Isolation and characterization of provisional serovar Shigella boydii E16553 from diarrhoeal patients in Bangladesh

Shigellosis is one of the major causes of childhood morbidity and mortality in many developing countries including Bangladesh (Kotloff et al., 1999). Shigellosis is caused by four species of Shigella, namely Shigella dysenteriae, Shigella flexneri, Shigella boydii and Shigella sonnei. Each serogroup contains multiple serotypes based on the structure of the O-antigen component (Simmons & Romanowska, 1987).

Strains that give biochemical reactions typical of Shigella species but do not belong to any of the recognized O-serogroups are described as provisional serovars after ratification by the WHO International Collaborating Center for Shigella at CDC, Atlanta. However, their status remains provisional until several laboratories have identified such serovars and assessed their virulence properties and epidemiological importance.

The number of S. boydii serotypes in the Shigella scheme was 15 in 1958. Since then, a number of provisional serovars of S. boydii have been proposed from reference laboratories (Gross et al., 1980, 1982, 1989). The WHO International Collaborating Center for Shigella in its last classification incorporated three of these provisional serovars into the serotyping scheme and designated them S. boydii serovars 16, 17 and 18 (Wathen-Grady et al., 1985). However, other provisional serovars of S. boydii have been reported but as yet have not been assigned an O antigen number in the international scheme (Matsushita et al., 2002).

The provisional serovar concerned in this present study is E16553 of S. boydii. Ten strains of E16553 were originally isolated from stools of patients in Britain, Finland, Iceland, Sweden and Japan who travelled mainly in India and Pakistan (Gross et al., 1982). The strains were also isolated from other geographic locations and designated S. boydii 19 (the designation S. boydii 19 is acknowledged in the 8^th Edition of the Manual of Clinical Microbiology, ASM Press, 2003). Since then we have recovered 17 isolates of this organism from the stools of diarrhoeal patients in Bangladesh and report here their characterization with respect to biochemical, serological and virulence characters. We recommend that their provisional status be changed and that they be accorded serotype status.

Bacterial strains.
Between 1995 and 1999 we isolated 15 Shigella-like organisms that did not agglutinate with antisera to any known serotypes of Shigella species from diarrhoeal stools of patients at the Clinical Microbiology Laboratory of the International Center for Diarrhoeal Disease Research, Bangladesh (ICDDR, B), Dhaka. Diarrhoeal stools were routinely tested for Salmonella, Shigella, Vibrio, Campylobacter and rotavirus by standard methods (WHO, 1987). Two isolates of a Shigella-like organism from 19931995 were previously identified as S. boydii provisional serovar E16553 by the Laboratory of Enteric Pathogens, Central Public Health Laboratory, Colindale, London, UK, during our studies of S. dysenteriae provisional serovars E22383 and E23507 (Ansaruzzaman et al., 1995). All 17 clinical isolates were Gram-negative rods that were non-motile and non-lactose fermenters.

The reference type strains of S. boydii provisional serovars E16553 (NCTC 11462; Gross et al., 1982) and E28938 (Gross et al., 1989) were obtained from Sweden (SBL, Stockholm, Sweden). One strain of provisional serovar E16553 isolated from a diarrhoeal patient who had travelled from Indonesia to Japan was collected by S. Matsushita (Ueda et al., 2001). Escherichia coli serogroups O18ac, O23, O38 and O52 (Gross et al., 1982) were obtained from J. P. Nataro, Centre for Vaccine Development, University of Maryland School of Medicine, MD, USA. Strains were stored in trypticase soy broth with 0.3 % yeast extract (TSYB; GIBCO) at 70 °C in 25 % glycerol.

Biochemical tests.
All 17 clinical isolates and the provisional serovar E16553 type strain were subjected to the biochemical tests listed in Table 1 by conventional methods as described by Ewing (1986), with some modifications as previously described (Ansaruzzaman et al., 1995). Isolates were also tested in API 20E microgalleries (BioMérieux).

