Molecular and Cellular Probes (2002) 16, 307±314doi:10.1006/mcpr.2002.0428, available online at http://www.idealibrary.com on 1

Application of a convenient DNA extraction method and

multiplex PCR for the direct detection of Staphylococcus

aureus and Yersinia enterocolitica in milk samples

A. Ramesh, B. P. Padmapriya, A. Chandrashekar and M. C. Varadaraj*

Department of Food Microbiology, Central Food Technological Research Institute,Mysore 570 013, India

(Received 4 January 2002; accepted for publication 22 May 2002)

The application of PCR for the direct and sensitive detection of food-borne pathogens is largely affectedby the quality of the template DNA prepared from food samples. In the present study, a chemicalextraction method of bacterial DNA from spiked milk samples for the direct detection ofStaphylococcus aureus and Yersinia enterocolitica was evaluated by PCR. Gene speci®c primerswere designed to target the nuclease (nuc) and the attachment invasion locus (ail ) genes of S. aureusand Y. enterocolitica, respectively and used in PCR. A combination of organic solvents, detergents andalkali in the DNA extraction method permitted a detection limit of 10 cfu mlÿ1 milk for both thepathogens. When equal numbers of S. aureus and Y. enterocolitica were spiked in milk samples, theindividual detection limit was determined to be 103 cfu mlÿ1 milk. Simultaneous ampli®cation of 482and 359 bp fragments of the nuc and ail genes was obtained using the primer pairs in a single reaction.Multiplex PCR enabled the detection of 104 cfu mlÿ1 milk of S. aureus and Y. enterocolitica withoutany pre-enrichment step. A combination of conventional isolation technique and PCR using DNAextracted by the proposed method was used to test raw milk samples for possible contamination withS. aureus and Y. enterocolitica. The presence of S. aureus in the tested samples was indicated by boththe methods while Y. enterocolitica could not be detected in any of the samples. The template DNAextraction method developed in this study is rapid, sensitive and avoids interference from potential PCRinhibitors and demonstrates the potential of detecting multiple pathogens in milk samples without anyenrichment. # 2002 Elsevier Science Ltd. All rights reserved.

KEYWORDS: PCR, Staphylococcus aureus, Yersinia enterocolitica, multiplex PCR, detection, milk.

INTRODUCTION

The development of rapid, sensitive and infalliblemethods of detecting food-borne pathogens hasreceived much impetus in the recent years owing toan increased public awareness of the health hazardsassociated with microbial contamination of food.Conventional methods of detecting bacteria in foodcomprises of propagation in selective enrichmentmedia followed by microbiological and biochemical

* Author to whom all correspondence should be addressed: Tel: �91-82

0890±8508/02/$ ± see front matter

tests, which are cumbersome and time-consuming.The advent of nucleic-acid based assay systems likethe polymerase chain reaction (PCR) has led to theemergence of improved, expedient and reliablemethods of microbial identi®cation and surveillance,capable of even detecting nonviable cells.1,2

The direct detection of pathogenic bacteria in foodsamples is a challenging task, hampered by thepresence of PCR-inhibitory substances frequentlyassociated with enrichment media, DNA isolation

1-571539; Fax: �91-821-517233; E-mail: mcvraj14@yahoo.com

# 2002 Elsevier Science Ltd. All rights reserved.

A. Ramesh et al.308

reagents and the food matrix itself and additionallycompounded by the presence of high numbers ofindigenous micro¯ora.3,4 Consequently, there is aneed to develop a sample preparation strategy thatcan effectively sequester the pathogenic bacteria orthe target DNA from the food sample. In order toachieve this improved sample preparation methodshave been designed based on enzymatic treatment,buoyant density centrifugation, DNA af®nity puri-®cation columns and magnetic beads coated withspeci®c antibodies or lectins.5±7

Amongst the food samples, milk has been shown tobe a natural reservoir for a number of pathogenicstrains and thus represents a signi®cant health risk.8

A number of PCR-based investigations for milksamples have aimed at tracing a single targetorganism, with special reference to pathogens likeListeria monocytogenes.9±11 However, an integratedapproach for the detection of multiple pathogenicstrains present in milk is desirable. This wouldyield adequate information about the microbialquality of milk and ensure public health.

