Curing piglets from diarrhea and preparation of a healthy microbiome with *Bacillus* treatment for industrial animal breeding

Piglet epidemic diarrhea (PED) spreads rapidly and causes severe dehydration, morbidity, and mortality in swine herds, challenging animal welfare and global meat production. Diarrhea in 5–15 day-old piglets is commonly associated with pathogens such as Isospora suis, Clostridia, Escherichia coli, Salmonella choleraesuis and Brachyspira spp., especially under poor housing/hygiene conditions. Widespread antibiotic use in livestock faces drawbacks: adverse effects, emergence of multi-drug resistance, and disruption of gut microbiota. There is a need for effective alternatives that both alleviate clinical diarrhea and restore a healthy gut ecosystem. This study tests whether a defined Bacillus probiotic cocktail (B. subtilis Y-15, B. amyloliquefaciens DN6502, B. licheniformis SDZD02) can cure piglet diarrhea and re-establish a healthy microbiome, compared against standard antibiotics (colistin and kitasamycin) and untreated controls, using fecal 16S rRNA gene sequencing to characterize microbiome changes.

Non-antibiotic strategies for pig production include acidifiers, prebiotics, yeasts, and plant oils, but some plant oils have side effects. Probiotics such as Bacillus and Lactobacillus strains have shown benefits for digestive enzyme activity, gut integrity, immunity, and growth in swine and rodent models. Bacillus species like B. subtilis, B. amyloliquefaciens, and B. licheniformis have been studied for immunomodulation, antimicrobial compound production, and environmental robustness, supporting their use as probiotics. Prior work also documents antibiotic-induced gut dysbiosis in animals and limited efficacy of antibiotics to restore eubiotic microbiota, underscoring the rationale for probiotic alternatives.

Design: Four groups of piglets (N=80 total; n=20 per group) were studied: Group I healthy control (normal), Group II ill untreated (diarrhea), Group III ill + antibiotics (AB), Group IV ill + Bacillus (microecosystem). Piglets (Duroc × Landrace × Yorkshire), 25±1 days old, 7.0–7.68 kg, were reared individually in nursery beds (24–31 °C; 32–67% humidity; constant daylight) in YuZhou city (Henan, China). Diarrhea occurred naturally; piglets with watery diarrhea were selected for ill groups. Treatments lasted 7 days. Diets: all groups received standard feed; AB group received 400 ppm colistin plus 100 ppm kitasamycin; microecosystem group received 3000 ppm of a direct-fed microbial containing Bacillus subtilis Y-15, B. amyloliquefaciens DN6502, and B. licheniformis SDZD02 (3.0×10^10 CFU/g total; 1.0×10^10 CFU/g each strain), produced by Shandong Global Biotech Co. Ltd.; healthy controls received a 3000 ppm placebo carrier. Diarrhea incidence was recorded; both AB and microecosystem groups achieved clinical recovery by day 7; no relapse up to 1 month (follow-up). Sampling: After 7 days of treatment, fecal samples (2.0 g) were collected (ten piglets per group for sequencing), stored at −80 °C, and processed for genomic DNA (QIAamp Fast DNA Stool Mini Kit). 16S rRNA gene amplification used universal primers (16F and 16R) with standard PCR cycling. Sequencing: Illumina HiSeq/MiSeq (BGI) with quality filtering (Phred Q20 sliding window; adapter and low-complexity removal). Paired-end reads were merged (FLASH; min overlap 15 bp, mismatch ≤0.1). OTU clustering: USEARCH/UPARSE at 97% similarity; chimeras removed with UCHIME (mapping to Gold/UNITE). Taxonomy: RDP Classifier v2.2 trained on Greengenes (201305), confidence 0.8; unassigned OTUs removed. Diversity and statistics: Venn diagrams (R VennDiagram), PCA on OTU relative abundance (R ade4), rank-abundance curves, taxonomic histograms (phylum to species), heatmaps (R gplots; Euclidean distance; complete linkage). Alpha-diversity indices (Observed species, Chao1, ACE, Shannon H, Simpson D) computed with mothur v1.31.2; rarefaction curves plotted in R. Group-wise comparisons used Wilcoxon rank-sum and Kruskal–Wallis tests. Additional analyses: Phylogenetic trees via QIIME/PyNAST/FastTree; genus-level phylogeny. MicrobiomeAnalyst used to assess alpha/beta diversity for grouped comparisons (A: healthy+microecosystem vs B: diarrhea+antibiotics); NMDS with Bray–Curtis distances.

