Characterization of boar semen microbiome and association with sperm quality parameters.

Elevated levels of bacteria within fresh extended boar semen are associated with decreased sperm longevity, therefore reducing the fertility of a semen dose. The objective of this study was to characterize the bacterial communities using 16S rRNA sequencing in freshly extended boar semen samples and relate the prevalence and diversity of the microbial population to sperm quality parameters 1) between studs, 2) between pooled and single-sire doses, and 3) over a 5-day period. Eight single-sire (n = 4 per stud) and eight pooled (n = 4 per stud) non-frozen extended semen doses were obtained from two boar studs (A and B). Pooled doses were the composite of the boar's ejaculates used in single-sire doses. Doses were subsampled for 5 d post-collection. Ten negative controls of each pooled dose (n = 2) and single-sire dose (n = 8) remained sealed until the last day. Microbiome analysis was achieved by examining the V4 hypervariable region of the 16S rRNA gene of flash-frozen samples. Two evaluators determined the average sperm motility and agglutination (0: no adhesion to 3: >50% adhesion) by averaging their estimates together at 10 random locations per slide. Stud A had greater sperm agglutination (1.6 vs. 1.0 ± 0.1; P < 0.01) than stud B. Sperm motility decreased over the 5-day period (P < 0.01) and tended (P = 0.09) to be greater in stud B than A (67.4% vs. 61.5% ± 0.02%). Compared with stud A, stud B had a greater relative abundance of Proteobacteria (60.0% vs. 47.2% ± 1.5%; P < 0.01) and a lower relative abundance of Firmicutes (22.5% vs. 31.9% ± 1.4%; P < 0.01). Moreover, stud A had a greater relative abundance of Bacteroidetes (6.3% vs. 5.3% ± 0.4%; P < 0.01) and Actinobacteria (11.5% vs. 10.1% ± 0.5%; P = 0.05) than stud B. Differences were found in alpha diversity for both Chao1 (P < 0.01) and Shannon (P < 0.01) diversity indexes among days 2, 3, 4, and 5 post-collection to day 1. For beta diversity, unweighted UniFrac metric on days 2, 3, 4, and 5 post-collection differed from those on day 1 (P < 0.01). There were significant correlations between sperm motility and relative abundance of Prevotella (r = -0.29), Ruminococcus (r = -0.24), and Bacteroides (r = -0.32). Additionally, there were significant correlations between sperm motility and Chao1 (r = -0.50) and Shannon's index (r = -0.36). These results demonstrate that differences in bacterial communities over time and between boar studs can be associated with variation in sperm quality.