Enterotoxigenic strains of Staphylococcus aureushave been reported to cause a number of milk-bornediseases.12 The proximity of dairy cows and theirhandlers coupled with poor conditions of sanitationand hygiene contributes to the prevalence of S. aureusin milk. Emerging pathogens like Yersinia entero-colitica are capable of surviving and growing in milkstored at low temperatures13 and have been asso-ciated with outbreaks of milk-borne diseases.14,15 Inthe light of the serious health-related and economicimplications caused by both these pathogenic strainsthe objective of the present study was as follows: (i) todevelop a simple and rapid method of preparingbacterial DNA templates from milk samples arti®-cially inoculated with S. aureus and Y. enterocolitica;(ii) use PCR to detect these pathogenic strains directlyin milk samples obviating a pre-enrichment step;(iii) assess the sensitivity of the developed protocol todetect these bacterial strains individually and inmixed cultures; (iv) use a multiplex approach forsimultaneous detection of these bacterial strains inmilk and (v) evaluation of the method in raw milksamples.

Table 1. Sequences of oligonucleotide primers used in the specenterocolitica

Primer Nucleotide sequence 5 0±3 0 Bacterial species/

Sa-1 GAAAGGGCAATACGCAAAGA S. aureusSa-2 TAGCCAAGCCTTGACGAACT Nuclease (nuc)

Ye-1 CTATTGGTTATGCGCAAAGC Y. enterocoliticaYe-2 TGCAAGTGGGTTGAATTGCA Attachment Invasio

MATERIALS AND METHODS

Bacterial strains and culture conditions

Staphylococcus aureus ATCC 14458 obtainedthrough the courtesy of Dr S. Notermans, NationalInstitute of public Health, the Netherlands andYersinia enterocolitica MTCC 859, procured fromInstitute of Microbial Technology, Chandigarh werechosen as model food-borne pathogens in this study.The strains were maintained on brain-heart-infusion(BHI) agar (HiMedia, Mumbai, India) slants at 4�C.S. aureus and Y. enterocolitica were grown in 10 mlBHI broth at 37�C and 32�C, respectively, for 18 h at150 rpm in an orbital shaker (Sciegenics Biotech Pvt.Ltd, Chennai, India). The cells were harvested bycentrifugation at 8000� g for 20 min at 4�C andresuspended in sterile 10 ml, 0�85% saline prior tobeing used in the experiments.

Oligonucleotide primers

The sequences of the oligonucleotide primers usedin the study are given in Table 1. Based on theGenBank data (V 01281) primers SA-1 and SA-2 weredesigned using the Primer 3�0 software programdevised by Eve Rosen and Helen J. Skaletsky (1996,1997). These primers were selected to demarcate a482 bp region of the staphylococcal nuc gene codingfor nuclease. The oligonucleotide primers Ye-1 andYe-2 which amplify a 359 bp fragment of thevirulence-associated attachment invasion locus(ail ) of Y. enterocolitica was selected based on thesequences reported earlier.16

Isolation of genomic DNA from pure cultures

Genomic DNA was isolated from overnight culturesof S. aureus and Y. enterocolitica by using the pro-cedure described previously.17 DNA isolated fromthese strains was used individually and in combin-ation to test the speci®city of the primers in aconventional and multiplex PCR respectively.

i®c detection of Staphylococcus aureus and Yersinia

target gene Amplicon size (bp) Reference/source

482 This study

359 Fenwick & Murray (1991)n Locus (ail )

Multiplex PCR for food borne pathogens 309

Sample preparation for pasteurized milk

Pasteurized milk samples were procured from thelocal retail outlets and steamed for 30 min prior touse. To investigate the compatibility of the DNAextractionmethodwithPCR-basedenumerationofcellnumber, 0�4 ml of saline suspensions having knownconcentrations of S. aureus and Y. enterocolitica inthe range of 108±101 cfu were added individually to0�4 ml of milk samples.

The speci®city of the chosen primers to discern theabove bacterial strains in a mixed population as wellas to facilitate their simultaneous detection wastested in another set of experiment wherein themilk samples (0�4 ml) were mixed with 0�2 ml of eachof the bacterial strains of known concentrations in therange of 106±102 cfu.

