Update on Streptococcus suis Research and Prevention in the Era of Antimicrobial Restriction: 4th International Workshop on S. suis ^†

Abstract

Streptococcus suis is a swine pathogen and a zoonotic agent afflicting people in close contact with infected pigs or pork meat. Sporadic cases of human infections have been reported worldwide. In addition, S. suis outbreaks emerged in Asia, making this bacterium a primary health concern in this part of the globe. In pigs, S. suis disease results in decreased performance and increased mortality, which have a significant economic impact on swine production worldwide. Facing the new regulations in preventive use of antimicrobials in livestock and lack of effective vaccines, control of S. suis infections is worrisome. Increasing and sharing of knowledge on this pathogen is of utmost importance. As such, the pathogenesis and epidemiology of the infection, antimicrobial resistance, progress on diagnosis, prevention, and control were among the topics discussed during the 4th International Workshop on Streptococcus suis (held in Montreal, Canada, June 2019). This review gathers together recent findings on this important pathogen from lectures performed by lead researchers from several countries including Australia, Canada, France, Germany, Japan, Spain, Thailand, The Netherlands, UK, and USA. Finally, policies and recommendations for the manufacture, quality control, and use of inactivated autogenous vaccines are addressed to advance this important field in veterinary medicine.

1. Introduction

Streptococcus suis is considered one of the most important bacterial swine pathogens leading to important economic losses to the porcine industry worldwide. S. suis has been reported globally in both traditional and intensive swine operations [1]. Control is based on an alarming overuse of antimicrobials, leading to a dramatic increase of the risk related to antimicrobial resistance. It is also an agent of disease in humans and considered in most OECD (Organisation for Economic Co-operation and Development) countries as an occupational disease affecting mostly swine industry workers. In Asia, this pathogen affects the general population and represents a significant public health concern [2]. After a deadly 2005 Chinese human outbreak, research teams worldwide turned their attention to S. suis with an explosion of published articles (Figure 1).

In 2013, the 1st International Workshop on S. suis was organized with the aim to increase international collaborations. Since then, the upsurge of studies certainly contributed to our understanding of bacterial–host interactions. However, the use of different research models resulted in misconceptions and complicated diagnostics and vaccine development [3,4]. Eight years after the 1st workshop and after two other workshops (2014 and 2016), the 4th International Workshop on S. suis was organized to strengthen scientific knowledge and provide, through international cooperation, relevant scientific information, and advice that will have a direct influence on the decisions made by the swine industry.

The diagnosis and epidemiology of the infection in humans and pigs; different aspects of the pathogenesis of the disease; antimicrobial resistance, prevention and control; and finally autogenous vaccine policy were addressed during the meeting and are further discussed below.

2. Diagnosis and Epidemiology of the S. suis Infection in Humans and Pigs

S. suis is an encapsulated pathogen, and the capsular polysaccharide (CPS) antigen is the basis of S. suis classification into serotypes [1]. Originally, 35 serotypes were identified. However, phylogenetic and/or sequence analyses showed that the reference strains of serotypes 20, 22, 26, 32, 33, and 34 should be taxonomically removed from the S. suis species. Serotypes 32 and 34 were reclassified as Streptococcus orisratti. Serotypes 20, 22, and 26 were proposed as Streptococcus parasuis, while serotype 33 was classified as Streptococcus ruminantium [5,6,7]. Nevertheless, more extensive studies using a higher number of S. suis and S. suis-like strains will provide additional insights into the classification of these strains and make the species boundaries clearer [6]. In fact, all 35 serotypes of S. suis (or S. suis-like strains) (with the exception of serotype 33) are isolated from diseased pigs [1], and from the clinical point of view, many diagnostic laboratories still identify the 35 serotypes. The tools for molecular epidemiology of S. suis have been recently reviewed in Reference [7].

The worldwide distribution of major S. suis serotypes involved in swine clinical cases is schematically represented in Figure 2. Among those reported, serotype 2 is considered the most common cause of infections in piglets worldwide and a major zoonotic agent [2]. Nevertheless, other serotypes are increasing in importance in different countries, as is the case of serotype 9, particularly in some countries of Western Europe. By means of novel animal models and diagnostic tools, de Greeff et al. (Appendix A) epidemiologically determined the population genetics of S. suis serotype 9 in The Netherlands [8]. Obtained data using comparative genome hybridization and whole genome sequencing suggest that clinical serotype 9 swine isolates are genetically very similar whereas serotype 9 isolates carried by healthy pigs are more heterogeneous. A few carriage isolates clustered together with clinical isolates; these carriage isolates probably reflect clinical isolates that are not causing any clinical outbreaks on the farms but do have virulent potential. By infecting Caesarean Derived Colostrum Deprived (CDCD) piglets intravenously with a high dose of bacteria, it was shown that, within S. suis serotype 9, the virulence of clinical and tonsillar carriage isolates differs significantly (Appendix A). Interestingly, a recent study indicated that S. suis isolates associated with disease in pigs comprise predominantly of serotypes 2, 3, and 1/2 (Appendix B), which is consistent with reports from other pig producing countries [2].

