Assistant Professor Mayo Clinic, Arizona, United States
Introduction: The major primary genetic events in multiple myeloma (MM) include translocations of immunoglobulin genes, and a hyperdiploidy (HRD) karyotype. HRD MM has historically been categorized as a subgroup of MM with a more favorable prognosis, but some patients may have more aggressive disease. Further research is needed in this subgroup. Proteogenomics integrates proteomic and genomic data; we employed it here to determine whether any given trisomy may lead to a selective advantage.
Methods: Protein and mRNA were extracted from CD138+ isolated cells from 47 HRD MM samples using the Qiagen AllPrep Kit. RNA-seq data was processed using MAPR-Seq pipeline using default parameters and gene-wise raw counts were extracted for each sample. Protein samples were digested with trypsin. Proteomics data was processed using Specranaut and protein-wise intensities were extracted for each sample. A multivariate analysis of gene expression data was performed using edgeR software configured to assess the effect of each chromosomal arm-level abnormality on each gene’s expression when accounting for other copy number variant (CNV) events. Genes with an adjusted p-value of < =0.05 were considered significantly associated with the corresponding CNV event. A similar method was followed for performing multivariate analysis with proteomics data with slight modifications.
Results: We found gene dosage effects with increased relative RNA expression and protein abundance of genes on trisomic chromosomes when compared to genes/proteins on all chromosomes (sum relative frequency 14.4 vs 8.8 for RNA and 49.0 vs 30.0 for protein, respectively). We focused on trisomies 5, 9, 11, 15, and 19 as these are common in HRD MM. KIF2A, which was found to be significantly more abundant in trisomy 5 samples (Rank 13.1, p=0.02), regulates spindle organization and chromosome movement during mitosis, and knockout experiments lead to chromosome misalignment and mitotic arrest. The mechanism for trisomy in HRD MM is unknown and defects in chromosome alignment genes may play a role. Trisomy 9 samples had significantly higher abundance of MLLT3, a known oncogene in AML (Rank 15.8, p=0.003). We found increased expression of CCND1 in samples with trisomy 11 compared to those without trisomy 11 (Rank 10.3, p=0.04) but this was not significant at the protein level (Rank 2.52, p=0.56). Trisomy 11 also had increased protein abundance of PSMD1 (Rank 14.1, p=0.007); mutations are increased in MM refractory to proteasome inhibitors. SNAPC5, a protein important for transcription, was elevated in samples with trisomy 15 (Rank 25.6, p=< 0.001). Trisomy 19 by far had the highest protein abundance in zinc fingers (n=7), which have a known role in MM progression.
Conclusions: Proteins are the final effectors of most cellular processes. Proteogenomics can improve our understanding of the drivers of HRD MM. Using a proteogenomic approach we found gene dosage effects and identified proteins that may lead to selective advantage of trisomies in HRD MM.
