1016

Journal of Food Protection, Vo/. 65, No. 6, 2002, Pages 1016-1023

Copyright ©, International Association for Food Protection

Molecular Identification of Nine Commercial Flatfish Species by

Polymerase Chain Reaction-Restriction Fragment Length
Polymorphism Analysis of a Segment of the
Cytochrome b Region
ANDRES SANJUAN* AND ANGEL S. COMESANA

Xentftica £volutiva Molecular, Facultade de Ciencias-Bioloxfa, Universidade de Vigo, £-36200 Vigo, Spain

MS 01-179: Received 18 May 2ool/Accepted 8 October 2001

ABSTRACT
Commercial refrigerated or frozen flatfish fillets are sometimes mislabeled, and identification of these mislabeled products
is necessary to prevent fraudulent substitution. Identification of nine commercial flatfish species (order Pleuronectiformes),
Hippoglossus hippoglossus (halibut), Lepidorhombus boscii (four-spotted scaldfish), Lepidorhombus whiffiagonis (megrin),
Platichthys flesus (flounder), Pleuronectes platessa (European plaice), Reinhardtius hippoglossoides (Greenland halibut), Sco-
. phthalmus maximus (turbot), Scophthalmus rhombus (brill), and Solea vulgaris (=Solea solea) (sole), was carried out on the
basis of the amplification of a 486-bp segment of the mitochondrial genome (tRNAGlu/cytochrome b) by using the polymerase
chain reaction (peR) and universal primers. Sequences of PCR-amplified DNA from the flatfish species were used to select
eight restriction enzymes (REs). The PCR products were cut with each RE, resulting in species-specific restriction fragment
length polymorphism. Seven species groups could be identified by application of the single RE DdeI and six species groups
by using HaeIII, Hinfl, MaeI, or Mbo!. Different combinations of only a couple of these REs could unambiguously identify
the nine flatfish species. Genetic polymorphisms of the target sequence were examined by comparison with previously pub-
lished DNA sequences, and the results of this comparison confirmed the usefulness of this technique in distinguishing and
genetically characterizing refrigerated or frozen pieces of these nine flatfish species.

The market for processed fishery products is steadily
growing, and the identification of commercial fish species
becomes a problem when the distinguishing features of fish
are removed at the processing stage. Commercial refrigerated
or frozen flatfish fillets are sometimes mislabeled. Several
products purporting to contain expensive and popular flatfish
species are highly susceptible to fraudulent or unintentional
substitution of cheaper fish species. Identification of these
products is necessary to prevent fraudulent substitution.
Numerous analytical methods for fish species authenti-
cation have been developed. Most of these methods rely on
the analysis of proteins and involve mainly electrophoretic
techniques, high-performance liquid chromatography, and
immunoassays (1, 15, 18, 19, 29 and references therein).
Advances in molecular biology techniques have allowed the
development of DNA-based methods (13,29). DNA carries
an organism's genetic information, which is the same in all
cell types, and it is a very stable and long-lived biological
molecule. Most of the genetic approaches to the determina-
tion of species identity are based on the amplification of a
region of mitochondrial or nuclear DNA by the polymerase
chain reaction (PCR) with conserved primers. Applications
of the peR to fish species identification have proliferated
because of the simplicity, specificity, and sensitivity of this
technique (24). One of the PCR-based methods used for fish
* Author forcorrespondence. Tel: 34 986 812569; Fax: 34 986 812556;
E-mail: asanjuan@uvigo.es.
species identification has been direct sequence analysis of
the amplified fragments (2, 3, 13, 18, 25, 26 and references
therein). Restriction fragment length polymorphism (RFLP)
of PCR products constitutes an alternative for the identifi-
cation of species that is simpler than sequencing (6-13, 18,
20, 25-27, 32 and references therein).
The mitochondrial encoded cytochrome b (cyt b) gene
is a useful genetic marker for the identification of the spe-
cies origin of a fish sample (for details, see Bartlett and
Davidson (3) and Davidson (13)). The cyt b gene is intra-
specifically conserved, but it contains enough interspecific
heterogeneity to produce species-specific restriction frag-
ment patterns (5, 21). Cyt b is among the most extensively
sequenced genes in vertebrates (16). Cyt b sequences can
be found in international data banks and can be used to
compare sequences and to obtain restriction patterns.
Different cyt b gene sequences have been studied to
identify several flatfish species (order Pleuronectiformes). A
359-bp amplified fragment was digested with three restric-
tion endonucleases (REs) to distinguish Platichthys fiesus,
Pleuronectes platessa, and Solea vulgaris (=Solea solea)
(11). This amplicon was not detected in Reinhardtius hip-
poglossoides, but a 20l-bp fragment was subsequently am-
plified and digested with four REs to identify these four flat-
fish species (12). Moreover, other flatfish species, including
Scophthalmus rhombus, have also been characterized by
PCR-RFLP of the 5' end of the cyt b fragment (9, 32).
The objective of this work was to develop a simple
J. Food Prol" Vol. 65, No.6 PCR-RFLPs OF FLATASH SPECr£S 1017

