PCR-ELISA detection of Escherichia coli in milk

PCR-ELISA de ec ion o Esche ichia coli in milk P. Daly, T. Collie and S. Doyle Depa men o Biology, Na ional Uni e si y o I eland, Maynoo h, I eland 2001/232: ecei ed 6 Augus 2001 and accep ed 19 Decembe 2001 P. DALY, T. COLLIER AND S. D O YLE. 2002. Aims: The pu pose o his s udy was o de elop a eliable molecula p ocedu e o he de ec ion o Esche ichia coli in milk. Me hods and Resul s: Robus and expedi ious DNA ex ac ion and PCR echniques we e e alua ed using Enzyme-Linked Immunoso ben Assay (ELISA) de ec ion o bio in-labelled amplicons o acili a e op imal de ec ion o E. coli DNA. Conclusions: I was ound ha 5 E. coli colony- o ming uni s (c u) could be de ec ed pe PCR eac ion using he PCR-ELISA sys em, equa ing o a sensi i i y o de ec ion o 100 E. coli c u ml )1 pas eu ized milk. Signi icance and Impac o he S udy: This app oach should acili a e e alua ion o milk con amina ion and enable apid de ec ion o E. coli mas i is, leading o co ec deploymen o ele an an ibio ic he apy and imp o ed animal wel a e. INTRODUCTION In ecen yea s Esche ichia coli has become ecognized as a se ious ood-bo ne pa hogen and has been associa ed wi h nume ous ou b eaks o disease in he UK, Japan and USA (Uy endaele e al. 1999; Sco e e al. 2000). Raw milk and dai y p oduc s, such as pas eu ized milk and yoghu con amina ed wi h E. coli, ha e been he main cause o se e al ou b eaks o milk-bo ne disease since he 1980s and hus emain a se ious heal h isk (Allmann e al. 1995; Seo e al. 1998). De ec ion and enume a ion o E. coli in ood p oduc s can be assessed by a ious me hods including pla e coun ing, immuno-magne ic sepa a ion, low cy ome y and ch omogenic- luo ogenic subs a e echnology (Venka es- wa an e al. 1996; de Boe 1998; Seo e al. 1998; Yu 1998). The s anda d de ec ion me hod o E. coli in ol es he use o subs a es ha de ec E. coli and can be ime-consuming. Up o 97% o E. coli p oduce b- D -glucu onidase which hyd olyses 4-me hylumbelli e yl-b- D -glucu onide (MUG) o o m a blue luo escen p oduc (Venka eswa an e al. 1996; Mana i 2000). Howe e , a mu a ion in he uidA gene in E. coli leads o alse-nega i e iden i ica ion when using b- D -glucu onidase as a me hod o de ec ion, while ce ain s ains o E. coli, such as O157, ail o p oduce his enzyme (Yokoigawa e al. 1999). In addi ion, b- D -glucu onidase is also p oduced by membe s o he Salmonella,Shigella and some Ye sinia spp., while S aphylococcus spp. can also hyd olyse MUG. Thus, mo e accu a e me hods o de ec - ing E. coli a e necessa y (de Boe 1998). Polyme ase chain eac ion (PCR) has led o apid and sensi i e de ec ion o E. coli om clinical samples and a ious ood ypes (F a amico e al. 2000; McKillip and D ake 2000; Hsu and Tsen 2001). Howe e , many ood ypes con ain PCR inhibi o s which co-pu i y wi h he a ge DNA (Gonza ´lez e al. 1999), he eby equi ing ex ensi e sample p epa a ion o emo e, dilu e o inac i a e inhibi o s p io o PCR ampli ica ion (F a amico e al. 2000). In many cases, he low le el o E. coli in ood ypes equi es a p e-en ichmen s ep o o e come