Correlation between PFGE Groups and mrp/epf/sly Genotypes of Human Streptococcus suis Serotype 2 in Northern Thailand

Streptococcus suis infection is a severe zoonotic disease commonly found in Northern Thailand where people often consume raw pork and/or pig's blood. The most frequent clinical presentations are meningitis, sepsis, and endocarditis with higher rate of mortality and hearing loss sequelae. To clarify the correlation between pulsed-field gel electrophoresis (PFGE) groups and mrp/epf/sly genotypes of S. suis serotype 2, 62 patient and 4 healthy pig isolates from Northern Thailand were studied. By PFGE analysis, at 66% homology, most human isolates (69.4%) and 1 pig isolate were in group A, whereas 14.5% of human isolates and 3 out of 4 pig isolates were in group D. According to mrp/epf/sly genotypes, 80.6% of human isolates were identified in mrp + epf − sly − and only 12.9% were in mrp − epf − sly + genotypes; in contrast, 1 and 3 pig isolates were detected in these two genotypes, respectively. Interestingly, all isolates of S. suis serotype 2 classified in PFGE groups A, B, and E were set in mrp + epf − sly − genotypes. These data show a close correlation between PFGE groups and mrp/epf/sly genotypes of human S. suis serotype 2.


Introduction
Streptococcus suis, recognized as a significant swine and human pathogen, mainly causes meningitis, sepsis, endocarditis, and septic shock [1]. It can be transmitted to humans by contact with sick or carrier pigs, pig-derived products [2], or eating undercooked pork [3,4]. Capsular polysaccharide (CPS) is the most important proven critical virulence factor, due to its antiphagocytosis activity [5]. Of 35 serotypes, serotype 2 is the most frequently isolated and associated with disease in both animals and humans [6]. In addition, virulence-related proteins, such as muramidase-released protein (MRP), extracellular factor (EF), and hemolysin (suilysin, SLY), are expressed by some strains of S. sui "as discussed by Gottschalk and Segura [1]. " These proteins are encoded by the genes mrp, epf, and sly, respectively. MRP/EF/SLY phenotypes or mrp/epf/sly genotypes of S. suis serotype 2 have been studied mostly in pig isolates with very little data for human isolates [1,7], especially in Northern Thailand. Pulsed-field gel electrophoresis (PFGE) of DNA restricted with SmaI has been confirmed to be valuable for evaluating the genetic diversity of S. suis [8]. The present study aims to clarify the correlation between PFGE and mrp/epf/sly genotypes of S. suis serotype 2 isolated from patients in Northern Thailand.

S. suis Serotype 2 Strains from Humans and Healthy Pigs.
A total of 66 S. suis serotype 2 (SS2) isolates were included in this study. Sixty-two human SS2 were isolated from 43 and 19 patients at Maharaj Nakorn Chiang Mai Hospital and Lamphun Provincial Hospital, respectively. To isolate S. suis from pigs, a total of 150 submaxillary lymph nodes of healthy pigs that were bought from two retail markets in the municipal area of Chiang Mai province between November and December 2002 were studied. From each homogenized sample, S. suis was isolated on NNCC agar (sodium azide, nalidixic acid, colistin-crystal violet agar) "as described by Kataoka et al. [9]" and selective blood agar (sheep blood agar + SR126E, oxoid). All -hemolytic grampositive cocci were screened first with standard biochemical tests and confirmed with an API 20 Strep (bioMérieux). S. suis serotype 2 was identified by the coagglutination test with antiserotype 1 (negative reaction) and 2 (positive reaction) antibodies "as discussed by Gottschalk et al. [6]. " P1/7 strain, a reference strain of serotype 2, which was isolated from a diseased pig with the mrp + epf + sly + genotype, was used as a positive control. For negative control, we used beta-hemolytic streptococci and distilled water.

