The Y chromosome as a battleground for intragenomic conflict

Abstract

Y chromosomes are typically viewed as genetic wastelands with few intact genes. Recent genomic analyses in Drosophila, however, show that gene gain is prominent on young Y chromosomes. Meiosis- and RNAi related genes often co-amplify on recently formed X and Y chromosomes, are testis-expressed, produce anti-sense transcripts and short RNAs. RNAi pathways are also involved in suppressing sex ratio drive in Drosophila. These observations paint a dynamic picture of sex chromosome differentiation, suggesting that rapidly evolving genomic battles over segregation are rampant on young sex chromosomes and utilize RNAi to defend the genome against selfish elements that manipulate fair meiosis. Recurrent sex chromosome drive can have profound ecological, evolutionary, and cellular impacts and account for unique features of sex chromosomes.

Gene loss as a paradigm of Y evolution.

In many animals and plants with separate sexes, sex is determined by a pair of heteromorphic X and Y chromosomes. Sex chromosomes (see Glossary) have evolved from ordinary autosomes [1], yet harbor unique characteristics. Male-specific Y-chromosomes are typically characterized as degenerate gene deserts: over millions of years of evolution, they have lost most of their ancestral protein-coding genes, accumulated repeat-rich, repetitive DNA and have often acquired a transcriptionally silent heterochromatin structure [2,3]. X-chromosomes typically maintain most ancestral genes and remain euchromatic, but in response to Y chromosome degeneration often evolve dosage compensation [2,3], to transcriptionally balance gene loss on the Y.

A defining feature of sex chromosomes is the lack of recombination between the X and Y over most or all of their length[4]. The complete lack of recombination on the Y renders natural selection inefficient, and is the ultimate cause for Y chromosome degeneration and loss of most ancestral genes [5].

Indeed, Y chromosomes of various species, ranging from plants to worms, insects and mammals, contain only few functional genes [2,6–9]. Y chromosomes are only transmitted through males, and the few genes that remain on degenerate Y chromosomes typically have male-specific functions [2,6–9]. For example, careful mapping and functional studies have identified only a dozen protein-coding genes on the 40-Mb large Drosophila melanogaster Y chromosome [10,11], all of which have spermatogenesis-specific functions [12]. Concurrent with gene loss, Y chromosomes in many taxa instead have accumulated massive amounts of repetitive DNA, and almost all of the D. melanogaster Y chromosome consists of heterochromatin, made up of such repeats. In some taxa, such as most grasshoppers, cockroaches, or dragonflies, Y chromosomes are lost entirely [2,13].

Their high repeat content combined with the notion of Y chromosomes being ‘genetic wastelands’ has deemed sequencing efforts of the Y chromosome both too challenging and pointless, and Y chromosomes are typically entirely ignored in many genome projects. The sequencing of a large portion of the human Y chromosome was a daunting challenge [14], given the difficulties of sequencing repeat-rich regions (note that the large heterochromatic part of the human Y has not been sequenced yet). Similar to D. melanogaster, the human Y contains only a small fraction of the number of genes on the X, confirming the view of the Y as a degenerate chromosome [14,15]. However, it also harbors nine gene families with testis-specific expression that mostly lack homologs on the X and are organized as palindromes [14,15]. The organization of the human Y chromosome thus reflects a combination of different evolutionary processes: the degeneration of the majority of genes originally common to the ancestral X/Y chromosome pair; and the occasional gain of genes with male-specific functions by sporadic transpositions from the X chromosome and autosomes. Other mammalian Y chromosomes that were subsequently sequenced showed a similar structure and gene content [8,16,17]. The mouse Y chromosome, however, revealed a highly unusual sequence composition [18], and detailed experiments pointed towards intragenomic conflict as a major player shaping sex chromosome evolution (Box 1). Indeed, the unique inheritance patterns of sex chromosomes make them a hotbed for intragenomic conflict [19,20]. Recent work in Drosophila suggests that the prevailing view of gene loss being the dominating evolutionary process on Y chromosomes may have been greatly oversimplified by focusing on highly degenerate Y chromosome. Instead, studies of young sex chromosomes in Drosophila point towards dynamic genomic battles on this unique part of the genome, which are the focus of this review.

