Veterinary Microbiology 97 (2003) 87101

Virulence genes of O149 enterotoxigenic Escherichia coli from outbreaks of postweaning diarrhea in pigs
Babak N. Noamani a , John M. Fairbrother b , Carlton L. Gyles a,
b

Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ont., Canada, N1G 2W1 Department of Pathology and Microbiology, School of Veterinary Medicine, University of Montreal, St. Hyacinthe, Que., Canada, J2S 7C6 Received 13 January 2003; received in revised form 14 August 2003; accepted 14 August 2003

Abstract The goal of this research was to determine whether isolates of O149 porcine enterotoxigenic Escherichia coli (ETEC) recovered from recent outbreaks of severe diarrhea in weaned pigs in Ontario, Canada, had virulence attributes different from those of isolates of the same serogroup from diarrhea of pigs in the 1970s and 1980s. Polymerase chain reaction amplication was used to determine the distribution of 11 virulence-associated genes in recent (100 isolates) and old (35 isolates) Ontario O149 porcine ETEC. These tests demonstrated that 92% of the recent isolates possessed the estA gene for STa enterotoxin, whereas none of the old isolates had this gene. H antigen determination showed that all the isolates which lacked the estA gene (all 35 old isolates plus 8 recent isolates) were H43, whereas isolates which had the estA gene were H10. The astA gene for enteroaggregative heat-stable enterotoxin (EAST1) and the K88ac antigen were present in all 135 isolates. Plasmid analyses identied a cryptic 5.1 kb plasmid in 99% of recent and 60% of old isolates. Suppressive subtractive hybridization associated several types of DNA fragments with the recent O149 ETEC, namely, fragments with no homology to DNA in databases, fragments of LPS biosynthesis genes, and F plasmid DNA. We conclude that the recent outbreaks of PWD in Ontario pigs were associated primarily with a new serotype of O149 ETEC and that isolates of this serotype possessed the estA gene that was not present in old O149 ETEC isolated from pigs in Ontario. 2003 Elsevier B.V. All rights reserved.
Keywords: Escherichia coli; Enterotoxigenic; O149; H10; H43; Pig; Bacteria; Diarrhea

Corresponding author. Tel.: +1-519-824-4120x4715l; fax: +1-519-767-0809. E-mail address: cgyles@ovc.uoguelph.ca (C.L. Gyles).

0378-1135/$ see front matter 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.vetmic.2003.08.006

88

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

1. Introduction Over the past 5 years, there has been a marked increase in frequency and severity of outbreaks of Escherichia coli postweaning diarrhea (PWD) in pigs in Ontario and Quebec (J. Fairbrother, personal communication; Josephson and Smart, 1998; Josephson et al., 1999, 2000). Investigations failed to identify management factors that could be associated with the severe outbreaks (Amezcua, 2001). The enterotoxigenic E. coli (ETEC) that were isolated from affected pigs were hemolytic K88-positive organisms that were not otherwise characterized in diagnostic laboratories. The present study was undertaken to determine whether the K88-positive ETEC isolated from these outbreaks possess unusual virulence attributes. The known virulence factors of ETEC are mbriae associated with colonization of the intestine, and enterotoxins which are responsible for inducing diarrhea. The mbriae that have been implicated in PWD are F18 and K88 (F4), each of which exists as two major variants, called ab and ac. Whereas F18 mbriae are associated almost exclusively with PWD, K88 mbriae are implicated in diarrhea of neonatal as well as weaned pigs (Gaastra and de Graaf, 1982; Wittig and Fabricius, 1992; Nagy and Fekete, 1999). F18ab-positive ETEC often produce verotoxin 2e (VT2e), are capable of causing edema disease, and are appropriately referred to as ETEC/VTEC (Nagy et al., 1997; Parma et al., 2000; Fekete et al., 2002). Typically, ETEC carry only a single type of colonization pilus, but occasionally isolates carry both F18 and K88 mbriae (Nagy and Fekete, 1999). Other types of mbriae, including 987P (F6) (Moon, 1990), K99 (F5) (Smith and Linggood, 1971; Orskov et al., 1975) and F41 (Vazquez et al., 1996) are usually found on porcine ETEC isolated from neonatal pigs (Wilson and Francis, 1986; Harel et al., 1991; Ojeniyi et al., 1994), and are only occasionally found on ETEC from weaned pigs (Wilson and Francis, 1986; Nagy and Fekete, 1999; Frydendahl, 2002). Four enterotoxin genes have been reported in ETEC from pigs with PWD: elt (heat-labile enterotoxin, LT), estA (heat-stable enterotoxin STa or STI), estB (heat-stable enterotoxin STb or STII), and astA (the enteroaggregative heat-stable enterotoxin, EAST1). Various combinations of elt, estA, and estB genes have been described in ETEC (Mainil et al., 1998; Osek, 1999; Amezcua et al., 2002; Frydendahl, 2002). The literature on astA genes is very recent and limited, but these genes have been particularly associated with K88-positive ETEC (Yamamoto and Nakazawa, 1997; Frydendahl, 2002; Menard and Dubreuil, 2002; Osek, 2003). In the case of O149 ETEC, most reports on isolates have indicated a high prevalence of the combination of elt and estB genes, but strains with elt, estA, and estB genes have also been reported (Mainil et al., 1998; Osek, 1999, 2000). Recently, An et al. (1999) described the paa gene which encodes the porcine attaching and effacing associated (Paa) protein, which is associated with the attaching and effacing phenotype in certain porcine strains of E. coli. Nothing is known of the prevalence or role of this gene in the virulence of porcine ETEC. Two potential advantages that ETEC strains have over non-enterototoxigenic E. coli are greater dispersion by pigs with diarrhea compared to pigs with normal feces, and greater opportunity for acquisition of genes due to the massive numbers of the bacteria in the intestinal tracts of affected pigs. It would, therefore, not be surprising if ETEC strains acquired genes which enhanced their virulence. Enhanced virulence of ETEC in recent

