Journal of Microbiological Methods 75 (2008) 506514

Contents lists available at ScienceDirect

Journal of Microbiological Methods

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j m i c m e t h

Analytical limits of four -glucuronidase and -galactosidase-based commercial culture methods used to detect Escherichia coli and total coliforms
Andre F. Maheux a,b, Vicky Hupp a,b, Maurice Boissinot a,b, Franois J. Picard a,b, Luc Bissonnette a,b, Jean-Luc T. Bernier a,b, Michel G. Bergeron a,b,
a b

Centre de Recherche en Infectiologie de l'Universit Laval, Centre Hospitalier Universitaire de Qubec (Pavillon CHUL), Qubec (Qubec), Canada Dpartement de Biologie Mdicale, Facult de Mdecine, Universit Laval, Qubec (Qubec), Canada

a r t i c l e

i n f o

a b s t r a c t
Colilert (Colilert), Readycult Coliforms 100 (Readycult), Chromocult Coliform agar ES (Chromocult), and MI agar (MI) are -galactosidase and -glucuronidase-based commercial culture methods used to assess water quality. Their analytical performance, in terms of their respective ability to detect different strains of Escherichia coli and total coliforms, had never been systematically compared with pure cultures. Here, their ability to detect -glucuronidase production from E. coli isolates was evaluated by using 74 E. coli strains of different geographic origins and serotypes encountered in fecal and environmental settings. Their ability to detect -galactosidase production was studied by testing the 74 E. coli strains as well as 33 reference and environmental non-E. coli total coliform strains. Chromocult, MI, Readycult, and Colilert detected glucuronidase production from respectively 79.9, 79.9, 81.1, and 51.4% of the 74 E. coli strains tested. These 4 methods detected -galactosidase production from respectively 85.1, 73.8, 84.1, and 84.1% of the total coliform strains tested. The results of the present study suggest that Colilert is the weakest method tested to detect -glucuronidase production and MI the weakest to detect -galactosidase production. Furthermore, the high level of false-negative results for E. coli recognition obtained by all four methods suggests that they may not be appropriate for identication of presumptive E. coli strains. 2008 Elsevier B.V. All rights reserved.

Article history: Received 14 July 2008 Received in revised form 4 August 2008 Accepted 5 August 2008 Available online 8 August 2008 Keywords: E. coli MI agar Colilert Chromocult Coliform agar ES Readycult Coliforms 100 Total coliforms

1. Introduction The multiple-tube fermentation and membrane lter techniques are classical reference methods used for water quality monitoring (APHA et al., 1998) that have been associated with their own limitations. The multiple-tube fermentation method provides results only after 3 to 4 days and the interference by a high number of noncoliform bacteria have been shown to alter the efciency of the analysis (Evans et al., 1981; Means and Olson, 1981; Seidler et al., 1981). For the membrane lter technique, the most widely used medium for drinking water analysis are m-Endo and mFC media in United States and Canada (APHA et al., 1998) and Tergitol-TTC medium in Europe (AFNOR, 1990). However, since these media lack specicity, coliform conrmation is required (APHA et al., 1998) which delays the results by 2 to 3 days. Also, the presence of a high number of background heterotrophic bacteria was shown to decrease coliform recovery (Burlingame et al., 1984; Clark, 1980). The inherent limita-

Corresponding author. Centre de Recherche en Infectiologie de l'Universit Laval, Centre Hospitalier Universitaire de Qubec (Pavillon CHUL), 2705 Laurier Blvd. Suite RC709, Qubec (Qubec), Canada G1V 4G2. Tel.: +1 418 654 2705; fax: +1 418 654 2715. E-mail address: Michel.G.Bergeron@crchul.ulaval.ca (M.G. Bergeron). 0167-7012/$ see front matter 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.mimet.2008.08.001

tions of these two methods make them unable to provide, within hours, useful public health information. To diminish background effects of heterotrophic bacteria and circumvent the need for a conrmation stage required by both multiple-tube fermentation and membrane lter techniques, methods based on the enzymatic properties of coliforms (-galactosidase for total coliforms and -glucuronidase enzymes for Escherichia coli detection) have been developed. These enzymes have been chosen because conventional coliform monitoring is based on detection of the presence of -galactosidase and because the gene encoding the glucuronidase enzyme (uidA) was found to be specic (Brenner et al., 1972) and present in more than 97% of E. coli isolates (Lupo and Halpern, 1970; Martins et al., 1993). Colilert (Colilert, IDEXX Laboratories, Westbrook, ME, USA), Readycult Coliforms 100 (Readycult; Merk KGaA, Darmstadt, Germany), Chromocult Coliform agar ES (Chromocult; Merk KGaA, Darmstadt, Germany), and MI agar (MI; BD, Franklin Lakes, NJ, USA) are four commercial test methods based on the determination of galactosidase and -glucuronidase enzyme activities which are used to detect, within 24 h, total coliforms and E. coli in water samples. These tests are easy to use, require no additional conrmatory step, and provide a more rapid estimate of indicators of bacteriological contamination of water as compared to classical techniques (Brenner

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514 Table 1 Ability of Colilert, Readycult Coliforms 100, Chromocult Coliform agar ES and MI agar test methods to detect Escherichia coli and Shigella sp. strains Strains (origin) Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia Escherichia coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli coli (n = 74) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) No. reference ATCC 11775 ATCC 23511 ATCC 35401 ATCC 43886 ATCC 43890 ATCC 43894 ATCC 43895 ATCC 43896 CCRI-1191 CCRI-1192 CCRI-1193 CCRI-1213 CCRI-2099 CCRI-2105 CCRI-2106 CCRI-2107 CCRI-2108 CCRI-2109 CCRI-2166 CCRI-2202 CCRI-8825 CCRI-8826 CCRI-8831 CCRI-8832 CCRI-8833 CCRI-8834 CCRI-8835 CCRI-8836 CCRI-8837 CCRI-8838 CCRI-8839 CCRI-8840 CCRI-8852 CCRI-9493 CCRI-14813 CCRI-14858 CCRI-14859 CCRI-14871 CCRI-14881 CCRI-16465 CCRI-16485 CCRI-16527 CCRI-16528 CCRI-16537 CCRI-16539 CCRI-16540 CCRI-16579 CCRI-16580 CCRI-17006 CCRI-17021 CCRI-17027 CCRI-17042 CCRI-17045 CCRI-17056 CCRI-17063 CCRI-17065 CCRI-17097 CCRI-17151 CCRI-17158 CCRI-17161 CCRI-17172 CCRI-17176 LSPQ 2086 LSPQ 2092 LSPQ 2096 LSPQ 2113 LSPQ 2115 LSPQ 2117 LSPQ 2118 LSPQ 2125 LSPQ 2127 LSPQ 3760 LSPQ 3761 Serotype Test methods Colilert O1:K1:H7 O16:K1(L):NM O78:H11 O25:K98:NM O157:H7 O157:H7 O157:H7 O78:K80:H12 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A O157:H7 O157:H7 O103:H2 O103:H2 O111:HO111:HO26:NM O26:NM O145:NM O145:NM N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A O8:H9 O18:NM O26:NM O111:NM O128:H8 O113:H21 O117:H4 O128:NM O157:H7 O157:H7 O157:H7 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Readycult Coliforms 100 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Chromocult Coliform agar ES + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + MI agar + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

