The human T-cell leukemia virus type-1 p30II protein activates p53 and induces the TIGAR and suppresses oncogene-induced oxidative stress during viral carcinogenesis

Abstract

In normal cells, aberrant oncogene expression leads to the accumulation of cytotoxic metabolites, including reactive oxygen species (ROS), which can cause oxidative DNA-damage and apoptosis as an intrinsic barrier against neoplastic disease. The c-Myc oncoprotein is overexpressed in many lymphoid cancers due to c-myc gene amplification and/or 8q24 chromosomal translocations. Intriguingly, p53 is a downstream target of c-Myc and hematological malignancies, such as adult T-cell leukemia/lymphoma (ATL), frequently contain wildtype p53 and c-Myc overexpression. We therefore hypothesized that p53-regulated pro-survival signals may thwart the cell's metabolic anticancer defenses to support oncogene-activation in lymphoid cancers. Here we show that the Tp53-induced glycolysis and apoptosis regulator (TIGAR) promotes c-myc oncogene-activation by the human T-cell leukemia virus type-1 (HTLV-1) latency-maintenance factor p30^II, associated with c-Myc deregulation in ATL clinical isolates. TIGAR prevents the intracellular accumulation of c-Myc-induced ROS and inhibits oncogene-induced cellular senescence in ATL, acute lymphoblastic leukemia, and multiple myeloma cells with elevated c-Myc expression. Our results allude to a pivotal role for p53-regulated antioxidant signals as mediators of c-Myc oncogenic functions in viral and non-viral lymphoid tumors.

Keywords: ATL; HTLV-1; Oncogene; ROS; TIGAR; p53.

Fig. 1

The HTLV-1 p30^II protein activates p53 and induces oncogenic cellular transformation with c-Myc which is dependent upon p53 transcriptional activity. (A) The predicted structure of the HTLV-1 p30^II protein with amino acid residues comprising the TIP60-binding domain (aa residues 99–154) highlighted in white (Awasthi et al., 2005; Romeo et al., 2015). (B and C) Primary hu-PBMCs were transduced with pLenti-HTLV-1 p30^II-GFP, pLenti-HTLV-1 p30^II (HA-tagged), pLenti-GFP control, or empty pLenti 6.2/V5-DEST vector and cultured in the presence of recombinant hu-IL-2 (50U/ml) and selected on Blasticidin (5 μg/ml). Untransduced mock cells are shown for comparison. The transduced cultures were repeatedly passaged and monitored for long-term proliferation, as defined by continuous growth beyond crisis (> 4 months). Data presented in the graph are from 192 experimental replicates. The scale bars represent 0.2 mm. (D) Immunoblot analysis of p30^II-GFP expression in Molt-4 lymphoblasts transduced with various dilutions of a concentrated lentiviral HTLV-1 p30^II-GFP virus stock. (E) The induction of p53 protein expression in human HFL1 fibroblasts that were transduced with pLenti-HTLV-1 p30^II-GFP or pLenti-HTLV-1 p30^II (HA) expression vectors, as compared to a pLenti-GFP negative control, was visualized by immunofluorescence-microscopy. Untransduced mock cells are also shown. DAPI-nuclear staining is provided in the DIC overlay images. The scale bars represent 20 μm. The nuclear colocalization between the HTLV-1 p30^II (HA-tagged, green) and p53 (red)-specific fluorescent signals in the bottom panels was quantified using Carl Zeiss Axiovision 4.8 software and is represented in the heat-map graph. (F-H) The induction of p53 protein expression by HTLV-1 p30^II-GFP or HTLV-1 p30^II (HA) in transfected HeLa cells, Jurkat lymphocytes, and cultured hu-PBMCs was detected by SDS-PAGE and immunoblotting. A pCEP4-wildtype p53 expression construct was included (F, right lane) as a positive control. Relative Actin levels are shown for reference. (I) HT-1080 fibrosarcoma cells were cotransfected with a pG13-luc reporter plasmid, which contains two copies of the consensus p53-binding sequence (El-Deiry et al., 1993), and increasing amounts of pCEP4-wildtype p53 (as a positive control) or pEGFP-N3-HTLV-1 p30^II-GFP, or the pLenti-HTLV-1 p30^II-GFP expression construct. Relative luciferase activities were measured and normalized for equivalent total cellular protein levels. The averaged data from three experiments are shown. (J) Oncogenic foci-formation was assessed by cotransfecting HFL1 fibroblasts with expression constructs for c-Myc and HTLV-1 p30^II-GFP, in the presence of increasing amounts of pCEP4-wildtype p53 or pCEP4-p53-R175H (a dominant-negative DNA-binding mutant of p53; Hermeking et al., 1997). The transfected cultures were monitored for foci-formation (i.e., loss of cellular contact-inhibition) over a three-week period (see s). Error bars represent the standard deviation between replicate data sets (n=3) for the duplicate experiments shown.

