Introduction

Streptococcus suis, a zoonotic gram-positive coccus bacterial pathogen, causes significant economic losses in the pig industry and invasive infections in humans who are in close contact with infected pigs or contaminated pork-derived products [1, 2]. Among the 29 true serotypes, serotypes 2 and 14 are the most pathogenic in humans [1, 2], and the serotypes most commonly isolated from diseased pigs are serotypes 1/2, 2, 3, 7 and 9 [1].

The identification and characterization of virulence factors or markers in S. suis would be a major advance in understanding the pathogenesis of infection and would aid in epidemiological surveillance. There is still no consensus on universal factors or markers that can clearly differentiate pathogenic from nonpathogenic isolates. Molecular tools to differentiate potentially pathogenic pathotype (disease-associated) strains from nonpathogenic pathotype (non-disease-associated) strains are of utmost importance [3]. Currently, there are several methods that claim to determine or predict pathogenic S. suis strains, including genotyping of epf, mrp, and sly [4, 5]; MLST [6]; whole-genome sequencing [7,8,9]; minimum core genome (MCG) [10]; serotyping PCR [11, 12]; five newly proposed virulence-associated genes (VAGs), including srtF, ofs, RNA-binding protein (SSU_RS09525), and two hypothetical proteins (SSU_RS09155 and SSU_RS03100) [13, 14]; and pathotyping PCR [15]. Some techniques are difficult to apply to the routine testing of a large number of isolates, have high costs and are labour intensive and time-consuming. Some techniques lack the discriminatory power to differentiate S. suis strains into virulent and avirulent subpopulations, thus limiting their usefulness in epidemiological studies.

Wileman et al. (2019) developed multiplex PCR pathotyping to target three genetic marker genes associated with observed clinical phenotypes, including the genes for copper-exporting ATPase 1, a type I restriction-modification system S protein, and a putative sugar ABC transporter [15]. That study used the selected genetic markers to differentiate S. suis into a disease-associated group (pathogenic pathotype/virulent) and a non-disease-associated group (nonpathogenic pathotype/nonvirulent) [15]. Multiplex PCR pathotyping worked well for S. suis strains from England and Wales [15], but contradictory results were observed with S. suis strains from Switzerland [16]. In addition, this multiplex PCR assay was not validated in S. suis strains from geographic regions other than Europe. This study was to evaluate the ability of the multiplex PCR pathotyping approach described by Wileman and colleagues [15] to distinguish pathotypes of S. suis isolates from Thailand. We used this method with S. suis isolates recovered from humans and clinically healthy pigs in Thailand. The current study aimed to determine whether PCR can predict the potential virulence of Thai isolates and provide evidence of the versatility of the multiplex PCR scheme in Thai S. suis strains in contrast to European strains.

Materials and methods

Bacterial strains and DNA extraction

The S. suis strains used in this study were collected from human patients and clinically healthy pigs. Human S. suis strains were randomly selected from a previous study [17] that collected strains between 2009 and 2012 in all parts of Thailand. Healthy pig S. suis strains were collected between April 2010 and March 2011 in northern Thailand [18]. The criteria for selection were based on the distribution of serotypes, sequence types (ST), clonal complexes (CC), area (provinces/regions), isolation years, and sources.

In total, 451 S. suis strains were selected and used in this study, consisting of human isolates (n = 278) and clinically healthy pig isolates (n = 173). Among these, 239 were serotype 2 isolates (human = 226; pig = 13), 47 were serotype 14 isolates (human = 46; pig = 1), 14 were serotype 9 isolates (human = 1; pig = 13), 7 were serotype 4 isolates (human = 1; pig = 6), 6 were serotype 7 isolates (pig only), 4 were serotype 5 isolates (human = 2; pig = 2), 4 were serotype 1 isolates (human = 2; pig = 2), 31 were serotype 16 (pig only), 19 were serotype 3 (pig only), 19 were serotype 31 (pig only), and 61 were other serotype isolates from pigs. Details of these isolates are shown in Tables 1 and 2.

Table 1 Distribution of pathogenic pathotypes in Streptococcus suis isolates from humans
Full size table
Table 2 Distribution of pathogenic pathotypes in Streptococcus suis isolates from pigs
Full size table

All S. suis strains were cultured on sheep blood agar at 37 °C for 24 h. S. suis DNA was extracted using the ZymoBIOMICS DNA Miniprep Kit (ZymoBIOMICS DNA Miniprep Kit; Irvine, CA, USA) following the manufacturer’s instructions. S. suis species and serotypes of the extracted DNA were confirmed by PCR [11].

