Identification of FoxM1 As Therapeutic Target in TKI-Resistant Ph⁺ ALL

Despite initial responsiveness of primary Philadelphia chromosome positive acute lymphoblastic leukemia (Ph⁺ALL) to tyrosine kinase inhibition (TKI), the majority of patients will relapse and develop TKI-resistant disease. Foxm1 belongs to the forkhead box transcription factor family and is a key regulator of malignant growth by promoting cell cycle and survival through increased DNA damage repair. Foxm1 has been implicated in the progression and chemoresistance in a wide range of solid tumors, including hepatocellular carcinoma and breast cancer. Foxm1 is expressed in dividing cells and regulates the expression of critical regulators for G2/M entry of the cell cycle including Cdc25B, cyclin-B1, Plk-1 and Aurora B kinase. In addition it decreases protein stability of p27kip and p21cip and regulates the expression of antioxidant defense machinery of the cell, e.g. by superoxide dismutase expression.

We compared Foxm1 expression levels in Ph⁺ ALL patient samples and CD19⁺ B cell precursors from healthy donors and found 12-fold higher levels in the leukemic cells (p=0.011). More importantly, Foxm1 levels at the time of diagnosis in a clinical trial for patients with high risk acute lymphoblastic leukemia (ALL) were predictive of poor outcome (COG P9906; n=207). Comparative analysis of microarray data from matched sample pairs at diagnosis and relapse revealed a significant upregulation of Foxm1 in the relapse samples (n=42; p=0.0025). To further study the role of Foxm1 in Ph⁺ ALL, we developed a genetic model for inducible inactivation of Foxm1 in Ph⁺ ALL. To this end, B cell precursors of Foxm1^fl/fl mice were transformed with BCR-ABL1 and transduced with a tamoxifen (4-OHT)-inducible Cre. Interestingly, 4-OHT-mediated deletion of Foxm1 resulted in reduced cell viability and an arrest in G0/G1 with a significant decrease of the S-phase of the cell cycle following deletion of Foxm1. The ability to form colonies in vitro was significantly decreased by deletion of Foxm1. In addition, Foxm1^−/− ALL cells revealed a strikingly higher sensitivity towards TKI-treatment (Imatinib dose-response curve) compared to the control cells. As a potential therapeutic agent to pharmacologically inhibit Foxm1 function, we evaluated the effects of a previously described ARF peptide that binds and inhibits Foxm1 function. We treated TKI-resistant (BCR-ABL1^T315I) and TKI-sensitive patient-derived xenograft Ph⁺ ALL cells with various ARF peptide concentrations and found significant growth inhibition after 72h (IC50 16.8±4.3μM, n=4), regardless of TKI responsiveness. In addition, treatment of ARF peptide in combination with TKI reduced the viability from 65.7%±1.7 after TKI treatment alone (10μM) to 19%±0.8 after 48h (TKI 10μM ARF peptide 12μM). Ph⁺ ALL cells treated with similar concentrations of a mutated ARF control peptide revealed 77%±0.9 viable cells and ARF peptide treatment alone decreased the viability to 29.6%±0.4. Hence treatment with the ARF peptide alone induces apoptosis in patient-derived Ph⁺ ALL cells and enhances the effect of TKI, which confirms the findings of the ALL mouse model for human Ph⁺ ALL xenografts. In a complementary approach, we used the natural antibiotic Thiostrepton, which functions via Foxm1 blockade. To test the ability of Thiostrepton as a potential anti-leukemia agent, we studied patient-derived TKI-resistant (BCR-ABL1^T315I) Ph⁺ ALL cells. Treatment of these patient derived Ph⁺ALL cells induced cytotoxicity in nanomolar concentrations of Thiostrepton along with a significant downregulation of Foxm1 protein levels. By contrast, Non-BCR-ABL1 tumor cells including lymphoma cells were not responsive to Thiostrepton treatment at similar concentrations.

Our analyses reveal that Foxm1 is a valid therapeutic target for the treatment of TKI sensitive and resistant Ph⁺ ALL, including BCR-ABL1^T315I. We show that Foxm1 has a crucial function in Ph⁺ ALL and impacts a) leukemia proliferation, b) colony formation, and c) TKI-resistance. These findings identify Foxm1 a rational target for combination therapy with TKI or as a single agent for TKI-resistant Ph⁺ ALL.

No relevant conflicts of interest to declare.