Adaptive evolution of testis-specific histones: a SeXY conflict?

From the Malik Lab, Basic Sciences Division

The genetic material of eukaryotic cells is organized into units known as nucleosomes, each containing ~150 base-pairs of DNA wound around a complex of eight histone proteins. Typical nucleosomes contain two copies each of four histones (H3, H4, H2A and H2B) that are post-translationally modified to regulate chromatin structure and gene expression. In certain contexts, variant histone proteins replace the canonical ones, often performing specialized functions. One example is the family of short H2A variants, which lack portions found in most histones and are expressed only in mammals during male germ cell development. Based on their amino acid sequences, short H2A variants are thought to promote looser wrapping of DNA around the histone octamer, but their functional role in gametogenesis and their evolutionary trajectory are only partially understood.

To better understand the diversity and function of short histone H2A variants, Dr. Antoine Molaro and Dr. Janet Young from the Malik Laboratory (Basic Sciences Division) used publicly available sequence databases to perform a detailed phylogenetic analysis of this protein family. Their work, recently published in Genome Research, identified two novel evolutionary branches (clades) of H2A variants and revealed that short H2As are evolving rapidly.

phylogenetic tree of H2A variants in mammals
Phylogenetic tree of short histone H2A variants in mammals Image provided by Dr. Antoine Molaro

Prior to the Malik lab’s work, three clades of short H2A variants had been identified in mammals. These clades are denoted H2A.B, H2A.L and H2A.P. By collecting and comparing sequences similar to the known H2A clades, Dr. Molaro and Dr. Young found an additional short H2A variant, designated H2A.Q, as well as a distinct, normal-length H2A gene in platypus and marsupial genomes, designated H2A.R, that likely evolved from the same common ancestor as the short H2As. Based on accepted mammalian phylogeny and the principle of parsimony (i.e., choosing the simplest possible solution), the authors concluded that the common ancestor of placental mammals had at least three copies each of H2A.B and H2A.L as well as one copy each of H2A.P and H2A.Q.

By analyzing publicly available RNA-seq datasets, the authors confirmed that H2A.Q is indeed expressed in extant dogs and pigs and H2A.R is expressed in platypuses and opossums. In all cases, expression was restricted to the testes, consistent with previous reports that the short H2A variants are only expressed in the male germ line. Interestingly, humans have an H2A.Q gene but the authors found no evidence of its expression in testes RNA-seq data from humans or primates.

Comparison of the amino acid sequences of each short H2A clade to canonical H2A revealed a number of differences with potential functional consequences. For example, loss of conserved arginine residues and other charge-altering changes are expected to weaken the interaction of short H2A variants with DNA and reduce the packing density of the corresponding nucleosomes. In addition, high divergence within an important loop region may indicate that short H2A variants cannot dimerize with canonical H2A within nucleosomes, while large differences in the C-terminal tail across all short H2A variants suggest that each one may interact with different regulatory proteins.

The researchers were surprised to find that none of the examined genomes contained the same combination of short H2A variants and that short H2A genes often harbor inactivating mutations. These results suggest that the short H2A variants may perform redundant functions with each other or be non-essential. In addition, duplicate alleles of the same short H2A variant within a given genome were found to frequently undergo gene conversion, meaning that over time these genes became more similar to each other than to the same genes in related species. An interesting exception to these observations is in rodents, in which H2A.L has not undergone gene conversion and male mice missing only one of the H2A.L alleles are sterile; in addition, rodents are the only case in which the short H2A genes are not located on the X chromosome.

Dr. Molaro found that the rate of amino acid evolution is quite high among short H2A variants, on par with the rapidly evolving centromeric histone cenH3. This result suggests that these genes are either evolving under positive selective pressures that favor diversity or that they are not under strong evolutionary constraints. For H2A.P and H2A.B, mutations were found to cluster at specific sites rather than be randomly distributed throughout the genes, indicating that positive selection is likely to be the correct explanation for these clades. "This result is quite intriguing since histones are thought to be some of the oldest and most conserved proteins in eukaryotes", says Dr. Molaro.

Together, the Malik lab's results indicate that the short H2A histone variants have undergone adaptive evolution and are thus likely to influence mammalian fitness. Given their testis-specific and X-linked expression, the authors propose that the short H2A variants may be participating in a genetic conflict, perhaps by antagonizing a Y chromosome-linked factor. As stated by Dr. Molaro, "these histones appear to be engaged in a new form of evolutionary arms race involving the X chromosome. This has the potential to change our perspective of mammalian reproduction." Future studies will characterize the functional role of short H2A variants in shaping nucleosome structure during male germ cell development in different mammalian species.

Molaro A, Young JM and Malik HS. (2018) Evolutionary origins and diversification of testis-specific short histone H2A variants in mammals. Genome Research. 28:460-473

This work was supported by the Damon Runyon Cancer Research Foundation, the National Institutes of Health, and the Howard Hughes Medical Institute.