In Figure 2, it can also be appreciated that the epidemiological situation in North America is different from other countries. In this part of the globe, multiple serotypes are found in swine clinical cases [2]. Indeed, recent works confirmed that a variety of S. suis serotypes can be found in commercial swine production systems in USA and Canada, with serotypes 1/2, 7, 2, 1, 3, and 5 commonly isolated from systemic infection sites, the frequencies depending on the study [9,10].

Besides this classification based on CPS antigen/gene locus, S. suis is genetically differentiated into sequence types (ST) by multilocus sequence typing (MLST) [7]. The distribution of most important STs within serotype 2 is represented in Figure 3.

It is important to note that epidemiological data is missing for several countries; for example, information relating to S. suis in the Australian pig herd is limited [11,12]. To fill this gap, O’Dea et al. (Appendix B) analysed and characterised a significant number of S. suis isolates from diseased pigs obtained across multiple production sites over a seven year period [13]. The most prominent MLSTs were, in decreasing order, ST27, ST25, ST28, ST483, and ST1. Analysis of serotype and MLST combinations showed a high proportion of isolates as serotype 2 ST25, serotype 3 ST27, and serotype 2 ST28. A deeper phylogenetic comparison of the S. suis isolates from Australian pigs against a global collection of S. suis strains showed that Australian clones of S. suis associated with clinical disease in pigs have maintained a stable core genome, mirroring the international seed-stock from which they were derived (mainly from the UK and North America). While Australian serotype 2 ST25 isolates could be distinguished from those of USA and Canada, only small single nucleotide polymorphism (SNP) differences are notable across the core genome. Despite the limited number, the characterisation of serotype 1/2 ST1 clones is significant, a finding also recently observed in North America. Indeed, Gebhart et al. (Appendix C) characterised the diversity of a contemporary collection of S. suis isolates across North America (mainly in USA) by serotyping and MLST to address the limited information on current S. suis strains circulating within this country [10]. The predominant ST was ST28, followed by ST94, ST1, and ST108 among multiple STs identified, illustrating the high diversity among S. suis isolates. Importantly, this study revealed the predominance of serotype 1/2 ST28 from clinically affected pigs in USA. In addition, the study also reported ST1 strains, containing the three classical virulence-associated genes (epf, sly, and mrp), which are considered highly virulent and potentially zoonotic, an unexpected finding based on previous epidemiological data [2]. Epidemiological surveillance of these strains in North America is recommended [10].

S. suis is a zoonotic disease of increasing awareness. Sporadic cases of human infection have been described worldwide since 1968; however, multiple outbreaks in some Asian countries highlight the public health importance of this infection [2,14,15,16] (Appendix D). Small skin wounds are the main route of entry in humans in Western countries, although in some cases no wound is evidenced [1]. In some Asian countries, such as Vietnam and Thailand, human S. suis is considered among the most frequent causes of bacterial meningitis in adults [17,18]. Cultural differences between Asian and Western countries likely impact the epidemiology of S. suis including lifestyle; common use of backyard production systems; close contact of humans with pigs; and, in countries such as Vietnam, Laos and Thailand, the common practice of consuming raw pork products [19,20,21]. It has been clearly demonstrated that the oral route of infection is the principal cause of human infection in these countries [1,20]. As aforementioned, serotype 2 is the main zoonotic serotype; however, serotype 14 has also been isolated from humans many times in Thailand and UK and uncommonly in France, Australia, and Canada [1,22]. After serotype 14, serotype 5 is the third serotype most commonly found from humans [23]. Sporadic cases due to other serotypes have also been reported [2]; among them, serotype 9, one of the most important serotype recovered from diseased pigs in Europe, has been recently shown to have zoonotic potential [24].

Within serotype 2, STs involved in human cases are represented in Figure 3 and parallel those found in swine clinical cases, highlighting their epidemiological link. Kerdsin et al. (Appendix D) have been extensively studying the epidemiological situation in Thailand [14,22]. To date, four outbreaks of S. suis infections in humans have been recorded in that country, with case fatality rates varying from 6.5% to 16.1%. The food safety campaign implementation in Phayao province during 2011–2013 showed a marked decrease of the disease incidence proportion and stresses the need for health policies to reduce the burden of this infection [20]. Microbiological characterisation of S. suis in Thailand showed that serotype 2 is the main serotype for human infections, followed by serotypes 14, 24, 5, 4, 9, and 31. Of note, ST1 and ST104 (serotype 2) are predominant in Thai human infections. Interestingly, the latter is almost exclusively found in that country. On the other hand, ST105 is the main ST among serotype 14 isolates from human cases. Kerdsin et al. also revealed that Thai serotype 9 isolates are of ST16 (Appendix D) [24], also commonly affecting pigs in some European countries [2,25]. Schultsz et al. (Appendix E) have been studying the emergence of zoonotic S. suis infections in The Netherlands. Zoonotic S. suis serotype 2 strains of ST20, isolated from human patients, were shown to be highly genetically related to serotype 9 strains of ST16 [25]. It has been shown that the CPS is a major virulence factor (see Section 3) and that the cps gene locus can be exchanged between S. suis of different serotypes. Interestingly, a study from The Netherlands [26] showed that such CPS switch may lead to an increase in zoonotic potential. In addition to cps locus acquisition, loss or acquisition of other genes or loci may contribute to changes in zoonotic potential. For example, the S. suis ST20 strains (typically found in The Netherlands) acquired not only the serotype 2 cps locus but also a type 1 Restriction-Modification (R-M) system. Schultsz et al. (Appendix E) suggested that S. suis ST20 strains may have acquired the ability to adapt to specific niches through the acquisition of a R-M system that can regulate expression of CPS and/or other (surface exposed) molecules.

Besides serotype 2 ST1 strains, which present high zoonotic potential worldwide, serotype 2 ST7 is only endemic to China, and it was responsible for human outbreaks in 1998 and 2005 in this country [16]. In a recent study addressing the evolution of ST7 strains in China, it was shown that, of 38 sporadic ST7 S. suis strains, which mostly caused sepsis, serotype 14 was the most frequent, followed by serotype 2. Compared to the genome of the epidemic strain (serotype 2, ST7), the major differences in the genomes of sporadic ST7 strains were the absence of the 89 kb pathogenicity island specific to the epidemic strain and insertion of mobile elements that play a significant role in the horizontal transfer of antimicrobial resistance genes [27]. This is the first study addressing the evolution of the ST7 strains and reporting serotype 14 ST7 isolates, highlighting the need to increase the surveillance of this human life-threatening lineage of S. suis.

In addition to the “classical” described serotypes, a novel variant (serotype Chz) and strains carrying 26 novel capsular polysaccharide loci (NCL1-26) have been identified recently [28,29,30,31,32]. These findings expand the views of the genetic diversity of S. suis cps loci. Novel cps loci are continually being found, and their discovery should eventually reduce the number of untypeable strains recovered from diseased animals. Nevertheless, their virulence potential and the role of these NCLs in the pathogenesis of the disease remain to be evaluated.

3. Virulence Factors and Pathogenesis of the Infection

In pigs, S. suis usually colonises the upper respiratory tract, in particular the pharyngeal and palatine tonsils, but alimentary and genital tracts can also be colonisation sites. In fact, piglets are first colonised by S. suis from birth as soon as they pass through the birth canal, since S. suis is found in the sow vagina. In humans, S. suis nasopharyngeal colonisation has been reported in people working in close contact with infected animals, such as butchers and abattoir workers [1,19,33]. The bacteria may also colonise the gastrointestinal tract in people consuming fresh/raw contaminated pork meat [19]. Consequently, the mucosal barrier is the first line of defence of the host against this pathogen. In swine, colonisation of the upper respiratory tract by S. suis may lead to an asymptomatic carriage but is also considered the first step for the development of an invasive disease, particularly in the context of coinfections with porcine respiratory viruses or polymicrobial infections [34,35,36,37,38,39,40,41,42].

Indeed, S. suis belongs to the “porcine respiratory disease complex”, a term used to describe a common multifactorial respiratory disease in swine that occurs as a result of polymicrobial infections, environmental stressors, and host factors such age and immunological status. Brockmeier et al. (Appendix F) have been analysing the complexity of these polymicrobial infections, which can be the result of multiple viruses, bacteria, or a combination of viruses and bacteria infecting the pig at the same time or in close succession. For example, the majority of cases where porcine reproductive and respiratory syndrome virus (PRRSV) or swine influenza virus (SwIV) was determined to be the primary aetiology, also presented with secondary bacterial pneumonia caused by S. suis, among other bacterial species. Experimental data suggest that coinfections with viruses, such as PRRSV, SwIV, porcine circovirus, and porcine respiratory coronavirus (PRCV), generally result in a more severe clinical disease outcome. There are several mechanisms that might contribute to increased susceptibility to secondary infection and the enhanced disease that often occurs with coinfections, including disruption of the epithelial barrier or alteration of the innate or adaptive immune responses by the primary pathogen, direct interactions between or among the pathogens such as the formation of multispecies biofilms, disturbances to the ecological niche such as the upper respiratory microbiota, or enhanced transmission. Brockmeier et al. (Appendix F) and others [39,43,44,45] have shown that coinfection with PRRSV, SwIV or PRCV and bacteria such as Bordetella bronchiseptica, S. suis or Glaesserella (Haemophilus) parasuis results in a greater incidence of disease, greater percentage of lungs affected, more severe lesions, and slower resolution of such lesions than occurs with infection with a single pathogen alone and that this is often correlated with an amplified local pro-inflammatory cytokine response.

