Proteomic Characterization of An Isogenic Multiple Myeloma Cell Line Model of Bortezomib Resistance

Abstract 1820

Proteasome inhibitors such as Bortezomib (Bort) represent a key drug class for the therapeutic management of multiple myeloma (MM). However, MM patients, even those who initially achieve complete clinical and biochemical remission with bortezomib-based regimens eventually relapse, and bortezomib-refractory disease is associated with short overall survival. Identifying the molecular basis of resistance to bortezomib is therefore crucial for the rational development of novel therapies to hopefully improve clinical outcome for advanced MM. To address this question, we generated an isogenic cell line model of bortezomib resistance by successive rounds of in vitro exposure of bortezomib-sensitive MM.1R cells to progressively increasing bortezomib concentrations. Serial dose-response analyses confirmed the generation of several clones with variable reduction in bortezomib-sensitivity (IC₅₀ range 40–80nM vs. <10nM for parental MM.1R). The proteomic profile of one of these clones, provisionally termed MM.1VDR, was compared to its isogenic bortezomib-sensitive lines MM.1S and MM.1R, using Liquid Chromatography-Mass Spectrometry (Orbitrap XL). Fold change, Mascot scores (as a measure of confidence for the identity of a given protein) and ANOVA scores for differentially expressed proteins were determined using the Progenesis LC-MS software. 386 proteins were determined to be differentially expressed, of which 154 demonstrated a 2–32 fold change, with p values <0.05. We reasoned that proteins or transcripts differentially expressed in multiple isogenic bortezomib-resistance models may be implicated in molecular mechanism(s) of this resistance. We therefore cross-referenced the list of proteins differentially expressed in MM.1VDR cells with a reanalysis of publically available gene expression profiling datasets of other isogenic models of bortezomib-sensitivity vs. resistance. These included the HT-29 colon cancer cell line (GSE29713) and the mantle cell lymphoma (MCL) lines HBL2-BR and JEKO-BR (GSE20915). In this integrative molecular profiling analysis, we identified no protein/transcript that was concordantly up- or down-regulated in all 4 isogenic models studied. However, we identified that eleven proteins differentially expressed in MM.1VDR cells had concordant differential expression of their transcripts in Bort-resistant HT-29, and either HBL2-BR or JEKO-BR MCL cells. Upregulated markers included FH, DDX46, PSMB4, AKR1A1, EIF3B, BLVRA, PSMB2, RPA3, HPRT1 and PSME3; while PPFIBP2 was down-regulated. These differentially expressed molecules are known to be involved in proteasome structure and function (PSMB4, PSMB2, PSME3, DDX46), mRNA binding and translation initiation (EIF3B/EIF3S9), DNA repair (RPA3), as well as drug metabolism and chemo-resistance (AKR1A1), while several are differentially expressed in diverse neoplasias vs. normal cells (e.g. PPFIB2P is differentially expressed in endometrial cancer, Colas et al, Int J Cancer 2011). Importantly, in gene expression profiling studies in CD138+ tumor cells from selected bortezomib-treated MM patients (treated as part of the APEX and/or SUMMIT clinical trials), high transcript levels for FH, EIF3B, PSMB4 or AKR1A1 or low transcript levels for PPFIBP2 correlate with shorter overall survival (p<0.05, log-rank tests), suggesting that these molecules may have a functional link with the mechanisms responsible for emergence of clinical resistance to bortezomib. Our data, taken together, therefore indicate that the mechanisms of bortezomib resistance are likely multifactorial and potentially tumor-type/specificity-dependent. However, through integrative proteogenomic analyses, we identified functionally and potentially clinically relevant candidate markers of bortezomib resistance. These markers may represent novel targets in efforts to overcome bortezomib-resistance and further improve outcome of MM patients. *Authors contributed equally.