Background: Relapsed/refractory diffuse large B-cell lymphomas (r/r-DLBCL) are a therapeutic challenge, especially in patients not suitable for high dose chemotherapy, stem cell transplantation or after CAR-T-cell therapy. Aims: We describe and characterize BTM-3566, a first-in-class compound that activates the OMA1-dependent mitochondrial stress pathway. BTM-3566 is active against a variety of B-cell malignancies but has the greatest effect in DLBCL, inducing robust therapeutic responses in vitro and in vivo. Methods: Chemical biology, animal models, functional genetics Results: BTM-3566 is a small molecule based on a pyrazolothiazol-backbone. BTM-3566 induces apoptosis and complete cell killing in DLBCL lines with an IC50 of ~300 nM, including ABC, GCB, double-hit and triple-hit lymphoma lines. BTM-3566 has > 50% of oral bioavailability and 6 hours of serum half-life. 14-day dosing in mice and dogs demonstrated excellent tolerability at therapeutic doses. In a dose-finding study using the DLBCL line SUDHL-10, once daily oral treatment with 20 mg/kg BTM-3566 for 21 days resulted in complete regression in all tumor-bearing animals. Importantly, no subsequent tumor growth was observed for 2 weeks after cessation of therapy. Expansion studies into human DLBCL PDX models harboring a range of high-risk genomic alterations, demonstrated response in 100% of the lines with complete tumor regression in 6 of 8 PDX models tested (Table 1). Transcriptome analyses revealed that BTM-3566 activates the ATF4-integrated stress response (ISR), indicated by phosphorylation of eukaryotic translation initiation factor 2α (eIF2α) and upregulation of the stress transcription factor ATF4. Of the four eIF2a-kinases in the human genome we determined that HRI (EIF2AK1) was uniquely required for BTM-3566 eIF2a phosphorylation, activation of the ATF4-ISR and induction of apoptosis. HRI is activated by mitochondrial-related stress, including heme depletion, ROS generation or blockage of mitochondrial ATP synthesis which all result in the activation of the mitochondrial protease OMA1. We found that BTM-3566 activates OMA1 in a manner unrelated to changes in mitochondrial ATP synthesis, reactive oxygen species generation or electron transport chain inhibition. OMA1 activation was required for the therapeutic effect and CRISPR-Cas9 depletion of OMA1 eliminated BTM-3566’s apoptotic activity. Substrates of OMA1 include the mitochondrial GTPase OPA1 and DELE1, a mitochondrial protein recently shown to act as a sensor of mitochondrial dysfunction and to signal through HRI kinase and ATF4. We show that downstream of OMA1 the apoptotic stress signal is relayed to HRI by DELE1, whereas OMA1-dependent OPA1 cleavage and mitochondrial fragmentation were dispensable for the therapeutic effect. Gene expression-based profiling of BTM-3566 sensitivity in over 400 cancer cell lines showed that FAM210B, a poorly characterized mitochondrial membrane protein, negatively correlated with response to BTM-3566. Overexpression of FAM210B prevented OMA1 activation and causes complete resistance to BTM-3566-induced apoptosis and cell cycle arrest. Thus, FAM210B serves as a strong predictor of BTM-3566 sensitivity, as well as revealing a novel mechanism of regulation of OMA1 activation. Image:Summary/Conclusion: We describe a novel antitumor mechanism in DLBCL, where BTM-3566 induces mitochondrial stress, activating the OMA1-DELE1-HRI-eIF2a-ATF4 pathway leading to apoptosis and tumor regression. An IND application in DLBCL has been completed with initiation of first in human clinical trials planned in 2022.
Read full abstract