Long-Term Interactions with Bone Marrow Stromal Cells Alter the Molecular Dependencies of Myeloma Cells: Insights from Genomewide CRISPR-Based Functional Genomic Screens

Bone marrow stromal cells (BMSCs) play important roles in the pathophysiology of multiple myeloma (MM), other hematologic malignancies and metastatic solid tumors, by supporting the survival, proliferation and treatment resistance of malignant cells. Prior molecular profiling studies from our group and others documented that a broad range of proliferative, oncogenic and anti-apoptotic transcripts are upregulated in MM cells by their in vitro coculture with BMSCs in short-term assays (e.g. <1 week of coculture). In this study, we hypothesized that longer-term co-cultures (e.g. over 4+ weeks) with BMSCs are capable of further remodeling of the molecular networks on which MM cells depend for their survival and proliferation.

To address this hypothesis, we performed a genome-wide CRISPR/Cas9-based gene editing screen, in which MM.1S cells engineered to stably express the nuclease SpCas9 were also transduced with a lentiviral particles for a pool of ~70,000 sgRNAs (Brunello library), targeting exons of ~20,000 genes (plus non-targeting control sgRNAs), under conditions of transduction which allow for an average of no more than 1 sgRNA to be incorporated in a given cell. This allowed us to convert the initial population of MM.1S cells into heterogeneous pools in which each gene is subject to individual loss-of-function, due to Cas9-mediated editing, by only 1 sgRNA (pool contains typically 4 sgRNAs per gene; 500+ cells per sgRNA; 4 replicates per experimental condition). The MM.1S-Cas9+ Brunello-transduced cells were then cultured in the absence (conventional monoculture) or presence of the immortalized human BMSC lines HS5 or HS27A (transduced with neither Cas9 nor Brunello sgRNAs), which are established BMSCs lines that can increase survival and proliferation of MM and many other tumor cell types in vitro. The monocultures or co-cultures of MM.1S cells were passaged serially over 4 and 6 weeks: each at those time points, MM cells were isolated from the cultures and processed for PCR and next generation sequencing studies to quantify the sgRNAs which exhibited depletion (vs. enrichment) in MM cells with vs. without co-culture with BMSCs.

Consistent with our initial hypothesis, we observed that MM.1S cells cocultured for 4 or 6 weeks with each of the 2 BMSC lines exhibited, compared to monocultures of the same duration, statistically significant depletion (at least 3 sgRNAs/gene, MAGECK algorithm) for several genes with known or biologically plausible roles in cell survival or proliferation, including inhibitors of apoptosis, proteasome subunits, ribosomal genes, transcription factors or cofactors, regulators of splicing, autophagy and intracellular trafficking; with higher numbers of significant "hits" after 6 vs. 4 weeks of co-culture with either HS5 or HS27A. Some of these essential genes exhibited sgRNA depletion both in MM cells cultured in the absence of BMSCs and a quantitatively more pronounced sgRNA depletion in co-cultures with BMSCs, suggesting that interaction with BMSCs enhances the functional dependence of MM cells on these genes. Interestingly, some genes (e.g. with roles in translational and transcriptional elongation; mitochondrial organization and metabolism) exhibited preferential depletion of sgRNAs only in the presence, but not absence of BMSCs, suggesting that these genes represent candidate dependencies with preferential role in the context of tumor-stromal interactions. Notably, among the 2500-3000 genes with significant sgRNA depletion in the absence of stromal cells, >200 of genes (including genes involved in RNA and mRNA processing, cell cycle control, epigenetic regulation and cellular response to stress) did not exhibit sgRNA depletion in the presence of BMSCs, suggesting that the BMSC-MM cell interaction may attenuate or abrogate the role of these genes as candidate molecular dependencies in MM.

Our study demonstrates the feasibility of applying functional genomic platforms, such CRISPR/Cas9, in the tumor cell compartment of co-culture systems. Our study also documents the ability of stromal co-cultures to remodel the molecular vulnerabilities of MM cells, and supports the notion that many genes play their role as dependencies and potential therapeutic targets preferentially or exclusively in the context of tumor-stromal interactions.