507

(continued on next page)

508 Table 1 (continued) Strains (origin) Escherichia coli (clinical) All E. coli strains All E. coli non-O157:H7 strains All environmental strains Shigella sp. (n = 8) Shigella boydii Shigella dysenteriae Shigella dysenteriae Shigella dysenteriae Shigella exneri Shigella exneri Shigella sonnei Shigella sonnei All Shigella strains No. reference LSPQ 3762

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514

Serotype O157:H7

Test methods Colilert 38/74 (51.4%) 38/65 (58.5%) 22/28 (78.6%) Readycult Coliforms 100 60/74 (81.1%) 58/65 (89.2%) 28/28 (100%) Chromocult Coliform agar ES 59/74 (79.7%) 59/65 (90.8%) 27/28 (96.4%) MI agar 59/74 (79.7%) 59/65 (90.8%) 28/28 (100%)

ATCC 9207 ATCC 11835 CCRI-8843 CCRI-8844 CCRI-2198 ATCC 12022 CCRI-2196 ATCC 29930

N/A type 1 N/A N/A N/A type 2b N/A N/A

0/8 (0%)

+ 1/8 (12.5%)

+ 1/8 (12.5%)

+ + + 3/8 (37.5%)

et al., 1993, 1996b; Edberg et al., 1988; Horman and Hanninen, 2006; Pitkanen et al., 2007). Different collections of strains were also tested with each commercial -galactosidase and -glucuronisade-based test methods to establish their ability to recover total coliforms and E. coli strains. All of these methods were found to be at least as efcient as classical reference methods (Landre et al., 1998; Rice et al., 1990, 1991, 1993). However, the expression of the -glucuronidase enzyme was found to be variable depending on the medium and technique used (Chang et al., 1989; Feng and Lampel, 1994; Shadix and Rice, 1991). Furthermore, there is no study comparing the analytical limits of these methods using pure cultures of bacteria, in order to determine their respective analytical ability to detect E. coli strains and total coliforms. In this study, we compared four commercial -galactosidase and -glucuronidase-based test methods (Colilert, Readycult, Chromocult agar, and MI agar) using a collection of bacteria representing strains of different geographical origins and serotypes obtained in both fecal and environmental settings. More specically, we have determined their ability to detect target E. coli and total coliform strains, as well as non-coliform strains (specicity testing). 2. Materials and methods

mentale du Qubec (Qubec, Qubec, Canada; n = 14). The analytical limits of the four culture methods were also assessed by using 8 Shigella strains representing four nomen species (Table 1). The ability of the four methods to detect total coliform strains was also veried by using 33 reference and environmental non-E. coli total coliform strains (Table 2) consisting of Citrobacter spp. (n = 12), Enterobacter spp. (n = 4), non-E. coli Escherichia spp. (n = 4), Hafnia spp. (n = 2), Klebsiella spp. (n = 5), Pantoae spp. (n = 1), Raoultella spp. (n = 3), Serratia spp. (n = 1), and Yersinia spp. (n = 1). The specicity of the four methods was demonstrated by using a battery of clinical and environmental strains consisting of 52 Grampositive bacterial species and 37 non-total coliforms Gram-negative bacterial species (Table 3). The identity of all reference, clinical and environmental strains used in this study was conrmed using a MicroScan Autoscan-4 system (Siemens Healthcare Diagnostic Inc., Newark, DE, USA) or with a Vitek 32 (bioMrieux SA, Marcy l'toile, France). Bacterial strains were grown from frozen stocks, kept at 80 C in brain heart infusion (BHI) medium (BD, Franklin Lakes, NJ, USA) containing 10% glycerol, and cultured on sheep blood agar. Three passages were performed prior to analysis of each strain with each culture method. 2.2. Bacterial cell suspension preparation

2.1. Bacterial strains The ability of the Colilert (Colilert), Chromocult coliforms agar ES (Chromocult), Readycult Coliform 100 (Readycult), and MI agar (MI) test methods to detect E. coli strains was veried using 74 strains of E. coli (Table 1). Eight (8) E. coli strains were obtained from the American Type Culture Collection (ATCC; Manassas, VA): ATCC 11775, ATCC 43886, ATCC 23511, ATCC 35401, ATCC 43890, ATCC 43894, ATCC 43895, and ATCC 43896. Clinical isolates of E. coli (n = 38) were obtained from Huashan Hospital (Shanghai, China; n = 1), Hpital Ambroise Par (Boulogne, France; n = 1), Institut fr Hygiene und Mikrobiologie der Universitt Wrzburg (Wrzburg, Germany; n = 10), Laboratoire de Sant Publique du Qubec (Sainte-Anne de Bellevue, Qubec, Canada; n = 12), National Institute of Public Health (Warsaw, Poland; n = 1), South African Institute for Medical Research (Johannesburg, South Africa; n = 2), Microbiology Laboratory of the Centre Hospitalier de l'Universit Laval (Qubec, Qubec, Canada; n = 5), University of Edinburgh (Edinburgh, Scotland; n = 5), and Wyeth-Ayerst Research (Pearl River, NY; n = 1). Environmental isolates of E. coli (n = 28) were obtained from various sources and isolated by different methods including Colilert, MI, Chromocult, mFC agar, and modied mTEC agar. These environmental strains were isolated from (i) drinking water samples obtained from the Service d'Analyse Environmentale Bodycote (Qubec, Qubec, Canada; n = 9), (ii) water samples from Bermuda (n = 5), (iii) river water samples across Canada, were obtained from the Centre d'Expertise en Analyse EnvironneEach bacteria described in Tables 13 was grown to logarithmic phase (0.50.6 OD600) and adjusted to a 0.5 McFarland standard, before being serially diluted ten-fold in phosphate-buffered saline (PBS; 137 mM NaCl, 6.4 mM Na2HPO4, 2.7 mM KCl, 0.88 mM KH2PO4, pH 7.4). For each bacterial strains, an aliquot of the 10 5 dilution was spiked in spring water (Labrador, Anjou, Qubec, Canada) ltered on a 0.22 m pore size membrane lter (Millipore Corporation, Billerica, MN, USA) to produce a suspension having approximately 100 colony forming unit (CFU)/100 mL of water. Bacterial count was veried by ltering 100 mL of each spiked water sample through a GN-6 membrane lter (47 mm diameter, 0.45 m pore size) with a standard platform manifold (Millipore Corporation, Billerica, MA, USA). Then, the lter was incubated on sheep blood agar plates for 24 2 h at 35.0 0.5 C prior to the determination of colony counts. Tests to conrm the sterility of lter membranes and buffer used for rinsing the ltration apparatus were also performed. 2.3. Membrane ltration methods For membrane ltration methods, two 100 mL volumes of each spiked water samples were ltered on GN-6 membrane lters with a standard platform manifold. One lter was incubated on Chromocult agar plates (Merck KGaA, Darmstadt, Germany) while the other lter was incubated on MI agar plates (BD) for 24 2 h at 35.0 0.5 C. Subsequently, colony count and color were determined for Chromocult

