eaea Genes in Escherichia coli Derived from Japanese Patients with Sporadic Diarrhea

eaea Genes in Escherichia coli Derived from Japanese Patients with Sporadic Diarrhea Mitsugu YAMAZAKI* & Makoto SAITO Aichi Prefectural Institute of Public Health (*Present adress: Hokubu Market Food Inspection Office, Aichi Prefectural Government) Kazuhisa INUZUKA Anjo Kosei Hospital Shogo SHIMA and Hiroshige TANIWAKI School of Medicine, Fujita Health University Kenichirou ITO National Institute of Infection Deseases (Received: May 8, 1997) (Accepted: July 25, 1997) Key words: eaea gene, attaching and effacing Escherichia coli, diarrheagenic E. coli, sporadic diarrhea Abstract To investigate the prevalence of attaching and effacing Escherichia coli, we examined 364 strains isolated from the feces of 9,684 patients with diarrhea at the Anjo Kosei Hospital in Japan for the presence of eaea. Twenty-nine (8%) of the strains were eaea positive. Of enteropathogenic E. coli (EPEC), 11 of the 87 (13%) strains were for the positive eaea gene. The serotypes and the numbers of eaea-positive strains among the strains tested were as follows: O26:H-(2/3), O55:H7 (4/4), O55:H-(2/ 2) and O128:H2(3/3). Two enterohemorrhagic E. coli (EHEC) strains (Verotoxin positive O157:H7) were also eaea positive. Among 260 non-epec strains that were not categorized as diarrheagenic E. coli, 16 (6%) were eaea positive. Those serotypes were as follows: O15:H2, O20:H6, O28:H28, O63:H6, : H7, O28:H6, O153:H19 and O157:H45. EPEC strains including O18:H7 and six other serotypes, enteroinvasive E. coli (EIEC), and enterotoxigenic E. coli (ETEC) were all eaea negative. Introduction Enteropathogenic Escherichia coli (EPEC) is a bacterial pathogen that causes infantile diarrhea. It includes specific serotypes that are epidemiologically confirmed pathogens that do not produce known enterotoxins and which are not enteroinvasive1,2). EPEC produce characteristic histopathogenic lesions in the infant intestine and in experimental animal models3,4). Moon and coworkers5) described this lesion as attaching and effacing (A/E), characterized by intimate adherence of bacteria to the enterocyte and dissolution of the brush border at the site of bacterial attachment. The A/E lesion appears to be associated with fluid secretion and diarrhea. Therefore Moon et al. designated this type of E. coli as attaching and effacing E. coli (AEEC). Non-enteropathogenic E. coli (non-epec)6, 7 and enterohemorrhagic E. coli (EHEC)8) `10) in fact Correspondence to : Mitsugu YAMAZAKI Hokubu Market Food Inspection Office, Aichi Prefectural Government 107, Hattan, Toyoba, Toyoyama-cho, Nishikasugai-gun, Aich 480-02, Japan