TABLE l. Sludied commercial f/a1fish species (order Pleuronec-
l.ifonnes) from Ihe North Allanlie. along lI'ilh Ihe Jlllmbers of se-
qllt'nced (wd RE-wllIlvzed il1di~'idlJals
Species
Family Scoplllhalmidae
LepidorllOmbw boseii
(four-spolled scaldfish)
L. whiffiagollis (megrjn)
Seopll/ha!mus nlLLrimllS (turbol)
S. rhombus (brill)
Family PleurOl1eelidae
Hippoglossus hippog!osslIs (halibul)
PialiehtIJ.I's jieJIl5 (flounder)
Plel.frOllectes plalessa (European plaice)
ReinhardtillS hippoglossoides
(Greenland halibul)
Family Soleidae
Solea vulgaris (=5. solea) (sole)
TOla)
peR. A 486-bp ponion of lhe milochondrial genome in Ihe
IRNAGlu/CYI b region was amplified by PCR using the universal
primers L14724 (5'-CGA AGC TTG ATA TGA AAA ACC ATC
GTT G-3' (23)) and H15l49 (S'-AAA erG CAG CCC CTC
Sequenced RE-anulyzed
AGA ATG ATA TTT GTC CTC A-3' (17). which work well for
indi viduals individuals
many fish taxa (22). TIle primers were synlhesized by Amersham
Phannacia Biotech (Buckinghamshire. UK). Double-slranded am-
pli /ications were carried OUI wil h a final volume of 50 to 100 IlJ
2 15
containing 9 mM Tris HCI (pH 9.0). 2 mM MgCI 2 , 45 mM KCI,
3 15
0.09% Triton X-lOa, 0.2 mM dATE 0.2 mM dCTP. 0.2 mM dGTP.
0.2 mM dTTP (Amersham Phannacia Biolech). 0.06 ,...,M each
3 15
primer; 1.5 U of Taq DNA polymerase (PTomega, Southampton.
2 15
UK). and 3 IJ-I of the above extracted genomic DNA as a templale.
A negalive conlroi reaclion (withoul DNA) was also included for
J 10 each PCR set 10 test for possible conlamination. The cycling con-
2 15 ditions on a Master Gradient lhennal cycler (Eppendorf. Hamburg,
2 \5 Gennany) consisted of denaturation 8t 95°C for 5 min. amplifi-
3 15 cation with 30 cycles of 94°C for 60 s. 50°C for 60 s. and noc
for 120 s. and a final extension at noc for 10 min.
PCR products were examined by eleclrophoresis through a
3.0% SeaKem LE and Nusieve GTG agarose (2: J) gel (FMC
15
BioProduClS. Rockland. Maine) In TBE buffer (89 mM Tris base.
22 130 89 mM boric acid, 2.5 roM ethylenediaminetelra-acetic acid [pH
7.5]) and Slained by ethidium blOmjd~. As a size reference. a 100-
or 50-bp ladder (Amersham Pharma\:ia BiOlech) was used.

method for the identjfication of refrigerated and frozen fil-
lets from nine commercial flatfish species by PCR-RFLP
analysis of a 486-bp fragment in the mitochondrial
tRNA Glu/cyl b region. This method can be applied to the
detection of fraudulent or unintentional mislabeling of pro-
cessed products containing these species.
MATERIALS AND METHODS
Samplt's and DNA extraction. Rcfrigeralcd or frozen sam-
ples of nine commerCial Nonh Allanlic tlmfish s.pecies (order
Pie II I'(>n ectijo I'm es). Lepidorholllbiis bO.tcii (follr,spolled scald-
fish). LepidorhOlllbus whi{fiOQonis (megrin). ScophrhalmflS maxi-
mIlS (rurbOI). S. rlwlI!bllS (brill). Hippoglos.W5 llippog/osms (hal-
ibut). P. jfesll.\ (flounder). P. plale.ua (European plaice). Reill-
hardlill.f llippoglossoides (Greenland halibut). and S. pulga!'it
(=5. solea) (sole). were collected (Table I). For most species.
whole individuals were oblailleo at fish markets in nonhwesl
Spain (Cangas. Ferrol. and Vigo). FroLen whole R. hippoglosso·
ides individuals were oblained dircclly from a scientific expedition
in Ihe nonhwcst Atlanlic waters. Fr07.en H. hippo/(Iossiis adults
and juveniles were oblained from a manufacturer in Vigo. Spain.
and from Newfoundland. Canada. walers, along with caudal piec-
es of previously identified indi viduals from Norwegian fish mar-
kel'. Every specimen was morphologically idenlified (28. 3/).
TOlal DNA was extracted from the muscle lissue of each
individual (/4). Approximately 100 mg of muscle lissue was di-
ge.qed with 500 j.ll of extraction buffer (50 lru\ll Tris, 25 mM
EDTA, 1% SDS) and 10 I-LI of 20 mg/ml proteinase K (Roche
Diagnostic GmbH. Mannheim. Germany). The mixture was in·
cubated overnight al 37°e. DNA was extracted once wilh 600 ,.d
of phenollchloroform/isoamytalcohol (25:24: I) and once with 550
fJ-l of chloroform and IVa, then precIpitated overn.igl11 Wilh 1 ml
of 95% ethanol at - 20°e. The tubes were centrifuged al 6,000 x
g for 10 min. and lhen the ethanol was discarded and the pellets
were dried in a vacuum concentrator syslem (SpeedVac plu , Sa-
vanl) for 10 min. The dried pellels were resuspended in 200 III
of TE buffer (10 mM Tris. I mM EDTA [pH 8.0)) and slOred al
-20°e.
DNA sequencing. Double-stranded DNA from PCR reac-
tions with sufficienl amountS of amplified DNA were cleaned us-
ing the Qiaquick PCR purification kit (Qiagen GmbH. Hilden,
Germany) according 10 lhe manufacturer's instrul·tions. The DNA
was eluted in 30 10 40 ).1.1 of lhe EB buffer provided with lhe kjt.
The concenlralion of the purified PCR produci was eSlimated by
agarose gel electrophoresis wilh a slandard (Mass Ruler. BioRad
Laboratories. Richmond. Calif.) as a reference marker.
Purified peR products were directly sequenced in a Beckman
CEQ2000 sequencer (Beckman Inslrument<,. Bucks. UK) ill the
Facuilade de Cicncias. Universidade de Vigo. Spain. DNA se-
quencing was accomplished with fluorescence dye labeled di-
deoxynucleolides and reagents provided with the CEQ Dye Ter·
minator Cycle Sequencing Kit (Beckman !nSlnlmcOls). For each
sequencing reaction, approximalely 30 ng (aboul 100 fmol) of
purified PCR product was added. To subslantiate correel DNA
sequences. restriction site locations. and fragment sizes. direcl and
reverse sequen{'es of PCR fragments for each of lhe primers used
in inilial PCR amplification were obtained. Sequences were ana-
lyzed and prepared for publicalion with the help of the ESEE
compuler program (.:f). All sequences are given as lheir non-lem-
plate-Slrand equivalents (Fig. 1) and have been deposited in
GenBank (accession numbers AF413798 10 AF41381O).
Restriction site analysis of PCR products. For Ihe reslne-
lion site analysis of the lRNAGlo/cyt b region. DNA sequences
were compared with other sequences previously reponed or in-
cluded in inlernational databases. Several REs were selected for
species-specific rcslriclion pallems with the DNASIS computer
program (Hitachi Software Engineering Europe, Ardon, France).
The punned peR products we.re separalely digested wilh
eighl REs (Roche Diagnostic: New England BioLabs. Beverly.
Mass.) Wilh 4- or 5-bp recognition ~equences (Table 2): AI,,1.
Ddel. HaellJ. HapU. Hinfl. ,Hat'L /I'lboT, and NeiL Saljsfactory
results were also obtained wilhout previous puri ficalion of Ihe
amplified DNA fragments. A 10-~1 aliquot of PCR product Wilh
about 100 to 150 ng of amplified DNA was di.rectly digested
through incubalion at J7°C overnighl wilh 5 U of an RE. The
resulting fragmenls were separated by eleclrophoresis on II 3.0%
1018 SANJUAN ANP COMESANA J. Food Prot.. Vol. 65. No.6

