Enumeration of heterotrophs, fecal coliforms and Escherichia coli

Enumeration of heterotrophs,fecal coliforms and Escherichia coli in water:comparison of 3M k Petrifilm k plates with standardplating proceduresH.Schraft *,L.A.WatterworthDepartment of Biology,Lakehead University,Thunder Bay,ON,Canada P7B 5E1Received 12July 2004;received in revised form 12October 2004;accepted 12October 2004Available online 10November 2004AbstractA total of 177naturally contaminated water samples were analyzed by membrane filtration according to the Standard Methods for the Examination of Water and Wastewater published by the American Public Health Association.Filters were incubated in parallel on mHPC-agar and 3M k Petrifilm k Aerobic Count Plates (Petrifilm k AC plates)for heterotrophic counts.Fecal coliforms and Escherichia coli were enumerated on mFC-agar and 3M k Petrifilm k E.coli /Coliform Count Plates (Petrifilm k EC plates).Typical colonies on each media type were confirmed following standard procedures.Heterotrophic counts were between 103and 104CFU/mL and the average log 10counts obtained on Petrifilm k AC plates were about two-fold lower than on mHPC-agar.Counts for fecal coliforms and E.coli were between 102and 103CFU/mL.Average log 10counts for confirmed fecal coliforms obtained on Petrifilm k EC plates were slightly lower than on mFC agar with a correlation coefficient of 0.949.The average log 10counts for confirmed E.coli on Petrifilm k EC plates and on mFC agar were statistically not different (P =0.126)with a correlation coefficient of 0.879.Specificity of Petrifilm k EC plates and mFC agar was evaluated by comparing typical colony counts with confirmed counts.On mFC agar,counts for typical colonies were by 2log 10CFU higher than the actual confirmed counts.In contrast,on Petrifilm k EC plates typical colony counts were almost identical to confirmed colony counts for both fecal coliforms and E.coli .This comparison illustrates the high specificity of Petrifilm k EC plates for enumeration of both fecal coliforms and E.coli in water.D 2004Elsevier B.V .All rights reserved.Keywords:3M k Petrifilm k plates;Water microbiology;Heterotrophic counts;Fecal coliforms;E.coli1.IntroductionThe waterborne outbreak linked to enterohemor-rhagic Escherichia coli in 2000in Walkerton,Ontario (Anonymous,2000)has resulted in heightened surveillance of the microbiological quality of surface0167-7012/$-see front matter D 2004Elsevier B.V .All rights reserved.doi:10.1016/j.mimet.2004.10.008*Corresponding author.Tel.:+18073438351;fax:+18073467796.E-mail address:heidi.schraft@lakeheadu.ca (H.Schraft).Journal of Microbiological Methods 60(2005)335–342and drinking water.New regulations are implemented in Canada and in many other countries and in Ontario, they have resulted in a significant increase of micro-biological water analyses(Ontario Safe Drinking Water Act,2002).Enumeration of total heterotrophic counts is commonly used as indicator of overall microbiolog-ical quality(APHA,1995;WHO,1993).To ensure absence of enteric pathogens,enumeration of appro-priate indicator organisms is required.Some contro-versy still exists as to which group of organisms would be most suitable,but total coliforms,fecal coliforms and E.coli have been suggested.Total coliforms are defined as aerobic or facultatively anaerobic,Gram-negative rod-shaped,non-spore-forming,oxidase negative bacteria that ferment lactose at358C with the production of acid and gas (APHA,1995;WHO,1993).Since coliforms are defined by biochemical properties,the group does not have taxonomic value and organisms of many genera, such as Escherichia,Citrobacter,Klebsiella or Serratia,are included.Total coliform counts are used to monitor treated water supplies with the objective to determine adequacy of the water treatment process and the integrity of the distribution system(APHA, 1995;WHO,1993).To avoid the limitations of total coliforms,monitoring of fecal(or