Introduction

Dioecy is the most common sexual type in animal kingdom. In contrast, only 5–6 % of angiosperm species are dioecious plants (Renner 2014). Similar to that in mammals, the sex chromosome determination system in plants is the main mechanism for dioecious plant reproduction (Westergaard 1958). However, sex chromosomes in plants have evolved much more recently; the sex chromosomes in mammals occurred approximately 300 million years ago (Graves 2006), whereas the sex chromosomes of Carica papaya and Coccinia indica evolved only less than 7 million years ago (Wang et al. 2012; Sousa et al. 2013). Such species with “young” sex chromosomes offer a unique opportunity in gaining access to the very early stages of sex chromosome evolution (Charlesworth 2013, 2015).

Repetitive DNA sequences, mainly transposable elements (TEs) and tandem repeats (satellites, minisatellites, and microsatellites), are a major fraction of eukaryotic genomes (Mehrotra and Goyal 2014). These sequences were previously regarded as “junk DNA.” At present, they are considered to have important functions in genome evolution (Cioffi et al. 2015; Contreras et al. 2015). Recent studies have revealed that repetitive sequences may play vital roles in many steps of sex chromosome evolution (Ellison and Bachtrog 2013; Zhou et al. 2013; Faber-Hammond et al. 2015). Thus, what roles do repetitive sequences play in plant sex chromosome evolution? What is the relationship between repetitive sequences and recombination suppression, the critical step of plant sex chromosome evolution? What mechanism of repetitive sequences is involved in plant sex chromosome evolution?

Recent efforts to characterize repetitive sequences in plant sex chromosomes add a new level of resolution to our understanding of the role of repetitive sequences in sex chromosome evolution. This review focuses on these advances in the function of repetitive sequences in the structure and evolution of plant sex chromosomes. Epigenetic modification is also discussed as a potential mechanism of the repetitive sequences that contributes to sex chromosome evolution. Given that knowledge on dioecious plants is limited, we include contributions from animals where appropriate.

Origin and evolution of plant sex chromosomes

In dioecious plants, most species belong to the XY chromosome system, in which males (XY) are heterogametic sex, and females (XX) are homogametic, such as in Silene latifolia, Rumex acetosa and C. papaya. In other species, such as Fragaria virginiana and Silene ottites, females are heterogametic (ZW) and males are homogametic (ZZ). Male heterogamety apparently evolves more often than female heterogamety both in animals and plants. For convenience, we shall mostly consider male heterogametic systems in the following text; similar considerations apply to female heterogamety.

All sex chromosomes are generally believed to be derived from pairs of autosomes (Ohno 1967). In plants, dioecy is usually derived in lineages from hermaphrodite plants. The evolution of hermaphroditism to dioecy may start with the emergence of two mutations, namely, a recessive loss-of-androecium mutation and a dominant gynoecium-suppressing mutation. These two loci must be linked at one chromosome pair to stabilize the sexes. Immediately following the emergence of the sex-determining gene(s) may be the suppression of the recombination between the two sex-determining loci and their vicinity regions. Once the recombination is stopped, the non-recombining region can expand rapidly probably because of the accumulation of repetitive sequences and duplications, thereby causing the Y chromosome to be larger than the X chromosome. However, with the continuation of the evolution process, the severe degeneration of the Y chromosome causes the loss of function of most genes; meanwhile, loss of nonfunctioning sequences results in Y chromosome shrinkage. Ultimately, the Y chromosome is reduced severely and may even disappear at some point in the future. Eventually, a new sex determination system that is based on the X-to-autosome ratio evolves (for review see Charlesworth et al. 2005; Charlesworth 2013).

The evolution of sex chromosomes is a gradual and continuing process. The key events of evolutionary processes include sex determination gene emergence, recombination suppression, repeat accumulation, Y chromosome degeneration, and X chromosome dosage compensation, among which, recombination suppression is a vital and early event of the sex chromosome evolution process.