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514 Table 2 Ability of Colilert, Readycult Coliforms 100, Chromocult Coliform agar ES and MI agar test methods to detect total coliforms Strains (origin) E. coli (n = 74) Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli Escherichia coli No. reference (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (environmental) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) (clinical) ATCC 11775 ATCC 23511 ATCC 35401 ATCC 43886 ATCC 43890 ATCC 43894 ATCC 43895 ATCC 43896 CCRI-1191 CCRI-1192 CCRI-1193 CCRI-1213 CCRI-2099 CCRI-2105 CCRI-2106 CCRI-2107 CCRI-2108 CCRI-2109 CCRI-2166 CCRI-2202 CCRI-8825 CCRI-8826 CCRI-8831 CCRI-8832 CCRI-8833 CCRI-8834 CCRI-8835 CCRI-8836 CCRI-8837 CCRI-8838 CCRI-8839 CCRI-8840 CCRI-8852 CCRI-9493 CCRI-14813 CCRI-14858 CCRI-14859 CCRI-14871 CCRI-14881 CCRI-16465 CCRI-16485 CCRI-16527 CCRI-16528 CCRI-16537 CCRI-16539 CCRI-16540 CCRI-16579 CCRI-16580 CCRI-17006 CCRI-17021 CCRI-17027 CCRI-17042 CCRI-17045 CCRI-17056 CCRI-17063 CCRI-17065 CCRI-17097 CCRI-17151 CCRI-17158 CCRI-17161 CCRI-17172 CCRI-17176 LSPQ 2086 LSPQ 2092 LSPQ 2096 LSPQ 2113 LSPQ 2115 LSPQ 2117 LSPQ 2118 LSPQ 2125 LSPQ 2127 LSPQ 3760 LSPQ 3761 Test methods Colilert + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Readycult Coliforms 100 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Chromocult Coliform agar ES + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + MI agar + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

509

(continued on next page)

510 Table 2 (continued) Strains (origin) Escherichia coli (clinical) All E. coli strains All environmental strains Non-E. coli total coliforms (n = 33) Citrobacter amalonaticus (clinical) Citrobacter braakii (clinical) Citrobacter farmeri (clinical) Citrobacter freundii (clinical) Citrobacter freundii (environmental) Citrobacter gillenii (clinical) Citrobacter koseri (clinical) Citrobacter koseri (clinical) Citrobacter murliniae (clinical) Citrobacter sedlakii (clinical) Citrobacter werkmanii (clinical) Citrobacter youngae (food) Enterobacter aerogenes (clinical) Enterobacter cloacae (clinical) Enterobacter cloacae (environmental) Enterobacter sakazakii (environmental) Escherichia blattae (animal) Escherichia fergusonii (clinical) Escherichia hermannii (clinical) Escherichia vulneris (clinical) Hafnia alvei (clinical) Hafnia alvei (environmental) Klebsiella oxytoca (clinical) Klebsiella pneumoniae (clinical) Klebsiella pneumoniae (environmental) Klebsiella pneumoniae (environmental) Klebsiella pneumoniae (environmental) Pantoea agglomerans (clinical) Raoultella ornithinolytica (clinical) Raoultella planticola (environmental) Raoultella terrigena (environmental) Serratia marcescens (environmental) Yersinia enterocolytica (clinical) All non-E. coli total coliforms All environmental non-E. coli total coliforms All total coliforms All environmental total coliforms

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514

No. reference LSPQ 3762

Test methods Colilert + 70/74 (94.6%) 28/28 (100%) Readycult Coliforms 100 + 71/74 (95.9%) 28/28 (100%) Chromocult Coliform agar ES + 72/74 (97.3%) 28/28 (100%) MI agar + 64/74 (86.5%) 28/28 (100%)

ATCC 25405 ATCC 43162 ATCC 51112 ATCC 8090 CCRI-14856 ATCC 51117 ATCC 27156 ATCC 27028 ATCC 51641 ATCC 51115 ATCC 51114 ATCC 29935 ATCC 13048 ATCC 13047 CCRI-17108 CCRI-17037 ATCC 29907 ATCC 35469 ATCC 33650 ATCC 33821 ATCC 13337 CCRI-16651 ATCC 13182 ATCC 27799 CCRI-17014 CCRI-17064 CCRI-17074 ATCC 27155 ATCC 31898 ATCC 33531 ATCC 33257 ATCC 13880 ATCC 9610

+ + + + + + + + + + + + + + + + + + + + 20/33 (60.6%) 6/7 (85.7%) 90/107 (84.1%) 34/35 (97.1%)

+ + + + + + + + + + + + + + + + + + + 19/33 (57.6%) 6/7 (85.7%) 90/107 (84.1%) 34/35 (97.1%)

+ + + + + + + + + + + + + + + + + + + 19/33 (57.6%) 7/7 (100%) 91/107 (85.0%) 35/35 (100%)

