Streptococcus suis infection on European farms is associated with an altered tonsil microbiome and resistome

Streptoccocus suis is a Gram-positive opportunistic pathogen causing systemic disease in piglets around weaning age. Outbreaks of S. suis disease are controlled by metaphylactic use of antibiotics, leading to high levels of antimicrobial resistance in S. suis isolates. This is an 30 issue for both animal and human health due to the zoonotic disease potential of S. suis . The 31 mechanisms facilitating invasive disease are not known but may involve host and 32 environmental factors. The palatine tonsils are considered a portal of entry for pathogenic 33 strains to cause systemic disease. We hypothesised that tonsil colonization by pathogenic and 34 commensal bacteria may impact on disease risk via colonization resistance and co-infections. We conducted a case-control study on 9 European farms, comparing the tonsil microbiome of 36 piglets with S. suis systemic disease with asymptomatic controls. We also compared these to 37 piglets on control farms and piglets reared naturally in a forest. We found a small but significant difference in the tonsil microbiota composition of case and control piglets. Case-control associations varied between amplicon sequence variants (ASVs) and Variants commensal Rothia case compared ,

identical to both G. parasuis and Actinobacillus indolicus) trended towards case-association 1 4 2 (0.27% vs 0.16%). Pasteurella multocida was relatively prevalent (50%) but at low 1 4 3 abundance (mean 0.06%), although two outlier case piglets had over 1% abundance. Actinobacillus pleuropneumoniae had 34% prevalence and 0.08% mean abundance but was 1 4 5 most abundant in control piglets. Bordetella bronchiseptica had lower prevalence (17%) but 1 4 6 higher abundance (0.10%) due to some control piglets having up to 12% abundance. Mycoplasma hyopneumoniae was virtually absent from the dataset, with only 6 reads from a 1 4 8 single control piglet.  The tonsillar microbiome of piglets is similar between farms but diverged in free-range The tonsillar microbiota composition was more similar between piglets on the same farm European farm piglets shared a large core microbiota with 83 ASVs being present in 80% of 2 0 9 piglets or more, 44 of these were not found in the free-range piglets. Vice versa of 157 ASVs 2 1 0 found in all 5 forest piglets, 117 were not found in any farm piglet. Free-range piglets had 2 1 1 low abundance of the genera most abundant in farm piglets, in particular Moraxella (0.6% vs 2 1 2 13%) and Streptococcus (0.9% vs 12%), and higher abundance of Acinetobacter (10% vs 2 1 3 9 2.3%), "Rikenellaceae RC9 gut group" (9.9% vs 0.1%), and Treponema pedis (3.6% vs 2 1 4 0.04%). The sample with the lowest S. suis abundance in the study, 0.3% abundance of a 2 1 5 single ASV (ASV 1050), was from a free-range piglet. This ASV did not have 100% identity 2 1 6 to the 16S rRNA gene V3-V4 region of any S. suis strain publicly available in SILVA or 2 1 7 NCBI assembly databases. The other free-range piglets were colonised by S. suis ASVs 2 1 8 shared with farm piglets.  log(1000*abundance+1)). Samples clustered broadly by country, but samples from farm DE6 2 2 3 clustered with NL1, and samples from ES4 clustered away from the other Spanish farms. B.

