Drug Combinations Co-Targeting Myeloid Cell Leukemia-1 (Mcl-1) Protein Can Overcome Microenvironmentally-Induced Multi-Drug Tolerance in Non-Hodgkin Lymphomas

Abstract

Ibrutinib (IBR), an inhibitor of Bruton's Tyrosine Kinase (Btk) and venetoclax (VEN), an inhibitor of Bcl-2 have been used in Chronic Lymphocytic Leukemia (CLL), but single agent responses to these drugs are often incomplete and not durable. We previously reported synergistic cytotoxicity for the IBR+VEN combination and have initiated a trial to test that combination (NCT02419560). However, we noted extensive variability in response to IBR+VEN between patient samples ex vivo, suggesting intrinsic tolerance even to this combination. CLL cells with an “activation” phenotype (CD5+/19+/69+) that occurs from interactions with the microenvironment were less sensitive to IBR+VEN, implicating the cancer microenvironment as an inducer of drug tolerance. This was supported by our finding that CLL cells display decreased sensitivity to IBR+VEN in co-culture with agonists and cells that emulate the cancer cell microenvironment. The combination of CpG-oligodeoxynucleotides (ODN), sCD40L, and IL10 (“agonist mix”) induced near complete loss of sensitivity to IBR+VEN and upregulated Mcl-1 and Bcl-xL expression in CLL cells. However, Mcl-1 or Bcl-xL inhibitors alone were weakly effective in these cells, suggesting that exploitation of these targets requires drug combinations.Here we report a combination drug screen with an Mcl-1 inhibitor in a CLL microenvironment model ex vivo and novel drug combinations that overcome multi-drug tolerance. CLL patient PBMCs were cultured in the microenvironment model consisting of HK follicular dendritic cells, HUVEC, Jurkat T cells expressing CD40L, and CpG-ODN. Markers characteristic of activation and drug tolerance in vivo were assessed in CLL (CD5+/19+) cells using flow cytometry. We found increased expression of Ki67, CD69, and anti-apoptotic proteins Mcl-1 and Bcl-xL and decreased CXCR4 expression, as expected upon microenvironmental stimuli in vivo.Also, CLL cells from 2 IBR-treated patients were still responsive to agonist-induced drug tolerance, suggesting Btk inhibition may be less effective in blocking these microenvironmental stimuli in vivo. By comparing the ratio of anti-apoptotic proteins with cognate pro-apoptotic proteins, we noted targetable dependence of CLL cells for diverse pro-survival proteins (Mcl-1 and Bcl-xL) in our model.Flow cytometry analysis demonstrated that inhibitors of Mcl-1 (S63845), Bcl-xL (A1155643), or Bcl-2 (VEN) when used as single agents failed to induce significant apoptosis in CLL cells cultured in our model, suggesting induction of multi-drug tolerance. Drug combinations inhibiting Mcl-1 with Bcl-xL, Bcl-2, or Bcl-xL and Bcl-2 have effectively overcome multi-drug tolerance. As direct and simultaneous inhibition of multiple pro-survival proteins could be fatal to healthy cells, we carried out a Mcl-1-anchored combinatorial drug screen with inhibitors of other survival pathways in 3 CLL patient PBMCs cultured in our ex vivo model to identify drug combination(s) that induce selective toxicity in multi-drug tolerant CLL cells. Cytotoxicity was determined by analyzing cleaved PARP and dead cell staining in CLL and healthy T (CD3+/5+) cells by flow cytometry. Results showed that inhibitors of apoptotic proteins (A1155643, VEN, ABT737), IRAK4 (CA-4948, compound 26, and AS2444697), intracellular TLRs (chloroquine), Hsp90 (ganetespib), CDK (ribociclib), proteasome (bortezomib), Btk (IBR), AKT (MK2206), and HDAC (SAHA and panobinostat) or activator of PP2A (DBK1532 and NZ8061) were synergistically toxic with Mcl-1 inhibitor (S63845) in CLL cells. By comparing toxicity in CLL and healthy T cells, drug combinations targeting Mcl-1 with proteasome, IRAK4, TLRs, Btk, Hsp90, and CDK showed selective toxicity in CLL cells, indicating a potential therapeutic window for these combinations. Drug combinations targeting Mcl-1 with other apoptotic proteins were highly toxic to healthy T cells.In conclusion, CLL cells in our ex vivo model display characteristics of microenvironmentally-induced multi-drug tolerance. A similar phenotype was noted using post IBR therapy samples, suggesting IBR alone may not be effective in overcoming this multi-drug tolerance in vivo. Drug combinations targeting Mcl-1 and proteasome, IRAK4, TLRs, Btk, Hsp90, or CDK selectively overcame multi-drug tolerance in CLL cells. Thus, these combinations may be effective in patients showing intrinsic tolerance to multiple drugs. DisclosuresPortell:TG therapeutics: Research Funding; AbbVie: Research Funding; Infinity: Research Funding; Genentech/Roche: Consultancy, Research Funding; Acerta: Research Funding; BeiGene: Research Funding; Kite: Research Funding; Amgen: Consultancy. Narla:University of Michigan: Patents & Royalties: Small molecule PP2A activators. Williams:Takeda: Research Funding; TG Therapeutics: Consultancy; Seattle Genetics: Consultancy; Verastem: Consultancy; Juno: Consultancy; Sandoz: Consultancy; Celgene: Consultancy, Research Funding; Novartis: Research Funding; Astra-Zeneca: Consultancy; Abbvie: Consultancy; Pharmacyclics: Research Funding; Janssen: Consultancy, Honoraria, Research Funding; Kite: Consultancy; Gilead: Consultancy, Research Funding.