poo sensi i i y, which inc eases he o e all assay ime by up o 24 h (Seo e al. 1998; Sco e e al. 2000). Thus, a mo e sensi i e, speci ic and apid me hod o de ec ing E. coli is equi ed. Alanine acemase is p oduced by all E. coli s ains, including O157, and is essen ial o cell-wall biosyn hesis by p o iding D -alanine o pep idoglycan. Thus, de ec ion o he alanine acemase (al ) gene may be use ul o he de ec ion o E. coli in oods (Yokoigawa e al. 1999). In he p esen s udy, he success ul design and applica ion o a obus PCR-ELISA o he de ec ion o low le els E. coli in milk is desc ibed. Co espondence o: D S. Doyle, Depa men o Biology, Na ional Uni e si y o I eland, Maynoo h, Co. Kilda e, I eland (e-mail: [email p o ec ed]). ª2002 The Socie y o Applied Mic obiology Le e s in Applied Mic obiology 2002, 34, 222–226 MATERIALS AND METHODS Bac e ial s ain and a i icial specimen p epa a ion Esche ichia coli s ain JM109 was ob ained om P omega (Madison, WI, USA). Esche ichia coli-adul e a ed milk specimens we e p epa ed as ollows. An E. coli cul u e was g own in Lu ia-Be ani (LB) B o h a 37°C o 16 h wi h shaking. A 25 ml aliquo o he cul u e was cen i uged a 3200 g o 15 min. A e emo al o he supe na an luid, he pelle was esuspended in 1 ml esh pas eu ized milk, aw milk o phospha e-bu e ed saline (0Æ05% ( / ) Tween- 20; PBST) and each sample se ially dilu ed in unadul e a ed ma ix o de e mine he sensi i i y o PCR ampli ica ion. A 100 ll aliquo o each dilu ion was also pla ed, in iplica e, on LB Aga and colony coun ing was pe o med using an Eagle-Eye II gel documen a ion sys em (S a agene, La Jolla, CA, USA). F esh pas eu ized milk and aw milk we e indi idually pla ed o de e mine he p esence o na u ally- con amina ing mic o lo a. DNA ex ac ion Milk samples (100 ll) con aining E. coli (0–10 4 c u ml )1 ) we e each mixed wi h 1 ml PBST, o exed, and cen i- uged a 10 000 g o 5 min. This p ocedu e was pe o med o pelle he bac e ial cells and o emo e in e e ing amoun s o p o ein and lipid in milk samples, which may in e e e wi h PCR ampli ica ion. Following emo al o he supe na an luid, DNA was ex ac ed using he ollowing p ocedu es. Alkaline ex ac ion. The alkaline ex ac ion p ocedu e was a modi ica ion o he me hod p oposed by Milla e al. (2000). The bac e ial pelle was esuspended in 1 ml alka- line ex ac ion solu ion (0Æ5moll )1 sodium hyd oxide, 0Æ05 mol l )1 sodium ci a e). This mix u e was shaken o 10 min and cen i uged a 13 000 g o 5 min. The supe - na an luid was again emo ed and he pelle esuspended in 500 ll0Æ5 mol l )1 T is-HCl pH 8Æ0, ollowed by cen- i uga ion a 13 000 g o 5 min. This s ep was epea ed. The inal pelle was esuspended in 100 ll 10 mmol l )1 T is-HCl, 1 mmol l )1 EDTA pH 8Æ0, and placed in a hea ing block o 1 h a 100°C. The sample was hen eeze– hawed wice, cen i uged a 13 000 g o 15 min and he supe na an luid emo ed o PCR analysis. De e gen ex ac ion. The ex ac ion me hod o Yokoig- awa e al. (1999) o de ec E. coli s ains in oods, including aw milk, was employed. He e, ollowing he ini ial cen i uga ion o milk samples in PBST, he