The suppression of the recombination between the sex determination loci is most likely selected to avoid the production of neuters or hermaphrodites (Charlesworth and Guttman 1999). The recombination of the male-specific region of the Y chromosome (MSY) is reduced or restricted even in species with incipient sex chromosomes (Ming et al. 2011). The extent of recombination suppression generally reflects the stage of sex chromosome evolution. For homomorphic sex chromosomes, which are considered relatively young sex chromosomes, the nonrecombining region is usually very small. For example, Asparagus officinalis, which has a pair of very young sex chromosomes, only have a very small male-specific region (Ming et al. 2011) or the male-specific region on the Y chromosome may be lacking entirely (Charlesworth 2015). Although its sex determination loci and adjacent markers could recombine, the recombination frequency is reduced (Jamsari et al. 2004; Telgmann-Rauber et al. 2007). Papaya also has homomorphic sex chromosomes, but they are more diverged than those of asparagus. High-density linkage mapping of the papaya genome has shown that 225 of the 347 markers in linkage group 1 (LG1) co-segregate with sex; this condition reveals the severe suppression of recombination around the sex determination locus (Ma et al. 2004). S. latifolia has a pair of heteromorphic sex chromosomes, which represents advanced evolutionary stages in plants. The recombination of almost all of the regions between X and Y chromosomes is suppressed, except for the pseudoautosomal region (Armstrong and Filatov 2008). However, it is worth noting that small MSY regions do not always represent young sex chromosomes because of the different evolutionary rates of the non-recombining regions in different species (Zhou et al. 2014; Geraldes et al. 2015).

Accumulation of repetitive DNA in plant sex chromosomes

The accumulation of repetitive sequences is a common feature of sex chromosomes in animals (Steinemann and Steinemann 1992; Erlandsson et al. 2000; Böhne et al. 2012; Yano et al. 2014; Matsuda 2015). For instance, high concentrations of repetitive sequences in the Y chromosome have been observed in humans (Erlandsson et al. 2000; Skaletsky et al. 2003). Even in the neo-Y of Drosophila miranda, which was formed only about 1–2 million years ago, has already accumulated a large number of TEs (Bachtrog 2005; Bachtrog et al. 2008). In the past several years, repetitive sequences have been identified from sex chromosomes in several dioecious plants (Table 1). The organization and characterization of repetitive sequences in sex chromosomes were extensively studied in model dioecious plants, such as C. papaya, S. latifolia, and R. acetosa.

Table 1 Summary of the repetitive sequences detected on the plant sex chromosomes
Full size table

C. papaya is a leading dioecious model largely due to its small and sequenced genome, particularly sex chromosomes (Ming et al. 2008; Wang et al. 2012; VanBuren et al. 2015). The Y chromosome has a small MSY region (8.1 Mb, represents only 13 % of the entire Y chromosome) (Yu et al. 2007). A high density of repetitive sequences has been observed in the MSY region (Liu et al. 2004; Yu et al. 2007). Further studies have shown that repetitive sequences occupy 79.3 % of the hermaphrodite-specific region on the Y chromosome (HSY) and 79.2 % of MSY, whereas 67.2 % in the X chromosome counterpart; these values are all obviously higher than the ratio of repetitive sequences in the entire genome (51 %) (Wang et al. 2012; VanBuren and Ming 2013a; VanBuren et al. 2015). The TEs of the HSY are mainly Ty1-copia and Ty3-gypsy retrotransposons, and the Ty3-gypsy retrotransposons in X chromosomes and HSY regions are nearly twice those in the autosomes (Wang et al. 2012). The expansion of the sex determining region may be associated with the accumulation of Ty3-gypsy elements (Na et al. 2014). Furthermore, 21 sex-specific repeats were obtained from the sex determining region: 20 from HSY and one from the X chromosome. The 21 repeats are absent in autosomes and are related to the expansion of the sex determining region, suggesting the crucial roles played by sex-specific repeats in sex chromosome differentiation (Na et al. 2014).

In another model species S. latifolia, the origin of sex chromosomes is less than 10 million years ago (Nicolas et al. 2005). The Y chromosome is the largest chromosome in male metaphase, and it has accumulated a large number of repetitive DNAs (Table 1). Transposon display analysis of S. latifolia Y chromosomes indicated that TE insertions are present at higher predicted frequencies at sites on the Y chromosome than elsewhere in the genome (Bergero et al. 2008b). This field was reviewed extensively by Kejnovsky et al. (2009) and Vyskot and Hobza (2015). Thus, they are not described in this article.