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

89

outbreaks of postweaning diarrhea could also be due to the presence of a new clone of O149 E. coli with additional virulence factors. The hypothesis that was tested by experiments described in this paper was that O149:K88ac ETEC from recent outbreaks of diarrhea in pigs in Ontario possess genes which have enhanced their virulence compared to older isolates. A collection of 100 O149 ETEC isolated from pigs with PWD during 19982001 was compared with a collection of 35 O149 ETEC isolated from pigs with PWD in the period 19741987. The two collections were compared for possession of genes for K88ac, F18, 987P, and F41 mbriae; LT, STa, STb, EAST1, and STx2e toxins; and Paa protein. Subtractive hybridization was also conducted between a recent and an old isolate of O149 ETEC to identify potential virulence or tness genes that have not previously been associated with porcine ETEC. 2. Materials and methods 2.1. Bacterial strains A total of 135 hemolytic O149 ETEC isolates from cases of PWD in pigs in Ontario were obtained from the Animal Health Laboratories (AHL, Guelph, Ontario) and Gallant Custom Laboratories (Cambridge, Ontario). The collection consisted of 100 isolates recovered from pigs with PWD in the period 19982001 (recent isolates) and 35 isolates recovered from pigs with diarrhea in the period 19741987 (old isolates). All the isolates were streaked on blood agar, checked for purity, tested biochemically to ensure that they were E. coli, then frozen at 70 C in a milk-based freezing solution (Harris, 1954). 2.2. Identication of O149 and K88ac antigens Both O149 and K88ac antigens were identied by slide agglutination tests. A small portion of a bacterial colony grown on blood agar was suspended in one drop of normal saline solution on a microscope slide. One drop of antiserum was added and mixed by gently rocking the slide. Agglutination within 30 s was recorded as a positive reaction. The O149 antiserum was produced by repeated intravenous inoculation of a rabbit with a strain of heat-killed ETEC O149:K91 which was negative for the K88ac antigen. The K88ac antigen was purchased from a commercial source (E. coli laboratory, Facult de mdecine vtrinaire, Service de diagnostic, Universit de Montral, St. Hyacinthe, PQ). 2.3. Hemolysin activity The growth from the frozen stock was touched with an inoculating loop, streaked on a blood agar plate (Columbia agar, supplemented with 5% sheep blood, Oxoid), and incubated overnight at 37 C. Hemolysis was evident as a zone of lysis surrounding the bacterial growth. 2.4. Antimicrobial sensitivity test Susceptibility of isolates to nine antibacterial drugs was determined by disc diffusion as per NCCLS methods, described by the disk manufacturer (BBL, Becton Dickinson

90

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

Microbiology Systems, Franklin Lakes, NJ, USA). The antibacterial drugs that were tested were ampicillin (Am), apramycin (Ap), ceftiofur (Cef), enrooxacin (En), gentamicin (G), neomycin (N), spectinomycin (Sp), tetracycline (Tc), and trimethoprim/sulfamethoxazole (T/S). These are the antimicrobial agents that are in common use in swine enterprises in Ontario and are the agents that are used for routine testing in the Animal Health Laboratory, Laboratory Services, University of Guelph. 2.5. Multiplex PCR assay for detection of nine virulence genes The multiplex PCR assay allowed amplication of nine virulence genes associated with porcine ETEC, namely genes for LT, STa, STb, STx2e, K88, K99, F18, F41 and 987P (Bosworth and Casey, 1997). DNA template was prepared according to the instructions of the manufacturer (Instagene, Bio-Rad Lab., Mississauga, ON). Each 25 l reaction contained 0.2 mM of each dNTP, 0.5 mM of each primer, 1/10 volume of 10 PCR buffer, 3.5 mM MgCl2 , 10 l template DNA, and 2.5 units of Taq polymerase. The samples were amplied in a GeneAmp PCR 2400 thermocycler (Perkin-Elmer). The PCR products were electrophoresed in 1% agarose (Invitrogen Life Technologies, Carlsbad, CA) in 0.5 tris-borate EDTA (TBE) buffer for 1 h at 100 V, stained with ethidium bromide, and photographed under UV light. 2.6. PCR assays for detection of genes for EAST1 and Paa The protocols for PCR amplication of the genes for EAST1 and Paa were as described by Yamamoto and Nakazawa (1997) and An et al. (1999), respectively. 2.7. PCR assays for detection of genes for H10 and H43 antigens A random sample of ve recent and ve old isolates were serotyped at the Health Canada Laboratory for Foodborne Zoonoses to conrm the identity of the O149 antigen and to determine the H antigen(s) present on the isolates. Subsequently, primers were designed to specically amplify sequences of the iC genes that encode the H10 and H43 agellins, and all isolates were tested by PCR to determine whether they possessed the iC gene that encodes H10 or H43 antigen. These tests were conducted as individual PCR reactions. The primers (Table 1) were synthesized by the Guelph Molecular Super Center, Laboratory Services, University of Guelph. DNA was obtained by boiling the cultures. Each 25 l
Table 1 Primers for amplication of the iC genes encoding E. coli H10 and H43 antigens Gene iC H10a Primer 10F 10R 43F 43R Primer sequence (5 3 ) GTAACTACTGTTGGCCGCGATG AACACCAGCATCACTGATATTAGCG AAATCGACTCTTCAACCCTGGG CAGTTTCAGGCCACTCGTGTC Size of amplicon (bp) 171 443

iC H43b
a b

Based on nucleotide sequence in AF169320. Based on nucleotide sequence in AF169323.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

