Sexual dimorphism in brain transcriptomes of Amami spiny rats (Tokudaia osimensis): a rodent species where males lack the Y chromosome

The undifferentiated gonad and brain are programmed to be either male or female in most sexually reproducing species. Brain sexual differentiation in therian mammals is generally influenced by a surge in gonadal steroid hormones during embryo development followed by a second surge later in adulthood. This paradigm is termed “organization-activational programming” and occurs in a sex-specific manner [1, 2, 3]. The organizational period is typified by spikes in androgens and/or estrogens during development and results in permanent masculinization and defeminization of the brain in males [4, 5, 6, 7]. Masculinization of several brain regions, including the hippocampus, is modulated by aromatization of testosterone to estrogen [4, 5, 6, 7]. Manifestation of sex-dependent behaviors, especially male-specific traits, requires a second spike in steroid hormones at adulthood (“activational period”) [8, 9, 10, 11, 12, 13]. The salient question remains as to the potential contribution of sex chromosome-associated genes in guiding brain sexual differentiation in therian mammals.

Sex-determining region Y (Sry) resides on the Y chromosome in therian mammals, which includes placental mammals and marsupials [14, 15, 16, 17, 18]. This gene serves as a transcriptional activator modulating testis development, and thereby, it was coined the testis determining factor (TDF) [19, 20, 21]. SRY includes an HMG box as a DNA-binding domain [22] that activates in the testes Sry-related box 9 (Sox9) [23], cerebellin 4 precursor gene (Cbln4) [24], transcription factor 21 (Tcf21, former also called Pod1) [25], and neurotrophin 3 (Nt3) [26]. However, the role of Sry in brain sexual differentiation is uncertain.

The four core genotype (FCG) mouse model has been used in elucidating the neural effects of Sry relative to other Y chromosome associated genes [27, 28, 29, 30, 31, 32, 33, 34]. In this model, Sry is deleted from the Y chromosome and re-inserted as a transgene on an autosomal chromosome in both XX and XY chromosome bearing mice [27]. The resulting breeding scheme gives rise to the FCG mouse model as offspring from these mated pairs can be one of four different genotypes: XXSry⁻(karotypically and gonadally female, naturally lacking Sry), XXSry⁺ (karoytypically female but gonadally male due to presence of autosomal Sry transgene), XYSry⁻ (karyoptically male but gonadally female due to deletion of endogenous Sry gene), and XYSry⁺ (karyotypically and gonadally male) [28]. Results to date with this model reveal that Sry and other sex chromosome associated genes interact with steroid hormones to affect neurobehavioral programming. Gonadectomized XX females consume more food and show increased adiposity relative to XY mice [29]. On the other hand, intact XXSry⁻ and XYSry⁻ mice have increased activity levels, consume less food, and show enhanced anxiety-like behaviors [30]. Social and parenting behaviors are also influenced by Sry and other Y chromosome associated genes [32, 35]. This model indicates Sry regulates neural expression of progesterone receptor (PR) in the anteroventral periventricular nucleus (AVPV), medial preoptic area (MPOA), and ventromedial nucleus, gamma-aminobutyric acid (GABA)/serotonin/dopamine-related genes within the frontal cortex, and growth hormone (Gh) expression in the hypothalamus [27, 33, 34].

To ascertain the potential role of the Y chromosome and Sry in regulating brain sexual differentiation, a better approach though would be to examine the brain transcriptome profile in males and females of therian mammals where the males lack a Y chromosome/Sry and both sexes possess an XO system. Such is the case with two Tokudaia (Amami [Ryukyu] spiny rat- Tokudaia osimensis and Tokunoshima spiny rat- T. tokunoshimensis) and two Ellobius (Transcaucasian mole vole- E. lutescens and Zaisan mole vole- E. tancrei) species [36, 37, 38, 39, 40, 41]. Both Tokudaia species are critically endangered, and thus, it is vital to begin to understand how gonad and brain sexual differentiation occurs in them. In Amami spiny rats, two possible genes that might guide testes formation are chromobox protein homolog 2 (Cbx2), which acts upstream of Sry, and Er71, also called Ets variant 2 (Etv2), which is upregulated by Sry in species containing this gene [42, 43]. However, it is not clear which genes or transcript isoforms show sex differences in this Tokudaia species. Based on the current status of Amami spiny rats, it is not permissible to obtain brain samples from embryonic or neonatal individuals. Thus, it is not possible to determine which gene(s)/transcript(s) might guide gonad and potentially brain sexual differentiation in these species. Sex differences in brain expression remain evident in adult mice, birds, and humans [44, 45, 46, 47, 48]. To begin to understand, neural sex differences in Amami spiny rat, the current studies sought to determine which genes/transcript isoforms show sexually dimorphic patterns of expression in the brain of this species. An ancillary goal of this study was to develop the first reference transcriptome library for Amami spiny rat that may aide in future studies designed to identify putative sex-determining gene(s)/transcript(s) in this species. RNAseq analyses was performed on whole brain samples from male and female adult individuals that died accidentally as part of a government-approved mongoose eradication project on the island in which this species inhabits. The overarching hypothesis was that there would be considerable sexually dimorphic differences in the brain transcriptomic profile with signature patterns identified in each of the sexes.

