomic DNA sequences. Journal of the
American Chemical Society, 141(39), 15519–15523. https://doi.org/10.1021/
jacs.9b08935.
Liu, H., Sun, Y., Ramos, K. R. M., Nisola, G. M., Valdehuesa, K. N. G., Lee, W. K., …
Chung, W.-J. (2013). Combination of Entner-Doudoroff pathway with MEP increases
isoprene production in engineered Escherichia coli. Plos One, 8(12), Article e83290.
https://doi.org/10.1371/journal.pone.0083290.
Liu, J., Zhang, W., Du, G., Chen, J., & Zhou, J. (2013). Overproduction of geraniol by
enhanced precursor supply in Saccharomyces cerevisiae. Journal of Biotechnology, 168
(4), 446–451. https://doi.org/10.1016/j.jbiotec.2013.10.017.
Loeschcke, A., & Thies, S. (2015). Pseudomonas putida—a versatile host for the
production of natural products. Applied Microbiology and Biotechnology, 99(15),
6197–6214. https://doi.org/10.1007/s00253-015-6745-4.
López, J., Essus, K., Kim, I.-K., Pereira, R., Herzog, J., Siewers, V., … Agosin, E. (2015).
Production of β-ionone by combined expression of carotenogenic and plant CCD1
genes in Saccharomyces cerevisiae. Microbial Cell Factories, 14(1), 84. https://doi.org/
10.1186/s12934-015-0273-x.
Lu, Z., Peng, B., Ebert, B. E., Dumsday, G., & Vickers, C. E. (2021). Auxin-mediated
protein depletion for metabolic engineering in terpene-producing yeast. Nature
Communications, 12(1), 1051. https://doi.org/10.1038/s41467-021-21313-1.
Lyu, X., Lee, J., & Chen, W. N. (2019). Potential natural food preservatives and their
sustainable production in yeast: Terpenoids and polyphenols. Journal of Agricultural
and Food Chemistry, 67(16), 4397–4417. https://doi.org/10.1021/acs.jafc.8b07141.
Ma, Y., Li, W., Mai, J., Wang, J., Wei, Y., Ledesma-Amaro, R., & Ji, X.-J. (2021).
Engineering Yarrowia lipolytica for sustainable production of the chamomile
sesquiterpene (-)-α-bisabolol. Green Chemistry, 23(2), 780–787. https://doi.org/
10.1039/D0GC03180A.
Matsuda, Y., Awakawa, T., & Abe, I. (2013). Reconstituted biosynthesis of fungal
meroterpenoid andrastin A. Tetrahedron, 69(38), 8199–8204. https://doi.org/
10.1016/j.tet.2013.07.029.
Meadows, A. L., Hawkins, K. M., Tsegaye, Y., Antipov, E., Kim, Y., Raetz, L., …
Tsong, A. E. (2016). Rewriting yeast central carbon metabolism for industrial
isoprenoid production. Nature, 537(7622), 694–697. https://doi.org/10.1038/
nature19769.
Meng, H., Wang, Y., Hua, Q., Zhang, S., & Wang, X. (2011). In silico analysis and
experimental improvement of taxadiene heterologous biosynthesis in Escherichia coli.
Biotechnology Bioprocess Engineering, 16(2), 205–215. https://doi.org/10.1007/
s12257-010-0329-z.
Mischko, W., Hirte, M., Fuchs, M., Mehlmer, N., & Brück, T. B. (2018). Identification of
sesquiterpene synthases from the Basidiomycota Coniophora puteana for the efficient
and highly selective β-copaene and cubebol production in E. coli. Microbial Cell
Factories, 17(1), 1–13. https://doi.org/10.1186/s12934-018-1010-z.
Montañés, F. M., Pascual-Ahuir, A., & Proft, M. (2011). Repression of ergosterol
biosynthesis is essential for stress resistance and is mediated by the Hog1 MAP
Z. Liang et al.
kinase and the Mot3 and Rox1 transcription factors. Molecular Microbiology, 79(4),
1008–1023. https://doi.org/10.1111/j.1365-2958.2010.07502.x.