91

reaction mixture contained 10 l of the template DNA, 0.2 mM de-oxynucleotide triphosphates (dNTPs), 0.4 pmol of each primer, 2.5 l of 10 PCR buffer (supplied with Taq polymerase), 2.0 mM MgCl2 and one unit of Taq polymerase (AmpliTaq Gold DNA polymerase, Perkin-Elmer, Foster City, CA). PCR conditions consisted of a denaturing step at 94 C for 3 min, followed by 35 cycles of a denaturing step at 95 C for 45 s, an annealing step at 65 C for 45 s and extension at 72 C for 45 s, followed by a nal extension at 72 C for 7 min. 2.8. Plasmid proles Plasmid DNA was prepared using the Qiagen Plasmid Mini Kit (Qiagen, Mississauga, ON). The protocol is based on alkaline lysis, followed by binding of plasmid DNA to Qiagen anion-exchange resin under low salt and pH conditions, elution, precipitation, and solution of the precipitated DNA in sterile distilled water. Samples were run on 0.7% agarose gels, stained with ethidium bromide, and photographed under UV light. 2.9. Subtractive hybridization One recent isolate (Ro8) and one old isolate (Old82) which had similar plasmid proles and a minimum number of drug resistance markers, were selected for subtractive hybridization. Isolate Ro82 was an O149:H10 isolate that was resistant to tetracycline only and possessed genes for LT, STb, STa, K88, and EAST1. Isolate Old 82 was an O149:H43 isolate that was also resistant to tetracycline only and possessed genes for LT, STb, K88, and EAST1. Genomic DNA preparations from both isolates were isolated using a Qiagen kit (Qiagen, Tip 100). DNA sequences common to a recent and an old O149:K88ac isolate were removed and sequences unique to the recent isolate were amplied by two rounds of PCR, following the instructions of the manual of the Clontech PCR-Select Bacterial Genome Subtraction Kit (Clontech, Palo Alto, CA, USA). PCR products were ligated to pGEM T-EASY vector (Promega, Madison, WI, USA) and the ligated products were transformed into XL1-Blue (Stratagene, La Jolla, CA, USA) competent cells (Nishimura et al., 1990). Transformants were selected for ampicillin resistance. Plasmid DNA was extracted (Qiagen minikit Tip 20), the concentrations of plasmid DNA were estimated (GeneQuant, Fisher), and 200 g quantities of plasmid DNA were submitted for sequencing (Laboratory Services, University of Guelph). Primers identical to those used in the nested PCR (ClonTech PCR-Select Bacterial Genome Subtraction Kit, ClonTech) were used for sequencing. The nucleotide sequences of DNA inserted in the plasmid vector were subjected to BLAST searches (National Center for Biotechnology Information, http://www.ncbi.nlm.nih.gov). The specicity of cloned DNA fragments that did not show homology to E. coli K12 sequence was determined by PCR amplication, using primers that hybridized specically to the cloned fragments. The PCR amplications were conducted with the cloned DNA fragments, genomic DNA of strain Ro8, and genomic DNA of strain Old82 as templates.

3. Results The slide agglutination tests conrmed that all 100 recent and 35 old isolates belonged to O149 and were K88ac-positive. The results of H antigen determination by serological

92

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

Fig. 1. PCR identication of genes for K88, LT, STa, and STb in old and new O149 ETEC. Lane 1: molecular weight markers (bp). Lane 2: PCR products of F41 (612 bp), K88 (499 bp), LT (272 bp), STa (158 bp), and STb (113 bp) genes. Lane 3: products of STx2e (733 bp), 987P (409 bp), F18 (313 bp), and STa (158 bp). Lane 4: products from negative control E. coli K12 strain 711. Lane 5: product from strain 711 which received spectinomycin resistance genes in mating with recent O149 ETEC strain JG280. Lane 6: PCR products from strain JG280. Lane 7: PCR products from an old O149 ETEC strain.

methods showed that all ve old isolates and one new isolate possessed the H43 antigen and that four recent isolates possessed the H10 antigen. The results of PCR determination of the H10 and H43 antigens showed that among the 35 old isolates, all possessed the gene for the H43 antigen, whereas among the recent isolates eight possessed the gene for the H43 antigen and 92 possessed the gene for the H10 antigen. Hemolytic activity on blood agar plates was detected in all 135 isolates. The results from multiplex PCR assay (illustrated in Fig. 1) showed that the only remarkable difference between recent and old isolates was the presence of the gene for STa heat-stable enterotoxin in 92% of the recent isolates, and its absence from the old isolates. All 92 recent isolates which possessed the iC gene for the H10 antigen also possessed the estA gene. Multiplex PCR amplication demonstrated that all of the isolates carried the faeG gene for K88 adhesin, and that none carried genes for the F18, F41, or 987P mbriae or the STx2e toxin. All of the old isolates and all but one of the recent isolates carried the genes for LT and STb. Single PCR assays showed that the astA gene for EAST1 was present in all of the isolates. The most common virulence pattern for recent isolates was K88/EAST1/LT/STa/STb, whereas the most common pattern for old isolates was K88/EAST1/LT/STb. The porcine attaching and effacing associated gene was present in 17% of the recent isolates and none of the old isolates.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

93

Fig. 2. Percentage of old and recent isolates resistant to each of eight antibiotics. All isolates were susceptible to enrooxacin. Tc: tetracycline, Sp: spectinomycin, Am: ampicillin, N: neomycin, G: gentamicin, Ap: ampicillin, T/S: trimethoprim/sulfamethoxazole, and Cef: ceftiofur.

