Role of human T‐cell leukemia virus type I Tax in expression of the human telomerase reverse transcriptase (hTERT) gene in human T‐cells

Abstract

The viral product Tax encoded by human T‐cell leukemia virus type I (HTLV‐I) is thought to play a central role in leukemogenesis. Clonal expansion of HTLV‐I‐infected cells requires the extension of cell division with telomere maintenance, which is regulated by the ribonucleoprotein enzyme telomerase. However, the roles of Tax in the expression of telomerase activity in T‐cells remains controversial. Our previous study indicated that expression of the human telomerase reverse transcriptase subunit (hTERT) gene, which determines telomerase activity, is tightly regulated in human T‐cells. In the present study, we investigated Tax‐mediated regulation of hTERT gene expression by Tax in human T‐cells. HTLV‐I Tax induced expression of the hTERT gene in human peripheral blood leukocytes. Reporter assays revealed that Tax activated the hTERT promoter in quiescent Kit 225 cells, while the promoter activity was repressed by Tax in proliferating Jurkat cells. Both up‐regulation and down‐regulation by Tax were mediated through the 43‐bp sequences in the promoter, which carried at least two elements that independently functioned as repressors. The two elements bound distinct factors. G1 to S phase transition induced by introduction of either cyclin D2 with cdk4 or p130‐specific shRNA also activated the hTERT promoter, implying that activation of the hTERT promoter in quiescent Kit 225 cells is associated with cell cycle progression. Our findings suggest that the cell cycle state critically influences Tax‐mediated regulation of hTERT expression. (Cancer Sci 2008; 99: 1155–1163)

Human T‐cell leukemia virus type I (HTLV‐I) is the etiological agent of adult T‐cell leukemia/lymphoma and HTLV‐I‐associated myelopathy/tropical paraparesis.^{( 1} , ^{2 )} Several lines of evidence indicate that Tax encoded by HTLV‐I plays a central role in leukemogenesis. Introduction of the Tax gene causes persistent growth of primary T‐cells in vitro, a process dependent on interleukin‐2 (IL‐2),^{( 3} , ^{4 )} and the induction of tumors and leukemia in mice in vivo.^{( 5} , ⁶ , ^{7 )} However, the molecular mechanism of leukemogenesis induced by HTLV‐I infection has not been completely elucidated.

Tax was initially identified as a trans‐acting transcriptional activator, which activates its own viral enhancer of the 21‐bp motif in the long terminal repeat (LTR).^{( 8} , ^{9 )} Subsequently, Tax was found to activate the transcription of cellular genes involved in cell growth and cell death signalings. These include genes for molecules promoting cell growth such as growth factors/cytokines, growth factor receptors, and cell cycle regulators, as well as for molecules preventing cells from apoptosis such as Bcl‐X_L, XIAP, and survivin.^{( 10 )} Tax‐modulated expression of these genes is thought to be mainly the consequence of Tax‐induced activation of enhancer elements of genes by host cellular transcription factors, such as cAMP‐responsive element–binding factor (CREB), serum‐responsive factor (SRF), and nuclear factor–kappa B (NF‐κB), which have been shown to be physically associated with Tax.^{( 11} , ¹² , ¹³ , ^{14 )} In addition, the activities of Tax that interfere with DNA polymerase β expression and functions of the NF‐κB regulator IκB, cdk inhibitor p16^INK4a, mitotic checkpoint protein MAD1, and tumor suppressor p53, have been thought to facilitate cell transformation and leukemia development.^{( 15} , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ^{20 )} Thus Tax is a potential oncoprotein that induces tumor development presumably through aberrant action of genes related to growth signaling.

To achieve clonal expansion of HTLV‐I‐infected cells, these cells must have the capacity to extend cell division beyond cellular senescence.^{( 9 )} Most human somatic cells have a limited replicative lifespan.^{( 20} , ²¹ , ²² , ^{23 )} Each replication cycle leads to a progressive telomere shortening, resulting in cellular senescence with chromosome instability.^{( 21} , ²² , ^{23 )} Transformed cells and germline cells appear to have certain compensatory mechanisms to avoid telomere shortening. One mechanism synthesizing terminal telomere sequences is mediated by the ribonucleoprotein enzyme telomerase, whose activity is restricted by expression of its catalytic subunit hTERT.^{( 24} , ^{25 )} Normal somatic cells usually express no or very low telomerase, but, in addition to germline cells, lymphocytes are exceptional in that they express certain levels of telomerase.^{( 26} , ²⁷ , ^{28 )} In fact, normal T‐cells show endogenous telomerase activity during immune reactions.^{( 29} , ^{30 )} However, the replicative senescence of normal T‐cells has been demonstrated in vivo, ^{( 31 )} suggesting that normal T‐cells keep cellular senescence, unlike HTLV‐I‐transformed T‐cells.

