www.elsevier.com/locate/foodres

Food Research International 39 (2006) 1023–1028

PCR-based methodology for the authentication of the Atlantic mackerelScomber scombrus in commercial canned products

Carlos Infante *, Aniela Crespo, Eugenia Zuasti, Marian Ponce, Laura Perez,Victoria Funes, Gaetano Catanese, Manuel Manchado

Laboratorio de Identificacion de Especies Pesqueras y Acuıcolas, CIFPA ‘‘El Toruno’’, I.F.A.P.A. C.I.C.E. (Junta de Andalucıa),

11500 El Puerto de Santa Marıa (Cadiz), Spain

Received 10 November 2005; accepted 20 February 2006

Abstract

A multiplex-PCR assay for the authentication of the Atlantic mackerel Scomber scombrus in commercial canned products has beendeveloped. This novel method consists of a S. scombrus-specific fragment (123 bp) corresponding to the mitochondrial NADH dehydro-genase subunit 5, and a positive amplification control corresponding to the small rRNA 12S subunit (188 bp). The system was assayedusing six different canned products labeled as S. scombrus. We observed a positive identification in all samples, revealing this metho-dology as a potential molecular tool for direct application in the authentication of S. scombrus canned products.� 2006 Elsevier Ltd. All rights reserved.

Keywords: S. scombrus; ND5; Multiplex-PCR; Canned products; Authentication

1. Introduction

The last few years have witnessed to a tremendousgrowth in demand of quality food due to changes in con-sumer attitudes toward health and nutrition, especially inthe area of raw and processed fish products. Nowadays,consumers demand quality well labeled seafood products.However, involuntary or deliberate mislabeling exists andmay go undetected, resulting in a misrepresentation ofthe actual quality of that product. Consumers are not ableto identify the fish species when the morphological charac-teristics such as shape, size or appearance are removed atthe processing stage. In addition, the increasing availabilityin markets of cuts, either fresh or frozen, processed orunprocessed, often makes the species unrecognisable andopens the possibility of fraudulent adulteration and substi-tution of high-priced species with others of lesser value. As

0963-9969/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.foodres.2006.02.006

* Corresponding author. Tel.: +34 956011315; fax: +34 956011324.E-mail address: carlos.infante.ext@juntadeandalucia.es (C. Infante).

a consequence, falsification becomes commonplace due tothe enormous profits that can be realized (Mackie et al.,1999).

The family Scombridae contains 15 genera and about 51species of epipelagic and generally migratory marine fish,characterized by an elongate and fusiform body althoughmoderately compressed in some genera (Collette, Reeb, &Block, 2001). This family includes mackerels, bonitos,and tunas, showing a worldwide distribution from tropicalto subtropical oceans (Collette, 2003). Scomber is a repre-sentative genus of the family Scombridae and currentlyincludes four species. They differ in a series of morphologiccharacters, including sculpturing of the skull, number ofprecaudal vertebrae, arrangement of palatine teeth, andnumber of first dorsal spines (Matsui, 1967). The Atlanticmackerel Scomber scombrus (Linnaeus, 1758) is mainlyfound in the north Atlantic Ocean and the Mediterraneanand Black Sea. The Pacific mackerel Scomber australasicus

(Cuvier, 1832) inhabits the west Pacific and the southeastIndian Oceans. The chub mackerel Scomber japonicus

(Houttuyn, 1782) is found antitropically in disjunct popu-

1024 C. Infante et al. / Food Research International 39 (2006) 1023–1028

lations in warm and temperate waters of the Indian andPacific Oceans and adjacent seas, with an overlapping dis-tribution with S. australasicus in the northwestern Pacific.Finally, Scomber colias (Gmelin, 1789) is distributed overthe eastern and western coasts of the Atlantic Ocean, over-lapping with S. scombrus in the north Atlantic.

