Lab-on-a-Chip PCR-RFLPAssay for the Detection of CanineDNA in Burger Formulations

Md. Mahfujur Rahman & Md. Eaqub Ali &Sharifah Bee Abd Hamid & Subha Bhassu &

Shuhaimi Mustafa & Md. Al Amin & Md. Abdur Razzak

Received: 7 October 2014 /Accepted: 7 January 2015# Springer Science+Business Media New York 2015

Abstract Canine species detection in foods is important inthe perspectives of health, religions, and fare-trade food busi-ness. This study describes a very short-amplicon length Poly-merase Chain Reaction (PCR)-Restriction Fragment LengthPolymorphism (RFLP) assay with lab-on-a-chip detectionplatform for the authentication of canine DNA in processedfoods. A 100-bp fragment of canine mitochondrial Cyto-chrome b (cytb) gene was selected and amplified using a pairof canine-specific primers. The amplified PCR products werevalidated by RFLP analysis using lab-on-a-chip microfluidicbioanalyzer kit. Both gel-image and electropherograms au-thenticated the canine-specific PCR products before (100 bp)and after restriction digestion (51, 30, and 19 bp). The assaysuccessfully detected 0.0001-ng canine DNA under pure stateand 0.01 % (w/w) canine meat spiked in chicken and beefburger formulations. Screening of eight commercial burgersacross Malaysia did not reveal any canine adulteration. Webelieve the assay would find potential applications in foodindustries, Halal food regulatory bodies and animal rightprotection authorities across the globe.

Keywords Burger formulation . Lab-on-a-chip bioanalyzerkit . Microfluidic capillary electrophoresis . PCR-RFLP

Introduction

It is a long-term envision of human civilization that food willbe compliant with health, religions, culture, and age (Ali et al.2014). While elders are concerned of healthy foods, youngeroften run after taste, appearance, and availability (Nam et al.2010). In contrast to the consumer ages, religions often play akey role in controlling the consumption, preparation, process-ing, and purchasing of foods (Bonne and Verbeke 2008; Namet al. 2010). Most religions have food taboos, for example,pork is not allowed to be consumed for the followers of Islamand Judaism (Bonne and Verbeke 2008). In Islam, meats ofthe ritually slaughtered animals with split hoof such as sheep,cattle, buffalo, and goat are allowed, but those of the carni-vores with sharp teeth such as dog and cat are forbidden to beconsumed (Khattak et al. 2011).

The term “Halal” is an Arabic word which defines thepermitted things for the Muslims by the Islamic law drawnfrom the divine book of Quran and the compilation of ProphetMuhammad (Hadith). The “Halal” status of processed foodscannot be verified by consumers using organoleptic test oreven after consumption since the processing treatments sig-nificantly modify the organophysical biomarkers, making thephysical identification extremely difficult (Bonne andVerbeke 2008). Therefore, “Halal” logo on food products aretrusted by the consumers, and it authenticates the halal statusof the food and its ingredients. Due to huge demand (US$700billion annually) and higher price of Halal foods, fraudulentlabelling of halal brands are frequently taking place (Ali et al.

M. M. Rahman :M. E. Ali (*) : S. B. A. Hamid :M. Al Amin :M. A. RazzakNanotechnology and Catalysis Research Centre, University ofMalaya, Kuala Lumpur 50603, Malaysiae-mail: eaqubali@gmail.com

S. BhassuInstitute of Biological Sciences, University of Malaya, KualaLumpur 50603, Malaysia

M. E. Ali : S. BhassuCEBAR Laboratory, IPS Building, University of Malaya, KualaLumpur 50603, Malaysia

S. MustafaInstitute of Halal Products Research, University of Putra Malaysia,Serdang 43400, Malaysia

Food Anal. MethodsDOI 10.1007/s12161-015-0090-1

2012b; Ismail 2014; Rohman et al. 2011). The recent humanmeat scandal in McDonald’s in USA (Olumide 2014); humanfetus consumption as sexual and medicinal drugs in China(Bird 2012), porcine DNA scandal in chocolate products inMalaysia (Chakravorty 2014); horse meat scandal in Europe,and rat meat scandal in China have given scientists in the fielda brainstorming apprehension on what to detect, when todetect, and how to detect (Ali et al. 2014). Thus, meat speciesauthentication has become an important issue and timely needto secure consumers’ trust on labeled brands (Ali et al. 2012c;Rohman et al. 2011). Recently, DNA-based techniques suchas PCR (Rahman et al. 2014), PCR-RFLP (Ali et al. 2012b),and DNA barcoding (Wong and Hanner 2008) have becomethe method of choice because of the higher stability anduniversal availability of the DNA molecule itself (Pereiraet al. 2008). PCR-RFLP is particularly interesting in thisendeavour since it allows the confirmation of the authenticPCR products through the analysis of the restriction-digestedPCR products (Ali et al. 2012b). PCR-RFLP assays withshorter DNA targets are advantageous due to higher availabil-ity and better stability of such biomarkers under extreme foodprocessing conditions. However, hundred nanometres poresize of traditional agarose gel facilitates only separation oflarger DNA targets (100 bp–10 kb) with good resolution(Salieb-Beugelaar et al. 2009). It cannot resolute restriction-digested smaller DNA targets, specially <50 bp. On the otherhand, DNA-based assays with micro-fluidic technologycoupled with a lab-on-a-chip bioanlyzer kit is partially auto-mated and can effectively separate shorter nucleic acid frag-ments ≥10 bp differences in length with good resolution(Ali et al. 2012b). This is an easy and user-friendly analysisapproach integrating automated capillary electrophoresis (CE)in chip, giving higher resolution and speed with better repro-ducibility while using less reagents and samples over thetraditional agarose gel electrophoresis (Funes-Huacca et al.2004).