+ + + + + + + + + + + + + + + 15/33 (45.5%) 0/7 (0%) 79/107 (73.8%) 28/35 (80.0%)

and MI, while uorescence under UV light ( = 365 nm) was measured for MI. Each preparation of Chromocult and MI plates was tested for performance using pure cultures of target and non-target microorganisms, as recommended by the USEPA microbiology methods manual. Tests to conrm the sterility of the lter membranes and buffer used for rinsing the ltration apparatus were also performed. 2.4. Liquid culture methods For liquid culture methods, preparation, validation, storage, and handling steps were all performed according to the manufacturer's instructions. For the Colilert method, one snap pack containing the Colilert reagent (IDEXX Laboratories Canada Corp., Toronto, Ontario, Canada) was dissolved in 100 mL of spiked water sample. The solution was then added to a Quanti-Tray, sealed, and incubated at 35.0 0.5 C for 24 h prior to the identication of samples presenting a yellow color and uorescence under UV light ( = 365 nm). All reactions remaining negative after 24 h of incubation were incubated at 35 C for an additional 4 h prior to a second analysis evaluation of the reaction color and uorescence signal. Positive and negative controls, using pure cultures of target and non-target microorganisms, were performed as recommended by the manufacturer. For the Readycult method, one snap pack containing the Readycult reagent (Merck, Darmstadt, Germany) was dissolved in 100 mL of spiked water and incubated at 35.0 0.5 C for 24 2 h prior to the identication of samples presenting a green color and uorescence under UV light ( = 365 nm). Positive and negative controls, using pure cultures of target

and non-target microorganisms, were also performed as recommended by the manufacturer. 3. Results 3.1. Detection of E. coli strains Seventy four (74) E. coli strains of different serotypes isolated from fecal and environmental settings as well as from different geographic origins were used to demonstrate the ability of the four culture methods to detect various E. coli strains (Table 1). For conrmatory purposes, all strains that have presented negative results were tested a second time with a different lot of kit/media. Fifty-nine (59) of the 74 E. coli strains tested (79.7%) yielded a glucuronidase-positive signal with both MI and Chromocult methods (Table 1). Fluorescence was observed for 60 of these strains (81.1%) tested with Readycult. Of the 74 E. coli strains tested, 55 (74.3%) were detected by all three methods, whereas 9 (12.2%) were undetectable by the three methods. In addition, discordant results between these three methods were observed for 10 strains (13.5%). The Colilert method gave the lowest percentage of detection since only 38 of the 74 E. coli strains (51.4%) showed uorescence after 24 h of incubation at 35 C. However, in two cases, Colilert detected a strain that was tested negative by one or two of the other methods. In the rst case, the E. coli strain detected by Colilert, Readycult, and Chromocult methods was not detected by MI and in the second, the strain detected by Colilert and Chromocult was not detected by both Readycult and MI methods. Chromocult, MI, and

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514 Table 3 Strains used for the specicity analysis Gram-positive bacteria (n = 52) Abiotrophia defectiva Enterococcus avium Enterococcus casseliavus Enterococcus cecorum Enterococcus columbae Enterococcus dispar Enterococcus durans Enterococcus faecalis Enterococcus faecium Enterococcus avescens Enterococcus gallinarum Enterococcus hirae Enterococcus mundtii Enterococcus pseudoavium Enterococcus rafnosus Enterococcus ratti Enterococcus saccharolyticus Enterococcus solitarius Enterococcus sulfureus Gemella haemolysans Granulicatella adiacens Lactobacillus acidophilus Leifsonia aquaticus Listeria grayi Listeria innocua Listeria ivanovii Listeria monocytogenes Listeria seeligeri Micrococcus luteus Staphylococcus aureus Staphylococcus capitis Staphylococcus epidermidis Staphylococcus haemolyticus Staphylococcus hominis Staphylococcus lugdunensis Staphylococcus saprophyticus Staphylococcus simulans Staphylococcus warneri Streptococcus agalactiae Streptococcus anginosus Streptococcus bovis Streptococcus constellatus Streptococcus cristatus Streptococcus gordonii Streptococcus intermedius Streptococcus mutans Streptococcus parasanguis Streptococcus pneumoniae Streptococcus pyogenes Streptococcus salivarius Streptococcus sanguinis Streptococcus suis ATCC 49176 ATCC 14025 ATCC 25788 ATCC 43198 ATCC 51263 ATCC 51266 ATCC 19432 ATCC 19433 ATCC 19434 ATCC 49996 LSPQ 3364 ATCC 8043 ATCC 43186 ATCC 49372 ATCC 49427 ATCC 700914 ATCC 43076 ATCC 49428 ATCC 49903 ATCC 10379 ATCC 49175 ATCC 4356 ATCC 14665 ATCC 19120 ATCC 33090 ATCC 19119 ATCC 15313 ATCC 35967 ATCC 9341 ATCC 25923 ATCC 27840 ATCC 14990 ATCC 29970 ATCC 27844 ATCC 43809 ATCC 15305 ATCC 27848 ATCC 27836 ATCC 13813 ATCC 33397 ATCC 33317 ATCC 27823 ATCC 51100 ATCC 33399 ATCC 27335 ATCC 25175 ATCC 15912 ATCC 6303 ATCC 19615 ATCC 7073 ATCC 10556 ATCC 43765 Gram-negative bacteria (n = 37) Acinetobacter baumanii Acinetobacter haemolyticus Aeromonas caviae Aeromonas hydrophila Burkholderia cepacia Haemophilus haemolyticus Haemophilus inuenzae Haemophilus parahaemolyticus Haemophilus parainuenzae Legionella pneumophila Moraxella atlantae Moraxella catarrhalis Neisseria caviae Neisseria elongata Neisseria gonorrhoeae Neisseria meningitidis Neisseria mucosa Pasteurella aerogenes Photorhabdus luminescens Proteus mirabilis Proteus vulgaris Providencia alcalifaciens Providencia rettgeri Providencia rustigianii Providencia stuartii Pseudomonas aeruginosa Pseudomonas uorescens Pseudomonas stutzeri Salmonella cholerasuis Salmonella indica Salmonella typhimurium Stenotrophomonas maltophilia Vibrio alginolyticus Vibrio cholerae Vibrio uvialis Vibrio parahaemolyticus Vibrio vulnicus