Table 1. Biochemical test reactions of all strains representing provisional serovar E16553

Antisera production.
Antisera were produced against the type strains of provisional serovars E16553 and E28938 (Gross et al., 1989) in adult New Zealand white rabbits by injecting heat-killed cells that were treated with alcohol and acetone as described previously (Ansaruzzaman et al., 1993; Ewing 1986). Antisera to other provisional serovars of S. dysenteriae (E22383 and E23507) were previously prepared by Ansaruzzaman et al. (1995).

Serological tests.
Isolates were subcultured on MacConkey agar (Difco) for 18 h at 37 °C and serotyping was performed by slide agglutination tests as described previously (Talukder et al., 2001). Isolates were tested initially with two commercially available antiserum kits (Denka Saiken) specific for all serovars and with monoclonal antibody reagents specific for S. flexneri and S. dysenteriae type 1 antigens (Reagensia AB). Isolates were checked with antisera to S. boydii provisional serovars E16553 and E28938 and S. dysenteriae provisional serovars E670/74, E22383 and E23507. Slide agglutination tests with an antiserum to E16553 were also performed with live and boiled cells of E. coli O- serogroups O18, O23, O38 and O52. Tube agglutination and agglutinin-absorption tests were carried out as described previously (Albert et al., 1995).

Antimicrobial susceptibility.
Susceptibility to antimicrobial agents was determined by the disk diffusion method as recommended by the National Committee for Clinical Laboratory Standards (NCCLS, 2000) with commercial antimicrobial discs (Oxoid). The antibiotics were tetracycline (30 µg), ampicillin (10 µg), sulfamethoxazole-trimethoprim (25 µg), nalidixic acid (30 µg), ciprofloxacin (5 µg) and mecillinam (25 µg). E. coli ATCC 25922 was used as a control.

Keratoconjunctivitis assay (Sereny test).
This test was performed according to the procedure described by Sereny (1957).

Isolation of plasmid DNA.
Plasmid DNA was prepared by the rapid alkaline lysis method of Kado & Liu (1981) with some modifications (Talukder et al., 2002) and separated by horizontal electrophoresis in 0.7 % agarose gels (Sigma) in a Tris-borate EDTA (TBE) buffer (Bio-Rad) at room temperature at 100 V (50 mA) for 3 h. After electrophoresis, the gel was stained with ethidium bromide and video images were obtained using a gel documentation system. The molecular mass of the unknown plasmid DNA was determined by comparison with plasmids of known molecular mass in E. coli PDK- 9 (Haider et al., 1989).

PCR for virulence genes.
All isolates and type strains were examined for the presence of the Shiga toxin genes (stx1 and stx2) following the procedure of Lin et al. (1993). Shigella enterotoxin genes, set1 (ShET-1) and sen (SHET-2) and ipaH genes were confirmed by PCR according to Vargas et al. (1999). All primers were synthesized in-house using an Oligo 1000 DNA Synthesizer (Beckman).

PFGE.
Chromosomal DNA for PFGE was prepared as described earlier (Talukder et al., 2002). DNA was digested with XbaI (Gibco-BRL) and restriction fragments separated in a CHEF-DRII system in 1 % pulsed-field certified agarose in 0.5 x TBE buffer. The gel was stained, destained and photographed on a gel documentation system (Talukder et al., 2002). The DNA size standards used were a bacteriophage lambda ladder, ranging from 48.5 to 1000 kb (Bio-Rad), and Saccharomyces cerevisiae chromosomal DNA, ranging from 225 to 2200 kb (Bio-Rad). Band patterns were interpreted according to the criteria of Tenover et al. (1995).

Invasive diarrhoea due to Shigella species remains an important public health problem in developing countries. Serological classification is insufficient for the identification of new Shigella species and a polyphasic approach utilizing biochemical, serological and molecular characterization is necessary to identify and confirm new Shigella species. Serological identification of Shigella with specific antisera, cross-absorbed with E. coli, is mandatory owing to their close relatedness to one other.

