SIRT3 Sumoylation Contributes to Chemoresistance in AML

Acute myeloid leukemia (AML) is a clonal disease originated from a rare population of malignant hematopoietic cells, called leukemic stem cells (LSCs), which is not only often resistant to standard chemotherapies, but also the major cause of relapse and eventual death of AML patients. The five-year survival of AML keeps as low as 27% for the last few decades. LSCs possess unique metabolism profiles such as higher rates of oxidative phosphorylation, and dependence on fatty acid oxidation for survival, which is distinct from normal hematopoietic cells, and, as a consequence, relatively low level of reactive oxygen species (ROS), a critical regulator for stemness maintenance. Therefore, targeting mitochondrial metabolism, especially ROS, may be a promising strategy to improve chemotherapy outcome for AML.

We have previously found in hepatocarcinoma cells that SUMOylation is one of the important post-translational modifications for a variety of cellular proteins, and is capable of regulating the enzymatic activity of some key mitochondrial enzymes involved in the metabolic control, one example of which is SIRT3, a NAD+-dependent protein deacetylase. SIRT3 is reported to influence cellular metabolism and downregulate ROS generation by deacetylating mitochondrial anti-oxidant enzymes. The targets of SIRT3 include superoxide dismutase 2 (SOD2), manganese superoxide dismutase (MnSOD) and isocitrate dehydrogenase 2 (IDH2), which have been shown closely related to leukemogenesis. Since sophisticated regulation of ROS production is required for the maintenance of LSCs, we reproduced SUMOylation of SIRT3, and investigated its role in the mitochondrial metabolism in AML. In fact, SIRT3 SUMOylation at lysine 288 was also found in AML cells.

To reveal the consequences of SIRT3 SUMOylation in AML, we constructed a plasmid expressing SIRT3-K288R that fails to be SUMOylated in AML cells. As a result, AML cells expressing SIRT3-K288R protected AML cells from as shown by apoptotic assays and quantitation of activated caspase 3 via reduction of not only total but also mitochondrial ROS production under chemotherapeutic agent-induced cell death comparing to those transfected with vector or overexpressing wild type SIRT3.

To further investigate the role of SIRT3 de-SUMOylation in AML, we examined the influence of mitochondrial metabolism and anti-oxidant enzymes by SIRT3-K288R. SIRT3-K288R significantly downregulated the acetylation of mitochondrial anti-oxidant enzymes, such as SOD2, leading to decreased NADP/NADPH ratio and increased GSH/GSSG ratio. SIRT3 de-SUMOylation enhanced OCR but impaired ECAR under both basic and cytarabine treated conditions. We analyzed 18 primary AML samples to evaluate the correlation among SIRT3 SUMOylation, ROS level and chemoresistance. As we expected, low level of SIRT3 SUMOylation correlates with low cellular ROS level in both bulk AML and CD34+CD38- AML stem cells, and less sensitivity to cytarabine.

Furthermore, MV4-11 cells bearing control vector, wild type SIRT3 or SIRT3-K288R were engrafted in NSG mice. Cytarabine was administered to the xenografts to evaluate the chemoresistance in these cell line-derived xenograft (CDX) mouse models. Consistent to the in vitro data, SIRT3-K288R reduced total and mitochondrial ROS in vivo, resulted in enhanced leukemogenesis and impaired survival.

Taken together, our study showed that SIRT3 can be SUMOylated in AML. De-SUMOylation enhances SIRT3 deacetylase activity, and contribute to the chemoresistance of AML cells via altered mitochondrial metabolism and reduced ROS generation. Thus, SIRT3 and its de-SUMOylase can be utilized as potential therapeutic targets to improve AML chemotherapy.