511

ATCC 19606 ATCC 17906 CCUG 44411 ATCC 7966 ATCC 25416 ATCC 33390 ATCC 9007 ATCC 10014 ATCC 7901 ATCC 33156 ATCC 29525 ATCC 25238 ATCC 14659 ATCC 25295 ATCC 35201 ATCC 13077 ATCC 19696 ATCC 27883 ATCC 43948 ATCC 25933 ATCC 29513 ATCC 9886 ATCC 9250 ATCC 33673 ATCC 43664 ATCC 35554 ATCC 13525 ATCC 17588 ATCC 7001 ATCC 43976 ATCC 14028 ATCC 13637 CCRI-14794 ATCC 25870 CCRI-14795 ATCC 17802 ATCC 27562

Readycult tests yielded a positive signal for all other strains found positive by Colilert. Overall, discordant results between Colilert and, at least one of the three other methods, were observed for 29 of the 74 strains panel (39.2%). An additional 4 E. coli strains showed positive with Colilert upon an extended incubation for an additional 4 h at 35 C thereby increasing its level of detection from 51.4 to 56.8%. When only non-O157:H7 E. coli strains were considered, 59 of 65 (90.8%) were found to be positive with both MI and Chromocult methods while 58 (89.2%) were positive with Readycult. Colilert presented the lowest level of detection by yielding a uorescent positive signal for only 38 of the 65 non-O157:H7 E. coli strains (58.5%) tested. Finally, when only the 28 environmental E. coli strains were considered, 22 (78.6%) were detectable with Colilert while 27 (96.4%) were detectable with Chromocult and 28 (100%) with both MI and Readycult. 3.2. Detection of Shigella strains Eight (8) Shigella spp. strains were used to demonstrate the ability of the four tests to detect Shigella species (Table 1). S. exneri and S.

dysenteriae strains were undetected by all 4 methods, while the S. sonnei strains were -galactosidase-positive with the four methods and also -glucuronidase-positive based on the MI test. S. boydii was detected as -galactosidase-positive by both Chromocult and MI tests and as glucuronidase-positive by Readycult, Chromocult, and MI test methods. 3.3. Detection of total coliform strains The ability of the four methods to detect -galactosidase production by total coliform strains was veried with the 74 E. coli strains (Table 1) as well as with the 33 clinical and environmental non-E. coli total coliform strains (Table 2). Colilert, Readycult and Chromocult showed comparable detection rates of respectively 20 (60.6%), 19 (57.6%), and 19 (57.6%) of the 33 non-E. coli total coliform strains tested, whereas the MI test method gave the lowest percentage by detecting only 15 (45.5%) of these strains. Only 4 strains (12.1%) were detected by the four test methods whereas 6 (18.2%) were undetectable by all methods. Thus, discordant results were observed for 23

512

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514

(69.7%) of the 33 non-E. coli total coliform strains tested (Table 2). Again, when considering all total coliform strains, including all E. coli strains, MI presented the lowest percentage of detection with only 80 (74.8%) of the 107 strains tested whereas Colilert, Readycult, and Chromocult detected -galactosidase production in respectively 90 (84.1%), 90, and 91 (85.0%) of all total coliform strains tested. Interestingly, we observed that with environmental non-E. coli total coliform strains only, 6 of the 7 (85.7%) total coliform strains tested positive for -galactosidase production with both Colilert and Readycult, while all were positive with Chromocult. MI did not detect any of these strains. However, when only the 28 environmental E. coli strains were considered, all were positive with MI as well as with the 3 other methods. 3.4. Specicity Specicity (i.e. the ability to detect only -galactosidase and glucuronidase production by the target species of the four test methods) was veried by testing 89 non-total coliform strains representing 52 species of Gram-positive and 37 species of Gramnegative bacteria that are frequently encountered in fecal or environmental settings (Table 3). Eighty-six (86) of these 89 (96.6%) bacterial species were undetectable with the 4 test methods. Providencia alcalifaciens and Providencia rustigianii yielded a -galactosidase-positive signal while Salmonella indica yielded positive results for both -galactosidase and -glucuronidase with Readycult, Chromocult and MI test methods. 4. Discussion By using a panel of 74 E. coli strains of different serotypes isolated from fecal and environmental settings, we have determined the ability of the test methods to detect -glucuronidase production. To our knowledge, this is the rst report on the comparison of these test methods, using pure cultures. We found that Chromocult, MI and Readycult present a similar percentage of detection varying from 79.9 to 81.1%, whereas Colilert detected only 51.4% of the E. coli strains. As opposed to Rice et al. (1990) who observed -glucuronidase production for 99.5% of the E. coli strains tested after 28 h, an additional 4 h of incubation for uorescent-negative Colilert strains allowed only a slight improvement in the detection of -glucuronidase production (increased to 56.8%). The lowest level of detection obtained by Colilert, as compared to the 3 other tests, is not explained by the presence of E. coli O157:H7 strains, that have been reported to be uniformly -glucuronidase-negative (Krishnan et al., 1987; Ratnam et al., 1988). Indeed, when these strains were excluded in the data set, Chromocult, MI and Readycult methods still showed a similar percentage of detection varying from 89.2 to 90.8%. Whereas the detection level with Colilert was 58.8%. The ability to detect E. coli was also veried with environmental isolates only. Surprisingly, and contrary to what have been published previously by Rice et al. (1993) and Shadix et al. (1993) who detected glucuronidase production for more than 95% of the E. coli strains tested, Colilert showed the lowest detection level of environmental E. coli strains (i.e. 78.6%), whereas almost all E. coli strains were detected with Chromocult, MI and Readycult (96.4 to 100%). Such discrepant results are not totally unexpected since studies previously realized by Doyle et al. (1955), Feng and Lampel (1994), Chang et al. (1989), and Shadix and Rice (1991) showed that the percentage of -glucuronidase-negative E. coli strains of a given nature source population will be variable and that the composition of the media will inuence the rate of detection. The results of this study, obtained by comparing a collection of E. coli strains, show that Colilert systematically detected approximately 20% less E. coli strains than the 3 other culture methods, even if they are based on the same enzymatic principles. A similar lower percentage of E. coli detection using minimal media o-nitrophenyl--D-galactopyranoside

