I previously described where in a genome we would expect to find sexually antagonistic genes. Briefly, depending on whether a gene is male-biased or female-biased and whether beneficial mutations are dominant or recessive, we can predict whether these sexually antagonistic genes will be on X chromosomes or autosomes. As I mentioned in that post, the theoretical results can only be translated into realistic predictions if we have reliable estimates of the relevant parameters. We do not have such estimates, but we can study the distribution of sex-biased genes throughout genomes. The results of such experiments can then be used to infer how selection acts upon sexually antagonistic genes.
With the publication of complete genome sequences from multiple different animals, gene content on X chromosomes on autosomes could be compared. Manyuan Long's research group performed some of these early studies, including one where they looked at duplicated genes in the Drosophila melanogaster genome [Retroposed New Genes Out of the X in Drosophila]. The found an excess of duplicated genes with an ancestral copy on the X chromosome and a derived copy on an autosome. When they examined the expression patterns of these genes, they found that many of the autosomal copies of genes duplicated from the X chromosome were transcribed in the testes. It appeared that male-biased genes were accumulating on the autosomes and selected against on the X chromosome.
However, to explain the pattern, Long and colleagues did not invoke sexual antagonism. Instead, they pointed out that the X chromosome is inactivated during spermatogenesis. Therefore, testes expressed genes will be selected against if they are located on the X chromosome because they won't be transcribed during some stages of spermatogenesis. But not everyone agreed with them. Chung-I Wu and Eugene Xu argued that selection against male-biased genes may have driven the testes expressed genes off the X, which then allowed for the evolution of X inactivation [Sexual antagonism and X inactivation -- the SAXI hypothesis].
We'll get back to the conflict between X inactivation and sexual antagonism in a little bit. But first we'll take a look at where the male genes are in other organisms. Manyuan Long's lab also looked at duplicated genes in the human and mouse genomes [Extensive Gene Traffic on the Mammalian X Chromosome]. Just like in Drosophila, they found lots of genes duplicated from the X chromosome to the autosomes, and the autosomal copies were expressed in the testes. But they also found an excess of genes duplicated onto the X chromosome. One model of sexual antagonism predicts that the X-linked copies will be female-biased. Long and colleagues did not find an excess of ovary expressed genes on the X chromosome, so it was unclear why an excess of genes were duplicated onto the X chromosome.
What happens when we explicitly look for male-specific genes? David Page and colleagues did that with mouse transcripts, and they found that genes expressed only in spermatogenesis tend to be found on the X chromosome [An abundance of X-linked genes expressed in spermatogonia]. While the studies on duplicated genes found that male genes accumulate on autosomes, this study discovered that the X chromosome is a disproportionate source of male genes . . . in mice.
But the same pattern does not hold in Drosophila. Brian Oliver's group compared gene expression in male and female flies and found an underrepresentation of male-biased genes on the X chromosome [Paucity of Genes on the Drosophila X Chromosome Showing Male-Biased Expression]. This is consistent with the studies of duplicated genes. There are two different explanations for this pattern: either sexual antagonism drives male-biased genes off the X chromosome or X inactivation during spermatogenesis prevents testis-expressed genes from accumulating on the X chromosome.
Oliver and colleagues came up with a clever way to compare those hypotheses [Demasculinization of X chromosomes in the Drosophila genus]. They collected gene expression data from males and females of six additional Drosophila species -- which was possible thanks to the sequencing of the genomes of those species. Just like in D. melanogaster, all species showed a paucity of X-linked male-biased genes. That can be seen in the first panel of the figure below.
a, Percentage of genes with female-biased (red), non-biased (grey) and male-biased (blue) expression on chromosome arms. Muller's elements and arms are indicated (X chromosomes in bold). Significant under-representation from a random distribution of genes in an expression class (chi-squared test) is noted (*P
[Click to enlarge.]
They then did something clever with the data; they excluded the testis-biased genes (that is genes expressed in testes significantly more than somatic tissue). If X-inactivation drove the male-biased genes off the X chromosome, excluding testis-biased genes should eliminate the discrepancy between the X chromosome and autosomes. If, however, selection against male-biased genes on the X chromosome drives the demasculinization of the X chromosome, removing testis-biased genes from the dataset should have no effect on the discrepancy. It turns out that removing the testis-biased genes does not eliminate the deficiency of male-biased genes on the X chromosome. Oliver and colleagues conclude that natural selection drives male-biased genes off the X chromosome independently of the effects of X inactivation during spermatogenesis.
What are we left with? Well, there is evidence that male-biased genes are selected against on the X chromosome. This supports the hypothesis that beneficial mutations in male-biased genes are dominant. However, this does not hold across all taxa -- see the mouse study where testis-specific genes were found on the X chromosome. Could it be that we can't generalize the dominance of beneficial mutations for different classes of genes (perhaps the testis specific genes have recessive beneficial mutations)? Or maybe the dominance of beneficial mutations varies across taxa.