DNA extraction from milk samples

To extract bacterial DNA from the spiked and controlmilk samples, 0�25 ml each of diethyl ether andchloroform were added to the samples and vortexedbrie¯y. The samples were centrifuged at 8000 g for20 min at 25�C. The aqueous phase was transferredto a fresh tube and 0�5 ml of 6 M urea and 0�1 ml of10% SDS were added. The samples were incubatedat 37�C for 20 min and then centrifuged at 80006 gfor 20 min at 25�C. The supernatant was discardedand 0�1 ml of 0�2 N NaOH was added to the samplesand incubated at 37�C for 10 min. DNA wasprecipitated by adding 1�0 ml of chilled absoluteethanol and one-tenth volume of 3 M sodium acetate(pH 4�8) and holding the samples at ÿ20�C for 2 h.Samples were then centrifuged at 80006 g for20 min at 4�C. Excess salt in the DNA preparationwas removed by 70% ethanol wash. The DNA pelletwas air-dried and resuspended in 15 ml of sterileultra®ltered water (Millipore, Bedford, MA, USA).

PCR conditions

Ampli®cation of bacterial DNA was performed in atotal reaction volume of 25 ml which contained 2 ml ofsamples comprising either genomic DNA from purecultures or DNA extracted from milk samples. Thereaction mixture consisted of 16 PCR buffer (10 mM

Tris-HCl pH 9�0, 50 mM KCl, 1�5 mM MgCl2, 0�01%gelatin), 200mM of each deoxynucleoside tri-phosphate, 50 pmol of each primer and 1�0 U ofTaq DNA polymerase (Bangalore Genei, Bangalore,India). Template DNA was initially denatured at94�C for 5 min. Subsequently, a total of 35 ampli-®cation cycles were carried out in a programmable

thermocycler (GeneAmp PCR System 9700, Perkin-Elmer, USA). Each cycle consisted of denaturation for1 min at 94�C, primer annealing for 1 min at 55�Cand extension for 1 min at 72�C. The last cycle wasfollowed by a ®nal extension at 72�C for 8 min.

To check the speci®city of the primers and forthe simultaneous detection of S. aureus andY. enterocolitica in milk samples, a multiplex PCRwas carried out using gene speci®c primers targetedto both the bacterial strains in the same reactionmixture. A total of 25 ml of the reaction mixtureincluded 16 PCR buffer, 3�0 mM MgCl2, 200mM ofeach deoxynucleoside triphosphate, 50 pmol of eachprimer, 3�0 U of Taq DNA polymerase and 4 ml oftemplate DNA (either from pure culture or frommilk). Prior to thermocycling, a `hot-start' format wasadopted, wherein the PCR mixture was held at 94�Cfor 5 min followed by addition of Taq DNA polymer-ase. Ampli®cation was carried out for 35 cycles andthe cycling pro®le was same as mentioned above.

PCR products were analyzed by agarose gelelectrophoresis. Aliquots (10 ml) of PCR productswere loaded onto 1.5% agarose gel and subjectedto electrophoresis for 2 h at 120 V in 16TAEbuffer.18 Gels were observed and documented in aCCD-camera based gel documentation system(E.A.S.Y-RH, Herolab, Wielsloch, Germany).

Analysis of raw milk samples

Raw milk samples from seven different sources werecollected from the local dairy farms. For the enu-meration of S. aureus, 1 ml aliquots of samples wereserially diluted in 0�85% saline and appropriate dilu-tions were plated onto Baird-Parker agar (HiMedia,Mumbai) plates. The plates were incubated at 37�Cfor 24±48 h and observed for characteristic colonymorphology. For the isolation of Y. enterocolitica,1 ml aliquot of the milk samples was enriched in99 ml phosphate buffered saline (PBS) for 48 h.A 0�5 ml aliquot of the enriched broth was mixedwith 4�5 ml of 0�25% KOH and incubated for 1 min.Subsequently, 0�1 ml of this broth was streaked ontoYersinia selective agar (HiMedia, Mumbai) supple-mented with Cefsulodin-Irgasan-Novobiocin (CIN).The plates were incubated at 32�C for 48 h andobserved for characteristic bulls-eye colonies.19

Aliquots of 1 ml of individual milk samples weresubjected to total DNA extraction using the methodas described earlier. The DNA preparations wereused as templates in PCR with nuc and ail primers ofS. aureus and Y. enterocolitica, respectively. Ampli-®cation conditions and analysis of the ampliconswere as described previously.