supe na an luid was emo ed and 100 ll lysis bu e (10% (w/ ) NaCl, 2% ( / ) Tween-20 and 2 mmol l )1 EDTA) was added. Samples we e hen ea ed a 100°C on a hea ing block o 10 min and cen i uged a 5000 g o 5 min. Supe na an luids we e hen used di ec ly as PCR empla es. Sodium iodide (NaI) ex ac ion. The NaI ex ac ion p ocedu e was ha p oposed by Ishizawa e al. (1991) o isola ion o DNA om se um. Oligonucleo ide p ime s and DNA ampli ica ion P ime s speci ic o a conse ed egion si ua ed wi hin he E. coli alanine acemase gene we e selec ed (Yokoigawa e al. 1999) using a modi ied ampli ica ion p ocedu e: o wa d: 5¢ Bio in-CTGGAAGAGGCTAGCCTGGACGAG 3¢and e e se: 5¢-AAAATCGGCACCGGTGGAGCGATC-3¢ (Sigma-Genosys, Camb idge, UK). A 5 ll olume o each ex ac ed sample was used o PCR (PCR Exp ess The mal Cycle ; Hybaid, Ash o d, Middlesex, UK), which also con ained 10 mmol l )1 T is-HCl pH 9Æ0, 50 mmol l )1 KCl, 0Æ1% ( / ) T i on X-100, 2Æ0 mmol l )1 MgCl 2 , 200 lmol l )1 o each dNTP (P omega), 1 mol l )1 be aine (Sigma) and 1Æ0lmol l )1 o he abo e p ime s in a o al olume o 49Æ5ll. Ho -s a PCR was pe o med on each sample by adding 1Æ25 U Taq polyme ase (P omega) a e he ini ial 6 min dena u a ion s ep a 95°C. The PCR eac ion was hen con inued wi h 35 cycles consis ing o 20 s dena u a ion a 95°C, p ime annealing/ex ension a 72°C o 90 s, and a inal ex ension o 5 min a 72°C. A 10 ll aliquo o each PCR p oduc was subjec ed o 1% (w/ ) aga ose gel elec opho esis con aining 0Æ5lgml )1 e hidium b omide (Sigma) o 30 min a 100 V. Amplicon isualiza- ion (366 bp) was pe o med using an Eagle-Eye II gel documen a ion sys em (S a agene). Mic opla e p epa a ion and ELISA o ma S ep a idin-coa ed mic opla es we e p epa ed by coa ing mic owells (Nunc Maxiso p, Roskilde, Denma k) wi h s ep a idin (2Æ5lgml )1 ) in 50 mmol l )1 sodium ca bona e pH 9Æ4, and s abilized by he addi ion o 1% (w/ ) albumin in he same diluen . Bio inyla ed PCR p oduc s we e dilu ed 1/20 in 6·SSC o gi e a inal olume o 200 ll; hey we e hen added o s ep a idin-coa ed mic owells and incuba ed a 37°C o 30 min. A e wo washes wi h PBST, 100 llo 125 mmol l )1 NaOH, 100 mmol l )1 NaCl we e added o he mic owells, incuba ed a oom empe a u e o 3 min and washed ou imes wi h PBST. A 100 ll olume o dini ophenol (DNP)-labelled oligonucleo ide (100 ng ml )1 ; 5¢(DNP) 3 -ATATTCACCGGCTGACGAAC-3¢)in6· SSC/0Æ1% (w/ ) sodium dodecyl sulpha e was added o each mic owell and incuba ed o 1 h a 60°C, ollowed by washing ou imes wi h PBST. Mic owells we e hen blocked wi h 2Æ5% (w/ ) milk powde in PBST a 20°C o 1 h. Following blocking solu ion emo al, IgG [an i-DNP] E. COLI PCR-ELISA 223 ª2002 The Socie y o Applied Mic obiology, Le e s in Applied Mic obiology,34, 222–226 ho se adish pe oxidase conjuga e was added, incuba ed a 20°C o 30 min and washed ou imes wi h PBST. Subs a e (100 ll e ame hylbenzidine) was hen added and incuba ed a 20°C o 15 min. The eac ion was e mina ed by he addi ion o 0Æ5 mol l )1 sulphu ic acid and measu ed spec opho ome ically a 450/630 nm (Dyna ech MRX Dynex Technologies, Ash o d, Middlesex, UK). RESULTS