Moser, S., & Pichler, H. (2019). Identifying and engineering the ideal microbial terpenoid
production host. Applied Microbiology Biotechnology, 103(14), 5501–5516. https://
doi.org/10.1007/s00253-019-09892-y.
Nagamine, S., Liu, C., Nishishita, J., Kozaki, T., Sogahata, K., Sato, Y., … Maruyama, J.-I.
(2019). Ascomycete Aspergillus oryzae is an efficient expression host for production
of basidiomycete terpenes by using genomic DNA sequences. Applied Environmental
Microbiology, 85(15), e00409–00419. https://doi.org/10.1128/AEM.00409-19.
Özaydın, B., Burd, H., Lee, T. S., & Keasling, J. D. (2013). Carotenoid-based phenotypic
screen of the yeast deletion collection reveals new genes with roles in isoprenoid
production. Metabolic Engineering, 15, 174–183. https://doi.org/10.1016/j.
ymben.2012.07.010.
Paramasivan, K., & Mutturi, S. (2017). Progress in terpene synthesis strategies through
engineering of Saccharomyces cerevisiae. Critical Reviews in Biotechnology, 37(8),
974–989. https://doi.org/10.1080/07388551.2017.1299679.
Pardo, E., Rico, J., Gil, J. V., & Orejas, M. (2015). De novo production of six key grape
aroma monoterpenes by a geraniol synthase-engineered S. cerevisiae wine strain.
Microbial Cell Factories, 14(1), 136. https://doi.org/10.1186/s12934-015-0306-5.
Peralta-Yahya, P. P., Ouellet, M., Chan, R., Mukhopadhyay, A., Keasling, J. D., &
Lee, T. S. (2011). Identification and microbial production of a terpene-based
advanced biofuel. Nature Communications, 2(1), 1–8. https://doi.org/10.1038/
ncomms1494.
Pitera, D. J., Paddon, C. J., Newman, J. D., & Keasling, J. D. (2007). Balancing a
heterologous mevalonate pathway for improved isoprenoid production in Escherichia
coli. Metabolic Engineering, 9(2), 193–207. https://doi.org/10.1016/j.
ymben.2006.11.002.
Prakash, I., Markosyan, A., & Bunders, C. (2014). Development of next generation stevia
sweetener: Rebaudioside M. Foods, 3(1), 162–175. https://doi.org/10.3390/
foods3010162.
Pramastya, H., Song, Y., Elfahmi, E. Y., Sukrasno, S., & Quax, W. J. (2020). Positioning
Bacillus subtilis as terpenoid cell factory. Journal of Applied Microbiology. https://doi.
org/10.1111/jam.14904.
Quin, M. B., Flynn, C. M., & Schmidt-Dannert, C. (2014). Traversing the fungal
terpenome. Natural Product Reports, 31(10), 1449–1473. https://doi.org/10.1039/
C4NP00075G.
Quin, M. B., Flynn, C. M., Wawrzyn, G. T., Choudhary, S., & Schmidt- Dannert, C. (2013).
Mushroom hunting by using bioinformatics: Application of a predictive framework
facilitates the selective identification of sesquiterpene synthases in Basidiomycota.
ChemBioChem, 14(18), 2480–2491. https://doi.org/10.1002/cbic.201300349.
Rodrigues, J. S., & Lindberg, P. (2021). Metabolic engineering of Synechocystis sp. PCC
6803 for improved bisabolene production. Metabolic Engineering. Communications,
12, Article e00159. https://doi.org/10.1016/j.mec.2020.e00159.
Schempp, F. M., Drummond, L., Buchhaupt, M., & Schrader, J. (2018). Microbial cell
factories for the production of terpenoid flavor and fragrance compounds. Journal of
Agricultural and Food Chemistry, 66(10), 2247–2258. https://doi.org/10.1021/acs.
jafc.7b00473.
Schilmiller, A. L., Schauvinhold, I., Larson, M., Xu, R., Charbonneau, A. L., Schmidt, A.,
… Pichersky, E. (2009). Monoterpenes in the glandular trichomes of tomato are
synthesized from a neryl diphosphate precursor rather than geranyl diphosphate.