Fig. 2. PCR based direct detection of S. aureus in spikedmilk samples. (a) Agarose gel electrophoresis of ampli®edPCR product; (b) Histogram showing quanti®cation ofPCR product. Lanes: M: molecular size marker; 1±8:108±101 cfu mlÿ1; 9: control.

A. Ramesh et al.310

RESULTS

Speci®city of oligonucleotide primers

The speci®city of the nuclease (nuc) and theattachment invasion locus (ail ) primers were evalu-ated by PCR using the genomic DNA isolated fromS. aureus ATCC 14458 and Y. enterocolitica MTCC859, respectively. Agarose gel electrophoresis of thePCR products revealed an amplicon size of 482 and359 bp (Fig. 1, Lanes 1 and 2). The size of theamplicons corresponded to the expected size and noadditional or nonspeci®c bands were observed.

The ability of the primer pairs to yield gene speci®camplicons in a single reaction was also assessed. Itwas evident from the result (Fig. 1, Lane 3) thatspeci®c and unique PCR products of 482 and 359 bpwere observed indicating no cross reactivity betweenthe primers.

Sensitivity of PCR

The results of PCR using the nuc and ail primers inconjunction with DNA extracts of milk sampleshaving known concentrations of S. aureus andY. enterocolitica are shown in Figures 2 and 3. The

Fig. 1. Agarose gel electrophoresis of the PCR productsobtained from S. aureus ATCC 14458 and Y.enterocolitica MTCC 859. Lanes: M: GeneRulerTM 100 bpladder plus; 1: PCR product from S. aureus usingnuc primers; 2: PCR product from Y. enterocolitica usingail primers; 3: multiplex PCR product using nuc and ailprimers.

Fig. 3. PCR based direct detection of Y. enterocolitica inspiked milk samples. (a) Agarose gel electrophoresis ofampli®ed PCR product; (b) Histogram showingquanti®cation of PCR product. Lanes: M: molecular sizemarker; 1±8: 108±101 cfu mlÿ1; 9: control.

Multiplex PCR for food borne pathogens 311

detection limit for both the strains was 10 cfu mlÿ1

which corresponded to 1�34 cfu per reaction tube(Figs 2a and 3a). There was a steady decrease in theintensity of the amplicons which correlated withdecreasing cell numbers for both the strains (Figs 2band 3b). The ampli®cation pro®le for S. aureus withnuc primers was brighter in intensity (Fig. 2b)compared to that obtained for Y. enterocolitica withthe ail primers (Fig. 3b). Ampli®cation products werenot observed on agarose gel for the DNA extractsfrom unspiked milk samples.

Impact of background ¯ora on PCR sensitivity

The sensitivity of PCR for enumerating a speci®cbacterial pathogen in presence of a co-existingmicrobe was investigated by extracting DNA frommilk samples spiked with equal numbers ofS. aureus and Y. enterocolitica in the range of 106±101 cfu mlÿ1 ). From Figures 4 and 5 it was evidentthat a detection limit of 103 cfu mlÿ1 could beachieved individually for both the strains, corre-sponding to 134 cfu per PCR tube. There was aquantitative decrease in the intensity of the

Fig. 4. PCR based direct detection of S. aureus in thepresence of Y. enterocolitica in milk samples. Lanes: M:molecular size marker; 1±6: 106±101 cfu mlÿ1; 7: control.

Fig. 5. PCR based direct detection of Y. enterocolitica inthe presence of S. aureus in milk samples. Lanes: M:molecular size marker; 1±6: 106±101 cfu mlÿ1; 7: control.

amplicons re¯ecting corresponding decrease in cellnumbers. Uninoculated milk samples failed to yieldany ampli®cation product.

Simultaneous detection of S. aureus andY. enterocolitica

The potential of the primer pairs for simultaneousdetection of both the pathogenic strains in a singletube reaction was assessed. Milk samples werespiked with equal numbers of both the strains in therange of 106±102 cfu mlÿ1. Using the multiplex PCRconditions stated before, the primer pairs were ableto generate gene speci®c amplicons, which corr-esponded to the expected size of 482 and 359 bp,respectively. The amplicons were visualized asseparate and explicit bands of unique size (Fig. 6).No extra bands or non-speci®c bands were observed.The strains of S. aureus and Y. enterocolitica could besimultaneously detected when they co-existed inmilk at a cell concentration as low as 104 cfu mlÿ1

(Fig. 6, Lane 3).