thermotolerant) coliforms has been suggested.This group includes those coliforms that ferment lactose at448C. Although many coliforms are not active at the elevated temperature,non-fecal organisms,such as some species of Klebsiella,Enterobacter or Citro-bacter,are detected by fecal coliform tests.Since levels of fecal coliforms are mostly directly correlated to those of E.coli,fecal coliform counts have been widely accepted for routine monitoring of water quality.However,confirmation of isolates as E.coli by additional biochemical tests is recommended (WHO,1993).Today,E.coli is generally recognized as a good indicator for fecal contamination and it has been recommended as the indicator of choice for monitoring of drinking water(WHO,1993).Standard methods for the enumeration of hetero-trophs,fecal coliforms and E.coli in water rely on membrane filtration with subsequent incubation of the filters on solid agar media.Preparation of such media is time consuming and expensive.Thus,ready-made media that would reduce time and labor required for analyses are needed,particularly in light of the drastic increase of requirements for microbiological water testing.3M k Petrifilm k plates provide such ready-to-use plating media which are widely accepted and approved for microbiological analysis in the food and beverage industry.In particular,3M k Petrifilm k Count Plates could be useful as alternative for heterotrophic counts and3M k Petrifilm k E.coli/ Coliform Count Plates for fecal coliforms and E.coli. However,only limited information has been published to date that allows an evaluation of Petrifilm k plates for microbiological analysis of water by membrane filtration.The objective of this study was therefore to evaluate the application of ready-to-use3M k Pet-rifilm k plates for the enumeration of heterotrophic counts,fecal coliforms and E.coli in naturally contaminated wastewater.2.Materials and methods2.1.Water samplesA total of177water samples were received from five different wastewater treatment facilities in South-ern Ontario.The water samples were collected by plant personnel into sterilized500mL plastic bottles, refrigerated and shipped to the laboratory in cooled insulated containers.To ensure countable levels of microorganisms,all samples were taken before any disinfection step.2.2.Membrane filtrationWithin24h of collection,samples were analyzed by membrane filtration method for heterotrophic counts and fecal coliforms following the Standard Procedures for Water Analysis published by the American Public Health Association(APHA)(1995).The samples were serially diluted in0.1% peptone water and the dilutions filtered onto mem-brane filters(47mm diameter and0.45A m pore size).Cellulose acetate filters(AMD Manufacturing, Mississauga,Ontario)were used for heterotrophic counts and fecal coliform counts were determined on nitrocellulose filters(Millipore).After filtration,each filter was aseptically placed on the appropriate growth medium.H.Schraft,L.A.Watterworth/Journal of Microbiological Methods60(2005)335–342 3362.3.Heterotrophic countsHeterotrophic counts were determined in parallel on mHPC agar(Difco)prepared according to manu-facturer’s instructions and3M k Petrifilm k Count Plates(Petrifilm k AC plates).Petrifilm k AC plates were hydrated by adding1mL of sterile0.1%peptone water1h prior to use.Each sample was plated in duplicate onto both media types.The plates were incubated at35F0.58C for48F2h.Plates of those dilutions yielding20–200 colonies were enumerated.2.4.Counts for fecal coliforms and E.coliCounts for fecal coliforms and E.coli were determined in parallel on mFC agar(Difco)prepared according to manufacturer’s instructions and on3M k Petrifilm k E.coli/Coliform Count Plates(Petrifilm k EC plates).Petrifilm k EC plates were hydrated by adding1mL of sterile0.1%peptone water one hour prior to use.Each sample was plated in duplicate onto both media types.Both mFC agar