Semen adaptation to microbes in an insect

Abstract

Sperm function is suggested to evolve by sexual selection but is also reduced by microbial damage. Here, we provide experimental evidence that male fertility can adapt to microbes. We found that in vivo, male fertility was reduced by one-fifth if sperm encountered microbes in the females that they had not previously been exposed to, compared to sperm from males that coevolved with these microbes. The female immune system activation reduced male fertility by an additional 13 percentage points. For noncoevolved males, fertility was larger if microbes were injected into females after they had stored away the sperm, indicating microbial protection as a previously unrecognized benefit of female sperm storage. Both medical and evolutionary research on reproductive health and fertility will benefit from considering our findings that the impact of microbes on sperm depends on their joint evolutionary history. Our results may assist in reconciling contradictory results of sexually transmitted disease effects on sperm and bring empirical realism to a recently proposed role of locally adapted reproductive microbiomes to speciation.

Keywords: coevolution; fecundity; host–parasite interactions; speciation.

Figure 1.

Simplified version of the experimental design used to disentangle female effects caused by microbe effects on sperm vs. immune system effects on sperm, as well as separating female sperm from microbe-sperm co-exposure. Untreated control (n = 33) and wounding control (n = 67) females are not presented graphically here. Note that immune system activation includes a separate treatment level, wounding only, subsumed here as an injection without live microbes (white arrows with gray stars at the tip representing dead microbes). A unique advantage of our study system is that fertility effects that arise from female resource allocation after infection can be controlled experimentally: An experimental infection before sperm storage (6 hr after mating, “direct treatment”) includes microbial impacts on the sperm plus collateral immune damage plus fertility reduction due to female defense allocation costs. Experimentally challenging females after sperm had been moved to sperm storage organs (24 hr after mating, “indirect treatment”) separates sperm from microbes and immune effectors present in the female hemolymph and only causes female defense allocation effects on fertility. Therefore, fitness costs of a microbial attack on sperm before storage (“direct treatment”) consist of (i) direct microbial plus immune costs on sperm and (ii) female defense and resource allocation costs.

Figure 2.

Female lifetime fertility in relation to infection treatments and microbe exposure of sperm. (A) Number of fertile eggs for the control (black), mating with a foreign male (very light gray), wounding controls (light gray), female immune system activation (gray), and microbial toxicity (A sperm in blue/B sperm in red). Sperm had either been directly exposed to microbes (“6 hr after mating”—filled coloration) or after they were in the female sperm storage organ (“24 hr after mating”—white circles). (B) Female lifetime fertility relative to the fertility of control females in relation to infection treatments and microbe exposure of sperm. Together, wounding, immune system activation, and microbial toxicity reduced fertility by 10% for co-exposed and by 35% for B sperm. The bars show the proportional decrease of lifetime fertility per female for mating with a foreign male (white), wounding only (light gray), female immune system activation (gray), and microbial toxicity (blue/red). Sperm had either been directly exposed to microbes (“6h after mating”—filled coloration) or after they were in the female sperm storage organ (“24h after mating”—hatched coloration). A fitness cost of a microbial attack on sperm before storage (6 hr after mating,” “direct treatment”) consists of (i) direct microbial plus immune costs on sperm and (ii) female defense and resource allocation costs. A microbial attack of females after sperm had been moved to storage (24 hr after mating, “indirect treatment”) only causes fitness costs of (ii). Deducting costs of (ii) from the total fitness costs [(i) and (ii)] will therefore produce the fitness costs (i), i.e., those arising from the direct microbial, or immune, impact on sperm. The average reduction in fertility caused by the immune system was 14.3% (mean of hatched and filled gray bars). The effect of microbial toxicity is only visible for direct contact (filled bars) and is much stronger if microbes and sperm are not coadapted. On average, B sperm caused a 36% reduction in female fertility (mean of hatched and filled red bars) compared to the mean number of eggs laid by untreated control females mated to males of their own population. The number below the error bars represents the absolute mean difference in the number of eggs per female. Error bars represent one standard error.

Figure 3.

Proportion of females laying fertile eggs under control and microbial treatments. Females mating with their own males (control treatment): black solid line; females mating with a foreign male (dashed black line). Wounded control females (light gray line), control females immune-activated (IS indirect and direct) with dead microbes (gray lines), and the females injected with live microbes (A sperm blue and B sperm red lines). Females used sperm that were co-exposed (blue lines) or B sperm (red lines) to microbes used, and that experienced direct exposure (solid lines) or indirect (after storage) exposure (dashed lines). The probability of becoming infertile differed between treatments (Cox proportional hazards: Wald χ²_8,170 = 15.420, p = 0.05). Females using B sperm had a 3.73× higher risk of becoming infertile than control females when sperm were directly exposed to microbes (Cox proportional hazards: z = 3.294, p < 0.001) but not when sperm had been stored in the seminal vesicles (1.7×; Cox proportional hazards: z = 1.258, p = 0.21). Wounding increased this risk by a factor of 1.9 (Cox proportional hazards: z = 2.076, p = 0.04). Effects of mating with a foreign male, direct and indirect immune system activation, and fertilization with co-exposed sperm after directly or indirectly encountering microbes were all nonsignificant.