The sex chromosomes of R. acetosa are XX/XY1Y2, with sex determination based on the X/A ratio; the sex chromosomes appeared about 15–16 million years ago (Navajas-Pérez et al. 2005). Early in 1994, a study on a tandemly arranged repetitive DNA sequence mapped onto both Y and X chromosomes was conducted (Rejón et al. 1994). Several Y chromosome-specific repetitive sequences, such as RAYSI, RAYSII, and RAYSIII, were also reported (Shibata et al. 1999; Mariotti et al. 2009). The three satellite DNAs are completely separated from one another. However, they evolved from an ancestral satellite DNA, suggesting the evolutionary role of satellite DNAs in Y chromosome differentiation and heterochromatization in the Rumes species (Mariotti et al. 2009). Another satellite DNA, namely, RAE180, was located in Y1, Y2, and one autosome with particular expansion on the Y1 chromosome (Shibata et al. 2000). RAYSI and RAE180 are the main components of the Y heterochromatin (Shibata et al. 1999, 2000). The distribution patterns of a large number of microsatellites were studied, and the results showed the expansion of most of these microsatellites on both Y1 and Y2 chromosomes of R. acetosa (Kejnovský et al. 2013). Analysis based on low-pass 454 sequencing revealed that copia and gypsy retrotransposons are the most abundant, and DNA transposons and non-long terminal repeat retrotransposons are relatively rare in the genome. A gypsy subfamily accumulated in Y1 and Y2 chromosomes. This research also detected two novel satellite DNAs, namely, RA160 and RA690. RA160 localizes mainly in the X chromosome, whereas RA690 dominates in the Y1 chromosome (Steflova et al. 2013).

Repetitive sequences were also discovered from the sex chromosomes of M. polymorpha (Okada et al. 2001), Cannabis sativa (Sakamoto et al. 2005), Bryonia dioica (Oyama et al. 2010), and Humulus lupulus (Divashuk et al. 2011). Recently, Akagi et al. (2014) detected a male-specific sex-determinant candidate gene OGI, which encodes a small RNA that regulates MeGI gene transcription in Diospyros lotus. Sequence analysis of 150 kb surrounding OGI revealed highly repetitive sequences and the presence of male-specific regions (Akagi et al. 2014). In our laboratory, we identified a number of TEs with sex-biased abundance in A. officinalis by sequencing the male and female genomes (Li et al. 2014). These sequences might be located in the sex chromosomes. We also identified several potential male-specific TEs from dioecious plants Spinacia oleracea and Humulus japonicas through subtracting hybridization and dot blot techniques; chromosome localization using FISH is ongoing. Preliminary results showed that one gypsy-like retrotransposon adjacent to a male-specific marker accumulated higher level in Y than in X chromosomes of S. oleracea (our unpublished data).

By comparing the information presented in Table 1, we found that Y chromosome-specific repetitive sequences are more abundant in the relatively ancient evolved sex chromosomes (Table 1). For example, the Y chromosomes of R. acetosa were established earlier than those of papaya and S. latifolia, and more Y chromosome-specific repetitive sequences were identified. Although we cannot exclude the influence of research extent on a given species, the findings indicate that the dynamic variation in repetitive sequences could have witnessed the differentiation and evolution of sex chromosomes. Thus, what about the roles played by repetitive sequences in plant sex chromosome evolution? Although the present information is far beyond complete understanding, some progress has been made in understanding the possible nature of repetitive sequences against plant sex chromosome evolution.

Roles played by repetitive sequences in plant sex chromosome evolution

One of the main stages of sex chromosome evolution is accumulation of repeats. Accumulated observations suggest that many of the other main stages of sex chromosome evolution, such as recombination suppression, gene degeneration, expansion or shrinkage of sex chromosomes, and dosage compensation, are all related directly or indirectly to repetitive sequences.

Repetitive sequences and recombination suppression

Two mechanisms mainly contribute to the recombination suppression of sex chromosomes; one is chromosome rearrangement, and the other is heterochromatization (Yu et al. 2007; Steinemann and Steinemann 2005a). Repetitive sequences, particularly TEs, may induce either chromosome rearrangement or heterochromatization.

