Abstract
Multiple myeloma (MM) remains an incurable malignancy with a clear need for novel therapeutic modalities. Moreover, acquired or de novo resistance to established or novel therapeutics remains a major challenge in this, and other, neoplasias. BET Bromodomain inhibitors (BBIs), including JQ1, have potent anti-MM activity in vitro and in in vivo, but do not provide curative outcome and do not induce apoptosis in many tumor cell types. Recently, a "next-generation" BBI, dBET, that causes degradation of BET Bromodomains (BRDs) through CRBN-mediated ubiquitination has been demonstrated to have potent activity in leukemia and myeloma. Here we sought to compare the mechanistic differences between BRD inhibition with BRD degradation in treatment-naive and drug-resistant MM. Additionally, we posited that resistance to dBET treatment could emerge through genetic perturbations and wished to uncover potential mechanisms prior to its clinical utilization. To address this, we compared effects of JQ1 with lead optimized compound dBET6, in a panel of human MM cell lines (± stromal cells), including clones resistant to JQ1, bortezomib and IMIDs, and assessed viability using CS-BLI/CTG assay. RNAseq and reverse phase protein arrays (RPPA) were employed to compare the transcriptional and translational effects of BRD degradation vs. inhibition. Using an open-ended unbiased genome-wide CRISPR (clustered regularly interspaced short palindromic repeats)-associated Cas9 approach, we examined whether we could uncover genes associated with resistance to dBET6. MM1.S cells were transduced with Cas9 and pooled lentiviral particles of the GeCKO library, consisting of 2 pooled sgRNA sub-libraries (~120,000 sgRNAs; targeting ~19,000 genes and ~1800 miRNAs). Using this CRISPR/Cas9-based approach we sought to expedite the isolation of MM cells resistant to dBET6. We treated the pool of cells thrice with dBET (250nM), allowing regrowth between treatments and maintaining a coverage of 1000 cells/sgRNA. dBET6-resistant cells were processed to quantify sgRNA enrichment or depletion, using deep sequencing. We observed dBET6 to have significantly greater potency against MM cells than JQ1, or its combination with lenalidomide, and that MM1S.CRBN-/- cells were resistant to dBET6. Resistance to neither JQ1 nor bortezomib conferred resistance to dBET6. We observed dBET6 to induce rapid and robust (<4hrs) degradation of BRD2, BRD3 and BRD4 and loss of c-MYC protein, compared with JQ1 which caused an apparent increase in BRD4 protein and significantly less c-MYC down-regulation. Interestingly, while dBET6 caused a time-dependent reduction in pro-survival Mcl-1 protein (among others) and increased cleavage of caspase-3/7, JQ1 caused Mcl-1 upregulation and did not induce cleavage of caspase-3/7. As predicted, our CRISPR/Cas9 screen identified significant enrichment of sgRNAs targeting CRBN, as well as several members of the Cullin-RING ligase (CRL) complex, known to play a critical role in E3 ubiquitin ligase activity. Preliminary experiments using individual sgRNAs appear to validate the role the CRL complex in dBET resistance. In summary, our data strongly support the development of dBET for the treatment of treatment-naive and drug-resistant MM. We demonstrate overlapping and distinct mechanisms of action between BRD inhibition vs. degradation and suggest that differential potencies of JQ1 vs. dBET is, at least in part, due to far greater loss of c-MYC and Mcl-1 expression, however further analysis is warranted. Additionally, our results demonstrate that loss of function of CRBN or the CRL complex induces dBET resistance by perturbing dBET-mediated BRD4 degradation. However, it is plausible that additional CRBN/CRL-independent mechanisms of dBET resistance exist that allow cells to survive despite complete degradation of BRDs and this will be a key question to be answered in future studies.
Disclosures
Bradner: Novartis Institutes for BioMedical Research: Employment.
Haploinsufficiency of autism candidate gene NUAK1 impairs cortical development and behavior
