Aberrant Epigenetic Suppression of the Ptprot Gene Encoding a Protein Tyrosine Phosphatase Targeting KIT in Acute Myeloid Leukemia

Abstract 1322

Recent studies have suggested that deregulated expression of tyrosine phosphatases and the resulting alteration in the phosphorylation status of their substrates play a significant role in the oncogenic function of tyrosine-phosphorylated proteins. Over the past decade we have been studying the tumor suppressive function of protein tyrosine phosphatase receptor-type O (PTPRO). The gene for PTPRO encodes two functional isoforms, full-length form (PTPRO-FL) and a truncated form (PTPROt) that are expressed in a tissue-specific manner. PTPROt is primarily expressed in hematopoietic cells. It is, however, transcriptionally and epigenetically suppressed in CLL-like cell lines and primary CLL [Clin Cancer Res. 2007 Jun 1;13(11):3174–81, Blood. 2011 Dec 1;118(23):6132–40]. We have now shown that PTPROt is also hypermethylated in a discovery set of primary AML samples (n=77) provided by the University hospital Ulm biobank relative to bone marrow from normal controls. Analysis of the AML cell lines Kasumi-1 and ME-1 cells also showed dramatically reduced PTPROt expression relative to THP-1 and MV4-11 cells. Treatment of Kasumi-1 cells with the hypomethylating agent decitabine led to re-activation of PTPROt at both RNA and protein levels. PTPRO CpG island (CGI), methylated in Kasumi-1 cells, became hypomethylated following treatment with decitabine. Similarly, bone marrow samples from elderly AML patients who received decitabine 20 mg/m2/day – 10 days on the OSU 07017 study exhibited hypomethylation and upregulation of PTPROt. To further evaluate the functional significance of hypermethylation and silencing of PTPROt in AML, we then searched for potential kinase substrates of the protein. Kasumi-1 and ME-1 cells (where PTPROt is suppressed) are both CBF cell lines characterized by RUNX1/RUNX1T1 [(8;21) translocation] and CBFB/MYH11 [inv(16)], respectively. While CBF AML patients are generally classified within a more favorable cytogenetic group, those cases harboring mutation in the KIT gene that results in constitutively active receptor tyrosine kinase have a poor outcome. In addition, increased expression of the encoded kinase protein occurs in ∼80% of CBF AML, regardless of KIT mutational status. Kasumi-1 and ME-1 cells are also characterized by mutated (constitutively active) and over-expressed KIT, respectively. Since tyrosine phosphorylation regulates the enzymatic activity and oncogenic function of KIT protein we hypothesized that this kinase could be a substrate of PTPROt and that expression of PTPROt would be critical to maintain control of its activity. Indeed, using in vitro substrate-trapping assay we demonstrated that KIT is a direct substrate of PTPROt. Further, in vivo studies conducted by co-transfecting KIT^D816V and PTPROt (wild type or catalytic site mutant) in H293T cells showed that phosphorylation of KIT at Y719, a read-out for KIT activity, was reduced when KIT was co-expressed with PTPROt-WT but not with vector control or catalytic site mutant of PTPROt. These observations suggest that suppression of PTPROt in CBF AML with over-expressed or mutated KIT could contribute to the leukemogeneic function of KIT. Given that the epigenetic suppression of PTPROt can be reversed by hypomethylating agent decitabine, it is possible that combination of decitabine and tyrosine kinase inhibitors with or without chemotherapy may represent a novel therapeutic approach in CBF AML. [Supported by grant CA101956]