The ability to subvert independent assortment of chromosomes is found in many meiotic drivers, such as the t haplotype in house mice Mus musculus, in which the t-bearing chromosomal homolog is preferentially transmitted to offspring. This is explained by a poison-antidote system, in which developing + and t sperm in testes of + /t males are exposed to ‘poison’ coded by t loci, from which t sperm are protected, allowing t sperm an overwhelming fertilisation advantage in monogamous matings. This system is thought to result in poorly and normally motile sperm subpopulations within + /t sperm, leaving t sperm unharmed. Conversely, we found that the fastest quartile of sperm from + /t males swam more slowly, both forwards and along their travel path, and had reduced straightness and linearity, compared to the fastest quartile of + / + sperm. Moreover, sperm from + /t males had shorter tails and narrower heads than + / + sperm, and these morphological differences covaried with motility differences. Finally, + /t traits did not show evidence of bimodal distributions. We conclude that the t haplotype drive results in lasting damage to the motility of both + and t developing sperm, although previous studies indicate that + must be more harmed than t sperm. This damage to all sperm may explain the low success of + /t males in sperm competition with + / + males, seen in earlier studies. We propose that the harm the t causes to itself could be termed ‘spiteful’, which may also be common to other gamete-harming meiotic drive systems.

Please see the Methods section of the article for complete information.

House mice Mus musculus domesticus for this study were obtained from the experiment of Runge & Lindholm (2021, https://doi.org/10.1098/rsos.202050). Following euthanasia, the cauda epididymis from 34 +/t and 46 +/+ males were placed in warm mHTF medium, opened to release sperm, and 10 minutes later removed. After 2 hours of incubation at 37°C, a sample of sperm was placed in a Leja4 slide and sperm motility was analysed with a MouseTraxx unit (Hamilton-Thorne). Sperm tracks were captured by recording 30 frames at a 60 Hz with minimum contrast set to 35 and minimum cell size at 5 pixels. The CASA velocity results were only used for analysis if at least 200 sperm were counted and 50 of those were classified as rapid.

We investigated whether the acrosome reaction had taken place after 2 hours of incubation by smearing a 20µl sperm sample on a slide and air drying it. The slides were stained using freshly made CBB stain (40ml H₂O, 50ml methanol, 10ml acetic acid and 220mg Coomassie blue) by incubating in a bath of CBB for 4 minutes and careful rinsing afterwards. After slides were air dried (min. 4 hours) samples were sealed with DPX mountant for histology (SIGMA) and 24 x 60 mm cover slips. Afterwards pictures were taken at a magnification of 400x with a Leica EC4 using the LAS EZ program (50% brightness, 0,7 gamma and 105 saturation). At least 100 and a mean of 229 sperm were pictured for each sample. The presence or absence of the acrosome was determined by eye.

We investigated sperm morphology by measuring head, midpiece, and tail lengths by hand in ‘ImageJ’ using the segmented line tool. The images were based on sperm stained to visualise acrosome reaction (above). Lengths were calibrated using a pictured micrometre (Neubauer/Scherf Präzision). We measured head length from the base of the head to the tip and head width at the widest part of the head (both excluding the sperm hook). The same calibration factor was used throughout all measurements. We measured a total of 1000 sperm from 50 individuals, 25 +/t and 25 +/+. 

Outliers of more than 10 times the standard deviation in sperm motility traits were excluded from the analyses (<0.1% of data points per trait). In the sperm morphology analyses, sperm that deviated more than 3 times the standard deviation from the mean were removed as outliers. 14, 16, 11 and 1 outliers were removed from midpiece length, tail length, head length and head width measures respectively. Removed outliers did not qualitatively alter the results.

Please see "README_Winkler_&_Lindholm_2022_Data.txt"