In the absence of appropriate models of polymicrobial infections typically seen in swine production facilities, mechanistic understanding of the underlying complex molecular and cellular interactions remains limited. The transition between S. suis colonisation of the mucosal barriers and systemic infection is only partially understood. In the studies performed by Wells et al. [46], a hypothetical model to explain systemic infection of piglets with S. suis based on the identification of S. suis in the tonsillar lymphoid tissue and presence of large numbers of CD169+ macrophages was proposed. In this model, CD169+ macrophages might act as reservoirs for replication of virulent strains of S. suis, whereby environmental and physical stressors that reduce competition from the tonsil microbiota lead to increased S. suis in the tonsil and escape into the bloodstream through the many small blood vessels permeating the lymphoid tissue. Such a model would also be compatible with the finding that serotype 2 strains interacting with the CD169 Siglec 1 receptor through their sialylated CPS are more commonly associated with invasive disease. Wells et al. also proposed that subversion of the innate functions of CD169+ macrophages by coinfecting viruses such as PRRSV or SwIV may also contribute to the survival of S. suis and spread to the bloodstream. In this context, the potential mechanisms used by S. suis serotypes without sialic acid in their CPSs are largely unknown. It is also possible that survival and or replication capacity of S. suis in this particular subset of macrophages is a characteristic associated with disease-causing strains.

In this regard, research performed by Valentin-Weigand et al. (Appendix G) has significantly contributed to our understanding of viral–bacterial interactions during airway coinfections. One important suggested virulence factor is the S. suis cholesterol-dependent cytotoxin, named suilysin, produced by many virulent strains. This toxin has multiple actions towards a variety of cells (reviewed in Reference [47]), including attachment of S. suis to the surface of the respiratory epithelium and/or direct lysis of epithelial cells [48,49], probably facilitating bacterial breach of this mucosal barrier. However, suilysin-negative clinical isolates are also found frequently in swine populations worldwide and especially in North America [10,50]. Valentin-Weigand et al. found that, during secondary bacterial infection, suilysin contributes to the damage of well-differentiated respiratory epithelial cells in the early stage of infection, whereas cytotoxic effects induced by SwIV became prominent at later stages of infection. Besides suilysin, it has been shown that a prior infection by SwIV enhances the adherence to and colonisation of porcine airway epithelial cells by S. suis in a sialic-acid dependent manner and facilitates invasion in a suilysin-independent fashion [51,52,53]. These findings might explain why suilysin-negative strains can also cause clinical infections (Appendix G). As aforementioned, the potential mechanisms used by S. suis serotypes without sialic acid in their CPSs remain to be discovered.

In humans, the intestinal route of infection seems to be an important port of entry after consumption of fresh/raw contaminated pork meat in some Asian countries. In pigs, this route of infection has also been investigated by direct inoculation in the intestine or oral capsule-mediated delivery of virulent strains of S. suis into the small intestine of piglets [54,55]. In view of the inoculum size, stress-inducing conditions, and the small proportion of challenged animals which developed invasive disease in those studies, translocation from the intestine appears to be a possible but not a very efficient route of infection. The conditions leading to sufficient passage of S. suis through the stomach are still unclear and might differ in neonatal, suckling, or weaning periods. More studies are required on oro-gastrointestinal S. suis infections in piglets to confirm the relevance of intestinal colonisation vs. infection (translocation followed by systemic dissemination) in swine clinical cases [1]. To enlighten this controversial aspect of S. suis disease pathogenesis, de Greeff et al. (Appendix A) studied the putative role of intestinal colonisation for serotype 9 and showed that, during a S. suis serotype 9 outbreak, tonsils of piglets were colonised with different serotypes. The highest percentage of colonisation was found for serotype 9, followed by serotypes 7 and 2 (Appendix A,

Table A1. Presence of S. suis serotypes 1, 2, 7, and 9 in tonsils and faecal samples of piglets at Sterksel farm as determined using qPCR.