Two Shigella-like organisms isolated during 19931995 were previously identified by the reference laboratory in London as strains of S. boydii provisional serovar E16553. An antiserum raised against the type strain of provisional serovar E16553 reacted to the homologous titre (2560) with each of the 17 isolates examined here. All of the isolates produced strong agglutination in slide tests with antiserum to E16553. Four strains of E. coli serogroups O18, O23, O38 and O52 reacted weakly with this antiserum but no cross-reaction was observed with any of the established or other provisional serovars of Shigella. Agglutinin-absorption tests confirmed minor antigenic cross-reactivity with the different O-serogroups of E. coli. All isolates, including the E16553 type strain, gave identical biochemical reactions (Table 1), supporting their phenotypic standing as S. boydii.

Production of keratoconjunctivitis in the guinea pig eye within 48 h was a uniform feature and all isolates were positive for ipaH gene and Shigella enterotoxin 2 gene (sen) but negative for Shiga toxin (stx1 and stx2) and Shigella enterotoxin 1 (set1) genes (Table 2). The isolates were all susceptible to ampicillin, sulfamethoxazole-trimethoprim, nalidixic acid, ciprofloxacin and mecillinam but eight exhibited resistance to tetracycline. Multiple plasmids of different molecular sizes were evident, of which the most common were 140, 3.4, 2.7 and 1.4 MDa. Two Bangladeshi isolates contained additional 1.2 and 1.6 MDa plasmids, and one of these isolates harboured a further plasmid of approximately 62 MDa. Antibiotic resistance did not correlate with the presence of any particular plasmid. Although pathogenicity tests are not used as criteria for the classification of members of the Enterobacteriaceae, the invasiveness of the isolates in the Sereny test, the presence of a 140 MDa plasmid (Sansonetti et al., 1982), and Shigella enterotoxin 2 (sen) gene and ipaH gene associated with invasiveness of the strains provide additional evidence that these isolates are representative of Shigella.

Table 2. Phenotypic, genotypic and invasive properties of isolates of provisional serovar E16553 +, Positive; , negative; ND, not done.

The clonal variation within the collection of isolates was explored by PFGE. XbaI-digested chromosomal DNA of the isolates yielded a total of 1516 reproducible fragments and two different profiles, A and B, were distinguished. Type A, which was further subdivided into subtypes A1 and A2, was found among the strains isolated in Bangladesh and Japan, and type B was restricted to the type strain (Fig. 1). This relationship may be explained by the fact that the Bangladeshi and Japanese isolates emerged recently compared to the type strain, which was isolated during 19791980.

(102K): Fig. 1. PFGE pattern of XbaI-digested chromosomal DNA from representative strains of E16553 along with a Japanese strain and the type strain. Lanes: A, Bacteriophage λ ladder (marker); B, type strain; C, Japanese strain; DK, Bangladeshi strains.

Recently, reports of the isolation of the provisional serovar E16553 have emerged from Japan and India (Matsushita et al., 1994; Dutta et al., 2003; Ueda et al., 2001), with increasing numbers of infections. Given its isolation in different geographical locations from several patients with diarrhoea and its pathogenicity supported by in vitro tests, we propose that this serovar be accorded full serotype status as S. boydii 19. This will serve to encourage other laboratories to look for these strains and help define their role as diarrhoeagenic agents worldwide.

This research was funded by the ICDDR, B, Centre for Health and Population Research and by other core donors of the Centre who share its concern for the health problems of developing countries. Current donors providing unrestricted support include the aid agencies of the governments of Australia, Bangladesh, Belgium, Canada, Japan, The Netherlands, Saudi Arabia, Sri Lanka, Sweden, Switzerland and the USA. We thank B. Rowe, Laboratory of Enteric Pathogens, Central Public Health Laboratory, Colindale, London, UK, for serotyping untypable Shigella isolates as provisional serovar E16553.