In the present study, we investigated HTLV‐I Tax‐mediated regulation of hTERT expression in human T‐cells. Our results clearly demonstrate that Tax activates the hTERT promoter in quiescent T‐cells, but mediates no, if any, effect or repression in growing T‐cells. We also found a novel element in the hTERT promoter involved in the Tax‐dependent modulation of promoter activity.

Materials and Methods

Cells and culture.  The IL‐2‐dependent human T‐cell line Kit 225 was maintained in RPMI‐1640 medium containing 10% fetal calf serum (FCS) and 1 nM IL‐2 (Imunace Shionogi, Osaka, Japan). The human acute lymphocytic leukemia T‐cell lines Jurkat, MOLT‐4, and the Tax‐expressing HTLV‐I‐infected T‐cell line MT‐2 and Tax nonexpressing HTLV‐I‐transformed T‐cell line ED40515(–) were maintained in RPMI‐1640 containing 10% FCS. The rat embryonic fibroblast cell line REF52 was cultured in Dulbecco's modified Eagle medium containing 10% FCS. Peripheral blood lymphocytes (PBL) were obtained from a consenting healthy adult by discontinuous density gradient sedimentation using Ficoll‐Paque PLUS (Amersham Pharmacia‐Biotech, Uppsala, Sweden). The cells were cultured in RPMI‐1640 medium containing 20% FCS with 10 µg/mL phytohemaggultinin (PHA) for 72 h. The cells were washed twice with RPMI‐1640 medium, cultured in RPMI‐1640 medium containing 20% FCS without PHA for 48 h, infected with recombinant adenoviruses, and further cultured in the absence or presence of 1 nM IL‐2 for 72 h.

Real‐time polymerase chain reaction (PCR).  Total RNA was extracted from cell pellets using Isogen (Nippon Gene, Tokyo, Japan) and polyA RNA purification was carried out using PolyATtract (Promega, Madison, WI, USA) according to the protocols recommended by the manufacturer. Quantitative detection for the telomerase component hTERT was performed with LightCycler (Roche Diagnostics, Mannheim, Germany) using the LightCycler FastStart DNA Master HybProbe and LightCycler Primer/Probe set (Roche Diagnostics) using the procedure outlined in the kit manual. The assay was repeated three times and the results were expressed as mean ± SE.

Exogenous expression of Tax.  Tax was exogenously expressed in normal cells and cell lines by infection with recombinant adenoviruses or transfection with expression plasmids, respectively. A series of expression vectors based on the human β‐actin promoter for Tax (pMT‐2Tax) and its mutants (TaxM22, Tax703, and Taxd3) were described previously.^{( 32 )} pHβAPr‐1‐neo was used as a control vector. A recombinant adenovirus for expression of Tax (AxCAIY‐Tax) and the control virus (Ad‐Con) were infected at MOI of 100 as described previously.^{( 32 )}

Plasmids.  Luciferase reporter plasmids carrying the hTERT promoter, pTERT‐S, and pTERT‐L have been previously described.^{( 33 )} The reporter plasmids pTERT(–281)‐S and pTERT(–281)‐L were generated by removing the EcoRI and BssHII, –3927 to –282, fragment at the 5′ side of the promoter from pTERT‐S and pTERT‐L, respectively. The reporter plasmids pTERT(–105)‐S and pTERT(–105)‐L were constructed by deleting pTERT(–281)‐S and pTERT(–281)‐L, respectively, using the ExoIII/Mung Bean nuclease deletion kit (Stratagene, La Jolla, CA, USA). PGL3‐Prom/43 × 3 carrying the 43‐bp (+9~+51) sequences in the hTERT promoter was generated by subcloning into the BglII site of the pGL3 promoter vector (Promega) of three tandem repeats of the 43‐bp sequences. The luciferase reporter plasmids pGL3/E2WTx4 and pGL3/E2MTx4 were generated by subcloning into the BglII site of the pGL3 promoter vector of the BglII fragments from pE2WTx4‐Luc and pE2MTx4‐Luc,^{( 32 )} which carry four tandem repeats of the adenovirus E2 enhancer with two copies of wild‐type or mutant E2F‐binding sites, respectively. PGL3/LTR‐Luc carrying HTLV‐I LTR was generated by subcloning of the HindIII fragment from pCHL4,^{( 34 )} into the HindIII site of the pGL3 basic vector. PGL3/κB‐Luc was generated by subcloning of the HindIII fragment from pκB‐Luc,^{( 16 )} into the MluI and HindIII sites of the pGL3 basic vector. PβA‐cyclinD2 were generated by subcloning the 1.0‐kb fragment from pR2^{( 35 )} into the HindIII site of pHβAPr‐1‐neo. PcDNA3‐ upstream stimulatory factor (USF)1 and pcDNA3‐USF2 were kindly provided by Dr T. Doi (Osaka University, Osaka, Japan).^{( 36 )} Expression plasmids for c‐Myc (pCMV5/c‐myc), cdk4 (pCMV/cdk4), β‐galactosidase (pCMV/β‐gal), and mutant pRb (pSM7‐LP) have been described previously.^{( 37} , ^{38 )} pshRNA‐p130 was constructed by inserting synthetic oligonucleotides complementary to a part of the p130 gene shown below into the BamHI and HindIII sites of pSilencer 2.0‐U6 (Ambion, Austin, TX, USA).