cause the A/E lesion. Jerse and colleagues" have identified the chromosomal attaching and effacing gene (eae), renamed eaea, which is necessary for causing A/E lesions. They cloned and sequenced this gene and found that it encodes a 94-kDa membrane protein12). To investigate the prevalence of AEEC among sporadic diarrhea cases in Japan, we examined diarrheal strains of E. coli for the presence of the eaea sequence and its relationship to serotypes. Materials and Methods Bacterial strains From January 1989 through December 1992, stools were obtained from 7,597 infants or young children and from 2,087 adults with acute diarrhea who visited at Anjo Kosei Hospital. The samples were inoculated onto DHL agar plates. Three to five coliform colonies on each plate were selected and identified. Other enteropathogens, namely, Vibrio sp., Shigella sp., Salmonella sp., Campylobacter sp., Aeromonas sp., Plesiomonas sp, and Yersinia sp., were simultaneously examined by standard procedures. The positive controls for eaea were E2348/69 and 886L, which show localized adherence to HEp-2 cells. The negative controls for eaea were three strains: 251 that shows diffused adherence13), as well as 17-2 and JM221 that show aggregative adherence14). The bacteria were stored at -80 Ž in L broth containing 50% glycerol. Serotyping antigens were determined by slide agglutination tests using boiled bacteria cells. H antigens were determined by tube agglutination tests. Forty-three OK antisera and 22 H antisera were purchased from Denka Seiken Ltd. (Tokyo, Japan). H1 and H14 were purchased from Difco Laboratories (Detroit, Mich. USA). We defined EPEC serotypes as follows: O18:H7, H14 and H-; O26:H11 and H-; O44: H18, H34, and H-; O55:H6, H7, and H-; O86:H2, H34 and H-; O111:H2, H12, H21 and H-; O114: H2; O119:H6 and H-; O125:H21 and H-; O126:H2, H27 and H-; O127:H6, H9, H21 and H-; O128:H2, H7, H6; O158:H23 as previously described7,15,16). DNA hybridization assay The alkaline phosphatase-conjugated oligo-nucleotid probes, eaeal (5'- CTGAAAGCGAAATGATGAAGGC-3'), were designed from nucleotide sequences of eaea in E2348/ 69 (GenBank accession No. M58154 and M34051), and prepared by Toyobo Ltd. (Tokyo, Japan). Bacteria grown on tryptic soy agar were spread onto nylon membranes (Hybond N+; Amersham International plc, UK), lysed with alkali and hybridized at 55 Ž for 15 min according to the manufacturer's manual (Toyobo). The membranes were washed, then the probes were detected by enzymatic eaea-pcr reaction. eaekl (5'-GCTTAGTGCTGGTTTAGGAT-3'), and eaek4 (5'-TCGCCGTTCAGAGAATCGC-3'), which were designed from nucleotide sequences near the 5'-terminal of eaea, were used as primers17). A small cluster of bacteria was suspended in 100ƒÊl of distilled water, and boiled for 10 min, then 4.8ƒÊl aliquots were added to 5.2ƒÊl of PCR mixture18), and amplified by 25 cycles of denaturation at 94 Ž for 30s, annealing at 50 Ž for 1 min,, and extension at 72 Ž for 1 min. The amplified products were resolved by polyacrilamide gel electrophoresis, and visualized by ethidium bromides fluorescence. All samples positive for the eaea-probe were amplified in this manner and compared with 56 eaea-probe negatives. Examination for other enteropathogenic factors Heat-labile enterotoxins were detected by means of reversed passive latex agglutination assay (Denka Seiken). Heat-stable enterotoxins were detected by enzyme-linked immunoassay (Denka Seiken). Invasion was examined by PCR with inve-primers18). The cytotoxcity tests with Vero cells

were conducted as described previously19). The tentative positives were confirmed by PCR with the primers designed by Karch and Meyer20). Results Serotyping and categorization of isolated E. coli strains Three hundred and sixty-four E. coli strains that reacted with 37 O:K antisera were isolated from 364 patients. Among these, 299 strains (82%) reacted with 21 H antisera and were classified into 89 serotypes. These strains were examined for enterotoxin production, cell invasiveness, and cytotoxicity. Finally, 104 strains (29%) were divided into the following pathogenic categories: 87 EPEC, 13 ETEC, 2 EIEC and 2 EHEC. Two hundred and sixty strains that did not fit into the four categories are referred to as non-epec in this paper. Prevalence of eaea sequences in EPEC, EHEC, ETEC, EIEC and non-epec strains All strains were tested for hybridization with probes for eaea. The eaea probe hybridized to E. coli strains E2348/69 and 886L (localized adhesion positive), but not with strains 251 (diffuse adhesion positive), 17-2 and JM221 (enteroaggregative) (data not shown). Twenty-nine strains (8%) derived Table 2 Serotypes and number of eaea positive strains in EPEC Table 1 Frequency of eaea-positive strains among 364 strains derived from patients with sporadic diarrhea Table 3 Serotypes and number of eaea-positive strains in non-epec *Untypable H antigens.

from diarrhea were eaea probe positive (Table 1). All of these positives were also PCR positive. In EPEC, 11 of the 87 strains (13%) were eaea positive (Table 2). All E. coli strains with O26:H-, 055: H7, O55:H- and O128:H2 serotypes possessed the eaea sequence. However, the eaea sequence was not detected in the frequently isolated serotype O18:H7, or in O86:H-, O111:H21, O111:H-, O119: H- O126: Table 4 Frequency of eaea-negative strains in non-epec *Untypable H antigen.