In Spain, the Atlantic mackerel S. scombrus is highlyappreciated by consumers owing to the excellent propertiesof the meat. To enforce the commercial importance of thisspecies, national legislation to protect the denomination of‘‘Caballa’’ when referring to S. scombrus canned productshas been developed. Yet, the similarity of flesh in bothappearance and texture makes Scomber species undistin-guishable from each other in the absence of morphologicalcharacters. Hence, substitution for other Scomber speciescan occur, and consequently the development of a methodto be directly applied in the authentication of S. scombrus

canned products is demanded.To date, most of the works aimed at identifying the spe-

cies origin of canned fish samples have employed the mito-chondrial genome as their target. Mitochondrial DNA hasseveral advantages over nuclear DNA, including its higherabundance, mutation rate, and number of copies inside thecell (Mackie et al., 1999). Several different techniques havebeen applied, including direct sequencing of PCR amplifiedproducts (Jerome, Lemaire, Verrez-Bagnis, & Etienne,2003; Terol, Mascarell, Fernandez-Pedrosa, & Perez-Alonso, 2002; Unseld, Beyermann, Brandt, & Hiesel,1995), PCR-restriction fragment length polymorphism(PCR-RFLP) (Jerome, Lemaire, Bautista, Fleurence, &Etienne, 2003; Pardo & Perez-Villareal, 2004; Quinteiroet al., 1998; Ram, Ram, & Baidoun, 1996), PCR-singlestrand conformation polymorphism (PCR-SSCP) (Rehbeinet al., 1999, 1998), or real-time PCR (Lopez & Pardo,2005). The multiplex-PCR method for the authenticationof S. scombrus in canned products here described relieson the simultaneous amplification of two mitochondrialregions, one corresponding to the small rRNA 12S subunitas a positive amplification control, and a S. scombrus-spe-cific fragment corresponding to the NADH dehydrogenasesubunit 5 (ND5), which is generated only in the presence of

Table 1List of primers used in this study

Name Sequence

RNA12S�1 50-GACAGCTACGACACAAACTGRNA12S�2 50-TGCACCTTCCAGTACACTTAC

ND5�1 50-ATTATGTCCTTCCTTCTCATCND5�2 50-AGCTAGTAAGGTAAGAGTAA

RNA12Spp�1 50-GCAAAATTGGCACAGCCCAGRNA12Spp�2 50-CGGTGTGTACGCACTTCAGA

SSCO�1 50-TGCCCTACTTCACTCTAGCACSSCO�2 50-GTGGTTAGTGCTCCGAGGCA

Upper: primer sequences for PCR amplification of partial ND5 (ND5�1 and NLower: primer sequences for multiplex-PCR amplification of the S. scombrus-fragment (RNA12Spp�1 and RNA12Spp�2). Amplicon size is indicated in ea

the correct DNA template. The obtained results show thatthis protocol might be a suitable methodology for theauthentication of S. scombrus canned products.

2. Material and methods

2.1. Fish sampling

Genetic analyses were conducted on 16 specimens ofS. scombrus, 15 individuals of S. colias, 10 individuals ofS. japonicus, and 10 specimens of S. australasicus. Samplesof S. scombrus and S. colias were collected in the Gulf ofCadiz (Spain; northeast Atlantic) during a fish samplingperformed throughout September 2004 as a part of the sci-entific project ‘‘Fisheries Resources of the Gulf of Cadiz’’,supported by the ‘‘Consejerıa de Agricultura y Pesca’’ ofthe ‘‘Junta de Andalucıa’’ (Spain). Pacific samples ofS. japonicus and S. australasicus were captured in theKochi prefecture (Japan) using a settle net in October2004. Skeletal muscle of Pacific individuals was preservedin 100% ethanol until receipt by the laboratory.

A total of six canned products of different weights andbrands, all labeled as ‘‘Caballa’’, were purchased at localsupermarkets. The content of each can consisted of a var-iable number of distinct fillets. Two randomly selected fil-lets were analyzed per can.

2.2. DNA isolation, amplification and sequencing

Total genomic DNA was isolated from 150 mg of raw orcanned muscle sections using a FastDNA� kit for 40 s atspeed setting 5 in the Fastprep� FG120 instrument(Bio101, Inc., Vista, CA, USA). All DNA isolation proce-dures were performed in accordance with the manufac-turer’s protocol.