Major patterns of resistance of the isolates to the antimicrobial drugs except enrooxacin are shown in Fig. 2. All isolates were sensitive to enrooxacin. Resistance to tetracycline was present in over 95% of the recent isolates and in 40% of the old isolates. Resistance to apramycin was not observed in old isolates and was less than 20% in the recent isolates. The numbers and sizes of plasmids were determined for the 135 isolates. All isolates carried two to eight plasmids whose sizes varied from 3 kb to greater than 167 kb. Most recent isolates possessed a cluster of three to ve large plasmid bands (90150 kb) and 99% of these isolates also had a cryptic plasmid of 5.1 kb (Fig. 3). Sixty percent of the old isolates had a 5.1 kb plasmid band. Southern hybridization showed that the recent ETEC isolates typically possessed a cluster of plasmid bands in the 7288 kb range that encoded LT, STb, K88, and STa, and a single band in the range of 5469 kb, which encoded alpha hemolysin and EAST1 (Gyles, unpublished). Among the old isolates, the cluster or a single band at 7288 kb usually encoded LT, STb, K88, and EAST1, whereas a single band at 5469 kb encoded alpha hemolysin. Bands in the range of 120167 kb were associated with drug resistance. The cryptic 5.1 kb plasmid was sequenced and shown to be highly related to the ColE1-related ColJs plasmid of Shigella exneri (Smajs and Weinstock, 2001) except that the region corresponding to the colicin proteins in pCG7 has no homology with sequences in the databases. There was no feature of the plasmid proles that distinguished recent and old isolates. Cloning of the secondary PCR products into the pGEM T-EASY vector resulted in 78 colonies on medium containing ampicillin, X-gal, and isopropyl thiogalactoside (IPTG). White colonies were subcultured on LB agar plus ampicillin and were screened for plasmid content by the cracking method (Nishimura et al., 1990). Screening showed that 75 of the colonies contained inserts, whose product sizes ranged from 250 to >724 bp. The plasmids in 46 randomly selected clones were extracted, and the inserted DNA fragments were sequenced. Data from 31 clones whose inserted DNA sequences showed high homology with E. coli K12 DNA or were present in both tester and driver genomic DNA are not

94

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

Fig. 3. Plasmid proles of recent and old O149 ETEC isolates. Lanes 1, 2: marker plasmids. Lanes 14: plasmids from four recent isolates. The 5.1 kb plasmid (lanes 13) was found in almost all recent isolates. Lanes 58: four old isolates.

shown. These 31 clones included clones with as many as six identical copies. The results of BLAST analyses of the remaining 15 sequences are shown in Table 2. Four clones had high homology with the msbA gene (multicopy suppressor of the HtrB temperature-sensitive phenotype) (Karow and Georgopoulos, 1993), which is involved in lipid A biosynthesis, and two clones showed high homology with waaT, another gene implicated in LPS biosynthesis (Table 2). One clone had an insert whose translated protein sequence was 50% identical to a putative acetyl transferase, which may be involved in LPS modication. Five clones contained DNA inserts for which no homology was detected in a BLAST search. Ten clones showed high homology (9199%) with sequences in the F plasmid. One clone had homology with the ColIb plasmid; one showed homology with DNA from the mobile genetic element, Tn1721; and one clone was highly homologous with a region in the S. exneri SRL pathogenicity island (PAI). The inserts in clones with homology to genes in E. coli K12 (msbA and waaT), and clones 13, 22, and 25 (Table 2) were detected in both driver and tester genomic DNA. Inserts in the other clones were unique to the tester genomic DNA.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101 Table 2 Analysis of selected DNA sequences recovered following subtractive hybridization Clone (insert in bp) 32, 38, 39, 45 (264) 9, 18 (515) 3 (501) 1 (>169) 2 (190) 29 (590) Homologous gene (% identity) E. coli (1101511272) msbA (98%) E. coli waaT (99%) None None None None K12a Homologous protein (% identity) MsbA (100%)

95

33 (420) 16 (458) 74 (>724)

None F plasmid genomic DNA (23642820) (98%) F plasmid genomic DNA (1034510546 and 1944919573) (92%) Transposon Tn1721 (99%) (gi48194) Shigella exneri 2a SRL pathogenicity island (98%) (gi21450881)

UDP-galactose:(glucosyl) LPS alpha1, 2-galactosyltransferase waaT (99%) Putative acetyltransferase of Actinobacillus actinomycetemcomitans (50%) (87/169) None None 110280: Conserved hypothetical protein of Clostridium perfringens (39%); 137310: Helicobacter pylori putative protein jhp0462 (43%) 101412: Borrelia burgdorferi plasmid partition protein (38%) YacA (81%) Transposon Tn1000 (405584) (91%); IncFIB replicon (1135) (96%); unknown (136404) Probable methyl accepting chemotaxis protein (Tn1721) (84%) 168269 is related to E. coli hypothetical membrane protein YcdU (96%)

27 (616) 7 (268)

All F plasmid DNA sequences are numbered according to gi:8918823 which lists the 99159 bp of the complete sequence of the F plasmid. a All E. coli K12 sequences are numbered according to gi:6626251 which lists the 4639221 bp of the complete sequence of strain MG1655.

