Targeted hyperactivation of oncogenic STAT5-signaling in acute lymphoblastic leukemia

Abstract Background and Significance: Oncogenic tyrosine kinases activate STAT5a and STAT5b to promote survival and proliferation in acute lymphoblastic leukemia (B-ALL). Counterintuitively, we found that high expression levels of negative STAT5-regulators (CISH and SOCS-family proteins) predict poor clinical outcomes for B-ALL patients. Likewise, our experiments based on genetic deletion of SOCS2 and SOCS3 showed that impaired STAT5 feedback inhibition resulted in loss of colony formation and cell death in B-ALL. While STAT5 functions as oncogenic driver in B-ALL, these results highlight a previously unrecognized role of STAT5-feedback inhibition in leukemia cells and suggest a “Goldilocks zone” to maintain optimal STAT5 signaling strength. Results: To assess STAT5 signaling thresholds in vivo, we modeled modest increases and decreases of STAT5-activity based on Stat5a gain- (GOF) and loss-of-function (LOF) point mutations. Stat5aGOF B-ALL cells showed increased cell mass, glycolysis (ECAR, lactate production) and accelerated proliferation. However, increased ER-stress and cellular senescence subverted colony formation and leukemia-initiation capacity of Stat5aGOF B-ALL cells. In contrast, Stat5aLOF leukemia cells were small and quiescent with 3.5-fold increased colony formation and a 21-fold increased the frequency of leukemia-initiating cells in transplant recipients. Studying Stat5aGOF- and Stat5aLOF-dependent transcriptional programs, we identified MYC- and BCL6-target genes as top-ranking gene sets. Metabolomic analyses revealed that Stat5aGOF increased 60 MYC-dependent metabolites across glycolysis, TCA cycle and amino acid metabolism pathways. In contrast, metabolomic and lipidomic analyses revealed that Stat5aLOF promoted BCL6-dependent phosphatidylethanolamine (PtdEtn) production, which is essential for autophagosome biogenesis. Stat5aLOF not only promoted PtdEtn synthesis but also increased autophagy (LC3B puncta formation and flux). Interestingly, defective autophagy, ER-stress, and senescence in Stat5aGOF B-ALL cells were largely mitigated by supplementation of PtdEtn. To leverage MYC and BCL6 as biomarker of increased and decreased STAT5-activity, respectively, we engineered patient-derived B-ALL xenografts (PDX) with N-terminal mNeonGreen-MYC and mScarlet-BCL6 fusion knockin-alleles by CRISPR and HDRT. Single-cell time-lapse experiments over 24 hours revealed that B-ALL cells with mutant JAK2 and BCR-ABL1 autonomously transitioned between cell states of high (MYC) and low (BCL6) STAT5-activity, with a period of four hours. Interestingly, inhibitors of JAK2 (ruxolitinib) and BCR-ABL1 (imatinib) suppressed STAT5 (MYC) activity and forced transition to a quiescent BCL6+ cell state. Conversely, treatment of B-ALL PDX carrying MYC- and BCL6-knockin fusions with the STAT5-agonist ABBV-CLS-484 had the opposite effect of imatinib, strongly activated STAT5 and forced transition of BCL6+ quiescent B-ALL cells into a MYC+ proliferative cell state. To assess potential therapeutic benefit of targeted STAT5-hyperactivation, we treated NSG mice bearing patient-derived BCR-ABL1 B-ALL xenografts with daily injections of 40 mg/kg ABBV-CLS-484 i.p. for two weeks. Recapitulating the effects of Stat5aGOF in murine B-ALL, ABBV-CLS-484 treatment subverted leukemia-initiation capacity and substantially prolonged survival of transplant recipient mice. Conclusions: The tyrosine kinase inhibition paradigm is based on the dependency of B-ALL cells on high-level STAT5-signaling and activation of MYC. Our findings reveal a previously unrecognized dependency of human B-ALL cells on negative STAT5-feedback regulation by CISH and SOCS and activation of BCL6. Every four hours, B-ALL cells transition between cell-states of higher (MYC) and lower (BCL6) STAT5 activity. While traditional tyrosine kinase inhibitors target the MYC-dependent cell proliferation, our findings support a rationale for targeting BCL6-dependent quiescence. The small molecule STAT5-agonist ABBV-CLS-484 was recently introduced in Phase 2 clinical trials as immunotherapy adjuvant to increase STAT5-activity in T-cells for enhanced T-cell antitumor immunity in patients with solid tumors (NCT04777994). Since ABBV-CLS-484 showed favorable safety and desirable activation of T-cell antitumor immunity, targeted STAT5-hyperactivation is amenable to near-term evaluation in patients with refractory B-ALL.