Rapid protamine evolution suppresses meiotic drive in Drosophila

Abstract

Many animal species replace histones with protamines during spermatogenesis. Despite their importance for sperm function, protamines rapidly evolve in many species; the biological causes behind their rapid evolution remain unknown. Here, using in vivo gene replacement, we investigated the causes and consequences underlying the rapid evolution of protamine Mst77F, which is essential for male fertility in D. melanogaster. Mst77F ortholog replacements led to defects in DNA compaction of X-chromosome-bearing sperm compared to Y-chromosome-bearing sperm during spermatogenesis, resulting in fewer X-bearing mature sperm and male-biased progeny. Unlike D. melanogaster, Mst77F is not essential for male fertility in D. yakuba but is still required to suppress sex-ratio distortion. Our results suggest that relentless pressure to suppress sex chromosomal meiotic drive drives the rapid evolution of protamines.

One-sentence summary:

A rapidly evolving essential protamine suppresses sex-chromosome meiotic drive in Drosophila

Most eukaryotes rely on histones and histone variants to package their DNA and regulate processes such as transcription, DNA repair, and chromosome segregation (1). However, various animal species have independently evolved distinct DNA-compaction proteins, collectively known as protamines, to facilitate the compact packaging of genomes inside sperm heads, reducing their volume by up to 100-fold compared to somatic cells (2–5). For example, mammalian protamines are likely derived from linker histones (6, 7), whereas Drosophila protamines (or sperm nuclear basic proteins, SNBPs) are derived from HMG-box encoding transcription factors (2, 8). Although protamines are absent in some animals, they are essential for male fertility in all studied species that encode them, including mammals and Drosophila(4).

Despite their critical roles in sperm genome packaging and male fertility, protamine genes are among the fastest-evolving genes across various animal species, such as mammals(9–11) and Drosophila species (2). Yet the driving forces behind their rapid evolution remain unclear. A prevailing hypothesis suggests that the rapid evolution of protamines results from their impact on the shape of sperm heads and their role in enhancing sperm success amid competition for fertilization (10, 12, 13). Indeed, protamine loss and mutations impair sperm head morphology and fertilization (14–16). However, there is no direct evidence that rapid protamine evolution is driven by selective advantage for successful fertilization (17). Moreover, the intensity of sperm competition is either poorly correlated or not correlated with protamine evolutionary rates (9, 10, 18, 19). Additionally, some studies have suggested that the rapid evolution of protamines in mammals results from hypermutation at their CpG-bearing arginine codons, followed by selection to restore optimal arginine content for protamine function (20). The biological causes underlying the rapid evolution of protamines remain experimentally unverified in any species.

Our recent phylogenomic analysis of protamine genes in Drosophila species, which encode at least 15 distinct protamine genes (21–24), proposed an alternative explanation for the rapid evolution of protamines (2). This analysis revealed extensive turnover of protamine genes associated with sex chromosome evolution, including multiple independent protamine gene duplications and amplification events on sex chromosomes across Drosophila species (2, 25–28). Such gene duplicates, including X-linked paralogs of ProtA/B and Y-linked paralogs of Mst77F (Mst77Y), could evolve roles as sperm killers – selfish genes that enhance their propagation at the expense of host fertility (25, 26, 29). Conversely, in the montium subgroup of Drosophila species, half of the ancestrally conserved protamine genes were lost coincident with an XŶ chromosomal fusion, which obviated sex chromosome competition (2). These findings suggest that rapid protamine evolution may be driven by the need to suppress post-meiotic competition or meiotic drive arising from conflicts between X- and Y-bearing sperm, thereby maintaining male fertility and preserving the Fisherian optimal 50:50 sex ratio in resulting progeny (30, 31).

