ES03.02 Targeting CDKN2A in Mesothelioma

Abstract

Interrogation of the genomic landscape in mesothelioma has revealed considerable inter-patient genomic heterogeneity, however development of molecularly targeted stratified therapy is currently in its infancy. Somatic copy number loss involving the chromosome 9p21-3 locus occurs at high frequency (45% deep deletion in the TCGA cohort), and its expression may be further suppressed by DNA methylation. 9p21-3 harbours CDKN2A. Conditional Knockout of CDKN2A is sufficient to initiate mesotheliomas in vivo consistent with a role in tumorigenesis. This is supported by emerging phylogenetic analysis in which 9p21-3 occurs as an early (clonal or truncal) homozygous event during tumour evolution, in around 20% of cases. Late homozygous and heterogzyous losses also occur, with evidence of parallel evolution. Importantly, CDKN2A is a major negative prognostic factor associated with earlier progression following surgical resection. CDKN2A encodes two genes; 1, the inhibitor of cyclin dependent kinases 4 and 6 known as p16ink4A, and 2. MDM2, an inhibitor of the p53-MDM2 interaction. Early preclinical studies showed that re-introducing p16ink4A transgene in a CDKN2A negative mesothelioma exhibited anti-tumour activity. Small molecule inhibition of CDK4/6 essentially phenocopies p16ink4a (in the presence of wild type RB1). Large scale drug-gene interaction studies have revealed CDKN2A dependent sensitivity to CDK4/6 inhibition suggesting a possible strategy in mesothelioma. We have therefore developed the MiST stratified umbrella trial (NCT NCT03654833) is therefore evaluating the CDK4/6 inhibitor abemaciclib in p16ink4a negative relapsed mesothelioma (arm 2) Loss of p14ARF expression promotes MDM2, an E3 ubiquin ligase targeting p53. Given the low mutation rate for p53 in mesothelioma (around 16% in the TCGA), MDM2 inhibitors may, in the context of p14ARF loss, de-repress p53 to mediate tumour suppression. Preclinical data supports this hypothesis which has translational potential. Co-deletion of the gene MTAP with CDKN2A is common, and coincident with around 80% of CDKN2A losses. MTAP (methylthioadenosine phosphorylase) may represent a potential molecular target in 9p21-3 deleted mesothelioma. This tumour suppressor salvages the essential amino acid methionine, as well as adenine, and plays a critical role in the polyamine salvage pathway. Recently it has been shown that loss of MTAP leads to an increase level in its substrate, methylthioadenosine (MTA). This metabolite directly interacts with and inhibits the epigenetic modifier, PRMT5 (protein arginine methyltransferase 5) that mediates symmetrical arginine di-methylation of histone H4 (H4R3me2s). A direct consequence of this is a reduction in the pool of functional PRMT5 enzyme, revealing a vulnerability to 1. Inhibition of PRMT5 expression or 2. Inhibition of PRMT5 substrate (SAM) synthesis via MAT2A (which converts methionine to SAM). 1st generation small molecule PRMT5 inhibitors are currently undergoing phase 1 clinical evaluation, however MTA dependent allosteric antagonism of such inhibitors presents a challenge to achieving efficacy. Small molecule transcriptional suppression may present a solution. Accordingly, using the connectivity map, we have identified quinacine as a PRMT5 transcriptional suppressor capable of suppressing PRMT5 transcription (which is c-jun dependent), mediating MTAP selective loss of viability with commensurate reduction in H4R3me2s, consistent with an on-target effect. In summary, 9p21-3 loss encompassing CDKN2A and MTAP is one of the most common genomic aberrations in mesothelioma. Novel strategies are emerging with significant translational potential to deliver targeted therapies for mesothelioma. CDKN2A MTAP synthetic lethality