4
Mean pairwise Bray-Curtis dissimilarity between samples from the different farms. C. PCA 2 2 5 on ARG abundance (transformed with log(1000*abundance+1)). The pre-and post-weaning  We quantified the abundance of antibiotic resistance genes (ARGs), collectively called the 2 2 9 resistome, by mapping metagenomic reads to the Resfinder database [26] and normalizing 2 3 0 abundance by fragments per kilobase reference per million fragments (FPKM). Piglets on 2 3 1 most farms received antibiotics via feed or water, and these were included for resistome 2 3 2 analysis. Twenty-two piglets received intramuscular injections, and these were excluded from 2 3 3 the main analysis. All case-control pairs included in the main analysis had received equal 2 3 4 antimicrobial treatment. Farms varied in total ARG abundance ( Figure 3D), but shared high  not by farm ( Figure 3C). sampled piglets. This study could not disentangle these two effects since most antimicrobial 2 4 6 treatments were given equally to all sampled piglets at each farm. Farm NL1, a high health 2 4 7 status research farm where the sampled piglets were not treated, and where piglets are rarely 2 4 8 treated with antimicrobials, had the lowest ARG abundance (except for the free-range forest 2 4 9 piglets). In Germany, high health status farms DE1 and DE2 had low ARG abundance, but so 2 5 0 did DE6 despite a history of severe S. suis disease. Farms DE3, ES3, and ES4 were assessed (supplementary file 1), and piglets on these farms were also administered antimicrobials 2 5 3 before sampling. These three farms had the highest ARG abundance ( Figure 3D). There was 2 5 4 no consistent link between the antimicrobials administered and the abundance of ARGs conferring resistance to these. On farm DE1, piglets that had received tetracycline had lower all piglets had received tetracycline, tetracycline ARG abundance was lower than on DE2 and   In total 89 ASVs were classified as S. suis, and 52 of these were present in 2 or more tonsil genomes. This likely relates to the fact that many clinical but few non-clinical strains have 3 1 0 been sequenced. ASVs are, however, not good markers for assessing S. suis strain diversity as 3 1 1 the 16S rRNA gene V3-V4 region correlates poorly with whole genome phylogeny, and as 21% of all S. suis in the microbiota but was only found in 0.08% of the assemblies. In 3 1 6 addition to strain DE512T1 collected in the present study, ASV 17 was only found in 2 3 1 7 isolates recently sampled from diseased pigs in China (GCF_019793915.1 and 3 1 8 GCF_019794525.1).

1 9
Further, we used metagenomic data to assess the relative proportion of commensal and 3 2 0 pathogenic S. suis in the tonsillar microbiota of each piglet. We created a S. suis pangenome 3 2 1 by clustering protein coding genes from a previously published genome collection [21] at 3 2 2 80% identity, and mapped metagenomic reads to the representative sequences of each cluster 3 2 3 to assess their abundance in each sample. We calculated the ratio of prevalence of each 3 2 4 cluster in clinical and non-clinical genomes to assess their putative association with 3 2 5 pathogenicity. We found that in the tonsillar microbiota, genes predominantly found in non- 0.01), due to higher abundance of commensal S. suis in control samples. One often used S. suis marker gene for strain virulence is the gene encoding suilysin (sly), similarly abundant in case and control piglets ( Figure 5D). On some farms we did not detect 3 3 5 sly in any piglets, but this may be due to small sample size and insufficient sequencing depth.  In this study, we found the tonsil microbiota composition of case piglets with S. suis clinical (opportunistic) pathogens [18,27]. We did not find disease-associations with the tonsil 3 5 3 microbiota abundance of any species linked to PRDC, such as G. parasuis or S. suis itself.

5 4
This study included only piglets with weaning age systemic S. suis disease, and PRDC 3 5 5 associated taxa may be more relevant to respiratory disease in older finisher pigs.

