Females translate male mRNA transferred during mating

Abstract

Although RNA is found in the seminal fluid of diverse organisms, it is unknown whether it is functional within females. We developed a proteomic method (VESPA, Variant Enabled SILAC Proteomic Analysis) to test the hypothesis that Drosophila male seminal fluid RNA is translated by females. We found 67 male-derived, female-translated proteins (mdFTPs) in female lower reproductive tracts, many with predicted functions relevant to reproduction. Knockout experiments indicate that mdFTPs play diverse roles in postmating interactions, affecting fertilization success, and the formation/persistence of the insemination reaction mass, a trait hypothesized to be involved in sexual conflict. These findings advance our understanding of reproduction by revealing a mechanism of postmating molecular interactions between the sexes that strengthens and extends male influences on reproduction in previously unrecognized ways. Given the diverse species that carry RNA in seminal fluid, this discovery has broad significance for understanding molecular mechanisms of cooperation and conflict during reproduction.

Keywords: Health sciences; Immunity; Machine learning; Mathematical biosciences; Virology.

Graphical abstract

Figure 1

Overview of VESPA VESPA differentiates proteins made by males and females by isotopic labeling and the source of RNA transcripts by fixed nucleotide (nuc) differences between species. After mating, female reproductive tracts are removed, proteins are extracted and digested, and samples are analyzed by liquid chromatography-tandem mass spectrometry. Peptides that match the amino acid (aa) sequence of the male but carry the heavy label of the female are diagnostic for mdFTPs.

Figure 2

mdFTPs are supported by multiple lines of evidence (A) All mdFTPs were identified by a minimum of two heavy peptides, with at least one being diagnostic. Figure indicates the number of genes per mdFTP type, and the filled boxes indicate the properties of each type. The most strongly supported mdFTPs include those identified by multiple diagnostic peptides, in two or more replicates, and/or by both MaxQuant and MSFragger. (B) Mean RNA transfer index (RTI) per gene was higher for high-support mdFTPs [yellow in (A), n = 67] and low-support mdFTPs (n = 99) compared to all other genes with RNA-seq data (n = 9727). (GLMM: gene category χ² = 812.8, p = 2.2e⁻¹⁶). Post hoc tests were performed using Tukey’s method (∗∗∗∗p ≤ 0.0001). Error bars represent standard error of the mean. (C) Venn diagram showing overlap of mdFTPs with D. arizonae SFPs. Low overlap suggests mdFTPs may perform different functions from other proteins in the ejaculate or were not detected by our methods.

Figure 3

mdFTPs have functional significance to reproduction (A) Enrichment of protein domains and GO terms for biological process and molecular function link mdFTPs to processes important for reproduction. This includes enrichment of CAP domains, which are linked to reproduction in diverse organisms, and two domains (serpin family and terpenoid cyclases/protein prenyltransferase alpha-alpha toroid) that have predicted involvement in coagulation. GO terms associated with energy production, oxidative stress response, and immunity were also enriched. Level of enrichment is indicated on the x axis and differentiated by color. (B) List of mdFTPs with protein domains associated with coagulation. Additional domains not necessarily linked to coagulation are also noted. High-support mdFTPs refer to the 67 in yellow from Figure 2A, whereas low support mdFTPs represent the remaining 99.

Figure 4

Gene KO experiments demonstrate mdFTPs have functional effects on diverse reproductive processes (A) D. arizonae females mated to D. arizonae males with the ARI26694 KO mutation had a smaller reaction mass initially, as measured by the average perimeter. However, it degraded more slowly compared to the reaction mass in females mated to WT males (two-way ANOVA: genotype × time interaction, F = 16.5, p = 0.0001). Post hoc comparisons were performed using Tukey’s method (∗∗p ≤ 0.01). KO-0 h: n = 26; WT-0 h: n = 35; KO-6 h: n = 24; WT-6 h: n = 15. Error bars represent 95% confidence intervals of estimated marginal means. (B) D. arizonae females mated to D. arizonae males with the ARI/11629 KO mutation had a larger reaction mass than females mated to WT males (two-way ANOVA: genotype, F = 12.0, p = 0.0007). KO-0 h: n = 42; WT-0 h: n = 41; KO-6 h: n = 41; WT-6 h: n = 50. The dashed line indicates the mean perimeter of unmated female lower reproductive tract. Error bars represent 95% confidence intervals of estimated marginal means. (C) D. arizonae females mated to D. arizonae males with a ARI11629 KO mutation laid more unfertilized eggs on days one and three postmating compared to females mated to WT males (GLM: genotype × day interaction, χ² = 17.2, p = 0.0002). Post hoc comparisons were performed using Tukey’s method (∗∗∗p > 0.001). Error bars represent 95% confidence intervals of estimated marginal means. WT-D1: n = 119; KO-D1: n = 44; WT-D3: n = 104; KO-D3 n = 360; WT-D5 n = 131; KO-D5 n = 90.