PI3Kδ inhibition enhances trametinib efficacy by suppressing feedback-activated PI3K/AKT signaling in PTPN11-mutant JMML

Abstract Juvenile myelomonocytic leukemia (JMML) is a rare but aggressive pediatric hematopoietic malignancy driven by mutations in the Ras signaling pathway. Gain-of-function mutations in PTPN11 occur in approximately 35% of patients and are strongly associated with poor prognosis. The most recent clinical trial demonstrates that the MEK inhibitor Trametinib shows efficacy in certain JMML subtypes; however, patients harboring PTPN11 mutations exhibit limited or no response (Stieglitz E, et al. Cancer Discovery, 2024), indicating the existence of alternative resistance mechanisms. The RAS signaling pathway is tightly controlled by negative feedback mechanisms that limit signal duration and amplitude. In addition to activating the canonical MAPK cascade, RAS also engages parallel pathways such as PI3K/AKT, allowing for compensatory signaling. MEK inhibition with Trametinib may relieve this feedback, resulting in sustained RAS activity and PI3K/AKT pathway activation. In this study, we investigated whether such feedback-driven signaling underlies Trametinib resistance in PTPN11-mutant JMML and evaluated a rational combination approach to counteract this resistance. To model PTPN11-driven JMML, we employed murine HCD-57 erythroleukemia cells stably expressing JMML-associated PTPN11 mutants, which recapitulate key disease features, including cytokine-independent growth and constitutive RAS pathway activation. Although Trametinib effectively suppressed ERK phosphorylation, it nevertheless enhanced AKT phosphorylation, indicating compensatory activation of the PI3K/AKT axis. Pulldown assays revealed elevated levels of intracellular RAS-GTP following Trametinib treatment, suggesting a disruption of negative feedback regulation. RNA-seq analysis further demonstrated downregulation of multiple RAS GTPase-activating proteins (RAS-GAPs), including Rasa2, Rasa3, and Rasal3, findings that were independently validated by quantitative PCR. Concurrently, expression of RAP1GAP was upregulated, which likely reduces RAP1-mediated sequestration of RAS effectors, thereby facilitating RAS activation. These data collectively support the notion that MEK inhibition relieves feedback inhibition on RAS, resulting in its hyperactivation and diversion of signaling through bypass pathways such as PI3K/AKT. To counteract compensatory PI3K/AKT activation under MEK inhibition, we tested Trametinib combined with Idelalisib, a selective PI3Kδ inhibitor. In PTPN11-mutant HCD-57 cells, the combination showed strong synergy by Bliss analysis, inducing greater G1 arrest and apoptosis than single agents, with concurrent suppression of p-ERK and p-AKT. Similar effects were seen in primary JMML cells ex vivo, with reduced viability, increased apoptosis, and impaired colony formation. In a PTPN11-mutant HCD-57 xenograft model, combination therapy significantly outperformed single agents by prolonging survival, decreasing leukemic burden in spleen and bone marrow, and reducing infiltration in spleen and liver, with leukemic cells showing dual pathway suppression in vivo. The SHP2-E76K transgenic mouse model, characterized by leukocytosis, thrombocytopenia, anemia, and splenomegaly, exhibited greater normalization of blood counts and spleen size after combination treatment compared to monotherapies. In a JMML patient-derived xenograft (PDX) model, limiting dilution assays revealed that combination therapy more effectively reduced leukemia-initiating cells than either drug alone, demonstrating superior stem cell targeting. Across models, combination treatment was well tolerated with no overt toxicity. Collectively, our findings demonstrate that PTPN11-mutant JMML cells activate compensatory PI3K/AKT signaling in response to MEK inhibition, driven by feedback disruption within the Ras pathway. Selective PI3Kδ inhibition with Idelalisib effectively suppresses this compensatory mechanism and enhances the antitumor efficacy of Trametinib. This dual blockade strategy provides a mechanistic rationale and preclinical proof-of-concept for combination therapy in PTPN11-mutant JMML, which is historically resistant to single-agent MEK inhibitors. These results support future clinical investigation of MEK and PI3Kδ co-inhibition to improve treatment outcomes in this high-risk JMML subgroup.