1 L14724 60
FIGURE I. Aligned DNA sequences of H. hjppoglossus Cl CGAAGCTTGATATG~~~CCATCGTTGttattcaac~acaa9aaccc~aA~a9tc
the amplified tRNACil'lcyt b region for H. hjppoglossus B1
L. bosC.li Vl , . . . . . . . • . . . . . . . . . . . . . . . . . . . . c •.•.............. t ....• a ... ,ca
nine flatfish species. Codes for each se- L. whiffiagonjs C5 , _ , •....... t a _, .. c.
quence include Ihe species name. one or L. whiffiagonjs GS7 . . . . . . • . . . , . . . . . . . . . . . . . . . • . . . . . . . • • . . . • . . . . . . . t. a .. .. C.
two letten fur the area of origin (B. Ber- P. flesus Cl ............................... , _.._._.. a ..
P. placessa [2 ................... , , '_''_'_' .ac.
gen. Norway; C. commercial from tile R. hippoglossoides Nl
norJlllvest Iberian .....arers. e:tcept H. hip- R. hippoglossoides N5 ................................................... ~ ....
s. maxim1..'s Cl ...........•.................•........... g t .. '" a c.
poglossus C I, which is of unknown origin; S. maximL's V1 ...........•........................... t.g t a c.
GS, Grand Sale bank: N, Ne....foundlaJld. S. rhombus V 1 .... , " , t a c.
S. vulgaris C1 0 •••••••••••••••••••••••• q. __ .--- a ~ .. , •••••••••••••• •

Canada. waters: V, RIa de Vigo. Spain) Hae!!!

{JIld (he number of lhe individual. III the !'l.3e

firSI DNA sequence (thaI of H. hippoglos- 120

sus C I J. the corresponding coding se- H. hippoglossus C1 taoqt&&6tecOAocotettctaaAaatcgC&6acqat90tttagtcqacctcccoqcce
H. hippoglossus 81
quence of Ihe cyt b gel/e is in boldface L. boscii V1 • • • • il. • • • • • • • • • • • ct.a .. t , c .. ~ . a., t .• a.
type, and the positiolls of primers LJ4724 L. whif{iagonis C5 o ••• a ct. a .. c . . , c .. CC---" • • • • • • • • • • , • t .. ,a.

and H 15149. used for PCR amp/ificatioll L. whiffiagonis GS7 · a c ( . Q •• c .. , ~ c , . c~ c. .. a .

P. flesus C1 .c c t ~ ~ t..:......:...!. .a ..•.
and sequencing. are indicated in capital P. platessa C2 • c: •••.•.•••••••• c t, ..:......:....! ,a .
fellers. A dOl indicates identity wilh the R. hippoglossoides N1 , , , t .a. , ..
R. hipPOylossoides NS · , t .a .
firsl DNA sequence. alld a dash indicates S, m3ximus C1 .... a a.c 9..:..:..-'. ••••• t a..:......:...!. t .
a deletion. ReslriC/ion recognitioll sites for S, maxirnr;s VI .... a a .c 9..:...:....:. ••••• t ...•....... 3..:......:...!. t .
S. rhombus V1 .. , .a .. g c . .c 2..:.:...:. ag .....••.. a ~ _ t.
REs With diagnostic capability are ullder- 5. 'Julqaris Cl · a t . aa. t. . , att aget. eg. ta. td.:. a .
lined (single or double). 111>01 Dde r MbQ I
l'-1ae

180
H. hippoglossus Cl ecteta&t4toeeqqeet9a~q&&ctttqqgtetetttta~t9tttA&ttaeeo
H. hippoglossus B1 ·~ , , , a .. ~ . , . ~ .
L. bosel i V1 .a .. a a a c .. c .. c c.g~.t. . .':..:...:..:..cS,:,,' .
L. wniffiagonis CS .a .. 0 ••c a ~a .. t, .c . . c .. c ~ ~c .
L. ~,iffiagonis GS7 .Cli •• 0 • •c 3 • • • • • .3. •• t. . . c .. C • • c ~~ ~c .. ~,.

P. flestls Cl .... a .. c t. .. C • • . • • a ' .. c .. c Q.d . • • •

P. placessa C2 . . . . • • C ••••• t .. C •• , .• a ~ .c, .c ..o.......!.... • • • • • • g.~
R. hippoglossoides N1 · c g .. a c. , , 9. a .
R. hippoglossojdes NS ............. c ....• q •. d c" , ••...... g.3 .
S. maxirnus C1 .......... ~ a., , .c c.t~ E...:.....:....cc .. t..
5. maxiJ/Ius V1 .......... ~ a., c c.t.:.....:...E:..., •. £.:....:...:...cc . . t .
S. rhombus V 1 " t .. 0 g .. a c .. c c.t~ ~ce .. ,
S. vulgaris C1 · ", .. g .. c .. t .. t.ea a. ,c . . c .. 21 ••••• C .• , •• ce ,.
DdeI Ddel i'iael rr OdeI DdeI
Mael Hinfl Hi.ofl

240
H. hlppoglossus C1 &&Attgegaee9gettatt~eataeaetaeaC&t0A9&oattqotaeCqeettca
H. h~ppoglossus Bl
L. bo.sci i V1 ..:...,gc. cet.a .. a .. gc. c .....•.•... , .. t: cq ea 9 .. t .. t .
J,. w!>Ufiagonls C5 · . c. cct,a .. a .. ac t t cg t .. c . . a .. a: •• t .
L. whltfiagonis GS7 · .c.ccta .. a .. ac t. . . . . . . . . • . t eg t . . c . . a .. a .. t .
? f1~sus Cl . , . >" .c .. ~ c t t .. t .
P. .old tessa C2 ....... t. ~ Q ••••• r. t..t. .
R. hippoglossoldes N1 ...... tt g ~ t t .........•......
R. hippoglossoides NS ...•... t g... . , r. t .
S. /Oaxi"us C1 · .C.Ceitt ~ .. t t .. g , .t .. cg.q .. t . . c .. 0 . . a t.
S. lDeximus V1 · .c. cat.t ~ .. t t. .• g t .. cg.q .. t . . C •• a . . a t.
S. rhombus Vl · .c. cctt gc. t t g t . . tg.g .. t. .c .. a .. a t.
S. VUlgaris Cl .... t.c .. a .. ac. c .. Q; • • c .. a .. g. ~ t ..• q . ... t.gcct .... c .. a .. '( t
Hlfltr Ddel
"'jaer