Here, by using gene knockouts, ortholog replacements, and engineered sex chromosome fusions, we demonstrate that the rapid functional divergence of the Mst77F protamine impacts male fertility in D. melanogaster. We show that the correct sequence and dosage of Mst77F are necessary to suppress sex-ratio distortion caused by unequal maturation rates of sperm carrying X or Y chromosomes during spermatogenesis. In D. melanogaster, transgenic flies with reduced Mst77F dosage or orthologs from other species exhibit chromatin condensation defects in sperm bearing X chromosomes, resulting in their underrepresentation in mature sperm and ultimately leading to male-skewed progeny. Our results suggest that the need to suppress competitive interactions between sperm with different genotypes, such as X- and Y-bearing sperm, during spermatogenesis may be one of the key forces driving the rapid evolution of protamines.

Result

Mst77F protamine replacements impair normal male fertility in D. melanogaster

Using CRISPR/Cas9 and two guide RNAs (gRNAs), we generated a complete Mst77F knockout (Mst77F KO) by replacing all protein-coding sequences with an attP site, enabling subsequent site-specific ΦC31 recombinase-mediated insertion (Fig. 1A; Fig. S1A–B). Recapitulating earlier findings (21, 32–34), we observed that Mst77F KO homozygous males are entirely sterile, producing no mature sperm or offspring (Fig. 1B; Fig. S1C). Spermatids in Mst77F KO males still appear to undergo the histone-to-protamine transition, removing most histones and incorporating other protamines, e.g., ProtA/B (34). However, we observed that some spermatids have larger nuclei (Fig. S1D), consistent with Mst77F’s proposed role in condensing genomic DNA in sperm.

Fig. 1: Functional divergence of the Mst77F essential protamine across Drosophila species.

(A) Schematic of the CRISPR/Cas9-mediated knockout and ΦC31 integrase-mediated replacement of Mst77F in D. melanogaster. The endogenous Mst77F is located within an intron of Pka-R1. It was first replaced with a DsRed marker (not shown) and an attP site, generating male-sterile Mst77F knockouts (KO). Orthologous Mst77F transgenes from various Drosophila species were then inserted into the attP site to test their ability to rescue male fertility. A phylogeny shows when these Drosophila species diverged from D. melanogaster (in millions of years ago, MYA) and the % protein identity between each ortholog and Mst77F-mel. (B) Comparison of the fertility of transgenic homozygous D. melanogaster males carrying two copies of Mst77F from different species. Each dot represents the total number of adult offspring produced by a single male. (C) Comparison of the fertility of transgenic hemizygous D. melanogaster males encoding only a single copy of Mst77F from the designated Drosophila species. Significance was determined by unpaired Student’s t-tests (*p<0.05; ****p < 0.0001; ns, not significant).

To rescue the male sterility caused by the loss of Mst77F, we cloned and reintroduced the wild-type D. melanogaster cDNA copy of Mst77F (Mst77F-mel) using ΦC31-mediated integration at the endogenous locus (Fig. 1A). The construct included 500 bp of endogenous upstream sequence as the promoter and carried a C-terminal 3×FLAG epitope tag. As previously reported (21, 32–34), D. melanogaster males carrying either one or two copies of Mst77F-mel in the endogenous locus are fertile, with each male capable of producing more than 100 offspring (Fig. 1B; Fig S1C). This shows that loss of Mst77F causes male sterility and that the 3×FLAG tag does not interfere with its essential function in male fertility. Immunostaining with a FLAG antibody revealed that transgenic Mst77F is translated and incorporated into spermatid chromatin during the late canoe stage of spermatogenesis (Fig. S2), similar to endogenous Mst77F (34).