shRNA for p130:

5′‐gatcccgtatattctcagcatttccattcaagagatggaaatgctgagaatatattttttggaaa‐3′;

3′‐ggcatataagagtcgtaaaggtaagttctctacctttacgactcttatataaaaaaccttttcgaa‐5′.

Transfection assay.  Reporter plasmids were introduced into cells by the diethylaminoethyl–dextran method as described previously.^{( 34 )} After transfection, cells were cultured in the absence of IL‐2 for 40 h to induce quiescence, and were harvested for luciferase assay using the Luciferase Assay System (Promega), which was performed according to the protocol recommended by the manufacturer. Luciferase activity was normalized relative to protein concentration. REF52 cells were introduced with reporter plasmids by the calcium phosphate method as described previously.^{( 32 )} Cells were then cultured in 0.1% FCS for 48 h. In case of serum stimulation, after culturing in 0.1% FCS, cells were further cultured with FCS at a final concentration of 10% for 18 h. Luciferase activity was normalized relative to amount of β‐galactosidase from pCMV/β‐gal. All assays were performed at least three times in duplicate, and the mean ± SE values are presented.

Electrophoretic mobility shift assay (EMSA).  DNA fragments used as probes and competitors (+9~+30, +20~+40, +31~+51, random43, 43‐bp [+9~+51], +12~+14 mut 43‐bp, +46~+48 mut 43‐bp, and mutx2 43‐bp) were produced from previously described respective plasmids (pTERT‐S/+9~+30, pTERT‐S/+20~+40, pTERT‐S/+31~+51, pTERT‐S/random43, pTERT‐S/FW43, pTERT‐S/+12~+14 mut, pTERT‐S/+46~+48 mut, and pTERT‐S/mutx2, respectively).^{( 33 )} Probes were labeled with (α‐³²P)dATP (3000 Ci/mmol; GE Healthcare Bioscience, Tokyo, Japan) using the Klenow enzyme. Whole cell extracts of Kit 225 were prepared from asynchronously growing Kit 225 cells as described previously.^{( 39 )} The +9~+30 probe was incubated with 20 µg of the Kit 225 cell extract in the presence of 1 µg poly (dI‐dC) (GE Healthcare Bioscience) in 20 µL of binding buffer (10 mM Tris‐HCl [pH 7.5], 50 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 5% glycerol, and 10 mM dithiothreitol). The +31~+51 probe was similarly incubated in binding buffer with 20 µg of the Kit 225 cell extract in the presence of 3 µg herring sperm DNA (Invitrogen, Carlsbad, CA, USA). The reaction mixture was incubated for 20 min at room temperature and a 15 µL portion of reaction mixture was separated on a 4% non‐denaturing polyacrylamide gel in 6.7 mM Tris‐HCl (pH 7.5), 3.3 mM sodium acetate, and 1 mM ethylenediaminetetraacetic acid. Electrophoresis was conducted at 250 V for 2 h at room temperature. For competition experiments, 200‐time molar excess of competitor DNA was added 20 min before the addition of DNA probes. In the supershift assay, antibodies were added to the reaction mixture before the addition of the DNA probe. Antibodies against USF1 (sc‐229 X), USF2 (sc‐862 X), c‐Myc (sc‐764 X), Mad1 (sc‐222 X), Mnt (sc‐769 X), Sp1 (sc‐14027 X), and E12 (sc‐762 X) were purchased from Santa Cruz (Santa Cruz, CA, USA).

DNA fragments and oligonucleotides used as probes and competitors are as follows (underlines indicate mutant nucleotides):

43‐bp (+9~+51): 5′‐gatctGGCGCGAGTTTCAGGCAGCGCTGCGTCCTGCTGCGCACGTGGGa‐3′;

3′‐aCCGCGCTCAAAGTCCGTCGCGACGCAGGACGACGCGTGCACCCtctag‐5′.

random43: 5′‐gatctTGCAATCCGTAGCTACAGTCCGATCTACAGTCTAGTACCATTAa‐3′;

3′‐aACGTTAGGCATCGATGTCAGGCTAGATGTCAGATCATGGTAATtctag‐5′.

+12~+14 mut 43‐bp: 5′‐gatctGGCTTAAGTTTCAGGCAGCGCTGCGTCCTGCTGCGCACGTGGGa‐3′;

3′‐aCCGAATTCAAAGTCCGTCGCGACGCAGGACGAC‐GCGTGCACCCtctag‐5′.