Sample	Time Point	S. suis 1	S. suis 2	S. suis 7	S. suis 9
Tonsil	1 day preweaning	10/132 (8%)	7/132 (5%)	43/132 (33%)	113/132 (86%)
	1 week postweaning	2/130 (2%)	10/130 (8%)	80/130 (62%)	124/130 (95%)
	4 weeks postweaning	3/109 (3%)	19/109 (17%)	80/109 (73%)	105/109 (96%)
Feces	1 day preweaning	0/132 (0%)	0/132 (0%)	1/132 (1%)	45/132 (34%)
	1 week postweaning	0/130 (0%)	0/130 (0%)	3/130 (2%)	31/130 (24%)
	4 weeks postweaning	0/109 (0%)	2/109 (0%)	0/109 (0%)	11/109 (10%)

). Similarly, serotype 9 was the major serotype colonising the intestine (Appendix A,

Table A2. Presence of S. suis serotypes 1, 2, 7, and 9 in specific organs (joints and central nervous system (CNS)) and intestinal samples of piglets at Sterksel farm as determined using qPCR.

Sample	Status	S. suis 1	S. suis 2	S. suis 7	S. suis 9
Joints/CNS	Diseased	0/20 (0%)	0/20 (0%)	0/20 (0%)	16/20 (80%)
	Healthy controls	0/20 (0%)	0/20 (0%)	0/20 (0%)	0/20 (0%)
Intestine	Diseased	0/20 (0%)	0/20 (0%)	5/20 (25%)	19/20 (95%)
	Healthy controls	0/20 (0%)	1/20 (5%)	5/20 (25%)	17/20 (85%)

). The data provided by this study suggest that serotype 9 is a better coloniser of the porcine intestine than the other serotypes. Interestingly, previous in vitro studies have shown that serotype 9 isolates can adhere to porcine intestinal cells more efficiently than the other serotypes [54]. Similarly, serotype 2 strains were shown to adhere better to a human intestinal epithelial cell line than serotype 9 strains, suggesting a relationship with the high zoonotic potential of the former serotype (Appendix E). In this in vitro model, the CPS was shown to interfere with these first steps of S. suis colonisation and invasion, as reported in other in vitro studies with epithelial cells [56,57]. Recent developments in stem cell research allowed intestinal organoids to be used as infection models for S. suis. Organoids contain most cell types present in the proximal and distal intestinal mucosa and can be dissociated and grown as monolayers on semipermeable membranes [58,59]. A zoonotic S. suis serotype 2 strain was shown to translocate in this model, which makes it a promising in vitro model for future study of intestinal infection with S. suis (Appendix E).

The lack of a well-standardized in vivo mucosal infection model slowed down research progress in this area. To develop such a model, de Greeff et al. (Appendix A) applied several mucosal infection routes to piglets from different backgrounds. It was shown that intranasal inoculation in combination with oral inoculation result in a reproducible colonisation of the tonsils and the intestine of piglets; however, this mild mucosal colonisation model failed to replicate clinical disease. Different stressors facilitate and increase colonisation by S. suis but are not strict requirements for colonisation. A harsher mucosal challenge can be achieved by combining oral inoculation with intratracheal inoculation. In this case, the level of intestinal colonisation is lower, but more clinical symptoms like lameness, central nervous signs, or septicaemia were induced in the piglets (Appendix A). Further optimization of the oro-gastrointestinal infection model will definitively impact research aimed to dissect how S. suis breaches the mucosal barriers.

Encapsulated extracellular S. suis is a highly invasive pathogen. After penetration of host mucosal barriers, it can reach and survive in the blood and finally invades multiple organs including spleen, liver, kidney, lung, and heart. In addition, S. suis is able to cross the brain microvascular endothelial cells (BMECs) and/or the epithelial cells of the choroid plexus at the blood–brain barrier (BBB) and/or the blood–cerebrospinal fluid barrier to gain access to central nervous system (CNS) [56,60,61]. Septicaemia, meningitis, endocarditis, pneumonia, and arthritis are the most common forms of S. suis invasive disease.

The CPS of S. suis is without doubt the major virulence factor allowing this bacterium evasion of immune-clearance mechanisms [56,62]. However, almost all of the studies on the role of CPS used serotype 2 strains, and little is known on the role of CPSs of other serotypes as a virulence factor. Therefore, it is unknown whether differences in serotype themselves (i.e., differences in CPS structure/composition) directly affect S. suis virulence. To answer this question, for the first time, Okura et al. (Appendix H) experimentally generated serotype switched mutants. Six serotype switched mutants were generated using the serotype 2 reference strain P1/7 (P1/7cps2to3, P1/7cps2to4, P1/7cps2to7, P1/7cps2to8, P1/7cps2to9, and P1/7cps2to14) by exchanging the CPS synthesis gene cluster for those of serotypes 3, 4, 7, 8, 9 and 14, respectively. Their virulence was compared in mice and pigs. Only a serotype 2 switch to CPS type 4 or to type 8 showed a marked and consistent impact of bacterial virulence traits. The CPS8 conferred to S. suis a hyper-virulent phenotype, whereas the CPS4 conferred to S. suis a non-virulent character. Serotype switch from CPS2 to CPS7 or to CPS3 had restricted impact, and serotype switch from CPS2 to CPS14 or to CPS9 had no significant effect on S. suis virulence (Appendix H). Taken together, these findings suggest that serotype switching can differentially modulate S. suis virulence depending on the CPS expressed and demonstrate its importance on S. suis pathogenesis and clinical disease.

