Targeting the IRE1α-XBP1s Pathway to Enhance Venetoclax Effectiveness in AML

The unfolded protein response (UPR) is a signal transduction network that regulates how cells negotiate ER stress, including chemotherapy, and has been implicated in multiple human cancers. We previously reported that key components of the UPR support the pathogenesis of acute myeloid leukemia (AML). Specifically, we found that inhibition of the UPR transcription factor XBP1s reduces AML cell survival and colony formation and significantly delays the onset of disease in a genetically engineered mouse model of AML driven by MLL-AF9. This current study aims to determine the functional role and therapeutic potential of targeting XBP1s and its upstream activator, IRE1α in human AML. We first performed a retrospective analysis of XBP1 expression as well as a gene signature associated with XBP1s activation (Reactome_Activation_of_Chaperone_Genes_By_XBP1s) across gene expression profiles of bone marrow (BM) or peripheral blood (PB) samples derived from AML patients and healthy donors. Compared to healthy donors, both XBP1 expression and XBP1s activity were significantly higher in multiple genetic subtypes of AML, such as those bearing MLL-rearrangements and normal karyotype AML bearing FLT3-ITD mutations. To assess the functional role of XBP1s in AML patient cell survival, we engineered patient-derived AML (PD-AML) samples (derived from BM or PB) to express control or human XBP1s targeting shRNAs and assessed their ability to survive in co-culture with the HS-27 stromal cell line. Inhibition of XBP1s significantly reduced AML cell survival in co-culture in 5 of the 8 patient samples that were evaluated. Furthermore, we found that XBP1s inhibition reduced the colony formation of 3 additional samples in human cytokine-enriched methylcellulose. Interestingly, we observed that 6 of the 8 samples that responded to XBP1s inhibition were positive for FLT3-ITD, whereas all of the samples that were insensitive to XBP1s inhibition were FLT3-ITD-negative. While direct inhibitors of XBP1s are unavailable, compounds that target its upstream activator, IRE1α are commercially available. We first assessed how genetic inhibition of IRE1α impacts AML cell biology. Similar to XBP1s inhibition, IRE1α depletion reduced human and mouse AML cell growth and colony formation. Next, we explored how small molecule inhibitors of IRE1α impacted the viability of freshly harvested PD-AML samples at diagnosis. As IRE1α is both a kinase and an RNase, we investigated the impact of 4μ8C (RNase inhibitor) and KIRA6 (dual kinase/RNase inhibitor). Cells were treated and then analyzed 5 days later by flow cytometry to assess AML cell number and viability. Both KIRA6 (IC50= 250-500 nM) and 4μ8C (IC50= 10-20 μM) displayed anti-leukemia activity in 8 of 18 patient samples with no effect on healthy donor cell viability. We also evaluated whether IRE1α inhibitors cooperate with FDA-approved AML drugs such as cytarabine, anthracyclines or venetoclax (VEN). Using Bliss-Loewe and HSA models via SynergyFinder to calculate drug cooperativity on data from 14 treated PD-AMLs, we observed that IRE1α inhibitors selectively cooperated with VEN in an additive (n=2) or synergistic (n=8) manner to eliminate blasts. Remarkably, we found that IRE1α inhibitors rendered VEN-resistant PD-AML samples sensitive to VEN treatment (n=4). We do not observe any relationship between IRE1α sensitivity (either as a single agent or in combination with VEN) and a particular AML genetic sub-type or lesion, however, this is likely due to the small sample size, which we are currently expanding. These data suggest that targeting the IRE1α-XBP1s pathway may influence VEN efficacy and that targeting this pathway may be an effective strategy for overcoming VEN resistance in AML.