Box 1. Convergent gene acquisition and amplification on mouse sex chromosomes

In contrast to the classical view of Y chromosomes being heterochromatic and gene poor, the mouse Y is almost entirely euchromatic and contains about 700 protein-coding genes [18]. The mouse Y chromosome consists of ancestral sequence which originated from the autosomal ancestor of the mammalian sex chromosomes, and acquired sequence not originally present on the ancestral autosomes. Ancestral sequence occupies only 2 Mb and its evolutionary history follows the classical trajectory of Y chromosome degeneration, with the bulk of the ancestral genes having been lost (only 9 ancestral genes remain). The remaining, acquired sequence consists almost entirely of repeated sequences made up of ~200 copies of a 500-kb unit that contains three rodent-specific protein-coding gene families. Members of the gene families are consequently massively amplified on the mouse Y (132, 197, and 317 copies of Sly, Srsy, Ssty, respectively, with intact ORFs), and their sequences are highly similar to each other. Intriguingly, the same gene families found on the mouse Y have been convergently acquired and amplified on the X chromosome, and both the X and Y acquired and amplified genes are expressed predominantly in the male germline [88]. Experimental manipulation supports segregation distortion driving the acquisition and amplification of these genes on the mouse X and Y: Sly is essential for normal spermatogenesis, and Sly-deficient male mice have reduced fertility, show widespread overexpression of X-linked genes, including its paralogs Slx/Slxl1, and produce female-biased litters [44]. In contrast, deficiency of Slx/Slxl1 reduces male fertility and results in male-biased litters [45], and simultaneous knockdown of all three genes significantly improves fertility, returns expression of sex-linked genes to wild-type levels, and restores equal sex ratio [45]. These antagonistic interactions suggest that Slx/Slxl1 and Sly are involved in an intragenomic conflict trying to cheat fair meiosis, and an evolutionary arms race between segregation distorter and repressor may drive their dramatic copy number expansion and amplification on the mouse sex chromosomes [89,90].

Box 2: Sex Chromosomes: a battle ground for meiotic drive?

Meiotic drive systems in flies and mammals involve a distorting locus that targets a sensitive responder locus on its homolog, so that the chromosome where the driver resides is transmitted to more than 50% of the offspring of a heterozygous carrier [39,91]. A meiotic drive system requires that (1) a distorter is able to discriminate its host chromosome from its homolog; (2) the distorter and responder loci must be in strong linkage to avoid the generation of suicide chromosomes that carry the distorter and a sensitive responder; and (3) the distorter must be able to trigger a molecular pathway that results in the destruction or unpreferential transmission of its homologs into gametes. All of these conditions are met on sex chromosomes: The X and Y chromosome can be highly differentiated, providing many potential targets for a driving allele, and they do not recombine over most or all of their length, thereby preventing the generation of suicide chromosomes. In addition, old Y chromosomes in many species are highly heterochromatic, and defects in heterochromatin formation can lead to chromosome mis-segregation [92]. This makes the Y chromosome especially vulnerable for mis-segregation, and the X chromosome—via manipulation of proteins involved in the regulation of the chromatin state—can take advantage of the heterochromatin state of the Y chromosome to induce meiotic drive. Indeed, most of the cellular meiotic drive phenotypes described in Drosophila species suggest a failure in chromatin state regulation [39,41,42,53,93,94]. One known component of a driving X chromosome in D. simulans consists of a partial deletion of the heterochromatin protein HP1D2 which disrupts segregation of Y chromatids in meiosis, leading to an excess of X-bearing sperm among males carrying the driving HP1D2 allele [95]. Thus, while segregation distorters can arise on all chromosomes, they are particularly prone to evolve on sex chromosomes [39].

Box 3: The seven stages of Y chromosome evolution.

Disparate evolutionary processes may dominate on Y chromosomes of different ages (Figure I). Conflict may dictate early Y chromosome evolution: young sex chromosomes contain many genes that are expressed in spermatogenesis, providing ample opportunity for amplification of dosage-sensitive drivers, and co-amplification of homologous suppressors that utilize RNAi. Rampant sex chromosome drive can account for massive co-amplification of X/Y genes on young sex chromosomes, and temporary increase in gene number on the Y [32]. Yet the Y is fighting a losing battle: both drivers and suppressors are less likely to emerge and fix on the Y from the beginning due to the decreased efficacy of selection, and this disparity only worsens over time due to Y degeneration: Gene loss on the Y means fewer loci can evolve or suppress drive [61], and repeat accumulation and heterochromatinization make the Y an easier target for drivers on the X [51], and the silencing effect of spreading heterochromatin can interfere with the expression of suppressors on the Y [28,96]. Unguarded drive of the X has dramatic fitness costs for the rest of the genome and transcriptional silencing of sex chromosomes (especially the X) during spermatogenesis may have evolved as a host genome defense against selfish meiotic drive elements [62]. Meiotic sex chromosome inactivation (MSCI) will silence meiotic drive elements in the right place and time, and transcriptional silencing of the X has been found in a wide range of taxa with heteromorphic sex chromosomes, including mice, chicken, flies and C. elegans [63,64,97–99]. Yet, MSCI will also silence other host genes on the sex chromosomes that are necessary for normal progression of spermatogenesis [63], which can drive gene trafficking of spermatogenesis genes off the sex chromosomes [74]. However, their peculiar inheritance also makes sex chromosomes a preferred location for male genes (the X is hemizygous in males, and the Y is male-limited), and the gene content of sex chromosomes can reflect these opposing selective pressures [100]. Gene content masculinization can accompany Y evolution: testis-expressed genes are less likely to be lost on the Y, or can be gained from autosomes [101,102]; this will result in an increase of genes with male-biased expression over time on the Y chromosome. Stasis of gene content can be observed on old, degenerate Y chromosomes in both mammals and fruit flies [11,46,103]. After the initial erosion of most genes, the small and highly specialized gene repertoire of the Y has remained considerably stable over millions of years of evolution.