Despite the religious taboos and huge protection fromanimal right groups, dog meat has been used for humanconsumption in many countries of the world including SouthKorea, China, and Vietnam (Podberscek 2009). Furthermore,availability of stray dogs in certain countries (Kumarapeli andAwerbuch-Friedlander 2009; Totton et al. 2010) has made dogmeat a potential adulterant in costly meats such as beef, lamb,and chicken. Henceforth, several PCR assays for canine spe-cies detection have been documented (Abdel-Rahman et al.2009; Abdulmawjood et al. 2003; Ali et al. 2013; Arslan et al.2006; Gao et al 2004; İlhak and Arslan 2007; Martín et al.2007; Rahman et al. 2014). However, all of these assays arebased on traditional agarose-gel electrophoresis for PCR end-point detection, and none of them have been tested in burgerformulation. Additionally, amplified PCR products have notbeen validated by RFLP analysis. To the best of our knowl-edge, the present study is the first effort to develop a short-

amplicon length (100 bp) PCR-RFLP assay for the separationof the digested fragments using an automated lab-on-a-chipbioanalyzer kit. We tested the assay under complex matricesand validated it in burger formulation.

Materials and Methods

Samples Collection

Raw meat samples of chicken (Gallus gallus), quail (Coturnixcoturnix), sheep (Ovis aries), goat (Capra hircus), beef (Bostaurus), buffalo (Bubalus bubalis), pig (Sus scrofa), cucumber(Cucumis sativus), and tilapia (Oreochromis niloticus) werepurchased in triplicates from the various wet markets in KualaLumpur and Selangor in Malaysia. Dog (Canis lupus) meatsamples were collected from three different dogs, officiallyeuthanized by Dewan Bandaraya Kuala Lumpur (DBKL)located in Taman Air Panas in Kuala Lumpur as a measuredof population control. Commercial samples of eight differentbranded chicken and beef burgers were purchased in tripli-cates from eight different outlets across Malaysia. All sampleswere transported under ice chilled condition (4 °C) and werestored at −20 °C for DNA extraction and future work.

DNA Extraction

DNA was extracted from 25 mg of raw meat samples usingNucleoSpin® Tissue extraction kit (Macherey-Nagel,Germany) according to the manufacturer’s instructions. Foradmixed and burger samples, DNAwas extracted from 1-gmsample using CTAB method following animal tissue extrac-tion protocol and was further purified using Promega Wiz-ard™ DNA isolation kit (Promega Corporation, Madison,USA). The concentration and purity of the DNAs were deter-mined using UV-vis spectrophotometer Biochrom Libra S70(Biochrom Ltd., UK).

Canine-Specific Primer

A pair of canine-specific primers (forward 5′ CCTTACTAGGAGTATGCTTG 3′ and reverse 5′ TGGGTGACTGATGAAAAAG 3′) were designed using on-line version of theprimer3plus software (www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) targeting a 100-bp fragment ofcanine cytb gene (GenBank AF034253.1) which included a30-bp AluI-cut site as internal oligo. First, the canine specific-ity of the primers was theoretically determined using NCBIblast analysis (www.ncbi.nlm.nih.gov). Furthermore, cytbgene sequences of 13 meat-providing animals and fish species(Chicken: EU839454.1, Duck: HQ122601.1, Turkey:HQ122602.1, Quail EU839461.1, Sheep: EU365990.1, Goat:EU130780.1, Beef: EU807948.1, Buffalo: D32193, Pig:

Food Anal. Methods

GU135837.1, Tuna: AM989973.1, Sardine: DQ197989.1, Ti-lapia: AF015020.1, Prawn: AF125382.1) and apocytochrome(cob) gene sequences of 6 plant (Tomato: XM004251454.1,Onion: GU253304.1, Wheat: AF337547.1, Maize: X00789.1,Potato: X58437.1, Cucumber XM004153108.1) were re-trieved from publicly available NCBI database and alignedby ClustalW multiple-sequence alignment tool (Thompsonet al. 1994). The consensus canine-specific 100-bp cytb genesequences were used to study pairwise distance, and adendogram was constructed with molecular evolution andphylogenetic analysis software, MEGA version 5 (Tamuraet al. 2011). Primers, AluI cut site positions, and number ofoligonucleotide mismatch in the primer binding sites wereanalysed by BioEdit software version 7.2 (Hall 2004). A 3Dplot was created using pairwise distances of 100-bp caninecytb gene-specific site and number of oligonucleotide mis-matches by XLSTAT software (Addinsoft 2013) to define thepotentiality of the primers for canine species detection. Forinternal control, a previously designed eukaryotic primer pairs(forward 5′ GGT AGT GAC GAA AAATAA CAA 3′ andreverse 5′ ATA CGC TAT TGG AGC TGG AATTA C CTACAGG AC3′) targeting 141-bp conserved fragment of the eu-karyotic 18S rRNA gene were used (Rojas et al. 2010). Allprimers were obtained from the 1st BASE Laboratories, Pte.Ltd. in Selangor, Malaysia and were tested in real-PCR run.