A fragment of the rRNA 12S (543 bp) and ND5(804 bp) mitochondrial genes was amplified in all the exam-ples using the primers shown in Table 1. Primer pairs weredesigned employing the software Oligo� v6.88 (Medprobe,Oslo, Norway) from known teleost sequences for these twogenes. Reactions were carried out in 25 ll of reaction vol-

Amplicon (bp)

CGATTAGATACC-3 0 543CATGTTACGAC-30

GGCTGATGGT-3 0 804GGGCTCAGGCGTT-30

AACGTCA-3 0 188GCCGATT-30

AATGGTCGTT-3 0 123GAGGCAG-30

D5�2) and rRNA 12S (RNA12S�1 and RNA12S�2) mitochondrial genes.specific product (SSCO�1 and SSCO�2) and for the universal rRNA 12Sch case.

C. Infante et al. / Food Research International 39 (2006) 1023–1028 1025

ume: 1 ll of DNA template (�30–50 ng) was added to 24 llof PCR mix consisting of 17.25 ll of sterile distilled water,2.5 ll of dNTP mix 10 mM, 2.5 ll of 10· buffer, 1 ll ofMgCl2 50 mM, 0.25 ll (1.25 units) of BioTaq� DNA poly-merase (Bioline, London, UK) plus 0.5 ll of each primer(10 lM). The thermal cycle profile was identical for allthe amplified fragments. An initial denaturation step of96 �C for 2 min was followed by 35 cycles of 96 �C for30 s, 60 �C (rRNA 12S) or 58 �C (ND5) for 30 s, and72 �C for 1 min. A final extension for 10 min at 72 �Cwas also included. PCR products were electrophoresedon a 1% agarose gel and visualized via ultraviolet trans-illu-mination before sequencing.

Double-stranded DNA products were purified using aPCR product purification kit (Marlingen Bioscience,Ijamsville, MD, USA) and subsequently used for directcycle sequencing with a BigDye� Terminator v3.1 kit(Applied Biosystems, Foster City, CA, USA). All sequenc-ing reactions were performed in accordance with the man-ufacturer’s instructions on a 377 DNA sequencer (AppliedBiosystems). Primers used were the same as those used forPCR. The sequences obtained have been deposited in Gen-Bank/EMBL/DDBJ with the following accession numbers:AB241438 to AB241453, and AB242178 to AB242196.

Nucleotide sequences were analyzed using the computerprograms Sequencing Analysis v3.4.1 (Applied Biosystems)and Seqman v5.51 (DNASTAR, Madison, WI, USA),respectively, and further aligned with the Megalign v5.51software (DNASTAR).

2.3. Multiplex PCR and electrophoresis

Reactions were carried out in a 25 ll reaction volumecontaining 1 ll of DNA template, 16.25 ll of sterilized dis-tilled water, 2.5 ll of dNTPs 10 mM, 2.5 ll of 10· buffer,0.75 ll of each of the primers RNA12Spp�1 andRNA12Spp�2 (10 lM), 0.5 lL of each of the S. scom-

brus-specific primers SSCO�1 and SSCO�2 (10 lM), and0.25 ll (1.25 units) of HotMaster� Taq DNA polymerase(Eppendorf, Hamburg, Germany). These primer pairs(Table 1) were also designed using the software Oligo�

v6.88 (Medprobe). Following an initial step of 2 min at96 �C to activate the HotMaster� Taq DNA polymerase,the thermal cycle profile was 40 cycles of denaturation at96 �C for 15 s, and annealing/extension at 70 �C for 15 s.

For analysis, PCR products (10 ll each) were electro-phoresed on a 2.5% Ultrapure� Agarose 1000 (Invitrogen,Carlsbad, CA, USA) gel using 1· TBE buffer, and visual-ized via ultraviolet trans-illumination after staining of thegel with ethidium bromide. The 25 bp DNA Ladder (Invit-rogen) was used as the molecular weight standard.