Table 5 Age of diarrheal patients from whom eaea-positive E. coli were isolated *H untypable. **Thirty -six-year old patient. H27 and O127:H21. Only one of the two O26:H-strains was eaea positive. Other serotypes were not isolated in this study. Two EHEC strains were isolated and both were of the O157:H7 serotype and eaea posiitve. All strains belonging to ETEC and EIEC were eaea negative. Two hundred and sixty strains were categorized as non-epec, of which the eaea positives are shown in Table 3. The eaea sequences were detected in serotypes O15:H2, O20:H6, O28:H28, O63:H6, O153:H7, O153:H19 and O157: H45, and in the serogroups O115, O125, O153 and O157, which did not react with any commercial H antisera. Serotypes of non-epec which did not possess eaea sequences are shown in Table 4. Two hundred forty-four strains were divided into 36 serogroups (65 serotypes), which did not possess eaea sequences. Among them, O1:H7, O1:H- and O86:H27, were frequently isolated. Other pathogens simultaneously isolated with eaea-positive E. coli were as follows: Campylobacter coli with O26:H- (one of two cases); Aeromonas caviae or Campylobacter jejuni with O128:H2 (1/4 each); Aeromonas hydrophila with O63:H6 (1/1). The ages of patients with the eaea-positive E. coli, including two with EHEC, ranged from 0 to 36 years (Table 5). Most of those were 10 years of age (90%), with the majority being 3 years old (66%). Discussion We investigated E. coli from patients with diarrhea in Japan for the prevalence of eaea sequences. We found that 13% of the EPEC strains possessed DNA sequences homologous to the eaea sequence of strain E2348/69. This frequency is much lower than that reported by Knutton et al.15) who stated that 11 of the 20 EPEC-serotype strains (55%) isolated from patients with diarrhea in London were AEEC according to the Fas test. This discrepancy may have arisen from the differences in EPEC serotypes derived in the two studies. Strains classified as O18:H7 and O111:H21, which were eaea negative, were particularly frequent in this study. However, Knutton et al., isolated only one

