TSPYL2 as a transcription regulator through histone posttranslational modification

Abstract

Testis-Specific Protein Y-encoded Like 2 (TSPYL2) is a member of the Nucleosome Assembly Protein (NAP)/SET/TSPY superfamily. It is highly expressed in neurons. Our group has generated Tspyl2 mutant mice to investigate its physiological functions. These Tspyl2 loss-of-function mutant mice are morphologically normal but have impaired long term potentiation, learning and memory defects, and decreased Grin2a and Grin2b transcription. The total level of trimethylation of histone H3 Lysine 27 (H3K27me3) increases in the Tspyl2 mutant hippocampus. H3K27me3 is one of the key histone marks and it is associated with Polycomb Group mediated transcriptional repression. With these evidences we hypothesize that TSPYL2 is involved in epigenetic regulation of gene expression in neurons. By comparing H3K27me3 ChIP-seq data of primary hippocampal neurons isolated from wild-type and Tspyl2 mutant embryos, I showed that loss of TSPYL2 in the mutant neurons increased both the level and distribution of H3K27me3 on a genome-wide scale. More regions in the genome and genes were marked by H3K27me3 in the mutant but there were also genes being marked only in wild-type samples. At genomic regions marked by H3K27me3 in both genotypes, the intensity of H3K27me3 modification was increased by 30 % in Tspyl2 mutant samples according to ChIP-seq analysis. Functional classification of the 476 H3K27me3 marked genes showed that they were mainly involved in RNA metabolism (44%), developmental processes (46%), and functioning as transcription factors (39%). Genes differentially marked in promoter regions (5 kb upstream TSS) were selected for validation of H3K27me3 marks. Six out of eleven selected sites were successfully validated by ChIP-quantitative PCR. Transcription levels of selected genes were tested, but no obvious correlation of the transcription level with the level of H3K27me3 was found. These data confirmed that TSPYL2 was involved in regulation of H3K27me3 although it was not the only factor to determine gene expression level. To understand how TSPYL2 might regulate transcription, its interaction with other transcription regulators was investigated. Based on the reports in other NAPs, histone acetyltransferase CBP and p300 were selected as candidate TSPYL2 interaction proteins. My data on mammalian-two-hybrid assay suggested that TSPYL2 interacted with CBP. ChIP analysis on wild-type and Tspyl2 mutant primary neurons suggested that TSPYL2 might improve binding of p300 and CBP to Grin2a and Grin2b promoters, respectively. Secondly, Tspyl4 shared a similar expression pattern with Tspyl2. In 293 cells, TSPYL2 interacted with TSPYL4 and altered its subcellular localization. Deletion analysis suggested that the C-terminal acidic tail of TSPYL2 was not required for the interaction. Thirdly, preliminary data of co-immunoprecipitation showed that TSPYL2 interacted with the specific H3K27 methyltransferase EZH2. These data suggested that TSPYL2 might regulate transcription via histone acetylation as well as methylation. In conclusion, TSPYL2 affected trimethylation of histone H3 Lysine 27 in neurons and this may involve EZH2. This is the first time a NAP was found to affect histone methylation and may give insights into the functions of other NAPs. Consistent with previous findings on other NAPs, TSPYL2 interacted with CBP. TSPYL2 may function as a heterodimer with TSPYL4.