Sperm Microbiota and Its Impact on Semen Parameters

Abstract

Compared to its female counterpart, the microbiota of the male genital tract has not been studied extensively. With this study, we aimed to evaluate the bacterial composition of seminal fluid and its impact on sperm parameters. We hypothesized that a dysbiotic microbiota composition may have an influence on sperm quality. Semen samples of 26 men with normal spermiogram and 68 men with at least one abnormal spermiogram parameter were included in the study. Samples were stratified based on total sperm count, spermatozoa concentration, progressive motility, total motility and spermatozoa morphology. Microbiota profiling was performed using 16S rRNA gene amplicons sequencing and total bacterial load was determined using a panbacterial quantitative PCR. Semen samples broadly clustered into three microbiota profiles: Prevotella-enriched, Lactobacillus-enriched, and polymicrobial. Prevotella-enriched samples had the highest bacterial load (p < 0.05). Network analysis identified three main co-occurrence modules, among which two contained bacteria commonly found in the vaginal flora. Genera from the same module displayed similar oxygen requirements, arguing for the presence of different ecological niches for bacteria that colonize semen through the passage. Contrary to our hypothesis, shifts in overall microbiota composition (beta-diversity) did not correlate with spermiogram parameters. Similarly, we did not find any difference in microbial richness or diversity (alpha-diversity). Differential abundance testing, however, revealed three specific genera that were significantly enriched or depleted in some of the sperm quality groups (p < 0.05). Prevotella relative abundance was increased in samples with defective sperm motility while Staphylococcus was increased in the corresponding control group. In addition, we observed an increased relative abundance of Lactobacillus in samples with normal sperm morphology. Our study indicates that overall bacterial content of sperm might not play a major role in male infertility. Although no major shifts in microbiota composition or diversity were found, the differential abundance of specific bacterial genera in the sperm suggests that a small subset of microbes might impact the spermatozoal physiology during sperm transition, more specifically motility and morphology. Further studies are required to challenge this finding and develop potential strategies to induce the formation of a healthy seminal microbiota.

Keywords: Lactobacillus; Prevotella; infertility; microbiota; spermatozoa.

Figures

FIGURE 1

Schematic representation of sample stratification. In addition to the spermiogram status, five categories, including total sperm count, sperm concentration, total motility, progressive motility and morphology were analyzed independently by comparing normal and abnormal values. For sperm concentration and progressive motility categories, abnormal values were further separated into two classes (abnormal and severe). Lower reference limits, defined by the WHO, are shown for each category.

FIGURE 2

Characterization of semen microbiota profiles. (A) Principal Coordinate Analysis (PCoA) score plot on the Bray-Curtis distance at genus taxonomic level with each dot representing an individual patient explaining 19% (x axis) and 12% (y axis) of variance, respectively. Colors indicate microbiota profile, as defined by unbiased Partitioning Around Medoid clustering (PAM) clustering with dashed line representing connection to the cluster centroid. (B) Boxplot comparing the richness of microbiota profiles, as measured by chao1 index. (C) Boxplot comparing the diversity of microbiota profiles, as measured by chao1 index. (D) Boxplot comparing the total bacterial load of microbiota profiles, as determined by Pan 16S qPCR. Each dot represents an individual patient with mean boxplot indicating the mean plus or minus SD. Statistics represents the result of post hoc one-tailed Wilcoxon rank sum test.

FIGURE 3

Interaction network of the most abundant genera (> 1% mean relative abundance) using SparCC algorithm. Connecting edges represent significant interactions (one-sided p-value < 0.05) with thickness proportional to SparCC co-occurrence values. Nodes are sized according to mean relative abundance of the corresponding genus in the data set. Members of interaction module 1, 2, and 3 are highlighted in red, green and yellow, respectively.

FIGURE 4

Differentially abundant bacterial genera with semen parameters. (A) Heatmap of the most abundant genera (> 1% mean relative abundance) with corresponding Ward linkage dendrogram based on the Bray Curtis distance matrix at genus level. Each column represents an individual sample and color-scale bar indicating operational taxonomic unit (OTU) relative abundance. Samples grouping according to spermiogram and microbiota profile (1-Prevotella-enriched, 2-Lactobacillus-enriched and 3-polymicrobial) are indicated as column annotations. OTUs are labeled based on their phylum (p), class (c), order (o), family (f), and genus (g). (B) Bar plot representation of differentially abundant bacterial genera as determined by linear discriminant analysis effect size (LEfSe) analysis comparing spermiogram groups. (C) Same analysis comparing total motility groups. (D) Same analysis comparing morphology groups. Each bar is colored according to its belonging to normal (green) or abnormal (red) group.