Mst77F is one of the youngest Drosophila protamine genes, likely originating in the common ancestor of D. melanogaster and D. ananassae ~ 20 million years ago (Fig. 1A) based on phylogenetic evidence (35). If the rapid evolution of Mst77F were tied to its roles in male fertility, we hypothesized that orthologs from various species might be maladapted for one or more of these roles in D. melanogaster, potentially revealing the selective pressures driving Mst77F’s rapid divergence. Therefore, we utilized the attP site introduced at the endogenous Mst77F locus to introduce an evolutionary allelic series of C-terminally 3×FLAG-tagged Mst77F orthologs from five divergent Drosophila species encoding proteins with increased amino acid divergence (Fig. 1A): D. simulans (2.5 mya, Mst77F-sim, 71% identity), D. yakuba (6 mya, Mst77F-yak, 64%), D. eugracilis (10 mya, Mst77F-eug, 35%), D. takahashii (15 mya, Mst77F-tak, 35%), and D. ananassae (25mya, Mst77F-ana, 22%). We only swapped the protein-coding sequences, retaining the regulatory sequences from the D. melanogaster Mst77F locus. We confirmed that all Mst77F transgenes were expressed at similar stages of spermatogenesis and comparable levels to the FLAG-tagged Mst77F-mel transgene using immunostaining (Fig. S2) and transcriptomic analyses (Fig. S3).

We first assessed whether Mst77F ortholog replacements affected male fertility in D. melanogaster. We found that D. melanogaster males homozygous for Mst77-ana were sterile. In contrast, transgenic males homozygous for other Mst77F orthologs had only a modest decrease in fertility at 25°C (Fig. 1B) or at 29°C (Fig. S4A), a temperature known to perturb sperm development (36). Thus, our findings show that homozygous Mst77F orthologs from closely related species can complement the crucial function of male fertility in D. melanogaster, despite sharing only 35–71% protein identity. We next reduced Mst77F dosage by generating hemizygous males with one Mst77F allele and one knockout allele. Whereas Mst77F-mel could restore wild-type male fertility even in a single copy, the Mst77F-sim, Mst77F-yak, Mst77F-eug, and Mst77F-tak alleles showed more than a ten-fold reduction in fertility in single copies (Fig. 1C). Thus, Mst77F orthologs from closely related species are haploinsufficient for normal male fertility in D. melanogaster.

Mst77F protamine replacements unleash sex-ratio distortion in D. melanogaster

We previously hypothesized that autosome-encoded protamines, such as Mst77F, might evolve rapidly to suppress competition between X- and Y-bearing sperm, thereby ensuring optimal sex ratios and high male fertility. To test this hypothesis, we evaluated how the Mst77F gene replacements affected the sex ratio of the resulting adult progeny. We crossed transgenic D. melanogaster (XY) males encoding two copies of Mst77F-mel to wild-type D. melanogaster females (Fig. 2A). These crosses produced ~50% males at both 25°C (Fig. 2B–C) and 29°C (Fig. S4B), consistent with Mendelian expectations, indicating equal success of X- and Y-bearing sperm.

Fig. 2: Mst77F gene replacements in D. melanogaster generate progeny with a biased sex ratio.

(A) Punnett squares illustrate the expected genotypes from crosses between transgenic Mst77F-encoding XY males and wild-type XX females. Sex ratios of resulting adult progeny from transgenic D. melanogaster males carrying (B) two copies of Mst77F from the designated species, or (C) one copy of Mst77F from the designated species. Each dot represents data from a single male (n ≥ 7). The dashed line indicates the expected 50:50 Mendelian ratio. In some cases, low fertility (<100 offspring/male) of hemizygous males precludes robust tests against 50:50 ratio using χ² test. (D) Punnett square illustrates the expected genotypes from crosses between transgenic Mst77F-encoding XY males and females carrying an attached-X chromosome (X^X). In matings with X^X females, fertilization by X-bearing sperm produces XO males, whereas fertilization by Y-bearing sperm produces X^XY females. Embryos carrying zero or three X chromosome are inviable. (E) In crosses with X^X females, transgenic homozygous D. melanogaster males carrying two copies of Mst77F-mel have higher fertility than D. melanogaster males carrying two copies of Mst77F-sim or Mst77F-yak. (F) Transgenic homozygous D. melanogaster males carrying two copies of Mst77F-mel produce progeny with expected 50:50 Mendelian ratios. In contrast, D. melanogaster males carrying two copies of Mst77F-sim or Mst77F-yak produce more female-biased progeny (Student’s t-tests; P <0.05). Statistics for progeny counts were carried out using the unpaired Student’s t-tests (*p<0.05; **p<0.01; ****p < 0.0001; ns, not significant).

