Altering Trends in the Dominance of Shigella flexneri Serotypes and Emergence of Serologically Atypical S. flexneri Strains in Dhaka, Bangladesh

Abstract

Of 469 recently isolated Shigella flexneri strains, 452 agglutinated with Shigella flexneri-specific monoclonal antibodies. Of these, 396 could be assigned to 10 of the currently recognized 15 serotypes, with S. flexneri 2b dominating (23.2%). Of the 56 untypeable strains which showed invasive properties, 17 were serologically atypical and the remaining 39 belonged to a new serotype.

Shigellosis is one of the major diarrheal diseases in Bangladesh and several other developing countries and is caused by any one of the four species or groups of Shigella, namely, S. dysenteriae, S. flexneri, S. boydii, and S. sonnei. Each serogroup contains multiple serotypes, and at least 47 serotypes are currently recognized based on their biochemical and/or lipopolysaccharide characteristics. Based on the structure of the O antigen, a component of the lipopolysaccharide present on the outer membrane of the cell (16), 15 serotypes of S. flexneri are currently recognized. Since S. flexneri serotype 2a was found to be the major endemic serotype in developing countries, the currently available candidate vaccine is directed against S. flexneri 2a (12). It is important to determine the prevalence of various serotypes of S. flexneri in different communities worldwide and to monitor the changes over time.

From 1 January 1997 to 30 June 2000, 469 clinical isolates of S. flexneri were isolated from patients attending the Dhaka treatment center operated by the International Centre for Diarrhoeal Diseases Research in Bangladesh. Shigella strains were isolated and identified in the Clinical Microbiology Laboratory following standard microbiological and biochemical methods (17). S. flexneri strains were serotyped using (i) a commercial antiserum kit (Denka Saiken, Tokyo, Japan) and (ii) monoclonal antibody reagents specific for all S. flexneri type and group factor antigens (2, 3). Serological reactions were performed by the glass slide agglutination test as described previously (6).

Plasmid DNA was prepared by the rapid alkaline lysis method of Kado and Liu (8) and analyzed as described previously (7). The Sereny test (9, 15) and determination of the Congo red binding ability of S. flexneri (the results of both indicate invasive property) were performed using procedures described elsewhere (13, 14).

During the study period, 469 strains were identified as S. flexneri based on agglutination with species-specific polyclonal rabbit antisera (Denka Saiken) and general biochemical tests. These strains were further serotyped using the commercially available rabbit antisera specific for all type and group factor antigens. Of the 469 strains, 359 could be serotyped with the commercial kit while 110 (23.4%) were untypeable. All of the 469 strains were then examined using a panel of monoclonal antibodies against S. flexneri (MASF) using the typing scheme shown in Table 1. All but 17 of the 469 strains agglutinated strongly with MASF B (4), an S. flexneri species-specific monoclonal antibody (Table 1). When retrospectively examined, the 17 strains that did not agglutinate with MASF B were found to agglutinate with both S. flexneri and S. boydii species-specific commercial polyclonal rabbit antisera.

TABLE 1.

Agglutination of S. flexneri isolates isolated from January 1997 to June 2000 with MASF antibodies

Serotype	No. of isolates (%) agglutinating with antibody
	Type antigen specificity					B	Group antigen specificity			Provisional serotype
	I	II	IV-2	V	VI		Y-5	6	7.8	IV-1	1c
1a	2					2 (0.4)	1
1b	69					69 (15.3)		69
1c						37 (8.2)					37
2a		70				70 (15.5)	70
2b		105				105 (23.2)			105
3a						70 (15.5)		70	70
3b						0
3c						0
4			17			17 (3.8)				17
4a						0
4b						0
5a				1		1 (0.2)	1
5b						0
6					22	22 (4.9)				6
X						1 (0.2)			1
Y						19 (4.2)	19			9
4X						39 (8.6)				39
Total						452 (100)