4. Discussion The major difference between recent and old isolates in known virulence genes was that 92% of the recent isolates possessed the estA gene (encoding STa), whereas all the old isolates lacked this gene. The nding that the recent STa-positive isolates were all H10 and the other isolates were all H43 was very helpful in clarifying the picture. It appears that the recent outbreak-associated isolates represent a new serotype and are distinctly different from the old isolates. To our knowledge, this is the rst report of the H43 antigen being associated with O149 ETEC. Interestingly, the O149 ETEC is frequently identied as STa-negative (Blanco et al., 1997; Sarrazin et al., 2000; Bischoff et al., 2002; Frydendahl, 2002), but there are some reports of STa-positive isolates of this serotype (Garabal et al., 1996; Frydendahl, 2002). The ndings in the present study underscore the importance of determination of H antigens in characterizing E. coli. In a study in Quebec, STa was also exclusively associated with recent O149 porcine ETEC, but the STa-positive isolates constituted a only 48% of the recent isolates (Fontaine et al., 2001). Sarrazin et al. (2000) found that, as in Ontario, ETEC of serotype O149:K91:K88ac

96

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

were the most prevalent in Swiss pigs. However, these O149 ETEC possessed genes for LT and STb but not for STa. Interestingly, STa is associated primarily with ETEC that cause diarrhea in neonatal pigs and calves (Burgess et al., 1978; Fairbrother et al., 1988; Ojeniyi et al., 1994), but pigs in the postweaning period have a greater susceptibility to this enterotoxin than do unweaned pigs (Stevens et al., 1972; Mezoff et al., 1991). Because LT, STb, and STa act on intestinal epithelial cells by different mechanisms (Nataro and Kaper, 1998), it is expected that the presence of the gene for STa would enhance the capacity of strains to cause diarrhea. The ndings with respect to the astA gene for EAST1 enterotoxin are similar to those of recent reports in which the gene has been found in a high percentage of porcine ETEC (Choi et al., 2001; Fontaine et al., 2002; Frydendahl, 2002). In Quebec, over 90% of O149 ETEC from pigs with diarrhea from 1986 to 2000 had the astA gene (Fontaine et al., 2001). In Denmark, the astA gene was detected in 100% of O149 porcine ETEC (Frydendahl, 2002). These ndings suggest that it may be worthwhile to determine the role of EAST1 in diarrhea caused by porcine ETEC. The gene for Paa was detected in 17% of recent isolates compared with 67% reported among O149 ETEC in Quebec (Fontaine et al., 2001). In both studies, there was a much higher prevalence among recent compared with old isolates. Since the Paa protein is secreted by a type III secretion system (An et al., 1999), it may play a role in pathogenesis. The high frequency of resistance to multiple antimicrobials among recent ETEC isolates is in agreement with recent reports (Fairbrother, 1999; Josephson et al., 1999; Amezcua et al., 2002). There is no indication that drug resistance enhances the virulence of ETEC, but virulence genes are sometimes associated with drug resistance genes (So et al., 1976; Gyles et al., 1977; Franklin et al., 1981; Hunter et al., 1994). Conjugation experiments with selected recent isolates showed that drug resistance and enterotoxin genes were co-transferred (Gyles, unpublished). Plasmids that carry virulence factors in porcine ETEC vary in size. The genes for LT and STb are often found on the same plasmid and range from 67 to105 kb (Gyles et al., 1977; Franklin et al., 1981; Franklin and Mollby, 1983; Wasteson and Olsvik, 1991). The K88 genes are carried on plasmids that range from 75 to 177 kb (Shipley et al., 1978; de Graaf, 1990; Wasteson and Olsvik, 1991; Bertin, 1992; Mainil et al., 1998). Genes for STa have been reported on plasmids of molecular sizes from 32 to142 kb (Gyles et al., 1974; Harnett and Gyles, 1985; So et al., 1979; Bertin, 1992). Yamamoto and Nakazawa (1997) showed that the gene for EAST1 was carried on a 58 kb plasmid in one isolate and on an 85 kb plasmid in another isolate. The gene for hemolysin in porcine E. coli has been located on a 50 kb plasmid (Beutin et al., 1986). These ndings indicate that it is difcult to associate specic virulence genes with a certain size of plasmid. The old O149 ETEC isolates that were available for examination constituted a smaller collection and had been stored for a longer time compared with recent isolates. One cannot rule out the possibility that, because of the low numbers examined, O149:H10 isolates were missed among the old isolates. However, it is clear that O149:H10 isolates were the dominant serotype among the new and O149:H43 was dominant among the old isolates. Consideration was also given to the possibility that old isolates may have lost plasmid-encoded estA genes during storage. While this is possible, it is highly unlikely. The plasmid-encoded genes for LT, STb, K88, EAST1, and alpha hemolysin were all maintained on the 35 old strains.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

