Jpn. J. Infect. Dis., 62, 318-323, 2009
Epidemiological Report
Prevalence and Properties of Diarrheagenic Escherichia coli among Healthy Individuals in Osaka City, Japan Sami Fujihara1,2,3, Kentaro Arikawa2, Tetsu Aota2, Hiroshi Tanaka2, Hiromi Nakamura3, Takayuki Wada3, Atsushi Hase3, and Yoshikazu Nishikawa2* 1
Urban Living and Health Association, Osaka 530-0027; 2Osaka City University Graduate School of Human Life Science, Osaka 558-8585; and 3Osaka City Institute of Public Health and Environmental Sciences, Osaka 543-0026, Japan (Received April 13, 2009. Accepted June 5, 2009)
SUMMARY: The etiological roles of diarrheagenic Escherichia coli (DEC), including enteroaggregative E. coli (EAggEC), diffusely adherent E. coli (DAEC) and EAST1EC---a strain of E. coli that possesses no diarrheagenic characteristics other than the EAggEC heat-stable toxin 1 (EAST1) gene---remain controversial. To clarify the prevalence of DEC among healthy individuals in Osaka City, Japan, and to compare the virulence properties of strains previously isolated from diarrheal patients, fecal specimens were examined for DEC. Isolation rates of Shiga toxin-producing E. coli, enterotoxigenic E. coli and EAggEC were significantly lower among healthy adults than sporadic adult patients. There were no differences in enteropathogenic E. coli (EPEC), DAEC and EAST1EC between patients and healthy carriers. Subtyping of the intimin gene (eae) of EPEC, and measuring the IL-8 inductivity of DAEC on epithelial cells could provide criteria to distinguish strains in diarrheal patients from those in healthy carriers. Proper criteria should be established in order to diagnose subtypes of DEC as causative agents. sporadic diarrheal patients in Osaka City (13). We then suggested that DAEC comprises heterogeneous groups of organisms with variable virulence and that the subgroup of DAEC possessing afimbrial adhesin (Afa) genes and/or inducing high levels of interleukin-8 (IL-8) secretion in epithelial cells plays a role in causing sporadic diarrheal illnesses, particularly in pediatric fields (14-16). However, the enteropathogenicity and etiological roles of these atypical EPEC (17,18), EAggEC (19,20), DAEC (18,21) and EAST1EC strains remain somewhat controversial. Clinical microbiologists often find it difficult to assess the etiological significance of these isolates, particularly when the organisms are isolated from sporadic patients. It is helpful for clinical laboratories to know the prevalence of DEC among healthy people beforehand. In this study, a total of 2,230 fecal specimens from healthy adults were examined for DEC. The isolation rates and virulence properties were discussed in comparison with those among diarrheal patients in our previous study (13). Salmonella was examined as a representative enteric pathogen for comparison with DEC.
INTRODUCTION Escherichia coli is a normal inhabitant of the intestinal tract in humans and warm-blooded animals; however, certain strains cause enteric disease in their hosts and are referred to as diarrheagenic E. coli (DEC). Based on distinct epidemiological and clinical features, specific virulence determinants and other characteristic markers, such as enterotoxins and adherence phenotypes, DEC strains have been classified into the following six pathotypes (1): enteropathogenic E. coli (EPEC); Shiga toxin-producing E. coli (STEC); enterotoxigenic E. coli (ETEC); enteroinvasive E. coli (EIEC); enteroaggregative E. coli (EAggEC); and diffusely adherent E. coli (DAEC). E. coli that has no diarrheagenic characteristics other than the EAggEC heat-stable toxin 1 (EAST1) gene is defined as EAST1EC, and is included in the possible seventh group of DEC in this investigation. Although EAST1 is reported to be an enterotoxin of EAggEC (2-4), it has not been well accepted as a virulence factor (5,6). However, we have focused our attention on EAST1EC because an outbreak due to EAST1EC O166:H15 occurred in Osaka City in 1996 (7,8). EPEC possessing both the bundle-forming pilus gene (bfpA) and intimin gene (eae) for E. coli-attaching and -effacing is a well-recognized pathogen in developing countries as class I EPEC (9) or typical EPEC (10). However, atypical EPEC organisms possessing eae alone have been reported to be more prevalent in both developing and developed countries (11), and animals can be reservoirs of atypical EPEC, in contrast to typical EPEC, in which humans are the sole reservoir (12). To accumulate precise information on the prevalence of DEC, we previously reported isolation rates of DEC among
MATERIALS AND METHODS Specimens: Between August 2006 and November 2007, a total of 2,230 stool specimens from healthy adult individuals, including food handlers, were examined. Bacterial examination: Fecal specimens were examined for the presence of E. coli, Salmonella and Shigella with desoxycholate hydrogen sulfide lactose agar (Nissui Pharmaceutical, Tokyo, Japan). More than two suspicious colonies were tested in TSI (Nissui) and LIM media (Nissui), and were identified based on biochemical tests. E. coli and Salmonella were serotyped with antisera (Denka Seiken Co. Ltd., Tokyo, Japan). Bacterial strains: DEC strains isolated from diarrheal patients throughout our previous study (13) were used to
*Corresponding author: Mailing address: Graduate School of Human Life Science, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan. Tel & Fax: +81-6-66052883, E-mail:
[email protected] 318
Pulsed-field gel electrophoresis (PFGE): PFGE was performed by the method of Pei et al. (26). Statistics: Differences between isolation rates were analyzed by performing chi-square tests with Yates’ continuity correction or Fisher’s exact probability test. The KruskalWallis test was used to analyze seasonal variations of the isolation rates. The chi-square statistic for an M × N contingency table was used to compare the overall distribution of intimin subtypes between strains isolated from healthy carriers and those from patients.