Of the 452 strains that agglutinated with MASF B, 396 could be assigned to 10 of the currently recognized 15 serotypes of S. flexneri. The predominant S. flexneri serotypes determined using MASF were 2b (23.2%), 2a (15.5%), 3a (15.5%), 1b (15.3%), and 1c (8.2%), which together accounted for 77.7% of the S. flexneri strains examined in this study (Table 1). Of the remaining 56 strains of S. flexneri, which could not be definitively serotyped, 17 agglutinated with type antigen factor IV-2 but did not agglutinate with group factor Y-5 or 6, thereby excluding their characterization into serotype 4a or 4b. These 17 strains instead agglutinated with a new antigenic determinant, E1037 (IV-1) (Table 1). The remaining 39 of the untypeable strains agglutinated only with the provisional antigen, MASF IV-1, specific for a new antigenic determinant (E1037), but did not react with any type or group antigen-specific antisera tested.

Plasmid profile analysis of the 56 untypeable strains of S. flexneri along with the 37 S. flexneri 1c strains showed that all the untypeable strains and 34 of the 1c strains harbored the 140-MDa invasive plasmid (Table 2). All of the 56 untypeable S. flexneri strains had the ability to bind Congo red (Table 2), reflecting the invasive property of these strains. Twelve strains of the 56 containing the 140-MDa plasmid and 3 strains of S. flexneri 1c which did not carry the 140-MDa plasmid were selected at random and subjected to the Sereny test. The strains containing the 140-MDa plasmid were positive by the Sereny test, while the strains without the 140-MDa plasmid were negative (Table 2).

TABLE 2.

Phenotypic, genotypic, and invasive properties of serotype 1c and serologically atypical isolates of S. flexneri

S. flexneri serotype	No. of isolates tested	ID by API-20E testa	No. Congo red positive (%)	No. with 140-MDa plasmid (%)	Sereny test results
					No. of isolates tested	No. positive	No. negative
1c	37	Shigella spp.	34 (92)	34 (92)	5	2	3
Type 4	17	Shigella spp.	17 (100)	17 (100)	5	5	0
4X	39	Shigella spp.	39 (100)	39 (100)	5	5	0

^aID, identification. 

The major observation in this study was that serotype 2b has now become the dominant serotype of S. flexneri prevalent in Dhaka. From a vaccine perspective, the shift in S. flexneri serotype dominance from 2a to 2b is not such a disconcerting event given the fact that S. flexneri serotypes, with the exception of serotype 6, have some degree of antigenic relatedness attributable to a common repeating tetrasaccharide unit (5). All previous studies on serotypes of S. flexneri in Bangladesh have shown 2a as the dominant serotype (1, 7, 10, 11), but this has now changed. In this study, E1037 was found in all of the type 4 strains and in a number of strains (designated as 4X) which reacted only with S. flexneri-specific MASF B (4). These appear to constitute a new serotype of S. flexneri with uncharacterized antigenic determinants. Serologically atypical strains displaying conflicting agglutination patterns have recently been reported among strains of S. flexneri isolated in rural Egypt, and this was attributed to limitations in the ability of available commercial antibody reagents to reliably detect the full diversity of serological variants of S. flexneri (6). We also observed that the commercially available antiserum kit could not type 23.4% of the S. flexneri strains, and 17 strains showed cross-reactions with S. boydii antisera. Another interesting trend was that the incidence of subserotype 1c is increasing, and the previously identified S. flexneri serotypes 3b, 3c, 4a, and 5b were not detected in this study while such serotypes as 1a, 5a, and X were disappearing. Recently El-Gendy et al. (6) reported that the antigenic determinant E1037 and MASF type antigen factor IV-2 are present in some strains of S. flexneri 1c isolated in rural Egypt. Overall, it appears that there is a changing profile in regard to the prevalence of various serotypes of S. flexneri.

The untypeable strains (those newly designated serotype 1c and the serologically atypically strains provisionally designated types 4 and 4X by use of monoclonal antibodies) were studied in order to determine whether these strains were invasive. With the exception of three, all the strains were found to harbor a large plasmid of approximately 140 MDa and had the ability to bind Congo red and produce keratoconjunctivitis in guinea pigs, attesting to their invasive properties. Three strains of serotype 1c did not harbor this plasmid and were avirulent.