PCR Amplification

We amplified 100-bp canine and 141-bp eukaryotic targetsusing a PCR assay of 20 μl reaction volume composed of 5×colourless GoTaq Flexi buffer supplied with the enzyme,1.25 μ GoTaq Flexi DNA Polymerase, 1.25 mM MgCl2,200 μM dNTP’s (Promega, Madison, USA), 0.4 μM eachprimer, and 20 ng of total DNA extracted from each sample.The cycling conditions used in a gradient thermocycler(Eppendorf, Germany) were initial denaturation at 94 °C for3 min followed by 30 cycles of denaturation at 94 °C for 30 s,annealing at 58 °C for 30 s, and extension at 72 °C for 1 min.The final extension was performed at 72 °C for 5 min. PCRproducts were detected on a microfluidic-based lab-on-a- chipusing Experion DNA 1 K analysis kit following manufac-turer’s instructions (Bio-Rad Laboratories, USA).

Preparation of Binary Admixtures

To evaluate the performance of the PCR assay in meat prod-ucts, binary admixtures of dog-chicken and dog-beef wereprepared by spiking 10, 1, 0.1, and 0.01 % dog meat in chickenand beef in a 100-g specimen (Ali et al. 2012b). Thus, preparedadmixtures were blended vigorously to make a homogeneousadmixture. All admixtures were prepared on 3 different days by3 independent analysts and were autoclaved at 120 °C under45-psi pressure for 2.5 h before extracting DNA.

Burger Preparation

To simulate commercial burger, model raw, autoclaved, andready to eat burgers were prepared using dog, chicken, andbeef meat following Ali et al. (2012a). For pure burger, 100-gspecimen of minced dog, chicken, and beef meats were addedto 1.2-g finely chopped tomato, 0.2-g onion, 0.2-g egg, 1-gcumin seed, and 0.25-g cayenne pepper and mixed well. Eachseparate mixture was given to a burger shape. To simulate dogmeat adulteration, chicken and beef burgers were prepared byspiking 10, 1, 0.1, and 0.01 % of dog meat into chicken andbeef meats, mixed and minced well. The raw burgers wereautoclaved at 133 °C at 45 psi for 20 min according toEuropean legislation (Commission 2002). The raw burgersmeat were grilled on both sides in an electrical oven at220 °C for 15 min and mixed with tomato, cucumber, andbuns to simulate ready to eat burger formulation. All burgerswere kept at −20 °C and were blended well into a homogenousmixture prior to DNA extraction.

RFLPAnalysis

For RFLP analysis, PCR products were digested with AluIrestriction enzyme in a 30-μl reaction mixture containing10 μl of PCR product, 1 μl of enzyme (1 FDU), 17 μl ofdistilled water, and 2 μl of 10× digestion buffer supplied withthe enzyme (New England Biolab, USA). Firstly, the mixtureswere gently mixed and spin downed and then incubated at37 °C in a water thermostat for 15 min for digestion. After15 min of digestion, the enzyme was inactivated by heatingthe mixture at 65 °C for 5 min. Finally, RFLP analysis wasperformed by running 1 μl of the restriction digested productsof each sample in Experion™ lab-on-a-chip well using 1 KDNA analysis kit (Bio-Rad Laboratories, USA).

Results and Discussion

Canine Species Specificity

Species-specific PCR assay is a commonly used technique forspecies authentication in raw and processed foods (Ali et al.2013; Mane et al. 2009; Rahman et al. 2014). The presence ofa single mismatch in the primer binding site may reduce theefficiency of PCR assay or may lead to PCR amplificationfailure (Wu et al. 2009). Therefore, calculation of oligonucle-otide mismatch is one of the key factors to be consideredwhiledesigning species-specific primers. Designing primers withperfect matching with the specific target and multiple-mismatches with non-target species would definitely increasethe specificity of the primer, decreasing chances of non-targetamplification.

Food Anal. Methods

Blast analysis against non-redundant nucleotide se-quences using NCBI database revealed that the primerswere 100 % specific for the canine cytb gene. Sequencealignment test against cytb/cob gene sequences of total 20potential animal and plant species for burger formulationusing ClustalW multiple sequence alignment program dem-onstrated 100 % similarity with the canine cytb gene andmultiple mismatches with those from other species(Fig. 1b). Pairwise distances for 100-bp canine-specificsites computed by the maximum composite likelihoodmethod (Tamura et al. 2004) ranged from 0.26 to 2.15.The lowest distance was between dog and sheep (0.26)and the highest was between dog and tomato (2.15). Con-struction of dendrogram also revealed the maximum se-quence similarity of dog with sheep (Fig. 1a). The numberof oligonucleotide mismatches of all other species exceptdog in the primer binding sites was 5–11 (Fig. 1c). Thus, a3D plot using pairwise distance and number of primermismatches reflected clear discrimination of the canine