Massive gene amplification on young Y chromosomes.

Drosophila neo-sex chromosomes have served as model systems to characterize the molecular and evolutionary forces driving sex chromosomes differentiation [21–26]. Here, novel sex chromosomes are formed independently in different lineages by fusions of autosomes to the ancestral sex chromosomes, allowing the reconstruction of how ordinary autosomes transition into heteromorphic sex chromosomes at various stages of differentiation (Figure 1). One species especially useful for characterizing this transition is D. miranda, whose 1.5MY old neo-sex chromosomes are at a sweet spot of divergence (Figure 1): they are similar enough so that homologous regions can readily be aligned (98.5% identical), yet with dramatic structural and epigenetic differentiation between the neo-X and neo-Y [23,24,26–28]. Decades of cytogenetic and molecular work have revealed that the neo-Y of D. miranda is degenerating; neo-Y genes show increased rates of protein evolution [23,29], including premature stop codons and frame-shift mutations at several hundred genes [26,30], and dozens of genes have been lost entirely [23,31,32]. The neo-Y also shows a dramatic accumulation of transposable elements (TEs) [27,33], and is starting to evolve a heterochromatic appearance [28], with the majority of genes being expressed at a lower level from the neo-Y [26,34]. In response, its neo-X has started to evolve dosage compensation [35–37].

Figure 1.

A. Drosophila neo-sex chromosomes. Neo-sex chromosomes in Drosophila are formed by fusions of autosomes with the ancestral sex chromosomes, and the neo-X and neo-Y carry identical genes at the time of their origination (0 MY). Comparing neo-sex systems of varying age provides a dynamic picture of the molecular changes associated with sex chromosome differentiation. Down-regulation of individual protein-coding genes is observed on the young neo-Y in D. albomicans (0.1MY) [22]. A general remodeling of the genome architecture of the neo-sex chromosomes has occurred in D. miranda. Its neo-Y underwent massive decay in gene function, is accumulating repetitive DNA, and is becoming heterochromatic [26,28]. Its neo-X is evolving partial dosage compensation [35]. After 15 MY (D. pseudoobscura), virtually no sequence similarity remains between old neo-sex chromosomes, and the neo-Y is entirely heterochromatic, and the neo-X is fully dosage compensated, resembling the general architecture of the ancestral sex chromosomes in Drosophila. B. Amplification of genes on the neo-sex chromosomes of D. miranda. Shown is a schematic representation of gene content of the D. miranda neo-Y/Y-chromosome. The D. miranda neo-Y/Y harbors 1697 multi-copy Y genes (derived from 363 distinct proteins), and 2036 genes (derived from 94 distinct proteins) that co-amplified on both the X/neo-X and Y/neo-Y. Most ampliconic Y genes were ancestrally present on the autosome that formed the neo-sex chromosomes.

While a draft genome for D. miranda was published several years ago [26], initial attempts to generate a contiguous neo-Y sequence using short-read sequencing failed due to the repetitive nature of the neo-Y. Application of long-read technology, however, allowed assembly of over 110 Mb of the Y/neo-Y into 3 large contigs, and examination of its gene and repeat content revealed several unexpected results [24,32]. Transposable elements had started to expand on the neo-Y [23,27], but just how much they had accumulated within 1.5 MY was astounding: the neo-Y was found to be about 3 times the size of the neo-X, mostly due to the invasion of tens of thousands of TEs [24]. Widespread gene loss on the Y is considered a hallmark of sex chromosome differentiation, and indeed, about 5% of genes (143 genes) that were ancestrally present on the neo-Y are completely absent [32], and hundreds of genes contain premature stop codons [26,31]. However, entirely unexpected and counter to prevailing views of Y evolution, the number of annotated protein-coding genes on the neo-Y (6,448 genes) dramatically outweighed the number of genes on the neo-X (3,253 genes), or the ancestral autosome from which the neo-sex chromosomes derived (3,087 genes) [32]. This implies that the neo-Y gained well over 3000 genes since its formation 1.5MY ago. Thus, rather than simply degenerating, the initial stages of Y evolution appear to be characterized by a massive accumulation of genes. This raises a series of questions: Where do these genes come from? How did they originate? And why do they accumulate on the Y chromosome?