Proceedings of the National Academy of Sciences, 106(26), 10865–10870. https://doi.
org/10.1073/pnas.0904113106.
Shaw, J. J., Berbasova, T., Sasaki, T., Jefferson-George, K., Spakowicz, D. J.,
Dunican, B. F., … Strobel, S. A. (2015). Identification of a fungal 1, 8-cineole
synthase from Hypoxylon sp. with specificity determinants in common with the plant
synthases. Journal of Biological Chemistry, 290(13), 8511–8526. https://doi.org/
10.1074/jbc.M114.636159.
Shi, T.-Q., Gao, J., Wang, W.-J., Wang, K.-F., Xu, G.-Q., Huang, H., & Ji, X.-J. (2019).
CRISPR/Cas9-based genome editing in the filamentous fungus Fusarium fujikuroi and
its application in strain engineering for gibberellic acid production. ACS Synthetic
Biology, 8(2), 445–454. https://doi.org/10.1021/acssynbio.8b00478.
Soliman, S. S. M., Mosa, K. A., El-Keblawy, A. A., & Husseiny, M. I. (2017). Exogenous
and endogenous increase in fungal GGPP increased fungal Taxol production. Applied
Microbiology and Biotechnology, 101(20), 7523–7533. https://doi.org/10.1007/
s00253-017-8509-9.
Song, Y., Guan, Z., van Merkerk, R., Pramastya, H., Abdallah, I. I., Setroikromo, R., &
Quax, W. J. (2020). Production of squalene in Bacillus subtilis by squalene synthase
screening and metabolic engineering. Journal of Agricultural and Food Chemistry, 68
(15), 4447–4455. https://doi.org/10.1021/acs.jafc.0c00375.
Song, Y., He, S., Abdallah, I. I., Jopkiewicz, A., Setroikromo, R., van Merkerk, R., …
Quax, W. J. (2021). Engineering of multiple modules to improve amorphadiene
production in Bacillus subtilis using CRISPR-Cas9. Journal of Agricultural and Food
Chemistry, 69(16), 4785–4794. https://doi.org/10.1021/acs.jafc.1c00498.
Sun, W., Zhao, Y.-J., & Li, C. (2019). De novo synthesis of plant natural products in yeast.
In T. P. Basso (Ed.), Yeasts in Biotechnology (pp. 105–122). IntechOpen.
Tagami, K., Minami, A., Fujii, R., Liu, C., Tanaka, M., Gomi, K., … Oikawa, H. (2014).
Rapid reconstitution of biosynthetic machinery for fungal metabolites in Aspergillus
oryzae: Total biosynthesis of aflatrem. ChemBioChem, 15(14), 2076–2080. https://
doi.org/10.1002/cbic.201402195.
Tholl, D. (2006). Terpene synthases and the regulation, diversity and biological roles of
terpene metabolism. Current Opinion in Plant Biology, 9(3), 297–304. https://doi.org/
10.1016/j.pbi.2006.03.014.
Tippmann, S., Ferreira, R., Siewers, V., Nielsen, J., & Chen, Y. (2017). Effects of
acetoacetyl-CoA synthase expression on production of farnesene in Saccharomyces
cerevisiae. Journal of Industrial Microbiology & Biotechnology, 44(6), 911–922. https://
doi.org/10.1007/s10295-017-1911-6.
17
Food Research International 147 (2021) 110487
Trikka, F. A., Nikolaidis, A., Athanasakoglou, A., Andreadelli, A., Ignea, C., Kotta, K., …
Makris, A. M. (2015). Iterative carotenogenic screens identify combinations of yeast
gene deletions that enhance sclareol production. Microbial Cell Factories, 14(1), 60.
https://doi.org/10.1186/s12934-015-0246-0.
Troost, K., Loeschcke, A., Hilgers, F., Özgür, A. Y., Weber, T. M., Santiago-Schübel, B., …
Drepper, T. (2019). Engineered Rhodobacter capsulatus as a phototrophic platform
organism for the synthesis of plant sesquiterpenoids. Frontiers in Microbiology, 10
(1998). https://doi.org/10.3389/fmicb.2019.01998.