plates and Petrifilm k EC plates were sealed in plastic bags and incubated immersed in a waterbath at44.5F0.2 8C for24F2h.Plates of those dilutions yielding20–200colonies were enumerated.On mFC agar,all blue colonies were counted and recorded as typical fecal coliforms.All other colonies (yellow or gray in color)were also counted and recorded as atypical colonies.On Petrifilm k EC plates,red colonies with gas bubbles were counted as typical fecal coliforms and blue colonies with gas bubbles as typical E.coli.Total fecal coliforms were calculated by adding numbers for blue and red colonies with gas.Colonies without associated gas bubbles(either red or blue)were recorded as atypical colonies.2.5.Biochemical identification of fecal coliforms andE.coliFor plates containing100colonies or less,10 typical and five atypical colonies were selected and subcultured on Nutrient Agar(NA)at35F18C for24 h.From plates containing more than100colonies, 10%typical and5%atypical colonies were selected. Pure sub-cultures on nutrient agar were grown in Tryptic Soy Broth(TSB)at35F18C for24h and then tested in lauryl tryptose broth and EC medium (Lauryl/EC test)according to the Standard Procedures for Water Analysis,sections9221B and9221E (APHA,1995).Pure cultures grown on NA were used for Oxidase test and determination of IMViC pattern(indole,methyl red,V oges Proskauer and citrate utilization test)following Standard Procedures for Water Analysis,section9225A(APHA,1995).Gram negative,oxidase negative rods with a positive Lauryl/EC test were considered confirmed fecal coliforms and Gram negative,oxidase negative rods with the IMViC pattern[++ÀÀ]were recorded as confirmed E.coli.2.6.Statistical analysisAll CFU counts per mL were converted to log10 counts before statistical analysis.For each plate-method combination,a paired t-test was performed on the differences in the average log10counts between each of the two media types.The software package SigmaStat,Version 2.03(SPSS)was used for all analyses.3.Results and discussion3.1.Heterotrophic countsA total of54samples were analyzed and the heterotrophic counts were between103and104 CFU/mL(data not shown).The average log10counts obtained on Petrifilm k AC plates were about two-fold(0.5log10)lower than on mHPC agar.Even though this difference is small,a paired t-test showed that it is statistically significant(P b0.001).In addition,the correlation coefficient between the two media types was only0.775.This is below the range of0.94–0.99found in previous studies with foods where aerobic colony counts on Petrifilm k AC plates were compared to counts obtained on plate count agar(Blackburn et al.,1996).The difference in counts observed with water samples is likely related to the composition of the two media compared.Petrifilm k AC plates have been designed specifically for the enumeration of aerobic bacteria in the food and dairy industry containing StandardH.Schraft,L.A.Watterworth/Journal of Microbiological Methods60(2005)335–342337Methods nutrients.Standard Methods agar contains much less enzymatic protein digests (5g/L)and sugar (1g/L)than mFC agar (15and 12.5g,respectively).Since the microflora in water is different from that in food,the lower protein digest concentration in Standard Methods plates may not be sufficient in supporting growth of a few organisms present in water samples.This is particularly important for analysis of treated drinking water,where heterotrophic counts are often used as indicator.Therefore,lower counts could be expected using a medium such as Petrifilm k AC plates,which was not designed for the growth of hetero-trophic bacteria found in water samples.3.2.Counts for fecal coliforms and E.coliCounts for fecal coliforms and E.coli of the 177samples analyzed were between 102and 103CFU/mL (Fig.1).Average log 10counts for confirmed fecal coliforms obtained on Petrifilm k EC plates were slightly (0.2log 10)lower than on mFC agar.The