member of the O111:H21 strain. The serotypes O26:H-, O55:H7, and O55:H- reported as AEEC, were detected in this study at a similar frequency and all strains except one O26:H- were eaea-positives. However, serotypes O114:H2, O125:H21 and O127:H6, also described as AEEC, were not isolated in the study. Furthermore, O119:H6 and O142:H6, which were identified as AEEC in Bangkok, Thailand, by Echeverria et al.7), were not isolated in our study. There may be a domestic difference in AEEC serotypes. In this study, two O157:H7 strains were eaea positive, which agrees with other reports10,11). EIEC and ETEC, whose virulence mechanisms differ from those of EPEC and EHEC, were eaea negative. A/E activities have been found in non-epec6,7). Sixteen of 260 non-epec strains (6%) isolated in Japan were found to be eaea positive. Albert et al.6) isolated O15:H2 from a child in Bangladesh that was AEEC. Echeverria et al.7) have also shown that O157:H45 is AEEC. We showed here that the O28: H28, O63:H6, O153:H7 and O153:H19 serotypes and the O115:HUT, O153:HUT and O157:HUT serogroups possess eaea sequences. In Japan, O1:H7 strains have been isolated during sporadic diarrhea outbreaks. However, whether or not these serotypes should be included in EPEC remains uncertain. Thirty-four of the strains isolated in this study did not possess the eaea sequence. Acknowledgment We thank Blay Kay for providing 886L, 251 and JM221 strains, and James Nataro strains 2348/69 and 17-2. for providing References 1) Edelman R & Levine MM: Summary of a workshop on enteropahogenic Esherichia coll. J Infect Dis 1983; 147: 1108-1118. 2) Robins-Browne RM: Traditional enteropathogenic Escherichia coli of infantile diarrhea. Rev Infect Dis 1987; 9: 28-53. 3) Ulshen MH & Rollo JL: Pathogenesis of Escherichia coli gastroenteritis in man another mechanism. N Engl J Med 1980; 302: 99-101. 4) Rothbaum R, McAdams AJ, Giannella R & Partin JC: A clinicopathologic study of enterocyte-adherent Escherichia coli: a cause of protracted diarrhea in infants. Gastroenterology 1982; 83: 441-454. 5) Moon HW, Whipp SC, Argenzio RA, Levine MM & Giannella RA: Attaching and effacing activities of rabbit and human enteropathogenic Escherichia coli in pig and rabbit intestines. Infect Immun 1983; 41: 1340-1351. 6) Albert MJ, Alam K, Ansaruzzaman M et al: Localized adherence and attaching-effacing properties of nonenteropathogenic serotype of Escherichia coli. Infect Immun 1991; 59: 1864-1868. 7) Echeverria P, Orskov F, Orskov I, et al: Attaching and effacing enteropathogenic Escherichia colt as a cause of infantile diarrhea in Bangkok. J Infect Dis 1991; 164: 550-554. 8) Tzipori S, Wachsmuth IK, Chapman C, et al: The pathogenesis of haemorrhagic colitis caused by Escherichia colt 0157:H7 in gnotobiotic piglets. J Infect Dis 1986; 154: 712-716. 9) Knutton S, Baldwin T, Williams PH & McNeish AS: Actin accumulation at sites of bacterial adhesion to tissue culture cells: Basis of a new diagnostic test for enteropathogenic and enterohemorrhagic Escherichia colt. Infect Immun 1989; 57: 1290-1298. 10) Gannon VPJ, Rashed M, King RK & Thomas EJG: Detection and characterization of the eae gene of shiga-like toxin-producing Escherichia colt using polymerase chain reaction. J Clin Microbiol 1993; 31: 1268-1274. 11) Jerse AE, Yu J, Tall BD & Kaper JB: A genetic locus of enteropathogenic Escherichia colt necessary for the production of attaching and effacing lesions on tissue culture cells. Proc Natl Acad Sci 1990; 87: 7839-7843. 12) Jerse AE & Kaper JB: The eae gene of enteropathogenic Escherichia colt encodes a 94-kilodalton membrane protein, the expression of which is influenced by the EAF plasmid. Infect Immun 1991; 59: 4302-4309. 13) Nataro JP, Scaletsky ICA, Kaper JB, Levine MM & Trabulsi LR: Plasmid-mediated factors conferring diffuse and localized adherence of enteropathogenic Escherichia colt. Infect Immun 1985; 48: 378-383. 14) Nataro JP, Kaper JB, Robins-Browne R, Prado V, Vial P & Levine MM: Patterns of adherence of diarrheagenic Escherichia colt to HEp-2 cells. Pediatr Infect Dis J 1987; 6: 829-831. 15) Knutton S, Phillips AD, Smith HR, et al: Screening for enteropathogenic Escherichia colt in infants with diarrhea

by the fluorescent-actin staining test. Infect Immun 1991; 59: 365-371. 16) Moyenuddin M, Wachsmuth IK, Moseley SL, Bopp CA & Blake PA: Serotype, antimicrobial resistance, and adherence properties of Escherichia coli strains associated with outbreaks of diarrheal illness in children in the United States. J Clin Micro 1989; 27: 2234-2239. 17) Nakazawa M & Ito K: Experimental infection of infant rabbits with verocytotoxin-producing Escherichia coli of bovine origin. Jpn J Assoc Infect Dis 1995; 69: 772-776. 18) Ito K, Watanabe H, Toyosato M, Kosaka T & Mizuno K: Genetic analysis of Shigella pathogenesis and rapid detection method of Shigella virulence gene by the polymerase chain reaction. Nippon rinsho (extra issue) 1992; 50: 368-372. 19) Konowalchuk J, Speirs JI & Stavric S: Vero response to a cytotoxin of Escherichia coli. Infect Immun 1977; 18: 775-779. 20) Karch H & Meyer T: Single primer pair for amplifying segments of distinct shiga-like-toxin genes by polymerase chain reaction. J Clin Microbiol 1989; 27: 2751-2757.