Ex ac ion o E. coli and de ec ion o al gene amplicons by aga ose gel elec opho esis Esche ichia coli was cul u ed in LB b o h and added o aw milk, esh pas eu ized milk and PBST. Following pla ing o each dilu ion on LB aga , he numbe s o colony- o ming uni s pe millili e (c u ml )1 ) we e de e mined. Esche ichia coli DNA was ex ac ed om adul e a ed samples anging om 10 0 o 10 4 c u ml )1 using he alkaline, de e gen and NaI ex ac ion p ocedu es. No E. coli DNA was de ec able in ei he unadul e a ed pas eu ized o aw milk, he la e ha ing been ob ained unde s e ile condi ions. Bac e ial DNA was success ully ex ac ed om adul e a ed aw and pas eu ized milk sam- ples, and PBST, by he alkaline ex ac ion and NaI me hods. A minimum o 50 E. coli c u we e ep oducibly de ec able by PCR using he al gene oligonucleo ide p ime s (Table 1 and Fig. 1). Since 5 ll ex ac ed sample (ini ial olume 100 ll) we e used in each PCR eac ion, he PCR de ec ion limi was de e mined o be 10 3 c u ml )1 . Signi ican ly, he de e gen ex ac ion me hod p oposed by Yokoigawa e al. (1999) ailed o esul in he gene a ion o any PCR p oduc (Table 1). In ac , ampli ica ion o a ge DNA ailed e en a e high le els o E. coli (10 8 c u ml )1 ) we e ex ac ed by he de e gen me hod (da a no shown). I was also ound Table 1 PCR ampli ica ion o Esche ichia coli DNA ollowing ex ac ion by sodium iodide, de e gen and hea om aw milk (milk (R), pas eu ized milk (milk (P) and PBST (con ol diluen ). The in ensi y o e hidium b omide amplicon binding is gi en by he ollowing abb e ia ions: high (3 +), mode a e (2 +), low (+) and nega i e ()) Ex ac ion me hods E. coli (c u 5 ll )1 Sodium iodide De e gen Hea PCR empla e) Milk (R) Milk (P) PBST Milk (R) Milk (P) PBST Milk (R) Milk (P) PBST 0))))))))) 0Æ5))))))))) 5))))))))+ 50 + 2 + 2 + )))))2+ 500 3+ 3+ 3+ )))))3+ Fig. 1 Aga ose gel elec opho esis o 366 bp PCR p oduc s om alkaline ex ac ion o aw milk, pas eu ized milk and PBST con aining 0–500 c u Esche ichia coli. Lane 1: nega i e con ol; lanes 2–4: aw milk, pas eu ized milk and PBST, ollowing alkaline ex ac ion, con aining no E. coli; lanes 5– 7: aw milk, pas eu ized milk and PBST, ollowing alkaline ex ac ion, con aining 5 c u E. coli; lanes 8–10: aw milk, pas eu ized milk and PBST, ollowing alkaline ex ac ion, con aining 50 c u E. coli; lanes 11–13: aw milk, pas eu ized milk and PBST, ollowing alkaline ex ac ion, con aining 500 c u E. coli; lane 14: pu i ied E. coli DNA (s ain JM109); lane 15: 1 kbp molecula size ma ke 224 P. DALY ET AL. ª2002 The Socie y o Applied Mic obiology, Le e s in Applied Mic obiology,34, 222–226 ha hea ea men alone o bac e ial cells (100°C o 15 min) ob ained om ei he aw o pas eu ized milk was insu icien o acili a e success ul PCR ampli ica ion o E. coli DNA, ye DNA was de ec able a e hea ea men only, a iden ical E. coli c u ml )1 , in PBST (Table 1). Aga ose gel elec opho esis o PCR p oduc s shows he de ec ion o 50 and 500 c u using he alkaline ex ac ion me hod o DNA ex ac ion om PBST and aw milk, espec i ely (Fig. 1). Occasionally, as ew as 5 E. coli c u (equi alen o 100 c u ml )1 ) we e de ec able a e alkaline ex ac ion om pas eu ized milk. Howe e , e hidium b o- mide s aining could no ep oducibly de ec E. coli a his low le el. PCR-ELISA o al gene amplicons Following amplicon de ec ion by gel elec opho esis, PCR p oduc s we e hen u he analysed by ELISA in an a emp o imp o e he sensi i i y o de ec ion. The esul s o he ELISA o PCR p oduc s om alkaline ex ac ion a e p esen- ed in Fig. 2. Resul s show speci ic de ec ion o 50–500 c u by he DNP-labelled oligonucleo ide p obe when E. coli was ex ac ed om aw milk and PBST. Pas eu ized milk sam- ples showed enhanced ex ac ion o E. coli DNA ela i e o aw milk and PBST, whe eby as ew as 5 c u could be eliably de ec ed using PCR-ELISA a e DNA pu i ica ion by he alkaline ex ac ion me hod; his con i med he inc eased sensi i i y o he ELISA de ec ion me hod o e con en ional aga ose gel elec opho esis and e hidium b omide de ec ion o DNA (i.e. 10 2 s 10 3 c u ml )1 , espec i ely). DISCUSSION A apid and sensi i e PCR-ELISA sys em o de ec ing E. coli in milk has been de eloped. Using se ial dilu ions o E. coli in aw and pas eu ized milk and PBST, he speci ic de ec ion o low le els o E. coli and he emo al o po en ial inhibi o y subs ances by alkaline ex ac ion o NaI ex ac- ion ha e been demons a ed. The PCR-ELISA used in his s udy de ec ed be ween 5 and 50 E. coli c u and did no equi e a p e-en ichmen s ep. Hsu and Tsen (2001) ound le els o 10 6 c u ml )1 (10 3 E. coli c u ml )1 using pla e coun aga ). Bee ens e al. (2000) obse ed a mean le el o 119 E. coli c u ml )1 in 58 samples o aw milk. The le el o endogenous mic o lo a disco e ed in aw milk in he p esen s udy was 2 ·10 2 c u ml )1 , wi h obse ed le els o E. coli be ween 1 and 10 c u ml )1 . This lowe le el o bac e ia in aw milk could be a ibu ed o ob aining he milk om cows unde asep ic condi ions and om heal hy cows wi h no e idence o mamma y in ec ion. Uy endaele e al. (1999) epo ed ha he p esence o 10 6 c u ml )1 o endogenous mic o lo a did no in luence he de ec ion limi o hei PCR assay, which was in ag eemen wi h ou obse a ions. I should be no ed ha he al p ime s used in his s udy de ec a wide ange o E. coli s ains bu also lead o he de ec ion o Shigella spp. (Yokoigawa e al. 1999). Howe e , he incidence le els o Shigella in oods is o dina ily e y low (Gonza ´lez e al. 1999) and he e o e would no be expec ed o complica e he de ec ion limi s o he PCR-ELISA assay p esen ed he e. Yokoigawa e al. (1999) quo e a sensi i i y o de ec ion o 10 3 E. coli cells ml )1 using de e gen ex ac ion and an iden ical p ime pai o ha employed in he p esen s udy. Howe e , i p o ed impossible o eplica e his inding, possibly due o he inhibi o y na u e o EDTA in he PCR eac ion. In o de o ob ain an equi alen le el o sensi i i y o ha o Yokoigawa e al., he ollowing modi ica ions we e equi ed. Fi s , an inc ease in PCR cycle numbe om 25 o 35 cycles had a signi ican e ec , esul ing in a leas a 10 6 inc ease in sensi i i y o de ec ion. Secondly, he Fig. 2 PCR-ELISA analysis o aw milk (j), pas eu ized milk ( ) and