When SRY was identified, it was presumed that it regulated sexual differentiation in all eutherian mammals [14, 15, 16, 17, 18]. However, those species that defy our expectations reveal the complexity and elegance of sexual differentiation that like other biological processes has been subject to evolutionary forces million years in the making. Such is the case for select species within Tokudaia and Ellobius. Intriguingly, 46, XX human male brothers lacking SRY possess testes but are azoospermic [57]. Overexpression of SOX9 in these SRY-negative men likely stimulates male sexual differentiation [58, 59]. Thirty-eight intersex, XX SRY-negative cases have also been reported in pigs [60].

How sexual differentiation occurs in rodent species where males lack a Y chromosome and SRY remains enigmatic. In Amami spiny rats, two possible candidates have been postulated to regulate testes formation: Cbx2, which acts upstream of Sry, and Er71 [42, 43]. The wave of sexual differentiation affects both the gonad and brain [1, 2, 3]. While in the latter steroid hormones, testosterone and estrogen, are the guiding factors, SRY might also be involved [61]. In those species that possess SRY, it continues to be expressed in the adult male brain [62, 63]. At the current time, we cannot obtain embryonic or neonatal brain samples from Amami spiny rat to identify putative sex determining gene(s)/transcript(s) in this species. Our prediction at the outset was that similar to mice, birds, and humans [44, 45, 46, 47, 48], sex differences would be identified in gene/transcript expression in adult Amami spiny rat, albeit many of these many not necessarily be sex-determining genes. The additional objective of the current work was to establish the first reference transcriptomic library for Amami spiny rat that would likely facilitate future work seeking to identify those gene(s)/transcript(s) compensating for the absence of the Y chromosome and SRY in this species. With these limitations and potential strengths in mind, we analyzed the transcriptomes of the adult male and female Amami spiny rat brains to determine which genes and transcripts show sexually dimorphic patterns of expression.

While certain genes differed in expression between male and female Amami spiny rat, as detailed further below, more prominent changes were discovered when comparing sex differences of individual transcripts. The aggregate characteristics of the differentially expressed transcripts are compelling. While the proportional contribution of each splice variant to the total mRNA for a given gene might be low, such variants might demonstrate different potencies in binding and activating various cellular pathways. Studies with mice and other taxa demonstrate that expression of such transcript variants can result in dramatic sex differences and potentially underpin sexual differentiation [64, 65, 66, 67]. In Amami spiny rat, additional phenotypic studies are needed to validate the significance of these findings and whether they broadly differ across various sexually dimorphic organs, including the gonad, and discrete brain regions.

Several transcripts upregulated in the whole brain of males compared to females encode zinc finger proteins. Additionally, gene set enrichment analyses of the genes from which the upregulated transcripts derive identified several zinc-finger related pathways, including Krueppel-associated box (KRAB). KRAB only (KRAB-O) serves as an SRY-interacting protein [68, 69]. Zinc finger proteins are polypeptides exhibiting sequence-specific, nucleic acid-binding role, and serve as trans-acting molecules they govern cellular growth and differentiation. Zinc finger protein Y-linked (Zfy) is another gene located on the Y chromosome, and while not the TDF, it is essential for normal spermatogenesis [70], although an isolated human case has been reported of normal sexual differentiation in the absence of ZFY [71]. Similar to SRY, this gene continues to be expressed in the brain of adult men [62]. In the Amami spiny rat, Zfy, along with eukaryotic translation initiation factor 2 subunit 3, Y-linked (Eif2s3y) and lysine Demethylase 5D (Kdm5d), have been translocated to the X chromosome [72]. The current brain transcriptome studies detected Zfy, but the gene and individual transcripts showed similar expression pattern in males and females.

A corollary gene is present on the X chromosome, Zfx with male mice harboring a mutation of this gene undergoing normal gonadal and male genitalia sexual differentiation but showing markedly reduced sperm counts [73]. Mutant Zfx female mice possess normal ovaries and external genitalia but show diminished oocyte populations and reduced fecundity culminating in abbreviated reproductive potential. Primordial germ cell populations prior to genital ridge migration are reduced in both XX and XY mice offspring and growth impairments are evident in both sexes. The collective data suggest that Zfx might thus be important in both male and female sexual differentiation in eutherian mammals. In Drosophila, the hermaphrodite (her) gene encodes for a C2H2-type zinc finger protein that demonstrates similar properties as Zfy and Zfx and presumably regulates terminal function in sexual differentiation [74]. In the current studies, one form Zfx was shown to be upregulated in the brain of males. However, several other genes encoding zinc finger proteins were also upregulated in males. Thus, we postulate that upregulation of Zfx acting possibly in concert with other C2H2-type zinc finger transcript forms in male Amami spiny rat might compensate for the absence of SRY and serve as the collective factors stimulating male sexual differentiation.

