DEK Represses hTERT Expression during Leukemogenesis

hTERT encodes telomerase reverse transcriptase which is the rate-limiting factor for telomerase activity (TA). Its expression is primarily regulated at the transcriptional level. High TA participates to the immortalization of malignant cells but the positive correlation between hTERT overexpression, increased TA and leukemogenesis does not seem to be a rule as hTERT underexpression has been evidenced in chronic myeloid leukemia, adult T-cell leukemia/lymphoma (ATLL), chronic lymphocytic leukemia, and in acute myeloid leukemia (AML). Certain oncogenes such as BCR-ABL or the HTLV-1-encoded oncoprotein Tax, which are respectively involved in the development of chronic myeloid leukemia and ATLL, negatively interfere with hTERT, resulting in its transcriptional repression. To investigate the molecular control of hTERT transcriptional repression during leukemogenesis, we carried out a proteomic screening of hTERT promoter occupancy in Tax+ and Tax− cells. Nuclear extracts (NEs) were prepared from HeLa cells transfected with a Tax or a control plasmid. NEs were incubated with the hTERT core promoter (hCP) and with an irrelevant control DNA fragment (CDF). For Tax+ and Tax− cells, protein interacting with the hCP or the CDF were purified and then analyzed by mass spectrometry. 270 proteins were identified from NEs incubated with hCP. Of those, 230 were present in both Tax+ and Tax− NEs whereas 22 and 18 were identified in NEs derived from Tax− and Tax+ cells. Values for CDF were 11, 1, and 0. The Tax+-hCP peptidome included DEK, a chromatin protein first identified as a fusion protein with CAN in AML with t(6;9). DEK is overexpressed in the majority of AML and displays multifunctional properties. Cotransfections and western blotting with proteins purified from NEs showed that DEK and Tax are present along the hCP in Tax+ cells but not in cells transfected with the control vector. DEK recruitment to the endogenous hTERT promoter was also evidenced by ChIP assays in Tax+- but not in Tax--cells. Cotransfection assays with a hTERT promoter-luciferase reporter plasmid and various concentrations of DEK and Tax expressors revealed that DEK inhibits hTERT transcription in a dose-dependent manner and that Tax and DEK synergize for repressing hTERT. Physical interaction between Tax and DEK was further demonstrated in vivo by co-immunoprecipitation assays. Cotransfections with modified reporter plasmids demonstrated that DEK represses hTERT transcription through the same E-box as Tax. We showed that a phosphorylation-dependent increased DNA binding activity of DEK stimulates its negative effect on hTERT transcription. SiRNA-mediated knockdown of DEK expression abolished the negative effect of Tax on hTERT, indicating that DEK is necessary for Tax-mediated hTERT repression. We quantified by qRTPCR the expression of 3 additional E-box containing genes known to be repressed by Tax, in cells stably expressing tax before and after siRNA-mediated knockdown of DEK: DEK knockdown resulted in a 3.5-, 2-, 1.5-, 14-fold increased expression of hTERT, lck, DNA polymerase beta and p18INK4C genes, respectively. Whether DEK could be involved in hTERT repression in other hematological malignancies than ATLL was finally demonstrated by identifying DEK as a member of the hCP peptidome in bone marrow blasts derived from patients with de novo AML with hTERT transcriptional repression but not in normal bone marrow mononuclear cells derived from donors. In conclusion our proteomic analysis of hTERT transcriptional regulation allowed the detection of specific preleukemic factors and provided evidence that in proliferating HeLa cells 15% of the hTERT promoter peptidome depends on the absence (9%) or the presence (6%) of Tax. The Tax-specific factor DEK physically interacts with Tax, is recruited to the hTERT promoter in Tax+ cells in which it represses hTERT transcription synergistically with Tax, through the same downstream E-box involved in Tax-dependent hTERT repression, and in a phosphorylation-dependent manner. The interplay between Tax and DEK in repressing the transcription of additional genes suggests that previous scenarios of Tax-mediated transcriptional repression in HTLV-1 should be revisited and updated to a new model in which DEK connects tax expression and gene repression. Finally, DEK and hTERT interactions appear to pertain to other malignancies than ATLL, such as AML.