Uribe, S., Ramirez, J., & Peña, A. (1985). Effects of beta-pinene on yeast membrane
functions. Journal of Bacteriology, 161(3), 1195. https://doi.org/10.1128/
JB.161.3.1195-1200.1985.
Vedula, L. S., Jiang, J., Zakharian, T., Cane, D. E., & Christianson, D. W. (2008).
Structural and mechanistic analysis of trichodiene synthase using site-directed
mutagenesis: Probing the catalytic function of tyrosine-295 and the asparagine-225/
serine-229/glutamate-233–Mg2+B motif. Archives of Biochemistry Biophysics, 469(2),
184–194. https://doi.org/10.1016/j.abb.2007.10.015.
Vieira, A. J., Beserra, F. P., Souza, M. C., Totti, B. M., & Rozza, A. L. (2018). Limonene:
Aroma of innovation in health and disease. Chemico-Biological Interactions, 283,
97–106. https://doi.org/10.1016/j.cbi.2018.02.007.
Wang, C.-Y., Chen, Y.-W., & Hou, C.-Y. (2019). Antioxidant and antibacterial activity of
seven predominant terpenoids. International Journal of Food Properties, 22(1),
230–238. https://doi.org/10.1080/10942912.2019.1582541.
Wang, C., Park, J. E., Choi, E. S., & Kim, S. W. (2016). Farnesol production in Escherichia
coli through the construction of a farnesol biosynthesis pathway–application of PgpB
and YbjG phosphatases. Biotechnology Journal, 11(10), 1291–1297. https://doi.org/
10.1002/biot.201600250.
Wang, K., & Ohnuma, S.-I. (1999). Chain-length determination mechanism of isoprenyl
diphosphate synthases and implications for molecular evolution. Trends in
Biochemical Sciences, 24(11), 445–451. https://doi.org/10.1016/S0968-0004(99)
01464-4.
Wang, Q., Quan, S., & Xiao, H. (2019). Towards efficient terpenoid biosynthesis:
Manipulating IPP and DMAPP supply. Bioresources and Bioprocessing, 6(1), 6. https://
doi.org/10.1186/s40643-019-0242-z.
Wang, Y., Lim, L., DiGuistini, S., Robertson, G., Bohlmann, J., & Breuil, C. (2013).
A specialized ABC efflux transporter GcABC-G1 confers monoterpene resistance to
Grosmannia clavigera, a bark beetle-associated fungal pathogen of pine trees. New
Phytologist, 197(3), 886–898. https://doi.org/10.1111/nph.12063.
Wawrzyn, G. T., Quin, M. B., Choudhary, S., López-Gallego, F., & Schmidt-Dannert, C.
(2012). Draft genome of Omphalotus olearius provides a predictive framework for
sesquiterpenoid natural product biosynthesis in Basidiomycota. Chemistry & Biology,
19(6), 772–783. https://doi.org/10.1016/j.chembiol.2012.05.012.
Wichmann, J., Baier, T., Wentnagel, E., Lauersen, K. J., & Kruse, O. (2018). Tailored
carbon partitioning for phototrophic production of (E)-α-bisabolene from the green
microalga Chlamydomonas reinhardtii. Metabolic Engineering, 45, 211–222. https://
doi.org/10.1016/j.ymben.2017.12.010.
Wu, W., Tran, W., Taatjes, C. A., Alonso-Gutierrez, J., Lee, T. S., & Gladden, J. M. (2016).
Rapid discovery and functional characterization of terpene synthases from four
endophytic xylariaceae. Plos One, 11(2), e0146983. https://doi.org/10.1371/
journal.pone.0146983.
Yang, X., Nambou, K., Wei, L., & Hua, Q. (2016). Heterologous production of α-farnesene
in metabolically engineered strains of Yarrowia lipolytica. Bioresource Technology,
216, 1040–1048. https://doi.org/10.1016/j.biortech.2016.06.028.
Yılmaz, F. M., Görgüç, A., Uygun, Ö., & Bircan, C. (2020). Steviol glycosides and
polyphenols extraction from Stevia rebaudiana Bertoni leaves using maceration,
microwave-, and ultrasound-assisted techniques. Separation Science and Technology,
1–13. https://doi.org/10.1080/01496395.2020.1743311.
