• Open Access

c-Maf suppresses human T-cell leukemia virus type 1 Tax by competing for CREB-binding protein

Authors


To whom correspondence should be addressed.
E-mail: hieshima@med.kindai.ac.jp

Abstract

Latent infection of human T-cell leukemia virus type 1 (HTLV-1) is considered to be preferentially associated with CCR4+ CD4+ T cells. Here we report that c-Maf, one of the critical transcription factors for Th2 differentiation, suppresses the transcriptional activity of HTLV-1 Tax by competing for CREB-binding protein. Notably, c-maf expression is selectively induced in a fraction of CCR4+ CD4+ T cells upon activation. Furthermore, c-Maf significantly decreases Tax-induced HTLV-1 envelope gp46 gene expression from an infectious HTLV-1 molecular clone and tax expression in a cell-free HTLV-1 infection system. Collectively, c-Maf may play a role in latent infection of HTLV-1 in CCR4+ CD4+ T cells by negatively regulating Tax activity. (Cancer Sci 2011; 102: 890–894)

Human T-cell leukemia virus type 1 (HTLV-1) is the causative agent of adult T-cell leukemia (ATL).(1) Although HTLV-1 infects both CD4+ and CD8+ T cells,(2,3) leukemic cells are mostly CCR4+ CD4+ T cells.(4) Given that CCR4 is selectively expressed by T-cell subsets such as Th2 cells, skin-homing memory/effector T cells and regulatory T cells (Treg), ATL cells may originate from one of these T-cell subsets.(5–7) HTLV-1 encodes the potent transcriptional activator Tax, which strongly induces the expression of viral genes including tax itself and also various cellular genes, leading to strong growth promotion.(8,9) However, Tax is also a good target for host cytotoxic T lymphocytes (CTL),(10,11) and circulating ATL cells usually do not express tax.(1) Thus, downregulation of tax by factors such as virally encoded HBZ may be critical for the survival and latency of HTLV-1-infected T cells in the presence of strong host immune responses.(12,13)

c-Maf is a transcription factor critically involved in the differentiation of Th2 cells and presumably in some Treg.(14,15) Approximately 30% of ATL cases were reported to be positive for c-Maf.(16) However, the role of c-Maf in latent infection of HTLV-1 in CD4+ T cells is not known. Here we report that c-Maf partially suppresses Tax activity by competing for CREB-binding protein (CBP), and thereby may promote latent infection of HTLV-1 in c-Maf-expressing CCR4+ CD4+ T cells.

Materials and Methods

Cells.  HTLV-1-transformed cell lines, SLB-1, C8166, MT-2 and MT-4, and a HTLV-1-negative human T-cell line, Jurkat, were used. The PBMC were isolated from heparinized venous blood samples of healthy donors and acute ATL patients with a high leukemic cell count (>90%) by using Ficoll–Paque centrifugation. This study was approved by the local ethics committees of Nagasaki University and Kinki University, and written informed consent was obtained from each patient.

RT-PCR.  RT-PCR was performed as described previously.(17) The following primers were used: 5′-gctggtgaccatgtctgtgc-3′ (forward) and 5′-tccttgtacgcgtccctctc-3′ (reverse) for c-maf; 5′-atcggctcagctctacagttcct-3′ (forward) and 5′-attcgcttgtagggaacattggt-3′ (reverse) for tax; and 5′-aggacagagcatggctcgcctacaga-3′ (forward) and 5′-taatggcagggaggtagggctcctga-3′ (reverse) for CCL22.

Western blot.  Whole cell lysates were prepared using RIPA buffer (Thermo Fisher Scientific, Rockford, IL, USA). Nuclear extracts were prepared and used for detecting c-Maf. Goat anti-c-Maf (N-15) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse monoclonal anti-Tax antibody (Ly-4, IgG3, originally established by Y. Tanaka, University of the Ryukyus)(18) and HRP-conjugated anti-goat or anti-mouse IgG were used. Bands were visualized using an ECL plus detection kit (GE Healthcare Bio-sciences, Piscataway, NJ, USA).