Comparative analysis revealed that most of the newly gained genes are derived from tandem amplification of a subset of protein-coding genes ancestrally present on the neo-sex chromosomes [32]. Overall, the neo-Y contains 3733 gene copies derived from 457 ancestral single-copy genes, which are now found mostly as clusters of gene families across the neo-Y [32] (Figure 1B). Intriguingly, 94 ancestral single-copy genes not only amplified on the Y/neo-Y, but independently co-amplified on the X/neo-X [32]. Gene amplifications on the neo-Y could be driven by various evolutionary processes. Multicopy Y genes might accumulate neutrally, or may be slightly deleterious. For example, repeats on the neo-Y can provide a substrate for non-allelic homologous recombination and thereby promote gene-family expansion [38], or heterochromatin spreading from repeats may silence gene duplicates on the neo-Y [28]. Gene family expansions on the Y can also be beneficial for males. Global transcription is lower from the neo-Y-chromosome of D. miranda [34], and gene amplification may help compensate for reduced gene-dose of neo-Y genes. Additionally, Y chromosomes are an ideal location for genes that specifically enhance male fitness [20]. Gene amplification on the Y could also be a signature of intragenomic conflicts. Y chromosomes compete with the X over transmission to the next generation [39,40], which has resulted in co-amplification of genes on the sex chromosomes in other species [41–45]. Under neutral/slightly deleterious models driving gene gain, amplified Y genes should be a random subset of the genes present on the Y, and not enriched for particular functional categories or expressed in any specific tissue. Functional genomic analysis, however, revealed unique characteristics of multicopy Y genes (which only amplify on the Y/neo-Y) versus co-amplified X/Y genes (co-amplified on both the X/neo-X and the Y/neo-Y) that suggest that selective processes contribute to gene family expansion. Multi-copy-Y genes comprise dosage-sensitive and testis-specific genes that appear to have amplified on the Y to increase male fitness, while co-amplification of X/Y genes may reflect ongoing conflicts over sex chromosome transmission [32] (see below).

The Y chromosome as a safe haven for dosage-sensitive and male-beneficial genes.

Survival of ancestral genes on the Y chromosome of mammals is non-random and dosage-sensitive genes are more likely to be retained [8,14,46]. In Drosophila, transcriptional up-regulation of X-linked genes in males assures equal amounts of gene product for X-linked genes in both sexes [47]. D. miranda has evolved only partial dosage compensation of its neo-X chromosome [35,36], and gene amplification may help compensate for reduced gene dose of neo-Y genes. Indeed, the neo-X homologs of multi-copy Y genes are less likely to be dosage compensated in males compared to single-copy Y genes [32]. This suggests that some multi-copy Y genes (those that only have a few copies and are ubiquitously expressed) are dosage sensitive, and additional gene copies on the Y may contribute to dosage compensation [32].

Y chromosomes are transmitted from father to son, and are thus an ideal location for genes that increase male fitness [20]. Old, degenerate Y chromosomes of several species, including humans, contain multi-copy gene families that are expressed in testis and contribute to male fertility [8,14,46]. Similarly, the few genes that remain on the old Y chromosome of D. melanogaster all function in spermatogenesis [12]. In both humans and D. melanogaster, most Y genes have originated by transposition from an autosome [14,48]. This suggests an ongoing evolutionary process of acquisition of genes with male-specific functions on old Y chromosomes (masculinization); if higher levels of expression of these genes are advantageous for males but disadvantageous for females, this would be favored by selection [49].

Analysis of Drosophila neo-sex chromosomes of various ages suggests that Y chromosomes can provide safe havens for male beneficial genes throughout the process of Y evolution. The very young neo-Y of D. albomicans (about 0.1 MY old) contains only a few dozen genes that already show evidence of decay [22] (Figure 1A), but genes that are down-regulated on the neo-Y are depleted for testis-expressed genes. Thus, during the very initial stages of Y differentiation, genes with testis-biased expression are maintained more effectively on a degenerating Y chromosome relative to random genes. The D. miranda neo-Y also shows evidence for preferential maintenance of genes with important male functions [26]. Neo-Y genes classified as male-beneficial more often undergo accelerated protein evolution in D. miranda, and neo-Y genes evolve biased expression toward male-specific tissues [26], suggesting that a subset of neo-Y genes improve or acquire male-related functions [26]. Finally, the neo-Y has acquired and amplified testis-expressed genes [32]; most multi-copy Y gene families with a high copy number on the neo-Y were expressed almost exclusively in testis of D. miranda [32], mimicking patterns of gene family amplification of male fertility genes found on old Y chromosomes in other species [8,14,46].

This demonstrates that natural selection maintains male beneficial genes more effectively during the early stages of Y degeneration, existing Y genes evolve male-specific functions, and genes with male functions are acquired secondarily on evolving Y chromosomes, supporting the notion of the Y as a safe haven for male genes.

X/Y gene co-amplification: a molecular signature of sex chromosome drive?

