Identification of Tumor Suppressors and Synthetic Lethal Targets in -7/del7q Myeloid Neoplasms By Iterative Computational Analysis Coupled with Experimental Validation

Complete or partial deletion of chromosome 7 (-7/del7q) is one of the most frequent classes of chromosomal abnormalities found in myeloid malignancies. -7/del7q is distinguished by its poor prognosis, high prevalence among pediatric cases, association with bone marrow failure syndromes, and widespread occurrence among various pathomorphological subtypes of myeloid malignancy (Mori et al.). Despite decades of study, however, the genes responsible for the leukemogenic properties of these deletions are incompletely identified and no targeted therapies matched to the lesion exist (Inaba et al.). Hypothetical candidate mechanisms of -7/del7q pathogenesis include haploinsufficiency and biallelic inactivation of tumor-suppressor genes (TSGs). Deletion may also create clonal vulnerabilities due to haploid loss of genes contained in deleted regions that render clonal cells vulnerable to synthetic-lethal gene (SLG) targeting treatment strategies while leaving normal cells unaffected. To date, these aspects of -7/del(7q) pathogenesis have been tested only on a piecemeal basis, with most studies focusing on identifying haploinsufficient TSGs. Our group has performed extensive computational analyses to query these with the intention of creating a comprehensive understanding of -7/del7q pathogenesis. We have further verified these analyses with selected knockout models (Vail et al, Durkin et al), initiated a wide scale screening effort to comprehensively assay all published -7/del7q potential leukomegenic drivers, and continue to iterate our computational analyses to evolve a detailed map of pathogenic pathways in this context.

To these aims, we queried an extensive in-house databank of myeloid neoplasia patients, as well as external data sources. Within this combined cohort are RNA-seq datasets from 29 - 7/del7q patients and WXS/WGS datasets from 47 - 7/del7q patients.

We conceptualized TSGs as either haploinsufficient genes or as genes that are biallelically inactivated, where one allele is lost through the deletion event and the other through germline or somatic mutation. To identify this overlooked class of TSGs, we created a biallelic inactivation test, wherein we queried chromosome 7 genes for preferential loss of either the wild-type or a previously mutated allele. We identified 84 genes with a significant preference for loss of the wild-type allele, suggesting these genes are TSGs physiologically inactivated by the second loss-of-function genetic lesion. Among these hits were four genes that are known tumor suppressors in other contexts: TRIM24, CCDC138, SFRP4, and PTPN12.

Leukemogenesis in the -7/del7 context may also be driven through the loss of specific combinations of genes. To explore combinatory inactivation of more than one chromosome 7 TSG as a potential leukemogenic mechanism, we expanded the biallelic inactivation test to check for co-mutation amongst chromosome 7 gene pairs. From this, we identified a set of 21 chromosome 7 genes with 33 co-mutation preferences amongst the set, including RAPGEF5: MIOS, TAF6:MCM7, and PMS2:AIMP2.

We also sought to identify non-TSG therapeutic targets in -7/del7q lesions. We analyzed expression and mutation signatures to detect compensatory genes that may represent vulnerabilities within cancer cells. We identified hemizygous mutations in our patient data and compared expression levels to diploid patients. As expected, most genes lost in a -7/del7q lesion were commensurately downregulated, but some showed no loss (or even gain) of expression. These upregulated genes may be required for cell survival in cancer cells, but not healthy cells, representing druggable SLG targets. Genes with expression unaffected (or upregulated) by deletion include those whose protein is targetable by an existing small molecule (ABCB5, ADCY1, CYP3A5, EPHA1, NOS3, NPC1L1, SMO), an existing biotherapeutic (monoclonal antibody/enzyme or other protein) (ABCB4, IL6, LAMB1), or both (ABCB1, EGFR, GPER1, MET).

This analysis, combined with others focused on identifying haploinsufficient TSGs, comprise a comprehensive and rational atlas of -7/del7q leukemogenic drivers and therapeutic targets. These targets are currently being experimentally confirmed and characterized with the goal of developing targeted small molecule therapies via our in-house drug developmental pipeline to address the unmet clinical need these lesions present.