Background
Shigellosis is an acute intestinal infection caused by bacteria of the genus
Shigella. The main symptom of this infection is bloody diarrhoea and the minimum infective dose is as low as 10–100 bacterial cells due to relative resistance to stomach acid [
1]. Shigellosis is a major public health concern worldwide, especially in developing countries [
2]. The infection is most frequent in children, the elderly and the immunocompromised [
2,
3]. A more recent annual estimate of the shigellosis burden was at 90 million incidences and 108,000 deaths [
4].
Shigella can be differentiated into four species or serogroups,
S. dysenteriae
S. flexneri
S. boydii, and
S. sonnei based on biochemical properties and group-specific O antigens in the outer membrane of the cell wall. There is a shift in
Shigella dominance from
S. flexneri to
S. sonnei in developing and developed countries [
4‐
6]. In Malaysia, Lee and Puthucheary [
7] and Banga Singh
et al. [
8] reported an increasing dominance of
S. sonnei as the etiologic agent of shigellosis. Apart from the apparent dominance shift, several studies have described the spread of a well-defined pandemic
S. sonnei clone since 1990s which is characterized by common features such as biotype
g, a particular PFGE
XbaI pulsotype, resistance pattern of streptomycin - trimethoprim-sulfamethoxazole - tetracycline, and the presence of a 2.2 kbp class 2 integron [
9,
10].
Bacterial subtyping is frequently applied for outbreak investigations and surveillance of infectious diseases. Strain-specific fingerprints generated are used to facilitate the identification of disease transmission routes and sources [
11]. Biotyping is one of the earlier phenotypic subtyping methods applied to
S. sonnei. The bacteria can be subdivided into five biotypes (a, d, e, f, and g) on the basis of biochemical properties. This method however is not sufficiently discriminative.
Several genotyping methods with higher discriminatory power such as pulsed-field gel electrophoresis (PFGE) have been applied to subtype
S. sonnei[
12‐
14]. Since PFGE is a gel-based method, it requires strict adherence to standardized protocol for reproducible results. Standardized PulseNet PFGE protocol has been useful for inter-laboratory comparison of results [
15]. However, due to the monomorphic nature of
S. sonnei, PFGE occasionally may not be able to distinguish epidemiologically unrelated
S. sonnei isolates and is not appropriate for phylogenetic analysis of strains that have evolved over a longer time span [
16,
17].
Multilocus variable-number tandem-repeat (VNTR) analysis (MLVA) is a highly discriminative sequence-based subtyping tool and has been used to study the genetic relatedness among
S. sonnei. This method is based on the inherent variability of short sequences that are organized as tandem repeats at multiple VNTR loci. The VNTR loci in
S. sonnei were found to have different degrees of variability [
17]. MLVA with four to eight highly variable VNTR loci exhibited a discriminatory power parallel to or higher than PFGE [
18]. Furthermore, MLVA based on the combination of VNTR loci with different variability has also been successfully used for phylogenetic analysis of
S. sonnei that have evolved over different timescales [
17].
Although S. sonnei is becoming an important etiologic agent of shigellosis in Malaysia, there is limited information on the genetic background of local strains. Therefore, the objective of the study was to characterize local S. sonnei strains by analyzing their biotypes, antimicrobial resistance patterns and genotypes. In addition, the prospect of using MLVA for routine subtyping of local S. sonnei in comparison with PFGE was also evaluated.
Discussion
The predominant biotype identified in this study was biotype
a (n = 29, 73%) and the remaining strains were of biotype
g (n = 11, 27%). These two biotypes were the most commonly reported biotypes in Australia from 1990–2009 [
23]. In many other countries, an increasing prevalence of
S. sonnei biotype
g was reported. Izumiya
et al.[
14] reported the prevalence of
S. sonnei biotype
g among travel associated cases in Japan while almost all biotype
a isolates were from patients with travel history to Southeast Asia. In Korea,
S. sonnei isolated during 1977–1986 were of biotype
a, whereas isolates in 1991–2000 were of biotype
g[
24]. Similarly,
S. sonnei isolated from cases acquired in Ireland, Italy, United States, and a few African countries also suggested the increasing prevalence of
S. sonnei biotype
g since the 1990s [
9,
25,
26]. More Malaysian strains however need to be analyzed to obtain a more accurate representation of the distribution and prevalence of
S. sonnei biotypes in Malaysia.
Overall, PFGE analysis in this study indicated that
S. sonnei biotype
a strains were genetically more diverse than biotype g strains. The non-rhamnose fermenting biotype
g strains were given further attention as there is a spread of a pandemic clone of biotype
g strains across different continents [
9,
10]. Ten out of 11 Malaysian biotype
g strains in this study exhibited similar characteristics of the pandemic clone that is biotype
g, resistance pattern of SSxT, and similar but distinguishable PFGE patterns. A number of these biotype
g strains were highly similar in PFGE banding patterns. The banding patterns, by visual comparison, were similar to some biotype
g strains from Ireland and Italy [
13]. These Malaysian strains could be phylogenetically linked to the pandemic biotype
g clone. The biotype
g strains were isolated from the period 1997 to 2000 and year 2008, indicating that these strains persisted in Malaysia for at least a decade.
