Abstract
Interventions targeting individual driver mutations in AML have not overcome resistant disease. Missing has been definition of a signal hub that integrates multiple oncogenic drivers for a composite pathway to resistance, where hub abundance determines signal(s) strength. We hypothesized that AML signal hub is p62SQSTM1, a nonmutant-active signal scaffolding for multiple pathways, including NFkB and MAPK (Moscat, Cell. 2016). Genetic epistasis in AML cell lines validated origin from P62 of such signals. Oxidative stress resistance pathways (OXPHOS) driven by p62SQSTM1 may also join with pathway(s) for leukemic progenitor replication, involving HOXA and beta-catenin within resistant AML's, but this has not been established. We sought merging pathways to OXPHOS-mediated resistance also serving to guard replicative potential through HOXA and beta-catenin. AMLs from patients were tested prior to treatment, and later failure, of induction chemotherapy, when given along with targeted agent (for cases with Flt3 mutation). Resistant AML blasts were found to have tandem overexpression of HOXAs and OXPHOS effectors p62SQSTM1, NRF2, NQO1 (NADPH oxidoreductase), BCL-2 ranging from 2-5-fold log10 relative quantity (RQ-Taqman) above chemo-sensitive core-binding factor AML, which lacks HOXA expression. We hypothesized that disrupting the downstream pathway from p62SQSTM1 to OXPHOS ending on MDM2 and/or MDMX (Todoric, Cancer Cell, 2017) is an approach to affect broad cell extinction. Indeed, besides destabilization/destruction of wild-type p53 affected by MDM2, the paralogue MDMX exists in an induced tripartite complex with CK1alpha (CK1) and p53, whose activity for restraining p53 and protecting beta-catenin from phosphorylation/inactivation is disrupted upon stress-induced CK1 activation (Huang, EMBOJ, 2020; Ueda, Cancer Cell, 2021), which we affected by using a proteasome inhibitor (PI). We found by immunoblotting AML blast lysates following PI treatment in vitro with Ixazomib, CK1 translocates to the nucleus and phosphorylation (S33,S37,T41,S45)/inactivation and cytoplasmic export of beta-catenin ensued, accompanying cell apoptosis, thus implicating MDMX inactivation resulting from PI. Primary p53-wild-type AML blasts or isogenic cell lines were treated in vitro with graded doses of the MDM2 inhibitors Pevonedistat (PEVO) or Siremadlin, either added alone, or in combination with Ixazomib at pharmacokinetically-achievable dose ranges typical in patients. Among high-risk primary AML blasts with Flt3ITD alone, or ITD along with NUP98-NSD1 or mDNMT3A, or Flt3neg with mutant EP300, or with PRDM16-SKI fusion as drivers, synergy/cooperativity was observed between MDM2 inhibitor and PI Ixazomib for inhibition of 3H-thymidine incorporation and visible apoptosis in vitro. Full characterization within PRDM16-SKI blasts by immunoblotting demonstrated expected posttranslational modification (PTM) on beta-catenin by PI, whereas distinct PTM for beta-catenin followed PEVO application, and the combination demonstrated further heterogeneity. Treatment of these blasts with Ixazomib or the combination with PEVO also led to induction of p53 in tandem with the destruction of MDM2/MDMX PTM isoforms and nuclear translocation of CK1. Strong upregulation/nuclear translocation of NRF2 resulted as part of the apoptotic pathway. To further discriminate the requirement for, and the nature of PTMs in the functional outcomes, FLAG-tagged (S33-A, S37-A,T41-A) mutant beta-catenin isoform, also lacking lysine ubiquitin target residues, was transfected in a lentiviral vector into p53-mutant K562 cells as host, and cells were compared to the same cells transfected with non-mutant FLAG-tagged beta-catenin. PEVO/Ixazomib did not induce PTM in mutant FLAG-beta-catenin transfectants, but the expected PTMs occurred in wild-type FLAG-tagged beta-catenin transfectants. In this context, PEVO became a stimulator of cell proliferation of the cells bearing mutant PTM-resistant beta-catenin, thus implicating beta-catenin disruption involving these residues as a requisite for optimal AML cell inhibition with the combination. In addition, a combination of PEVO/Ixazomib was as effective for apoptosis as was Flt3 inhibitor among certain Flt3mutant primary AML blasts. A clinical test of this strategy is warranted in high-risk relapsed/refractory p53WT/p62SQSTM1-expressing AML.
Disclosures
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