TEs can induce gene mutation (Morales et al. 2015), chromosome rearrangement (Xuan et al. 2011), ectopic recombination, and genome remodeling (Tian et al. 2009; Schaack et al. 2010). TE insertions can promote chromosome rearrangements via ectopic pairing and crossing over between repeated sequences in different locations in the Y chromosome (Tenaillon et al. 2012). A higher degree of rearrangement of Y than X chromosomes has been observed in mammals (Repping et al. 2006) and in the plants papaya (Gschwend et al. 2012; Wang et al. 2012) and S. latifolia (Hobza et al. 2007; Zluvova et al. 2007; Bergero et al. 2008a). Subsequent structural rearrangement could further lead to the suppression of the recombination between the rearranged regions (Charlesworth et al. 2005).

In addition to chromosome rearrangement, heterochromatization is another prominent feature of the Y chromosome; it apparently plays a dominant role in recombination suppression. Heterochromatization may have originated simply as a means to defend the genome from invasive parasitic DNA. The accumulation of repeats induces heterochromatization of the genome region to eliminate the damage caused by TE activity (Steinemann and Steinemann 2005a). Steinemann and Steinemann (2005b) suggested that conversion of the Y chromosome from euchromatin to heterochromatin is probably due to the cumulative retrotransposon insertions. They also considered retrotransposons an ideal driving force of heterochromatization of the Y chromosome of D. miranda. Direct evidence on the claim that TEs cause the heterochromatization of the Y chromosome was provided by a recent study. Zhou et al. (2013) studied the genome, epigenome, and transcriptome of D. miranda and confirmed that heterochromatin formation is triggered by the presence of repetitive DNA derived from TEs. The heterochromatization of newly formed sex chromosomes can not only suppress recombination (Bergero et al. 2008a) but also recruit more TE insertions that can cause the non-recombining region to spread into the flanking regions (Zhu et al. 2003).

The heterochromatization of Y or W chromosomes has been observed in many animals (Skaletsky et al. 2003; Steinemann and Steinemann 1992, 2005a; Cioffi et al. 2012). In plants, the two large Y chromosomes of R. acetosa are cytologically heterochromatic (Shibata et al. 1999; Mariotti et al. 2009). DAPI/C-banding staining indicates that the Y chromosome of C. sativa carries a fully heterochromatic DAPI positive arm (Divashuk et al. 2014). The Y chromosome of M. polymorpha is small and largely heterochromatic (Yamato et al. 2007). In H. japonicas, two large chromocenters observed in some male nuclei probably represent two Y chromosomes condensed to form heterochromatic interphase bodies (Grabowska-Joachimiak et al. 2011). Even in the primitive Y chromosome of papaya, knob-like heterochromatin structures specific to MSY have been observed (Zhang et al. 2008). However, it is surprising that no sign of heterochromatization of the Y chromosome of S. latifolia exists (Vyskot and Hobza 2015). This situation indicates that plant sex chromosome evolution typically follows a similar trajectory. However, different mechanisms may exist, which is in accordance with the independent origin of sex chromosomes in different plant lineages.

Repetitive sequence and morphology and structure of sex chromosome

Expansion ability is an intrinsic property of repetitive sequences, and the non-recombining regions of the young Y chromosome provide an opportunity for expansion (Charlesworth et al. 1994). The accumulation of repetitive DNAs in sex chromosomes may be both the cause and consequence of recombination suppression (Bergero and Charlesworth 2009; Cioffi et al. 2012). Once the recombination is suppressed, repetitive sequences are predicted to accumulate rapidly in the sex chromosomes. Kejnovský et al. (2013) suggested that microsatellite expansion is an early event shaping the Y chromosome, and microsatellites are probably targets for TE insertions. The massive accumulation of repetitive sequences may thus cause structural and morphological diversification of sex chromosomes. Chromosome rearrangements induced by repetitive sequences can change chromosome morphology and lead to visible heteromorphism in sex chromosomes. Furthermore, the abundance of repetitive sequences could cause sex chromosome expansion. The young Y chromosomes found in Drosophila and in several plants are often larger than their X homologues. Even in the homomorphic sex chromosomes of papaya, HSY (8.1 Mb) is larger than its counterpart in the X region (3.5 Mb) mainly because of retrotransposon insertions (Wang et al. 2012). Ectopic recombination based on repetitive sequences may also cause region deletion (Charlesworth et al. 1994), which probably contributes to Y chromosome shrinkage in animals (Fig. 1).