Betran E, Thornton K, Long M. 2002. Retroposed New Genes Out of the X in Drosophila. Genome Res 12: 1854-1859 doi:10.1101/gr.6049
Emerson JJ, Kaessmann H, Betran E, Long M. 2004. Extensive Gene Traffic on the Mammalian X Chromosome. Science 303: 537-540 doi:10.1126/science.1090042
Parisi M, Nuttall R, Naiman D, Bouffard G, Malley J, et al. 2003. Paucity of Genes on the Drosophila X Chromosome Showing Male-Biased Expression. Science 299: 697-700. doi:10.1126/science.1079190
Sturgill D, Zhang Y, Parisi M, Oliver B. 2007. Demasculinization of X chromosomes in the Drosophila genus. Nature 450: 238-241 doi:10.1038/nature06330
Wang PJ, McCarrey JR, Yang F, Page DC. 2001.An abundance of X-linked genes expressed in spermatogonia. Nat Genet 27: 422-426 doi:10.1038/86927
Wu C-I, Xu EY. 2003. Sexual antagonism and X inactivation -- the SAXI hypothesis. Trends Genet 19: 243-247 doi:10.1016/S0168-9525(03)00058-1
It's hard to draw exact parallels between Drosophila and mouse sex chromosome meiotic behaviour. In particular, while there does seem to be meiotic sex chromosome inactivation in Drosophila, it's not clear that this process corresponds directly to mammalian meiotic inactivation.
In particular, it has recently become clear that in mammals, meiotic sex chromosome inactivation is a special case of a more generalised silencing of unsynapsed chromatin during meiosis. Anything unsynapsed during mammalian meiosis becomes transcriptionally silenced, and this silencing is at least partially maintained into postmeiotic stages. Since Drosophila male meiosis is asynaptic and achiasmate, the same process clearly cannot be occurring!
In mammals, where the silencing applies more widely than to just the sex chromosomes, the SAXI argument cannot apply. There is no reason why sexual antagonism should remove genes from asynaptic autosomal loci and thus enable the silencing to evolve.
Looking at the mouse X in more detail, the enrichment for testis specific genes applies only in early (premeiotic) germ cell stages, and the X becomes then becomes inactive in spermatocyte stages. This inactivation is largely maintained into haploid spermatid stages, however some germ cell specific genes on the X do reactivate. How these genes are selected for reactivation is an open topic.
There is also an important distinction to be drawn here between male-biased genes which are neutral in females (e.g. genes which are expressed specifically in male tissues and never expressed in females) and genes which are actually selectively deleterious in a female context. There would seem no reason to remove the former class of genes from the X chromosome - if they're never switched on at all in female tissues, it's not going to matter which chromosome they're located on.
Excellent points about the difference between Drosophila and mammalian sex chromosomes. Not only are there striking differences in X inactivation between the two taxa, it wasn't even clear that there was X inactivation in Drosophila spermatogenesis until very recently. These are definitely not homologous systems.
The issue with difference between sexually antagonistic and male-specific genes is important. I agree that there won't be selection against male-specific genes while there in females (although there may have been ancient selection on these genes to silent them in females). However, how do you explain the excess of testis-specific genes on the X chromosome in mouse? Recessive beneficial mutations?
Hemizygous exposure of recessive beneficial mutations is certainly a plausible mechanism.
I guess you could also argue that there may be a class of genes that are necessary for testis function but that must be switched off during meiosis. It might be more efficient to relocate these genes to a chromatin region that's silenced in meiosis, rather than evolving a specialised promoter for each individually.
One thing that's not widely appreciated is just how strange the testis transcriptome is: something like 5% of the entire genome is specifically expressed in male germ cells!
5% of the entire genome is specifically expressed in male germ cells!
Is this for mammals? Because, in Drosophila, a huge portion of the genome is expressed in testes, not-specifically. Transcription is up-regulated across the entire genome.
That's in mice, but I'd be surprised to see anything different in other mammals.
Schultz N, Hamra FK, Garbers DL. A multitude of genes expressed solely in meiotic or postmeiotic spermatogenic cells offers a myriad of contraceptive targets. Proc Natl Acad Sci U S A. 2003 Oct 14;100(21):12201-6.
What proportion of Drosophila genes are specific to the testis transcriptome? I've not personally worked in flies, but my feeling is that it wouldn't be wildly different from mammals.
Sperm really are strange cells. They have an acrosome and a flagellum - found in no other cell type in mammals or Drosophila. The genome itself is packaged up in novel histones/protamines. All sorts of epigenetic signals get erased and reset specifically in the germline. All that work's got to be done by something! So I don't find it particularly surprising that a large proportion of the genome is testis-specific. And that's bound to have knock-on effects on genome evolution as a whole, particularly on the sex chromosomes.