Pathotyping PCR

The multiplex PCR procedure was slightly modified from the previously described procedure [15]. The reaction mixture (15 µl) contained 1X PCRBIO HS Taq Mix Red master mix (PCRBIO Taq DNA Polymerase; London, UK), 0.7 µM of primers SSU1589_0460F and SSU1589_0806R, 0.3 µM of other primers, and 15 ng of template DNA. The primers for pathotyping (pathogenic pathotype [disease-associated] and nonpathogenic pathotype [non-disease-associated]) and their sequences were described in a previous report [15]. The primers included SSU0207_0735F (5’-TTACAAGAACAGGGCAAGACAGTCGCC-3’), SSU0207_0945R (5’-GCTGCTTTATAATCTGGGTGTTCGTTG-3’), SSU1589_0460F (5’- CCTTTAATGCAGGGGACAAAAGTGAGCTC-3’), SSU1589_0806R (5’- CCCATAATCTTACAGTTAACTTCCTTGC-3’), SSUST30534_0368F (5’- ATCCCCTCCCAATAAAAGATTTGGATGC-3’), SSUST30534_1259R (5’- TTTTCGAGCTCTCCATACACTGCTTCTG-3’), SSU0577_0086F (5’- CAGGTAGTTTGGGCTTAGCTTCATCAGG-3’), and SSU0577_0807R (5’- TGGATGCTGAATTCGCAACTGGGCAATC-3’).

The PCR amplification conditions were described previously [15] and consisted of 95 °C for 5 min; 35 cycles of 95 °C for 30 s (denaturation), 66 °C for 90 s (annealing), and 72 °C for 90 s (extension); and a final extension of 68 °C for 10 min. The PCR products were evaluated in 2% agarose gels at 100 V for 30 min. S. suis strain P1/7 was used as a positive control for the multiplex PCR. A negative control consisting of the same reaction mixture but with water instead of template DNA was included in each run. The gel was stained with ethidium bromide for 20 min. DNA bands were visualized and photographed under ultraviolet light using gel documentation equipment (SynGene; Cambridge, UK). The sizes of the PCR products were determined by comparison with a molecular size standard (GeneRuler 100 bp Plus DNA ladder; Thermo Fisher Scientific, CA, USA).

As shown in Fig. 1, PCR product sizes of both 347 bp and 211 (or 190 bp) or single products of either 347 bp or 211 bp (or 190 bp) were indicative of the disease-associated group (SSU0207: copper exporting ATPase 1 and SSU1589: type I restriction-modification system S protein), while an 892 bp product corresponded with the non-disease-associated group (SSUST30534: putative sugar ABC transporter) [15]. Amplicons of 722 bp corresponded to the S. suis species-specific marker (SSU0577: WhiA sporulation regulator) [15]. Diagnostic accuracy was measured in terms of sensitivity, specificity, positive predictive value, and negative predictive value compared to the origin of isolates from either human patients or clinically healthy pigs. Statistics were analysed by Fisher’s exact test with Stata version 17.0 software (StataCorp, College Station, TX, USA). Data were considered significant at p < 0.01.

Fig. 1
figure 1

Agarose gel electrophoresis of representative Streptococcus suis pathotypes. Lane 1 = disease-associated group I, Lane 2 = disease-associated group II, Lane 3 = disease-associated group III, Lane 4 = disease-associated group IV, Lanes 5 and 6 = undetermined, Lane 7 = non-disease-associated group, Lane 8 = negative control, and Lane M = 100 bp DNA ladder

Full size image

Results and discussion

This is the first study to utilize PCR pathotyping to test S. suis isolates from a region (Thailand) other than Europe (England, Wales, and Switzerland) [15, 16]. Collectively, the current study revealed that the Wileman multiplex PCR pathotyping assay could distinguish all 451 S. suis strains into 4 types of disease-associated groups: type I (211 and 347 bp), type II (190 and 347 bp), type III (211 bp), and type IV (190 bp), as shown in Fig. 1. In the case of a PCR product size of either 190 or 211 bp, 892 bp was considered undetermined in this study (Fig. 1). Types III and IV revealed only a single band of either 211 or 190 bp, with no amplification of 347 bp. Two previous studies have also detected 211 or 190 bp bands without a 347 bp band [15, 16].

The human S. suis isolates (n = 278) contained 226 serotype 2 isolates, 46 serotype 14 isolates, 2 serotype 1 isolates, 2 serotype 5 isolates, and one isolate each of serotypes 4 and 9. Of these, 99.3% (n = 276) belonged to the disease-associated group, including 52.6% type IV (n = 145), which was most commonly found in the serotype 2-ST104 strain (n = 40); 36.2% type I (n = 100), which was most commonly found in the serotype 14-ST105 strain (n = 43); 7.9% type III (n = 22), which was found in the serotype 2-ST104 strain (n = 11); and 3.3% type II (n = 9) (Table 1; Fig. 2). We also detected 0.7% (n = 2) that seemed to be undetermined, as shown in Table 1. None of the human isolates were in the non-disease-associated group. This confirmed the PCR results that these groups of isolates were pathogenic and had potential virulence. Hence, the diagnostic sensitivity was 99.3% (276/278), and the positive predictive value was 100% (276/276), whereas the specificity and negative predictive value could not be calculated because all tested S. suis isolates were from human patients and were considered pathogenic or virulent. A limitation of this study is that almost 98% of human isolates belonged to either serotype 2 or 14, which represents a relatively homogenous population. It is unknown if this test can be successfully used with isolates recovered from diseased pigs and those of different serotypes. Indeed, Scherrer et al. (2020) demonstrated that the multiplex PCR pathotyping method may work for serotypes 2 and 9 but not for serotypes 1, 3, 6, 7, 15 and 16 [16].