Evaluation of raw milk samples

The viable counts of mesophilic aerobes, S. aureusand Y. enterocolitica in raw milk samples arepresented in Table 2. The total count in the milksamples varied from 5�0±7�12 log10 cfu mlÿ1. With-out any prior enrichment, all the milk samplesdisplayed the presence of S. aureus with a highestcount of 6�03 log10 cfu mlÿ1 for sample number 1.Sample numbers 3, 4 and 6 had the least count of4�0 log10 cfu mlÿ1. DNA was extracted from the milksamples as described earlier and subjected to PCR

Fig. 6. Multiplex PCR for simultaneous detection ofS. aureus and Y. enterocolitica present as mixed culturein milk samples Lanes: M: Molecular size marker;1±5: 106±102 cfu mlÿ1.

Table 2. Analysis of raw milk samples for S. aureus andY. enterocolitica by conventional plating and PCRmethods

Sample no. Total countlog10 cfu mlÿ1

S. aureus countlog10 cfu mlÿ1

PCR withnuc

1 7�12 6�03 �2 5�84 4�30 �3 5�00 4�00 �4 6�00 4�00 �5 5�77 4�47 �6 5�30 4�40 �7 6�88 4�77 �

A. Ramesh et al.312

using nuc set of primers. Speci®c amplicon corre-sponding to the expected size of 482 bp was obtainedfor all the samples. No characteristic colonies wereobtained on YSM agar plates supplemented with CINwhen a PBS enrichment and alkali treatment step wasadopted to isolate Y. enterocolitica from raw milksamples. Ampli®cation was not observed when theDNA extracts of the milk samples were subjected toPCR using ail set of primers.

DISCUSSION

The polymerase chain reaction (PCR) has beenwidely accepted as the method of choice for rapidand reliable detection of microbes in food. Thistechnique can be extremely useful for pure microbialcultures, but when applied directly to food samplesits ef®ciency can be markedly reduced. The crux ofthis problem lies in the sample preparation method,which may inadvertently introduce inhibitory sub-stances and preclude the PCR. Appropriate methodshave thus been devised to overcome this problemand facilitate the development of PCR based routinemethods for bacterial detection in food samples.6,7,20

The present study illustrates a procedure for directPCR ampli®cation of bacterial DNA extracted frommilk samples. Presence of fat and other interferingsubstances in milk can foil efforts to obtaina consistent homogenate and extract adequate DNAfrom target bacterial cells. In the proposed method,a single step of organic extraction with diethyl etherand chloroform enabled the effective removal of fatand the method was relatively simple, compared tothe exhaustive steps of organic extraction.10 Theinclusion of urea facilitated the dissolution of themilk proteins. Proteinases present in milk have beensuggested to be potential inhibitors of PCR and thiseffect was previously obliterated by the use of BSAor proteinase inhibitors.5 In our method of DNAextraction, the treatment of samples with SDS and

NaOH served the purpose of alleviating the inhi-bitory effect of proteinases present in milk. Theinherent robustness of certain bacterial strains withspecial reference to Gram positive bacteria and theelaboration of nucleases either from the bacterialcells or from the food matrix itself necessitates the useof enzymes such as lysozyme and proteinase K in theDNA extraction method.21,22 In our study, ef®cientcell lysis was accomplished with the use of SDS andNaOH. Finally the DNA released from the lysed cellswas precipitated with ethanol in presence of sodiumacetate, avoiding the use of additional carriers suchas etachinmate.10 Essentially, the DNA extractionmethod proposed in this study does not require anyenrichment step and reduces sample loss by encom-passing only three basic steps.

Staphylococcal food poisoning out breaks hasbeen well documented from dairy products.23 Milkis often stored in refrigeration conditions and theability of pathogenic serotypes of Y. enterocolitica tosurvive and grow at low temperatures makes it apublic health concern for dairy industries. Consider-ing the above mentioned facts, S. aureus andY. enterocolitica were used as model pathogens inour investigation. The direct detection of thesepathogens in milk was ascertained by PCR usingthe DNA extracts from milk and gene speci®c primersunique to both the strains.

