BCR-ABL kinase inhibitors have dramatically improved treatment of patients with chronic myeloid leukemia (CML). Nevertheless, the disease is associated with genomic instability, resulting in unresponsiveness to ABL inhibitors and resistance ensues, indicating a need for additional therapeutic options. CML and other malignancies are associated with metabolic reprogramming, allowing transformed cells to cover their enhanced energy needs and to provide biomolecules for anabolic processes. The mechanisms that result in altered energy metabolism in cancers are diverse. In CML, BCR-ABL has been shown to induce a substantial increase in glucose uptake and lactate production. Malignant cells may also increase the conversion of glucose to glycogen through activation of specific glycogen synthases. We found that in Ph+ KU812 cells, BCR-ABL is responsible for a substantial increase in glycogen and ABL kinase inhibition reduces glycogen production by >61%, (p<0.05, n=3). Similar changes were also found in cell lines derived from AML patients transformed by JAK2.V617F (HEL) and FLT3-ITD (MOLM13) in response to their respective tyrosine kinase inhibitors. Further, glycogen production in KU812 cells was strictly dependent on the presence of glucose in the culture medium. Our analysis suggests that the rate-limiting glycogen synthase 1 (GYS1) is highly expressed in hematopoietic cells, while the liver glycogen synthase 2 (GYS2) is not. Using a lentiviral-based RNAi approach, we found GYS1 to be required for glycogen synthesis in KU812 cells. GYS1 knockdown with three different shRNA constructs also reduced cell growth by 28.7 to 47.3% (p<0.05, n=3) compared to scrambled shRNA. Whereas suppression of GYS1 lowered cellular glycogen levels, it did not significantly alter lactate production or mitochondrial metabolism. GYS1 is thought to be regulated mostly through inhibitory serine phosphorylation by glycogen synthase kinases 3α or 3β (GSK3), which are in turn inhibited by the AKT kinase. This is supported by our findings that glycogen levels in AKT inhibitor treated KU812 cells are reduced (77.9% of control, p<0.05, n=3). Also, inhibitory GSK3 phosphorylation was reduced in response to inhibition of ABL kinase activity in KU812 cells and AKT kinase activity in BaF3 cells expressing myristoylated AKT. Interestingly, reducing the inhibitory GSK3 phosphorylation in either model did not increase phosphorylation of the GSK3 substrate GYS1 at its inactivation site Ser641. Instead, our results show alternative regulation of GYS1 by BCR-ABL through increased mRNA and protein expression. In addition, our preliminary data have already demonstrated that BCR-ABL is required for increased glycolytic flux and lack of its activity reduces glucose uptake and would therefore also reduce levels of the allosteric GYS1 activator glucose-6-phosphate. Overall, these data suggest that GYS1 expression is required for increased growth and is regulated at least in part through GSK3-independent events, hinting at novel targets for drug development. Small molecule drugs that target GYS1 would be expected to specifically inhibit this pathway and to have activity in drug-resistant CML and other malignancies with high GYS1 activity.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.