3. Results and discussion

The aim of this work has been the developing of a mul-tiplex PCR-based method to authenticate processed com-mercial products of S. scombrus. This multiplex-PCR

system is lower-cost and less time-consuming than othermolecular techniques such as sequencing, PCR-RFLP orPCR-SSCP, making it of practical value to all laboratoriesconcerned with the authentication of S. scombrus cannedproducts.

3.1. Sequence data and primers design

For the development of the multiplex PCR-basedmethod of authentication presented in this study, two pri-mer pairs (Table 1), corresponding to highly conservedregions of rRNA 12S and ND5 mitochondrial genes, wereused. PCR fragments of both genes were further sequenced(in all Scomber samples) and aligned for pairwisecomparisons.

It is well known that processes like canning involve athermal treatment for sterilization, in which the DNA isseverely degraded. In fact, only residues of an average sizeranging from 100 to 200 bp are found (Pardo & Perez-Vil-lareal, 2004; Quinteiro et al., 1998). In order to authenti-cate commercially canned products, this restriction hasbeen overcome with the amplification and further analysisof short DNA fragments. Taking into account this issue,ND5 specific primers for S. scombrus were designed(Fig. 1). Positions 290 and 356 were chosen as optimalfor two reasons. Firstly, they allowed for observation ofthe S. scombrus specific nucleotide differences at the critical3 0-end in both primers SSCO�1 (T in S. scombrus, C in theother three Scomber species) and SSCO�2 (C in S. scom-

brus, A in the other Scomber species), respectively. Sec-ondly, these primers showed appropriate thermodynamiccharacteristics as selected by the program Oligo� v6.88(Medprobe). The amplicon expected was 123 bp in length(Table 1).

The existence of PCR inhibition has been reported whendealing with canning products (Pardo & Perez-Villareal,2004; Ram et al., 1996). Consequently, for the multiplex-PCR reaction, a second primer pair was included toamplify a fragment of the highly conserved mitochondrialrRNA 12S gene to provide positive control of DNA ampli-fication (Fig. 2). There were no nucleotide differencesamong sequences with regard to these primers to ensureefficient amplification in all Scomber species. The ampliconexpected was 188 bp in length (Table 1).

3.2. Test development

Once the two primer pairs were designed, the PCR con-ditions were first optimized using fresh fish DNA samples.A two-temperature PCR profile, with extension/annealingtemperature at 70 �C and 15 s of extension time, producedthe best results, enabling completion of PCR in less than anhour. Using these PCR conditions, no amplification of theS. scombrus-specific fragment was observed in any of theother Scomber species (Fig. 3).

The same PCR conditions were then assayed in sixS. scombrus commercially canned products. Two distinct

Fig. 1. Location of the forward (SSCO�1) and reverse (SSCO�2) primers for the amplification of the S. scombrus-specific PCR product. Positions 290 and356, corresponding to the 3 0-end of the primers, are shaded. Dots indicate identity.

Fig. 2. Location of the forward (RNA12Spp�1) and reverse (RNA12Spp�2) primers for the amplification of the universal PCR product. Shading showsthe total coincidence of the rRNA 12S sequences with those of the primers. Dots indicate identity.

Fig. 3. Multiplex-PCR products from fresh DNA samples subjected toS. scombrus authentication. Electrophoresis was performed on a 2.5%Ultrapure� Agarose 1000 (Invitrogen) and stained with ethidiumbromide. Std: Lanes containing the 25 bp DNA Ladder (Invitrogen).Lanes 1–4: S. scombrus; lanes 5–6: S. colias; lanes 7–8: S. japonicus; lanes9–10: S. australasicus. Band sizes of the 25 bp DNA Ladder are indicatedon the left. Sizes of the generated products are shown on the right.

1026 C. Infante et al. / Food Research International 39 (2006) 1023–1028

fillets were analyzed per can. As shown in Fig. 4, all sam-ples tested positive for S. scombrus. To ascertain the exis-tence of false positives, identifications were confirmed byamplification (in a single PCR) and further sequencing ofall the rRNA 12S fragments (not shown).