In support of this hypothesis, other zinc finger proteins exert essential sexual differentiation roles in lower animal species and mice after initial sexual differentiation. For instance, the flexuosa (fle1), encoding aC2H2 zinc finger protein, is an important repressor of female sexual differentiation in Podospora anserina [75]. The byr3 zinc finger protein (byr3) gene encoding a protein with seven finger domains affects sexual differentiation pathways in Schizosaccharomyces pombe [76]. In the Japanese wrinkled frog (Rana rugosa), an increase in testosterone causes sex reversal from female to male and accompanying decreases in zinc finger proteins, zinc finger protein 64 (Zfp64) and zinc finger protein 112 (Zfp112) [77]. In sex-reversed and normal mice, zinc finger protein 35 (Zfp35) is essential for normal spermatogenesis [78, 79]. Zinc finger proteins are elevated in human testicular cell lines overexpressing SRY, further highlighting the potential role of such proteins in maintaining sexual differentiation [80].

Sex differences in transcript isoforms encoding zinc finger proteins and other transcripts might originate due to epigenetic variation. In this aspect, intragenic DNA methylation and histone protein modifications can affect RNA polymerase II activity and thereby stimulate creation of alternative splice forms [81, 82, 83, 84, 85]. Heterochromatin protein 1 (HP1) might also play a role in DNA methylation effects on mRNA splicing [86]. DNA methylation and histone protein changes are evident in turtle species demonstrating temperature sex determination [87, 88, 89, 90, 91]. In the case of doublesex and mab-3 related transcription factor 1 (Dmrt1), DNA methylation status tightly associates with temperature and may thus be the master male sex determining gene in red-eared slider turtles (Trachemys scripta) [87].

RNAseq analyses revealed other genes and transcripts that were differentially expressed in the brain of males and females. It is not clear if such genes contribute to the initial brain sexual differentiation process or maintaining sex differences. Due to limited tissue, we could only validate three differentially expressed genes that were chosen based on there highly significant adjusted p value and overall fold change: Svs5, Serpina3n, and Cyp1b1. Of these, Svs5 and Serpina3n expression patterns were identical to that observed with RNAseq with an increase in males and females, respectively. Further work is needed to determine the significance of these and other gene expression changes.

The current studies with adult Amami spiny rat yielded less genes that show sex-dependent brain expression than past studies with mice, birds, and humans [44, 45, 46, 47, 48]. This difference could be attributed to the fact that from an algorithmic standpoint, potentially high level of intra-sample variability would decrease the number of inter-sample genes whose differential expression fell below a FDR of 0.05. The current work presumably identified those genes that show the greatest differential expression between sample types and are potentially the most important.

One of the primary limitations of this study is that in this endangered species, we could only screen random brain tissues in limited number of adult male and female Amami spiny rats. It is not clear if genes and transcripts guiding initial brain sexual differentiation continue to be expressed. The fact that ZFY and SRY are expressed in the brain of adult men shores up this possibility [62, 63]. Usage of induced pluripotent stem cells (iPSC) generated from this species might be useful in validating the current findings [92, 93]. Future work should compare current RNAseq results in the brain of Amami spiny rat to those obtained in Tokudaia species that possess sex chromosomes, e.g., Okinawa spiny rat (Tokudaia muenninki) [94, 95, 96], as this approach may reveal those genes/transcripts essential in sex determination in Amami spiny rat. Should the population of these Tokudaia species recover and/or permission be granted in the future to attempt to breed them in captivity, follow-up transcriptome studies should be conducted with discrete brain regions identified to show sex differences in other species, such as the hypothalamus, and with individuals spanning from the embryonic to adult period. As detailed in the methods, one of the other limitations of this study is that it depends upon whole brain samples from animals who died of accidental causes during a mongoose eradication project. In these cases, brain and other tissues were collected several hours to half a day post-mortem. Thus, these studies should be considered a first pass attempt to obtain a reference transcriptome for this species and examine for sex-dependent differences in gene and transcript expression in the brain of adult individuals. A better understanding of underpinning biological mechanisms in this species may aide in recovery efforts to the point where sufficient animals can be bred in captivity and more precise studies in the brain, gonad, and other sex-dependent organs performed.

In conclusion, transcriptomic analyses in the brain of adult male and female Amami spiny rats shows that sex differences are observed in select genes with more being upregulated in males than females. Comparison of individual transcript forms between males and females yielded more differences and potentially might provide mechanistic insight into how male sexual differentiation occurs in this Y chromosome deficient and SRY-negative rodent species. The most attractive candidates are upregulation of a variety of transcript forms encoding zinc finger proteins, including Zfx. Such sex differences in these and other alternative splice forms might originate due to epigenetic modifications. It remains to be determined if any of these identified gene(s)/transcript(s) modulate initial sexual differentiation in this species.