primers from all other species including sheep (Fig. 1c). Areal-PCR analysis was done against the total DNA of 7commonly used meat animals (chicken, quail, sheep, goat,beef, buffalo, and pig), 1 plant (cucumber), and 1 fish(tilapia) species. These species demonstrated minimumnumber of oligonucleotide mismatches in theoretical anal-ysis (Fig. 1c). The assay amplified only 100-bp canine cytbgene target at an optimized annealing temperature of 58 °Cand a primer concentration of 0.4 μM (Fig. 2a). Further-more, a universal 141-bp fragment of eukaryotic 18S rRNAgene target was amplified from all species (Fig. 2a), dem-onstrating both the specificity of the canine primers andeminence of good quality DNA from all species used in thisstudy (Rojas et al. 2010).

Previously proposed canine-specific PCR assays based onmitochondrial whole genome (322 bp) (İlhak and Arslan2007), mitochondrial cytb (808 bp) (Abdel-Rahman et al.2009), and D-loop (213 bp) gene (Gao et al. 2004) used largeramplicon size (≥213 bp) which may break down under

Fig. 1 In-silico analysis of thecanine specific primers. In a,Dendogram built from the 100-bpregions of cytb/cob-genesequences of dog and other 19animals, plant, and fish speciesusing neighbor-joiningmethod. In b,Mismatch bases ofstudied species with caninespecific primers and the positionsof AluI restriction sites. In c,Analysis of the primer mismatchand pairwise distance using 3Dplot

Food Anal. Methods

commercial food processing conditions, causing failure ortruncated PCR products. We used here shorter DNA target(100 bp) and found that it was stable under various foodprocessing treatments (Ali et al. 2013; Rahman et al. 2014).

Assay Sensitivity

The 100-bp canine-specific site was tested to determine thesensitivity of the assay using 10-fold serial dilutions (100, 10,1, 0.1, 0.01, 0.001, 0.0001, and 0.00001 ng) of the extractedcanine DNAwith distilled water. Clear gel-bands (Fig. 2b) andelectropherograms (data not shown) for canine-specific PCRproducts were observed with as low as 0.0001 ng of canineDNA extracted from raw meat. Thus, the detection limit(LOD) of the assay was determined to be 0.0001-ng canineDNA (Fig. 2b) which was equivalent to an automated andexpensive real-time PCR (Yusop et al. 2012).

The high sensitivity of this PCR assay might be due toshorter amplicon-length (100 bp). Rodríguez et al. (2005)obtained 0.01 ng sensitivity for 411-bp fragment of por-cine mt-12S rRNA gene. On the other hand, using a

shorter amplicon (119 bp) and molecular beacon real-time PCR, Yusop et al. (2012) recognized 0.0001 ngLOD for porcine mt-cytb gene. Ali et al. (2012b) alsodocumented 0.0001 ng LOD for a 109-bp site of porcinecytb gene using conventional PCR coupled to themicrofluidic bioanalyzer chip. Thus, the 0.0001 ng LODfor a 100-bp target of canine cytb gene using lab-on-achip technique for PCR end-point detection was accept-able and the assay would be useful to determine a traceamount of canine DNA in processed foods.

Binary Admix Analysis

To simulate commonly practiced meat adulteration, we pre-pared two sets of admixtures following Ali et al. (2012b). Set1 was dog-chicken and set 2 was dog-beef binary admixturescontaining various percentages (10–0.01 %) of spiked dogmeat. We subjected sets 1 and 2 to extensive autoclaving(2.5 h) to test the stability of the target under extensive heatand pressure treatments. We obtained canine-specific (100 bp)PCR products from all chicken and beef admixed containingas low as 0.01% (w/w) (20 ng total and 0.002 ng canine DNA)of spiked dog meat (Fig. 3a). An endogenous 141-bp eukary-otic targets were amplified from all admixed, reflecting goodquality DNA in all admixtures (Fig. 3a).

For canine-species detection under binary matrix, Martínet al. (2007) documented a species-specific PCR assaytargeting 101-bp fragment of 12S rRNA gene in meat-oatsbinary admixtures under normal autoclaving condition. Thesensitivity of that assay was 0.1 % (w/w) (about 0.125 ngDNA) for dog meat detection. The assay was less sensitivethan the present one in terms of the amount of template DNAused (125 vs. 20 ng) and the limit of detection (0.1 vs 0.01%).Recently for dog meat detection, Ali et al. (2013) documenteda cytb-based PCR assay using agarose gel end-point detectiontechnique with a sensitivity of 0.1 % (w/w) (0.02 ng DNA).Compared to this, CE-based end-point detection scheme ofthe present assay provided a higher sensitivity of 0.01% (w/w)(0.002 ng canine DNA) using similar admixtures of dog meatwith chicken and beef.