One of the major problems with techniques like PCR-RFLP and PCR-SSCP is the intra-specific variability, giv-ing rise to possible mutations and genetic variability withina species and, consequently, to the loss or gain of restric-tion recognition sequences (Quinteiro et al., 1998; Sanjuan& Comesana, 2002) and changes in the representative pat-tern of single-stranded DNA (Rehbein et al., 1999), respec-tively. Yet, it should be highlighted that our analyticalmethod would be affected only if the base at the 3 0-endof any of the primers corresponding to the ND5 genewas altered in relation to the template DNA ofS. scombrus. Nevertheless, no intra-specific variation atthese positions has been found in the sequence analysis car-

Fig. 4. Multiplex-PCR products from six different commercial cannedDNA samples (C1 to C6) subjected to S. scombrus authentication. Twodistinct fillets (F1 and F2) were analyzed per can. Electrophoresis wasperformed on a 2.5% Ultrapure� Agarose 1000 (Invitrogen) and stainedwith ethidium bromide. Std: Lanes containing the 25 bp DNA Ladder(Invitrogen). Band sizes of the 25 bp DNA Ladder are indicated on theleft. Sizes of the generated products are shown on the right.

C. Infante et al. / Food Research International 39 (2006) 1023–1028 1027

ried out in this work, neither fresh nor canned samples,despite the fact that the 3 0-ends of both primers are locatedin the third position of a codon, where the majority of thepolymorphism in protein-coding genes is detected. How-ever, the possibility of obtaining false negatives as a conse-quence of intra-specific variability cannot be rejected. Atthis point, we want to note that our protocol constitutesa rapid screening method for detecting positive signals.Samples giving a negative result require S. scombrus

authentication by a different DNA analysis technique ofthe rRNA 12S fragment (sequencing, PCR-RFLP orPCR-SSCP). In this context, similar authentication systemsbased on the species-specific PCR amplification of DNAproducts have been also designed for other commerciallyimportant species (Calvo, Rodellar, Zaragoza, & Osta,2002; Calvo, Zaragoza, & Osta, 2001; Infante, Catanese,Ponce, & Manchado, 2004; Lockley & Bardsley, 2000).

It has been stated that the analytical methods employedin species identification should be rapid and easy to per-form, without being cost prohibitive (Mackie et al.,1999). The protocol presented here fulfils all these criteria.Given the possibility of intraspecific variability, the identi-fication of species by two independent DNA analysis tech-niques is encouraged (Mackie et al., 1999). The assay herepresented constitutes a molecular tool for the detection ofmislabeling or fraudulent substitution of S. scombrus incanned products, and represents a valuable addition tothe range of methodologies currently available for S. scom-

brus identification.

Acknowledgments

We thank Direccion General de Pesca (Consejerıa deAgricultura y Pesca, Junta de Andalucıa), Empresa Publicapara el Desarrollo Agrario y Pesquero de Andalucıa (Dap),and Jose Marıa Naranjo Marquez for their continuous

support to the laboratory. This work has been funded byIFOP (EU).

References

Calvo, J. H., Rodellar, C., Zaragoza, P., & Osta, R. (2002). Beef- andbovine-derived material identification in processed and unprocessedfood and feed by PCR amplification. Journal of Agricultural and Food

Chemistry, 50, 5262–5264.Calvo, J. H., Zaragoza, P., & Osta, R. (2001). Technical note: a quick and

more sensitive method to identify pork in processed and unprocessedfood by PCR amplification of a new specific DNA fragment. Journal of

Animal Science, 79, 2108–2112.Collette, B. B. (2003). Family Scombridae Rafinisque 1815 – mackerels,

tunas, and bonitos. California Academic Sciences, 28. AnnotatedChecklist of Fishes No. 19.

Collette, B. B., Reeb, C., & Block, B. A. (2001). Systematics of the tunasand mackerels (Scombridae). In B. A. Block & E. D. Stevens (Eds.),Tuna: Physiology ecology, and evolution. Fish Physiology (Vol. 19,pp. 1–33). San Diego: Academic Press.