97

Besides, a similar study of 66 O149 ETEC in Quebec found that the estA gene was absent from strains in the years 19781989. Subtractive hybridization identied a number of DNA fragments that may be unique to the recent O149 isolate. One set of these fragments consisted of sequences that are related to LPS biosynthesis. Four clones had sequences that were 98% identical to that of E. coli K12 msbA and two had homology to the waaT gene (Heinrichs et al., 1998). The msbA gene encodes the proposed inner membrane lipid ippase which is an essential transporter involved in LPS biosynthesis (Doerrler and Raetz, 2002). Although it was present in both tester and driver strains, it is possible that the tester O149 strain, like O157:H7 EHEC, possesses additional plasmid-borne genes homologous to chromosomal genes involved in LPS biosynthesis. Interestingly, Janke et al. (2001) also identied clones with genes involved in LPS biosynthesis when they conducted subtractive hybridization involving the uropathogenic E. coli strain 536 as tester and E. coli K12 as driver. Perhaps these pathogenic E. coli possess additional mechanisms for modulation of the LPS that is produced. The sequences for four clones had no homology with sequences in the databases. The translated protein from the insert in one of these clones (#3) was most closely related to a putative acetyl transferase involved in the serotype-specic polysaccharide antigen of Actinobacillus actinomycetemcomitans (gi:9309325) and to acetyl transferases implicated in O antigen biosynthesis in other bacteria including Vibrio cholerae (Li et al., 2002). Two clones contained DNA that is highly related to sequences in the F plasmid (Table 2). One of these clones encoded a portion of the IncFIB replication region, which is related to regions in the F plasmid and the virulence plasmid of E. coli O157:H7. The arrangement is different from that of the F plasmid in which the transposon Tn1000 interrupts the FIC region, whereas in this clone this transposon is associated with the FIB replication region. These two replication regions are separated by about 10 kb of DNA in the F plasmid (Firth et al., 1996). The presence of this transposon in a recent O149 ETEC could promote cointegrate formation and transposition (Berg and Berg, 1996). It is likely that the recent isolate possessed an F-related plasmid which was not present in the old isolate. One clone (#7) had sequences with high homology to a region of the Shigella resistance locus (SRL) pathogenicity island, a novel complex of genes which encode a ferric dicitrate transport system and a cluster of multiple antibiotic resistance determinants (Luck et al., 2001). Although the short sequence available provides no information on the genes encoded by a potential PAI in the O149 ETEC, it is of interest that the site of insertion, namely the serX tRNA gene (which is the site of insertion of the SRL PAI) was a part of the sequence encoded by clone #7. In conclusion, the O149 E. coli isolated in recent years (19982001) from weaned pigs with diarrhea possess the gene for one additional enterotoxin (STa) compared with old isolates which lacked this gene. These STa-positive ETEC belonged to a different serotype (O149:H10) from the old isolates (O149:H43). A cryptic 5.1 kb plasmid was detected in 99% of recent O149 ETEC isolates but only in 60% of old isolates. Subtractive hybridization yielded a number of unknown gene sequences and F plasmid-related sequences that were present in a recent isolate and absent from an old isolate. Genes implicated in LPS

98

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

biosynthesis were also identied, although they were not all unique to the recent isolate. These studies suggest a possible basis for an increased virulence of recent isolates and have provided several leads for genes which may contribute to virulence or survival properties of recent O149 porcine ETEC. Although the estA gene was present in 92% of recent isolates and absent from the old isolates, the evidence suggests that the estA-positive new isolates represented a new clone rather than old isolates which had acquired the estA gene. Furthermore, although the estA gene could contribute to enhanced virulence of new isolates there is presently no evidence to indicate whether it does contribute. Deletion of the estA gene from a new isolate could allow for tests of virulence of an isogenic pair of estA-positive and estA-negative O149:H10 ETEC.

Acknowledgements The research was supported by the Natural Sciences and Engineering Research Councils Canadian Research Network on Bacterial Pathogens of Swine. The authors are grateful to Janet Liao for technical assistance and to Dr. Jan MacInnes for helpful advice.

References
Amezcua, R., 2001. Post-weaning diarrhea and mortality caused by Escherichia coli: investigation of risk factors and control methods. M.Sc. Thesis, University of Guelph. Amezcua, R., Friendship, R.M., Dewey, C.E., Gyles, C., Fairbrother, J.M., 2002. Prevalence of hemolytic E. coli serogroups involved, and their antimicrobial resistance patterns. Can. J. Vet. Res. 66, 7378. An, H., Fairbrother, J.M., Desautels, C., Harel, J., 1999. Distribution of a novel locus called paa (porcine attaching and effacing associated) among enteric Escherichia coli. Adv. Exp. Med. Biol. 473, 179184. Berg, C.E., Berg, M.D., 1996. Transposable element tools for microbial genetics. In: Neidhardt, F.C., Curtiss, R. III, Gross, C., Ingraham, J., Lin, E.C.C., Low, K.B., Magasanik, B., Reznikoff, W.S., Riley, M., Schaechter, M., Umbarger, H.E. (Eds.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 2. American Society for Microbiology, Washington, DC, pp. 25882612. Bertin, A., 1992. Plasmid content and localisation of the STaI (STaP) gene in enterotoxigenic Escherichia coli with a non-radioactive polynucleotidegene probe. J. Med. Microbiol. 37, 141147. Beutin, L., Montenegro, M., Zimmermann, S., Stephan, R., 1986. Characterization of hemolytic strains of Escherichia coli belonging to classical enteropathogenic O-serogroups. Zentralbl. Bakteriol. Mikrobiol. Hyg. 261, 266279. Bischoff, K.M., White, D.G., McDermott, P.F., Zhao, S., Gaines, S., Maurer, J.J., Nisbet, D.J., 2002. Characterization of chloramphenicol resistance in beta-hemolytic Escherichia coli associated with diarrhea in neonatal swine. J. Clin. Microbiol. 40, 389394. Blanco, M., Blanco, J.E., Gonzalez, E.A., Mora, A., Jansen, W., Gomes, T.A., Zerbini, L.F., Yano, T., de Castro, A.F., Blanco, J., 1997. Genes coding for enterotoxins and verotoxins in porcine Escherichia coli strains belonging to different O:K:H serotypes: relationship with toxic phenotypes. J. Clin. Microbiol. 35, 2958 2963. Bosworth, B.T., Casey, T.A., 1997. Identication of toxin and pilus genes in porcine E. coli using polymerase chain reaction (PCR) with multiple primer pairs. In: Proceedings of the Abstracts of the 97th Annual General Meeting, Abstract B-509. American Society for Microbiology, Miami Beach, FL. Choi, C., Cho, W., Chung, H., Jung, T., Kim, J., Chae, C., 2001. Prevalence of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1) gene in isolates in weaned pigs with diarrhea and/or edema disease. Vet. Microbiol. 81, 6571.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