5 6
We identified novel disease-associations with Fusobacterium gastrosuis, Bacteroides Alloprevotella/Prevotella/Bacteroides species lack isolate genomes and are unknown in 3 6 2 relation to S. suis disease, possibly due to being unculturable. Case-associated taxa may also 3 6 3 interact with the host to facilitate S. suis to cross epithelial barriers without themselves 3 6 4 entering the bloodstream, thus remaining undetected by necropsy. Alternatively, disease-3 6 5 associated taxa may increase in abundance due to host immune status and dysbiosis, as  Case piglets had lower tonsillar S. suis abundance than control piglets. We assessed that this 3 6 8 was due to a reduced abundance of strains from commensal clades, most of which are poorly represented among sequenced S. suis strains. While S. suis genes predominantly found in 3 7 0 non-clinical isolates were more abundant in control piglets, genes predominantly found in cases and controls. This shows that the majority of S. suis colonising tonsils are commensal, 3 7 3 lacking genes required for invading the host, but also confirms that strains carrying genes 3 7 4 most prevalent in clinical strains are also colonising asymptomatic piglets at low abundance. Based on these results, we conclude that tonsillar colonisation by S. suis itself cannot be used 3 7 6 to reliably predict or even confirm ongoing S. suis invasive disease. Associations between S. suis disease and the tonsillar microbiota have been investigated Lachnospiraceae were found to be more abundant in controls, while case piglets had higher  Differences between studies may largely be due to methodological differences, but in the 3 8 3 present study we included one US farm and found the sampled piglets to have low diversity, 3 8 4 and that while microbiota members were shared at the ASV level, composition was swabs. Furthermore, piglets suffering from other pathologies such as rectal prolapse and 3 8 8 hernia were sampled as controls, so the case piglets were not compared with healthy controls 3 8 9 as in the present study. While their study may include control-associated taxa associated with 3 9 0 other disease, our results may not be specific only to S. suis disease but include microbiome  We found case piglets to have higher abundance of antimicrobial resistance genes than 3 9 5 controls, despite case-control pairs being treated with the same antimicrobials. ARGs conferring resistance to doxycycline and tetracycline had the strongest case-association, and 3 9 7 ARGs conferring resistance to other antimicrobials used against S. suis also trended towards   Farms had significant differences in microbiota composition, but clustered by country. This  We sampled the tonsillar microbiota of 5 piglets living outdoors in a forest in The on intensive farms, including S. suis associated disease. We found the tonsillar microbiota of  We conducted a longitudinal sampling on farm ES2, collecting prospective samples from pre-  This suggests that maternal effects involving early life immunity-or microbiota may be 4 4 9 important in predisposing to S. suis disease and potential dysbiosis. This may be due to 4 5 0 vertical transmission of a disease-prone microbiota, but also differences in maternal 4 5 1 immunity, with antibodies depleting prematurely [46]. Streptococcus suis disease most 4 5 2 commonly occurs around the time colostral antibodies to S. suis start becoming depleted.

5 3
Various studies have found other colostral antibodies to have similar half-lives as well as piglets in some litters lack a sufficient level of maternal immunity against S. suis.

5 8
In conclusion, there are small but significant differences between the tonsillar microbiota of S. suis case piglets and asymptomatic controls. We discovered novel taxa associated with 4 6 0 case piglets, while S. suis abundance was higher in controls. The microbiota differences may 4 6 1 originate from dysbiosis starting during early life prior to disease outbreak, but further 4 6 2 research is needed to assert this. It is also not conclusively known whether S. suis invades resistance against classes commonly used to treat S. suis disease. This may be linked to high 4 6 5 ARG prevalence in case-associated taxa driven by more frequent exposure antimicrobial 4 6 6 treatment than control-associated taxa. We utilized a case-control study design to assess the association between tonsillar microbiota 4 7 0 composition and incidence of S. suis invasive disease. To search for consistent correlations 4 7 1 between microbiota composition and S. suis disease incidence we included farms from 4 7 2 different countries with different livestock management systems (supplementary text 1).

7 3
Tonsil swabs were obtained from 3-to 10-week-old piglets at 13 farms, of which 9 had 4 7 4 ongoing S. suis disease outbreaks. Three sampled farms had no history of S. suis disease. We  We selected 45 pairs of case-control piglets for metagenomic sequencing (table 1). The pairs were selected to be as equal as possible, coming from the same pen, room, and/or sow, and  This study and all animal procedures were approved by the appropriate ethical committees.   Whole genome sequencing of bacterial isolates 4 9 9 We selected 9 clinical (isolated from lesions observed at necropsy on the farms, but not  statistics is shown in Table S5.

0 7
Sample collection 5 0 8 The palatine tonsil microbiota of piglets was sampled by gently scraping the tonsillar surface 5 0 9 with HydraFlock swabs (Puritan, ME, USA) for 10 seconds. Swabs were immediately put in The V3-V4 region of the 16S rRNA gene was amplified with primers 341F (5′- compositional differences between groups. Vegan function RDA was used for principal  Metagenomic libraries were prepared with NEB Next Ultra DNA Library Prep Kit (New   5  3  3 England Biolabs, ME, USA) following the manufacturer's instructions. DNA was fragmented bp paired-end sequencing on an Illumina NovaSeq 6000 machine.

6 1
We mapped metagenomic reads to the representative sequence of each cluster as described above and accepted reads mapping at 80% identity and 80% length.