Characterization of the Genetic Diversity of S. suis Isolates by PFGE.
This method was modified from Son et al. [10]. For each S. suis serotype 2 strain, two independent extractions of DNA were performed to verify the reproducibility of patterns. Briefly, the cell suspension was prepared from overnight Todd-Hewitt Broth (Difco) culture and mixed with 1% agarose to form the agarose plugs. The cells were lysed with lysozyme and then treated with proteinase. The DNA was digested with SmaI and its fragments were resolved by PFGE within 1% electrophoresis grade agarose gel using a CHEF-DR II system (Bio-Rad). The gels were stained with ethidium bromide and photographed under a UV light. The cluster analysis was provided by Bionumeric software (Applied Maths) and transformed into an agglomerative cluster using the unweighted pair group method with arithmetic averages (UPGA) (based on Dice coefficients).
Amplification reaction for each gene was individually performed in a Perkin Elmer Cetus DNA thermal cycler with the reaction mixtures including the four deoxynucleotide triphosphates, each primer, and Taq DNA polymerase. The PCR products were run in a 2% agarose gel (mrp, epf, and epf * ) or a 0.8% agarose gel (sly), stained with ethidium bromide, and visualized by UV transillumination. Lambda DNA digested with EcoRI and HindIII and 100 bp DNA ladder markers (all from Promega Co.) were used to estimate fragment size.

Isolation of S. suis from Submaxillary Lymph Nodes of Healthy Pigs.
Of 150 submaxillary lymph node samples, 68 S. suis isolates (45.3%) were recovered. However, only 6 isolates (4%), detected from 4 samples (2.6%; 3 isolates were recovered from the same sample), were confirmed to belong to serotype 2. Hence, the four isolates recovered from 4 animals were included in this study.

Characterization of the Genetic Diversity of S. suis Isolates by PFGE.
Reproducible results were observed because an identical pattern was shown for each S. suis isolate after the two independent DNA extractions (results not shown). PFGE patterns after digestion with SmaI were characterized by 6 to 10 major bands with sizes ranging from 48.5 to 436 kb. Of the 66 isolates, cluster analysis by Bionumerics software revealed 5 PFGE groups (A to E), at 66% similarity ( Figure 1 and Table 1). Most patient isolates (43 of 62 or 69.4%) were expressed in PFGE group A, followed by PFGE group D (9 of 62 or 14.5%). The 4 pig isolates presented 2 different PFGE groups (1 in A and 3 in D) and the reference strain P1/7 belonged to pulsotype l group C.

Detection of mrp, epf, and sly Putative Virulent Genes by PCR.
When the three genes were composed as mrp/epf /sly genotype, 3 different types were analyzed ( Table 1). Most of the S. suis serotype 2 isolates from humans (50 of 62 or 80.6%) and only 1 pig isolate presented the mrp + epf − sly − genotype whereas 3 of the 4 pig isolates presented the mrp − epf − sly + genotype, which was also found in 8 human isolates. Four human isolates with the mrp + epf + sly + genotype corresponded to the genotype of a P1/7 reference strain.

Correlation between PFGE Groups and mrp/epf/sly Genotypes.
The association between PFGE groups and mrp/epf/sly genotypes shows that all isolates of SS2 in PFGE group A, B, and E were classified in the mrp + epf − sly − genotype. Most of the PFGE group D strains belong to the mrp − epf − sly + genotype, except for 1. All isolates of PFGE group C belong to mrp + epf + sly + or mrp + epf * sly + genotypes (Table 1).