Mendel’s law of segregation states that the two copies of each chromosome in a diploid organism are transmitted with equal probability to its offspring. However, equal segregation may be constantly challenged by intragenomic competition for transmission to the next generation. Meiotic drive elements (also known as segregation distorters) are selfish genetic elements that manipulate meiosis or gametogenesis to increase their own transmission [50].Theory suggests that sex chromosomes are particularly prone to evolve meiotic drive, since they lack recombination, are differentiated at the molecular level and the Y chromosome may provide a relatively easy target for segregation distorters [39,51,52] (Box 2). Known sex ratio drive systems in Drosophila either induce meiotic failure [53], or seem to operate as spermatids differentiate into mature sperm [42].

Meiotic drive favors the evolution of loci that suppress distortion among regions linked to the sensitive responder allele. Disruption of equal transmission of the X and Y chromosomes, however, has the added burden of distorting population sex ratios [39,51,52] which could lead to population collapse and extinction. Reduced fertility and distorted sex ratios thus create strong selective pressure to evolve loci that repress sex ratio drive [39,54,55]. Sex chromosome drive could thus be prevalent in evolution but often short-lived and cryptic, since bursts of meiotic drive elements arising on sex chromosomes should be followed by the quick invasion of suppressor alleles that restore equal sex ratios. Cryptic drive systems are typically only revealed through detailed genetic manipulations or crosses between populations or species [41,56,57], and little is known about their abundance in nature. Recurrent bursts of invading drivers and repressors, however, may sometimes leave characteristic footprints at patterns of genome evolution: past meiotic drive in both mammals and fruit flies has repeatedly led to duplications or co-amplifications of homologous genes on both the X and the Y chromosome [41,42,44,45]. Gene duplications have generated distorting chromosomes several times in Drosophila [52], and one of best-characterized meiotic drivers is the Winters sex-ratio system within D. simulans. Here, two X-linked distorters, Dox and MDox (with Dox being derived from MDox), are both necessary for drive [41,42], and two autosomal genes (called Nmy and Tmy) that originated from Dox/MDox, suppresses this drive [43]. Often, both the distorter and suppressor alleles are not only homologous to each other, but also amplified on both the X and Y chromosome [18,44,58]. This is the case for the Slx/Sly system on the mouse sex chromosomes (Box 1), whereby the X-and Y-linked copies of this co-amplified gene family directly silence each other [44,45]. Similar X/Y co-amplification has occurred in the Stellate/Suppressor of Stellate (Ste/Su(Ste)) system in D. melanogaster, a suspected now defunct sex ratio driver [59]. The expression of the X-linked multi-copy gene Ste leads to the production of defective sperm, and Su(Ste), which is a multi-gene copy of Ste that moved to the Y-chromosome, silences Ste [60]. Sequence homology between the distorter and the suppressor immediately suggest a molecular mechanism of how suppression might operate, either through competition of similar proteins (as for Slx/Sly[44,45]), or through silencing of the driver through RNAi (i.e. Dox/Nmy or Ste/Su(Ste); Figure 2, see below). If co-amplified X/Y genes are involved in a battle over fair transmission of the sex chromosomes, changes in gene copy number may tip the balance over inclusion of a chromosome into functional sperm, and could result in repeated co-amplification of dosage-sensitive distorters and suppressors on the sex chromosomes [32,61].

Figure 2.

RNAi and sex ratio drive in D. simulans. Natural populations of D. simulans harbor multiple genetically independent sex-ratio systems. Two major components of the Winters drive system are the X-linked distorter on the X (Dox) and a dominant autosomal suppressor (Nmy) [41,42]. Males homozygous for a non-functional allele of nmy fail to suppress the function of Dox and produce defective Y-bearing spermatids and thus an excess of daughters. Molecular characterization of these genes revealed that Nmy shares sequence homology with Dox and likely arose through retrotransposition of Dox itself [41,42], while Dox originated by duplication of another X-linked locus, MDox [42]. Nmy contains a near perfect inverted repeat [41], and it was shown that Nmy hairpin RNA produces small RNAs and these small RNAs are absent in non-functional nmy genotypes [43]. Additionally, another autosomal repressor (Tmy) was identified that shows sequence homology to Nmy and also produces hairpin RNAs, and Nmy and Tmy have partially overlapping capacity to suppress both Dox as well MDox [43]. Intriguingly, null mutants of two RNAi pathway components, dcr-2 and ago2, are male-sterile due to profound defects in spermatogenesis progression [43]. These mutants harbor massive de-repression of both Dox and MDox transcripts, consistent with loss of collaborative suppression by Nmy and Tmy small RNAs [43]. Thus, RNAi plays a critical role in intragenomic conflict over sex ratio in D. simulans.