(MMO)-MUG preparations, such as Colilert, was also observed by Martins et al. (1993). The dened substrate technology of Colilert is based on the choice and the amount of ingredients for providing a strict requirement for E. coli specic growth. Our results and those of Martins et al. (1993) suggest that such minimal media could be lacking some essential nutrients. Thus, the low recovery of some E. coli strains by minimal medium may not be totally attributed to the strain itself but may also be inuenced by the composition of the medium. In this study, S. dysenteriae and S. exneri strains were not detected by any of the four methods. For Readycult and Colilert, S. boydii and S. sonnei are recognized as total coliforms, whereas MI yielded an E. coliphenotype result for these two species. These results are in accordance with those published by McDaniels et al. (1996), where positive signal for -glucuronidase were observed within members of the genus Shigella. The ability of the four test methods to detect -galactosidase production was veried by using, in addition to the 74 E. coli strains, 33 reference and environmental non-E. coli total coliform strains found in fecal and environmental settings of different geographic origins. For the non-E. coli total coliforms, Chromocult, Colilert and Readycult showed a similar percentage of detection ranging from 57.6 to 60.6% whereas MI detected only 45.5% of the strains tested. It is well known in environmental microbiology that the total coliforms group, based on phenotypic characteristics, is not well dened. In accordance with Olstadt et al. (2007), the results of our study show that, in addition to this fact, there is a total lack of correlation between test methods based on the same enzymatic principle to recognize a strain as non-E. coli total coliform. Indeed, our results showed that there is no correlation between the 4 methods tested either within the same genera or within the same species. Furthermore, when only environmental non-E. coli total coliforms are considered, Chromocult achieved to detect all strains, while Colilert and Readycult detected 85.7% of them (7/8). MI was unable to detect any of these environmental isolates. When all total coliforms strains are considered, Colilert, Chromocult, and Readycult showed similar percentages of detection ranging from 94.6 to 97.3% of E. coli strains, whereas MI detected 86.5% of these strains. Interestingly, the four methods detected -galactosidase production in all environmental E. coli strains. However, studies using a higher number of environmental strains are required to extend this observation. Finally, no correlation was observed between galactosidase production results obtained by the Microscan identication test and the results obtained by the four other methods tested. Indeed, the orthonitrophenyl--D-galactopyranoside (ONPG) test performed on the Microscan system yielded -galactosidasepositive signal for 94.4% of total coliform strains tested. These results suggest that automated phenotypic identication tests may also have difculties to induce production of -galactosidase. Thus, identication methods, solely relying on the activity of a single enzyme, are subject to a lack of robustness and may lead to misinterpretations since enzymatic activity can be transient and highly regulated by environmental factors. By comparing the -glucuronidase and the -galactosidase production detection between environmental and clinical E. coli strains, one could be tempted to conclude that these 4 methods are more efcient to detect environmental E. coli strains than clinical E. coli strains. However, we must keep in mind that the environmental strains used in this study were isolated from recommended media used for the recovery of E. coli in water. Thus, environmental E. coli strains that cannot be detected by those media could not have been included in the strain collection used for in this study. This bias limits the interpretation of the difference between the rate of detection of environmental E. coli strains and strains isolated by clinical techniques. The specicity was veried by testing 89 non-total coliform strains frequently encountered in fecal or environmental settings. Only P.

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514

513

alcalifaciens, P. rustigianii, and S. indica yielded positive results. It was previously reported that species of Providencia yielded false-positive results with Colilert especially in a marine environment (Pisciotta et al., 2002). Furthermore, some strains of Salmonella species are known to express -galactosidase and -glucuronidase enzyme (Feng and Hartman, 1982; Kaluzewski and Tomczuk, 1995). Globally, all four tests methods studied here presented a low level of false-positive results based on testing of a bacterial strains collection representing 52 Grampositive and 37 Gram-negative species. These results are in accordance with studies claiming that a further systematic identication and conrmation for these techniques are not necessary for water samples monitoring (Edberg et al., 1988; Rompre et al., 2002), except for marine water monitoring in the case of Colilert (Landre et al., 1998; Pisciotta et al., 2002). This study examined the relative performance of the Colilert, MI agar, Chromocult agar and Readycult test methods by comparing their respective analytical limits, using a collection of strains consisting of E. coli, Shigella, total coliforms, and other non-target bacteria. These results do not necessarily correlate with their individual efciency to detect E. coli and total coliforms in natural water samples. This is explained by the fact that environmental samples present a heterogenic population of bacteria and the chance that a contaminated water sample only contains strains that are undetectable by one of these methods is quite low. Indeed, the majority of published studies comparing these methods, using natural water samples, showed a similar efciency between them and classical reference methods (APHA et al., 2005; Bernasconi et al., 2006; Brenner et al., 1993, 1996a, b; Buckalew et al., 2006; Clark and el-Shaarawi, 1993; Colquhoun et al., 1995; Cowburn et al., 1994; Eckner, 1998; Edberg et al., 1988, 1990; Fricker and Fricker, 1996; Fricker et al., 1997; Horman and Hanninen, 2006; Macy et al., 2005; Niemela et al., 2003; Schets et al., 2002). However, it cannot be excluded that a lower rate of detection of E. coli could be observed with Colilert when natural water samples analysed present low diversity of strains or when the water samples only contain low levels of bacterial contamination. Indeed, this may provide an explanation for the unexpected level of false-negative results observed with Colilert method by Clark et al. (1991), Hall and Moyer (1989), Lewis and Mak (1989), and Pitkanen et al. (2007), when it was compared to the standard membrane ltration fecal coliform (mFC) and thermotolerant E. coli (mTEC) test methods. Finally, classical microbiology techniques for water quality monitoring, such as mFC and Tergitol-TTC media, require strain conrmation. Thus, to conrm the identity of their E. coli presumptive strains, laboratories using classical techniques could be tempted to use one of the four commercial methods tested in this study. The relatively high level of false-negative results reached by the four methods suggests that they should not be used to conrm the identity of presumptive E. coli strains. Tests asserting multiple phenotypes are recommended for species identication. Future alternatives may include molecular tests for conserved species-specic genetic targets (Martinez et al., 2006). Acknowledgements We thank ve Brub and Marie-Claude Hlie for technical assistance. We also thank Drs Louise Ct, director of the Microbiology Laboratory of CHUL (Centre Hospitalier Universitaire de Qubec), Philippe Cantin (Centre d'Expertise en Analyse Environnementale du Qubec), Pierre Simard and Lynda Rodrigue (Bodycote Canada), Pierre Harbec (Laboratoire de Sant Publique du Qubec), Wang Fu (Huashan Hospital), Helge Karch (Institut fr Hygiene und Mikrobiologie der Universitat), Jordy Vila (Servei de Microbiologia, Centre de Diagnstic Biomdic, Universitat de Barcelona), Nicolas Chamoine (Hpital Ambroise Par), Patricia Bradford (Wyeth-Ayerst Research), Sebastian G. B. Amyes (University of Edinburgh), Marek Gniadkowski (National Institute of Public Health), and Mignon du Plessis (South African Institute for Medical Research) for providing E. coliShigella strains.

