The Molecular Landscape and Essential Pathways Involved in Bone Marrow-Mediated Carfilzomib Resistance of Multiple Myeloma

Background

Multiple myeloma (MM) remains an incurable malignancy, with most patients relapsing and dying from the disease. Anti-myeloma drugs, such as proteasome inhibitors (PIs) bortezomib and carfilzomib (CFZ), have considerably improved prognosis in myeloma. Despite these advances, disease heterogeneity, early relapse and treatment resistance still pose major challenges in MM treatment. Understanding the mechanisms that mediate PI resistance provide a key to targeting both, PI-resistant minimal residual disease that drives relapsed MM after prolonged PI-containing frontline therapy, as well as PI-refractory, aggressive advanced MM. While key mechanisms of the in vitro-generated PI resistance in MM have been revealed in cell line models, we lack understanding of PI resistance in vivo, where in particular clonal heterogeneity and the tumor microenvironment (TME) within the bone marrow (BM) add additional levels of complexity. Therefore, the aim of our study was to analyze the molecular landscape and changes occurring during MM progression under CFZ treatment in vivo and to identify key molecular processes contributing to CFZ-resistance of MM cells in the presence of stromal cells in vitro, to ultimately identify new molecular pathways and develop innovative treatment strategies in PI-resistant MM.

Methods

The NSG mice intrafemorally engrafted with human RPMI-8226 cells were either untreated or treated long-term with 4 mg/kg CFZ (intravenously) until they became drug resistant. At this point, CFZ naïve and CFZ-resistant cells were isolated and processed for single-cell RNA sequencing (scRNA-seq, 10x Genomics) with the aim to characterize a transcriptional CFZ-resistance signature in refractory cells. To investigate the role of the TME as well as the importance of cell-cell interactions in CFZ-resistance in vitro, we performed two independent genome-wide CRISPR/Cas9 library screenings. In the first one, Brunello library transduced RPMI-8226 cells were co-cultured with human stromal cells (HS5) and treated with CFZ to identify CFZ sensitivity/resistance candidate genes. In the second experiment, Brunello library and synthetic Notch (synNotch) receptor transduced HS5 cells were co-cultured with synNotch ligand transduced RPMI-8226 cells to identify genes that are essential for establishing cell-cell contacts between stromal and MM cells. Subsequent functional analysis of the highest-ranking CFZ sensitivity/resistance candidates in the RPMI-8226+HS5 co-culture included shRNA-silencing, single-gene knockouts, viability assays, cell cycle analysis and protein synthesis analysis using the SUnSET assay.

Results

ScRNA-seq analysis of CFZ-refractory RPMI-8226 cells growing in the BM of NSG mice showed a different transcriptional landscape, compared to CFZ-naïve cells isolated from the BM of untreated mice. The unsupervised clustering analysis, using UMAP, revealed that cells exposed to CFZ show distinct populations with a strong increase in the OXPHOS and protein folding capacity as well as down-regulation of several genes involved in proliferation and apoptosis, when compared to naïve cells. The CRISPR/Cas9 library screening where RPMI-8226 cells were co-cultured with HS5 cells and exposed to CFZ revealed several CFZ sensitivity candidates at the cut-off of false discovery rate (FDR) < 0.01 and fold change above 1.5-fold. Those genes are involved in cytokine signaling, cell growth, invasion, metastasis and quality control of translational elongation. At the same time, the CRISPR/Cas9 library screening, where synNotch receptor transduced HS5 cells were co-cultured with synNotch ligand transduced RPMI-8226 cells revealed gene candidates at the cut-off of FDR < 0.01 and fold change greater than 1.5-fold, which mediate stronger or weaker cell-cell interaction. Those genes are particularly involved in cytokine signaling and mitochondrial metabolism.

Conclusion

In conclusion, MM cells that acquired CFZ-resistance upon cell-cell contact with certain cell types within the TME, such as stromal cells, differ significantly from CFZ-naïve cells. CFZ-resistance, caused by cell-cell contact with stromal cells, is presumably mediated via decreased proliferative as well as protein synthesis capacity of MM cells. Therefore, stimulation of MM cells to proliferate and synthesize more proteins may be a key to targeting CFZ-resistance in vivo.