Small interference RNA (siRNA) and nucleofection.  Synthetic c-maf siRNA (ON-TARGET plus Set of 4) was purchased from Thermo Scientific Dharmacon (Lafayette, CO, USA). The GFP-22 siRNA (Qiagen, Valencia, CA, USA) was used as a negative control. MT-4 cells (1 × 106) were suspended in 100 μL of Nucleofector Solution (Cell Line Nucleofector Kit T: VCA-1002; Amaxa, Gaithersburg, MD, USA). Four micrograms of siRNA was added and mixed well. Cell suspensions were transferred to an electroporation cuvette and placed in the Nucleofector II device (Amaxa). Nucleofection of the cells was accomplished using the O-17 program. Immediately after nucleofection, 500 μL of prewarmed medium (RPMI 1640, 10% FCS, 2 mM glutamine and 50 μM 2-ME) was added to the cuvette, and cells were transferred to 12-well plates containing 1.5 mL of prewarmed medium. Cells were incubated for 48 h in a 37°C incubator containing 5% CO2 and subjected to real-time PCR analysis.

Promoter-luciferase assay.  Promoter-luciferase assay was performed as described previously.(17) pGL3-HTLV-1-LTR, pGL4.23[luc2/minP]-3 × 21-bp repeats, pGL4.23[luc2/minP]-2 × CRE, pGL3-CCL22(-722/-11), pGL3-2 × κB and pGL4.23[luc2/minP]-MARE were the reporter vectors used. The expression vectors for Tax, c-Maf, mouse CBP, TaxM22, Tax703 and TaxK88A were pHβPr.1-TaxMT-2, pCEFL-c-Maf, pcDNA3-mCBP, pHβPr.1-TaxM22, pHβPr.1-Tax703 and pHβPr.1-TaxK88A, respectively. Luciferase activity was determined using the Dual-Luciferase Reporter Assay System (Promega K.K., Tokyo, Japan). Transfection efficiency was normalized by using Renilla luciferase.

Real-time PCR.  CD4+, CD8+ and naïve CD45RO CD4+ T cells were negatively selected from PBMC by using the IMag Cell Separation System (BD Biosciences, San Jose, CA, USA). Polarized Th2 cells were obtained from naïve CD4+ T cells as previously described.(19) CD4+ and CD8+ T cells were treated with PHA plus IL-2 (20 U/mL) for 0, 3 or 10 days. Quantitative PCR was performed as previously described.(20) The primers for c-maf were 5′-aagaggcggaccctgaaaa-3′ (forward) and 5′-tccgactccaggacgtgtct-3′ (reverse). The probe for c-maf was 5′-cggctatgcccagtcctgccg-3′ (forward). The primers and probe for GAPDH were purchased from Applied Biosystems (Carlsbad, CA, USA).

Confocal microscopy.  Immunofluorescence staining was performed as previously described.(17) CCR4 was stained using anti-CCR4-APC (R&D Systems, Minneapolis, MN, USA). CCR4+ and CCR4 fractions were isolated from CD4+ T cells using FACSVantage (BD Biosciences). Uncultured CCR4+ CD4+, CCR4 CD4+ and CD8+ T cells and those cultured for 1 day with PHA were used. c-Maf was stained with rabbit anti-c-Maf (Santa Cruz Biotechnology), followed by goat anti-rabbit IgG F(ab′)2-biotin (Southern Biotechnology, Birmingham, AL, USA) and streptavidin-Alexa Fluor 546 (Invitrogen, Carlsbad, CA, USA). The PBMC treated with PHA plus IL-2 (20 U/mL) for 3 days were double stained with anti-CCR4-FITC (R&D Systems) and anti-c-Maf. They were then counterstained with TO-PRO-3 (Invitrogen) and observed under a confocal microscope (Carl Zeiss LSM 510 META, Jena, Germany).

