A Proteogenomic Approach to Identify Potential Drivers of Hyperdiploid Multiple Myeloma

Introduction

The major primary genetic events in multiple myeloma (MM) include translocations of immunoglobulin heavy chain (IgH) and a hyperdiploidy (HRD) karyotype (Kuehl, 2002). The main drivers of pathogenesis in HRD MM are unknown (Fonseca, 2003), therefore further research is needed in this subgroup. Translational research in HRD MM has relied on analysis of genomic and transcriptomic alterations, however changes at the genome and transcriptome level do not always correlate directionally with protein abundance. Because proteins are considered the main and final effectors of most cellular processes, proteogenomics can improve our understanding of the drivers of HRD MM. We employed it here to identify potential biomarkers in HRD MM and determine whether any given trisomy may lead to a selective advantage.

Methods:

Protein and mRNA were extracted from CD138+ cells from 47 HRD MM and 14 t(11;14) MM samples. RNA-seq data was processed using MAPR-Seq pipeline using default parameters and gene-wise raw counts were extracted for each sample. Proteomics data was processed using Specranaut and protein-wise intensities were extracted for each sample. A generalized linear model configured with negative binomial distribution was used to analyze gene-wise count data in edgeR software. Genes with an adjusted p-value of <=0.05 were considered significantly differentially expressed. Protein intensity data were normalized using trimmed mean of M-values and log2 transformed. A generalized linear model configured with Gaussian distribution was used to compare protein intensities between the groups. Proteins with an adjusted p-value <= 0.05 were considered significantly differentially expressed. Two separate rank lists for protein and mRNA data were generated for pathway analysis.

Results

When comparing our small cohort of HRD MM patient samples to those with t(11;14) MM, out of approximately 16,600 total transcripts and 9,700 proteins evaluated we identified approximately 2,100 significantly variant transcripts and 500 significantly variant proteins in HRD MM. GSEA pathway analysis of the transcriptome and proteome revealed significantly upregulated translational and ribosomal processing pathways in HRD MM. Transcriptional MYC target pathways were also significantly upregulated. RNAseq revealed significantly higher expression of MYC (Rank 24.5, p=5.3E-6) in HRD MM. It is known that patients with t(11;14) MM have increased expression of CCND1 (Kuehl and Bergsagel, 2012). As expected, we found that the HRD MM subgroup had significantly lower expression of CCND1 compared to the t(11;14) MM subgroup (Rank –79.9, p=2.8E-21) and this was also true at the protein level (Rank –21.7, p=2.2E-5). The most significantly abundant protein in the HRD MM subgroup compared to t(11;14) MM was BAHD1, which is a protein that targets H3K27me3 and recruits transcriptional co-repressors (Fan, 2020). BAHD1 has been shown to repress erythroid genes and may recruit HDAC5, which plays a role in osteoblast differentiation (Rank 72.3, p=2.9E-20; Xu, 2021). The second most abundant protein in HRD MM was CKS2, which binds to the catalytic subunit of cyclin-dependent kinases (CDKs) and therefore positively regulates cell proliferation, invasion, and migration and has been shown to be implicated in tumor progression in several solid tumors (Rank 72.3, p=2.9E-20; You, 2015). TRIM proteins promote tumorigenesis and cancer development. TRIM14 mediates cancer development through mediating the JAK/STAT, PI3K/AKT, NFkB, and p53 pathways and was found to be significantly higher in HRD MM (Rank 70.6, p=9.0E-20; Huang, 2022). We also found increased protein abundance of LAMP5 in HRD MM (Rank 20.1, p=6.0E-5). LAMP5 may play an important role in MM progression and has been shown to be highly expressed in patients with mixed-lineage leukemia and is associated with a poor prognosis (Wang, 2023; Gracia-Maldonado, 2022). We identified many proteins (not all discussed here) that have not been well described in HRD MM and thus may be biomarkers/targets of disease pathogenesis of this subgroup.

Conclusion

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 identified proteins that may lead to selective advantage and/or pathogenesis of trisomies in HRD MM.