Heterogeneity in semen-mediated tolerogenic responses: Defining immunological determinants of reproductive compatibility

Abstract STUDY QUESTION Do particular pairs of individuals modeling sexual pairs exhibit differences in the induction of semen-mediated long-lasting tolerance in support of pregnancy? SUMMARY ANSWER Yes, factors from both sexes contribute to the variability in the induction of long-lived regulatory T cells (Tregs) in response to semen, with factors related to the recipient seeming to slightly outweigh the factors related to the semen donor. WHAT IS KNOWN ALREADY Semen contains multiple immunomodulatory factors that can influence reproductive tract immunity in the recipient, promote tolerance to antigens from the semen donor (designated henceforth as “paternal”), and ultimately support successful pregnancy. These factors include the soluble fraction containing prostaglandins and cytokines, as well as a very high concentration of extracellular vesicles (EVs). Both soluble and vesicular fractions of semen have been shown to promote tolerogenic phenotypes and migration in antigen-presenting cells (APCs), and to induce Tregs. This study was undertaken to understand the variation in tolerance induction between pairs of individuals modeling sexually active couples. STUDY DESIGN, SIZE, DURATION We designed an in vitro pairing study using random semen donors (N = 11) and peripheral mononuclear cells (PBMCs) from random self-identified female recipients (N = 3). Semen was fractionated into vesicle-enriched (semen extracellular vesicles; SEV) and soluble (vesicle-depleted semen plasma; VDSP) fractions. Monocyte-derived dendritic cells (MoDCs) from each of the recipients were exposed to semen fractions from each donor, then cocultured with autologous naïve T cells to assess tolerogenic responses, specifically, generation of Tregs (FOXP3+ CD25Hi CD127Lo CD4+ T cells) and their phenotypic differences (TIGIT+, PD-1+ CTLA-4+, or AREG+ Tregs). These cells were then screened for T cell receptor-dependent activation by activation-induced marker (AIM) expression assay. These responses were compared across all pairs of semen donors and recipient cells. In addition, semen fractions were screened for candidate mediators of tolerance, such as human leukocyte antigen G (HLA-G) and adenosine-generating enzymes. PARTICIPANTS/MATERIALS, SETTING, METHODS This study used an in vitro human cell culture approach in which primary blood-derived immune cells were exposed to semen-derived vesicular and non-vesicular fractions. Blood samples were obtained from healthy self-identified female donors, and semen samples were obtained from healthy participants who provided initial screening specimens for a contraceptive study. Semen-derived vesicular and non-vesicular fractions were characterized by western blot analysis for markers associated with immune tolerance. Primary immune cells were exposed to these semen fractions, and the major endpoints included changes in Treg phenotypes and subset frequencies, as well as activation-induced marker expression. These outcomes were assessed by flow cytometry. No animal models were used. MAIN RESULTS AND THE ROLE OF CHANCE HLA-G content in semen (SEV and VDSP) varied markedly across donors, ranging from undetectable to high concentrations. Similarly, semen’s capacity to generate adenosine from ATP differed by approximately fourfold among donors. Induction of Tregs by SEV and VDSP from eleven semen donors was highly variable, from 2.8% to 63.4% of CD4+ T cell population. Semen fractions from a few donors consistently elicited lower tolerogenic responses across all recipients. However, when we considered the combined effects of both semen fractions, none of the semen donors failed to elicit a measurable tolerogenic response. On the other hand, one recipient was a consistently low responder to tolerance induction. Thus, both recipient immune cells’ features and semen-intrinsic properties are significant determinants of tolerogenic responsiveness. Semen re-exposure enhanced Treg AIM expression. The strongest effect was observed when Tregs were exposed to VDSP-loaded MoDCs. In contrast, SEV-loaded MoDCs caused a much weaker response. LARGE SCALE DATA N/A LIMITATIONS, REASONS FOR CAUTION One limitation of our study is that we used semen and circulating recipient cells from random donors not selected for any clinical characteristics, such as history of infertility. In addition, we did not correlate HLA-G content and adenosine-generating enzyme expression levels with the Treg and AIM readouts. We show, via Treg activation, that semen induces tolerance to its antigens, but we do not demonstrate which specific antigens are recognized. We posit that the variation in semen-induced tolerance observed between couples plays a role in reproductive outcomes; however, larger studies in clinically characterized cohorts and using cells from the tissue microenvironment will be required to establish this link. WIDER IMPLICATIONS OF THE FINDINGS Our findings demonstrate that semen-induced tolerance is highly variable and may be more strongly shaped by recipient factors than by semen donor factors. In addition, the generation of Tregs with enhanced activation to previously encountered semen supports the concepts of trained immunity and antigen-specific tolerance that arise cumulatively following exposure to semen from the partner. These Tregs are poised to expand and support pregnancies with fetuses expressing paternal alloantigens. FUNDING This work was supported by the following grants: NIH/NICHD R01AI153342 to LV, NIH/NIDA R01DA040386 to FH, NIH/NCATS KL2TR002317 to GG. DISCLOSURES The authors have no conflicts of interest to declare.