Based on biochemical, bioinformatics and in vitro and in vivo gene expression studies, Ferrando et al. [63] proposed a biological model that postulates the effect of carbon catabolite repression on expression of virulence genes in the mucosa, organs, and blood. In the oropharyngeal cavity, where glucose is rapidly absorbed but dietary α-glucans persist, there is a profound effect of carbohydrate availability on the expression of virulence genes. Several virulence factors involved in adherence to host cells, degradation of connective tissue (spreading factors), and avoidance of phagocytic killing, including suilysin, are upregulated when glucose is diminished. As discussed above, suilysin may facilitate dispersion of bacteria in mucosal tissues due to loss of barrier integrity. Once S. suis reaches the bloodstream, metabolism is adapted for optimal growth on glucose and the expression of virulence factors is reduced. In infected organs, glucose levels are lower than in the blood and are further reduced by inflammation and utilization by S. suis, leading to upregulation of suilysin and other virulence factors. These studies have important implications for the design of future control strategies including the development of anti-infective strategies by modulating animal feed composition.

To further understand the genetic basis of disease in S. suis, Weinert et al. [64] studied the genomic signatures of human and pig clinical isolates from the United Kingdom and Vietnam. Isolates associated with disease were shown to contain substantially fewer genes than nonclinical isolates but are more likely to encode virulence factors. Human disease isolates are limited to a single-virulent population, originating in the 1920s, when pig production was intensified, but no reliable genomic differences between pig and human isolates were observed, suggesting lack of consistent genomic adaptation of S. suis to the human population. The authors also reported little geographical clustering of different S. suis subpopulations and high rates of recombination in S. suis worldwide, implying that an increase in virulence anywhere in the world could have a global impact over a short timescale [64].

4. Antimicrobial Resistance

S. suis infections are one of the main causes of antimicrobial usage in piglets. Indeed, the incidence of the disease may be as high as 20%, although it is usually kept lower than 5% in the field due to the extensive and routine prophylactic and/or metaphylactic use of antimicrobials. Data from antimicrobial resistance of S. suis worldwide are alarming, and restriction of antimicrobials as a preventive measure must be a primary concern [65]. In addition, the most effective drugs against S. suis are those in categories 1 and 2 (critically or highly important). The industry is trying to reduce the use of these drugs given their importance in human medicine. Indeed, S. suis is considered a niche for antimicrobial resistance and represents a high risk of transmission of resistance to other pathogens [65].

Surprisingly, despite the worldwide use of beta-lactams in pigs for over 50 years, the majority of clinical S. suis remains sensitive to these antibiotics. However, beta-lactam resistant strains do exist and are primarily found in commensal sites (Appendix I, Figure A1). O’Dea et al. (Appendix B) reported clinical resistance to penicillin G, albeit at a relatively low level in 8.1% of isolates. In addition, they reported similar levels of antimicrobial resistance in Australian strains compared to those overseas with regards to tetracycline (99.3%), erythromycin (83.8%), and trimethoprim/sulfamethoxazole (0.7%). Resistance to florfenicol was 14.9%, while all isolates were clinically susceptible to enrofloxacin, likely due to this being banned from use in food producing animals in Australia. Therefore, this is an aspect of S. suis in Australia that must be carefully monitored from both an animal and public health point of view.

In a recent work, Libante et al. [66] made a comprehensive in silico search and analysis of Integrative Conjugative Elements (ICEs) and Integrative and Mobilizable Elements (IMEs) and extensive identification of antimicrobial resistance genes present in 214 S. suis draft genomes. Almost 400 antimicrobial resistance genes were detected in the 214 genomes analysed. A huge amount of ICEs, IMEs, and derived elements were detected at various chromosomal sites. High diversity was observed in recombination but also in conjugation/mobilization modules. Besides ICEs, IMEs appear to be major vehicles of antimicrobial resistance genes. The authors concluded that further studies are needed to evaluate such gene fluxes inside and between ecosystems and the contribution of ICEs and IMEs in these gene transfers keeping in mind a one health global perspective.