compare subtypes of the eae of EPEC and IL-8 induction due to DAEC. Detection of enterovirulent genes and production of enterotoxin: Our newly developed real-time PCR was used to examine the presence of 10 enterovirulence genes (eae, stx1, stx2, elt, est for STh, est for STp, virB, aggR, astA and afaB) in each isolate (22). Briefly, bacterial cells were recovered from a 0.5-ml pure culture of each strain grown for 18 h at 37°C, and genomic DNA was extracted with a Genomic DNA Purification kit (Gentra Systems, Minneapolis, Minn., USA) according to the manufacturer’s protocols. Duplex or triplex primer-probe reactions were performed in QuantiTect Multiplex PCR solution (QIAGEN, Hilden, Germany). Reaction mixtures (48 μl) were dispensed into 96-well, thin-wall PCR plates (Applied Biosystems, Foster City, Calif., USA), and 2 μl of genomic DNA solution was added as a template. Plates were covered with optically clear sealing film and briefly centrifuged. PCR was performed with the ABI PRISM 7000 Sequence Detection System under the following cycle conditions: denaturation at 95°C for 15 min, followed by 40 cycles of 95°C for 1 min and 60°C for 1 min. Shiga toxin (Stx) production was examined using Duopath Verotoxins (Merck Ltd., Tokyo, Japan). Subtyping of enterovirulent genes: According to the report of Blanco et al. (23), the eae genotype was identified using 26 types of intimin type-specific PCR primers complementary to the heterogeneous 3´-end of the genes. DAEC strains were assigned to subgroups according to the scheme described by Le Bouguénec et al. (24). Strains that amplified the primer set for afaC were identified as afapositive, and these strains were assigned to Afa/Dr+ and Afa/ Dr– groups on the basis of positive or negative reactions to primers for afaBC. The Afa/Dr– group also belongs to the Afa group, but the Dr antigen was not recognized. Organisms that were Afa/Dr+ but did not react with primers for any afa subtypes (afaE1, 2, 3, 5 and daaE) were classified as afaEX; bacteria possessed genes amplified with primers for Afa, but the gene was not identified as part of a particular subgroup. Tissue culture adhesion tests and IL-8 assay: Bacterial strains that showed positive reactions on PCR for eae, aggR or afaB were further examined by adhesion tests, as described previously (25). Monolayers of HEp-2 cells grown on cover slips (diameter 13 mm) in 24-well plates were prepared in the absence of antibiotics. Two-day-old monolayers of HEp2 cells were used for the tests. Bacterial strains were grown statically overnight at 37°C in 1% buffered peptone water (Oxoid, Basingstoke, England). Before the test, monolayers were washed once with Dulbecco’s PBS. One milliliter of Basal Eagle’s medium containing D-mannose (1% w/v) without antibiotics or sera was added to each well. Overnight bacterial culture (20 μl) was inoculated into each well, and plates were incubated at 37°C for 3 h. Monolayers were washed three times with PBS, and 1 ml of medium was added to each well. After a further 3-h incubation period, monolayers were washed thoroughly three times with PBS, fixed with absolute methanol, and stained with 10% (v/v) Giemsa. For an enzyme-linked immunosorbent assay of IL-8, adhesion tests were performed as described above, but the epithelial cells were not washed at 3 h, in contrast to the ordinary adhesion tests. A sample of 100 μl of the culture medium was taken after 6-h incubation, and a commercial kit “IL-8, Human, Assay” (GE Healthcare UK Ltd., Buckinghamshire, UK) was used according to the manufacturer’s instructions.
RESULTS Prevalence of DEC in healthy individuals: A total of 2,230 specimens from healthy individuals were examined between August 2006 and November 2007. DEC strains were isolated from 109 stool samples (Table 1). EPEC, EAST1EC, and DAEC were prevalent and did not show significant differences in isolation rates from those among diarrheal patients in our previous study. By contrast, no STEC, ETEC, or EIEC strains were isolated from healthy individuals in this study, and there were significantly fewer EAggEC organisms. The cumulative isolation rates of EPEC, DAEC and EAST1EC among healthy individuals in this study were assessed by season. Isolation rates of EPEC showed significant seasonal variation (Table 2; Kruskal-Wallis, P < 0.001). Characteristics of DEC in healthy individuals: Varieties of DEC serogroups were isolated through this surveillance, as shown in Table 3. Possession of eae, which plays an important role in intimate adhesion to intestinal epithelial cells and in producing attaching and effacing lesions, was thought to be the primary criterion for defining isolates as EPEC in this investigation. Consequently, 18 strains for which seroTable 1. Prevalence of each subpopulation of diarrheagenic E. coli1) Healthy adult Total no. of specimens 2,230 (%) Age (Mean ± SD, Median) 38.5 ± 15.7, 34 EPEC STEC ETEC EIEC EAggEC Afa(+)DAEC EAST1EC Subtotal
Diarrheal adult patient2) 126 (%) 37.0 ± 14.7, 32
32 (1.4) 0* 0*** 0 5 (0.2)*** 19 (0.9) 53 (2.4) 109 (4.9)
Salmonella spp.