Using the nuc and ail primers, amplicons of theexpected size were obtained from pure cultures ofboth the strains indicating the speci®city of thechosen primers. The possibility of amplifying targetsequences from both the strains was explored ina multiplex PCR. The `hot-start' format facilitatedcomplete denaturation of the template DNA andreduced the exposure of the polymerase to the hightemperature. Using a mixture of genomic DNAisolated from pure cultures of each strain, speci®camplicons of 482 and 359 bp were obtained in amultiplex PCR. This indicated that the nuc and ailprimers shared common melting properties andannealed uniquely to the target sequences, allowingidenti®cation of the bacterial strain based onmolecular size difference of the amplicons.

Although there has been a preponderance ofreports on PCR based detection of S. aureus inclinical samples,24±26 studies dealing with the quan-titative detection of S. aureus in food samples havebeen scanty. In our earlier investigation, the use ofnuc gene as a target for the detection of S. aureus infood samples has been reported.27 In the presentinvestigation, the detection limit for S. aureus was10 cfu mlÿ1 milk using nuc primers in PCR, withoutany prior enrichment of the samples. This level ofdetection was superior than that reported for

Multiplex PCR for food borne pathogens 313

skimmed milk and cream, respectively.28,29 Usingthe ail primers for Y. enterocolitica PCR applied toDNA extracted from milk samples enabled thedetection of 10 cfu mlÿ1milk. The detection limitsfor Y. enterocolitica attained in our study was similarto that observed earlier.30 But in the latter case, acolony hybridization technique was used followinggrowth of Yersinia colonies on CIN agar.

The presence of a co-existing non-target bacteriumin a food sample can attenuate the speci®c detectionof the target bacterial species. Strains of S. aureusand Y. enterocolitica were spiked together in milksamples to create an analytical system similar to thatin a food sample, wherein multiple pathogens arelikely to be present. A detection limit of 103 cfu mlÿ1

was achieved for both the pathogens when theycoexisted in milk in equal numbers. The sensitivityachieved was less than that obtained for individualpathogens. This can be accounted by the presence ofnon target bacterial DNA which can partition thePCR components and reduce the ampli®cationef®ciency. It is possible that at a cell concentrationbelow 103 cfu mlÿ1, the amplicons produced couldnot be detected on agarose gels.

In the multiplex PCR using nuc and ail primers ina single tube, 104 cfu mlÿ1 of S. aureus andY. enterocolitica could be simultaneously detected.The assay was speci®c and reliable since theidentities of the strains were indicated by thecharacteristic size of the amplicon. The detectionlimit attained in our study was less than that reportedearlier for Salmonella and Listeria monocytogeneswhere 103 cfu mlÿ1 milk could be detected.31 How-ever, this level of detection required slot blothybridization of the PCR products following ampli-®cation. The sensitivity of the multiplex PCR devel-oped in our study could possibly be enhanced byeither using primers designed for high copy numbertarget gene or by probe hybridization of theamplicons.

For routine use in food analysis, detection methodsneed to be sound and reliable. All the replicaexperiments conducted with the arti®cially inocu-lated milk samples yielded the same results indicat-ing the reproducibility of the method developed. Amajor drawback of PCR is its lack of discriminatorypower to differentiate live and dead cells. This couldlimit its use in the quantitative enumeration of bact-erial cells. An enrichment step on the other hand canobviate this problem, but is time consuming. In ourinvestigation S. aureus and Y. enterocolitica werefreshly grown prior to experiments, used for inocu-lating milk samples, which were immediately sub-jected to DNA extraction. Since the uninoculatedsamples failed to give any ampli®cation, the PCR

data obtained for the spiked samples re¯ected thecount of the added bacterial cells.

The raw milk samples procured from the localdairy farms showed the prevalence of S. aureusre¯ecting poor conditions of sanitation and hygiene.In our study PCR as well as the conventional platingfailed to show the presence of Y. enterocolitica in theraw milk samples tested. A similar observation wasalso reported earlier.30 As preliminary experiments,we had selected only a limited number of samples.

The implementation of PCR-based surveillanceprogram for detecting pathogens in milk samplescan be hindered by either the physical deteriorationof milk or the presence of an over-whelmingbackground population. Thus, there is a need todevelop an ef®cient protocol applicable directly tomilk samples, which remains unaffected by matrixderived factors, potential inhibitors and indigenousmicro¯ora. The DNA extraction method developedin the present study ful®ls these requirements andgenerates PCR compatible templates without anyenrichment of the samples. The method is rapid andthe level of sensitivity achieved in our experiments isapplicable to the practical survey of microbialcontamination in milk. A major outcome of thestudy is the development of a multiplex PCR to detectmultiple pathogens using compatible primers and theDNA extracts from the milk samples. The implica-tions of the present study are promising and the DNAextraction method in combination with judiciouschoice of primers in PCR can be extended to detectpathogens present in milk based products.