correlation coefficient between the two methods was 0.949(Fig.2).This is better than coefficients reported for method comparisons enumerating coliforms in foods,which ranged from 0.85to 0.93(Blackburn et al.,1996).Nevertheless,a paired t -test showed that the small difference between the two procedures was statistically significant (P b 0.001).The average log 10counts for confirmed E.coli on Petrifilm k EC plates were also lower (0.04log 10)than on mFC agar.However,this difference was not significant based on a paired t -test (P =0.126).The correlation coefficient between these two media types was 0.879(Fig.2).Baumgartner et al.(1993)have compared E.coli counts of surface water obtained on E.coli Direct (ECD)agar and Petrifilm k EC plates.They also found lower E.coli counts on Petrifilm k plates than on ECD agar.To estimate the relevance of this difference,the authors calculated that the ratio of log 10CFU on ECD/log 10CFU on Petrifilm k plates.Since this ratio was below three for 91.6%of the samples,it was concluded that this variation was irrelevant for microbiological analysis of water.The ratios of log 10CFU on mFC agar/log 10CFU on Petrifilm k plates are shown in Fig.3for both confirmed fecal coliforms and confirmed E.coli counts.In both cases,98%of the samples differed by only a factor of two or less.Thus,based on the criteria suggested by Baumgartner et al.(1993),Petrifilm k EC plates may be a useful alternative medium for the enumeration of fecal coliforms and E.coli in water.Petrifilm k EC plates have been designed for direct enumeration of coliforms and E.coli in food,eliminating the need for colony confirmations.The plates contain violet red bile nutrients,a gellingagent,Fig.1.Counts for confirmed fecal coliforms and E.coli obtained on mFC and Petrifilm k EC plates.mFC=mFC agar plates.EC=Petrifilm k EC plates.H.Schraft,L.A.Watterworth /Journal of Microbiological Methods 60(2005)335–342338an indicator of beta-glucuronidase activity and a metabolic indicator to facilitate colony enumeration.Typical E.coli on Petrifilm k EC plates appear as blue colonies with gas based on the facts that about 97%of E.coli produce beta-glucuronidase (resulting in a blue precipitate associated with a colony)and that about 95%of E.coli produce gas.Coliform colonies appear on Petrifilm k EC plates as red colonies with gas.Acid and gas production from lactose during metabolic fermentation cause the pH indicator to change the color of the gel to a darker red and the gas surrounding red coliform colonies indicatescon-Fig.2.Correlation of counts for confirmed fecal coliforms and E.coli obtained on mFCagar and Petrifilm k EC plates.H.Schraft,L.A.Watterworth /Journal of Microbiological Methods 60(2005)335–342339firmed coliforms.To obtain total coliform counts,numbers for typical E.coli and typical coliforms are added.Since the Petrifilm k EC plates were incu-bated at 44.58C in this study,the result of the added typical colony counts represents fecal coliforms.The specificity of the Petrifilm k EC plates and the mFC agar was evaluated by comparing typical colony counts with confirmed counts (Fig.4).On mFC agar,counts for typical colonies were by 2log 10CFU higher than the actual confirmed counts.In contrast,on Petrifilm k EC plates,typical colony counts were almost identical to confirmed colony counts for both fecal coliforms and E.coli .This comparison illus-trates the high specificity of Petrifilm k EC plates for enumeration of both fecal coliforms and E.coli in water.3.3.Biochemical reactions of isolates from mFC agar and Petrifilm k EC platesThe biochemical tests performed on typical colonies isolated from each media type substantiate the high specificity of Petrifilm k EC plates.Colonies from Petrifilm k EC plates were identified as fecal coliforms at a much higher rate (87.1%)than colonies from mFC plates (68.5%).The rates of E.coli among the typical fecal coliform colonies were 62.1%for mFC agar