Revealing mechanisms of mating plug function under sexual selection

Significance Promiscuous mating by females leads to competition between males for fertilization success. When fertilization is internal, this means that rival males’ sperm must compete within the female reproductive tract to reach the eggs. Males of diverse species deposit a mating plug during copulation, which is hypothesized to assist in the race for fertilization following multiple mating. Here, we tested this by using stable isotope labeling to discriminate the ejaculates of competing male voles in direct competition. This revealed that the mating plug simultaneously inhibits the sperm of rival males while promoting transport of a male’s own sperm, both of which are beneficial in the competition for fertilizations.

. Recovered mating plugs. Plugs were recovered from female bank voles immediately after sequential copulation with two males, during which each male ejaculated once.
Images of these plugs are numbered 1 to 17. Males mating in first or second mating roles had been fed either 'heavy' (H) or 'light' (L) diets to facilitate discrimination of their ejaculates, with the 'heavy' diet containing a stable isotope labelled amino acid ([ 13 C 6 ] lysine) at a relative isotope abundance of 0.5. For each plug, the identity and labelling status of the first male to mate is presented first in square brackets, with H1-4 indicating the identity of a 'heavy' first male, and L1-L14 indicating the identity of a 'light' first male. The identity and labelling status of the second male is then presented, using the same abbreviations. Mating plugs were cut to analyse the outer two quartiles (1-4, 8,9, 12-17), or if plugs split into two naturally, each part was cut in half (6,7,11, labelled as parts 1 and 2).
A further two plugs were also cut in half (5,10 -see main text). The ratio of H/(H+L) was calculated based on the area of a potential SVS4 tryptic peptide SASGSSTSYSLDK (see main text). In most cases (2-5, 8-10, 12-17), proteomic analysis revealed that the recovered plug material originated entirely from the second male to mate: when the second male was 'light' (coloured blue in the figure) the plug contained no [ 13 C 6 ] lysine (H/(H+L)<0.04), and when the second male was 'heavy' (coloured red in the figure) the plug contained half (50% [ 13 C 6 ] lysine (H/(H+L)>0.47), consistent with full labelling. In four cases (1,6,7,11) proteomic analysis revealed that recovered plug material originated from both the first and second males to mate (coloured purple in the figure). Plugs 6,7 and 11 each split naturally into two on recovery. In these cases the section of the plug positioned closest to the cervix (part 1) contained protein from the first (heavy) male combined with protein from the second (light) male, whereas the plug section positioned behind this (part 2) contained protein only from the second (light) male. Plug 1 did not split naturally on recovery but contained protein from both the second (heavy) and first (light) male.  Table S1. B. The mass of the mating plug that males produced explained significant variation in the number of their own sperm recovered from the uterus immediately after ejaculation. More sperm were recovered for males producing larger plugs, when mating in either a first mating role (as the only male) or second mating role (in a double mating). For statistical analysis, see Table 2.  Table S4. B. Following double copulations, the number of the second males' sperm reaching the uterus was predicted by the mass of their seminal vesicles, and hence by inference, by the size of their mating plug. Larger seminal vesicles of the second male were associated with more of their sperm reaching the uterus. For statistical analysis with control for second males' cauda sperm count (as measure of sperm production) see Table 3A.  S5. Sperm competition and paternity outcomes. Following double copulations, the relative proportion of sperm from two competing males in the uterus explained significant variation in paternity outcomes when the same males mated again with a different female.
S2 is the proportion of sperm present in the uterus from the second male immediately after a double copulation, and P2 is the proportion of offspring sired by the second male in a subsequent double copulation. Linear regression analysis: r 2 = 0.56, F 1,7 = 8.9, P= 0.02. males consumed a diet containing 50% [ 13 C 6 ] lysine, and 'light' (L) males consumed a control diet with [ 12 C 6 ] lysine for a minimum of 40 days prior to experimental mating. To investigate the consequences of natural variation in mating plug characteristics, a series of double and single copulations were conducted using differentially labelled males. A total of 60 copulations were achieved with 40 females, 20 of which mated with a single male (4H and 16L), and 20 of which mated sequentially with two differentially labelled males (16HL, 4LH). Immediately after each single or double copulation, female voles were humanely killed to recover ejaculates and mating plugs. Mating plugs were removed from the vagina, the contents of the uterus were removed, sperm counts were performed and the ejaculated seminal fluid and sperm were separated for subsequent proteomic analysis. Ejaculates were collected from unlabelled or 'heavy' labelled males, following a single copulation with a previously unmated female, and analysed by proteomics. The raw data files were interrogated to display the mass spectrum of a single peptide (from seminal vesicle protein IV (SVS4 peptide SASGSSTSYSLDK, m/z 645.296 light, m/z 648.306, heavy) to illustrate the differences in labelling profiles. Samples are from 4 'heavy' labelled males (H1-H4), and 14 'light' males (L1-L14). Note the absence of any peak corresponding to the heavy peptide in the light labelled samples (L1 to L14).

Filename: Data files_mating plugs.xlsx
This multi-tab spreadsheet contains raw data used for each of the analyses. Tab 1 is the index relating Tabs 2-15 to the location of results for each dataset, with a description of the analyses used, including number of observations, dependent variables, factors and random effects in models.

Filename: All plug proteome data.xlsx
This multi-tab spreadsheet is the output of Proteome Discoverer (Thermo). Tab 1 is the index relating individual sample IDs to proteomics file numbers. Tab 2 is the overall proteome abundance for the protein profile, heavily dominated by SVS4. Finally, a standard Proteome Discoverer analysis is present in tab 3.

Filename: BankVoleProteinDB_fasta.txt
This database was constructed by reference to other rodent protein databases, as described in the text. Where proteins were manually searched and annotated, these are indicated by short accession numbers (e.g. 'BV4', the most abundant protein in plugs, equivalent to SVS4 in other rodents).

Figure S1.
List of additional peptides used in Figure S1, with the calculated RIA values:        A) Copulation duration (from first intromission to ejaculation) was significantly longer for the second male to mate; B) the second male to mate performed significantly more intromissions per copulation, but C) the intromission rate (number of intromissions divided by copulation duration) and D) copulation latency (from introduction to initiation of copulation) were not significantly different for the first and second copulation. Results are shown for linear mixed models fit by REML, using data for 60 copulations (40 with males mating first or alone, and 20 with males mating second in a double copulation), by 24 males with 40 females. The analysis includes male and female ID as random factors. Copulation duration is log transformed. Intromission rate and copulation latency are square root transformed.  with an adjusted significance level of 0.001. Second, remaining proteins were individually examined to check that they were likely to be found in sperm. The remaining 21 proteins were used to calculate sperm numbers from competing males (see Table S7 for a list of peptides used).