Feed additives for the control of post-weaning *Streptococcus suis* disease and the effect on the faecal and nasal microbiota

The study addresses alternatives to antimicrobial metaphylaxis for controlling post-weaning Streptococcus suis disease in pigs. Routine antimicrobial use in swine, particularly around weaning, has contributed to antimicrobial resistance. Weaning is a high-risk period due to mixing of litters, dietary transition, and immature immunity, predisposing piglets to pathogens such as S. suis, an early colonizer of the upper respiratory tract that can cause severe disease (meningitis, septicemia) post-weaning. Historically, in-feed beta-lactams (e.g., amoxicillin) are used to mitigate disease, but stewardship efforts demand reduction of antimicrobial use. The objective was to evaluate three feed additive strategies—lysozyme, medium-chain fatty acids (FA) with lysozyme, and FA with a natural anti-inflammatory—versus amoxicillin and a non-treated control, measuring clinical outcomes, growth, and effects on faecal and nasal microbiota.

Background highlights include: widespread antimicrobial use has selected for multidrug-resistant bacteria. Weaning is a critical period for pig health requiring frequent antimicrobial interventions. S. suis is prevalent and pathogenic in nursery pigs, with farm-level diagnosis often clinical. Beta-lactams are commonly used against S. suis but stewardship policies encourage alternatives. Nutritional strategies can modulate gut microbiota and support immunity post-weaning. Prior studies show that microbiota diversity, including nasal communities, associates with health, and nasal microbiota may influence initial steps of S. suis infection. Longitudinal studies document rapid increases in piglet microbiota diversity post-weaning, with diet-driven shifts (e.g., increases in Prevotella, decreases in Lactobacillus) and links to performance. Some taxa (e.g., Weeksellaceae, Fusobacterium, Bergeyella) have been associated with respiratory disease, whereas Lachnospiraceae and Ruminococcaceae often associate with health.

Study site: A conventional pig farm with recurrent S. suis disease. Ethical compliance per EU Directive 2010/63/EU and Spanish RD 53/2013; faecal/nasal sampling classified as minimally invasive. Design: 569 piglets (21 days old) from 54 dams were enrolled. One-week pre-weaning, piglets were weighed (~4 kg mean), ear-tagged, and randomized into five groups, balancing dam of origin and weight: Lysozyme (Lys, n=113); medium-chain fatty acids plus lysozyme (FA + Lys, n=110); amoxicillin (Amox, in-feed; n=115); FA plus a natural anti-inflammatory (FA + antiinf; “Naturporc”; n=119); and negative control (no additives; n=112). Treatments were assigned across rooms (two per room). Farm personnel and veterinarian were blinded to treatment allocation. Clinical monitoring: Animals were observed daily; a veterinarian examined every 2 days. Clinical signs (limping, nervous signs, coughing, wasting, diarrhea) were recorded individually. Animals requiring antibiotic treatment were removed; removals and deaths were recorded with presumptive cause. At day 43 (end of nursery), remaining animals (n=434) were weighed to compute average daily weight gain (ADWG). Three animals (from control, Amox, FA + antiinf groups) were euthanized for diagnostics. Pathogen detection and culture: From a subset of 125 piglets (one per litter per group; cross-fostered excluded), nasal swabs were taken at weaning (D0) and pre-departure (D21/43); 123 at D0 due to 2 mislabeled. PCR detected S. suis and virulence-associated vtaA genes of Glaesserella (Haemophilus) parasuis. From euthanized pigs with CNS/lameness, foramen magnum swabs were cultured on chocolate agar; S. suis identification and serotyping (serotypes 2 and 9) performed; genotyping by ERIC-PCR. Antimicrobial susceptibility by broth microdilution (CLSI) against a panel of antimicrobials; breakpoints per CLSI. Microbiota sampling and sequencing: For microbiota, faecal content and nasal swabs from 12 piglets at D0 (pre-treatment), and 10 piglets per treatment group at D21 were analyzed. DNA extraction: faeces (NucleoMag VET), nasal swabs (NucleoSpin Blood). 16S rRNA V3–V4 region amplified (Klindworth primers) and sequenced on Illumina MiSeq 2x250. Rectal and nasal libraries prepared at Illinois Carver Biotechnology Center and UAB Genomics Service, respectively. Bioinformatics: QIIME2 v2019.10 pipeline; demultiplexing and QC (q2-demux), DADA2 denoising to generate ASVs, MAFFT alignment, FastTree2 phylogeny, taxonomy via q2-feature-classifier (naïve Bayes) against Greengenes 13.8. Rarefaction: faecal 8,500 reads, nasal 19,000 reads. Alpha diversity (Shannon, Chao) via non-parametric t-tests (Monte Carlo, 999 permutations). Beta diversity via Bray–Curtis and Jaccard; PCoA visualization with EMPeror; variance explained (R2) by Adonis (vegan), P-values via permutations (999). PERMANOVA for group comparisons (q2 diversity beta-group-significance). Differential abundance by ANCOM on frequency-filtered (n>10) genus-level tables and at ASV level; analyses repeated considering dam parity (gilts vs sows). Statistics: Clinical prevalence comparisons by Pearson’s Chi-squared. ADWG normality checked by Shapiro–Wilk; if normal, ANOVA with Tukey post hoc; if non-normal, Kruskal–Wallis with Dunn’s test and Bonferroni correction. Significance threshold P≤0.05. Analyses in R.