PBST (h) con- aining 0–500 c u Esche ichia coli ob ained ollowing alkaline ex ac ion p ocedu e. Neg- a i e and posi i e con ols a e also indica ed. Assay cu -o (ho izon al line) was calcula ed by eplica e analysis o eigh nega i e speci- mens and was de e mined o be 0Æ105 O.D. uni s (mean nega i es + 2 s anda d de ia ions, i.e. 0Æ065 + 2 (0Æ02)) E. COLI PCR-ELISA 225 ª2002 The Socie y o Applied Mic obiology, Le e s in Applied Mic obiology,34, 222–226 wash/cen i uga ion s ep p io o ex ac ion was included o enhance he eco e y o bac e ia and also o help emo e he high le els o a and p o ein. In e es ingly, he elimina ion o po en ial PCR inhibi o s by he inclusion o a washing s ep while eco e ing he bac e ial pelle p io o DNA ex ac ion was also achie ed by Uy endaele e al. (1999). Thi dly, ho -s a PCR was used o p e en educed ac i i y o Taq polyme ase du ing he ini ial dena u a ion s ep. Finally, be aine (1 mol l )1 ) was included as a PCR addi i e o enhance ampli ica ion. In es iga ion in o he ailu e o ampli ica ion using he me hod o Yokoigawa e al. (1999) illus a ed ha e en when 10 6 c u E. coli we e added o he PCR mas e mix con aining 5 ll o de e gen ex ac ion bu e , inhibi ion o ampli ica ion esul ed. The highe sensi i i y o E. coli de ec ion in pas eu ized milk samples using alkaline ex ac ion can be explained by enhanced emo al o inhibi o s, he eby acili a ing inc eased ampli ica ion e iciency. Fu he mo e, ailu e o sample hea ing alone o elease de ec able E. coli DNA om milk u he highligh s he impo ance o he emo al o inhibi o y compounds p io o ampli ica ion. The de elopmen o his no el PCR-ELISA me hod o e s se e al ad an ages o e adi ional mic obiological cul u e echniques. The PCR s ep akes jus o e 1 h o comple e due o he simul aneous annealing and polyme - iza ion a 72°C. Speci ic amplicon de ec ion ia he DNP oligonucleo ide p obe used in he ELISA con e s inc eased assay speci ici y wi h op imal sensi i i y. The me hod is s aigh o wa d, does no equi e expensi e equipmen and should be capable o de ec ing he sugges ed E. coli O157 in ec i e dose (4–24 o ganisms) as p oposed by S achan e al. (2001). Au oma ion and mul iple sample analysis can also be mo e easily pe o med on s ep a idin-coa ed mic o i e pla es in compa ison wi h aga ose gel elec o- pho esis. In addi ion, he subjec i i y o band iden i ica ion and he use o ca cinogenic DNA binding agen s a e also elimina ed. This no el PCR-ELISA should help in he assessmen o he hygienic quali y o milk collec ed om a ms and acili a e apid diagnosis o E. coli mas i is, which would allow o ea ly an ibio ic he apy helping o educe loss in milk p oduc ion and damage o mamma y issue. ACKNOWLEDGEMENTS The inancial assis ance o En e p ise I eland and T idel a De elopmen Limi ed is g a e ully acknowledged. REFERENCES Allmann, M., Ho ¨ elein, C., Ko ¨ppel, E. e al. (1995) Polyme ase chain eac ion (PCR) o de ec ion o pa hogenic mic oo ganisms in bac e iological moni o ing o dai y p oduc s. Resea ch in Mic obio- logy 146, 85–97. Bee ens, H., Hass B ac de la Pe ie e, B. and Ga ini, F. (2000) E