ACKNOWLEDGEMENTS

The authors are thankful to Dr V. Prakash, Director, CFTRI,Mysore for providing the facilities and also for interest in thepresent work. We thank Dr M. S. Prasad, Head of thedepartment, Food Microbiology. AR and BPP are thankfulto Council of Scienti®c and Industrial Research, New Delhi,India for Research Associateship and Senior ResearchFellowship, respectively. We are grateful to the Departmentof Biotechnology, Government of India for their ®nancialsupport.

REFERENCES

1. Scheu, P. M., Berghof, K. & Stahl, U. (1998). Detectionof pathogenic and spoilage microorganisms infood with the polymerase chain reaction. FoodMicrobiology 15, 13±31.

2. Josephson, K. L., Gerba, C. P. & Pepper, I. L. (1993).Polymerase chain reaction detection of nonviablebacterial pathogens. Applied and EnvironmentalMicrobiology 59, 3513±15.

A. Ramesh et al.314

3. Rossen,L.,Norskov,P.,Holmstrom,K.&Rasmussen,O.F.(1992). Inhibition of PCR by components of foodsamples, microbial diagnostic assays and DNA-extraction solutions. International Journal of FoodMicrobiology 17, 37±45.

4. Wilson, I. G. (1997). Inhibition and facilitation ofnucleic acid ampli®cation. Applied and EnvironmentalMicrobiology 63, 3741±51.

5. Powell, H. A., Gooding, C. M., Garrett, S. D., Lund, B. M.& McKee, R. A. (1994). Proteinase Inhibition of thedetection of Listeria monocytogenes in milk using thepolymerase chain reaction. Letters in Applied Micro-biology 18, 59±61.

6. Lindqvist, R., Norling, B. & Lambertz, S. T. (1997).A rapid sample preparation method for PCR detectionof food pathogens based on buoyant density centri-fugation. Letters in Applied Microbiology 24, 306±10.

7. Par-Gunnar, L., Hahn-Hagerdal, B. & Radstrom, P.(1994). Sample preparation methods in PCR-baseddetection of food pathogens. Trends in Food Scienceand Technology 5, 384±9.

8. Champagne, C. P., Laing, R. R., Roy, D., Mafu, A. A. &Grif®ths, M. W. (1994). Psychrotrophs in dairy pro-ducts: Their effects and their control. Critical Reviewsin Food Science and Nutrition 34, 1±30.

9. Starbuck, M. A. B., Hill, P. J. & Stewart, G. S. A. B.(1992). Ultra sensitive detection of Listeria mono-cytogenes in milk by the polymerase chain reaction(PCR). Letters in Applied Microbiology 15, 248±52.

10. Herman, L. M. F., De Block, J. H. G. E. &Moermans, R. J. B. (1995). Direct detection of Listeriamonocytogenes in 25 milliliters of raw milk by atwo-step PCR with nested primers. Applied andEnvironmental Microbiology 61, 817±19.

11. Nogva, H. K., Rudi, K., Naterstad, K., Holck, A. &Lillehaug, D. (2000). Application of 5 0-nuclease PCRfor quantitative detection of Listeria monocytogenes inpure cultures, water, skim milk, and unpasteurizedwhole milk. Applied and Environmental Microbiology66, 4266±71.

12. Adesiyun, A. A., Webb, L. A. & Romain, H. T. (1998).Prevalence and characteristics of Staphylococcusaureus strains isolated from bulk and composite milkand cattle handlers. Journal of Food Protection 61,629±32.

13. Stern, N. J., Pierson, M. D. & Kotula, A. W. (1980).Effects of pH and sodium chloride on Yersiniaenterocolitica growth at room and refrigeration tem-perature. Journal of Food Science 45, 64±7.

14. Black, R. E., Jackson, R. J., Tsai, T., Medvesky, M.,Shayegani, M., Feeley, J. C., Macleod, K. I. E. &Wakelee, A. M. (1978). Epidemic Yersinia entero-colitica infection due to contaminated chocolate milk.New England Journal of Medicine 298, 76±9.