5. Prevention and Control of S. suis Diseases

As S. suis is a very early coloniser of piglets, it cannot be eliminated by early weaning [1]. To reduce antimicrobial use, S. suis disease prevention should concentrate on management of predisposing factors and, mainly, vaccination. Despite intensive research leading to different vaccine-candidate antigens [67], no universally efficacious S. suis vaccine has been commercialised so far. The dream of having a universal cross-protective vaccine is highly challenging due to the high genomic diversity of S. suis. As discussed in the work performed by Weinert et al. (Appendix I), considerable genetic diversity of S. suis exists. Nonetheless, they have shown many trends in genetic and phenotypic differences between strains isolated from the upper respiratory tract of pigs without S. suis clinical signs (referred as “nonclinical”) and strains isolated from the lungs or systemic sites of pigs with S. suis clinical signs (referred as “clinical”) (Appendix I, Figure A1). Overall, the genetic diversity of clinical S. suis is less than that of nonclinical strains, implying that a design of a universal vaccine to control S. suis clinical infection might be possible.

Over 38 subunit vaccine candidates have been reported [67,68] and new ones are continuously being characterised, yet homologous protection is still controversial and cross-protection (either against other serotypes or at least using heterologous strains) was evaluated in few of these studies [69,70,71,72,73,74,75]. Lack of well-standardized animal models for immunization and challenge, especially for serotypes other than 2, is one of the reasons of limited progress towards a subunit vaccine, which also results in contradictory results for a same vaccine candidate [67]. Amongst other confounding factors are the use of mouse models (for pre-screening) without confirmation in swine trials, number of vaccine doses, and different adjuvants (which are not always compatible with a future use in swine medicine). More research, with better standardized protocols, would certainly advance subunit vaccine development.

One common vaccine target is, without doubt, the CPS because the majority of clinical strains have a capsule (Appendix I, Figure A1). Nevertheless, the vaccine-induced protection will be serotype-restricted. Multiple different polysaccharide epitopes should be selected in a vaccine to target different serotypes, and vaccination-driven strain replacement of the population might be expected. Finally, in addition to their diversity and as aforementioned, cps loci are often recombination “hot-spots” and this gives rise to CPS “switching” between strains. Highly conserved proteins might help overcome these limitations and fight strain replacement/evolution. As S. suis is part of the commensal flora, any preventive/control strategy should eliminate clinical isolates without altering the precious balance of the pig mucosal microbiota. Albeit possible disease-protective limitations, a systemic vaccination approach is to be prioritized in this context (Appendix I).

Due to the young age of the piglets affected and the lack of an effective commercial vaccine (as discussed above), many farms use metaphylactic perinatal antimicrobials to control S. suis disease. Besides the public health concern related to antimicrobial use in production animals and the increase in antimicrobial resistance, an additional problem of antimicrobial usage is that these treatments can also affect the beneficial bacteria of the microbiota. In an attempt to understand the harmful effect of overuse of antimicrobials at early age, Aragon et al. (Appendix J) have evaluated the outcome of perinatal antimicrobial treatment on the nasal microbiota at weaning. Elimination of perinatal antimicrobials resulted in an increase in bacterial diversity in the nasal microbiota at weaning. Furthermore, management of lactation without perinatal antimicrobials had a beneficial impact later in life in terms of animal health and productivity in the nursery phase. Albeit more detailed studies are required, Aragon et al. also observed that perinatal ceftiofur administration might favour colonisation with potentially clinically relevant serotypes of S. suis. As such, the precise effects of antimicrobial metaphylaxis are difficult to predict (Appendix J).

The only available vaccines used in the field are autogenous, which consist of killed bacteria (“bacterin”) from the predominant strain(s) recovered in an affected farm, produced by licenced laboratories and given back to the same farm only. However, there are very few scientific studies demonstrating whether the use of such vaccines in the field correlates with a reduced mortality and curative antimicrobial use. Indeed, field peer-reviewed reports on autogenous vaccines are almost nonexistent (only 2 published papers in the last 30 years) [76,77]; others are incomplete and, in most of them, a control (non-vaccinated) group is missing, which may preclude any scientifically sound conclusions [68]. On the other hand, controlled experimental (laboratory) studies have shown contradictory results concerning bacterin-induced protection [67,68]. That controversial protective response could be attributed, among others, to a loss of antigenicity caused by the killing procedure and/or production of antibodies against antigens not associated with protection. Moreover, autogenous vaccines are “manufacturer-related” (each licenced laboratory uses different protocols, antigen concentration, adjuvants, etc.) and “farm-specific” (useful against homologous—same strain—challenge only) [67]. Finally, the correct diagnosis of S. suis as a primary cause of disease may complicate the choice of the strain(s) to be included in the autogenous vaccine. There is still an unresolved issue concerning isolates recovered from lungs, which are considered by many researchers as secondary invaders and may or may not (depending on the laboratory) be included in the vaccine as antigens. Definitively, more studies are required to generate scientific knowledge to improve this important preventive tool and to help reduce the use of antimicrobials.