In contrast, fertile males with two copies of Mst77F orthologs, even from closely related species, produced over 60% male offspring. The sex-ratio distortion was further exacerbated by halving the dosage of Mst77F. Indeed, even D. melanogaster males carrying a single copy of Mst77F-mel produced 62% male progeny (Fig. 2C) though these males appear to be highly fertile (Fig. 1C). Males carrying only a single copy of Mst77F orthologs from other species yielded approximately 70% male progeny (Fig. 2C). Both the fertility defects and the sex-ratio distortion were alleviated in heterozygous males carrying one copy of Mst77F-mel along with a single copy of an Mst77F ortholog (Fig. S5A–B), which performed better than Mst77F-mel hemizygous males, suggesting a modest additive effect of the Mst77F orthologs (Fig. S5A–B). Thus, unlike Mst77F-mel, interspecific Mst77F orthologs cannot fully support function in male fertility or suppress sex-ratio distortion, but they do not act as dominant negatives in D. melanogaster.

Sex-ratio distortion can arise from differential production or fertilization success of X- or Y-bearing sperm, or from the varying viability of female or male offspring. To differentiate between these possibilities, we crossed the Mst77F-mel, Mst77F-sim, or Mst77F-yak homozygous male (XY) flies with females possessing attached-X (or X^X) chromosomes (Fig. 2D). In this cross, viable progeny can only result from nullo-X eggs fertilized by X-bearing sperm, which develop into XO males, or X^X eggs fertilized by Y-bearing sperm, which develop into X^XY females (Drosophila sex is determined by X:autosome ratio, rather than the Y chromosome). We observed slightly lower fertility among transgenic males homozygous for heterospecific Mst77F orthologs compared to Mst77F-mel homozygous males (Fig. 2E). We found no skew in progeny sex ratios in crosses between Mst77F-mel homozygous males and X^X females (Fig. 2F). In contrast, we noted a higher proportion of X^XY females among the progeny from crosses of Mst77F-sim or Mst77F-yak homozygous males with X^X females compared to Mst77F-mel homozygous males (Fig. 2F).

Thus, both crosses to XX and X^X females produce fewer progeny resulting from X-bearing sperm. This indicates that the observed sex-ratio distortion is not due to differential viability of the resulting male and female progeny. Instead, an overrepresentation of Y-bearing over X-bearing sperm must underlie this skew, occurring either during spermatogenesis or at fertilization. Our experiments further demonstrate that optimal dosage and sequence of Mst77F are necessary to suppress the sex-ratio distortion of the resulting progeny.

X-chromosome-specific DNA condensation defects during spermatogenesis underlie the sex-ratio distortion in Mst77F replacement males

To investigate whether X- and Y-bearing gametes progress similarly through spermatogenesis, we used fluorescent in situ hybridization (FISH) to detect satellite DNA located on either Y chromosomes (AATAC) or autosomes (Prod-sat) (Fig. 3A). This strategy allowed us to distinguish between X-bearing mature sperm, which hybridize solely with the Prod-sat probe, and Y-bearing mature sperm, which hybridize with both the Prod-sat and AATAC probes. We then calculated the proportions of X- and Y-bearing mature sperm in transgenic D. melanogaster males encoding two copies of different Mst77F orthologs. We found no X-versus-Y skew in mature sperm from Mst77F-mel homozygous males (Fig. 3B–C), consistent with no observed sex-ratio distortion in progeny from these males (Fig. 2). In contrast, we identified a significant overrepresentation of Y-bearing sperm in males encoding a single copy of Mst77F-mel, or in Mst77F-sim, Mst77F-yak, Mst77F-eug, or Mst77F-tak homozygous males (Fig. 3B–C), corresponding with the sex-ratio bias in the resulting adult progeny (Fig. 2). These findings suggest that the under-representation of X-bearing sperm during spermatogenesis is sufficient to explain the sex-ratio distortion observed in the progeny of Mst77F replacement males.

Fig. 3: Mst77F gene replacements preferentially disrupt sperm chromatin condensation and decrease the proportion of X-bearing sperm.