SeaKem LE and Nusieve GTG agarose (2: I) gel (FMC Bio-
Produc(~) containing ethidium bromide in TBE buffer al 10 V/cm
for 60 or 90 min, depending on the lengths of the fragmenls. Gels
were sometimes incubated in an elhidium bromide Solulion (I 1J.g/
ml) for 10 min after electrophoresis. The resuiClion fragments
were visualized by UV tranSillumination and photographed with
a Polaroid camera, and their sizes were delermined by comparison
with a 50- or 100-bp ladder (Amersham Pharmacia Biotech) with
I D Manager DNA analysis software (IDt Alcobendas, Madrid,
Spain). Only fragments larger than 70 bp were considered in the
RFLP analysis, as smaller fragmenls were not always clearly vis-
ible.
RESULTS AND DISCUSSION
DNA sequence anaJysis. Twenty-two PCR products of
the tRNAGlu/cyt b region for nine flatfish species were di-
rectly sequenced for both senses with universal primers
Ll4724 (23) and HI5149 (17) (Table I). Alignment of the
486-bp sequences showed 13 different sequences for all
species combined (Fig. I). The four species for which three
individuals were sequenced (H. hippoglosslJs, L. whiffi-
agonis. R. Izippoglossoides, and S. maximus; Table L)
showed two different DNA sequences each. A 3-bp deletion
before the coding sequence of cyt b gene was found for S.
vulgaris. Published cyt b sequences for P. jlesus. P. plotes-
so, R. hippoglossoides, S. rhombus. and S. vulgaris (II, 30,
32) were compared with sequences obtained in this work,
and no large differences were found, as the inferred poly-
peptide sequences were very close for each species.
Identification of flatfish species by RFLP. The re-
striction maps of the sequences were compared to deler-
J. Food Pro!.. Vol. 65. No.6 PCR-RFLPs OF FLATFJSH SPECiES 10)9

JOO
H. ~jpP091ossus Cl QQtcc~~ea~AtctqtogaqAC9~a~o9qctqace~tceqa·9eAtteAtq
FIGURE l. COli Iill ued.
H. hippo91oS5US B1
L. boscii Vl .~ ~.a.~ , . t •. G'I~ • • • • • • • , ••• t , a, .. " .. a,c .c .
L. ~~if!iagonis C5 . . .;t . • • . • • , .c •. c . . . . . . . . • . , t .. t .9..:...:.....:, .•• a,.;;.g .
L. whiffi.gonis GS7 •••• , •• 5., •••••• • t,.c , r. .. t 1...:....:....:., •• d.c . ..:. .
P. fles~s Cl ....... L . . a g .....•....••.. t . ~ .
P. pl.tessa C2 . . . . , .. t, .4., c~.g .. , .. ~, 1 •••• '.

1'<. hippoglossoides Nl ••• 1 •• t., , .. c .. g., , , , ..

k. hippDgJC~'soidi!S N~ ...... c., c .. g., , , .
s. arax-imus C1 · .. c 1 • C ••• ace, , t. •• t . , t t. a d. c. a .. C.
s. maxirnus '.11 · .. c C ••• ace, t: .. t , t t . 9...:..:...:. •.•• a . C. a .. ~ .
s. ~holl'..b"s VI ... c t .. c, ,t t,. t., t a.., d.C.d • • ::.
S, V'l.Jlqaris Cl ... c. tg c, .g t. •• t .. t . , Cl ••• t.o , .atg.c .. c.
Haerl I /finfl :-1bOI

36;)
H. hippoqJossus Cl ecaaeqqcgcatcattctttttcat~t9CctetAcctCcatat~a99aCt5eaCe
H. hi~ooglossus Bl
L. boscjj Vl · t , , , , L •• tc. a. .. a. I . a. g .. c: •• c. .. .., ct.. g ..
L, whirfiagonis C5 ·., , ,c .. tc. a .. Cd, ••• , C1. a, ,C_.'_'_' •••,ct. g .
~. whiffioqonis GS7 .......... , ,c . . Le.a, .ca, ,.=. • .11, ••• ,c_.'_'_'" .. ct.g .
p, {Jesus CI · .. t .. , , .. a . t. .• , ••• , ••• ,C_._.'_' •••• t . , ,
P. plHess. C2 • ••••••••• , •• , •••• , ••••••• ,. ~a.< •.•••.• , •. •• c~ .. .. t .. C ••• ,
R. hippoglossoides Nt • ••••••••• , •••••••••••••••••• a, ••• , •• a, .C. ,c...:...:-:..:.. .• , .9. ,c ••• ,
R. hippo~lossoid~s NS • •••••••• , , •• , ••••••••••••••• a, •• , , •• a, •• c •• c..:...:....:... •• '~.:..~.' •••
S, rn~:dIlu..'S ;:::1 .. c c t..a~. c.a.t. .. tgca c...:....:...:.. .. , . t .. g .. ,.
5. r.:L~ximvs Vl · , , .. c ,.c t. a .. t:a. t. , tqca c..:.....:.....:' t , . g .
S. rhombus Vl • •....... , , .. t c c. c:: •. to.t,. a. a ~ .. ~ .. t t .
S. t'Ulgd.ris Cl · , , , .. , ,. c .. t .. , a. t.., •••••• L:, • C. ,iI, , . t. .
HoelIl H,nfl

