Targeting Co-Regulatory Feedback Loop of MYC and GSPT1 By MYC/GSPT1 Dual Degradation Is Synthetic Lethal in Combination with Ven/Aza in TP53 Mutant Acute Myeloid Leukemia Stem and Progenitor Cells

The MYC gene governs numerous biological processes including protein translation through regulating genes involved in ribosome biogenesis (RPSs, RPLs), translation initiation (eIFs) and translation elongation (eEFs). However, little is known about the roles of MYC in translation termination. Eukaryote releasing factors (eRFs), including eRF1 and eRF3 (also known as GSPT1) sense stop codons and terminate translation to release the completed peptides through GTP hydrolysis. MYC directly binds to the TSS of GSPT1 in multiple leukemia and lymphoma cells in the IGV dataset. We observed that MYC knockdown reduced transcription of GSPT1 in HL-60 cells. On the other hand, GSPT1 knockdown also reduced MYC protein levels and double knockdown of MYC and GSPT1 further reduced protein levels of both MYC and GSPT1, suggesting a positive co-regulatory feedback loop between MYC and GSPT1. To investigate the translational control of MYC by GSPT1, we developed a reporter system to detect stop codon readthrough (SCR) with a MYC-stop-EGFP construct in HL-60 and HEK293T cells, where GFP signals serve as readout for SCR. G418, an aminoglycoside antibiotic that induces SCR by ribosome stress, induced GFP signals in both cell types. GT19715, the first-in-class cereblon modulator (CELMoD) of MYC and GSPT1 (Nishida et al. ASH 2022, 2023), induced significantly greater GFP signals than CC-90009, a selective GSPT1 CELMoD, suggesting that dual MYC/GSPT1 degradation enhances SCR on MYC gene. GT19715 significantly upregulated integrated stress response (ISR) pathway and induced ATF3 and ATF4 protein levels in HL-60 cells. The data suggest that MYC and GSPT1 have a positive co-regulatory feedback loop that can be targeted by the dual MYC and GSPT1 degrader GT19715 with ISR induction as the mode of action.

We identified highly enriched stem-like populations overexpressing MYC in TP53 mutant AML samples compared to normal bone marrow and TP53 wild-type samples by single-cell RNA seq (Nishida et al. ASH 2023). Within CD34+ AML leukemia stem progenitor cells (LSPCs), MYC and Ki-67 protein levels, determined by single-cell mass cytometry, were elevated in CD34+CD38+ leukemia progenitors (LPs) and CD34+CD71+ megakaryoerythroid progenitors (MEPs) compared to CD34+CD38- leukemia stem cells (LSCs) or CD34- blasts in TP53 mutant AMLs, suggesting that LPs and MEPs, cell populations downstream from LSCs, utilize MYC for proliferation. Indeed, CD34+CD38+ LPs were more sensitive to GT19715 compared to CD34+CD38- LSCs in primary AML samples. Quiescent LSPCs exhibit sensitivity particularly to Venetoclax (Ven) (Zeng et al. Nat Med 2022), providing the hypothesis that the combinatorial approach of dual MYC/GSPT1 degradation with inhibition of BCL-2 targets both cell-kinetically active LPs and MEPs, and quiescent LSCs. Indeed, the combinatorial treatment of GT19715 with Ven/5’-azacitidine (Aza) exhibited synergistic cell kill (Bliss index 67.67, P = 2.11e-10) in CD34+CD38- LSCs in primary TP53 mutant AML samples (N = 7). The combinatorial treatment significantly prolonged survival of mice with PDX AML cells (PDX358) harboring TP53 p.Y220C and p.P151A compared to GT19715 or Ven/Aza monotherapy. Furthermore, GT19715 profoundly reduced AML blasts but spared CD34+CD38-CD71+CD235a+ AML cells. The combinatorial treatment nearly eradicated these remaining AML cells in bone marrow samples in another ongoing PDX AML experiment (PDX824) carrying TP53 p.Y220C and p.G105fs, suggesting that the combinatorial approach of dual MYC/GSPT1 degradation with Ven/Aza can be an effective therapeutic approach for TP53 mutant AML.

Conclusion: we identified a novel positive co-regulatory feedback loop between MYC and GSPT1, and found that dual MYC/GSPT1 degradation by GT19715 profoundly synergized in the induction of cell death in combination with Ven/Aza in both TP53 mutant quiescent and proliferating AML stem and progenitor cells.