Heat-stable enterotoxin-producing Escherichia coli O169:H41 in Japan.

To the Editor: Enterotoxigenic Escherichia coli (ETEC) cause diarrhea by producing either a heatlabile enterotoxin, a heat-stable enterotoxin (ST), or both. Consequently, ETEC can be identified either by detecting the enterotoxin in the culture fluid by immunologic assays or by detecting enterotoxin-coding genes with DNA probes or polymerase chain reaction (PCR) amplification. For many clinical laboratories, however, serologic typing is the most common test used to determine if isolates are members of known pathogenic groups. Although E. coli serotype O169:H8 has been recognized as one of the ETEC strains (1), serotype O169:H41 is not established worldwide as one of the diarrheagenic E. coli of the ETEC group. Ando et al. first reported that an outbreak at a school for physiologically handicapped children in Saitama Prefecture was due to ST-producing E. coli serotype O169:H41 (2). We report an outbreak of diarrhea caused by E. coli O169:H41 that predates the outbreak reported by Ando et al. and information about additional outbreaks in Japan since 1991. In June 1991, we isolated toxigenic E. coli from the stool of two of three ill members of an eight-member family during an outbreak of diarrheal illness in Osaka, Japan. An epidemiologic investigation implicated pickles (kimchi) purchased during a visit to Korea; only the three members of the family who ate the pickles became ill. The major symptoms were diarrhea (3/3), abdominal pain (2/3), and fever (2/3) of 38° C. The incubation period was estimated at 33 hours. The serotype of these ST-producing E. coli isolates was not recognized immediately because the cultures were non-typable by the lot of E. coli antiserum available when the cultures were first isolated. The cultures were identified as E. coli O169:H41 when a new lot of antiserum became available. In another outbreak investigated by the Osaka City Department of Environment and Health and the Osaka Prefecture Department of Environment and Health, food poisoning occurred among 776 of 1,242 guests of wedding receptions held at a wedding facility during September 1993. The main symptoms were diarrhea (98%) and abdominal cramps (74%), and the mean incubation period was 40.5 hours. E. coli O169:H41 was isolated from the stool specimens of 7 of 14 patients. A strain of E. coli O169:H41 was isolated from frozen, ready-to-eat seafood recovered from a distributor who provided foods to the wedding facility. In addition to the outbreaks mentioned above, Japanese surveillance reports describe foodborne outbreaks in different prefectures between January 1991 and September 1994. In addition to being cultured for E. coli, stools were also routinely cultured for Shigella, Salmonella (including typhi and paratyphi), Vibrio, Clostridium, Aeromonas, Plesiomonas, Bacillus cereus, and Staphylococcus aureus. Stools were also examined for rotaviruses and small round viruses by electron microscopy. E. coli serotype O169:H41 was isolated from patients’ stools in 11 of 40 outbreaks; recovery rates were 10%-100%. In 7 of the 11 outbreaks, recovery rates of serotype O169:H41 exceeded 75%. PCR was used to examine the diarrheagenicity of 31 E. coli isolates selected from reported outbreaks that occurred from 1991 to 1994 (3). ST production of the 31 isolates was also examined by COLI ST EIA (Denka Seiken Co., Ltd., Tokyo, Japan), a competitive enzyme-linked immunosorbent assay (ELISA) for toxigenic and invasive strains of E. coli. Strains were grown in Casamino Acids-Yeast Extract broth shaken at 37° C for 18 hours. The supernatant obtained after the centrifugation of cells was used for the test according to the manufacturer’s instructions. Thirty of the 31 E. coli O169:H41 isolates tested demonstrated toxigenicity by both PCR and ELISA. Collaborative studies are in progress to further characterize these isolates and to study the relationships between different isolates by molecular epidemiologic methods. Five cultures of E. coli O169:H41 have been ribotyped by a digoxigenin-labeled (Genius System, Boehringer Mannheim) probe prepared from pKK3535 according to the manufacturer’s instructions. The resulting patterns were indistinguishable when the restriction enzymes EcoRI, SmaI, BglII, BamHI, SalI, PstI, or HindIII were used to digest chromosomal DNA. We suggest that this comparatively new serotype of ETEC may be spreading across Japan and urge that studies be conducted to determine its distribution and association with gastroenteritis worldwide.