Abstract A42: Epigenetic silencing of Brahma, a subunit of the chromatin remodeling complex SWI/SNF

Abstract

Abstract Both BRM and BAF47 are subunits of the SWI/SNF complex and have increasingly clear roles in cancer development. While BAF47 is inactivated via deletions/mutations, BRM is epigenetically silenced. Central to this silencing of BRM are two polymorphisms in the BRM promoter, which have been found to correlate with BRM loss in both cancer cell lines and primary lung tumors. In addition, the induction of BRM by pan-HDAC inhibitors, lead to the discovery that HDAC3 and HDAC9 are regulators of BRM. As these polymorphism sites are 6-7 base pair AT-rich insertions, they have >90% homology with MEF2, POU, and FOX transcription factors. Knockdown experiments of a number of transcription factors lead to uncovering GATA3, MEF2D, and BRN2 as additional regulators of BRM. Both GATA3 and HDAC9 are highly overexpressed in both BRM-deficient cancer cell lines as well as primary lung tumors. Knocking down of GATA3, MEF2D, or both leads to down regulation of HDAC9 and the up regulation of BRM. To determine if any of these proteins might be binding the BRM promoter we conducted ChIP experiments and found that both MEF2D and HDAC9 bind to the BRM promoter in the areas of both polymorphic sites. Since HDACs remove histone acetylation, we looked to see if several chromatin marks changed with the induction of BRM expression. We found that H3K9 and H3K27 become acetylated when BRM is induced. These data suggest a functional role for HDACs in silencing BRM by removing these chromatin marks. To determine if these polymorphic sites have a direct regulatory function on BRM, we conducted BRM promoter-swap experiments where the BRM promoter (and into exon 2) was linked to a luciferase-IRES-neomycin construct. Through homologous recombination we replaced the endogenous BRM promoter with this BRM reporter containing either none or both BRM polymorphic sites and found that removal of the BRM polymorphisms caused about 7-8 fold induction in luciferase indicating a role of these polymorphic sites in BRM regulation. To determine which proteins bind these polymorphic sites we first conducted EMSA experiments and found that these polymorphic sites, compared to wild-type sequence, yielded a gel shift. To determine which sequence elements of these polymorphisms are involved in this gel shift or protein binding, we sequentially changed each of three identical TTAAAA sequences to GCGCGC. When the center sequence was changed from TTAAAA to GCGCGC, no change in the gel shift was observed; however, when either these outer TTAAAA sequences were changed to GCGCGC, the gel shift was lost. This gel shift was also not observed if the center sequence was doubled in length. These results indicate that the outer sequences are involved in protein binding while the center sequence appears to function as a spacer. To further determine what proteins might actually be binding these polymorphic sites, we conducted DNA affinity purification experiments and compared the binding of 40mer with and without the -741 polymorphic sites. This analysis showed four bands uniquely binding these polymorphic sites. By mass spec, we identified the 100kd band as HDAC9 and the 30-35kd band was identified as the scaffold proteins 14-3-3 gamma, delta, and epsilon. Knockdown of either of the 14-3-3 proteins in BRM-deficient cell lines induced BRM and caused growth arrest in a BRM-dependent manner. As such, the involvement of these 14-3-3 proteins suggests a large regulatory complex is involved in the binding to the polymorphisms affecting the regulation of BRM. As we observed experimentally essentially identical results from BRM-deficient Rhabdoid, cervical, and lung cancer cell lines, the findings suggest a conservation of this epigenetic mechanism across very different cancer types. Citation Format: Bhaskar Kahali, Sarah Gramling, Ken Thompson, Stefanie Marquez, Lu Lu, David Reisman. Epigenetic silencing of Brahma, a subunit of the chromatin remodeling complex SWI/SNF. [abstract]. In: Proceedings of the AACR Special Conference on Chromatin and Epigenetics in Cancer; Jun 19-22, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2013;73(13 Suppl):Abstract nr A42.