As mentioned, D. miranda harbors a large number of co-amplified gene families on its young neo-sex chromosomes, and several features indicate ongoing conflicts over sex chromosome transmission. In particular, many of the co-amplified X/Y genes are ancestrally expressed in gonads and have well-characterized functions in meiosis, and gene ontology analysis revealed that co-amplified X/Y genes are significantly overrepresented in biological processes associated with meiosis, chromosome segregation and chromatin condensation [32]. In particular, co-amplified X/Y genes are significantly enriched for GO categories including “spindle assembly”, “DNA packaging”, “chromosome segregation”, or “male gamete generation”. As expected for meiotic drivers and their suppressors, both the X- and Y-linked copies of co-amplified X/Y genes are typically highly expressed during spermatogenesis, with little expression in other tissues [32]. Functional enrichment patterns thus indicate that putative drive genes often have well-characterized functions in meiosis ensuring precise chromosome segregation and condensation, and selfish elements have recurrently succeeded in manipulating these normally tightly regulated cellular processes for their own advantage. Most importantly, however, almost all co-amplified X/Y gene families generate short RNAs and are overall down-regulated compared to single-copy genes [32]; this argues against their amplification being simply driven to increase their gene product in testis, but instead is strong evidence for intragenomic battles between the sex chromosomes utilizing RNAi (see below). However, direct evidence is needed to demonstrate the involvement of co-amplified X/Y genes in drive.

Interestingly, co-amplification of testis-expressed genes on young neo-sex chromosomes appears to be the norm rather than the exception [61]. A conservative genomic analysis probing for co-amplified X/Y genes across 26 Drosophila species (15 with young neo-sex chromosomes and 11 without) identified co-amplified X/Y genes in ten species, nine of which have neo-sex chromosomes [61]. Similar to D. miranda, co-amplified X/Y genes show testis-specific expression and several have well-characterized meiosis-related functions and generate short RNAs [61]. Preferential occurrence of co-amplified X/Y genes in species with recently formed neo-sex chromosomes suggests that sex ratio distorters have repeatedly evolved to exploit genomic vulnerabilities associated with the formation of new sex chromosomes (Box 3). Meiotic sex chromosome inactivation (MSCI) is a peculiar feature suggested to act as an epigenetic form of host genome defense against selfish meiotic drive elements [62]. Sex chromosomes are precociously heterochromatinized and transcriptionally silenced during spermatogenesis in several organisms [63,64], which is the time and place during which meiotic drive elements are active. Suppression of transcription during spermatogenesis may not yet have fully evolved on young sex chromosomes, allowing the expression of sex-linked drivers (Box 3).

Sex Chromosome Drive and RNAi

How could co-amplification of meiosis-related genes on the X and Y result in meiotic drive and its suppression at the cellular level? Detailed molecular studies, functional enrichment patterns and gene expression profiles suggest that some meiotic drive elements and/or their suppressors employ RNAi mechanisms to battle over inclusion into functional sperm [52]. In D. melanogaster, piRNA pathway mutants enhance the drive caused by the well-characterized autosomal driver Sd [65]. In addition, the Ste/Su(Ste) system in D. melanogaster uses RNAi; Y-linked Su(Ste) multicopy genes silence homologous X-linked Ste transcripts through anti-sense expression and generation of short RNAs [60]. The most direct evidence for an involvement of RNAi in sex ratio drive comes from detailed molecular characterization of the Dox/Nmy sex-ratio driver in D. simulans [43]. Here, an X-linked driver (Dox/MDox) has been shown to be suppressed by homologous sequences on an autosome (Nmy/Tmy) via short RNA-based silencing [41,42] (Figure 2). Thus, the emergence of homologous repeated sequences can resolve an intragenomic conflict, through triggering the RNAi pathway.

RNAi is also thought to be involved in the putative sex chromosome drive in D. miranda [32]. Most importantly, the vast majority of co-amplified X/Y genes in D. miranda are targeted by short RNAs in testis, directly linking RNAi to regulation of co-amplified X/Y genes [32]. In addition, co-amplified X/Y genes show a significant overrepresentation of GO terms associated with piRNA metabolism and the generation of short RNAs [32]. Noteworthy genes in the RNAi pathway that are typically single-copy in insects but co-amplified on the X and Y of D. miranda include Dicer-2 (a endonuclease that cuts long double-stranded RNA into short RNAs), cutoff (a gene involved in transcription of piRNA clusters), or shutdown (a co-chaperone necessary for piRNA biogenesis). Co-amplification of RNAi genes is expected under recurring sex chromosome drive where silencing of distorters is achieved by short RNAs, since compromising the RNAi pathway would release previously silenced drive systems [43]. Indeed, knock-downs of Dcr-2 or Ago2 (required for short RNA generation) in D. simulans create female-biased sex ratio [43] (Figure 2).