+46~+48 mut 43‐bp: 5′‐gatctGGCGCGAGTTTCAGGCAGCGCTGCGTCCTGCTGCGCAATCGGGa‐3′;

3′‐aCCGCGCTCAAAGTCCGTCGCGACGCAGGACGACGCGTTAGCCCtctag‐5′.

mutx2 43‐bp: 5′‐gatctGGCTTAAGTTTCAGGCAGCGCTGCGTCCTGCTGCGCAATCGGGa‐3′;

3′‐aCCGAATTCAAAGTCCGTCGCGACGCAGGACGACGCGTTAGCCCtctag‐5′.

+9~+30: 5′‐gatctGGCGCGAGTTTCAGGCAGCGCTa‐3′; 3′‐aCCGCGCTCAAAGTCCGTCGCGAtctag‐5′.

+20~+40: 5′‐gatctCAGGCAGCGCTGCGTCCTGCTa‐3′; 3′‐aGTCCGTCGCGACGCAGGACGAtctag‐5′.

+31~+51: 5′‐gatctGCGTCCTGCTGCGCACGTGGGa‐3′; 3′‐aCGCAGGACGACGCGTGCACCCtctag‐5′.

Statistical analysis.  A paired t‐test was performed for statistical analysis.

Results

Effect of HTLV‐I Tax on hTERT expression in T‐cells.  The effects of Tax on expression of telomerase activity in human cells remains controversial. Tax reduced telomerase activity in HeLa and Jurkat cells,^{( 40 )} while hTERT expression was enhanced by Tax in 293T‐cells and primary T‐cells.^{( 41 )} The discrepancy prompted us to assume that the effects of Tax on hTERT expression depend on the phases of the cell cycle. To assess the hypothesis, we measured mRNA levels of hTERT in Tax‐introduced normal PBL by real‐time PCR. After exposure of T‐cells to PHA‐containing medium for 72 h, the cells were infected with a Tax‐expressing recombinant adenovirus, and further cultured in the absence or presence of IL‐2. Tax induced hTERT gene expression in the absence of IL‐2 (Fig. 1a). On the other hand, hTERT expression was not significantly changed by Tax in cells cultured in the presence of IL‐2 (Fig. 1a). These results suggest that Tax differentially modulates expression of endogenous hTERT expression in human normal PBL. To gain further insight into molecular mechanism of Tax‐mediated modulation of hTERT gene expression, we performed luciferase reporter assays to monitor hTERT promoter activity in human T‐cell lines with or without Tax. The advantage of Kit 225 cells is that their growth is arrested at the G1 phase by depletion of IL‐2. Jurkat cells were used as autonomous proliferating T‐cells. We have previously demonstrated that the hTERT promoter includes 43‐bp sequences with at least two DNA elements, which suppress hTERT expression in quiescent Kit 225 cells which are activated in response to IL‐2.^{( 33 )} A reporter plasmid pTERT(–281)‐L carries the 43‐bp sequences, but another plasmid pTERT(–281)‐S does not.

Figure 1.

Differential effects of Tax on the human telomerase reverse transcriptase (hTERT) promoter in resting and growing T‐cells. (a) Resting peripheral blood lymphocytes (PBL) were infected with a Tax‐expressing or control adenovirus, and cultured with or without interleukin‐2 (IL‐2). Levels of hTERT mRNA were measured with real‐time polymerase chain reaction (PCR). (b) Schematic structure of the hTERT promoter. The major transcription initiation site was indicated as +1. (c–h) PTERT(–281)‐L and (c–f) pTERT(–281)‐S or (g) pGL3‐promoter and pGL3‐Prom/43 × 3 were transfected along with the Tax expression plasmid (c,g) pMT‐2Tax into Kit 225, (d) Jurkat, (e) ED40515(–), or (f) MOLT‐4. MT‐2 cells were transfected with pTERT(–281)‐L or pTERT(–281)‐S (h). Cells were cultured in the absence of IL‐2 for 40 h and harvested for luciferase assay. Luciferase activity was normalized by protein content. Data are mean ± SE. *P < 0.05. RLU, relative light unit.