Fig. 1
figure 1

Model depicting that TEs promote heterochromatin formation and influence the morphology of the Y chromosome. TEs (red boxes) are inserted into the Y chromosome. Then, heterochromatin is formed in the TE region and later spreads to the flanking regions. The level of heterochromatization is high when chromatin is close to the TE region. The gradient black-grey boxes represent the extent of heterochromatization, and black indicate high heterochromatization level. Adjacent insertion of a TE can result in the spreading of heterochromatization to a gene and may lead to transcriptional silencing (“on” indicates that the genes are active, whereas “off” indicates that the genes are inactive). A large number of TE insertions can cause the Y chromosome to enlarge. Later ectopic recombination among homologous TEs results in deletions of the chromosome regions, thereby leading to small Y chromosomes

Full size image

Repetitive sequence and Y chromosome degeneration

Y chromosome degeneration is characterized by the formation of a heterochromatic chromatin structure and erosion of genetic activity (Steinemann and Steinemann 2005a). If TEs invade Y-linked genes or are near them and selection against their insertion is ineffective, function loss or reduced expression levels may occur, thereby contributing to genetic degeneration (Hollister and Gaut 2009). Generally, evolutionary old Y chromosome degenerates severely. For instance, the human MSY retains only 3 % of the genes of ancestral autosomes because of genetic decay (Skaletsky et al. 2003; Bellott et al. 2010). Although plant Y chromosomes may not degenerate as completely as the mammal Y chromosomes, a clear sign of degeneration exists in plant Y chromosomes. Analysis of sex-linked genes showed that genes are lost from the Y chromosome or gene expression levels are significantly reduced in Y-linked genes of S. latifolia (Bergero and Charlesworth 2011; Bergero et al. 2015; Toups et al. 2015; Blavet et al. 2015). Minimal but clear evidence showed that TEs participate in the genetic degeneration of the Y chromosome in S. latifolia. Miniature inverted repeat TEs from two active subfamilies that invade the evolving S. latifolia Y chromosome change gene expression and reduce the functions of Y-linked genes (Bergero et al. 2008b). Furthermore, several Y-linked genes having a lower expression level than their X-linked homologues showed evidence for TE insertion (Marais et al. 2008; Cegan et al. 2010).

Repetitive sequences may also be involved in the initial stage of sex chromosome evolution, e.g., repeat insertion is perhaps one of the causes of gene mutation, which leads to the development of the sex-determining gene(s). In addition, TE invasion may facilitate the evolution of dosage compensation. Analysis of the different ages of three X chromosome segments in D. miranda indicated that early TE invasion provides important regulatory sequences that can efficiently attract the dosage compensation complex (Ellison and Bachtrog 2013). However, whether the sex chromosome of plants gives rise to the dosage compensation effect remains a controversial problem.

Although it is revealed that the repetitive sequences play various roles in plant sex chromosome evolution; how the repetitive sequences are involved in these events remains unclear.

Epigenetic modification: possible mechanism of TEs involved in plant sex chromosome evolution

Epigenetic modification, such as DNA methylation and histone tail modifications, represents a defense mechanism to inactivate mobile genetic elements and other types of intrusive DNA (e.g., transgenes), plays a pleiotropic role in eukaryotic cells and organisms (Chinnusamy et al. 2013; Zeng and Cheng 2014; Li et al. 2015b). For example, DNA methylation is essential for the regulation of genes and genomic sequences (Secco et al. 2015; Wilkinson 2015). Epigenetic modification affects nearly all aspects of plant development, physiology, and reproduction (Wollmann and Berger 2012; Kawashima and Berger 2014). TE regulation is commonly involved in almost all silencing processes (Lisch and Bennetzen 2011). Transposition and epigenetics have been intertwined and inseparable since the discovery of TEs. In fact, the first evidence on epigenetic silencing of transposons was from studies that led to their discovery (McClintock 1948). TEs and the derived repetitive sequences are the preferred target of epigenetic modification (Hollister and Gaut 2009; Slotkin et al. 2009). In plants and mammals, the majority of methylated sequences can be transposons and other repetitive DNA elements, and epigenetic modification is often involved in the condensed, heterochromatic structure of repetitive DNA sequences (Vyskot 2005). The epigenetic silencing through the formation of heterochromatin contributes to the prevention of TE proliferation and suppression of unfavorable gene transcription (Lisch 2009; Zhang and Zhu 2011). Heterochromatin formation in Arabidopsis thaliana is related to DNA methyl transferase, H3K9 histone methylase, and histone deacetylase (H4K16) (Soppe et al. 2002; Zemach et al. 2013). Treatment with a demethylation reagent could lead to the activation and transposition of TEs in A. thaliana and fungus (Akiyama et al. 2007).