- Clinical outcomes (S. suis-like signs: limping/nervous): Prevalence—Lys 17.7% (20/113 initial); FA+Lys 12.7% (14/110); Control 22.3% (25/112); Amox 9.6% (11/115); FA+antiinf 9.2% (11/119). Compared to Amox: FA+Lys (P=0.45) and FA+antiinf (P=0.93) not significantly different; Lys higher with trend (P=0.07). Compared to Control: FA+antiinf significantly lower (P=0.006); FA+Lys showed a trend (P=0.06); Lys not significant (P=0.39). - Overall disease removals (any clinical sign): Removed 135/569 (23.7%). Group-level prevalence—Lys 21.2% (24 removed); FA+Lys 18.2% (20); Control 34.8% (39); Amox 19.1% (22); FA+antiinf 25.2% (30). Versus Amox: no significant differences (Lys P=0.69; FA+Lys P=0.86; FA+antiinf P=0.26). Versus Control: Lys lower (P=0.02) and FA+Lys lower (P=0.005); FA+antiinf not significant (P=0.11). - Growth: ADWG 305.2–320.1 g/day across groups; no significant differences (Shapiro–Wilk normal; ANOVA P=0.67). - Pathogen occurrence: S. suis colonization at D0: 121/123 positive; at D43: 101/102 positive (only one negative in Amox group). G. parasuis virulent strains: 37.2% (D0) increasing to 96.1% (98/102) at study end. No association between treatment and colonization by either pathogen. Two lesion-derived S. suis serotype 2 isolates were susceptible to penicillin, ampicillin, ceftiofur, and resistant to chlortetracycline (MIC ≥16), oxytetracycline (MIC ≥16), clindamycin (>16), neomycin (MIC ≥32), tylosin (>32), tulathromycin (>64); both shared ERIC-PCR profiles. - Microbiota dynamics over time: Alpha diversity increased from D0 to D21 in faeces (Shannon P=0.000001; Chao P=0.00002) and nasal (Shannon P=0.000006; Chao P=0.017). Beta diversity separated D0 vs D21: faeces Bray–Curtis R2=10.10%, Jaccard R2=8.58% (P=0.001); nasal Bray–Curtis R2=33.50%, Jaccard R2=12.54% (P=0.001). ANCOM identified 20 faecal and 37 nasal genera changing over time. - Microbiota after treatments (D21): Faecal alpha diversity lower in Amox vs Control (Shannon P=0.01) and trended lower vs FA+antiinf (P=0.06). Faecal richness: Amox lower than Control (Chao P=0.016); trended higher than Lys (P=0.08). Parity effects: piglets from gilts showed lower faecal richness than those from sows (P=0.037). Differences by treatment were more apparent among piglets from gilts. - Nasal alpha diversity: FA+antiinf highest vs Control (P=0.0004); Amox trended higher than Control (P=0.06); FA+antiinf trended higher than Lys and FA+Lys (P=0.013 and P=0.003 reported in figure comparisons). By parity: in sows, Amox > Control (P=0.012) and FA+antiinf highest; in gilts, FA+Lys < FA+antiinf (P=0.05). Nasal richness was highest for FA+antiinf and Amox; piglets from gilts consistently showed lower richness (P=0.021). - Beta diversity by treatment (D21): Faecal PCoA: treatment explained ~9% (Bray–Curtis R2=9.50% P=0.05; Jaccard R2=9.21% P=0.01). Including parity improved Jaccard (R2=19.61% P=0.036) but Bray–Curtis not significant (R2=19.75% P=0.127). Among piglets from gilts, treatment explained more variance (Bray–Curtis R2=40.25% P=0.05; Jaccard R2=38.04% P=0.08). Nasal PCoA: treatment effects stronger—Bray–Curtis R2=23.71% P=0.001; Jaccard R2=15.04% P=0.001. Including parity increased explained variance (Bray–Curtis R2=33.26% P=0.001; Jaccard R2=25.88% P=0.001). Pairwise PERMANOVA significant among all treatments (P<0.01). FA+antiinf group showed more homogeneous nasal microbiota. - Differential abundance (ANCOM, D21): Faeces—genus Mitsuokella differed among treatments (detected mainly in FA+Lys). No genus differences when including parity; one low-abundance ASV (Elusimicrobiaceae) differed by treatment when stratified by parity (present only in gilts of Lys and FA+Lys). Nasal—genera Helicobacter and an unclassified Weeksellaceae differed among treatments; with parity, Lachnospira and an unclassified Weeksellaceae differed. ASV-level nasal differences: two Lachnospiraceae and one Bacteroidales ASVs by treatment; with parity, one Clostridiales, one Lachnospiraceae, and one Ruminococcaceae ASVs differed. In FA+antiinf, an unclassified Weeksellaceae genus was relatively lower, while a Ruminococcaceae ASV was higher.

