Abstract

The SWI/SNF chromatin remodelling
complex plays an important role in tumorigenesis, both due to its broad
mutation status across cancers as well as specific dependencies imparted by
these mutations.  In particular, we and
others have previously uncovered a synthetic lethal relationship between SMARCA2
(also known as BRM) and SMARCA4 (also known as BRG1), the two mutually
exclusive catalytic subunits of the mammalian SWI/SNF complexes, in which
SMARCA4-deficient cancer cells rely on the remaining SMARCA2 ATPase activity for
survival .  SMARCA4 mutation or loss of
expression occurs in a variety of cancers, and is predominant in non-small cell
lung cancers (NSCLC) typically lacking targetable lesions such as EGFR, thereby
highlighting the need for a therapy that can exploit such a synthetic lethal
relationship.

Previously, DNA-encoded library and biochemical screens were
performed to identify SMARCA2 inhibitors but yielded few tractable hits.  One possible reason for the failure of hit finding
may be due to lack of context around the SWI/SNF complex with the isolated
enzyme, and as such, a cell-based assay may be more pharmacologically relevant.  From literature and in house studies it was
observed that knockdown of SMARCA2 in SMARCA4-deficient cells lines produced
several phenotypic effects including reduced proliferation, increased
senescence and increased global H3K9me3 and these end-point were investigated
as potential assay end-points. 
Additionally, RNAseq studies were also performed to identify other
biomarkers of SMARCA2 knockdown in a variety of SMARCA4 deficient lung cancer
cell lines.

A phenotypic cell-based screen was prosecuted through the
CRUK / AstraZeneca Screening Alliance whereby CRUK investigators work
collaboratively with AstraZeneca to access their HTS expertise and compound
collection.  A 14K compound annotated phenotypic
screening collection was tested in a proliferation end-point assay using H1299 paired
cell lines proficient and deficient for SMARCA4.  Two potential ‘hits’ were identified which
appeared to decrease proliferation in the SMARCA4-deficient H1299 cell line selectively
versus the proficient line.  A computational
chemistry review identified 84 compounds for further testing in a beta-galactosidase senescence cell assay and
biochemical ATPase assay.  Disappointingly
no compounds were identified with the desired profile to progress to further
studies.

Targeting the SWI/SNF complex and SMARCA2 in a relevant
synthetic lethal background remain attractive options for cancer therapy.  Recent publications describing the identification
of allosteric and protac inhibitors of SMARCA2/4 indicate this remains a
competitive area for drug discovery.