This research was supported by grants PA-15586 from the Canadian Institutes of Health Research (CIHR) and FCI-5251 from the Canadian Foundation for Innovation (CFI). Andre Maheux, Vicky Hupp, and Jean-Luc Bernier received a scholarship from Nasivvik (Center for Inuit Health and Changing Environment; Canadian Institutes for Health Research). Disclosure of interests: The authors of this study do not have any links with companies manufacturing or commercializing water testing products.

References
AFNOR, (Association Franaise de Normalisation), 1990. Eaux-Mthodes d'essais, 4th ed. Recueil de Normes Franaises, La Dfense, Paris. 735 pp. APHA, AWWA AEF, 1998. Standard methods for the examination of water and wastewater, 20st ed. American Public Health Association, Inc., Washington. D. C. APHA, AWWA AEF, 2005. Standard methods for the examination of water and wastewater, 21st ed. American Public Health Association, Inc., Washington. D. C. Bernasconi, C., Volponi, G., Bonadonna, L., 2006. Comparison of three different media for the detection of E. coli and coliforms in water. Water Sci. Technol. 54, 141145. Brenner, D.J., Fanning, G.R., Skerman, F.J., Falkow, S., 1972. Polynucleotide sequence divergence among strains of Escherichia coli and closely related organisms. J. Bacteriol. 109, 953965. Brenner, K.P., Rankin, C.C., Roybal, Y.R., Stelma Jr., G.N., Scarpino, P.V., Dufour, A.P., 1993. New medium for the simultaneous detection of total coliforms and Escherichia coli in water. Appl. Environ. Microbiol. 59, 35343544. Brenner, K.P., Rankin, C.C., Sivaganesan, M., 1996a. Interlaboratory evaluation of MI agar and the US Environmental Protection Agency-approved membrane lter method for the recovery of total coliforms and Escherichia coli from drinking water. Appl. Environ. Microbiol. 27, 111119. Brenner, K.P., Rankin, C.C., Sivaganesan, M., Scarpino, P.V., 1996b. Comparison of the recoveries of Escherichia coli and total coliforms from drinking water by the MI agar method and the U.S. Environmental Protection Agency-approved membrane lter method. Appl. Environ. Microbiol. 62, 203208. Buckalew, D.W., Hartman, L.J., Grimsley, G.A., Martin, A.E., Register, K.M., 2006. A longterm study comparing membrane ltration with Colilert dened substrates in detecting fecal coliforms and Escherichia coli in natural waters. J. Environ. Manage. 80, 191197. Burlingame, G.A., McElhaney, J., Bennett, M., Pipes, W.O., 1984. Bacterial interference with coliform colony sheen production on membrane lters. Appl. Environ. Microbiol. 47, 5660. Chang, G.W., Brill, J., Lum, R., 1989. Proportion of beta-D-glucuronidase-negative Escherichia coli in human fecal samples. Appl. Environ. Microbiol. 55, 335339. Clark, D.L., Milner, B.B., Stewart, M.H., Wolfe, R.L., Olson, B.H., 1991. Comparative study of commercial 4-methylumbelliferyl-beta-D-glucuronide preparations with the standard methods membrane ltration fecal coliform test for the detection of Escherichia coli in water samples. Appl. Environ. Microbiol. 57, 15281534. Clark, J.A., 1980. The inuence of increasing numbers of non-indicator organisms by the membrane lter and presenceabsence test. Can. J. Microbiol. 26, 827. Clark, J.A., el-Shaarawi, A.H., 1993. Evaluation of commercial presenceabsence test kits for detection of total coliforms, Escherichia coli, and other indicator bacteria. Appl. Environ. Microbiol. 59, 380388. Colquhoun, K.O., Timms, S., Fricker, C.R., 1995. Detection of Escherichia coli in potable water using direct impedance technology. J. Appl. Bacteriol. 79, 635639. Cowburn, J.T., Goodall, T., Fricker, E.J., Walter, K.S., Fricker, C.R., 1994. A preliminary study of the use of Colilert for water quality monitoring. Lett. Appl. Microbiol. 19, 5052. Doyle, M.L., Katzman, P.A., Doisy, E.A., 1955. Production and properties of bacterial betaglucuronidase. J. Biol. Chem. 217, 921930. Eckner, K.F., 1998. Comparison of membrane ltration and multiple-tube fermentation by the Colilert and Enterolert methods for detection of waterborne coliform bacteria, Escherichia coli, and enterococci used in drinking and bathing water quality monitoring in southern Sweden. Appl. Environ. Microbiol. 64, 30793083. Edberg, S.C., Allen, M.J., Smith, D.B., 1988. National eld evaluation of a dened substrate method for the simultaneous enumeration of total coliforms and Escherichia coli from drinking water: comparison with the standard multiple tube fermentation method. Appl. Environ. Microbiol. 54, 15951601. Edberg, S.C., Allen, M.J., Smith, D.B., Kriz, N.J., 1990. Enumeration of total coliforms and Escherichia coli from source water by the dened substrate technology. Appl. Environ. Microbiol. 56, 366369. Evans, T.M., LeChevallier, M.W., Waarvick, C.E., Seidler, R.J., 1981. Coliform species recovered from untreated surface water and drinking water by the membrane lter, standard, and modied most-probable-number techniques. Appl. Environ. Microbiol. 41, 657663. Feng, P., Lampel, K.A., 1994. Genetic analysis of uidA expression in enterohaemorrhagic Escherichia coli serotype O157:H7. Microbiology 140 (Pt 8), 21012107. Feng, P.C., Hartman, P.A., 1982. Fluorogenic assays for immediate conrmation of Escherichia coli. Appl. Environ. Microbiol. 43, 13201329. Fricker, E.J., Fricker, C.R., 1996. Use of two presenceabsence systems for the detection of E. coli and coliforms from water. Water Res. 30, 22262228. Fricker, E.J., Illinhworth, K.S., Fricker, C.R., 1997. Use of two Colilert and Quantitray for assessment of the bacteriological quality of water. Water Res. 31, 24952499.

