Vκ*MYC Mouse Model Recapitulates Key Genomic Defining Events in Multiple Myeloma

Background: The Vk*MYC transgenic model of multiple myeloma (MM) is based on the sporadic AID-induced activation of MYC in a single germinal center (GC) B-cell, in a mouse strain (C57Bl6) that spontaneously develops monoclonal gammopathy. With age, Vk*MYC mice universally develop a progressive expansion of class-switched, somatically hypermutated, monoclonal plasma cells. Myeloma defining events occur only after long latency, suggesting that additional mutations are required to induce malignancy. Here, we interrogated the genomic landscape of Vk*MYC MM to capture somatic mutations selected during progression.

Methods: We interrogated whole genome (WGS) and exome sequencing (WES) from 55 purified plasma cell (PC) from Vκ*MYC transgenic mice tumors. Additional 27 and 12 Vκ*MYC mice were investigated with array comparative genomic hybridization (aCGH) and mate-pair low coverage WGS, respectively. Our analysis included 34 individual de novo tumors from aged Vk*MYC mice, 37 individual tumors from recipient mice transplanted with tumor cell lines established from de novo donors, and 13 tumors selected to propagate in vitro.

Results: Tumor mutational burden was lowest in de novo samples (median 2.3 mutations/Mb, range 0.3-4.3 mutations/Mb) and highest in in vitro samples (median 5.8 mutations/Mb, range 1.2-12.7 mutations/Mb), similar to that of human MGUS and MM, respectively. To identify genes affected by nonsynonymous mutations more frequently than expected by chance (i.e., driver genes under positive selection), we ran dndscv. A total of 11 significant driver genes were detected, 7 of which are known driver genes in human MM: Dusp2, Trp53, Nfkbia, Pim1, Tent5c (Fam46c), Hist1h1c, Hist1h1e. Driver genes not previously described in MM were: Apobec3, Pten, Cd74, Cd79b. Additional nonsynonymous mutations were detected in driver genes frequently altered in MM including Kras, Nras, Dis3, as well as histone modifiers and NF-kB regulators. Differential mutation analysis revealed bi-allelic inactivation of Trp53 and Pten were more frequent in in vitro samples compared to de novo and transplanted tumor, in line with the almost universal p53 inactivation reported in human MM cell lines.

As in human hyperdiploid MM, GISTIC and molecular time analyses revealed evidence of simultaneous multi-gain events of chromosomes 1, 9, 11, and 15. As reported previously, we observed frequent deletion of chromosome 5. There were numerous recurrent focal CNAs involving oncodrivers. These included focal deletion of tumor suppressors genes Cdkn2a, Cdkn2b, Cdh11, Ncor1, Rb1, Trp53, Kdm6a and Pten, as well as amplification of oncogenes Mdm4, Mcl1, and Kras.

We detected several structural variants and complex events throughout the Vκ*MYC genome, similar to human MM. Importantly, we found chromothripsis in five (14%) cases. Additionally, in three Vκ*MYC mice we detected chromosomal translocations involving the immunoglobulin heavy or light chain loci that partnered and induced overexpression of MM associated genes such as Irf4, Map3k14 and Pou2af1. Overall, mutations activating the NF-kB and STAT3 pathways were the most common events, each present in more than half MM. Overall, the genomic landscape of the Vκ*MYC model resembles hyperdiploid MM in humans, with MYC translocations and enrichment of events affecting the NF-κB pathway.

Finally, analysis of mutational signatures uncovered similar signatures present in human MM, including aging, APOBEC, and poly-eta in the GC, among others. In particular, APOBEC mutational activity was detected in more than half of Vκ*MYC MM, further supporting the similarities with human MM where APOBEC mutational activity is detectable in 80-90% of human MM and progressive precursor, but it is undetectable in the stable precursor conditions.

Conclusions: Overall, these data reveal a progressive accumulation of genomic events from de novo, to transplant, to in vitro tumors that mimic the natural progression from newly diagnosed, to relapsed disease, to cell lines noted in human MM. Thus, our results highlight how the transformation pathways to malignant PC are conserved between mice and men. Given the long lead time between precursor diagnosis and malignant transformation to MM that limit the ability to investigate the earliest phases of tumorigenesis in human MM, the Vκ*MYC transgenic mouse model of human MM represents a unique opportunity to overcome this temporal limitation.

Landgren:Leukemia & Lymphoma Society: Research Funding; Janssen: Honoraria, Other: Independent Data Monitoring Committee (IDMC) member for clinical trials, Research Funding; Pfizer Inc: Consultancy; Aptitude Health: Honoraria; MMRF: Honoraria; Merck & Co., Inc.: Other: Independent Data Monitoring Committee (IDMC) member for clinical trials; Rising Tide Foundation: Research Funding; Theradex: Other: Independent Data Monitoring Committee (IDMC) member for clinical trials; NCI/NIH: Research Funding; Riney Foundation: Research Funding; Tow Foundation: Research Funding; Amgen: Honoraria, Research Funding. Bergsagel:Oncopeptides: Consultancy; Janssen: Consultancy; GSK: Consultancy; Novartis: Consultancy; Pfizer: Consultancy. Chesi:Abcuro, Palleon Pharmaceuticals, Pi Therapeutics.: Patents & Royalties: Genetically engineered mouse model of myeloma.; Pfizer, Novartis.: Consultancy.