420
H. hippogiossus CI a~otetueet<:ta.ua.a9' e"tgll"etljttg9'g~ttAtt<ltteteott<:togta.ll
H. hippoglo55US Bl ........ , , , , t , ,
L. boscLi VI .c a .. c .. , .. c., g .. c clC ••a .. a ct.a .. t ~.c.
L. whiffi~gonjs CS · a~ t .. c . , g .. c ae .. a .. C, • c .. c .. c . , c, t ,
L. ~hifi;agon1s GS7 · a ~ t .. c~ g •. C . . • • ac .. a .. c .. cg ~ c .. c c t .
p, fl€5US Cl ....... , ,t, 9 .. , ...• il a g.a. .. d. ,g t,
P. pldtessa C2 , , , t, q .. , , a.C . . a, •• • g
, a .. g
R~ hippoglossoides Nl , a , , , , a ~g.c .. c
c ....•..
.R" hippoglossoides N5 , •••• ,a •••• , , ••••••••• " ••. a .:_~g c .. c c ,
5. ma.--:i,.lIus C1 · c, ,t , c, , .. ac .. g. , i1 • • • • ,c t .. , , t ,
S. ma."::rnus V1 · c, , ,t , c, ae .• g .. a ,.c , . t.. , , , _.
S. rhombus Vl .c,~c " t .. c, ,g." .. t" •• qc .• g .. <3 ••• , ,. t. ,<;I •• c, ,c.
S. '-'uJ garis Cl · c .~a .. tq .. a. c ..... 9 . c .... aco. c .. a. . . . . ta. a .. a .. t.
,'!bol H1Ml Mdel

H15149 486
H. hipp glossus CI taatABoa9'eQttC9teg9'atae9't<:OtcQetTGAGGAC~TATCATTCTGAGGGGCTGCAGTTT
H. hippogioS5US 81
r... bosci.l V1 .. 9<0 <0· .. l . . . . . C .g .. t. _ _
, , ,
--
L. :.,Jh~=f.la.gonis C5 .. gc c. . .. c~~ a. __
L, whiffiagonis GS7 .. gc.,.C , , .. c " .. ~.,'- , , ....••... ,
p, (lesus C1 .. 9 .. c ... ...q, ,......... , ", ..
p. pld{';ess~ C2 .. , .g., , c, , , , . . . . . . . • • • . . . . . . . . . , , ...••••. ,
R. hippoqlossoid@s Nl ....... ~ c __ .. __ __ .......
R. hippoglossoides N5 ••••••• ~ ••••• C ••• , . . • , •••••••• , •••.•••••••. __
S, maxi,'rIus C1 ,gc ... t ... , . . ~ .. c . t. •• t . . t. .. ~.

s. r.raximvs \'1 · '9c. , , t .. , , .... c .. c' . t.. , t . . t. , . £.:....:.....:. ••• , , ••••••••• ,

5. rhc,nbus VI .~ .. c.. .c .. c .. t. .. , .. ~ .. ~ , ...................•••.. ,
s. "rulgarl~ Cl ·,9"c .. ~ a. •• , •••• " ••• ,.c " , , .
Haell! HaelU ~ 02iil

mine which REs could distinguish and identify the peR DdeI for L. whiffiagonis. with Hinfl for H. hippoglosslIs
products of each flatfish species. Eight REs with 4- or 5- and R. hippoglossoides, and with MboI for S. maximus (Ta-
bp recognition sequences were selected for screening, and ble 3).
different electrophoretic restriction patterns were detected The RE with the strongest diagnostic capability was
for each RE (Table 2). Ddel. followed by Hinfl, Hae11I, Mael, and Mbol (Table 3
Agarose electrophoresis following digestion of peR and Fig. 2). Ddel produced eight different restriction pat-
products from 130 individuals of the nine flatfish species terns (haplotypes) for an
species (Table 3). Five of the nine
showed band sizes that were in agreement with the ex- studied species (L. bosch. P flews. P. plaressa, S. IIwxilllllS.
pected sizes for the restriclion fragmenls inferred from lhe and S. vulgaris) and two species groups (one group includ-
sequence analysis. To visualize and clearly disti nguish frag- ed H. hippoglossus and R. II ippoglossoides. and the other
ments of similar lengths for our electrophoretic conditions. included S. rhombus and several L. whiffiagonis individu-
agarose electrophoresis was carried out for 90 min. e.g .. for als) showed different haplotypes. Severnl L. whiffiagonis
the 470-bp fragment (haplotype A) and the 449-bp frag- individuals yielded haplotype E instead of haplotype F ac-
ment (haplotype B) with Ddel, or for the long fragment of cording to the expected sizes by the absence of a restriction
haplotype B (323 bp) and lhe 310-bp fragment of haplotype sile in position 453 (300 bp = 282 bp + 18 bp; see Fig.
E with Hinfl. Each RE yielded different restriction patterns, I and Table 2). Consequently, the numbers of species
dependi ng on the studied flatfish species (Table 3). intra- groups and species identified with only DdeI can be con-
specific restriction polymorphism was found only with sidered 7 and 5. respectively (Table 3). The enzyme Hinfl
J. Food Pro!.. Vol. 65. NO.6 I'CR-RFLPs OF FLATFlSH SPECIES 1021

MW Rh Lb MW Sv Sm Ode MW Pp Lb MW Sr $v Haeltl
a Hh MW Lw Pt MW Sr bp b Hh MW Lw 8m MW Rh bp
.......- -....

->800 ->800
->500 -)500

->300

-140
-) 100 ->100

B A DD D E F C
A A H E C B G F
MW Pf Lb MW Rh 8m Hinfl d MW Lw Pt MW Sv 8m Mbol
C Hh MW Lw Pp MW Sr bp Lb MW Rh MW Sr bp
.....
------_-.:..._--...,....---....,..~