- Clinical outcome: Both antibiotics and Bacillus groups showed resolution of diarrhea by day 7; no diarrhea observed up to 1 month post-treatment, especially in the Bacillus group. - Overall microbiome similarity: Venn diagrams, PCA, heatmaps, and phylogeny showed microecosystem-treated piglets clustered with healthy controls, while antibiotics clustered with ill-diarrhea. - OTU richness (Table 1): Microecosystem 335 OTUs; Normal 301; Antibiotics 227; Diarrhea 160. Tag numbers per sample: Antibiotics 18,549; Diarrhea 19,477; Microecosystem 16,848; Normal 17,942. - Alpha diversity (Table 2): • Antibiotics: Sobs 227.0; Chao1 249.5; ACE 249.48; Shannon H 3.559; Simpson D 0.0506. • Diarrhea: Sobs 160.0; Chao1 172.0; ACE 177.37; Shannon H 2.950; Simpson D 0.1134. • Microecosystem: Sobs 335.0; Chao1 343.125; ACE 347.31; Shannon H 4.148; Simpson D 0.0363. • Normal: Sobs 301.0; Chao1 331.03; ACE 331.23; Shannon H 3.701; Simpson D 0.0875. Species richness and diversity were highest in microecosystem, closely matching healthy controls and exceeding antibiotics and diarrhea. - Phylum-level composition (Figure S2; Fig. 3A): • Diarrhea: ~63.32% Bacteroidetes; 17.04% Fusobacteria; 14.27% Firmicutes. • Antibiotics: 61.19% Bacteroidetes; 30.91% Firmicutes; 7.08% Proteobacteria. • Healthy and Microecosystem: Firmicutes 46.73–50.10%; Bacteroidetes 32.77–39.27%; plus Tenericutes 5.62–8.06% and Spirochaetes 1.10–1.68%. Microecosystem had 2.72% Verrucomicrobia; healthy had 9.27% Actinobacteria. Comparative tests: Ill vs healthy showed decreased Firmicutes and Actinobacteria and increased Bacteroidetes and Fusobacteria in ill (p<0.01). Antibiotics vs microecosystem: antibiotics associated with decreased Firmicutes (p<0.05), Tenericutes (p<0.01), Verrucomicrobia (p<0.05) and increased Bacteroidetes (p<0.05), Proteobacteria (p<0.01). - Class/order/family-level patterns (Fig. 3B–D; Fig. 4B–D): Illness associated with high Fusobacteria, Gammaproteobacteria, and reduced Clostridia; antibiotics failed to normalize Clostridia/Mollicutes and increased Gammaproteobacteria. Healthy/microecosystem maintained higher Clostridiales and balanced Bacteroidales; diarrhea/antibiotics showed low Clostridiales and high Bacteroidales. Families characteristic of ill/antibiotics included Fusobacteriaceae, Enterobacteriaceae, Odoribacteraceae, Paraprevotellaceae; microecosystem/healthy had higher Ruminococcaceae, S24-7, Prevotellaceae (balanced), Pirellulaceae, and others. - Genus/species-level (Fig. 3E–F; Fig. 4E–F): Diarrhea dominated by Prevotella, Odoribacter, Fusobacterium, Escherichia, Treponema, with species Ruminococcus gnavus, Prevotella stercorea, and E. coli elevated. Antibiotics reduced Fusobacteria but left high Escherichia and Prevotella. Microecosystem increased overall genus diversity, associated with beneficial genera (e.g., Bilophila) and restored diversity. Species-level: Microecosystem eliminated or markedly reduced E. coli and P. stercorea and increased Lactobacillus reuteri and Bacillus cereus; antibiotics increased Sharpea azabuensis and P. copri but did not reduce E. coli. - Beta diversity: NMDS (Bray–Curtis) showed clear separation of healthy/microecosystem (Group A) from diarrhea/antibiotics (Group B). Alpha-diversity indices (Chao1, ACE, Shannon, Simpson) significantly higher in Group A than Group B. - Diagnostic profile: Illness associated with low Firmicutes/high Bacteroidetes, high Fusobacteria/Proteobacteria, and specific taxa (e.g., E. coli, P. stercorea); Bacillus treatment reversed these patterns toward healthy baselines, antibiotics largely did not.

The study addressed whether a defined Bacillus probiotic can both cure piglet diarrhea and restore gut eubiosis better than antibiotics. Across multiple diversity metrics and taxonomic levels, Bacillus treatment shifted the fecal microbiota of ill piglets to closely resemble healthy controls, with higher richness/evenness and restoration of beneficial phyla/classes (Firmicutes, Clostridia, Tenericutes, Spirochaetes, Verrucomicrobia). Conversely, antibiotics resolved clinical symptoms but left microbiota profiles similar to illness, including persistently high Bacteroidetes/Proteobacteria, low Clostridia, and elevated Escherichia/Prevotella. Importantly, Bacillus treatment markedly reduced pathogenic/negative taxa (E. coli, P. stercorea) and elevated beneficial species (Lactobacillus reuteri), supporting a mechanistic link between microbiota normalization and sustained clinical recovery. The findings support using probiotic Bacillus consortia as effective alternatives to antibiotics for managing piglet diarrhea and maintaining a healthy gut ecosystem, with implications for animal health, productivity, and antibiotic stewardship.

A three-strain Bacillus probiotic (B. subtilis Y-15, B. amyloliquefaciens DN6502, B. licheniformis SDZD02) effectively cured piglet diarrhea and, unlike antibiotics, restored the gut microbiome to a healthy, diverse state closely matching controls. Microbiome analyses (OTUs, alpha/beta diversity, taxonomic profiles, heatmaps, phylogeny) consistently clustered microecosystem-treated piglets with healthy animals and antibiotics with ill-diarrhea. Key detrimental taxa (E. coli, Prevotella stercorea) decreased with Bacillus, while beneficial taxa (e.g., Lactobacillus reuteri) increased. These results advocate for Bacillus-based probiotics as natural, sustainable alternatives to antibiotics in industrial swine production. Future research should extend to additional livestock species, assess long-term and multi-timepoint effects, refine probiotic consortia (potentially adding other beneficial taxa identified), and evaluate translational relevance in controlled clinical trials.

- Single post-treatment sampling timepoint (day 7) for microbiome sequencing limits temporal dynamics assessment; rarefaction and follow-up were reported, but longitudinal microbiome profiling beyond day 7 was not performed. - Natural infection without controlled pathogen challenge may introduce variability in etiologic agents. - 16S rRNA gene sequencing (Greengenes/RDP classification) provides relative abundances with limited strain-level resolution and no functional metagenomics; absolute abundances were not measured. - Antibiotic regimen tested was limited to colistin plus kitasamycin; findings may not generalize to other antibiotics or dosing regimens. - Sequencing was performed on a subset of piglets (ten per group), potentially limiting statistical power at finer taxonomic levels.