Discussion
PFGE is a high-resolution genotypic technique originally proved by Allgaier et al. [14] to be most suitable for differentiating single isolates, thus making it the most reliable in epidemiological studies with S. suis. Berthelot-Hérault et al. [8] confirmed the PFGE analysis to be a very useful tool for such investigations. They reported on the SmaI restriction PFGE patterns (A, B, and C) of 123 S. suis isolates from pigs in France and from humans worldwide. In the present study, while most of S. suis serotype 2 isolates from humans were identified as PFGE groups A and D (69.4% and 14.5%, resp.),  Figure 1: Dendrogram showing the relationship between PFGE groups, mrp-epf-sly genotypes of S. suis serotype 2 isolates, and P1/7 reference strain.  Journal of Pathogens  Toxic shock syndrome  Total  A  2 4  1 3  3  3  4 3  B  2  0  0  0  2  C  0  3  1  0  4  D  4  2  3  0  9  E  0  1  3  0  4  Total  30  19  10  3  62 isolates from healthy pigs were included in group A and D (1 and 3 isolates, resp.). Interestingly, S. suis isolates from humans and healthy pigs were clustered in the same PFGE subgroups ("b" and "f "), except for 1 pig isolate (NL 110.1). This evidence may raise the possibility that healthy carrier pigs could transmit S. suis to humans. The MRP/EF/SLY phenotypes and/or mrp/epf/sly genotypes of S. suis serotype 2 have been used as a tool to determine genetic diversity among human and pig isolates as well as in epidemiological studies. In addition, a clear association between either MRP + , EF + , SLY + or mrp + , epf + , sly + phenotypes/genotypes with virulence has largely been established in many countries. Most S. suis invasive serotypes 2 strains isolated from diseased pigs in Europe were included in the mrp + /mrp s epf + /epf * /sly + genotype [15]. Most serotype 2 strains isolated from ill pigs or human patients in China and Vietnam also present such a mrp + , epf + , sly + phenotype [16,17]. However, Gottschalk et al. [6] first described a high frequency of MRP − EF − SLY − phenotype of most swine strains (72%) and two human isolates. In addition, 16 strains (18.8%) from diseased pigs were MRP + /MRP * EF − SLY − . These phenotypes seem to be common in North America, since similar results were confirmed with strains recovered from diseased pigs in the United States [18].

Journal of Pathogens
However, the situation seems to be different in Thailand. Our results are similar to those presented by Takamatsu et al. [7] who studied the mrp/epf /sly genotypes of 19 human SS2 isolates from Thailand, showing that most of them (68.4%) were assigned to sly − epf − mrp * * + genotypes. Healthy pig isolates from this study showed the same genotypes as the human isolates (either mrp + epf − sly − or mrp − epf − sly + genotypes). Moreover, the 3 isolates from the same pig's lymph node all showed the same PFGE pulsotype b and mrp − epf − sly + genotype. This usually indicates that this animal is probably still recovering from its illness. It is interesting to note that the reference isolate P1/7 showed a high genetic similarity with a human isolate, CSF060544 (PFGE group C pulsotype l and mrp + , epf + , sly + phenotype). This may indicate the possible transmission of diseased pig isolate to human.
Our team, Wongsawan et al. [19] had analyzed the genetic diversity of S. suis isolated from humans by both PFGE and mrp/epf/sly genotypes; however, they have not been compared one each other. In the present study, we wanted to address the correlation between different groups and genotypes of S. suis serotype 2 for future epidemiological studies. Interestingly, we found that all isolates of S. suis serotype 2 classified in PFGE groups A, B, and E were included in the mrp + epf − sly − genotype. In addition, all isolates in PFGE groups C and D except one were expressed in the genotypes mrp + epf + /epf * sly + and mrp − epf − sly + , respectively. Kerdsin et al. [20] reported the significant association between ST1 strain and the meningitis category ( < 0.0001); they did not discuss the correlation between PFGE groups and clinical features. In our disease outcomes (Table 2), we found that most of Bacteremia (24 of 30 or 80.0%) and meningitis (13 of 19 or 68.4%) cases were in PFGE group A. Fascinatingly, all 3 cases of toxic shock syndrome were also included in PFGE group A. For endocarditis, each 3 of 10 (30%) cases were distributed in PFGE groups A, D, and E.
In conclusion, mrp/epf/sly genotypes of S. suis serotype 2 isolated from humans are strongly correlated with PFGE analysis. The few isolates from pigs analyzed in this study were similar to those isolated from patients. It will be necessary to further study a higher number of swine isolates recovered from ill animals in Thailand.