Study of dynamics of histone H3 variants and H3 proteolysis during mouse spermatogenesis

Spermatogenesis is a complex differentiation process in which male gametes, known as spermatozoa, are produced from spermatogonial stem cells in the seminiferous tubules of the testis. The spermatogenesis process is typically divided into three phases: a mitotic phase, a meiotic phase and post-meiotic spermiogenesis. During mammalian spermiogenesis, haploid round spermatids undergo remarkable morphological changes and an extensive reorganization of chromatin to differentiate into mature spermatozoa. As part of the chromatin reorganization, most histones in round spermatids are replaced by transition proteins and subsequently by protamines. This histone-to-protamine exchange is required for efficient compaction of paternal genome into the sperm head and implicated in male fertility. Nonetheless, previous studies found that 1-10 % histones are still retained at specific loci, particularly at unmethylated CpG-rich promoters, in mouse and human sperm. How spermatid chromatin is reorganized genome-wide during spermiogenesis while some loci are exempted from histone eviction is still elusive.
Our previous study has shown that the residual nucleosomes in mouse sperm largely contain the histone H3 variant, H3.3. The study also revealed differential histone turnover of canonical and variant H3 in round spermatids, which may underlie the final histone composition in mature sperm. In order to determine the dynamics of H3 variants during mouse spermatogenesis, I analyze protein expression of canonical and variant H3 proteins at different stages of male germ cells by triton-acetic acid-urea gel-Western blotting. Surprisingly, I find that mouse testis-specific H3 variant (H3t), not canonical H3, is the most abundant H3 protein from spermatogonia to spermatids and that most canonical H3 is replaced by H3.3 during meiosis. I further observe that a relatively large portion of H3t is removed from chromatin during the process of histone-to-protamine exchange compared with H3.3, which is consistent with that H3.3 is the predominant H3 in residual sperm nucleosomes. Taken together, the first part of my thesis reveals important findings on chromatin composition and dynamics of histone H3 variants during mouse spermatogenesis.
In the second part of my thesis, I describe the discovery that histone H3 is cleaved at its N-terminal tail by a serine protease activity in nuclei of the late-stage mouse spermatids. Arginine 26 and lysine 27 on H3 are important to the H3 protease activity. This proteolytic cleavage of H3 tail may result in nucleosome destabilization and then contribute to nucleosome eviction during spermiogenesis. Interestingly, I find that the acetylation on H3 can prevent H3 from proteolytic cleavage in vitro and that the genome-wide distribution of H3 lysine 27 acetylation (H3K27Ac) is positively correlated to the occupancy of nucleosomes containing transcriptionally active mark in sperm, suggesting that the inhibition of H3 cleavage by acetylated lysine 27 in late-stage spermatids may lead to the nucleosome retention at specific loci in mature sperm. Overall, these findings provide novel insights into the mechanism of nucleosome eviction and retention during spermiogenesis through the regulation of H3 proteolytic cleavage.