1 (0.04)***
0 1 (0.8) 4 (3.2) 0 4 (3.2) 2 (1.6) 6 (4.8) 17 (13.5) 10 (7.8)
1)
: P values were determined by chi-square tests with Yates continuity correction. * and *** mean significantly fewer among healthy individuals compared to patients at P < 0.05 and 0.001, respectively. 2) : The data of adult patients were extracted from our previous study (13). Table 2. Cumulative number of strains of each subgroup of diarrheagenic E. coli isolated during this study by season Summer
Autumn
Winter
No. of specimens
Spring
330 (%) 768 (%)
242 (%)
890 (%)
EPEC1) EAST1EC DAEC
1 (0.3) 7 (2.1) 0
5 (2.1) 5 (2.1) 4 (1.7)
5 (0.6) 15 (1.7) 11 (1.2)
1)
319
21 (2.7) 26 (3.4) 4 (0.5)
: Seasonal change in the isolation rates of EPEC was statistically significant (Kruskal-Wallis, P < 0.001).
Table 3. O antigen serotype and number of strains isolated throughout this study1) STEC stx1/stx2 O103
(1)
Subtotal (1)
EPEC eae O15 (1) O27 (1) O55 (1) O74 (3) O103 (1) O124 (1) O127a (1) O128 (1) O145 (1) O166 (1) O167 (1) O168 (1) UT (18)3) (32)
EAggEC aggR/astA
DAEC afaB
EAST1EC astA
O78 (4) O126 (1)
O1 (2) O25 (3) O74 (1) O86a (3) O151 (1) UT (9)2)
(5)
(19)
O1 (2) O8 (2) O15 (2) O18 (5) O20 (1) O25 (5) O28ac (1) O74 (2) O86a (1) O121 (1) O166 (1) O169 (1) UT (29) (53)
Fig. 1. PFGE patterns of enteroaggregative E. coli O78 strains. Lanes M, Marker; 1, 5 and 9, strains SK34; 2, 6 and 10, strain SK37; 3, 7 and 11, strain SK50; 4, 8 and 12, strain SK60. DNA on lanes 1 - 4, 5 8 and 9 - 12 was digested by XbaI, BlnI and SfiI, respectively.
1)
: Numbers in parentheses indicate number of strains. UT; could not be serotyped with commercial antisera. 2) : One strain possessed both afaB and astA. 3) : One strain possessed both eae and astA.
Subtyping (23,27-29) of the eae assigned the strains isolated from carriers and from patients into 14 and 9 groups, respectively (Table 4): strains of patients were all from nonadults (13) since no EPEC was isolated from adult patients. Five EPEC strains did not produce amplicons with the typing primers of Blanco et al. (23) used in this study, and were designated as UT (untypable). No statistically significant difference was recognized in the overall distribution of intimin subtypes between healthy carriers and patients (chi-square M × N method, P = 0.23). However, both subtypes of β1 and γ1 were prevalent (Fisher’s test, P = 0.047 and 0.050, respectively) and were present among 61% of EPEC strains from patients, while among strains isolated from healthy carriers, θ/γ2 was the most prevalent (24%; Fisher’s test, P = 0.050), followed by ξR/β2B (12%), γ1 (12%), δ/ κ/Β2O (9%) and ι1 (9%). Six EPEC strains from healthy carriers showed positive reactions with two sets of typing primers: one O15 strain was γ1- and ι1-positive; one O74 was γ1- and δ/κ/Β2O-positive; one O168 was β1- and νR/ ε2-positive; one O128 and another O-untypable strain were ε1- and η-positive; and one O-untypable strain was θ/γ2and μR/ι2-positive. One O-untypable strain isolated from a patient was β1- and ι-positive. Nineteen strains of DAEC were isolated (19/2,230, 0.9%); three strains each of O86a and O25, two strains of O1, and one strain each of O74 and O151 were assigned into DAEC with another nine O-untypable strains because of their positive reactions with primers for afaB in our real-time PCR system (Table 5). Three strains were not reconfirmed as being afa-positive when afaC primers were used according to the scheme of Le Bouguénec et al. (24). These strains were assigned as putative new afa-positive strains because of their clear diffuse adhesion on HEp-2 cells. Two strains did not show clear diffuse adhesion on HEp-2 cells, despite the positive results in PCR. Another two strains were Afa/Dr–. Afa/ Dr+ DAEC strains isolated from healthy carriers induced less IL-8 secretion (
