Pyruvate Dehydrogenase (PDH) Activity Is a Source of Resistance to Sorafenib in FLT3-ITD Expressing Acute Myeloid Leukemia (AML)

Background:

AML is characterized by poor prognoses and resistance to therapy. Despite decades of research long term survival remains 30-40%, and little is known about the mechanisms involved in resistance. Cells that are resistant to cytarabine have increased oxidative metabolism and mitochondrial mass, thus suggesting a role of mitochondrial metabolism in resistance to therapy. Our lab has previously shown that genetic and pharmacologic inhibition of pyruvate dehydrogenase (PDH) sensitizes AML cells to DNA damaging agents, specifically cytarabine and doxorubicin. CPI-613 is a first in class agent that inhibits PDH and has been shown to sensitize AML cells to DNA damaging agents. How AML cells alter their mitochondrial metabolism in response to targeted therapies is not well known. Sorafenib is a tyrosine kinase inhibitor that has been shown to inhibit FLT3. Activating mutations in FLT3 are found in about 30% of AML patients with the most common mutation being an internal tandem duplication (FLT3-ITD). Sorafenib has limited activity as a single agent against FLT3 mutated AML. The degree to which mitochondrial metabolism contributes to resistance to sorafenib is currently unknown.

Methods:

A murine AML cell line expressing the FLT3-ITD (MFL2) was used to knockout PDH through the CRISPR-Cas9 system, and the safe-harbor ROSA26 locus was used as control. Knockout of PDH was confirmed by Western blot analysis. PDH knockout cells were treated with sorafenib at different doses for 24 and 72 hours. Viability assays were done using the EZQuant assay. Mitochondrial mass was assessed on PDH KO and ROSA26 control cells that were treated with sorafenib after 24-hours by staining the cells with Mitotracker and measuring uptake by flow cytometry. Apoptosis was assessed through annexin V and Propidium Iodide staining. Oxygen consumption was assessed using the Seahorse XF24 Extracellular Flux Analyzer. MOLM-13 (human AML cell line that expresses the FLT3-ITD) cells were treated with sorafenib with and without CPI-613 and combinatorial index values determined by the method of Chou and Talalay.

Results:

PDH deleted cells (PDH KO) were significantly more sensitive to sorafenib than ROSA control cells at several doses (p value <0.0001). This sensitivity is in part due to increased apoptosis, which was shown by a significant increase of Annexin V positive PDH KO cells compared to controls (p value <0.001). Interestingly, when mitochondrial oxygen consumption rates (OCR) were assessed after 24-hour sorafenib exposure, ROSA controls showed decreased OCR implying a direct effect on mitochondrial function. PDH KO cells showed low basal mitochondrial OCR that did not change with treatment of sorafenib, consistent with previous data. Mitochondrial mass was also assessed and ROSA controls showed an increase of mitochondrial mass with treatment of sorafenib at 24-hours while PDH KO cells showed no increase. CPI-613 synergized with sorafenib in MOLM-13 cells with a combinatorial index value of 0.581 (95% CI 0.4520-0.7107).

Conclusion:

Here we show that genetic deletion of PDH sensitizes FLT3-ITD positive cells to sorafenib. Mechanistically, sorafenib inhibits mitochondrial function as indicated by ROSA cells showing decreased OCR while increasing mitochondrial mass. PDH KO cells lack the mitochondrial reserve to counteract this affect resulting in increased cell death. A FLT3-ITD expressing cell line could be sensitized to sorafenib treatment by exposure to the TCA cycle inhibitor CPI-613. Considered in conjunction with previous findings, these results suggest that the PDH activity is an important factor in developing resistance to multiple classes of targeted and chemotherapeutic agents and suggests that CPI-613 could increase responses to targeted agents in AML.