Sex determination is the process by which organisms develop as either male or female. Sex determination mechanisms differ among species. Genome-level studies in many model and non-model species have elucidated the genetic mechanisms associated with sex chromosome differentiation. The process of sex chromosome degeneration and rearrangement, in which an obvious increase in repetitive non-coding sequences and transposable elements (TEs) is observed for most sex chromosomes [17,18,19,20,21], is an important biological constraint that sheds light on the evolution of sex chromosomes [35,42,45,46]. Using comparative genomic tools, the dynamics of functional transposable elements (TEs) on sex chromosomes can be identified in two aspects: (1) sequence conservation and (2) co-localization with regions with a known genomic function [79,80]. The conservation of TE sequences likely passively contributes to sex chromosome degeneration [81]. TE accumulation on the W/Y sex chromosome may start from several insertions in the close vicinity of crucial heterogametic sex-linked genes/loci where TEs can escape removal from the population due to stochastic processes, such as methylation/heterochromatization, ectopic recombination between TEs, or size reduction on W/Y sex chromosomes [62,82,83,84].

In contrast to mammals, birds and many snakes, which have small and degenerate W/Y chromosomes, the heterogametic sex chromosome (W/Y) is substantially larger than the Z/X chromosome in many fishes, reptiles and amphibians, indicating that sex chromosomes in these groups are usually younger, with frequent turnover [20,44,85,86,87,88,89,90]. In the medaka fish Y chromosome, the male-specific sex-determining region (SDR) has undergone duplication after TE insertion into the proto-Y chromosome [91]. In platyfish (Xiphophorus maculatus, Günther, 1866) [92], a higher repeat content was found in the SDRs of the X (43% repetitive sequences with 36% TEs) and Y (49% repetitive sequences with 32% TEs) chromosomes than in the whole genome (23% repeats with 21% TEs). The X and Y SDRs have higher densities of repeats and TEs than the whole X chromosome [93,94]. This suggests that TEs occupy the whole Y SDR, but are more compartmentalized on the X SDR [95]. A similar phenomenon occurs in the W chromosome of the half-smooth tongue sole (Cynoglossus semilaevis, Günther, 1873) [96], which has a substantially higher TE content than the Z chromosome (Chen et al. 2014) [97]. This indicates that TEs are recently strongly active in the X and Y or Z and W SDRs, thus potentially increasing the size and divergence of the currently small non-pseudo-autosomal region [98,99]. In the genome of the African clawed frog (Xenopus laevis, Daudin 1802) [100], recombination between the W and Z sex chromosomes stopped recently, with a strong accumulation of TEs in W-specific regions [101]. Potentially, the accumulation of TEs interfered with chromosome pairing during prophase I of meiosis and suppressed recombination, leading to an increase in sex chromosome size during the early phase of their differentiation, whereas a size reduction occurred later in their evolution and thus the W/Y sex chromosomes became smaller [102,103,104,105,106].

The transposition of a pre-existing SD locus onto a new chromosome may lead to the emergence of a new sex chromosome [107]. TEs displace DNA and may promote the emergence of new SD loci [108,109,110]. Recent studies have shown that the master sex-determining (SD) gene, sdY, is conserved in many species. However, it is not consistently located on the same linkage homology but seems to behave like a “jumping gene” [109,111,112]. Analysis of the boundaries of the moving region that harbors sdY revealed the presence of several TE sequences, based on which a mechanism involving TE-associated transduction has been proposed [113]. Similarly, in medaka, TEs co-localize with regions that have a known genomic function and play an active role in the evolution of sex chromosomes, and the integration of DNA via TEs contributed to the regulation of the newly emerged SD gene, dmrt1bY. The SD gene of medaka arose from a duplication event of the autosomal dmrt1a gene [11]. The two dmrt1 genes exert their functions at different times during gonad development: dmrt1bY in the SD stage and dmrt1a in the differentiating and adult testes [114]. This phenomenon may be linked to a rapid turnover of the sex chromosomes in the lineage, e.g., the genus Oncorhynchus has six independent sex chromosome pairs that originated 6–8 million years ago (MYA) [112].

The variety of Y chromosomes probably results from TEs (TC1-like transposase and non-LTR retrotransposons) in the flanking regions of the SDRs, which can move throughout the genome [110]. The movement of male/female SDRs among autosomes may prevent sex chromosome degradation and deleterious TE accumulation [95]. Alternatively, TEs have been shown to play a role in heterochromatization of the W/Y chromosomes and dosage compensation mechanisms [97,115,116,117,118,119,120,121]