Effect of c-Maf on viral gene expression from a HTLV-1 molecular clone.  Tax-dependent expression of the HTLV-1 envelope protein gp46 gene from a HTLV-1 molecular clone pFL-MT2(21) was studied. Jurkat cells were transfected with pFL-MT2 with or without a Tax expression vector (pHβPr.1-TaxMT-2) and/or c-Maf expression vector (pCEFL-c-Maf) using the DMRIE-C reagent (Invitrogen). After 24 h, the cells were stained for gp46 with mouse monoclonal anti-HTLV-1 gp46 antibody (clone 67/5.5.13.1; Abcam, Cambridge, MA, USA) or control mouse IgG1 followed by goat anti-mouse IgG-PE. Stained cells were immediately analyzed with FACSCalibur (BD Biosciences).

Effect of c-Maf on tax expression after infection with HTLV-1 virions.  8C feline kidney cells and HTLV-1-producing 8C/HTLV-1 cells (clone c77) were previously described.(22) The HTLV-1-negative 8C cells were transfected with c-Maf-eGFP or control eGFP expression vector, and eGFP-positive cells were sorted on FACSVantage 48 h post-transfection. Cell-free HTLV-1 infection was carried out as previously described.(22) Quantitative PCR for tax was performed by using the following primers and probe: the primers were 5′-gccgatcccaaagaaaaagac-3′ (forward) and 5′-cgaaaagaagactctgtccaaacc-3′ (reverse); the probe was 5′-tccaacaccatggcccacttccc-3′ (forward).

Statistical analyses.  Data are represented as mean ± SD. The Student’s t-test was used to determine the level of significance. A P value < 0.05 was considered significant.

Results

Inverse relationship between c-maf and tax expression in HTLV-1-infected T cells.  RT-PCR showed weak but consistent expression of c-maf mRNA in fresh ATL cells (Fig. 1A), corroborating the frequent expression of c-maf by ATL cells.(16) This may suggest that ATL is mostly derived from T-cell subsets expressing c-maf. The ATL cells from some patients also expressed tax, possibly due to a long lag time with transportation (>1 day) between blood drawing and PBMC preparation (Fig. 1A). We also examined c-maf expression in HTLV-1-transformed T cell lines. As shown in Figure 1(B), only some transformants strongly expressed c-maf. However, in such cases we noted that tax expression appeared to be inversely correlated with c-maf expression at both the mRNA and protein levels (Fig. 1B,C). Similarly, the expression of CCL22, one of the Tax target genes,(17) appeared to be inversely correlated with c-maf expression (Fig. 1B). We therefore examined the effect of c-maf expression on tax expression by using c-maf siRNA. Knockdown of c-maf expression in MT-4 cells significantly upregulated tax expression (= −0.68, = 0.0439, = 9), supporting the inverse correlation between c-maf and tax expression (Fig. 1D).

Figure 1.

 (A) RT-PCR analysis of fresh primary adult T-cell leukemia (ATL) cells. The PBMC obtained from acute ATL patients were examined. GAPDH was used as an internal control. (B) RT-PCR analysis of human T-cell leukemia virus type 1 (HTLV-1)-transformed T-cell lines. Jurkat is a T-ALL cell line. All other cell lines are HTLV-1-transformed T-cell lines. GAPDH was used as an internal control. (C) Western blot analysis. Nuclear extracts (20 μg) were used to detect c-Maf. Whole cell extracts (20 μg) were used for Tax detection. As an internal control, β-actin was used. (D) Effect of c-maf knockdown on tax expression. MT-4 cells were transiently transfected with c-maf-specific siRNA, cultured for 48 h and subjected to real-time PCR analyses. Data shown were relative expressions against GAPDH.