Fig. 2
figure 2

A chord diagram showing the interrelationship between clonal complexes (CCs) and the pathotype of human S. suis strains in the current study

Full size image

Furthermore, statistical analysis revealed that human S. suis CC1 isolates were significantly more likely to be type I disease-associated group rather than any other type. Conversely, CC104 and CC25 were significantly more likely to be type IV than any other type (Table 3). A previous study identified a gene variant of copper ATPase 1 in the pig S. suis isolates CC28, CC1109, and serotype 9-ST1105 [16]. In contrast, in the current study, the human serotype 2 isolates ST1, CC25, CC28, CC104, and CC233/379 were more likely to contain the copper ATPase 1 gene variant (Type II or IV) than other ST or CC.

Table 3 Relationship of pathotypes and clonal complexes of human Streptococcus suis isolates
Full size table

Surprisingly, when studying the clinically healthy pig isolates (n = 173), 71.1% (n = 123) were classified in the disease-associated group, consisting of 50.4% type IV (n = 62), followed by 41.5% type III (n = 51), 6.5% type I (n = 8), and 1.6% type II (n = 2). Type IV was most commonly found in serotypes 16 (n = 15) and 3 (n = 13). In addition, 17.3% (n = 30) of the isolates were undetermined, mostly in serotype 31 (n = 7), and only 11.6% (20 isolates) were classified in the non-disease-associated groups, as shown in Table 2. Statistical analysis showed that there was no significant association between serotypes and types of disease-associated groups.

In some cases, pigs may be carriers of virulent isolates of S. suis without presenting clinical signs, but this bacterial species is a normal inhabitant of the upper respiratory tract, and most isolates are usually considered nonpathogenic or non-disease-associated. The results of the current study indicate that multiplex PCR failed to identify this group of clinically healthy pig isolates, and its use in routine diagnosis may lead to a high percentage of false-positive results. Indeed, the original study reported that 77% of pig isolates in the UK were associated with disease and 3% were non-disease-associated according to their origins [15]. However, Scherrer et al. (2020) obtained contradictory results, showing that 34.7% of diseased pig isolates from Switzerland were classified as disease-associated, whereas 65.3% were non-disease-associated using this multiplex PCR method [16]. Our results suggest that the pathotyping PCR method described originally in the UK [15] is probably not useful to clearly differentiate between disease-associated and non-disease-associated strains in clinically healthy pig isolates. It would be interesting to include isolates collected from diseased pigs in future studies to confirm this. It is highly important that multiplex PCR be validated using more diverse S. suis strains from different geographic areas and origins of isolation.

Furthermore, the disease-associated marker genes described by Wileman et al. (2019) should be reconsidered, and additional markers should be identified. Recent studies based on genomic analysis proposed new genetic markers that could distinguish between pathogenic and nonpathogenic pathotypes, including ofs, srtF, RNA-binding protein (SSU_RS09525), and two hypothetical proteins (SSU_RS09155 and SSU_RS03100) [13, 14]. Marker genes SSU_RS09155, SSU_RS09525, and SSU_RS03100 demonstrated strong associations with pathogenic S. suis strains (96%) in North America [14]. The ofs+/srtF + genotype was also present in 74% of the US pathogenic S. suis isolates [13]. These marker genes should be evaluated in further studies with different geographic strains. A previous study demonstrated that the CDS2157 gene, which belongs to the Tex protein family, was absent in all highly virulent ST1 isolates and was specific to the I/WV (intermediate/weakly virulent) group [19]. This gene may be a candidate for evaluation.

An effective pathotyping scheme for distinguishing the pathotypes of S. suis is needed. This could lead to the establishment of a public health surveillance program by promoting a focus on strains with virulence and pathogenic potential from either swine or humans. In particular, it has potential application for predicting pathogenic pathotype strains in pig farms or pig production companies. It also couples monitoring pathotypes of S. suis isolates from raw pork products for food safety purposes and a reduction in zoonotic transmission of this pathogen through improved surveillance programs because poor quality food safety controls for raw pork products at slaughterhouses and wet markets have been suggested as a source of infection in humans [3]. Therefore, pathotyping tools should be continuously developed and further evaluated.

Conclusions

The multiplex PCR pathotyping method for S. suis strains described by Wileman and colleagues could differentiate isolates into four types of disease-associated strains. Some types of disease-associated groups were significantly related to some CCs, such as type I for CC1 and type IV for CC25 and CC104. However, this multiplex PCR method is not useful for differentiating non-disease-associated from disease-associated isolates in the case of Thai clinically healthy pig S. suis isolates, although it works well for human isolates.