TEs may play a role in resolving dosage-related gene expression problems in several species by promoting the silencing and condensation of sex chromatin in X (or Z) chromosomes, known as “the hitchhiking effect of favorable mutations” [35]. Interspersed repeat elements such as L1s in humans and mice have been suggested to enhance the inactivation process, thus promoting the heterochromatin state [122]. X chromosome inactivation in mammals, also termed “lyonization”, is a dosage compensation process in which one of the two X chromosomes is inactivated in XX females during early embryogenesis, preventing gene overexpression in comparison to males, which have a single X chromosome [123]. In marsupials, such as opossums, the paternal X chromosome is inactivated, whereas in placental mammals, a random X chromosome is inactivated by a long non-coding RNA produced by Xist [124]. Interestingly, in opossums, L1s show equally frequent interruptions on the X chromosome and autosomes, whereas in humans, L1s are less frequently interrupted on the X chromosome than on the autosomes [125]. On the human X chromosome, L1-poor regions contain genes that escape X inactivation and are physically distant from Xist-silenced regions [126]. This suggests that L1s play a role in the spreading of X chromosome silencing by recruiting Xist RNAs, which play a general role in the regulation of X-gene expression. This hypothesis has been tested in the spiny rat (Tokudaia osimensis, Abe, 1934) [127], of which males and females are XO [128,129]. A similar high concentration of LINEs was observed on both male and female X chromosomes, whereas no dosage compensation by X inactivation is required. This suggests that LINEs are not required on the X chromosome [130]. It is unclear whether the unique TE distributions patterns on the X chromosome are the cause or consequence of inactivation (or both). One possibility is that L1 accumulation on the X chromosome may be only a by-product of reduced recombination. Further investigation of the role of LINEs in X chromosome inactivation in mammals and other organisms is necessary.

Recently, the TE drbx1 was found to be inserted in an intron of the X-linked region encoding the SD gene dmrt1 in Siamese fighting fish (Betta splendens, Regan, 1910 [131]) [132]. This structural change was associated with a shift in the epigenetic silencing of X-dmrt1 during the critical sex determination stage. Similar mechanisms have been found in plants, e.g., in melon (Cucumis melo L.), and TE-induced methylation of the promoter of the transcription factor CmWIP1 has been shown to suppress expression and thus realize sex determination [133].

Some TE families are differentially expressed in specific tissues or conditions between sexes [81]. In platyfish, an accumulation of Texim genes is observed on the Y chromosome [134]. These genes are physically associated with a Helitron, which may have spread the Texim sequences on Y, but not X. The higher density of TEs on Y may suggest that they regulate some key sexual developmental genes and, consequently, impact sexual development [81]. In the medaka genome, a dmrt1-binding element in the promoter of the SD gene dmrt1bY mediates downregulation through its own gene product and autosomal dmrt1a. This dmrt1-binding silencer was introduced in the dmrt1bY promoter through the insertion of a novel TE, termed Izanagi, which is present in multiple copies in the genome and acts as a transcription factor-binding site [65]. This event contributed to the transcriptional rewiring of the new SD gene that created evolutionary innovations [135,136]. A recent study assessed TE expression on sex chromosomes of Drosophila melanogaster (Meigen, 1830) [137] and found specific enrichment of expressed TEs on the Y chromosome that were depleted on the X chromosome [138]. This result suggested that high-level expression of Y-specific TEs was associated with the activation of spermatocyte-specific and Y chromosome-specific transcriptional pathways.

The discovery of sex determination in the domestic silk moth (Bombyx mori, Linnaeus, 1758 [139]) demonstrated the regulatory function of the W chromosome, where TEs are involved in physical and biochemical interactions with thousands of autosomal protein-coding genes [116,118,140]. In the house fly (Musca domestica, Linnaeus, 1758 [119]), approximately two-thirds of the Y-linked scaffolds contain sequence similarities with TEs (Meisel et al. 2017 [120]). Neo-X chromosome formation through a domesticated non-autonomous Helitron has been identified in Drosophila miranda (Dobzhansky, 1935 [121]), and its role in the expression of X-linked genes has been revealed [117]. By contrast, the formation of male-specific lethal binding sites contributes to the dosage compensation process [117]. In mammalian genomes, TEs are hypomethylated in females compared to males, implying that oocytes are more resilient to TE transposition than the male germline. This may be linked to the lifelong division and numerous cell divisions of spermatogonial cells in contrast to oocytes. More cell divisions may allow for many deleterious insertions in the male germline [141].

In addition to the effects of TEs on host gene expression, there exist genomic differences between males and females in terms of TE profiles that impact sexual development [81]. Subsequent cycles of W/Y chromosome degeneration and rejuvenation may differ in length of evolutionary time. During W/Y chromosome degeneration, the appearance of a large number of TE transcripts in a cell may be more or less sudden [95]. In species/populations with a fixed W/Y chromosome loss or with frequent turnover, the period of TE transcript increase may be too short to induce trans-generational and constant preparedness of the genome defense system for TE invasion. When genome defense is inefficient, W/Y chromosome degradation may proceed at a faster rate, and the cycle of chromosome rejuvenation may be shorter [95]. The W/Y chromosome may become a substantial source of functional TE transcripts in the cell and threaten whole-genome stability [142,143,144,145]. The high density of TEs on the W/Y chromosome may serve as a hallmark for heterochromatin marks, which result in different levels of chromatin repression in the rest of the genome and in differential gene expression between males and females [35,36]. TE accumulation has been observed on specific Y or W sex chromosomes and on the corresponding regions of the X and Z chromosomes [83,146]. It is well established that TEs are abundant and play roles in both heterogametic and homogametic sex chromosomes; however, the question remains as to whether TEs are passive or active drivers of sex chromosome differentiation and evolution.

2. Ty3/Gypsy TEs and Sex Chromosome Differentiation