To examine effects of Tax on the hTERT promoter, these reporter plasmids were transfected along with the Tax expression plasmid into Kit 225 cells and cultured in IL‐2‐depleted medium. When pTERT(–281)‐S was transfected, background levels of luciferase activity in quiescent Kit 225 cells were relatively high without Tax (Fig. 1c). Similar high background levels were seen in Jurkat cells (Fig. 1d). These background levels were not significantly altered by ectopic expression of Tax. These results reflect the lack of the 43‐bp sequences functioning as repressive elements, as shown previously.^{( 33 )} On the other hand, quiescent Kit 225 cells transfected with pTERT(–281)‐L expressed a low background level of luciferase activity without Tax. Luciferase activity was significantly enhanced upon Tax introduction in Kit 225 cells even in the absence of IL‐2 but not in the presence of IL‐2 (Fig. 1c; and data not shown). The increase in hTERT promoter activity was reminiscent of that induced by IL‐2.^{( 33 )}

Jurkat cells transfected with pTERT(–281)‐L gave a high background level of promoter activity in the absence of Tax, like pTERT(–281)‐S (Fig. 1d). These results imply autonomous relief of 43‐bp sequence‐dependent repression in proliferating Jurkat cells. Interestingly, Tax significantly reduced activity of the hTERT promoter with the 43‐bp sequences, but not of the promoter lacking the 43‐bp sequences in Jurkat (Fig. 1g). Tax‐mediated reduction of hTERT promoter activity was also seen in other autonomously growing human T‐cell lines, MOLT‐4 and ED40515(–) (Fig. 1e,f). The reduction in these cell lines was dependent on 43‐bp sequences as well. Similarly 43‐bp sequence‐dependent repression was observed in Tax constitutively expressing MT‐2 cells (Fig. 1h).

Conjugation of the isolated 43‐bp sequences with the SV40 promoter in the pGL3‐Promoter vector markedly reduced promoter activity in quiescent Kit 225 cells in the absence of Tax. Cotransfection with the Tax expression plasmid induced significant enhancement of promoter activity (Fig. 1g). In proliferating Jurkat cells, activation by the isolated 43‐bp sequences was not induced irrespective of Tax expression (data not shown).

Tax‐expressing T‐cell lines have been reported to show lower telomerase activity than Tax‐negative T‐cell lines.^{( 40 )} Luciferase activity of pTERT(–281)‐L was significantly lower than that of pTERT(–281)‐S in the Tax‐expressing T‐cell line MT‐2 (Fig. 1h). Reduced luciferase activity of pTERT(–281)‐L, as compared to that of pTERT(–281)‐S, was not seen in other Tax‐negative cell lines (data not shown). Taken together, HTLV‐I Tax activates the hTERT promoter in human quiescent T‐cells; however, Tax represses the promoter already activated in proliferating cells. The Tax‐mediated modulation is via the 43‐bp sequences of the hTERT promoter.

Link between hTERT promoter activation and cell cycle progression by Tax.  We have previously demonstrated by flow cytometry that Tax is able to progress the cell cycle of quiescent PBL and Kit 225 cells in an NF‐κB pathway–dependent manner.^{( 32 )} To examine a link between hTERT promoter activation and cell cycle progression by Tax, Tax mutants were cotransfected along with pTERT(–281)‐L into quiescent Kit 225 cells. Tax mutants, Tax703 and Taxd3, were able to activate the hTERT promoter, whereas the Tax mutant TaxM22 was not effective in activation of the hTERT promoter (Fig. 2a). The same results were reproduced with a reporter plasmid with the typical E2F binding site (Fig. 2b), as shown previously.^{( 32 )} TaxM22 activated the CREB/ATF‐binding site in HTLV‐I LTR and the SRF‐binding site (CArG box), but not the NF‐κB binding site in quiescent Kit 225 cells (data not shown). Thus Tax‐mediated activation of the hTERT promoter seems to be associated with the ability of Tax to activate NF‐κB.

Figure 2.

Similarity in activation between the human telomerase reverse transcriptase (hTERT) promoter and E2F binding sites. (a,b) Kit 225 cells were transfected with expression plasmids for Tax and a series of Tax mutants TaxM22, Tax703, and Taxd3, along with the luciferase reporter plasmids (a) pTERT‐L and (b) pGL3/E2WTx4‐Luc. (c,d) REF52 cells were transfected with luciferase reporter plasmids pGL3/LTR‐Luc, pGL3/κB‐Luc, pGL3/E2WTx4, pGL3/E2MTx4, pTERT(–281)‐L, and pTERT(–281)‐S, along with (c) pMT‐2Tax or a backbone vector. After being cultured in Dulbecco's modified Eagle medium containing 0.1% fetal calf serum (FCS) for 48 h, cells were stimulated by the addition of FCS at a final concentration of 10% for 18 h (d). Luciferase assay was performed as described in the legend for Fig. 1. Data are mean ± SE. RLU, relative light unit.

Activation of the authentic NF‐κB binding site by exogenous introduction of a NF‐κB subunit p65 into Kit 225 has been shown previously.^{( 32 )} Under the same conditions, no or little, if any, activation of the hTERT promoter was observed (data not shown). These results suggest that the ability of Tax to activate NF‐κB may be essential but not enough for Tax‐mediated activation of the hTERT promoter.

