Evaluating differences in uterine microbiome and inflammatory status at 1 month postpartum associated with metritis and antibiotic treatment.

Ceftiofur treatment in cows with metritis enhances the likelihood of clinical cure but does not consistently improve subsequent pregnancy rates compared with healthy (nonmetritis) cows. We tested the hypothesis that ceftiofur treatment has no effect on uterine microbiota or endometrial inflammation at 1 mo postpartum in cows that were either healthy or diagnosed with metritis at 5 to 10 d postpartum. Cows diagnosed with metritis were matched with healthy cows in a 2 × 2 factorial design where they were either treated with an antibiotic (ceftiofur hydrochloride) or left untreated. Cows were slaughtered to collect the uterus at 1 mo postpartum, and the uterine microbiota was measured using bacteriology and 16S rRNA gene sequencing. The lumen of the uterus was examined visually for the presence of purulent material, and inflammation within the endometrium was assessed histologically. When bacteria were cultured from the lumen of the uterus at 1 mo postpartum, a greater number of species was isolated and a greater number of colonies formed from cows that were previously diagnosed with metritis compared with healthy cows. Antibiotic treatment administered at the time of disease diagnosis (5 to 10 d postpartum) had no effects on the abundance of culturable bacteria at 1 mo postpartum in healthy cows but reduced the number of culturable bacteria from the uterus of cows previously diagnosed with metritis. Based on 16S rRNA gene sequencing, cows originally diagnosed with metritis had an increased number of sequencing reads 1 mo postpartum than cows initially deemed healthy. Furthermore, antimicrobial treatment 5 to 10 d postpartum decreased the number of sequencing reads 1 mo postpartum irrespective of the initial disease diagnosis. The presence of purulent or metritis-like discharge in the uterine lumen 1 mo postpartum had a major effect on the uterine microbiota. Histopathological analysis revealed that the presence of purulent material in the uterus or signs of acute infection at 1 mo postpartum was associated with greater uterine inflammation compared with a clear uterine flush but minimally affected by treatment. Furthermore, antibiotic treatment had no main effect on α-diversity (Faith's phylogenetic diversity or Pielou's evenness), and no differentially abundant taxa were detected by analysis of composition of microbes. However, β-diversity analyses (unweighted UniFrac) showed a small but significant effect of treatment, with treated cows having lower within-group variability than untreated cows, suggesting a modest homogenizing effect on the uterine microbiota. A multinomial regression analysis revealed a marked reduction (145-fold) in the relative abundance of Fusobacterium in treated cows. This shift suggests a partial restoration of the microbial community in cows with metritis toward a profile resembling that of their healthy counterparts. Additionally, the analysis indicated that the uterine microbiota at 1 mo postpartum was collectively influenced by visual detection of purulent material, occurrence of metritis postpartum, antibiotic treatment, and estrus cyclicity. We rejected the hypothesis that early postpartum ceftiofur treatment did not affect uterine microbiota at 1 mo postpartum and concluded that ceftiofur treatment had a measurable effect on uterine microbiota and inflammation.