Yoshikuni, Y., Ferrin, T. E., & Keasling, J. D. (2006). Designed divergent evolution of
enzyme function. Nature, 440(7087), 1078–1082. https://doi.org/10.1038/
nature04607.
Zhang, C., Chen, X., Orban, A., Shukal, S., Birk, F., Too, H.-P., & Rühl, M. (2020).
Agrocybe aegerita serves as a gateway for identifying sesquiterpene biosynthetic
enzymes in higher fungi. ACS Chemical Biology, 15(5), 1268–1277. https://doi.org/
10.1021/acschembio.0c00155.
Zhang, D. L., Jennings, S. M., Robinson, G. W., & Poulter, C. D. (1993). Yeast squalene
synthase: Expression, purification, and characterization of soluble recombinant
enzyme. Archives of Biochemistry and Biophysics, 304(1), 133–143. https://doi.org/
10.1006/abbi.1993.1331.
Zhang, G., Cao, Q., Liu, J., Liu, B., Li, J., & Li, C. (2015). Refactoring β-amyrin synthesis
in Saccharomyces cerevisiae. AIChE Journal, 61(10), 3172–3179. https://doi.org/
10.1002/aic.14950.
Zhang, P., Fuentes, S., Siebert, T., Krstic, M., Herderich, M., Barlow, E. W. R., &
Howell, K. (2016). Terpene evolution during the development of Vitis vinifera L. cv.
Shiraz grapes. Food Chemistry, 204, 463–474. https://doi.org/10.1016/j.
foodchem.2016.02.125.
Zhang, Y., Nielsen, J., & Liu, Z. (2017). Engineering yeast metabolism for production of
terpenoids for use as perfume ingredients, pharmaceuticals and biofuels. FEMS Yeast
Research, 17(8), fox080. https://doi.org/10.1093/femsyr/fox080.
Zhao, J., Bao, X., Li, C., Shen, Y., & Hou, J. (2016). Improving monoterpene geraniol
production through geranyl diphosphate synthesis regulation in Saccharomyces
cerevisiae. Applied Microbiology Biotechnology, 100(10), 4561–4571. https://doi.org/
10.1007/s00253-016-7375-1.
Zhao, J., Li, C., Zhang, Y., Shen, Y., Hou, J., & Bao, X. (2017). Dynamic control of ERG20
expression combined with minimized endogenous downstream metabolism
contributes to the improvement of geraniol production in Saccharomyces cerevisiae.
Microbial Cell Factories, 16(1), 17. https://doi.org/10.1186/s12934-017-0641-9.
Z. Liang et al.
Zhao, Y.-J., & Li, C. (2018). Biosynthesis of plant triterpenoid saponins in microbial cell
factories. Journal of Agricultural and Food Chemistry, 66(46), 12155–12165. https://
doi.org/10.1021/acs.jafc.8b04657.
Zhou, F., & Pichersky, E. (2020). More is better: The diversity of terpene metabolism in
plants. Current Opinion in Plant Biology, 55, 1–10. https://doi.org/10.1016/j.
pbi.2020.01.005.
Zhou, P., Xie, W., Li, A., Wang, F., Yao, Z., Bian, Q., … Ye, L. (2017). Alleviation of
metabolic bottleneck by combinatorial engineering enhanced astaxanthin synthesis
18
Food Research International 147 (2021) 110487
in Saccharomyces cerevisiae. Enzyme and Microbial Technology, 100, 28–36. https://
doi.org/10.1016/j.enzmictec.2017.02.006.
Zhuang, X. (2013). Engineering novel terpene production platforms in the yeast
Saccharomyces cerevisiae ((Ph.D)). Lexington, Kentucky, USA: University of
Kentucky.
Zong, Z., Hua, Q., Tong, X., Li, D., Wang, C., Guo, D., & Liu, Z. (2019). Biosynthesis of
nerol from glucose in the metabolic engineered Escherichia coli. Bioresource
Technology, 287, 121410. https://doi.org/10.1016/j.biortech.2019.121410.