Reconstructing Catastrophic Chromothripsis Events Using Multiomic Data Reveals Their Functional Impact in Multiple Myeloma

Introduction: Complex structural variants (SVs) are important markers of high-risk disease in multiple myeloma, and include chromoplexy, templated insertions, and chromothripsis. Despite being able to identify these events in whole genome sequencing data, we still know relatively little about how these are generated and their selective advantage. Chromothripsis is particularly complicated and often involves dozens of SVs and copy number changes, though the functional impact has not been systematically investigated.

Methods: Paired short-read (Illumina, average depth 75x), long-read HiFi (PacBio, 16x) and Micro-C (Cantata Bio, 934M reads) data were sequenced from 13 patient-derived xenograft (PDX) samples and 2 cell lines, covering t(11;14), t(4;14), and hyperdiploid subgroups. Reads were aligned to the hg38 reference while haplotype phasing was conducted using small variants detected from HiFi data (deepvariant). De novo assembly was conducted for haplotypes using Micro-C and HiFi data. Chromothripsis events were reconstructed by integrating short-read (Manta), HiFi (pbsv), Micro-C (EagleC) data and assembled haplotypes. Dysregulated genes at the SV region were subsequently identified with RNA-sequencing. Moreover, reconstructed chromosome organization including loops/topologically associated domains (TADs) were identified from Micro-C data (Neoloopfinder). CUT&Tag-IT was conducted for six histone marks to determine chromatin states as well as identifying enhancers/repressors. Additionally, haplotype-level DNA methylation was determined from HiFi data while differential methylation regions were calculated and mapped onto breakpoints.

Results: Three chromothripsis events were found in three PDX samples. One occurred on chromosome 2p which was shattered into 32 large fragments (>500 kb), 10 and 13 of which were further amplified or deleted, respectively. 66% of breakpoints did not have homologous joins (N<=1), which, combined with pathway analysis indicated that non-homologous end joining was the dominant DNA damage repair mechanism. SNP B allele frequencies indicated chromothripsis only occurred on one allele.

Analysis of chromosome 2 breakpoints and interactions indicated that the chromothripsis event occurred through chromosome shattering, shuffling, and amplification of regions. This event was followed by a swap of material resulting in the insertion of the chromothripsis event into chr 18 and a 10 Mb region of chr 18 into chr 2, thereby affecting multiple chromosomes. Interestingly, chromothripsis-derived insertions into other chromosomes were also observed in 67% of 133 samples with chromothripsis events in the CoMMpass cohort, indicating that duplication of chromothripsis SVs onto other chromosomes is a common mechanism.

Among 236 protein coding genes in the region affected by chromothripsis, 109 were significantly dysregulated (|Log₂FC|>0.49, p<0.05, T-test). 84 were affected by copy number (CN) loss or gain, including 3 tumor suppressor genes (ASXL2, CN=1) and 4 oncogenes (MYCN, CN=4). Super enhancers (SE) dysregulated 59 genes including a de novo SE defined by H3K27ac detected in a neo-chromatin loop with MYCN which was also hypomethylated at the DNA level (10% vs. 23%, p=0.02, Mann Whitney U test). Also, a SE near KCNS3 was juxtaposed to CAD (an oncogene involved in nucleotide synthesis and proliferation), leading to its over-expression (Log₂FC=0.7 p=0.01, CN=2); another SE near MTA3 was juxtaposed to EPCAM, a cell surface adhesion protein, leading to a significant upregulation (Log₂FC=3.7, p=1x10^-8, CN=4); a t(2;18) breakpoint was found in the promoter of the tumor suppressor FOXN2, resulting in downregulation (Log₂FC=-0.5, p=0.08, CN=2); significant allele-specific hypomethylation (2% vs. 37%, p=0.008) was found on the chromothripsis allele in the promoter of oncogene SOS1, which was significantly over-expressed (Log₂FC=2.1, p=3x10^-6, CN=4).

Conclusion:

Reconstruction of a chromothripsis event revealed various mechanisms that lead to the dysregulation of multiple oncogenes and tumor suppressors. Our analysis indicated a highly dynamic environment that chromothripsis created to allow MM cells to manipulate gene regulation to gain a proliferative advantage, explaining the negative prognostic impact of these events.