c-Maf inhibits Tax-dependent promoter activation.  Luciferase reporter assay showed that c-Maf significantly inhibited Tax transactivation of the HTLV-1 long terminal repeat (LTR; Fig. 2A). Furthermore, since transcription by Tax mutant M22 (defective for NF-κB activation) was inhibited significantly, it was suggested that CREB-dependent LTR activation was inhibited. On the other hand, Tax mutant 703 (defective for CREB activation) was unable to active LTR as expected. The HTLV-1 LTR contains three repeats of 21-bp harboring a CRE-like site, which constitutes the major Tax-responsive element.(9,23) We confirmed that c-Maf inhibited Tax-transactivation of the 3 × 21-bp as well as the consensus 2 × CRE-Luc (data not shown). c-Maf also inhibited Tax transactivation of the CCL22 gene promoter (Fig. 2B), one of the Tax target genes.(17) Previously, we have demonstrated that the CCL22 promoter is regulated partly by NF-κB but not CREB.(24) Therefore, we next tested the effect of c-Maf on a NF-κB p65-induced promoter activation. We found that c-Maf also significantly inhibited the NF-κB promoter activation by NF-κB p65 (Fig. 2C). So far, we consider that Tax-dependent NF-κB activation is mainly regulated by IκB degradation in the cytoplasm.(1) However, it has also been reported that phosphorylated NF-κB p65 can interact with CBP, and CBP-mediated acetylation of NF-κB p65 also enhances promoter activation.(25) Therefore, c-Maf might also inhibit Tax-dependent CCL22 promoter activation through CBP. Collectively, these results have demonstrated that c-Maf is capable of inhibiting the transcriptional activity of Tax.

Figure 2.

 c-Maf inhibits Tax-dependent promoter activation. Jurkat cells were co-transfected as indicated. The promoter luciferase vectors are: (A,D) pLTR-GL3 (LTR-Luc, 3 μg or 1 μg); (B) pGL3-CCL22(-722/-11) (CCL22-Luc, 3 μg); (C) pGL3-2 × κB (2 × κB-Luc, 3 μg); and (E) pGL4.23[luc2/minP]-MARE (MARE-Luc, 1 μg). The expression vectors pHβPr.1-TaxMT-2, pHβPr.1-TaxM22, pHβPr.1-Tax703, pCEFL-c-Maf, pcDNA3.1-mCBP and pHβPr.1-TaxK88A were used as indicated. Promoter activity was defined as the ratio of firefly luciferase to Renilla luciferase. All studies were performed at least three times with three independent transfections each time. *< 0.05 (Student’s t-test).

c-Maf competes with Tax for the transcriptional co-activator CBP.  CBP is essential for the Tax transactivation via CRE.(26) Likewise, c-Maf requires CBP for its transcriptional activity.(27) Simultaneous interactions of multiple transcription factors with CBP may lead to transcriptional synergy(28) or transcriptional repression, as reported in the case of c-Myb-dependent transcriptional activation.(29,30) c-Maf binds to a region of CBP that contains the first zinc finger and CREB-binding domains, where Tax also binds.(26,27) To test a possible competition mechanism, we examined whether CBP could rescue c-Maf-mediated inhibition of Tax-dependent LTR activation. CBP did indeed significantly reverse c-Maf-mediated suppression (Fig. 2D). Conversely, c-Maf-dependent activation via the Maf recognition element (MARE) was significantly inhibited by Tax but not by TaxK88A,(31) a mutant Tax not binding to CBP (Fig. 2E). These findings support that c-Maf competes with Tax for CBP.

c-maf is expressed in Th2 and PHA-activated CD4+ T cells but not PHA-activated CD8+ T cells.  Next, we examined c-maf expression in primary CD4+ and CD8+ T cells by real-time PCR (Fig. 3A). As reported previously,(14) Th2-polarized cells strongly expressed c-maf. Furthermore, activation with PHA plus IL-2 strongly upregulated c-maf expression in CD4+ T cells with a peak on day 3, but only weakly in CD8+ T cells. We also confirmed c-maf expression at the protein level (Fig. 3B). We next performed c-Maf staining in fractionated T cells. As shown in Figure 3(C), activated CCR4+ CD4+ but not CCR4 CD4+ T cells were strongly positive for nuclear c-Maf staining. Although we could not accurately count c-Maf+ cells because of the formation of tight cell clusters, they were up to 20% of PHA-treated CCR4+ CD4+ T cells. Activated CD8+ T cells also showed a marginal c-Maf staining, but mostly in the cytoplasm. We also carried out double staining for c-Maf and CCR4 in PHA-stimulated PBMC. As shown in Figure 3(D), approximately 70–80% of the CCR4+ cells co-expressed c-Maf in the nucleus, whereas CCR4 cells were hardly stained for c-Maf. Thus, a fraction of CCR4+ CD4+ T cells strongly express c-Maf upon activation.