Unlike Kit 225 cells, the rat embryonic fibroblast cell line (REF)52 has been shown to fail to activate the E2F‐binding site by Tax, indicating cell‐line specificity in Tax‐mediated cell cycle progression.^{( 32 )} Based on this notion, we considered whether Tax could activate the hTERT promoter in REF52 cells. The hTERT promoter, as well as the E2F‐binding site, was not affected by exogenous introduction of Tax, whereas Tax greatly activated the authentic NF‐κB binding site and HTLV‐I LTR in REF52 cells cultured without serum (Fig. 2c). Serum stimulation, which induces progression of the cell cycle, was able to activate the hTERT promoter in REF 52 cells (Fig. 2d). Serum‐dependent activation was not seen with the hTERT mutant promoter lacking the 43‐bp sequences (pTERT[–281]‐S). These results paralleled those with the wild‐type and mutant E2F binding sites. We thus assumed that activation of the hTERT promoter, which is dependent on the 43‐bp sequences, is associated with cell‐cycle progression.

Association of hTERT promoter activation with cell‐cycle progression.  To verify the assumption that G1 to S phase transition initiates activation of the hTERT promoter, we designed an experiment in which the cell cycle of quiescent Kit 225 cells was artificially prompted by ectopic introduction of either cyclin D2 with cdk4 or shRNA specific for p130. P130 is a member of the pRb family that keeps cells G1‐arrested. Both treatments were confirmed to induce G1 to S phase transition of the cell cycle by means of E2F activation (Fig. 3a,b). Those cells facilitated promoter activity of pTERT(–281)‐L, while little, if any, up‐regulation of pTERT(–281)‐S was seen. The results suggest that the 43‐bp sequences are responsible for activation of the hTERT promoter in association with G1 to S phase progression.

Figure 3.

Association of human telomerase reverse transcriptase (hTERT) promoter activation with cell‐cycle progression. Kit 225 cells were transfected with the luciferase reporter plasmids pTERT(–281)‐S, pTERT(–281)‐L, pGL3/E2WTx4, and pGL3/E2MTx4, in addition to expression plasmids for cyclin D2 with (a) cdk4, (b) shRNA for p130, and (c) pRb mutant. Luciferase activities were determined after culture for 40 h, and normalized by protein content. Data are mean ± SE. *P < 0.05. RLU, relative light unit.

We further investigated hTERT promoter activity in growing cells ectopically introduced with a pRb mutant (pSM7‐LP), which is mutated at the phosphorylation sites necessary for regulation of E2F activation. PSM7‐LP was transfected along with pTERT(–281)‐L or pTERT(–281)‐S into Jurkat cells. Introduction of the pRb mutant induced down‐modulation of E2F activity, indicating induction of cell‐cycle arrest (Fig. 3c). No or little, if any, effect of the pRb mutant was seen on the repression of the hTERT promoters in pTERT(–281)‐L and pTERT(–281)‐S (Fig. 3c). These data indicate distinct mechanisms in repression between the hTERT promoter and the E2F binding sites.

Interaction of cellular factors with the 43‐bp sequences.  To gain insight into molecular mechanisms of Tax‐mediated regulation of the hTERT promoter, EMSA was performed with two probes in the 43‐bp sequences, since our previous study showed that the 43‐bp sequences contained two subregions that function independently with similar kinetics in activation in response to IL‐2.^{( 33 )} When the +9~+30 and +31~+51 subregion fragments were incubated with the cell extract of Kit 225, both the +9~+30 and +31~+51 probes formed complexes with cellular factors (Fig. 4). Complex formation was abolished by the addition of unlabeled cognate competitors, but not of non‐cognate or mutant competitors, indicating that complex formation is sequence‐specific, and the +9~+30 and +31~+51 subregions bind different factor(s). The +31~+51 subregion carries an E‐box, which has been shown to bind c‐Myc and USF in EMSA.^{( 42} , ⁴³ , ⁴⁴ , ^{45 )} In our experimental conditions including the Kit 225 cell lysate, the addition of anti‐c‐Myc antibody did not change any pattern of EMSA bands, while supershifted bands with the +31~+51 probe were evident in the presence of anti‐USF1 and anti‐USF2 antibodies. Other antibodies for Mad1, Mnt, Sp1, and E12 did not appreciably affect complex formation of the +31~+51 probe. The addition of these antibodies, including anti‐c‐Myc, anti‐USF1, and anti‐USF2, appeared to have little effect on complex formation with the +9~+30 probe (data not shown). In addition, no competition was seen when oligonucleotides of binding sites for Sp1, Ets/PEA3, AP‐1, and Stat5 were added. These results indicate that the element, perhaps the E‐box, in the +31~+51 subregion primary binds to USF and an element in the +9~+30 subregion associates with unidentified factor(s).

Figure 4.