Model Burger Analysis

Food adulteration is often performed in minced meat productssuch as burgers (Ali et al. 2012a), sausages, and meatballs.Among these items, burgers, especially the beef and chickenburgers, are very popular in all continents of the world (Aliet al. 2012a). Recent documentation on rat and horse meatscandal in China and Europe (Ali et al. 2014) has furtherprompted us to investigate whether canine meat is beingsubstituted in popular burger formulations. Therefore, wetested the performance of the PCR assay using lab-on-a-chipend-point detection, under complex background of minced

Fig. 2 Specificity (a) and sensitivity (b) of canine specific assay underpure state. In a, 100-bp canine PCR product (lane 1) and 141-bp eukary-otic endogenous controls (lanes 1–10) from different species aredisplayed. Lanes 1–10 demonstrate PCR products from dog, chicken,quail, sheep, goat, beef, buffalo, pig, cucumber, and tilapia, respectively.In b, lanes 1–7 represent canine PCR products from 100, 10, 1, 0.1, 0.01,0.001, 0.0001, and 0.00001 ng of canine templates. Lane 7 defines thatthe limit of detection is 0.0001 ng genomic DNA from pure canine meatunder raw state. Lane L, DNA ladder and lane N, negative control

Food Anal. Methods

meat products in pure as well as in deliberately contaminatedformat by spiking different percentages of dog meat (0.01 to10 %) with chicken and beef burger. We found that the assaycan detect as low as 0.01% of caninemeat contamination bothin chicken and beef burgers (Fig. 3b).

Previously, Ali et al. (2012b) reported a LOD 0.01 % forternary admixtures composed of pork, chicken, and wheatusing microfluidic detection method for a 109-bp site ofporcine cytb gene. For burger analysis using the same target(109 bp) and real-time PCR assay, Ali et al. (2012a) obtained0.01 % LOD. Recently, 0.1 % (w/w) and 0.2 % (w/w) LODhave been documented for canine species detection in frank-furter (Ali et al. 2013) and meatball (Rahman et al. 2014)formulations, respectively, using conventional PCR andagarose-gel electrophoresis with a short target of cytb gene.Thus, a 0.01 % LOD for canine-specific target in burgerformulation using a highly sensitive microfluidic lab-on-a-chip detection technique and shorter (100 bp) target was anexpected and logical outcome.

Burger Authentication by RFLP

RFLP sites in 100-bp target of canine cytb gene wassearched by using NEB cutter version 2.0 and two cutsites (Fig. 1b) for AluI restriction enzyme were obtainedwith 3 fragments (19, 30, and 51 bp). The AluI digestedPCR products of raw and autoclaved meats along withraw, autoclaved, and ready to eat dog burgers were sepa-rated by lab-on-a-chip microfluidic separation technologyand two clear fragments of 30 and 51 bp were visualizedboth in the gel-image and electropherograms (Fig. 4). Thegel-bands and electropherograms for the19 bp fragmentwere merged with the 15-bp lower marker and wereappeared as a thicker band (Fig. 4), since the techniquehave the limitation to resolute ≤5-bp difference in frag-ment length. The molecular size statistics of the 100-bpcanine-specific site from raw and autoclaved meat alongwith raw, autoclaved, and ready to eat model dog burgersare given in Table 1.

Conventional PCR is an essential tool for routine analysisin bioanalytical, clinical, and research laboratories. However,certain difficult-to-control features of the amplification pro-cess of the PCR have yet to be optimized (Ali et al. 2012b;

Fig. 3 Specificity and sensitivity analysis in admixed meats (a) andmodel burgers (b). Shown are in a, lanes 2–5: 10, 1, 0.1, and 0.01 %canine meat admixed with chicken meat and lanes 7–10: 10, 1, 0.1, and0.01% canine meat admixed with beef meat. Lane 1 and 6: 100% chickenand 100% beef, respectively. In b, lanes 1–4: 10, 1, 0.1, and 0.01% caninemeat spiked chicken burger and lanes 6–9: 10, 1, 0.1, and 0.01% dogmeatspiked beef burger. Lanes 5 and 10: 100 % chicken and beef burgers,respectively. Lane L: DNA ladder. The 100- and 141-bp PCR productsrepresent canine and eukaryotic endogenous control, respectively

Fig. 4 PCR-RFLP analyses, showing 100-bp PCR and AluI restrictiondigestion products obtained from raw, autoclaved dog meat alongwith raw, autoclaved, and ready to eat model dog burger. In gel image,lane L: DNA ladder; lanes 1, 3, 5, 7, and 9 before restriction digestion and

lanes 2, 4, 6, 8, and 10 after restriction digestion of PCR product obtainedfrom raw, autoclaved meat; and raw, autoclaved and ready to eat modeldog burger, respectively. Corresponding electropherograms are demon-strated with labels in insets on the right

Food Anal. Methods

Yang et al. 2005). These features generally may derive fromthe artifacts of simultaneous amplification of small contami-nants at high magnitudes and non-specific target amplification(Doosti et al. 2011; Yang et al. 2005). PCR product architec-ture with restriction site can unambiguously differentiate theoriginal targets. Thus, both in-silico digestion and real RFLPanalysis of 100-bp PCR product digested by AluI restrictionenzyme yielded three canine-specific fragments, reducing theprobability of misleading results by non-specific contaminantamplification. Quantitative real-time PCR assay with automat-ed detection technique by real-time monitoring of the ampli-fication cycle is interesting, but it often produces artifacts andincurs excessive costs (Bustin et al. 2009) in terms of instru-ment, reagents, and probes. On the other hand, PCR-RFLPanalysis produces enhanced specificity of the PCR assay witha unique restriction pattern from each species of animals (Aidaet al. 2005; Ali et al. 2012b; Doosti et al. 2011; Sait et al.2011). Previously, Abdulmawjood et al. (2003) documented aPCR-RFLP assay of 808-bp amplicon length target for caninespecies detection. Such a large-size amplicon had the possi-bility of breaking down during food processing, causing PCRfailure while testing the commercial foods. Thus, the advan-tage of our PCR-RFLP assay is easily understandable in termsof sensitivity, amplicon length, and stability.