Infante, C., Catanese, G., Ponce, M., & Manchado, M. (2004). Novelmethod for the authentication of frigate tunas (Auxis thazard andAuxis rochei) in commercial canned products. Journal of Agricultural

and Food Chemistry, 52, 7435–7443.Jerome, M., Lemaire, C., Bautista, J. M., Fleurence, J., & Etienne, M.

(2003). Molecular phylogeny and species identification of sardines.Journal of Agricultural and Food Chemistry, 51, 43–50.

Jerome, M., Lemaire, C., Verrez-Bagnis, V., & Etienne, M. (2003). Directsequencing method for species identification of canned sardine andsardine-type products. Journal of Agricultural and Food Chemistry, 51,7326–7332.

Lockley, A. K., & Bardsley, R. G. (2000). Novel method for thediscrimination of tuna (Thunnus thynnus) and bonito (Sarda sarda)DNA. Journal of Agricultural and Food Chemistry, 48, 4463–4468.

Lopez, I., & Pardo, M. A. (2005). Application of relative quantificationTaqMan real-time polymerase chain reaction technology for theidentification and quantification of Thunnus alalunga and Thunnus

albacares. Journal of Agricultural and Food Chemistry, 53, 4554–4560.Mackie, I. M., Pryde, S. E., Gonzales-Sotelo, C., Medina, I., Perez-Martın,

R., Quinteiro, J., et al. (1999). Challenges in the identification ofspecies of canned fish. Trends in Food Science & Technology, 10, 9–14.

Matsui, T. (1967). Review of the mackerel genera Scomber and Rastrel-

liger with description of a new species of Rastrelliger. Copeia, 1967,71–83.

Pardo, M. A., & Perez-Villareal, B. (2004). Identification of commercialcanned tuna species by restriction site analysis of mitochondrial DNAproducts obtained by nested primer PCR. Food Chemistry, 86,143–150.

Quinteiro, J., Sotelo, C. G., Rehbein, H., Pryde, S. E., Medina, I., Perez-Martın, R. I., et al. (1998). Use of mtDNA direct polymerase chainreaction (PCR) sequencing and PCR-restriction fragment lengthpolymorphism methodologies in species identification of canned tuna.Journal of Agricultural and Food Chemistry, 46, 1662–1669.

Ram, J. L., Ram, M. L., & Baidoun, F. F. (1996). Authentication ofcanned tuna and bonito by sequence and restriction site analysis ofpolymerase chain reaction products of mitochondrial DNA. Journal of

Agricultural and Food Chemistry, 44, 2460–2467.Rehbein, H., Mackie, I. M., Pryde, S., Gonzales-Sotelo, C., Medina, I.,

Perez-Martin, R., et al. (1999). Fish species identification in cannedtuna by PCR-SSCP: validation by a collaborative study and investi-gation of intra-species variability of the DNA-patterns. Food Chem-

istry, 64, 263–268.Rehbein, H., Mackie, I. M., Pryde, S., Gonzales-Sotelo, C., Perez-Martin,

R., Quinteiro, J., et al. (1998). Comparison of different methods toproduce single-strand DNA for identification of canned tuna by single-strand conformation polymorphism analysis. Electrophoresis, 19,1381–1384.

1028 C. Infante et al. / Food Research International 39 (2006) 1023–1028

Sanjuan, A., & Comesana, A. S. (2002). Molecular identification of ninecommercial flatfish species by polymerase chain reaction-restrictionfragment length polymorphism analysis of a segment of the cyto-chrome b region. Journal of Food Protection, 65, 1016–1023.

Terol, J., Mascarell, R., Fernandez-Pedrosa, V., & Perez-Alonso, M.(2002). Statistical validation of the identification of tuna species:

bootstrap analysis of mitochondrial DNA sequences. Journal of

Agricultural and Food Chemistry, 50, 963–969.Unseld, M., Beyermann, B., Brandt, P., & Hiesel, R. (1995). Identification

of the species origin of highly processed meat products by mitochon-drial DNA sequences. PCR Methods and Applications, 4, 241–243.