EMSA with two subfragments in the 43‐bp sequences. (a) The +9~+30 and (b) +31~+51 subfragments in the 43‐bp sequences were used as probes. (³²P)‐labeled probes were incubated with the cell lysates of Kit 225 cells cultured with interleukin‐2 (IL‐2). Competitors, +9~+30, +20~+40, +31~+51, +9~+51, random 43, +12~+14 mut 43‐bp, +46~+48 mut 43‐bp, and mutx2 43‐bp, were added at 200‐fold molar excess 20 min prior to the addition of the probe. Antibodies for the transcription factors indicated were incubated in the reaction mixture for 20 min at 4°C and the probe was then added.

Effects of transcription factors on hTERT promoter activity.  We then asked whether these factors influenced Tax‐induced activation of the hTERT promoter. Expression plasmids for c‐Myc and USF were transfected along with the Tax expression plasmid into quiescent Kit 225 cells for reporter assays of the hTERT promoter. Introduction of either c‐Myc or USF alone induced activation of pTERT(–281)‐L (Fig. 5a). Such activation was not seen with pTERT(–281)‐S. Interestingly, cotransfection of the Tax and c‐Myc expression plasmids profoundly elevated the hTERT promoter activity (Fig. 5a). A similar combination of Tax with USF did not change hTERT promoter activity in Kit 225 cells. The synergistic effect of the combination of Tax and c‐Myc was also seen with the E2F‐binding site, although c‐Myc alone was not effective in the activation of the E2F‐binding site in quiescent Kit 225 cells (Fig. 5b). Consistent with this, neither DNA synthesis as measured by BrdU incorporation was induced by introduction of c‐Myc or USF alone, nor was additional DNA synthesis seen in Kit 225 cells transfected with Tax and c‐Myc as compared with cells with Tax alone (data not shown).

Figure 5.

Effects of c‐Myc and upstream stimulatory factor (USF) on the human telomerase reverse transcriptase (hTERT) promoter. The reporter plasmids of either pTERT(–281)‐S, pTERT(–281)‐L (a), pGL3/E2WTx4, pGL3/E2MTx4 (b), or pTERT(–108)‐L (c) were transfected into Kit 225 cells along with the Tax expression plasmid or the expression plasmids for E‐box binding proteins USF1, USF2, and c‐Myc. USF1 and USF2 were simultaneously expressed and, in the experiments indicated, a combination of Tax with either c‐Myc or USF was used. Luciferase activities were determined after culturing for 40 h, and normalized by protein content. Data are mean ± SE. RLU, relative light unit.

It is known that the hTERT promoter carries two E‐boxes: one is in the 43‐bp sequences (+44~+49) and the other is –162~−167 upstream of the transcription initiation site. In order to evaluate the implications of the upstream E‐box, a reporter plasmid (pTERT[–105]‐L) lacking the –281~–106 region was transfected into quiescent Kit 225 cells. Either Tax or c‐Myc activated the upstream E‐box deletion mutant of the hTERT promoter, whereas USF did not (Fig. 5c).

Discussion

One major finding of this study is that HTLV‐I Tax‐mediated regulation of the hTERT promoter depends on the states of the cell cycle. Tax induced endogenous hTERT gene expression in normal resting PBL, while IL‐2‐induced growth of PBL failed Tax‐mediated induction. Further, in resting T‐cells, Tax up‐regulated the hTERT promoter in association with the cell‐cycle progression. On the contrary, Tax down‐regulated the hTERT promoter activated in growing T‐cells. Both up‐regulation and down‐regulation were mediated by the 43‐bp (+9~+51) sequences in the hTERT promoter. We have previously reported that the +9~+51 region of the hTERT promoter was an IL‐2‐responsive element.^{( 33 )} The present study demonstrates that the same region functions as a Tax‐responsive element. These results suggest that regulation of the hTERT promoter with the 43‐bp sequences may be predominantly associated with cell‐cycle state.

Tax has been shown to differentially regulate endogenous hTERT expression in human primary T‐cells; Tax‐mediated down‐regulation was observed in PHA‐stimulated primary T‐cells, while PHA‐unstimulated primary T‐cells showed Tax‐mediated up‐regulation.^{( 40} , ^{41 )} The contradiction may be reasoned by the present study. The assumption of cell cycle–dependent effects of Tax is further supported by the observations that, in REF52 cells, Tax, which does not induce G1 to S phase transition,^{( 32 )} exhibited trans‐activation of HTLV‐I LTR and the typical NF‐κB binding site, whereas Tax‐mediated induction of the hTERT promoter and the E2F binding site, which were activated by serum stimulation, was marginal.

Tax‐promoted cell‐cycle progression is thought to result from induction of a large number of genes mainly required for G1 and S phases in the cell cycle.^{( 10 )} Indeed, Tax directly activates the human cyclin D2 promoter, at least in part, in an NF‐κB‐dependent manner.^{( 46 )} Tax‐mediated activation of cyclin D2 is thought to induce the cell‐cycle progression through activation of E2F. Apart from Tax and IL‐2, we examined the effects of the induction of either cyclin D2 with cdk4 or p130 shRNA on promotion of the G1 to S phase transition in T‐cells. Both treatments were effective in activation of E2F, and subsequently hTERT promoter activation was induced (Fig. 3). Collectively, these results unequivocally demonstrate that the cell‐cycle progression is tightly associated with activation of the hTERT promoter in quiescent human T‐cells.