The study demonstrates that specific feed additives can reduce clinical signs compatible with S. suis disease to levels comparable to amoxicillin while avoiding antimicrobial metaphylaxis. FA + anti-inflammatory supplementation most closely matched amoxicillin for reducing S. suis-like signs and produced the highest nasal microbiota diversity and richness, which is often associated with healthier respiratory ecosystems. Although none of the interventions altered S. suis or G. parasuis colonization prevalence, clinical outcomes improved, implying modulation of microbial community structure and host resilience rather than pathogen eradication. Microbiota analyses showed more pronounced treatment effects in the nasal cavity than in the gut, despite oral administration, suggesting differential ecological responsiveness of the respiratory microbiome and potential relevance to S. suis pathogenesis. Dam parity emerged as an influential factor; piglets from gilts exhibited lower diversity and richness in both faecal and nasal microbiota, potentially contributing to increased susceptibility to disease. Unexpectedly, nasal diversity tended to be higher under amoxicillin than control, possibly reflecting transient colonizers in a depleted ecosystem, warranting further study. Overall, results support nutritional strategies, particularly FA + anti-inflammatory supplements, as promising components of antimicrobial stewardship programs targeting post-weaning disease.

Feed supplementation with medium-chain fatty acids plus a natural anti-inflammatory reduced clinical signs compatible with S. suis to levels similar to amoxicillin and yielded the highest nasal microbiota diversity and richness. FA + lysozyme and lysozyme alone improved some general disease indicators compared with control. Treatments had limited impact on faecal microbiota composition, while nasal microbiota showed clearer, treatment-specific shifts. Sow parity significantly influenced offspring microbiota, with piglets from gilts showing reduced diversity and richness. These findings indicate that certain feed additives can serve as alternatives to in-feed amoxicillin for controlling post-weaning disease signs in S. suis-affected farms and highlight the importance of considering dam parity in microbiota-targeted interventions. Future research should evaluate long-term growth and health outcomes to market weight, validate findings across diverse farms and health statuses, elucidate mechanisms linking specific taxa (e.g., Weeksellaceae decrease and Ruminococcaceae increase) to disease resistance, and test targeted interventions for gilts (e.g., pre/probiotics, stress reduction) to improve offspring microbiota development.

- Single-farm study under known S. suis infection pressure; external validity may be limited. - Colonization by S. suis and G. parasuis was not reduced by any treatment; clinical improvements may not generalize to settings with different pathogen dynamics. - Growth analysis (ADWG) excluded diseased/removed animals and ended at nursery departure (day 43), limiting inference to finishing performance. - Some statistical results were trends (P between 0.05 and 0.1); multiple comparisons may affect robustness despite corrections. - Microbiota sequencing targeted 16S V3–V4 with Greengenes taxonomy; taxonomic resolution to genus/ASV limits functional inference. - Parity effects were observational (gilts vs sows) without controlled interventions; potential confounders (e.g., prenatal stress, milk composition) not directly measured.