->800
->800
->500
->500
->300
->300

-163
->100
-> 100 A C B B A F D G
B F C A E G o A
FIGURE 2. EleCirophorelic l'eslriCiiol1 pallems of Ihe peR products of fhe cyt b region digested with different restrictirm enzymes: (a)
Ddd; (b) HaellJ: (c) Hi'll7: (d) Mbol. MW. moleclIlal' weigh I IIImker (/OO-bp ladder). Species codes: Hh. Hippoglo.<,sus hippoglos.<,u$:
Lb. Lepidorhombus boscii; Lw. Lepidorhombus whiffiagonis: Pf. Plalichlhys l1esus: Pp. Pleuronectes platessa; Rh, Reinhardtius hip-
poglossoides: Sill, Scophlhalmu$ maximus: Sr. Scophlhalmus rhombus: Sv, Solea vulgaris. H{/plorypes (/( the bOIlOIll of each figure (Ire
as in Table 2,

distinguished four species (L. boscii. P. fiesus. P. platesso.
and R. hippoglossoides), two species groups (one group in-
cluding S. maximus and the S. vulgaris, and the other group
including L. whiffiagonis. S. rhombus. and several H. hip-
poglossus individuals with haplotype A), and several H.
hippoglossus Individuals with haplotype B. Consequently.
the numbers of species groups and species identified with
Hinfl can be considered 6 and 4, respectively. HaeI1I, Maer,
and MboI also distinguished six species groups and four
I';pecies. each involving different species in the species
groups and in the species identified (Table 3).
Each studied flatfish species can be identified with spe-
cific parterns of different REs (Table 3 and Fig. 2). H. hip-
pog/osslls was clearly identified with Haell! (haplotype B)
(Fig. 2b); L. boscii was identified with Allil. Ddel (Fig. 2a),
Hint? (Fig. 2c), and MaeI; L. whiffiagonis was identified
with All/I, OdeI (haplotype E) (Fig. 2a), Maer, and MhoI
(Fig. 2d); P. jesus was identified with Odel (Fig. 2a) and
Hil7fl (Fig. 2c): P. platessa was identified with DdeI and
Hintl (Fig. 2c); R. hippoglossoides was identified with
Haem (Fig. 2b), Hintl (haplotypes G and H; Fig. 2c), and
Maer: S. moxinws was identified with DdeI (Fig. 2a) and
MhoI (haplolypes 0 and E; Fig. 2d); S. rhomblls was iden-
tified with Haem (Fig. 2b). MboI (Fig, 2d), and Neil: and
S. vulgaris was identified with AluL DdeI (Fig. 2a). Haem
(Fig. 2b), HaplI. MaeI, and MboI (Fig. 2d). No single en-
zyme could distinguish the nine studied Aatfish species. but
with several combinations of only a couple of REs, tissues
of these nine flatfish species can be identified. Any com-
bination of two of the most diagnostic REs (DdeI, HinfI,
Haem, lvlael, and MboI) except the combination of OdeI
with Hinfl and MboI and the combination of Haem with
lvlhol can unambiguously identify all nine flalfish species
(Table 3). Nevertheless, these results are based on a rela-
tively small number of individuals, and the consideration
of a more extensive number of individuals of each species
from different origins might increase the confidence leveL
These results are in agreement with those reported for
digestion of a 359-bp amplified fragment of the cyt b gene
with HinfI, Neil, and Sau3AI (isoschizomer of MboI) for
1022 SANJUAN AND COMESANA
P. flest/s, P. plolessa, and S. vuLgaris (J I). The Neil en-
donuclease cleaves the P. platcssa and S. vulgaris peR
products in Cespedes et al.'s (11) position 152, which cor-
responds to our position 262, but it does not cleave for P.
jfesus (Fig. I); Sau3AI (=MIJOI) cleaves the S. vulgaris
PCR product in their position 254, which corresponds to
our position 364, but it does not cleave for P. j1es/./s and P.
platessa (Fig. I). However, in the present work, the MboI
enzyme cleaved the PCR products of these three species in
position 108, but this restriction recognition sequence
comes before nucleotide 1 in Cespedes et at's (J 1) 359-bp
fragments. Digestion with Hinff cleaves PCR products in
different sites in each species; Cespedes et al.'s (1 J) restric-
lion position 54 (cleaving for P. flesus and P. plalessa PCR
products) corresponds to our position 164, their position 67
(cleaving for P. platessa) corresponds to our position 177,
their position 84 (cleaving for S. vulgaris) corresponds to
our position 194, and their position 171 (cleaving P.jlestls)
corresponds to our position 2R I (Table I). The absence of
intraspecific polymorphism for the recognition sequences of
these three REs in the report by Cespedes et al. (J 1) and
in the present work, with different individuals collected in
different places and years, supports the usefulness of these
REs in distinguishing and genetically characterizing P. jle-
SU5. P. p!ales.I"a, and S. vulgaris.
Intraspecific restriction polymorphism. One of the
problems with this technique is the existence of intraspe-
cific variability, giving rise to the possibility of mutations
and genetic variability within a species and, consequently,
the loss or gain of restriction recognition sequences. Ge-
netic database and population studies on DNA variation for
each species may establish the degrees of confidence for
different REs. Results of the PCR-RFLP analysis of a 464-
bp fragment of cyt b (practically the same region used in
thi S work but 14 and 8 bp short at the 5' and 3' ends,
respectively) for S. rhombU5 and P. plaressa with eight REs
(32) are in agreement with the present results except for S.
rhombus with HaeIII (32); a typographical mistake was
found for S. rhombus with HaeIIl (299 bp is written instead
of 199 bp). Of the five reported fragments, three have the
expected lengths in the present work-the reported 5' end
fragment of 149 bp (corresponding to our 163 bp length;
149 + 14 = 163), the 3' end fragment of 55 bp (corre-
sponding to our 63 bp; 55 + 8 = 63), and the 77-bp frag-
ment-but two fragments (144 and 39 bp in length) cor-