Figure 3.

 Selective expression of c-maf in activated CCR4+ CD4+ T cells. (A) Real-time PCR analysis. The indicated cell populations were examined for c-maf expression. (B) Western blot analysis. Nuclear extracts (20 μg) and whole cell extracts (20 μg) of CD4+ and CD8+ cells were prepared at the indicated days. c-Maf and β-actin were detected. (C) Immunological staining of c-Maf. CCR4+ CD4+, CCR4 CD4+ and CD8+ T cells were sorted and stained for c-Maf immediately (day 0) or 1 day after PHA treatment. c-Maf was visualized with Alexa Fluor 546 (red), while the nuclei were counterstained with TO-PRO-3 (blue). Transmitted light images are shown on the right. (D) Double staining for c-Maf and CCR4. The PBMC were stained without (day 0) or with PHA-stimulation for 3 days (day 3). c-Maf and CCR4 were visualized with Alexa Fluor 546 (red) and FITC (green), respectively. Nuclei were counterstained with TO-PRO-3 (blue). Transmitted light images are also shown on the right.

c-Maf suppresses the HTLV-1 envelope gp46 expression from an infectious HTLV-1 molecular clone.  We wanted to compare HTLV-1 gene expression between CCR4+ and CCR4 T cells freshly infected with HTLV-1. However, the efficiency of HTLV-1 transmission to normal T cells by the standard MT-2 co-culture system was very low in our experiments. We therefore used a HTLV-1 molecular clone pFL-MT2 that contains the entire HTLV-1 genome derived from MT-2 cells and exhibits viral gene expression upon Tax expression.(21) We transfected Jurkat cells with pFL-MT2 and monitored HTLV-1 gene expression by examining cell-surface expression of the envelope protein gp46 using flow cytometry (Fig. 4). When Jurkat cells were co-transfected with the Tax-expression vector, a small but consistent fraction of cells became positive for gp46 after 24 h (Fig. 4C; 4.0 ± 0.2%, n = 4). c-Maf, while not directly affecting the appearance of gp46+ cells (Fig. 4D), significantly decreased the appearance of gp46+ cells induced by Tax (Fig. 4E; 2.4 ± 0.2%, < 0.05, = 4). Thus, c-Maf suppressed Tax-induced gene expression from the HTLV-1 molecular clone.

Figure 4.

 Effect of c-Maf on Tax-induced gene expression from a human T-cell leukemia virus type 1 (HTLV-1) molecular clone. Jurkat cells were transfected with pFL-MT2 with or without a Tax-expression vector (pHβPr.1-TaxMT-2) and/or a c-Maf-expression vector (pCEFL-c-Maf). After 24 h, the viral expression level was monitored by examining cell-surface expression of the envelope protein gp46 using flow cytometry. The representative results of four independent experiments are shown. (A) no vector; (B) pFL-MT2 + pHβPr.1 + pCEFL; (C) pFL-MT2 + pHβPr.1-TaxMT-2 + pCEFL; (D) pFL-MT2 + pHβPr.1 + pCEFL-c-Maf; (E) pFL-MT2 + pHβPr.1-TaxMT-2 + pCEFL-c-Maf.

c-Maf suppresses the tax gene expression in cell-free HTLV-1 infection.  To corroborate the suppressive effect of c-Maf on the Tax-dependent gene expression from the HTLV-1 genome, we also tested the effect of c-Maf on tax expression using a cell-free HTLV-1 infection system, where cell-free HTLV-1 virions produced by HTLV-1-infected 8C (8C/HTLV-1) cells were used to infect HTLV-1-negative 8C cells.(22) The 8C cells transiently expressing c-maf with eGFP (8C-c-Maf-eGFP, purity >97%) or eGFP alone (8C-eGFP, purity >95%) were cultured with the supernatant of 8C/HTLV-1 cells for 24 h and analyzed for tax expression. Real-time PCR revealed that tax expression was significantly reduced in c-maf-expressing 8C cells compared with control 8C cells (Fig. 5). Thus, c-Maf was capable of suppressing tax expression from freshly infecting HTLV-1 virions.