Fig. 2

HTLV-1 p30^II induces the expression and mitochondrial localization of TIGAR which is required for its oncogenic cooperation with c-Myc. (A) Human HFL1 fibroblasts were transduced with pLenti-HTLV-1 p30^II-GFP, pLenti-GFP, or the empty pLenti 6.2/V5-DEST vector and the expression of TIGAR was visualized by immunofluorescence-microscopy. The HTLV-1 p30^II-GFP and GFP were detected by direct-fluorescence. DAPI nuclear-staining and Texas Red-Phalloidin staining of the Actin cytoskeleton are provided for reference. The scale bars represent 20 μm. (B) The mitochondrial localization of TIGAR in HFL1 fibroblasts transduced with pLenti-HTLV-1 p30^II-GFP was confirmed by immunofluorescence-microscopy in cells that were co-labeled with MitoTracker Orange. DAPI nuclear-staining is provided in the merged image. (C and D) The induction of p53 and TIGAR protein expression in 293 HEK cells and Jurkat lymphocytes expressing HTLV-1 p30^II-GFP was detected by immunoblotting. The cells were also transfected with pCEP4-wildtype p53 as a positive control (Hermeking et al., 1997). (E) Densitometry quantification of the TIGAR protein expression in 293 HEK cells shown in C. (F) HT-1080 fibrosarcoma cells were repeatedly transfected with an siRNA targeted against tigar transcripts (siRNA-tigar) or a scrambled RNA (scrRNA) control and the knockdown of endogenous TIGAR expression was detected by immunoblotting. (G) HT-1080 cells were transfected with a CMV-TIGAR (FLAG-tagged) expression construct (Bensaad et al., 2006) and siRNA-tigar or a scrRNA control, and the knockdown of FLAG-tagged TIGAR was detected by immunoblotting. (H) The effects of TIGAR overexpression or siRNA-tigar knockdown of TIGAR expression upon oncogenic foci-formation by HTLV-1 p30^II-GFP and c-Myc were determined by cotransfecting HFL1 fibroblasts and then monitoring the formation of transformed colonies over a three-week period. The scrRNA was included as a negative control. The averaged data from three experiments are shown. (I) The expression of the HTLV-1 p30^II-GFP fusion was visualized in the transformed colonies by direct-fluorescence microscopy. DIC phase-contrast images (right panels) are provided for reference.

Fig. 3

The induction of TIGAR by HTLV-1 p30^II prevents the accumulation of oncogene-induced cytotoxic ROS. (A-C) The inhibition of c-Myc-induced ROS by HTLV-1 p30^II (HA) or TIGAR (FLAG) was determined by transfecting HT-1080 cells with CβF-c-Myc and then transducing the cultures with pLenti-HTLV-1 p30^II (HA) or the empty pLenti 6.2/V5-DEST vector. In certain samples, CMV-TIGAR (FLAG), siRNA-tigar, or the scrRNA control was also included. The cells were stained with the fluorescent ROS-specific chemical probe, CM-H₂DCFDA, and the number of ROS-positive cells per field was counted in duplicate or triplicate using fluorescence-microscopy. DIC images are provided for comparison. Replicate data sets are shown for single representative experiments in panels B and C. (D) The expression of TIGAR, c-Myc, and p53 proteins in HT-1080 cells containing HTLV-1 p30^II and/or oncogenic c-Myc, and cotransfected with siRNA-tigar or the scrRNA negative control, was detected by SDS-PAGE and immunoblotting. Relative tubulin levels are shown as a protein-loading control. (E) The levels of intracellular ROS in HT-1080 cells expressing various combinations of c-Myc, HTLV-1 p30^II (HA), TIGAR (FLAG), or the empty pLenti-6.2/V5-DEST vector and either siRNA-tigar or a scrRNA control were determined by measuring the relative fluorescence-intensities of the CM-H₂DCFDA fluorescent probe within individual cells using Carl Zeiss Axiovision 4.8 software. Each data point in the graph represents an average of 19 cells.

