Structural and biochemical study of BAF chromatin remodelling complex

Abstract

Mammalian BAF complex, as the first identified chromatin remodelling complex, utilizes the energy of ATP hydrolysis to alter histone-DNA contacts in order to slide, eject or restructure nucleosomes, which regulates a variety of fundamental biological processes including gene transcription, DNA recombination and repair, cell cycle control, and so on. Recent exome sequencing studies have revealed that ~20% human cancers bear alterations of genes encoding BAF subunits, making BAF among the most frequently mutated complexes in cancers. Highly polymorphic BAF complex is composed of a mutually exclusive ATPase, either BRG1 or BRM, and other 7 to 14 BRG1- or BRM-associated factors (BAF subunits). Among all subunits, BRG1/BRM, BAF155/BAF170, BAF47 and BAF60 are evolutionarily conserved from yeast to human, which are regarded as essential components and have attracted much attention. BAF155/BAF170, BAF47 and BRG1 form the core regulatory machinery to modulate chromatin structure with the remodelling activity comparable to the intact complex. BAF60a interacts with BAF155 and various transcription factors to target BAF complex to specific genomic loci. In spite of being extensively studied for many years, the molecular mechanism how these BAF essential subunits interact is still lacking. During my Ph.D. study, I have applied structural biology tools and biochemical assays to investigate the physical interaction among BAF essential subunits and determined three crystal structures including BF155-BAF47, BAF155-BRG1 and BAF155-BAF60a complexes. In Chapter 3, I presented the high-resolution crystal structure of BAF155 SWIRM domain in complex with BAF47 RPT1 domain. BAF155 SWIRM domain mediates the interaction with BAF47, which is functionally distinct from other characterized SWIRM domains that possess DNA binding activity. It is worth noting that I am the first to present the crystal structure of BAF47 RPT1 domain, which adopts a novel structural fold. In Chapter 4, I narrowed down BAF155 SANT-binding region in BRG1 to the BH motif, and further solved the crystal structure of BH-SANT complex. BRG1 BH motif is first characterized to have extensive interaction with BAF155 SANT domain. Crystal structures of BAF155- BAF47 and BAF155-BRG1 complex provide the detailed molecular basis of BAF core regulatory machinery assembly. In Chapter 5, I mapped that BAF155 L2 loop directly interact with BAF60a SWIB and YEATS domains, and further determined the crystal structure of BAF155-BAF60a complex. This is the first time the SWIB and YEATS domains have been demonstrated to mediate protein-protein interactions. Taken together, BAF155, as the scaffold protein, interacts with BAF47, BAF60a and BRG1 to facilitate the chromatin remodelling at target loci. Crystal structures presented in this dissertation provide critical insights into the structural assembly of BAF complex and paved the way for future functional analysis. Overall, this study should advance our understanding of chromatin remodelling mechanisms of BAF complex, and provide potential therapeutic strategies to treat cancers associated with BAF subunit mutations.