Figure 5.

 Effect of c-Maf on Tax expression by infecting human T-cell leukemia virus type 1 (HTLV-1) virions. HTLV-1-negative 8C cells were transfected with the expression vector for c-Maf with eGFP or eGFP alone, and eGFP-positive cells were sorted on FACSVantage after 24 h. The sorted eGFP-positive cells were infected with cell-free HTLV-1 virions derived from HTLV-1-infected 8C (8C/HTLV-1, c77). After 24 h, tax and GAPDH expressions were analyzed by real-time PCR. The representative results of three independent experiments are shown.

Discussion

Although HTLV-1 encodes a potent transcriptional activator Tax that strongly promotes proliferation of HTLV-1-infected T cells, ATL develops after a long period of latency in approximately 5% of carriers.(1) Furthermore, ATL predominantly develops from CD4+ T cells expressing CCR4.(4) Previously, we have shown that Tax strongly induces CCL22, which encodes a CCR4 ligand, in HTLV-1-infected T cells.(17) We have further shown that the abundant production of CCL22 by HTLV-1+ T cells, together with the Tax-induced upregulation of ICAM-1 in HTLV-1+ T cells,(32) leads to preferential cell-mediated transmission of HTLV-1 to CCR4+ CD4+ T cells.(17) CCR4 is known to be selectively expressed by Th2 cells, skin-homing memory/effector T cells and Treg.(5–7) c-Maf is a transcription factor known to be expressed by Th2 cells and presumably by some Treg.(7,14,15) Here, we have shown that c-Maf suppresses the transcriptional activity of Tax by competing for CBP (Fig. 2). We have further shown that c-Maf is selectively induced in CD4+ T cells, especially in the CCR4+ fraction (Fig. 3), and 70–80% of CCR4+ CD4+ T cells express c-Maf in the nucleus. We also confirmed that c-Maf suppressed Tax-dependent HTLV-1 gene expression from a HTLV-1 molecular clone and from freshly infecting HTLV-1 virions (Figs 4,5).

Although Tax is essential for the growth promotion of HTLV-1-infected T cells, it is also a preferential target for host CTL.(10,11) Thus, regulated expression of Tax would be important for the survival of HTLV-1-infected T cells in vivo. The HTLV-1 accessory proteins Rex, HBZ and p30 are known to suppress tax expression.(1) Furthermore, several cellular factors such as CCAAT/enhancer binding protein β (C/EBPβ) and inducible cAMP early repressor (ICER) suppress the transcriptional activity of Tax.(33,34) Here we have shown that c-Maf is yet another cellular factor that is capable of suppressing the transcriptional activity of Tax by competing for CBP. Notably, c-Maf is selectively induced in CCR4+ CD4+ T cells, the preferential targets of HTLV-1 infection(17) and apparently the eventual cellular origin of ATL.(4,7) Thus, c-Maf may promote latent infection of HTLV-1 in CCR4+ CD4+ T cells by negatively regulating the expression of tax from the proviral HTLV-1, which together with multiple genetic alterations accumulated during the long latency period may lead to the development of ATL.

Acknowledgments

We thank Drs Masahiro Fujii, Jun-ichi Fujisawa, Xiaojing Ma, Shunsuke Ishii, Takeo Ohsugi and Chou-Zen Giam for providing the expression vectors for Tax, LTR-Luc, c-Maf, CBP, HTLV-1 and the TaxK88A, respectively. This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports and Technology, Japan, and the High-Tech Research Center Project for Private Universities: Matching Fund Subsidy from the Ministry of Education, Culture, Sports, Science and Technology of Japan, 2002–2009.

Disclosure Statement

The authors have no conflict of interest.

Ancillary