On the other hand, Tax has been known to inhibit proliferation in human cells, resulting in cell‐cycle arrest presumably, in part, due to Tax‐dependent stabilization of cdk inhibitors p21 ^CIP1/WAF1 .^{( 47 )} Tax‐mediated down‐regulation of the hTERT promoter in Jurkat cells may be attributable to these effects of Tax. It may be interesting to note that Tax and the pRb mutant without the phosphorylation sites caused different responses in hTERT promoter activity in Jurkat cells. Tax down‐regulated the hTERT promoter, while the pRB mutant, though down‐regulated the E2F binding site, did not give appreciable modulation in hTERT promoter activity.

Tax‐dependent up‐regulation and down‐regulation of the hTERT promoter were mediated by the 43‐bp sequences. The 43‐bp sequences contain a novel element and an E‐box (+44~+49). Myc/Max, Mad1/Max, and USF have been shown to bind to the E‐box and to activate the hTERT promoter in human cell lines.^{( 44} , ⁴⁵ , ^{48 )} Myc/Max binding to the E‐box is shown in human ovarian cancer cell lines and normal ovarian cells.^{( 42} , ^{43 )} Our experiments of EMSA, however, showed that the addition of anti‐c‐Myc antibody had no or little, if any, effect on the complex formation (Fig. 4). The difference might be attributable to different cell lines used.

Our EMSA and reporter assays demonstrate that USF bind to the downstream E‐box and activate the hTERT promoter, as shown previously.^{( 44 )} Without the upstream E‐box (–165~–160), we could not have observed USF‐dependent activation of the hTERT promoter including the downstream E‐box in Kit 225 cells, presumably indicating the requirement of USF association with the upstream and downstream E‐boxes for activation. The small contribution of USF to promoter activation in combination with Tax may implicate USF in the maintenance, rather than induction, of hTERT promoter activity.

Elimination of the 43‐bp sequences from the hTERT promoter caused elevation of the basal level of promoter activity in quiescent Kit 225 cells. The elevated level was as high as that of the wild‐type stimulated with IL‐2 or Tax. Thus the 43‐bp sequences seem to function as a repressor in the promoter. The 43‐bp sequences, in particular the +9~+30 subregion, are a human‐specific region of the TERT promoter. Though hTERT is strictly repressed in normal somatic cells, mouse TERT (mTERT) is expressed in normal somatic cells.^{( 49 )} Conjugation of the isolated 43‐bp sequences with the SV40 core promoter in the pGL3‐promoter vector also markedly reduced promoter activity in quiescent Kit 225 cells (Fig. 1g). Repression of the hTERT promoter has been previously reported to be mediated by two elements (GC‐box) of the hTERT promoter,^{( 50 )} one of which corresponds to an element in the +9~+30 subregion in our studies. The element (GC‐box) in the +9~+30 subregion has been reported to be affected by E2F in non‐lymphoid cell lines.^{( 51 )} We, however, observed that introduction of the E2F expression plasmid did not change hTERT promoter activity in quiescent Kit 225 and proliferating Jurkat cells (data not shown). This discrepancy may be due to different experimental conditions. In quiescent cells, the repressive mechanism may be essential for protection of cells from unnecessary elongation of telomere, which leads to chromosomal instability. Transition from the G1 to the S phase induced by Tax or IL‐2 in T‐cells may relieve the repression. Two independent elements in the 43‐bp sequences may secure the mechanism by binding distinct factors to each element.

In this context, it is important to identify those factors; however, so far, we have observed no distinct complex formation quiescent cell lysates with elements from that of growing cell lysates (data not shown). Alternatively, methylation of CpG sites in the 43‐bp sequences may be responsible for repression of hTERT gene expression through interference with the accession of factors to the elements in quiescent cells. This may also be the reason that Tax modulated no or little, if any, repression of the endogenous hTERT gene in Kit 225 and Jurkat cells. Experiments to clarify the molecular basis of the repressive and relief regulation involving the elements in the 43‐bp sequences are underway.

Our results suggest that in early stages of HTLV‐I infection, Tax‐mediated repression of telomerase activity may cause increases in genomic instability, resulting in the transformation of T‐cells. On the other hand, Tax‐mediated induction of the hTERT gene may contribute to extending the division capacity of HTLV‐I infected cells. Constitutive expression of the hTERT gene seen in adult T‐cell leukemia/lymphoma cells without expression of Tax may be mainly due to epigenetic modulation in hTERT expression.