(A) Schematic of DNA FISH probes used to distinguish X- and Y-bearing sperm based on the Y-linked AATAC (yellow), X-linked 359-bp satellite (black), and autosomal Prod-sat (cyan) satellite repeats. Y-bearing spermatids and sperm hybridize to both AATAC and Prod-sat probes, while X-bearing spermatids and sperm hybridize to Prod-sat and 359-bp probes. DNA was stained with Hoechst 33342 (magenta). (B) FISH was performed using the Y-linked AATAC (yellow) and Prod-sat (cyan) probes on seminal vesicles from homozygous D. melanogaster males encoding two copies of Mst77F-mel, Mst77F-sim, Mst77F-yak, Mst77F-eug, or Mst77F-tak orthologs or hemizygous D. melanogaster males encoding a single copy of Mst77F-mel. (C) Quantification of the proportion of Y-bearing sperm among total mature sperm in each of the six transgenic Mst77F gene replacement genotypes from (B). Each dot represents data from one seminal vesicle, and the dashed line represents the expected 50:50 Mendelian ratio. (n = total number of surveyed sperm in each genotype; χ² test against 50:50 ratio). (D) Immunostaining using an α-dsDNA antibody (yellow) was combined with FISH analyses using a probe against X-linked 359-bp satellite repeats (cyan) on late-stage spermatids from Mst77F-tak replacement males. DNA was stained with Hoechst 33342(magenta). White arrows highlight spermatids with signals from both the X-linked 359 probe and α-dsDNA antibody. The results from other genotypes are shown in Fig. S6A. (E) Mature sperm were stained with an α-dsDNA antibody (yellow) to assess their degree of chromatin compaction. Representative images of single sperm are shown here; seminal vesicle images are shown in Fig. S6B. (F) α-DsDNA antibody staining was carried out on sperm from the various Mst77F replacement D. melanogaster males. Images were quantified and normalized to the signal of Hoechst 33342. A minimum of 60 nuclei were assayed for each genotype. Statistical analyses were performed using unpaired Student’s t-tests (*p<0.05; ***p<0.001; ****p < 0.0001; ns, not significant). Scale bar = 10 mm for all images.

Since Mst77F is critical for DNA condensation, we hypothesized that delays or defects in DNA condensation might underlie reduced fertility and sex-ratio distortion in males with improper dosage or sequence of Mst77F. Under this hypothesis, X-bearing spermatids are more prone to DNA condensation defects than Y-bearing spermatids. To test this possibility, we focused on Mst77F-tak homozygous D. melanogaster flies, which have the most severe male bias in their progeny (Fig. 2B). In Drosophila species, X- and Y-bearing sperm resulting from a single meiotic event are produced simultaneously and undergo synchronous replication, resulting in a cyst containing 32 X-bearing and 32 Y-bearing spermatids. We performed immunostaining of spermatid cysts from Mst77F replacement homozygous D. melanogaster flies using an antibody to double-stranded DNA (α-dsDNA), which identifies decondensed spermatids (37). Under normal circumstances, chromatin in late spermatids and mature sperm is so tightly compacted that it completely excludes the α-dsDNA antibody. However, DNA condensation defects would result in staining with the α-dsDNA antibody (Fig. 3D–F; Fig S6). To distinguish whether spermatids containing X- or Y-chromosomes might be more prone to condensation defects, we combined immunostaining using the α-dsDNA antibody with FISH using a X-chromosome-specific 359-bp probe (Fig. 3A) in Mst77F-tak homozygous D. melanogaster flies. We found that, prior to spermatid individualization, ~84% of decondensed spermatids are X-bearing (n=57; p<0.001; Fig. 3D). In some cysts, well-condensed Y-bearing spermatids appear to individualize earlier than neighboring X-bearing spermatids, which remain not well-condensed. We investigated whether chromatin decondensation defects persist in mature sperm, even after sperm individualization. We observed no or minimal α-dsDNA antibody staining in mature sperm from Mst77F-mel homozygous D. melanogaster males (Fig. 3E–F; Fig. S6B). In contrast, we observed significant α-dsDNA antibody staining in late spermatids either from hemizygous Mst77F-mel males or from homozygous Mst77F-sim, Mst77F-yak, Mst77F-eug, or Mst77F-tak males, consistent with defective DNA condensation persisting even in mature sperm (Fig. S6B). We conclude that the inadequate dosage of conspecific Mst77F-mel or inappropriate sequence of heterospecific Mst77F leads to defective chromatin decondensation and differential loss of X-bearing sperm during sperm development in D. melanogaster, resulting in meiotic drive. Such DNA condensation defects are reminiscent of those observed in sperm from D. melanogaster males that encode the Segregation Distorter (SD) autosomal meiotic driver, which is known to selectively kill sperm that encode a target satellite (37). Similar chromatin defects might also explain the gross alterations in sperm nuclear morphology we observed in Mst77F-mel hemizygous males and Mst77F ortholog homozygous males, but not in Mst77F-mel homozygous males (Fig. S7A).

