A Unique Strain of Escherichia coli O157:H7 That Produces Low Verocytotoxin Levels Not Detected by Use of a Commercial Enzyme Immunoassay Kit

In June 1996, a food-borne outbreak due to Escherichia coli O157:H7 swept through Japan affecting more than 9,000 people [1]. This outbreak showed no sign of subsiding until late August 1996, and the government encountered major difficulties in controlling the outbreak. Because of the paucity of epidemiological information concerning this pathogen in Hong Kong, small outbreaks and sporadic cases of infection due to E. coli O157:H7 may go undiagnosed and unrecognized. To monitor the role of this pathogen in human disease in Hong Kong, including hemorrhagic colitis, uncomplicated diarrhea, and hemolytic uremic syndrome, screening for E. coli O157:H7 by use of MacConkey sorbitol agar was initiated in major hospitals in late 1996 [2]. We describe the first reported cases of infection due to E. coli O157:H7 in Hong Kong. In July and October 1997, sorbitol-negative E. coli colonies were isolated from the fecal samples obtained from two patients Figure 1. Pulsed-field gel electrophorewith diarrhea at Queen Elizabeth Hospital and Tuen Muen Hospisis (PFGE) of Escherichia coli O157:H7 tal. Serological evaluation of isolated colonies by use of latex strains with use of restriction XbaI endonuclease. Lane 1, control strain of E. coli agglutination assay (Remel, USA) confirmed E. coli O157:H7. O157:H7 from Public Health Laboratory However, strains from both patients were found to be nontoxigenic Services (PHLS) (Colindale, UK). Lane 2, by use of the Premier EIA for verotoxins I and II (Meridian DiagE. coli O157:H7 strain from Queen Elizanostics, USA). Cytotoxigenicity on vero cells was demonstrated beth Hospital. Lane 3, E. coli O157:H7 only when bacterial cells were treated with polymyxin B [3], and strain from Tuen Muen Hospital. neutralization of cytotoxicity was also observed with use of verotoxin neutralizing antibodies (Meridian Diagnostics). Genes encoding verotoxin II were detected by use of a PCR assay, and subsequent DNA hybridization also identified both strains as verotoxin II producers [4]. In addition, both strains contained the eae gene, which is associated with the attaching-effacing ability evident on a fluorescent actin staining (FAS) assay and DNA hybridization using specific probe from recombinant plasmid pCVD434 as described previously [5]. Pulsed-field gel electrophoresis (PFGE) showed identical patterns for both strains when digested with restriction endonuclease XbaI (figure 1). To our knowledge, we have described the first reported cases of infection due to E. coli O157:H7 in Hong Kong. Cytotoxicity is a major virulence factor of the pathogen, yet the commercial EIA kit failed to identify the verotoxins in the bacterial isolates from both patients. These false-negative results are probably attributable to the low level of extracellular toxins that are secreted by these two strains. Molecular diagnostic evaluation was a more effective approach for the identification of the pathogen in these two cases. Both strains possess the essential pathogenic attributes of E. coli O157:H7 that cause hemorrhagic colitis and hemolytic spread of a unique strain of E. coli O157:H7 in our community, uremic syndrome. There was no epidemiological link between the although no epidemics have been recorded. two patients; the identical PFGE patterns may indicate the clonal W. C. Yam, D. N. C. Tsang, T. L. Que, M. Peiris, W. H. Seto, and K. Y. Yuen Department of Microbiology, Queen Mary Hospital, The University of Hong Kong, Department of Pathology, Queen Elizabeth Hospital, and Financial support: The Research Grants Council of The University of Hong Department of Pathology, Tuen Muen Hospital, Hong Kong Kong. Reprints or correspondence: Dr. W. C. Yam, Department of Microbiology, References The University of Hong Kong, Pokfulam, Hong Kong. 1. Watanabe H, Wada A, Inagaki Y, Itoh K, Tamura K. Outbreaks of enteroClinical Infectious Diseases 1998;27:905–6 haemorrhagic Escherichia coli O157:H7 infection by two different genoq 1998 by the Infectious Diseases Society of America. All rights reserved. 1058–4838/98/2704–0046$03.00 type strains in Japan. Lancet 1996;348:831–2.