Commercial Burger Analysis

In the food industry, replacement of costly meats by cheaperproducts is relatively common in order to reduce cost andincrease profit. Therefore, we screened here commercial bur-ger samples using 0.01 % dog meat spiked dummy burgers asa positive control. Total eight different “Halal” branded

chicken (A–D) and beef burgers (A′–D′) were purchased fromdifferent Malaysian outlets and were tested. While the caninePCR product was obtained from all positive controls, nocommercial burgers collected from different outlets werefound to be positive for dog meat (Table 2), reflecting theabsence of dog-meat adulteration in burger formulations inMalaysia. Amplification of endogenous eukaryotic controlreflected good quality DNA in all commercial products. Thefindings are acceptable in Malaysian perspectives since thecountry is committed to develop Halal-hub industry and strict-ly monitoring the Halal status of foods.

The usual Halal burger formulation is composed ofchopped meat of chicken, beef, lamb or fish along with starch,seasonings, and salt. Replacement of dog meat in the com-mercial burger may add an extra profit. Species-specific PCRis one of the widely used methods for meat species authenti-cation. To detect fraudulent admixing of canine meat, a num-ber of species-specific assays have been reported (Abdel-Rahman et al. 2009; Abdulmawjood et al. 2003; Gao et al.2004; Martín et al. 2007). However, none of these assays weretested in the commercial burger background, while reportshave been made about the PCR inhibition possibilities incommercial or process food due to numerous factors originat-ed from multiple ingredients (Bottero et al. 2002; Di Pintoet al. 2005). Recently, canine-specific PCR assay was testedfor commercial frankfurter and meatball with a positive con-trol of 0.1 % and 0.2 % (w/w) deliberately spiked dog meatsample using agarose gel end-point detection (Ali et al. 2013;Rahman et al. 2014). For the first time, we documented here aPCR-RFLP assay with lab-on-chip microfluidics-based en-hanced detection platform and obtained extremely highsensitivity of 0.01 % (w/w).

Table 1 Molecular size statistics of lab-on-a chip based canine specific PCR-RFLP assay

Size (bp) Raw dog meat withAluI digestion

Autoclaved dog meatwith AluI digestion

Raw dog burger withAluI digestion

Autoclaved dog burgerwith AluI digestion

Ready to eat dog burgerwith AluI digestion

Before After Before After Before After Before After Before After

100 101±0.5 – 100±1.2 – 100±0.8 – 100±0.5 – 100±0.8 –

51 – 52±0.5 – 51±0.5 – 51±0.0 – 51±0.5 – 51±0.5

30 – 30±0.5 – 29±0.8 – 30±0.5 – 30±0.8 – 29±0.9

19 – – – – – – – – – –

Table 2 Analysis of canine species in commercial chicken and beef burgers

Chicken burger No. of sample ≥0.01 % detection Beef burger No. of sample ≥0.01 % detection Replicate Detection probability

A 3 0/9 A′ 3 0/9 3 100 %

B 3 0/9 B′ 3 0/9 3 100 %

C 3 0/9 C′ 3 0/9 3 100 %

D 3 0/9 D′ 3 0/9 3 100 %

Positive control 3 9/9 Positive control 3 9/9 3 100 %

Food Anal. Methods

Conclusion

Analytical method described here utilized both canine-specific PCR and RFLP analysis coupled with enhancedmicrofluidics-based lab-on-a-chip detection platform to deter-mine dogmeat adulteration in raw, autoclaved as well as readyto eat processed burgers. The technique combined the use ofcanine-specific primers together with eukaryotic endogenouscontrol to probe the quality of DNA used in each experiment.The efficiency and sensitivity of the assay were tested in raw,admixed and commercial burger backgrounds and high sensi-tivity (0.0001 ng DNA for raw pure meats and 0.01% (w/w or0.002 ng DNA for binary admixtures and beef and chickenburgers) were obtained. Additionally, the canine-specific PCRproduct was confirmed by RFLP analysis, eliminating confu-sion about the authentic target. The assay was validated inburger formulation. The screening of Halal-branded commer-cial burgers did not show any positive results for canine meatadulteration in Malaysian outlets.

Acknowledgments This research was jointly supported by Universityof Malaya research grant no. GC001A-14SBS to M.E. Ali and grant no:RU002-2014 to S.B.A. Hamid. M.M. Rahman is a recipient ofMalaysianInternational Scholarship no. KPT.B.600-18/3 Vol2 133.