and 79.2%for Petrifilm k EC plates (Table 1).Representatives of atypical colonies grown on mFC and Petrifilm k EC plates were also isolated and subjected to biochemical testing (Table 2).Of the 796atypical colonies tested from mFC plates,12.8%were identified as fecal coliforms and 29.8%as E.coli .On Petrifilm k EC plates,larger proportions of atypical colonies were identified as fecal coliforms (30.5%)and E.coli (73.3%),respectively.The higher rateofFig.3.Distribution of ratio log 10CFU on mFC agar/log 10CFU on Petrifilm k EC plates for confirmed fecal coliforms and E.coli.Fig.4.Difference between typical colony counts and confirmed colony counts obtained on mFC agar and Petrifilm k EC plates.H.Schraft,L.A.Watterworth /Journal of Microbiological Methods 60(2005)335–342340false negatives on Petrifilm k EC plates is likely related to the definition of b typical colonies Q used for these plates.On Petrifilm k EC plates,a colony typical for E.coli(blue with gas)is determined by a positive beta-glucuronidase activity associated with gas pro-duction.However,beta-glucuronidase negative E.coli have been reported for6–15%of human isolates (Rompre´et al.,2002),for22%of isolates from humans and various foods(Huang et al.,1997)and for6%of water isolates.In addition,rates for anaerogenic E.coli are reported to reach up to74%, depending of source(Kornacki and Johnson,2001).The limitations of the definitions for coliforms and E.coli used in this and many other studies are illustrated by discrepancies observed for identifica-tions of coliforms and E.coli.Since E.coli is considered a subgroup of coliforms,one would expect higher confirmation rates for coliforms than for E. coli.However,of the1326colonies appearing as typical E.coli on Petrifilm k EC plates(blue colonies with gas),93.4%were confirmed biochemically as E. coli,but only91.6%were confirmed as fecal coli-forms(Table1).Similarly,for atypical colonies from mFC and Petrifilm k EC plates,17%(mFC)and 42.7%(Petrifilm k EC plates)of confirmed E.coli were not confirmed as fecal coliforms(Table2).This discrepancy may be due to the generally accepted biochemical protocols used to determine whether an isolate is E.coli or a coliform.According to these procedures,which are based on the Standard Methodsfor the Examination of Water and Wastewaterpublished by the American Public Health Association(APHA)(1995),only isolates producing gas inLauryl/EC broth are considered coliforms.b True Q E.coli are Gram negative rods with a negative oxidasereaction and the IMViC pattern[++ÀÀ].However,O’Toole and Chiang(1999)have reported that15.2%of E.coli isolates from marine waters did not producegas in Lauryl/EC broth.Thus,it is possible that thosecolonies identified as E.coli,but not as coliforms,belonged to the group of anaerogenic group E.coli,which have the IMViC pattern[++ÀÀ],but do notproduce gas in Lauryl/EC broth.During routine water analysis,laboratories do notalways confirm the identity of typical isolates andresults would be reported based on b typical counts Qrather than b confirmed counts Q.When comparing theconfirmed E.coli and fecal coliform counts,a muchhigher percentage of typical colonies grown onPetrifilm k EC plates were confirmed as fecal coli-forms than on mFC agar.This will result in higheraccuracy of routine water analysis.Assuming a limitfor fecal coliforms of100CFU/mL,78.1%of thesamples tested would have exceeded the limit usingb typical counts Q on mFC plates,but only69.1%when using b confirmed counts Q.With Petrifilm k EC plates,this difference is only0.5%(65.2%for b typicalcounts Q and65.7%for b confirmed counts Q).Table1Identification of typical colonies isolated from mFC and Petrifilm k EC platesMedium Colony type Number of colonies tested Number(%)of colonies identified asFecal coliforms E.colimFC plates Blue19681348(68.5%)1222(62.1%) Petrifilm k EC plates Blue and red with gas16391428(87.1%)1298(79.2%) Red with gas313259(82.7%)203(64.9%)Blue