In order to trigger the RNAi response, the production of double-stranded RNA (dsRNA) is necessary [66]. This can be achieved through anti-sense transcription of the target genes and indeed, the vast majority of co-amplified X/Y (but not single-copy X/Y) genes in D. miranda produce both sense and antisense transcripts [32]. Antisense transcription and short RNA production has also been detected at co-amplified X/Y genes in D. pseudoobscura, suggesting similar mechanisms to silence sex chromosome drive in other Drosophila species with neo-sex chromosomes [61]. The production of dsRNA can also occur via hairpin RNAs. In the D. simulans Dox/Nmy system, the two suppressor genes both encode related long inverted repeats that can form hairpin RNAs, which are then processed by the RNAi machinery to generate short RNAs that repress the paralogous distorters [43].

Thus, the production of antisense transcripts or hairpin RNAs and short RNAs may be a common feature of meiotic drive elements [32,43]. Intriguingly, genes in the RNAi pathway often evolve rapidly and show frequent gene duplication and loss over evolutionary time periods. Argonaute 2 (Ago2), for example, is one of the key RNAi genes in insects, and has repeatedly formed new testis-specific duplicates in the recent history of the D. pseudoobscura group [67], and analysis of additional RNAi-pathway genes confirms that they undergo frequent independent duplications in this clade [68]. The presence of young neo-sex chromosomes in this species group might make them vulnerable to the invasion of meiotic drive elements and account for the rapid evolution of RNAi genes in this clade [69].

Evolutionary implications of sex chromosome drive

Meiotic drive elements can spread through populations even if they reduce organism fitness, and account for a multitude of molecular and phenotypic phenomena [39,54]. Strong selective pressure to amplify Y-linked suppressors of meiotic drive may indirectly account for the complete genetic decay of the Y chromosome [49]. Since the Y chromosome lacks recombination, strong positive selection for meiotic drive suppressors can drag along linked deleterious mutations to fixation [3]. The ongoing degeneration of ancestral Y genes may thus be a by-product of silencing recurrent meiotic drivers arising on the X. Natural lines of D. miranda show a wide range of sex-ratio bias (typically female-biased [70]) and patterns of molecular variation support episodes of positive selection shaping neo-X and neo-Y evolution of D. miranda [71,72]. This is consistent with recurrent and ongoing conflicts over segregation affecting the genomic architecture of sex chromosomes in this species.

Rampant sex chromosome drive can also shape the genomic distribution of genes and patterns of gene expression on sex chromosomes [62]. In Drosophila, de novo genes that arise from ancestral non-coding sequence are frequently expressed in testis and preferentially located on the X [73]. In contrast, duplicate genes that are formed by retrotransposition of an existing gene are often autosomal but derived from a parental X-linked gene and also show testis-biased expression [74]. The evolution of sex-ratio distorters and suppressors could contribute to these patterns, resembling evolutionary events of the Dox/Nmy sex ratio system in D. simulans. In particular, X-linked de novo genes with testis expression could be X-linked distorters, and autosomal retrotransposed copies of X-linked genes might function as suppressors of sex-ratio distorters via RNAi mechanisms.

In many species, including Drosophila, expression from the X chromosome is reduced during spermatogenesis [64,75], which may have evolved as a genome defense against driving X [62]. Gene co-amplification on the X and Y is common in flies with young neo-sex chromosomes, but becomes less frequent on older sex chromosomes [61]. Young X chromosomes may not yet be transcriptionally inactive during spermatogenesis and thus more vulnerable to meiotic drive. Genetic conflict between X-Y ampliconic genes may also contribute to hybrid sterility and consequent reproductive isolation [40,76,77], but can also impede molecular divergence between species [78]. Segregation distortion can result in male hybrid sterility in Drosophila [57], and further functional characterization of co-amplified, lineage-specific X-Y gene families will be needed to test the proposed link between X-Y genetic conflict and hybrid sterility.

Concluding Remarks

X-Y interchromosomal conflict and concurrent gene co-amplification on sex chromosomes may be widespread. In human and mouse—two species with high-quality reference assemblies for both sex chromosomes—the X and Y have co-acquired and amplified genes, and in both cases, meiotic drive has been invoked to explain this co-amplification [18,44,45,79,80]. Highly amplified gene families have also been detected in other mammals [81] and are widespread across fruit flies with young sex chromosomes [32,58,61], suggesting that sex chromosome drive may be prevalent in evolution. High-quality sex chromosome assemblies across more taxa are needed to determine the true phylogenetic range of lineage-specific acquisition and amplification of X-Y genes.

The prevalence of cryptic sex ratio drive systems in insects and animals may account for several evolutionary and molecular phenomena [62]. Sex ratio distorters can fuel the rapid turn-over of sex determination mechanisms across species [82], and the transcriptional inactivation of sex chromosomes during spermatogenesis may have evolved as a defense against meiotic drive elements. The recurrent fixation of cryptic drive systems on sex chromosomes might explain the prominent role of the X chromosome in the evolution of hybrid sterility in a wide range of species [40,76,83–85], and contribute to genomic biases in the location of sex-biased genes or gene duplicates [74,86]. The short RNA pathway is emerging to play an important role in controlling selfish genetic elements that try to exploit highly regulated cellular processes such as chromosome segregation [87]. Future characterization of putative drive systems will provide a full picture of how distorting elements manipulate and cheat meiosis, what molecular pathways or developmental processes are particularly vulnerable, and how the genome has launched evolutionary responses to counter distortion (See also outstanding questions).