Conflict of Interest Md. Eaqub Ali declares that he received fundingsupport from the University of Malaya, supervised this work, and editedmanuscript, and he has no conflict of interest to publish this paper. Md.Mahfujur Rahman declares that he performed this work under the super-vision of Md. Eaqub Ali, and he has no conflict of interest to publish thispaper. Sharifah Bee Abd Hamid declares that she paid salary to Md.Mahfujur Rahman from her research grant, and she has no conflict ofinterest to publish this paper. Subha Bhassu declares that she providedsome reagents to Md. Mahfujur Rahman, and she has no conflict ofinterest to publish this paper. Shuhaimi Mustafa declares that he ensuredlab facilities for this work, and he has no conflict of interest to publish thispaper. Md. Al Amin declares that he helped Md. Mahfujur Rahman insample collection, and he has no conflict of interest to publish this paper.Md. Abdur Razzak declares that he helped Md. Mahfujur Rahman inexperiments, and he has no conflict of interest to publish this paper.

Compliance with Ethics Requirements Ethical clearance of ref. no:NANOCAT /25/04/3013/ MMR (R) was obtained from the InstitutionalAnimal Care and Use Committee, University of Malaya (UM IACUC),and all experiments were conducted and animal meats were handledfollowing the national and institutional guideline.

References

Abdel-Rahman S, El-SaadaniM, Ashry K,HaggagA (2009)Detection ofadulteration and identification of cat’s, dog’s, donkey’s and horse’smeat using species-specific PCR and PCR-RFLP techniques. Aust JBasic Appl Sci 3:1716

Abdulmawjood A, Schönenbrücher H, Bülte M (2003) Development of apolymerase chain reaction system for the detection of dog and catmeat in meat mixtures and animal feed. J Food Sci 68:1757

Addinsoft S (2013) XLSTAT-PLSPM. Paris, France

Aida AA, CheMan YB,Wong CMVL, Raha AR, Son R (2005) Analysisof raw meats and fats of pigs using polymerase chain reaction forHalal authentication. Meat Sci 69:47–52

Ali ME, Hashim U, Dhahi TS, Mustafa S, Man Y, Latif MA (2012a)Analysis of pork adulteration in commercial burgers targetingporcine-specific mitochondrial cytochrome B gene by TaqManprobe real-time polymerase chain reaction. Food Anal Methods 5:784–794

Ali ME, Hashim U, Mustafa S, Che Man YB (2012b) Swine-specificPCR-RFLP assay targeting mitochondrial cytochrome B gene forsemiquantitative detection of pork in commercial meat products.Food Anal Methods 5:613–623

AliME,KashifM,Uddin K,HashimU,Mustafa S, CheManYB (2012c)Species authentication methods in foods and feeds: the present, past,and future of Halal Forensics. Food Anal Methods 5:935–955

Ali ME, Rahman MM, Hamid SBA, Mustafa S, Bhassu S, Hashim U(2013) Canine-specific PCR assay targeting cytochrome b gene forthe detection of dog meat adulteration in commercial Frankfurters.Food Anal Methods 1–8

Ali ME, Razzak MA, Hamid SBA (2014) Multiplex PCR in speciesauthentication: probability and prospects—a review. Food AnalMethods 1–17

Arslan A, Ilhak OI, Calicioglu M (2006) Effect of method of cooking onidentification of heat processed beef using polymerase chain reac-tion (PCR) technique. Meat Sci 72:326–330

Bird S (2012) A truly monstrous medicine: the mail investigates China’sgruesome human baby flesh pills.Mail online. Retrieved from http://www.dailymail.co.uk/news/article-2143294/A-truly-monstrous-medicine-The-Mail-investigates-Chinas-gruesome-human-baby-flesh-pills.html

Bonne K, Verbeke W (2008) Religious values informing Halal meatproduction and the control and delivery of halal credence quality.Agric Hum Values 25:35–47

Bottero M, Civera T, Anastasio A, Turi R, Rosati S (2002) Identificationof cows milk in “buffalo” cheese by duplex polymerase chainreaction. J Food Prot 65:362–366

Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M,Mueller R, Nolan T, Pfaffl MW, Shipley GL (2009) The MIQEguidelines: minimum information for publication of quantitativereal-time PCR experiments. Clin Chem 55:611–622

Chakravorty S (2014) Pork DNA found in two chocolate products ofCadbury Malaysia: Report. Reuters. Retrived from: http://www.reuters.com/article/2014/05/26/us-mondelez-intl-recall-idUSBREA4P0CH20140526

Commission E (2002) Commission Decision 2002/657/EC of 12 August2002 implementing Council Directive 96/23/EC concerning theperformance of analytical methods and the interpretation of results.Off J Eur Communities L 221:8–36

Di Pinto A, Forte VT, Conversano MC, Tantillo GM (2005) Duplexpolymerase chain reaction for detection of pork meat in horse meatfresh sausages from Italian retail sources. Food Control 16:391–394

Doosti A, Ghasemi Dehkordi P, Rahimi E (2011) Molecular assay tofraud identification of meat products. J Food Sci Technol 1–5

Funes-Huacca M, de Almeida C, Regitano L, Mueller O, Carrilho E(2004) Semiquantitative determination of Alicyclobacillusacidoterrestris in orange juice by reverse-transcriptase polymerasechain reaction and capillary electrophoresis—laser induced fluores-cence using microchip technology. Electrophoresis 25:3860–3864