In this study, more than half of the strains were resistant towards streptomycin (67.5%) and all strains were susceptible to kanamycin. This is in agreement with an earlier study in Malaysia by Hoe
et al.[
12] on Malaysian strains collected during the period 1997–2000. Resistance to trimethoprim-sulfamethoxazole and tetracycline are commonly reported in
S. sonnei[
5,
14,
24,
27]. However the resistance rates to trimethoprim-sulfamethoxazole and tetracycline in this study were relatively low at 37.5% and 40%, respectively as compared to those observed in Taiwan, Thailand, Japan and Korea [
5,
14,
24,
27]. Most of the strains in this study remained susceptible to ceftriaxone, ciprofloxacin, ampicillin, chloramphenicol and nalidixic acid. These results were consistent with a study on
S. sonnei from Northeast Malaysia [
8]. The low resistance rate to nalidixic acid in Malaysia is in concordance with a report by Izumiya
et al.[
14] where resistance to nalidixic acid was less frequent in
S. sonnei originated from South-east Asia. MDR
S. sonnei strains persisted throughout the study periods. Selective antibiotic pressure may lead to the persistence of MDR
S. sonnei strains in Malaysia [
12] as ampicillin and trimethoprim-sulfamethoxazole are used for the treatment of
Shigella infection locally [
28]. To the best of our knowledge, all the 40 infected patients recovered from their illness.
PFGE and MLVA showed comparable discrimination in subtyping of S. sonnei. Both techniques further differentiated the S. sonnei biotype a and g strains. However MLVA was better at grouping the strains on the basis of biotypes, and the overall percentage of similarity among strains was low when subtyping was done using MLVA. Furthermore, MLVA subtyping was more rapid and less laborious compared to PFGE. Interpretation of result was less subjective and results were more readily comparable between laboratories. All these suggest that MLVA may be a suitable complement to PFGE or even an alternative for routine subtyping of S. sonnei.
Most of the PFGE and MLVA clusters contained strains from multiple geographical locations, some clusters even contained strains isolated from distant parts of the country although they were epidemiologically unrelated
. This concurred with the report of Pichel
et al.[
3] where epidemiologically unrelated Argentinean
S. sonnei from very distant geographical areas were clustered together by PFGE. Epidemiologically unrelated MLVA single-locus variants, and strains isolated seven years apart at different locations yet with highly similar PFGE pattern were also observed. These observations where strains with no apparent epidemiological linkage were clustered together may be due to travel of individuals within Malaysia, long-persisted dissemination of different clones of
S. sonnei throughout the country, person-to-person transmission of a particular strain over an extended period with minor genetic changes, or a combination of these events.
Based on the MST analysis,
S. sonnei in the present study were heterogeneous, and a number of strains in this study were related to
S. sonnei from Taiwan and to a lesser extent, to those from neighbouring countries. Since
S. sonnei is frequently found responsible for travel-associated diarrhoea, it is not uncommon that transmission of
S. sonnei across countries and even continents occurred easily. A study on the global distribution of
S. sonnei clones divided the strains from 50 countries, including Malaysia into three major clonal groups [
29]. The 173 strains from five countries that were used for comparison in this study belonged to clonal groups A (n = 148) and C (n = 17) which were globally spread and clonal group B (n = 8) which was found in Europe, Africa and Asia [
29] (Figure
3). One hundred and fifty one of these strains originated from 10 shigellosis outbreaks (O1 – O10) in Taiwan between 1996 and 2004. Another 22 strains were collected from nine epidemiologically unrelated episodes (E1 – E9) in Taiwan, China, Indonesia, Vietnam and Cambodia between 1998 to 2005 [
16,
18]. Based onVNTR loci SS1 and SS6, the Malaysian strains in the present study also belonged to these three major clonal groups. These observations suggest the
S. sonnei strains circulating in Malaysia belonged to different clones that were spread worldwide. Although our observation may not represent the precise clonal structure of
S. sonnei in Malaysia due to the limited sample size, the present study has nevertheless demonstrated the wide genetic diversity of
S. sonnei circulating in this country.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
XPK carried out the experiments, analysis and interpretation of data and drafted the manuscript. CSC provided the information for VNTR primers and participated in design and interpretation of data and helped in drafting of the manuscript. HW provided partial funding for the project and helped in editing of the manuscript. NA co-supervised the project, and participated in the design of project, and helped in drafting of the manuscript. Norazah A provided some of the strains and helped in editing the manuscript. KLT conceived study, participated in its design, coordinated and supervised the project and co-wrote the manuscript. All authors read and approved the final manuscript.