respond to our 183-bp fragment (Tables 2 and 3). This dif-
ferent result is the consequence of the substitution of a G
in our position 309 for a C in the reported DNA sequence
(EMBL accession number AF165084) and, consequently,
the addition of a recognition site for Haelll. Actually, there
are two changes: our 5'-GC-3' (our positions 309 and 310)
corresponds to 5' -CG-3' in the reported sequence, and the
second G represents a change of a conservative amino acid.
This means that more than one individual must be analyzed
to obtain a high level of confidence.
On the other hand, the mtDNA sequence variation of
a 401-bp portion (without primers) of the cyt b gene (prac-
tically the same sequence as that used in the present work)
J. Food Prot., Vol. 65, No.6
for 280 R. hippoglo550ide5 individuals throughout the
NOl1h Atlantic was studied (30), and 22 different sequences
were found (GenBank accession numbers L77930 to
L7795l). These sequences were aligned with our sequences
for a reanalysis to obtain restriction maps with the DNASIS
computer program. Results of the reanalysis show that for
the AluI restriction enzyme, only 2.5% of 280 individuals
had a restriction pattern different from that found in the
present work (with a new cleavage site in aUf position 250);
for HaplI, 0.4% were different (with the loss of a cleavage
site in position 261); for Hinff, 1.1 % were different (with
a new cleavage site in position 281), with 45.7% being
analogous to our haplotype A and 53.2% being analogous
to our haplotype B; for Neil, 0.36% were different (with
the loss of a cleavage sile in position 262); and for DdeI,
HaelJI. MaeI, and Mbor, no differences were found. Taking
these results into account. at first sight, the probability that
one of the restriction patterns (using a single RE) shown in
this work is not present in the R. hippoglossoides species
is very low (with the highest percentage being 2.5% with
A/ul). However, this probability may be reduced with a cou-
ple of selected enzymes, e,g., Ddel and Haem. DdeI and
Maer, Haem and MaeI, or MaeI and MboI. Consequently,
results of the reanalysis of the previously reported genetic
data (30) clearly suppon the usefulness of the PCR-RFLP
analysis of the cyt b region in identifying with strong re-
liability and confidence R. hippogLos50ides in particular and
any fish species in general. When two REs are used, it is
enough to obtain an unambiguous identification, but often
it is sufficient to show the absence of a putative species in
a flatfish fillet (e.g., no sale present). Consequently, the
method described here is a powerful method for the detec-
tion of mislabeling or fraudulent SubSlicution of these spe-
cies. This method provides an alternative for flatfish species
identification that is simpler than sequencing. Nevertheless,
because of the existence of variability that has not been
previously detected, as shown above, more studies of in-
dividuals of most of these flatfish species are necessary [0
determine the genetic variability of each species throughout
its range of distribution and, consequently. to establish a
high degree of confidence.
ACKNOWLEDGMENTS
We are indebted to Franc Saborido (In,tltulo de Inve,tigaci6ns Ma-
rinas. Vigo. Spain). Angel Mendizabal (Men:aMadrid. Madrid. Spain).
Joseph Brown (Ocean Science Center. 51. John's. Canada), Tere,a Marti-
nez Areal. Ana Gago, and Jose A. Gradin Maninez for supplying (hh
samples and to F. Luis Venttn Sanchez and Paz Abella for technical sup-
pon. We also lhank two anonymous reviewers for useful suggestions. This
work was supported by grants ALJ98-0535 (Direcci6n General de Enscli-
anza Superior e tnvestigaci6n Clcnlffica. Ministerio de Educnci6n y Cul-
lura. Spain) and 64502.C347 (Univcrsidade de Vigo, Spain) to A. Sanjuan
and an equipment grant for a DNA sequencer (Xunln de Galicia. Spain).
We lllank the Universidade de VIg.O, Spain, for the technician fellowship
of E Luis Ventin S~nc:hez.
REFERENCES
1. Andrews, A. T. 1998. Electrophoretic methods, p. 204-239. In P. R.
Ashurst and M. J. Dennis (ed ), Analytical melhod\ of food authen·
lication. Blackie Academic and Profe>sional, Thomson Science.
London.
1. Food Prol.. Vol. 65. No.6
2. Bankll. S, E.. "nd W. S. D'lVidson, 1991. Identilicalion of Tlmllllll;'
/IIna species by Ihe polymerase chJin reaclion and dinxt sequence
analy 'is of Iheir milochondrial cytochrome b genes, Can. J. Fish.
AquJ!. Sci. 48:309-3l7.
). Banlell. S. E.. and W. S. Davidson. 1992. FINS (forensically infor-
mative nucleolide sequencing): a procedure for identifying the ani-
mal origin of biological specimens BloTcclmiques 12:408-411.
4, Cabot. E, L.. and A T Be<:kcnb~ch. 1989, Simuhaneo\l5 wiling of
multlpk nucleIC acid sequenccs wHh ESEE. COmpUI, App!. BioSci.
5'233-234.
5. Carr. S. M.. and H D. Marshall, 1991 A direct Jpproach to me
measuremenl of gen~lic \'arialion in lisll populallons: applicalions of
Ihe polymerase chain reaction to studies of Atlant;c cod. Gudus mor.
IlIIa L. J. Fish BIOI. 39(SuppJ. A): 101-107.
6. Carrera. E.. T Garda, A. Cespedts. J. Gonz:l.lez. A. Fernaodez, P E.
H~rniindez. and R, Manin, 1999. PCR-RFLP of the mllochondrial
cylochrol1.\e oxid~se gene: a. simplc melhod for discrimination be·
tween Athnlic salmon (Salmo safar) and rainbow trout (Olleorllyll'
elllls m)'''I55). J, Sci. Food Agric. 79: 1654-1658,
7, Carrera. E.. T Garcia. A. ce5ped~s, I. Gondlez. A, FemiindeL. P E.
Hernandez, and R. Manin. 1999. Salmon and trout analYSIS by PeR-
RFLP for identity aUlhcllIiC;llion. J. Food Sci. 64:410-41.3.
S. Carrera, E.. T. Garda. A. Cespedes. I. Gondlez. B Sanz, P E. Her-