Gao H-W, Xu B-L, Liang C-Z, Zhang Y-B, Zhu L-H (2004) Polymerasechain reaction method to detect canis materials by amplification ofspecies-specific DNA fragment. J OAC Int 87:1195–1199

Hall T (2004) BioEdit version 7.0. 0. Distributed by the author, website:www.mbio.ncsu.edu/BioEdit/bioedit.html

İlhak Oİ, Arslan A (2007) Identification of meat species by polymerasechain reaction (PCR) technique. Turk J Vet Anim Sci 31:159–163

Food Anal. Methods

Ismail ZI (2014) Islamic capital mart a viable alternative. New StraitsTime. Retrived from http://www.nst.com.my/node/29680

Khattak JZK, Mir A, Anwar Z, Abbas G, Khattak HZK, Ismatullah H(2011) Concept of Halal food and biotechnology. Adv J Food SciTech 3

Kumarapeli V, Awerbuch-Friedlander T (2009) Human rabies focusingon dog ecology—a challenge to public health in Sri Lanka. ActaTrop 112:33–37

Mane BG, Mendiratta SK, Tiwari AK (2009) Polymerase chain reactionassay for identification of chicken in meat and meat products. FoodChem 116:806–810

Martín I, García T, Fajardo V, Rojas M, Hernández P, González I, MartínR (2007) Technical note: detection of cat, dog, and rat or mousetissues in food and animal feed using species-specific polymerasechain reaction. J Anim Sci 85:2734–2739

NamK-C, Jo C, Lee M (2010) Meat products and consumption culture inthe East. Meat Sci 86:95–102

Olumide (2014) Human meat discovered in McDonald’s meat factory.Informationnigeriaorg Retrived from http://www.informationng.com/2014/08/human-meat-discovered-in-mcdonalds-meat-factory.html

Pereira F, Carneiro J, Amorim A (2008) Identification of species withDNA-based technology: current progress and challenges. RecentPatents DNA Gene Seq 2:187–200

Podberscek AL (2009) Good to pet and eat: the keeping and consumingof dogs and cats in South Korea. J Soc Issues 65:615

Rahman MM, Ali ME, Hamid SBA, Mustafa S, Hashim U, Hanapi UK(2014) Polymerase chain reaction assay targeting cytochrome bgene for the detection of dog meat adulteration in meatball formu-lation. Meat Sci 97:404–409

Rodríguez MA, García T, González I, Hernández PE, Martín R (2005)TaqMan real-time PCR for the detection and quantitation of pork inmeat mixtures. Meat Sci 70:113–120

RohmanA, Sismindari, Erwanto Y, CheManYB (2011) Analysis of porkadulteration in beef meatball using Fourier transform infrared(FTIR) spectroscopy. Meat Sci 88:91–95

Rojas M, González I, Pavón MÁ, Pegels N, Lago A, Hernández PE,García T, Martín R (2010) Novel TaqMan real-time polymerasechain reaction assay for verifying the authenticity of meat and

commercial meat products from game birds. Food Addit Contam27:749

Sait M, Clark EM,Wheelhouse N, Spalding L, LivingstoneM, Sachse K,Markey BK, Magnino S, Siarkou VI, Vretou E, Caro MR, Yaga R,Lainson FA, Smith DGE, Wright F, Longbottom D (2011) Geneticvariability of Chlamydophila abortus strains assessed by PCR-RFLPanalysis of polymorphic membrane protein-encoding genes. VetMicrobiol 151:284–290

Salieb-Beugelaar G, Dorfman K, Van den Berg A, Eijkel J (2009)Electrophoretic separation of DNA in gels and nanostructures. LabChip 9:2508–2523

Tamura K, Nei M, Kumar S (2004) Prospects for inferring very largephylogenies by using the neighbor-joining method. Proc Natl AcadSci U S A 101:11030–11035

Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011)MEGA5: molecular evolutionary genetics analysis using maximumlikelihood, evolutionary distance, and maximum parsimonymethods. Mol Biol Evol 28:2731–2739

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTALW: improvingthe sensitivity of progressive multiple sequence alignment throughsequence weighting, position-specific gap penalties and weight ma-trix choice. Nucleic Acids Res 22:4673

Totton SC, Wandeler AI, Zinsstag J, Bauch CT, Ribble CS, Rosatte RC,McEwen SA (2010) Stray dog population demographics in Jodhpur,India following a population control/rabies vaccination program.Prev Vet Med 97:51–57

Wong EH-K, Hanner RH (2008) DNA barcoding detects market substi-tution in North American seafood. Food Res Int 41:828–837

Wu JH, Hong PY, Liu WT (2009) Quantitative effects of position andtype of single mismatch on single base primer extension. JMicrobiol Meth 77:267–275

Yang I, Kim Y-H, Byun J-Y, Park S-R (2005) Use of multiplex polymer-ase chain reactions to indicate the accuracy of the annealing tem-perature of thermal cycling. Anal Biochem 338:192–200

Yusop M, Mustafa S, Che Man Y, Omar A, Mokhtar N (2012) Detectionof raw pork targeting porcine-specific mitochondrial cytochrome Bgene by molecular beacon probe real-time polymerase chain reac-tion. Food Anal Methods 5:422–429

Food Anal. Methods