Fig. 4

The infectious HTLV-1 ACH.p30^II mutant provirus, defective for p30^II production, is impaired for the induction of TIGAR. (A) The expression of TIGAR in HT-1080 cells stably expressing the HTLV-1 ACH.wildtype and ACH.p30^II mutant proviral clones was detected by immunoblotting. Actin protein levels are shown for comparison. (B) The induction of TIGAR expression in HT-1080 cells containing the HTLV-1 ACH.wildtype or ACH.p30^II mutant provirus was visualized by immunofluorescence-microscopy. Mitochondrial localization of the TIGAR protein was detected by co-labeling the cells with MitoTracker Orange. The expression of HTLV-1 gp21 was detected by immunofluorescence-microscopy. DAPI nuclear-staining is provided in the merged images. The scale bars represent 10 μm. (C) The relative levels of TIGAR protein in cells containing the HTLV-1 ACH.wildtype or ACH.p30^II mutant provirus (gp21-positive cells) was measured using Carl Zeiss Axiovision 4.8 software. The graphed data represent 10 individual cells per sample; and the relative fluorescence-intensities of the TIGAR signal were quantified over a constant 5 μm² surface area in each cell. (D) Equivalent virus production by HT-1080 clones stably expressing the HTLV-1 ACH.wildtype or ACH.p30^II mutant provirus was confirmed by quantifying the amounts of extracellular p19^Gag core antigen by ELISAs (Zeptometrix). Virus-containing supernatants were collected at 1-week and 2-week intervals; and data from two representative clones are shown relative to an HTLV-1 p19^Gag protein standard curve. The SLB1 (HTLV-1-producing) lymphoma cell-line was included as a positive control. (E) The relative levels of TIGAR and c-Myc were compared in HTLV-1-transformed lymphoma T-cell-lines (MJG11 and SLB1) and activated cultured hu-PBMCs. Actin is shown for reference. (F) TIGAR protein expression in HTLV-1-transformed SLB1 lymphoblasts was compared to activated cultured hu-PBMCs obtained from three different donors. (G and H) The TIGAR is overexpressed in cultured (ATL-1, ATL-4, and ATL-7) and primary (ATL-8, ATL-9, ATL-10, ATL-11, and ATL-12) HTLV-1-infected ATL clinical isolates and correlates with increased c-Myc expression, as compared to activated cultured hu-PBMCs. (I) The relative expression of TIGAR in HTLV-1-infected ATL-1 lymphoblasts was compared to uninfected, activated cultured hu-PBMCs by admixing the samples and then immediately performing immunofluorescence-microscopy. The HTLV-1-infected ATL-1 samples were differentially identified by immunostaining to detect gp21. DAPI nuclear-staining is provided in the TIGAR signal (red) micrographs. DIC images are also shown for reference. (J) The relative levels of TIGAR were compared in admixed hu-PBMCs that were either transduced with pLenti-HTLV-1 p30^II-GFP or mock treated. DAPI nuclear-staining is provided in the TIGAR signal (red) micrographs; and the HTLV-1 p30^II-GFP and TIGAR signals are merged in the DIC overlay images. The scale bars represent 10 μm.

Fig. 5

siRNA-knockdown of TIGAR expression sensitizes viral and non-viral hematological tumor cells to oncogene-induced ROS. (A and B) The HTLV-1 p30^II and TIGAR proteins prevent c-Myc-induced cellular senescence in HT-1080 cells. The cells were first transfected with CβF-c-Myc and/or CMV-TIGAR (FLAG) and then transduced with pLenti-HTLV-1 p30^II (HA) or the empty pLenti-6.2/V5-DEST vector. After five days, the cultures were fixed and stained using X-Gal to detect senescence-associated Beta-galactosidase expression (blue signal). Certain samples were also cotransfected with siRNA-tigar or a scrRNA control. The scale bars represent 20 μm. Replicate data sets are shown for a single representative experiment. (C-E) HTLV-1-transformed SLB1 lymphoma cells were repeatedly transfected with siRNA-tigar or a scrRNA control and the cultures were stained using the fluorescent ROS-specific chemical probe, CM-H₂DCFDA, and fluorescence-microscopy was performed to visualize and quantify intracellular ROS accumulation (C, top panels). The transfected SLB1 cells were then stained with X-Gal to detect senescence-associated Beta-galactosidase expression (C, lower panels). Wide-field images of cellular senescence (dark stained cells in DIC + X-Gal) are provided below the color micrographs. The chemical uncoupler, CCCP, was included as a positive control. (F) The expression of TIGAR and c-Myc in acute lymphoblastic leukemia (ALL: CCRF-CEM, CCRF-HSB2, Molt-4, and Rs4;11), multiple myeloma (MM: MM.1R, NCI-H929, RPMI 8226, and U266B1), and Sezary syndrome (SS: HuT-78) tumor cell-lines, compared to activated cultured hu-PBMCs, was detected by immunoblotting. Actin protein levels are shown for reference. (G and H) ALL (RS4;11) and MM (NCI-H929) tumor cell-lines were repeatedly transfected with siRNA-tigar or a scrRNA control, and then stained with CM-H₂DCFDA to detect intracellular ROS by fluorescence-microscopy. CCCP was included as a positive control. Replicate data sets for representative individual experiments are shown in the graphs.