nandez. ~nd R. M~r1in. 1998. Identlncallon of Atlanlic salmo/l (Sal.
mo salar) and r:Jinbow IrOUI (Ollcorhynchus ",)'kl,\,') using PCR am-
phfic31ion and reSlriClion ilnalysis of Ih~ mllochondml cylochrome
b gene. J, Food Pro!. 61 :482-486,
9. Cespedes. A.. T Garda. E, Carrera. J. Gonziilel. A, Fernandez. L.
Asensio. P. E. l-Iern;indez. and R. Manin. 1999. Genetic discrimi·
nallon among Solea solea and M,crocillms a<:evra by RFLP analy,is
of PCR ;unpJlfied roitochondnal DNA fragmenlS. Arch. Lebensmlt-
leJhyg, 50:67-70.
10. Cespcdes. A.. T. Garcia. E, c,mera. I. Gonzalez. A, Fernandez. P. E,
Hernandez. ~nd R. Manin. 1999. Identllication of sole (So/eo solea)
and Greenland halibut (ReillilardtlllS IJippogfos50ides) by PCR am-
plification of the 5S rDNA genc. J. Agric Food Chem. 47:1046-
1050.
II. Cespedes. A .. T Garcia, E. Carrera. r. Gonzalez. B Sanzo P. E. Her-
nandel. and R. Martin. 1998, Idenlillcallon of flatfish sp("ies using
polymc,rase ch~in reaClion (PCR) ~nd rcsu;C1ion analySIS of the cy·
10chrome b gene J. Food ScI. 63:206-209.
12. Cespedes. A.. T. Garcia. E. Carrera. J. Gonlala. B. Sanl. P E. Her-
nandez, and R. Martin. 1998. Polymerasc chain reaWon re,tr1clion
fragment length polymorphism analySIS of a shon fragment of the
cylOclJrome b gene (or ,dent; lication of flatfish species. J. Food Prol.
61 1684-1685.
]J Davidson. IN. S. 1998. DNAlPCR techniques. p. 182-203. In P. R.
AShurq il\d M J Denni, (cd.). Analytical method, of food authen·
tICatton, Blackie Aca,Jem IC and Profes,ional, Thomson Science.
London,
14 DeSalle, R" A. K. Wllhams, and M. George. 1993. (solation and
char~cl~rizallon of animal nltlochondrial DNA. Methods Enzymol.
224: 176-204.
15. Elienne. M" M. Jerome, J Fleurence. H. Rehbein. R. Kundinger, R.
Mendes. H. Costa. R. pcrez-Milflin. C. Pi.1eiro-Gonl;5let. A. Craig.
I. Mackie.!. M. Ymnn. M. Ferm. l. Manintz. f Jessen. A. Smell,
and J. B, Luten, 2000. Idemification of fish ,pccles ~fter cooking by
SDS-PAGE and urea IEF' ~ collaborative siudy. J, Agric. Food
Chem. 48:2653-2658,
16, John,. G C.. and J, C. Avise. 1998, A comparative sununary of
PCR-RFLPs OF rLATFISH SPECIES 1023
g 'nel ic dlstance-s in Ihe venebrales trom thc m llochondnal cylo-
c11rome /; gene. Mol. Riol. E\,ol. 15:1481-1490.
17. Kocher. T D., W. K. Thomas. A. Mcyer. S. V. Edwar Is. S. P;iiibo.
EX, Vtllablanca. and A, C. Wilson. 1989. Dynamics of milOchon·
drial DNA e\'ollllion in "nimals: amplification and sequencing wilh
conserved pnmcrs. Proe. Nail. Acad Sci. USA 86:6196-6200.
18. MacklC. 1. M.. S. E. Pryde. C. Gonwlcs·Sotelo. I. Medina. R. Perez-
Manin. J. Quinleiro. M, Rey-Mendez. and H. RehbeIn. 1999. Chal-
lenges in the identificalion of species o( canned thlJ, Trends Food
Sci. Teehnol. 10:9-14.
19, Mitnlbauer. E. 1998, Antibody lechniqucs. p, 241-268, III P. R,
A,hur>l and M, J. Dennis (ed,), Analyl,cal melhods of food aUlllen-
licalion. Blacklc Acadcmic aod Professlon,,1. Thomson Science.
London,
20, McKay, S.l .. M. J, Smilh. and R. H. Devlin, 1997, Polymerase chain
reaCtion-based speCIes idenllfication of salmon and coastal lrOUl in
Brilish Columbia. Mol. Mar. BIOI. BiOlechnol. 6: 131-140.
21. McVeigh. H. P. S, E, Bar1ell, and W. S, Davidson. 1991. Polymerase
chain re2clionldircct sequence analysis of the cytochrome b gene in
SalolO safar. Aquaculture 95:225-233,
22. Meyer. A, 1994. Molecular phylogenetic studies of fish. p, 2 \9-249.
III A R. Beaumonl (ed,). Genetics and evolUlion of aquatic organ-
ism•. Chapman and Hall. London.
23, Pa::ibo. S. 1990. Amplifying ancienr DNA. p. 159-166. 111 M. A.
Innis, D, H, Gelfand, J, J. Sninsky, and T. J. White (ed.), PCR pro-
locol<. A guide to metbods and applications, Academic Pre". San
Diego. Calif.
24, Palumbi. S. R, 1996. Nucleic aCid 11: the polymerase cham reaction,
p. 205-247, 11/ D, M. Hilhs, C.. MonIZ. and B, K, Mable (ed,).
Molecular systematics. 2nd ed. Sinauer Associales. Sunderland.
Mass.
25. QUlnteiro. J, C. G Sotelo. H. Rehbein. S. E. Pryde, I. Medina. R.
l. Perez·Martin. M, Rey-Mendez. and '- M. Mackie. 199B. Use of
mlDNA dlfect polymerase chain reaction (peR) sequencing and
PeR reslriction f ragmenl length polymorphism melhodologies In
species ldemilkalion of canned luna. J. Agnc, Food Chern. 46: 1662-
1669,
26 Ram. J. L .. M. L. Ram. and F. F. Baidoun, 1996. AUlhentlCalion of
canned lun~ and bonilO by sequence and resmction site analysis of
polymerase chain reaction produclS of nlltochondrial DNA. J. Agoc,
Food Chern. 44:2460-2467.
27. Russell. V. J., G. L. Hold. S. E. Pryde, H. Rehbein. J. Quintciro. M.
Rey-Mendez. C. G. SOlel0. R. perez-Man!n. A, T. Santos. and C.
Rosa. 2000, Use of reSlriction length polymorphjm 10 dl stingul sh
between salmon species. J. Agrie. Food Chem. 48:2184-2188.
28. Scoo. W. B . and M. G Scott. 1988. Atlanlic fishes of Canada, Can.
Bull. FIsh. Aqual. SCI. 219: 1-731.
29. SOlelO. c. G" C. Pineiro. J. M, Gallardo. and R. r. Perez-Marlin.
1993, FIsh species Idenlificatlon 111 sc~food products. Trends Food
SCI. Techno!. 4:395-401.
30. Vis. M. L.. S. M. Carr, W. R, Bowenng, and W. S, DaVIdson. 1997.
Greenland halibut (Reinhard/lus hlppogfossoidesl in Ihe Nonh AI-
lanllc are genetically homogenous. Can. J. Fl,h. Aqua!. SCI. 54:
1813-)821.
31 Whilehead. P. J. p.. M.. L. Bauchol, J.-c. Hureau, J, Nielsen. and E,
Tononese (ed.). 1986, Fishes of Ihe north-eastern Atlantic and the
Medilerranean. vol. 3, Clofnam. UNESCO, Pam.
32. Wolf, c., M, Burgener. P Hubner. and J. LUlhy. 2000. PCR-RFLP
analysis of milOchondrial DNA: dltlerenliatlon of fish species. Le-
bensm.-Wiss. Tcchnol. 33 144-150.