P-373: Functional genomics of genes residing in the minimally deleted chromosome 17p13 region in multiple myeloma: therapeutic implications

Introduction: Chromosome 17p deletion (del(17p)), a high-risk feature in multiple myeloma (MM), is typically monoallelic. Its biological/clinical significance involves complete loss-of-function (LOF) of the tumor suppressor gene (TSG) TP53, through deletion of one TP53 copy and inactivating mutation (double-hit lesion) of the other copy, whereas biallelic loss is rare. While this combined copy number (CN) loss and mutation of TP53 appear to be favored over biallelic deletion of TP53/17p13, the mechanisms for this phenomenon have not been elucidated. We hypothesized that 17p13 contains multiple genes essential for MM cell survival, thereby favoring retention of at least one copy of this region in MM cells.

Methods: We curated genes residing in chr17p13 through the UCSC Genome Browser and examined their essentiality scores (CHRONOS 22Q2 or CERES scores from DepMap) in 19 MM cell lines (including 3 TP53 wild-type [WT] lines), to identify genes are that “core”/”pan-cancer” dependencies (CERES/CHRONOS scores <-0.4 in >90% of tumor lines across all lineages); preferentially essential for MM cells (compared with all non-MM lines) vs recurrently (but not preferentially) essential for MM cells vs non-essential. Functional assessment of key genes was complemented by in silico analyses of MMRF CoMMpass data (IA22, n=1050) and CRISPR-based LOF studies (of individual genes or in pooled format) or cDNA-overexpression studies.

Results: Of the 108 chr17p13 genes, we identified 7 “core” essential genes (RPL26, AURKB, EIF4A1, MED31, EIF5A, TRAPPC1, ELP5), 2 MM-preferential dependencies (the glycosylation regulator MPDU1, and the purine biosynthesis gene PFAS); and other genes (e.g. CTDNEP1, CTC1) that were recurrent essentiality (e.g. in >3/19 MM lines). Enrichment of CRISPR KO was observed for TP53 in 2 of 3 TP53 WT MM lines; and 1 for candidate TSG, SAT2. This latter gene had enrichment of its CRISPR-KO in a pooled screen in NCI-H929 cells (not studied in DepMap). Supporting this result, targeted CRISPR-KO and overexpression of SAT2 increased vs decreased, the growth of H929 cells and their sensitivity to bortezomib or lenalidomide treatment. CN losses, all monoallelic, of SAT2 were observed in 11.4% of samples, suggesting that the preference for monoallelic CN loss of 17p13 applies not only for TP53 but also other putative TSGs of this locus. In contrast to 17p13, TSGs with recurrent biallelic deletion (e.g. CDKN2C on chr1p32; CN loss in 12% of pts, with 23/152 [15%] biallelic events) have fewer essential genes and longer genomic distances from them (e.g. MPDU1 is within 85kbp of TP53, while the closest essential gene to CDKN2C (BTF3L4) is >1Mbp away).

Conclusions: 17p13.1 contains several “core” or MM-preferential (MPDU1, PFAS) dependencies, some located close to TP53. This may explain why monoallelic del(17p) events predominate in MM and suggests that therapies targeting MPDU1, PFAS or other dependencies of this locus may have applications for high-risk 17p(del) MM.