99

de Graaf, F.K., 1990. Genetics of adhesive mbriae of intestinal Escherichia coli. Curr. Top. Microbiol. Immunol. 151, 2953. Doerrler, W.T., Raetz, C.R., 2002. ATPase activity of the MsbA lipid ippase of Escherichia coli, J. Biol. Chem. Fairbrother, J.M., 1999. Neonatal E. coli diarrhea. In: Straw, B.E., DAllaire, S., Mengeling,W.L., Taylor, D.J. (Eds.), Diseases of Swine. Iowa State University Press, Ames, IA, pp. 433441. Fekete, P.Z., Gerardin, J., Jacquemin, E., Mainil, J.G., Nagy, B., 2002. Replicon typing of F18 mbriae encoding plasmids of enterotoxigenic and verotoxigenic Escherichia coli strains from porcine postweaning diarrhoea and oedema disease. Vet. Microbiol. 85, 275284. Firth, N., Ippen-ihler, K., Skurray, R.A., 1996. Structure and function of the F factor and mechanism of conjugation. In: Neidhardt, F.C., Curtiss III, R., Gross, C., Ingraham, J., Lin, E.C.C., Low, K.B., Magasanik, B., Reznikoff, W.S., Riley, M., Schaechter, M., Umbarger, H.E. (Eds.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 2. American Society for Microbiology, Washington, DC, pp. 2377 2401. Fontaine, F., Peres, C., Gyles, C.L., Fairbrother, J.M., 2002. Trends in O149:K91 enterotoxigenic Escherichia coli from pigs in Quebec. In: Proceedings of the 17th Congress of the International Pig Veterinary Society. Ames, Iowa, IA, p. 70. Franklin, A., Mollby, R., 1983. Concurrent transfer and recombination between plasmids encoding for heat-stable enterotoxin and drug resistance in porcine enterotoxigenic Escherichia coli. Med. Microbiol. Immunol. 172, 137147. Franklin, A., Soderlind, O., Mollby, R., 1981. Plasmids coding for enterotoxins, K88 antigen and colicins in porcine Escherichia coli strains of O-group 149. Med. Microbiol. Immunol. 170, 6372. Frydendahl, K., 2002. Prevalence of serogroups and virulence genes in Escherichia coli associated with postweaning diarrhoea and edema disease in pigs and a comparison of diagnostic approaches. Vet. Microbiol. 85, 169182. Gaastra, W., de Graaf, F.K., 1982. Host-specic mbrial adhesins of noninvasive enterotoxigenic Escherichia coli strains. Microbiol. Rev. 46, 129161. Garabal, J.I., Gonzalez, E.A., Vazquez, F., Blanco, J., Blanco, M., Blanco, J.E., 1996. Serogroups of Escherichia coli isolated from piglets in Spain. Vet. Microbiol. 48, 113123. Gyles, C., So, M., Falkow, S., 1974. The enterotoxin plasmids of Escherichia coli. J. Infect. Dis. 130, 4049. Gyles, C.L., Palchaudhuri, S., Maas, W.K., 1977. Naturally occurring plasmid carrying genes for enterotoxin production and drug resistance. Science 198, 198199. Harel, J., Lapointe, H., Fallara, A., Lortie, L.A., Bigras-Poulin, M., Lariviere, S., Fairbrother, J.M., 1991. Detection of genes for mbrial antigens and enterotoxins associated with Escherichia coli serogroups isolated from pigs with diarrhea. J. Clin. Microbiol. 29, 745752. Harnett, N.M., Gyles, C.L., 1985. Linkage of genes for heat-stable enterotoxin, drug resistance, K99 antigen, and colicin in bovine and porcine strains of enterotoxigenic Escherichia coli. Am. J. Vet. Res. 46, 428 433. Harris, R.J.C., 1954. Biological Application of Freezing and Drying. Levin and Munksgaard, New York. Heinrichs, D.E., Yethon, J.A., Amor, P.A., Whiteld, C., 1998. The assembly system for the outer core portion of R1- and R4-type lipopolysaccharides of Escherichia coli. The R1 core-specicbeta-glucosyltransferase provides a novel attachment site for O-polysaccharides. J. Biol. Chem. 273, 2949729505. Hunter, J.E., Bennett, M., Hart, C.A., Shelley, J.C., Walton, J.R., 1994. Apramycin-resistant Escherichia coli isolated from pigs and a stockman. Epidemiol. Infect. 112, 473480. Janke, B., Dobrindt, U., Hacker, J., Blum-Oehler, G., 2001. A subtractive hybridisation analysis of genomic differences between the uropathogenic E. coli strain 536 and the E. coli K-12 strain MG1655. FEMS Microbiol. Lett. 199, 6166. Josephson, G., Smart, N., 1998. K88 strains of E. coli. AHL Newslett. 2, 45. Josephson, G., Smart, N., McEwen, B., van Dreumel, T., Gough, J., Hazlett, M., 1999. K88 coliform enteritis in swine. AHL Newslett. 3, 23. Josephson, G., Archambeault, M., Gyles, C.L., 2000. Colibacillosis in pigs in 1999. AHL Newslett. 4, 89. Karow, M., Georgopoulos, C., 1993. The essential Escherichia coli msbA gene, a multicopy suppressor of null mutations in the htrB gene, is related to the universally conserved family of ATP-dependent translocators. Mol. Microbiol. 7, 6979.