15. Tacket, C. O., Narain, J. P., Sattin, R., Lofgren, J. P.,Konigsberg, C. H., Rendtorff, R. C., Rausa, A., Davis, B.R. & Cohen, M. L. (1984). A multistate outbreak ofinfections caused by Yersinia enterocolitica transmittedby pasteurized milk. Journal of American MedicalAssociation 251, 483±6.

16. Fenwick, S. G. & Murray, A. (1991). Detection ofpathogenic Yesinia enterocolitica by polymerase chainreaction. Lancet 337, 496±7.

17. Ausubel, F. M., Brenl, R. & Kingstone, R. E. (1987).Preparation of genomic DNA from bacteria, In Current

Protocols in Molecular Biology, Unit 2�4, section 2�4�1.New York: John Wiley and Sons.

18. Sambrook, J., Fritsch, E. F. & Manniatis, T. (1989).Molecular cloning ± a laboratory manual. 2nd edition.Cold Spring Harbor, NY: Cold Spring Harbor LaboratoryPress.

19. Doyle, M. P. & Hugdahl, M. B. (1983). Improvedprocedure for recovery of Y. enterocolitica from meats.Applied and Environmental Microbiology 45, 127±35.

20. Lampel, A. K., Orlandi, P. A. & Kornegay, L. (2000).Improved template preparation for PCR-based ssays fordetection of food-borne bacterial pathogens. Appliedand Environmental Microbiology 66, 4539±42.

21. Gutierrez, R., Garcia, T., Gonzalez, I., Sanz, B.,Hernandez, P. E. & Martin, R. (1997). A quantitativePCR-ELISA for the rapid enumeration of bacteria inrefrigerated raw milk. Journal of Applied Microbiology83, 518±23.

22. Romero, C. & Lopez-Goni, I. (1999). Improved methodfor puri®cation of bacterial DNA from bovine milk fordetection of Brucella spp. by PCR. Applied andEnvironmental Microbiology 65, 3735±37.

23. Genigeorgis, C. A. (1989). Present state of knowledgeon Staphylococcal intoxication. International Journalof Food Microbiology 9, 327±60.

24. Monday, S. R. & Bohach, G. A. (1999). Use ofmultiplex PCR to detect classical and newly describedpyrogenic toxin genes in Staphylococal isolates.Journal of Clinical Microbiology 37, 3411±14.

25. Brakstad, O. G., Aasbakk, K. & Maeland, J. A. (1992).Detection of Staphylococcus aureus by polymerasechain reaction ampli®cation of the nuc gene. Journal ofClinical Microbiology 30, 1654±60.

26. Barski, P., Piechowicz, L., Galinski, J. & Kur, J. (1996).Rapid assay for detection of methicillin-resistantStaphylococcus aureus using multiplex PCR. Molecularand Cellular Probes 10, 471±5.

27. Padmapriya, B. P., Radhika, B., Keshava, N.,Chandrashekar, A. & Varadaraj, M. C. Polymerasechain reaction for the detection of food isolates ofStaphylococcus aureus using nuclease (nuc) genespeci®c primers. Abstract H24 of the 69th AnnualMeeting of Society of Biological Chemists (India),Calcutta, India, Dec. 7±9, 2000.

28. Wilson, G. I., Cooper, J. E. & Gilmour, A. (1991).Detection of enterotoxigenic Staphylococcus aureus indried skimmed milk: Use of the polymerase chainreaction for ampli®cation and detection of staphylo-coccal enterotoxin genes entB and entC1 and thethermonuclease gene nuc. Applied and EnvironmentalMicrobiology 57, 1793±8.

29. Straub, J. A., Hertel, C. & Wallter, P. H. (1999). A23S rDNA-targeted polymerase chain reaction-basedsystem for detection of Staphylococcus aureus in meatstarter cultures and dairy products. Journal of FoodProtection 62, 1150±6.

30. Durisin, M. A., Ibrahim, A. & Grif®ths, M. W. (1997).Detection of pathogenic Yersinia enterocolitica in milkand pork using a DIG-labeled probe targeted againstthe yst gene. International Journal of Food Micro-biology 37, 103±12.

31. Li, X., Boudjellab, N. & Zhao, X. (2000). CombinedPCR and slot blot assay for detection of Salmonella andListeria monocytogenes. International Journal of FoodMicrobiology 56, 167±77.