6. Policy on Autogenous Vaccine Manufacturing

The wide use of autogenous vaccines and cross-border movement of animals vaccinated with autogenous vaccines is now a common practice within Europe. Therefore, harmonising quality of these products has been considered necessary by the Coordination Group for Mutual Recognition and Decentralised Procedure—Veterinary (CMDv) and the national competent authorities for Veterinary Medicinal Products (VMPs) in Europe. M. Saléry discussed in Appendix K the regulatory framework of the manufacture, control, and use of autogenous vaccines within the European Union. Recently, a new legislation on VMPs has been adopted in European Union, applicable in January 2022. In this new regulation, autogenous vaccines have an updated definition and fall under European regimen. Article 2(3) of this regulation requires that some articles of the regulation apply to “inactivated VMP manufactured from pathogens and antigens obtained from an animal or animals in an epidemiological unit and used for the treatment of that animal or those animals in the same epidemiological unit or for the treatment of an animal or animals in a unit having a confirmed epidemiological link”. The introduction of the concept of epidemiological link allows now the use of autogenous vaccines in parental lines or for animals prior their introduction in fattening sites where they will be in contact with new pathogens. In addition, different laboratories working under different conditions presently exist in Europe. From January 2022, all autogenous vaccines will have to be done under Good Manufacturing Practices (GMP) or GMP-like requirements. The manufacturers will have to be authorized, and the compliance with the requirements will be controlled by inspection. As a consequence, autogenous vaccines will be produced under similar quality conditions. However, among others, differences in antigen preparation, antigen concentration, method of killing, and type and concentration of the adjuvants will still exist. In Appendix K, M. Saléry summarized key points from the CMDv’ recommendation paper [78], which deals with the manufacture, the control, and the use of viral and bacterial autogenous vaccines in Europe. In the framework of the new regulation, the CMDv’ recommendations will be updated in the coming years. The current and future regulatory framework ensures those products are of high quality. Nevertheless, safety and efficacy are not regulated. Safety and, more importantly, efficacy of S. suis autogenous vaccines need to be better demonstrated through field use and laboratory research.

7. Conclusions

Intensification of animal food production and emergence of new production systems (such as raised without antibiotics or organic systems as well as “Humane” animal farming) has resulted in emergence or reemergence of pathogens. In this context, coinfections are more likely to occur, increasing the incidence and/or enhancing clinical disease with certain pathogens, including S. suis. Continued development of suitable models for polymicrobial infections and a better understanding of the underlying pathological mechanisms are required to develop effective intervention strategies to prevent the effects of these diseases on swine production.

Mucosal infection models still require optimization and have the potential to improve our knowledge of the pathogenesis of S. suis-induced disease. These models would significantly contribute to vaccine development as well. Research on mucosal infection and immunity will also help address the potential use of nutrition management and/or microbiota enhancement to improve swine health and, consequently, S. suis control strategies. Early medication with antimicrobials has to be carefully considered since antimicrobials may interfere with the establishment of the microbiota and, in consequence, with immune maturation and other microbiota functions, which may have a lasting health effect later in life.

S. suis importance as a zoonotic pathogen is continuously increasing due to, at least in part, the pressure to reduce antimicrobial use in livestock. This public health threat is enhanced by the lack of universally effective vaccines that might reduce the infection load in swine and thus the risk for humans. Global strengthening of swine trade amongst countries has the potential to facilitate exchange of genetic material through recombination and mobile genetic elements, which may result in selective advantage and niche adaptation. It is crucial to understand the factors that are involved in the ability to cause zoonotic infection and to monitor if these may eventually result in S. suis strains that are more human adapted.

Indeed, a better worldwide surveillance system of S. suis-related disease in swine and humans would improve our understanding of the epidemiological evolution of this pathogen. New molecular tools have been developed; however, due to the diversity of S. suis and lack of appropriate markers to differentiate virulent strains from commensal ones, alternative techniques might still be required to achieve a comprehensive understanding of the S. suis bacterial community, including virulence-associated gene profiling. The development of such tools that might eventually allow prediction of virulence potential of a strain might be compromised by sampling strategies and the rich S. suis community in the upper respiratory tract of healthy animals.

In the era of antimicrobial restrictions and new social meat consumption trends, improvement and/or development of vaccination strategies is of utmost importance. In spite of decades of research on S. suis vaccines, autogenous bacterins are the only control strategy available that swine producers have access to. Due to intensive use of these vaccines in Europe, new policies are in place that would improve and normalize their manufacturing. The implantation of similar policies worldwide, including North America, where the use of autogenous vaccines is also widespread, would be expected or sought after. Nevertheless, more field studies are essential to scientifically validate their protective effect and, consequently, their cost–benefit impact for swine producers.