Outstanding questions.

How common is cryptic sex chromosome drive?

Sex chromosomes are prone to the invasion of meiotic drive elements, yet severe fitness consequences associated with sex ratio distortion select for the rapid fixation of suppressor alleles. A lack of MSCI may render young sex chromosomes particularly vulnerable to transmission distortion. Studying the prevalence of segregation distorters on X and Y chromosomes of different ages will inform us on the prevalence and temporal dynamics of sex chromosome drive.

What type of genes and pathways are involved in meiotic drive?

Drivers can increase their transmission in males by interfering with fair meiosis, or by preventing the maturation or function of sperm that do not contain it. A driving X could for example interfere with the correct packaging of the highly heterochromatic Y. Are there general themes in what kind of genes can evolve a drive phenotype? Is there some general weakness in spermatogenesis that drive can exploit? What type of genes, and what molecular functions are particularly prone to be manipulated in the intragenomic battle over transmission?

Are RNAi pathways typically involved in meiotic conflicts?

Data from Drosophila suggest that the emergence of homologous repeated sequences can resolve an intragenomic conflict, by silencing the initial driver through RNAi mechanisms. How often is the RNAi pathway employed to silence drivers on sex chromosomes? What other mechanisms exist for silencing a drive?

What are the evolutionary consequences of sex chromosome drive?

Recurrent conflict over the transmission of sex chromosomes could have shaped widespread cytological and evolutionary patterns, including the epigenetic regulation of sex chromosomes, the genomic distribution of genes expressed in the germline, and the evolution of hybrid sterility between species. Sex chromosome drive could result in population extinction, and may be a major reason for recombination itself. Has drive really had this much impact?

Figure I.

The seven stages of Y evolution. Shown are different evolutionary or cellular processes that dominate Y differentiation at different time points of Y evolution.

Highlights.

• Sex chromosomes have originated independently many times from ordinary autosomes, and widespread gene loss is considered a hallmark of Y chromosome evolution.

• Genomic analyses in Drosophila have revealed that gene (co)-amplification is common on recently formed X/Y chromosomes which may gain 1000s of genes.

• Co-amplified X/Y genes are derived from well-characterized meiosis genes involved in chromosome segregation and chromatin condensation, or in RNAi, are typically expressed in testis, and often produce antisense transcripts and short RNAs.

• Knock-down of RNAi genes in Drosophila can produce female-biased offspring, due to a failure of producing silencing short RNAs that target an X-linked driver.

• Sex chromosomes are susceptible to an evolutionary tug-of-war over segregation and intragenomic conflicts can utilize RNAi to counter sex ratio distortion.

Glossary

Sex chromosome

A type of chromosome that participates in sex determination

Autosomes

A chromosome that is not a sex chromosome

Repetitive DNA

DNA sequences that occur in multiple copies throughout the genome. Repetitive DNA is mostly composed of transposable elements and satellite DNA

Heterochromatin

A densely packed form of DNA that is associated with gene silencing and typically is composed of repetitive DNA

Euchromatin

A less densely packed form of DNA that is associated with gene transcription

Dosage compensation

A mechanism by which species with sex chromosomes equalize gene expression between both sexes

Y chromosome degeneration

A process in which the Y chromosome loses most of its original genes over evolutionary time

Palindromes

Two very similar long sequences point in opposite directions, connected by a “spacer”

Intragenomic conflict

An evolutionary phenomena where genes residing in different parts of the genome promote their own transmission in detriment of the transmission of other genes that reside in the same genome

Transposable elements (TEs)

DNA sequences that are able to move from one location to another in the genome

Dosage-sensitive gene

A gene where two copies are required for normal function

Masculinization

The maintenance or acquisition of genes on the Y chromosome with male-beneficial functions

Meiotic drive

A type of intragenomic conflict, whereby one or more loci within a genome will manipulate gametogenesis in such a way as to favor their own transmission

Segregation distorters

Selfish genetic elements that manipulate meiosis or gametogenesis to increase their own transmission

Sex ratio drive

Meiotic drive occurring on the sex chromosomes, and thus biasing the offspring sex ratio

Meiotic sex chromosome inactivation (MSCI)

Sex chromosomes are precociously heterochromatinized and transcriptionally silenced during spermatogenesis

RNAi mechanisms

A cellular mechanism in which RNA molecules inhibit gene expression or translation, by neutralizing targeted mRNA molecules