Proteogenomic Screens Identify Plasma Cell Specific Vulnerabilities to Halt Oncogenic Transcription in Multiple Myeloma

Multiple myeloma (MM) relies on oncogenic transcription programs that build upon transcriptional networks found in normal plasma cells. IRF4 serves as the master transcription factor (TF) to establish plasma cell identity. In myeloma cells, transcription is amplified by the recruitment of additional TFs, most prominently MYC, to drive the outgrowth and survival of malignant plasma cells. Targeted interference with this plasma/myeloma cell intrinsic transcription program would mark a precision-medicine approach with unprecedented activity but remains elusive in the absence of effective IRF4- or MYC-directed compounds. We therefore employed large-scale proteogenomic screens in combination with high-throughput combinatorial drug profiling to identify novel targets required for oncogenic transcription in MM.

Building upon our knowledge of central TFs in MM we first aimed to define essential regulators of both, IRF4 and MYC. To this end, we performed genome-wide CRISPR screens in three myeloma cell line models (H1112, KMS12PE, SKMM1) using endogenous IRF4 and MYC protein levels as a readout. Our screens identified 645 and 576 essential genes for IRF4 and MYC expression, respectively. An integrated analysis of the dual screens resulted in only 79 common genes driving protein expression of both TFs. This included established drug targets such as EP300 and the SWI/SNF chromatin remodeling complex (SMARCA4, SMARCB1, SMARD1), as well as plasma cell specific dependencies (IRF4, XBP1), validating our screening approach.

Gene set enrichment analysis of common screening hits prioritized general regulators of RNA polymerase II-based transcription and translation initiation factors. However, these gene sets were strongly comprised by pan-essential genes, prompting us to filter our data for common essential genes. Surprisingly, this analysis ranked the mediator complex gene MED12 at the top of the list, followed by IRF4. MED12 is involved in transcriptional regulation as part of the CDK8 subcomplex, composed of MED12, MED13, CDK8 and CCNC. We identified MED13 and CCNC as individual hits in our MYC and IRF4 screen, respectively (CDK8 was not included in the sgRNA library). Moreover, cyclin C emerged as another myeloma specific dependency. To corroborate this finding, we generated an IRF4-BioID2 fusion construct to identify proteins in close proximity (approx. 40nm) to IRF4 via mass spectrometry. Indeed, we identified not only the SWI/SNF chromatin remodeling gene family and common mediator complex members, but also CDK8 suggesting the CDK8 subcomplex as potential target to interfere with the IRF4-MYC autoregulatory loop in MM.

We next tested three different CDK8-targeting small molecule compounds that demonstrated single-agent activity in myeloma cell line models. To reinforce this finding, we conducted high-throughput drug screens in three myeloma cell lines with different genetic background. Using recently discovered inhibitors of SWI/SNF activity as backbone drugs for interfering with IRF4/MYC activity (Bolomsky et al., Cancer Cell, 2024), we screened against 2804 mechanistically annotated compounds. Enrichment analysis for top synergy hits revealed three common drug classes that scored consistently: MEK inhibitors, mTOR inhibitors, as well as all CDK8 inhibitors. Validation experiments in multiple cell lines confirmed synergistic activity between CDK8 inhibitors and other IRF4/MYC targeting compounds (e.g. IMiDs, SWI/SNF inhibitors) independent of the genetic background (e.g. RAS mutation status). These data suggest that MED12 and the CDK8 subcomplex participate in oncogenic transcription in MM, and selective small molecule inhibition of this subcomplex in combination with established agents represents a potential novel precision medicine strategy in MM.