Multiple domains of Mst77F contribute to its dual functions in D. melanogaster

Mst77F-yak is impaired for both fertility-essential and meiotic drive suppressor functions in D. melanogaster (Fig. 4A), even though it is separated from Mst77F-mel by merely six million years. Both Mst77F-mel and Mst77F-yak proteins are of similar length (215 and 219 amino acid residues, respectively) but differ by 78 amino acids, highlighting the remarkably rapid sequence evolution of Mst77F. These amino acid differences are distributed throughout the proteins, except for the relatively conserved high mobility box (HMG) domains and predicted nuclear localization signals (NLS; Fig. 4A). To identify the domains associated with fertility-essential or meiotic drive suppressor functions, we constructed chimeric proteins comprising different combinations of four segments of Mst77F-mel and Mst77F-yak.

Fig. 4: Fertility and progeny sex ratios of D. melanogaster-D. yakuba Mst77F chimeras.

(A) Schematic representation of predicted protein domain structures of Mst77F in D. melanogaster (2, 46), including a coiled-coil domain (aa 26–49), a high mobility group (HMG) box DNA-binding domain (aa 52–94), and two putative nuclear localization signals (NLS, aa 171–191, 200–214). Numbers indicate the positions of three breakpoints used to generate chimeric proteins for D. melanogaster and D. yakuba. Colors in the chimeric bars represent the species’ origin of the Mst77F segment, with D. melanogaster in blue and D. yakuba in magenta. (B) We measured the fertility of hemizygous D. melanogaster transgenic males expressing a single copy of Mst77F-mel, Mst77F-yak, or chimeric genes. Each dot represents data from an individual male. (C) We measured the adult progeny sex ratio of D. melanogaster transgenic males expressing two copies of Mst77F-mel, Mst77F-yak, or chimeric genes. Statistical comparisons of transgenic flies encoding chimeric genes were carried out to transgenic flies encoding either Mst77F- mel (above dashed line) or Mst77F-yak (below dashed line) using unpaired Student’s t-tests (*p<0.05; **p<0.01; ***p<0.001; ****p < 0.0001; ns, not significant).

Single hemizygous Mst77F-mel males have average more than 200 offspring, while hemizygous Mst77F-yak males have fewer than 20 progeny (Fig. 1). Therefore, to provide the most sensitive assay for male fertility, we tested each chimeric Mst77F gene in a single copy. Swapping the first, second, or fourth segment from Mst77F-yak into Mst77F-mel resulted in a significant reduction in fertility, whereas swapping in the third segment did not impact fertility. We also conducted a reciprocal analysis and discovered that the first or fourth segment from Mst77F-mel could only marginally rescue the low fertility of Mst77F-yak males (Fig. 4B). Hence, optimal fertility requires multiple domains of Mst77F-mel (Fig. 4B).