Given that TEs are inseparable from epigenetic modifications, we posit that TEs participate in the evolution of plant sex chromosomes probably via epigenetic modification. TE insertions can induce the host genome to inhibit the TE activity via epigenetic modification by a set of mechanisms that involve DNA methylation and histone modification, which cause the heterochromatin formation of the TE region. Then, the heterochromatin region of the neo-sex chromosome could expand to adjacent regions, where the active genes involved can be silenced (Fig. 1). Although no direct evidence exists, several studies have provided support for this hypothesis. Recent data demonstrate that epigenetic modification is involved in the regulation of sex expression and differentiation in plants and animals (Martin et al. 2009; Piferrer 2013; Kuroki et al. 2013; Kubat et al. 2014; Li et al. 2015b; Janoušek et al. 1996). A study that implemented chromatin immunoprecipitation analysis of D. melanogaster revealed that the local chromatin environment is substantially enriched for HP1a and K3K9me2 in the presence of a transposon 1360, whereas in the absence of 1360, the levels are low. This finding indicates that the ectopic assembly of heterochromatin in D. melanogaster is triggered by TEs via epigenetic modification (Sentmanat and Elgin 2012). Takata et al. (2007) analyzed 12 TEs of rice through high-throughput sequencing and found that the adjacent sequences of the retrotransposons have a high level of methylation. The extent of methylation depends on the proximity to the TE regions. The S1 retrotransposon is mainly inserted in the hypomethylation regions and serves as the target of methylase to promote the methylation of the flanking sequences (Arnaud et al. 2000). Recent studies that implemented genome-wide epigenetic analyses of maize revealed that DNA methylation and H3K9me2 initiated within TEs can spread beyond the borders of TEs, thereby influencing the chromatin state of nearby sequences (Eichten et al. 2012; Gent et al. 2013). More importantly, Martin et al. (2009) found that a TE is inserted into the transcription factor CmWIP1, leading to the methylation of the flanking transcription factor sequence of the TE and causing the production of unisexual male flowers in the melon. Although this study did not show the contributions of epigenetic modification to sex chromosome evolution, it linked TE, epigenetic modification, and plant sex determination together.

Early in 2004, Jablonka discussed several epigenetic-biased interpretations of the mammalian sex chromosome evolution, such as initiation of the sex determination locus by epigenetic inactivation (Jablonka 2004). Although there was still no direct evidence of the epigenetic modification involving in sex chromosome evolution currently, Jablonka’s opinion provides powerful support for us. Thus, what is the mechanism involved in the epigenetic modification induced by TE insertion in the sex chromosomes?

Almost all eukaryotic systems employ small interfering RNAs (siRNAs) derived from TEs to target TE mRNA for expression silencing and TE chromatin for heritable epigenetic modification (Slotkin and Martienssen 2007; Pontes et al. 2009). Approximately 80 % of siRNA clusters show an overlap with CG methylation, and a good correlation exists between siRNA abundance and CG methylation in the transcribed regions of inverted and tandem repeats (Cokus et al. 2008). In plants, the major siRNA-mediated epigenetic pathway is RNA-directed DNA methylation (RdDM), which function in TE silencing and heterochromatin induction (Matzke and Mosher 2014; Feng and Michaels 2015). Briefly, DICER-LIKE RNase enzymes produce 24 nt siRNAs that guide ARGONAUTE (AGO) and other downstream proteins to the siRNA-generating genomic loci and other loci homologous to siRNAs, thereby promoting and maintaining DNA and histone methylation (Matzke and Mosher 2014; Cui and Cao 2014; Feng and Michaels 2015; Matzke et al. 2015). A recent study on D. melanogaster revealed that siRNAs from an X-linked satellite repeat promote X chromosome recognition, indicating the important roles of siRNAs in sex chromosome evolution (Menon et al. 2014).