alua ion o he hygienic quali y o aw milk based on he p esence o bi idobac e ia: he cow as a sou ce o aecal con amina ion. In e na ional Jou nal o Food Mic obiology 54, 163–169. de Boe , E. (1998) Upda e on media o isola ion o En e obac e iaceae om oods. In e na ional Jou nal o Food Mic obiology 45, 43–53. F a amico, P.M., Bagi, L.K. and Pepe, T. (2000) A mul iplex polyme ase chain eac ion assay o apid de ec ion and iden i ica ion o Esche ichia coli O157:H7 in oods and bo ine eces. Jou nal o Food P o ec ion 63, 1032–1037. Gonza ´lez, I., Ga cia, T., Fe nandez, A., Sanz, B., He nandez, P.E. and Ma in, R. (1999) Rapid enume a ion o Esche ichia coli in oys e s by a quan i a i e PCR-ELISA. Jou nal o Applied Mic obiology 86, 231– 236. Hsu, S.-C. and Tsen, H.-Y. (2001) PCR p ime s designed om malic acid dehyd ogenase gene and hei use o de ec ion o Esche ichia coli in wa e and milk samples. In e na ional Jou nal o Food Mic obiology 64, 1–11. Ishizawa, M., Kobayashi, Y., Miyamu a, T. and Ma suu a, S. (1991) Simple p ocedu e o DNA isola ion om human se um. Nucleic Acids Resea ch 19, 5792. Mana i, M. (2000) New de elopmen s in ch omogenic and luo ogenic cul u e media. In e na ional Jou nal o Food Mic obiology 60, 205– 218. McKillip, J.L. and D ake, M. (2000) Molecula beacon polyme ase chain eac ion de ec ion o Esche ichia coli O157:H7 in milk. Jou nal o Food P o ec ion 63, 855–859. Milla , B.C., Ji u, X., Moo e, J.E. and Ea le, J.A.P. (2000) A simple and sensi i e me hod o ex ac bac e ial, yeas and ungal DNA om blood cul u e ma e ial. Jou nal o Mic obiological Me hods 42, 139–147. Sco e , S., Ald idge, M. and Capps, K. (2000) Valida ion o a me hod o he de ec ion o E. coli O157:H7 in oods. Food Con ol 11, 85–95. Seo, K.H., B acke , R.E. and F ank, J.F. (1998) Rapid de ec ion o Esche ichia coli O157:H7 using immuno-magne ic low cy ome y in g ound bee , apple juice, and milk. In e na ional Jou nal o Food Mic obiology 44, 115–123. S achan, N.J.C., Fenlon, D.R. and Ogden, I.D. (2001) Modelling he ec o pa hway and in ec ion o humans in an en i onmen al ou b eak o Esche ichia coli O157. FEMS Mic obiology Le e s 203, 69–73. Uy endaele, M., an Boxs ael, S. and Debe e e, J. (1999) PCR assay o de ec ion o he E. coli O157:H7 eae-gene and e ec o he sample p epa a ion me hod on PCR de ec ion o hea -killed E. coli O157:H7 in g ound bee . In e na ional Jou nal o Food Mic obiology 52, 85–95. Venka eswa an, K., Mu akoshi, A. and Sa ake, M. (1996) Compa ison o comme cially a ailable ki s wi h s anda d me hods o de ec ion o coli o ms and Esche ichia coli in oods. Applied and En i onmen al Mic obiology 62, 2236–2243. Yokoigawa, K., Inoue, K., Okubo, Y. and Kawai, H. (1999) P ime s o ampli ying an alanine acemase gene agmen o de ec E. coli s ains in oods. Jou nal o Food Science 64, 571–575. Yu, H. (1998) Compa a i e s udies o magne ic pa icle-based solid phase luo ogenic and elec ochemiluminescen immunoassay. Jou - nal o Immunological Me hods 218, 1–8. 226 P. DALY ET AL. ª2002 The Socie y o Applied Mic obiology, Le e s in Applied Mic obiology,34, 222–226