514

A.F. Maheux et al. / Journal of Microbiological Methods 75 (2008) 506514 Niemela, S.I., Lee, J.V., Fricker, C.R., 2003. A comparison of the International Standards Organisation reference method for the detection of coliforms and Escherichia coli in water with a dened substrate procedure. J. Appl. Microbiol. 95, 12851292. Olstadt, J., Schauer, J.J., Standridge, J., Kluender, S., 2007. A comparison of ten USEPA approved total coliform/E. coli tests. J. Water Health 5, 267282. Pisciotta, J.M., Rath, D.F., Stanek, P.A., Flanery, D.M., Harwood, V.J., 2002. Marine bacteria cause false-positive results in the Colilert-18 rapid identication test for Escherichia coli in Florida waters. Appl. Environ. Microbiol. 68, 539544. Pitkanen, T., Paakkari, P., Miettinen, I.T., Heinonen-Tanski, H., Paulin, L., Hanninen, M.L., 2007. Comparison of media for enumeration of coliform bacteria and Escherichia coli in non-disinfected water. J. Microbiol. Methods 68, 522529. Ratnam, S., March, S.B., Ahmed, R., Bezanson, G.S., Kasatiya, S., 1988. Characterization of Escherichia coli serotype O157:H7. J. Clin. Microbiol. 26, 20062012. Rice, E.W., Allen, M.J., Brenner, D.J., Edberg, S.C., 1991. Assay for beta-glucuronidase in species of the genus Escherichia and its applications for drinking-water analysis. Appl. Environ. Microbiol. 57, 592593. Rice, E.W., Allen, M.J., Edberg, S.C., 1990. Efcacy of beta-glucuronidase assay for identication of Escherichia coli by the dened-substrate technology. Appl. Environ. Microbiol. 56, 12031205. Rice, E.W., Johnson, C.H., Dunnigan, M.E., Reasoner, D.J., 1993. Rapid glutamate decarboxylase assay for detection of Escherichia coli. Appl. Environ. Microbiol. 59, 43474349. Rompre, A., Servais, P., Baudart, J., de-Roubin, M.R., Laurent, P., 2002. Detection and enumeration of coliforms in drinking water: current methods and emerging approaches. J. Microbiol. Methods 49, 3154. Schets, F.M., Nobel, P.J., Strating, S., Mooijman, K.A., Engels, G.B., Brouwer, A., 2002. EU Drinking Water Directive reference methods for enumeration of total coliforms and Escherichia coli compared with alternative methods. Lett. Appl. Microbiol. 34, 227231. Seidler, R.J., Evans, T.M., Kaufman, J.R., LeChevalier, M.W., 1981. Limitations of standard coliform enumeration techniques. J. AWWA 73, 538542. Shadix, L.C., Dunnigan, M.E., Rice, E.W., 1993. Detection of Escherichia coli by the nutrient agar plus 4-methylumbelliferyl beta-D-glucuronide (MUG) membrane lter method. Can. J. Microbiol. 39, 10661070. Shadix, L.C., Rice, E.W., 1991. Evaluation of beta-glucuronidase assay for the detection of Escherichia coli from environmental waters. Can. J. Microbiol. 37, 908911.

Hall, N.H., Moyer, N.P., 1989. Evaluation of multiple tube fermentation test and the autoanalysis Colilert test for the enumeration of coliforms and Escherichia coli in private well water samples. AWWA Technol. Conf. Proc., pp. 479496. Horman, A., Hanninen, M.L., 2006. Evaluation of the lactose Tergitol-7, m-Endo LES, Colilert 18, Readycult Coliforms 100, Water-Check-100, 3M Petrilm EC and DryCult Coliform test methods for detection of total coliforms and Escherichia coli in water samples. Water Res. 40, 32493256. Kaluzewski, S., Tomczuk, D., 1995. [Evaluation of the usefulness of tests for production of Beta-D-glucuronidase and propylene glycol utilization for the differentiation of enterobacteriaceae rods]. Med. Dosw. Mikrobiol. 47, 155168. Krishnan, C., Fitzgerald, V.A., Dakin, S.J., Behme, R.J., 1987. Laboratory investigation of outbreak of hemorrhagic colitis caused by Escherichia coli O157:H7. J. Clin. Microbiol. 25, 10431047. Landre, J.P., Gavriel, A.A., Lamb, A.J., 1998. False-positive coliform reaction mediated by Aeromonas in the Colilert dened substrate technology system. Lett. Appl. Microbiol. 26, 352354. Lewis, C.M., Mak, J.L., 1989. Comparison of membrane ltration and Autoanalysis Colilert presenceabsence techniques for analysis of total coliforms and Escherichia coli in drinking water samples. Appl. Environ. Microbiol. 55, 30913094. Lupo, M., Halpern, Y.S., 1970. Gene controlling L-glutamic acid decarboxylase synthesis in Escherichia coli K-12. J. Bacteriol. 103, 382386. Macy, J.T., Dunne, E.F., Angoran-Benie, Y.H., Kamelan-Tano, Y., Kouadio, L., Djai, K.A., Luby, S.P., 2005. Comparison of two methods for evaluating the quality of stored drinking water in Abidjan, Cte d'lvoire, and review of other comparisons in the literature. J. Water Health 3, 221228. Martinez, G., Bruant, G., Brousseau, R., Masson, L., Harel, J., 2006. Development of a new integrated diagnostic test for identication and characterization of pathogens. Dev. Biol. (Basel) 126, 213218. Martins, M.T., Rivera, I.G., Clark, D.L., Stewart, M.H., Wolfe, R.L., Olson, B.H., 1993. Distribution of uidA gene sequences in Escherichia coli isolates in water sources and comparison with the expression of beta-glucuronidase activity in 4-methylumbelliferyl-beta-D-glucuronide media. Appl. Environ. Microbiol. 59, 22712276. McDaniels, A.E., Rice, E.W., Reyes, A.L., Johnson, C.H., Haugland, R.A., Stelma Jr., G.N., 1996. Conrmational identication of Escherichia coli, a comparison of genotypic and phenotypic assays for glutamate decarboxylase and beta-D-glucuronidase. Appl. Environ. Microbiol. 62, 33503354. Means, E.G., Olson, B.H., 1981. Coliform inhibition by bacteriocin-like substances in drinking water distribution systems. Appl. Environ. Microbiol. 42, 506512.