Deciphering the biochemical crosstalk of histone sumoylation in human chromatin

Abstract

The histone code hypothesis states that posttranslational modifications (PTMs) of amino acids at the histone protein termini can combinatorially regulate key DNA processes such as replication, repair, and transcription. The flexible histone termini extend beyond the globular octameric core of the nucleosome core particle, composed of a tetramer of histones H3 and H4, and two dimers of histone H2A and H2B. This octameric protein core tightly wraps about ~147 bp of DNA, and helps to package the vast eukaryotic genome in the small nuclear volume. Proteins that install, remove, or bind to histone PTMs, or marks, work together to modulate the histone PTM landscape and to regulate DNA-templated processes. The biochemical relationship between histone marks is also called crosstalk, and this may be positive or negative, depending on if a histone mark promotes the installation or removal of subsequent marks, respectively. The dysregulation of these proteins and the histone PTM landscape can have deleterious effects on human health, including cancer development and neurodegenerative diseases. Therefore, understanding the biochemical mechanisms underlying the histone code will serve to improve our understanding of how genomic processes are regulated to maintain cellular homeostasis. One histone mark that is poorly studied in cells is histone sumoylation. It was first reported in 2003 as a modification of H4 in human cells and associated with transcriptional repression. The Chatterjee lab has found that H4 Lys12 sumoylation (H4K12su) stimulates demethylase and deacetylase activities within the transcriptionally repressive LSD1-CoREST1-HDAC1 complex. Key to this stimulation is the non-canonical sumo-interacting motif (SIM) within CoREST1. I have characterized the SUMO-CoREST1 SIM interaction using two-dimensional NMR and found that despite differing in sequence from typical SIMs, the CoREST1 SIM binds to the same cleft on SUMO that all SIMs have been observed to bind. Importantly, mutation of the hydrophobic CoREST1 SIM residues to alanine resulted in significant inhibition of binding to SUMO in NMR. We are currently using NMR to investigate the effect of mutations within the CoREST1 SIM that have been observed in somatic cancers. To directly address the function of H4K12su in transcription, I investigated SUMO’s biochemical crosstalk with histone acetylation, a histone mark associated with and critical for active gene transcription. Using in vitro histone acetyltransferase assays and mass spectrometry I found that H4K12su inhibits p300 acetyltransferase activity on H4 in octamers, nucleosomes, chromatinized plasmids, and in cells. Chromatin incorporating H4K12su was also a poor template for in vitro transcription in human nuclear extracts. Additionally, I discovered a new negative biochemical crosstalk between H4K12su and transcriptionally relevant H3 Lys4 methylation by the COMPASS/Set1 methyltransferase in vitro. Cellular experiments confirmed this negative crosstalk between the transcription start and end sites of annotated protein-coding human genes. Collectively, my thesis work has revealed that H4K12su directly inhibits transcription, and may do so in part by the inhibition of histone acetyl- and methyltransferases. Lastly, many mass spectrometric studies have identified sites of sumoylation by SUMO2/3 on human histones but none of them have been well biochemically validated. I have therefore sought to demonstrate histone sumoylation in vivo on both H4 and H2B, that latter which has not been previously shown in the literature. Incorporation of a trypsin cut site at the C-terminus of SUMO will allow us to discretely identify sites of sumoylation and validate them in cellular assays.