Thus, we speculate that siRNAs might be involved in epigenetic modification triggered by TE in sex chromosome evolution. TE insertion into sex chromosomes could give rise to siRNAs. The TE sequences could be transcripted by a plant-specific RNA polymerase, Pol IV. This transcribes is acted upon by an RNA-dependent RNA polymerase (RDR), generating a complementary RNA strand. The double-stranded RNAs formed by sense and antisense transcripts can be cleaved by DCL proteins into siRNAs. The siRNAs are incorporated into AGO proteins and are recruited to target loci by transcripts produced by another plant-specific RNA polymerase Pol V. Subsequently, the siRNA-loaded complex targets the region of chromatin for modification through the recruitment of histone methyltransferase, de novo DNA methyltransferase, and other chromatin-modifying proteins. Modifications in the TE euchromatin result in condensed heterochromatin, which is inaccessible to transcription. This heterochromatin can recruit numerous chromatin-modifying elements, thereby causing the expansion of heterochromatin to the flanking regions (Fig. 2).

Fig. 2
figure 2

Possible mechanism of heterochromatin formation in the sex chromosome via epigenetic modification triggered by TE insertion. TE insertion into sex chromosomes can give rise to siRNAs. Then, siRNAs can be loaded into protein complexes, and the siRNA-loaded protein complex can recruit various chromatin-modifying elements for modification of the target region of chromatin. The epigenetic modifications thus lead to the transition of euchromatin to condensed heterochromatin. The heterochromatin can further expand to the adjacent regions. See text for details

Full size image

The model we propose is mainly focused on plant sex chromosome heterochromatization based on epigenetic modification triggered by TE insertions, in which siRNAs play an important role. Although increasing studies on the mechanism of siRNAs in transposon silencing have been reported, many aspects remain unclear. For example, how is the RNA polymerase responsible for producing the precursor of siRNAs recruited to target sequences? On the other hand, transposons are remarkably diverse within genomes and sex chromosomes. It is unlikely that any one epigenetic pathway exclusively contributes to transposon silencing. Therefore, we emphasize epigenetic modification in sex chromosome evolution, whereas the details of the mechanism are not the point we focused on.

Conclusions and perspectives

The hypothesis that TE insertion triggers the epigenetic modification of the Y chromosome and thus leads to heterochromatization and recombination suppression requires more reliable cellular and molecular evidence. With the modern ability to sequence the sex chromosomes of dioecious plants and their homologous of relative monoecism species, it should be possible to compare the distribution pattern of the repetitive sequences and locate repetitive sequences and genes combined with genetic linkage map. Then, high-resolution fiber-FISH and chromosome and histone methylation detection techniques can be utilized to compare and analyze the DNA and histone methylation profiles of TEs and their flanking regions of sex chromosomes. A synthetic analysis of such data can help elucidate the mechanisms of TEs involved in heterochromatization and recombination suppression.

Many issues about sex chromosomes remain unclear, particularly with regard to the evolution mechanisms. Epigenetic modification is expected to become an active area of the research of sex chromosome evolution in the near future. Amazingly, presently the only sex-determining gene identified in dioecious plants encodes a small RNA (Akagi et al. 2014), implying the contributions of epigenetics to sex determination and sex chromosome evolution.

The sequences that accumulate in sex chromosomes were various. Except for repetitive sequences, it is interestingly to observe that chloroplast DNA accumulates preferentially in the non-recombining region of Y chromosomes (Steflova et al. 2014; VanBuren and Ming 2013b; Hobza et al. 2015). Moreover, it is unexpected to find that several repetitive sequences are ubiquitously distributed in autosomes but are absent in sex chromosomes (Cermak et al. 2008; Kralova et al. 2014; Steflova et al. 2013). Thus, the shaping and evolution of plant sex chromosomes appear to be more complex than what is recognized at present, with a plethora of evolutionary processes are involved, not only the accumulation of repetitive DNAs but also organelle DNAs and not only the accumulation but also the depletion of repetitive sequences.

Author contribution statement

SFL and WJG designed the outline of the article, SFL and WJG wrote the first draft. WJG and GJZ prepared the tables and figures. All authors were involved in revision of the draft manuscript and have agreed to the final content.