We next investigated the Mst77F chimeras for their ability to suppress sex-ratio distortion. Mst77F-mel homozygous males show no sex-ratio distortion, whereas Mst77F-yak homozygous males yield approximately 65% male progeny (Fig. 4C). Thus, to enhance the sensitivity of our comparison, we tested all chimeras as homozygotes, which also provided the added benefit of relatively higher overall fertility. We found that no single segment from Mst77F-yak significantly impaired Mst77F-mel’s capacity to suppress sex-ratio distortion (Fig. 4C). Conversely, introducing either the first or the fourth segment from Mst77F-mel into Mst77F-yak significantly alleviated but did not eliminate sex-ratio distortion (Fig. 4C). These results confirm that multiple segments of Mst77F-mel contribute to suppressing sex-ratio distortion.

By comparing all seven Mst77F transgenes, including Mst77F-mel, Mst77F-yak, and their chimeras, we identified a strong anti-correlation between male fertility and sex-ratio distortion (Pearson’s R=−0.82, p=0.01). This anti-correlation suggests that the depletion of X-bearing sperm directly leads to both sex-ratio distortion and decreased male fertility.

D. yakuba Mst77F knockouts are fertile but undergo sex-ratio distortion

The recent emergence of Mst77F and its loss in at least one lineage since its origin (the montium group) suggests that its crucial function in male fertility is unlikely to be universally conserved across Drosophila species. Our replacement experiments in D. melanogaster further demonstrate that Mst77F orthologs have functionally diverged in their roles related to male fertility and meiotic drive suppression. To clarify whether—fertility or suppression of meiotic drive—is the primary selective pressure driving the rapid evolution of Mst77F, we generated a CRISPR-Cas9-mediated knockout of Mst77F in the closely related species, D. yakuba (Fig. S8), which diverged from D. melanogaster only six million years ago (35, 38). To our surprise, we discovered that, unlike in D. melanogaster, Mst77F-yak knockouts in D. yakuba exhibit reduced fertility but are not sterile (Fig. 5A). Together with previous observations of the recent emergence and loss of Mst77F in some Drosophila species, our findings in D. yakuba suggest that Mst77F’s essential role in male fertility is likely a very recently derived function in D. melanogaster.

Fig. 5: Mst77F knockouts in D. yakuba are fertile but have sex-ratio distortion.

(A) Fertility of D. yakuba males carrying no, one, or two copies of wild-type Mst77F-yak. (B) Adult progeny sex ratios from D. yakuba males carrying no, one, or two copies of Mst77F-yak. Deviation from a 50:50 Mendelian expectation was measured using a χ² test. (C) Late spermatids and mature sperm from different D. yakuba males were stained with an α-dsDNA antibody (yellow) to assess their degree of chromatin compaction. Scale bar = 10 mm for all images. (D) Quantification of α-dsDNA antibody staining in sperm from wild-type, hemizygous, and Mst77F-yak knockout D. yakuba males. Statistical analyses were carried out using unpaired Student’s t-tests. (*p<0.05; ***p<0.001; ****p<0.0001; ns, not significant).

Since Mst77F-yak D. yakuba males are fertile, we next investigated whether Mst77F-yak plays a role in suppressing sex-chromosome meiotic drive, like it does in D. melanogaster. Indeed, we found that D. yakuba males lacking one or both copies of Mst77F-yak produced male-biased offspring, with homozygous knockout males showing a more pronounced bias than hemizygous males (Fig. 5B). This phenotype resembles the male-biased offspring observed in crosses involving either hemizygous Mst77F-mel or homozygous Mst77F-yak D. melanogaster males, both of which are fertile yet exhibit sex-ratio distortion (Fig. 2). Like our findings in D. melanogaster, we find that the depletion or absence of Mst77F-yak in D. yakuba males leads to sperm DNA compaction defects (Fig. 5C–D) and abnormal nuclear sizes in mature sperm (Fig. S7B). Based on the male-biased progeny, we infer that these chromatin condensation defects are also preferentially affecting the X-chromosome-bearing spermatids, just like in D. melanogaster. These findings suggest that the two copies of Mst77F-yak are required to suppress delays or defects in sperm DNA compaction, which in turn results in sex-ratio distortion in the progeny. Based on our detailed investigation of Mst77F in both D. melanogaster and D. yakuba males, we conclude that its rapid evolution across Drosophila species is driven by its newly discovered role in suppressing meiotic drive, rather than its previously described essential function in male fertility.