100

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

Li, M., Shimada, T., Morris Jr., J.G., Sulakvelidze, A., Sozhamannan, S., 2002. Evidence for the emergence of non-O1 and non-O139 Vibrio cholerae strains with pathogenic potential by exchange of O-antigen biosynthesis regions. Infect. Immun. 70, 24412453. Luck, S.N., Turner, S.A., Rajakumar, K., Sakellaris, H., Adler, B., 2001. Ferric dicitrate transport system (Fec) of Shigella exneri 2a YSH6000 is encoded on a novel pathogenicity island carrying multiple antibiotic resistance genes. Infect. Immun. 69, 60126021. Mainil, J.G., Daube, G., Jacquemin, E., Pohl, P., Kaeckenbeeck, A., 1998. Virulence plasmids of enterotoxigenic Escherichia coli isolates from piglets. Vet. Microbiol. 62, 291301. Menard, L.P., Dubreuil, J.D., 2002. Enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1): a new toxin with an old twist. Crit. Rev. Microbiol. 28, 4360. Mezoff, A.G., Jensen, N.J., Cohen, M.B., 1991. Mechanisms of increased susceptibility of immature and weaned pigs to Escherichia coli heat-stable enterotoxin. Pediatr. Res. 29, 424428. Moon, H.W., 1990. Colonization factor antigens of enterotoxigenic Escherichia coli in animals. Curr. Top. Microbiol. Immunol. 151, 147265. Nagy, B., Fekete, P.Z., 1999. Enterotoxigenic Escherichia coli (ETEC) in farm animals. Vet. Res. 30, 259284. Nagy, B., Whipp, S.C., Imberechts, H., Bertschinger, H.U., Dean-Nystrom, E.A., Casey, T.A., Salajka, E., 1997. Biological relationship between F18ab and F18ac mbriae of enterotoxigenic and verotoxigenic Escherichia coli from weaned pigs with oedema disease or diarrhoea. Microb. Pathog. 22, 111. Nataro, J.P., Kaper, J.B., 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11, 142201. Nishimura, A., Morita, M., Nishimura, Y., Sugino, Y., 1990. A rapid and highly efcient method for preparation of competent Escherichia coli cells. Nucleic Acids Res. 18, 6169. Ojeniyi, B., Ahrens, P., Meyling, A., 1994. Detection of mbrial and toxin genes in Escherichia coli and their prevalence in piglets with diarrhoea. The application of colonyhybridization assay, polymerase chain reaction and phenotypic assays. Zentralbl. Veterinarmed. 41, 4959. Orskov, I., Orskov, F., Smith, H.W., Sojka, W.J., 1975. The establishment of K99, a thermo-labile, transmissible Escherichia coli K antigen, previously called Kco, possessed by calf and lamb enteropathogenic strains. Acta Pathol. Microbiol. Scand. 83, 3136. Osek, J., 1999. Prevalence of virulence factors of Escherichia coli strains isolated from diarrheic and healthy piglets after weaning. Vet. Microbiol. 68, 209217. Osek, J., 2000. Clonal analysis of Escherichia coli strains isolated from pigs with post-weaning diarrhea by pulsed-eld gel electrophoresis. FEMS Microbiol. Lett. 186, 327331. Osek, J., 2003. Detection of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1) gene and its relationship with mbrial and enterotoxin markers in E. coli isolates from pigs with diarrhoea. Vet. Microbiol. 91, 6572. Parma, A.E., Sanz, M.E., Vinas, M.R., Cicuta, M.E., Blanco, J.E., Boehringer, S.I., Vena, M.M., Roibon, W.R., Benitez, M.C., Blanco, J., Blanco, M., 2000. Toxigenic Escherichia coli isolated from pigs in Argentina. Vet. Microbiol. 72, 269276. Sarrazin, E., Fritzsche, C., Bertschinger, H.U., 2000. Main virulence factors in Escherichia coli isolates from swine over two weeks old with edema disease and/or E. coli diarrhea. Schweiz. Arch. Tierheilkd. 142, 625630. Shipley, P.L., Gyles, C.L., Falkow, S., 1978. Characterization of plasmids that encode for the K88 colonization antigen. Infect. Immun. 20, 559566. Smajs, D., Weinstock, G.M., 2001. Genetic organization of plasmid ColJs, encoding colicin Js activity, immunity, and release genes. J. Bacteriol. 183, 39493957. Smith, H.W., Linggood, M.A., 1971. Observations on the pathogenic properties of the K88, Hly and Ent plasmids of Escherichia coli with particular reference to porcine diarrhoea. J. Med. Microbiol. 4, 467485. So, M., Heffron, F., McCarthy, B.J., 1979. The E. coli gene encoding heat stable toxin is a bacterial transposon anked by inverted repeats of IS1. Nature 277, 453456. Stevens, J.B., Gyles, C.L., Barnum, D.A., 1972. Production of diarrhea in pigs in response to Escherichia coli enterotoxin. Am. J. Vet. Res. 33, 25112526. Vazquez, F., Gonzalez, E.A., Garabal, J.I., Blanco, J., 1996. Fimbriae extracts from enterotoxigenic Escherichia coli strains of bovine and porcine origin with K99 and/or F41 antigens. Vet. Microbiol. 48, 231241. Wasteson, Y., Olsvik, O., 1991. Specic DNA fragments coding for ST1 and LT1 toxins, and K88 (F4) adhesin in enterotoxigenic Escherichia coli. Zentralbl. Veterinarmed. 38, 445452.

B.N. Noamani et al. / Veterinary Microbiology 97 (2003) 87101

101

Wilson, R.A., Francis, D.H., 1986. Fimbriae and enterotoxins associated with Escherichia coli serogroups isolated from pigs with colibacillosis. Am. J. Vet. Res. 47, 213217. Wittig, W., Fabricius, C., 1992. Escherichia coli types isolated from porcine E. coli infections in Saxony from 1963 to 1990. Zentralbl. Bakteriol. 277, 389402. Yamamoto, T., Nakazawa, M., 1997. Detection and sequences of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 gene in enterotoxigenic E. coli strains isolated from piglets and calves with diarrhea. J. Clin. Microbiol. 35, 223227.