with gas13261215(91.6%)1238(93.4%) On mFC plates,all blue colonies are considered fecal coliforms.On Petrifilm k EC plates,red colonies with gas are considered fecal coliforms and blue colonies with gas are considered E.coli.Total fecal coliforms are thus represented by blue and red colonies with gas.Table2Identification of atypical colonies isolated from mFC and Petrifilm k EC platesMedium Colony type Number of colonies tested Number(%)of colonies identified asFecal coliforms E.coli mFC Plates atypical796102(12.8%)237(29.8%) Petrifilm k EC plates atypical674206(30.6%)494(73.3%)H.Schraft,L.A.Watterworth/Journal of Microbiological Methods60(2005)335–342341In conclusion,Petrifilm k EC plates seem more selective for fecal coliforms and E.coli.They also allow the simultaneous identification of fecal coli-forms and E.coli and a very high percentage of suspect colonies are confirmed by standard biochem-ical tests.Petrifilm k EC plates facilitate separate enumeration of colonies positive for lactose(red), beta-glucuronidase(blue),and gas-production(gas associated with colony)and any combination of characteristics can be used to identify suspect colo-nies.In particular,Petrifilm k EC plates allow detection of beta-glucuronidase positive isolates as in other proposed media for direct enumeration of E. coli in water(Ciebin et al.,1995).Based on their high specificity and ability to simultaneously identify fecal coliforms and E.coli,Petrifilm k EC plates can be used as a cost-effective alternative to screen water samples for these organisms. AcknowledgementsThis work was supported by3M Canada(London, ON).We are grateful to H.Chung and J.Narayansingh for excellent technical assistance.Michael G.Williams, Kathryn Lindberg and Theresa Hackett have provided valuable input for the design and evaluation of the study,as well as for the preparation of the manuscript. ReferencesAmerican Public Health Association(APHA),1995.Standard Methods for the Examination of Water and Wastewater,19th ed.APHA,Washington,DC.Anonymous,2000(15October).Waterborne outbreak of gastro-enteritis associated with a contaminated municipal water supply, Walkerton,Ontario,May–mun.Dis.Rep.26-20,170–173.Baumgartner, A.,Grand,M.,Simmen, A.,Halvax,M.,1993.Quantitative analysis of E.coli in water-comparison of ECD-agar and Petrifilm k.Mitt.Geb.Lebensm.unters.Hyg.84, 382–387.Blackburn,C.de W.,Baylis,C.L.,Petitt,S.B.,1996.Evaluation of Petrifilm k methods for enumeration of aerobic flora and coliforms in a wide range of foods.Lett.Appl.Microbiol.22, 137–140.Ciebin,B.W.,Brodsky,M.H.,Eddington,R.,Horsnell,G.,Choney,A.,Palmateer,G.,Ley, A.,Joshi,R.,Shears,G.,1995.Comparative evaluation of m-FC and m-TEC media for membrane filter enumeration of Escherichia coli in water.Appl.Environ.Microbiol.61,3940–3942.Huang,S.W.,Chang,C.H.,Tai,T.F.,Chang,T.C.,par-ison of the beta-glucuronidase assay and the conventional method for identification of Escherichia coli on eosin-methyl-ene blue agar.J.Food Prot.60,6–9.Kornacki,J.L.,Johnson,J.L.,2001.Enterobacteriaceae,coliforms and Escherichia coli as quality and safety indicators.In: Downes,F.P.,Ito,K.(Eds.),Microbiological Examination of Foods,4th edition.American Public Health Association, Washington,DC.O’Toole,D.K.,Chiang,M.M.P.,1999.The use of MUG supplement to detect Escherichia coli by the multiple tube method in marine waters of Hong Kong.Mar.Pollut.Bull.38,921–924. Ontario Safe Drinking Water Act.2002.Statutes of Ontario 2002,Chapter32.Available at:http://192.75.156.68/DBLaws/ Statutes/English/02s32_st accessed October3,2003. Rompre´,A.,Servais,P.,Baudart,J.,de-Roubin,M.,Laurent,P., 2002.Detection and enumeration of coliforms in drinking water: current methods and emerging approaches.J.Microbiol.Methods49,31–54.WHO,1993.WHO Guidelines for Drinking Water Quality, 2nd ed.Recommendations,vol. 1.WHO,Geneva.H.Schraft,L.A.Watterworth/Journal of Microbiological Methods60(2005)335–342 342。