Discussion

Novel biological pressures shaping cellular processes can lead to unexpected signatures of genetic innovation, even in genes critical for viability or fertility (39). Despite their essential roles in male fertility in mammals and Drosophila species (2), protamine genes are among the most rapidly evolving protein-coding genes in many animals. Yet, the underlying causes of protamine evolution have remained unclear. Here, using genetic and cytological tools available in D. melanogaster and related species, we performed in vivo gene replacements to show that the rapid evolution of the male-fertility-essential protamine Mst77F ensures proper DNA condensation in sperm carrying either X or Y chromosomes during spermatogenesis. Failure to do so results in decreased fertility and sex-ratio distortion. Our findings represent the first experimental investigation into the causes of rapid evolution of protamines in any species, demonstrating that suppression of competition between X and Y chromosomes may be a primary driver of the rapid evolution of rapidly evolving protamines.

Our findings indicate that Mst77F protamine, which plays a crucial role in sperm genome compaction, may suppress meiotic drive by ensuring balanced representation of X and Y chromosomes in mature sperm. Disruptions in Mst77F dosage or binding specificity predominantly affect X chromosomes with unique satellite sequences, enabling Y-bearing sperm to outcompete X-bearing sperm and skew sex ratios. In contrast, a complete loss of Mst77F impairs condensation for all chromosomes, potentially leading to complete male sterility, as seen in D. melanogaster. Y-linked duplications, including Mst77Y in D. melanogaster (2) or Y-linked Mst33A in D. yakuba (2), could interfere with Mst77F function, thereby favoring their transmission by disrupting X-bearing sperm, prompting rapid evolutionary adaptation of Mst77F to evade such antagonism and maintain sex-ratio parity. Alternatively, Mst77F might suppress Y-linked sperm killers that target X-bearing sperm; Mst77F loss could unleash these selfish elements, distorting sex ratios and potentially causing sterility. Thus, the widespread and rapid evolution of Mst77F and other protamines might be driven by an ongoing evolutionary arms race with such genetic elements. Future studies that investigate the relationship between protamine evolution and the presence of sperm killers may offer insights into the mechanisms underlying selfish genetic elements and their effects on male reproductive success and sperm chromatin integrity.

Our findings reveal that both Mst77F-mel and Mst77F-yak act as suppressors of meiotic drive in their respective species. In contrast to their roles in male fertility functions, both genes are haploinsufficient to suppress sex-ratio distortion, even in the presence of other protamines. Previous findings have shown that ProtA and ProtB may also suppress meiotic drive; their knockdown intensifies the degree of meiotic drive conferred by Segregation Distorter (40). Moreover, loss of competition between X- and Y-bearing sperm due to a rare evolutionary XŶ fusion in the montium group of Drosophila species correlates with the loss of Mst77F and other protamines (2). These findings suggest that the retention, dosage, and rapid evolution of Mst77F and likely other protamine genes are associated with their role in suppressing meiotic drive during spermatogenesis. Our work highlights the power of an in vivo gene replacement strategy to investigate long-lasting conundrums about the biological causes and consequences of rapid evolution of genes essential for fertility and viability. It also highlights that, despite its ubiquity, Mendelian inheritance is ultimately a highly fragile truce.

Supplementary Material

Materials and Methods

Figs. S1 to S8

Data S1 to S3

Funding Statement

Damon-Runyon Cancer Research Foundation postdoctoral fellowship DRG 2438-21 (CHC)

National Institutes of Health grant R01-GM74108 (HSM)

Howard Hughes Medical Institute Investigator (HSM)

Data and materials availability:

Raw sequencing data were deposited in the NCBI Sequence Read Archive under BioProject accession no. PRJNA1290820. All data are available in the main text or the supplementary materials.