Resistance to chemotherapy is a leading cause of treatment failure. Drug resistance mechanisms involve mutations in specific proteins or changes in their expression levels. It is commonly understood that resistance mutations happen randomly prior to treatment and are selected during the treatment. However, the selection of drug-resistant mutants in culture could be achieved by multiple drug exposures of cloned genetically identical cells and thus cannot result from the selection of pre-existent mutations. Accordingly, adaptation must involve the generation of mutations de novo upon drug treatment. Here we explored the origin of resistance mutations to a widely used Top1 inhibitor, irinotecan, which triggers DNA breaks, causing cytotoxicity. The resistance mechanism involved the gradual accumulation of recurrent mutations in non-coding regions of DNA at Top1-cleavage sites. Surprisingly, cancer cells had a higher number of such sites than the reference genome, which may define their increased sensitivity to irinotecan. Homologous recombination repairs of DNA double-strand breaks at these sites following initial drug exposures gradually reverted cleavage-sensitive “cancer” sequences back to cleavage-resistant “normal” sequences. These mutations reduced the generation of DNA breaks upon subsequent exposures, thus gradually increasing drug resistance. Together, large target sizes for mutations and their Top1-guided generation lead to their gradual and rapid accumulation, synergistically accelerating the development of resistance.

The major heat shock protein Hsp70 forms a complex with a scaffold protein Bag3 that links it to components of signaling pathways. Via these interactions, the Hsp70-Bag3 module functions as a proteotoxicity sensor that controls cell signaling. Here, to search for pathways regulated by the complex, we utilized JG-98, an allosteric inhibitor of Hsp70 that blocks its interaction with Bag3. RNAseq followed by the pathway analysis indicated that several signaling pathways including UPR were activated by JG-98. Surprisingly, only the eIF2α-associated branch of the UPR was activated, while other UPR branches were not induced, suggesting that the response was unrelated to the ER proteotoxicity and ER-associated kinase PERK1. Indeed, induction of the UPR genes under these conditions was driven by a distinct eIF2α kinase HRI. Hsp70-Bag3 directly interacted with HRI and regulated eIF2α phosphorylation upon cytoplasmic proteotoxicity. Therefore, cytosolic proteotoxicity can activate certain UPR genes via Hsp70-Bag3-HRI-eIF2α axis.

Finding synergistic drug combinations is an important area of cancer research. Here, we sought to rationally design synergistic drug combinations with an inhibitor of BTK kinase, ibrutinib, which is used for the treatment of several types of leukemia. We (a) used a pooled shRNA screen to identify genes that protect cells from the drug, (b) identified protective pathways via bioinformatics analysis of these gene sets, and (c) identified drugs that inhibit these pathways. Based on this analysis, we established that inhibitors of proteasome and mTORC1 could synergize with ibrutinib both in vitro and in vivo. We suggest that FDA-approved inhibitors of these pathways could be effectively combined with ibrutinib for the treatment of chronic lymphocytic leukemia (CLL).

Resistance to chemotherapy is a leading cause of treatment failure. Drug-resistance mechanisms involve mutations in specific proteins or changes in their expression levels. It is commonly understood that resistance mutations happen randomly prior to treatment and are selected during the treatment. However, selection of drug-resistant mutants in culture could be achieved by multiple drug exposures of cloned genetically identical cells, and thus cannot result from selection of pre-existent mutations. Accordingly, adaptation must involve generation of mutations de-novo upon drug treatment. Here we explored the origin of resistance mutations to a widely used Top1 inhibitor irinotecan, which triggers DNA breaks, causing cytotoxicity. Resistance mechanism involved gradual accumulation of recurrent mutations in non-coding regions of DNA at Top1-cleavage sites. Surprisingly, cancer cells had higher number of such sites than reference genome, which may define their increased sensitivity to irinotecan. Homologous recombination repair of DNA double strand breaks at these sites following initial drug exposures gradually reverted cleavage-sensitive “cancer” sequences back to cleavage-resistant “normal” sequences. These mutations reduced generation of DNA breaks upon subsequent exposures, thus gradually increasing the drug resistance. Together, large target size for mutations and their Top1-guided generation lead to their gradual and rapid accumulation, synergistically accelerating development of resistance. Abstract Figure: