HTLV-1 Tax deregulates signal transduction pathways, transcription of genes, and cell cycle regulation of host cells, which is mainly mediated by its protein–protein interactions with host cellular factors. We previously reported an interaction of Tax with a histone methyltransferase (HMTase), SUV39H1. As the interaction was mediated by the SUV39H1 SET domain that is shared among HMTases, we examined the possibility of Tax interaction with another HMTase, SMYD3, which methylates histone H3 lysine 4 and activates transcription of genes, and studied the functional effects. Expression of endogenous SMYD3 in T cell lines and primary T cells was confirmed by immunoblotting analysis. Co-immuno-precipitaion assays and in vitro pull-down assay indicated interaction between Tax and SMYD3. The interaction was largely dependent on the C-terminal 180 amino acids of SMYD3, whereas the interacting domain of Tax was not clearly defined, although the N-terminal 108 amino acids were dispensable for the interaction. In the cotransfected cells, colocalization of Tax and SMYD3 was indicated in the cytoplasm or nuclei. Studies using mutants of Tax and SMYD3 suggested that SMYD3 dominates the subcellular localization of Tax. Reporter gene assays showed that nuclear factor-κB activation promoted by cytoplasmic Tax was enhanced by the presence of SMYD3, and attenuated by shRNA-mediated knockdown of SMYD3, suggesting an increased level of Tax localization in the cytoplasm by SMYD3. Our study revealed for the first time Tax–SMYD3 direct interaction, as well as apparent tethering of Tax by SMYD3, influencing the subcellular localization of Tax. Results suggested that SMYD3-mediated nucleocytoplasmic shuttling of Tax provides one base for the pleiotropic effects of Tax, which are mediated by the interaction of cellular proteins localized in the cytoplasm or nucleus. (Cancer Sci 2011; 102: 260–266)
Human T-cell leukemia virus type 1 (HTLV-1) is the causative agent of an aggressive leukemia known as adult T-cell leukemia (ATL).(1,2) The viral protein Tax plays a central role in the development of ATL in HTLV-1-infected carriers. Tax is known to enhance gene expression of HTLV-1 through target sequences in the U3 region of 5′LTR by interacting with cellular factors such as cAMP response element binding protein (CREB) and transcriptional co-activator CBP/p300.(3,4)
Tax also activates intracellular signal transactivation pathways that normally play a crucial role in cellular responses to various extracellular stimuli, resulting in the activation of several transcription factors, including nuclear factor (NF)-κB, serum response factor, and CREB. Consistent with both its cytoplasmic and nuclear activities, Tax was shown to be distributed in both compartments in HTLV-1-infected and Tax-transfected cells. Initial studies reported that Tax is localized predominantly in the nucleus and specifically accumulated in the nuclear speckled structures.(5,6) However, depending on the cell type, significant amounts of Tax have also been found in the cytoplasm.(7–9) The mechanism by which Tax was localized in cytoplasm was reported recently,(10–16) although the detailed mechanisms regulating Tax subcellular localization remain to be elucidated.
Epigenetic control of gene expression is mediated by chemical modifications of histone tails, such as acetylation, phosphorylation, and methylation, and DNA methylation such as CpG methylation, both leading to the regulation of chromatin structure and function.(17–19) Previous studies reported that Tax interacts with various histone modifying enzymes.(20–27) We previously reported that Tax directly interacts with SUV39H1.(27) The interaction of Tax with SUV39H1 was mediated by the SET domain of SUV39H1. The SET domain is shared among histone methyltransferases and possesses methyltransferase activity, and functions as lysine methylases that add multiple methyl groups to specific lysines in histones H3 and H4. As the SUV39H1 SET domain mediated the interaction with Tax, it was suggested that Tax is able to interact with a variety of histone methyltransferases.
SET and MYND domain-containing protein 3 (SMYD3) is unique in that it promotes di- and tri-methylation in H3-K4 and is frequently overexpressed in human colorectal, liver, and breast cancers where its enhanced expression is essential for the growth of cancer cells.(28,29) SMYD3 encodes a 428-amino acid protein containing a SET domain, a zf-MYND domain, and a SET-N region.
Thus, we speculated that Tax may interact with SMYD3 and the interaction may play roles in proliferation and immortalization of Tax-expressing T cells. In the present study, we examined direct interaction between Tax and SMYD3 and the subcellular localization of the Tax–SMYD3 complex.
Materials and Methods
Cell cultures. Jurkat, CEM, TIG1, Molt-4, HUT102, and MT-2 cells were cultured in RPMI-1640 supplemented with 10% FCS and antibiotics. HeLa, HEK293, and HEK293T cells were cultured in DMEM supplemented with 10% FCS and antibiotics. Peripheral blood mononuclear cells from healthy volunteers were prepared by centrifugation of peripheral blood with Ficoll-Paque (GE Healthcare UK Ltd. Buckinghamshire, UK) and used as control PBMC. Activated control T cells were prepared by stimulation of control PBMC with phytohemagglutinin (PHA) (10 μg/mL) for three days and cultured in RPMI-1640 containing 10% FCS, 1000 U/mL ampicillin, and 1 mg/mL streptomycin.
Antibodies. Lt-4 is a mAb that reacts with Tax.(30) Antibodies against SMYD3 (ab16027) were purchased from Abcam (Cambridge, UK). Anti-Flag antibody M2 was purchased from Sigma-Aldrich (St. Louis, MO, USA). Antibodies against IκBα (C-21) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies against HA (6E2) and Phospho-IκBα (Ser32/36) were purchased from Cell Signaling Technology (Beverly, MA, USA).
Plasmids. The wild-type Tax expression plasmid, pCG-Tax, pET3d/Tax, and a mutant Tax plasmid, TaxC29A, were kind gifts from Prof. Fujisawa (Kansai Medical University, Moriguchi, Japan) and Dr. Tsuji (National Institute of Infectious Diseases, Tokyo, Japan), respectively. The SMYD3 expression plasmid, p3xFLAG-CMV-SMYD3, was described previously.(28) Wild-type and mutant SMYD3 expression vectors were prepared by PCR using primers described below, and PCR products were cloned into pcDNA-HA or pGEX5X-1 (GE Healthcare UK Ltd. Buckinghamshire, UK). Forward primers: 5′-CCCGAATTCATGGAGCCGCTGAAGGTGGAAAAG-3′, 5′-CCCGAATTCTATCCTCCAGACTCCGTTCG-3′, 5′-CCCG-AATTCGAGCGCCGGAAGCAGCTGAGG-3′, 5′-CCCGAAT-TCATTAACAAACTGACTGAAGATAAG-3′; reverse primers: 5′-CCCCTCGAGTCAGGATGCTCTGATGTTGGCGTC-3′, 5′-CCCCTCGAGTCACTCACTGGTCATCAGCATATC-3′, 5′-CCCCTCGAGTCATCTGGGTTTGCAGCTTTTAAG-3′.
Using pET3d/Tax plasmid, histidine-tagged wild-type Tax was bacterially expressed and purified as described previously.(27)
Quantitative RT-PCR analysis. Quantitative RT-PCR was carried out to analyze SMYD3 mRNA levels. Total RNA was isolated from the cells by Isogen (Wako Pure Chemical Industries, Osaka, Japan) and cDNA was synthesized using SuperScript2 (Invitrogen, Carlsbad, CA, USA), followed by real-time PCR using SMYD3 primers (5′-AGGGGTTCAAGTGATGAAAG-TTG-3′, 5′-GCTGTGTTCTCTGCCATGTGT-3′). Levels of β-actin mRNA were measured as an internal control.
In vitro transcription and translation. For in vitro translation of the wild-type and mutant Tax proteins, the cDNA was amplified by PCR and cloned into pBluescript II SK (−). In vitro transcription and translation of the indicated cDNA was done using TNT Quick Coupled Transcription/Translation Systems (Promega, Madison, WI, USA) as described previously.(27)
GST pull-down assay. Wild-type and mutant GST-SMYD3 proteins (1 μg) bound to glutathione–Sepharose 4B (GE Healthcare and Biosciences) were mixed with in vitro translated wild-type and various mutants of Tax proteins. Binding reactions, SDS-PAGE, and visualization were carried out as described previously.(27) Relative intensities of the bands were determined using the NIH Image software (National Institutes of Health, Bethesda, MD, USA).
Co-immunoprecipitation and immunoblotting. For co-immunoprecipitation analyses, transfection was carried out by the standard calcium phosphate precipitation method. Cell lysates were prepared in TNE buffer (10 mM Tris–HCl [pH 7.8], 1% Nonidet P-40, 150 mM NaCl, 1 mM EDTA). When indicated, aliquots were removed for immunoblots of whole cell lysates. Immunoblots were carried out to detect co-immunoprecipitated or GST pull-down proteins, as described previously.(27) Primary antibodies used in this assay are described above. Alkaline phosphatase-conjugated anti-mouse immunoglobulin sheep and anti-rabbit donkey antibodies (both from Promega) were used as secondary antibodies.
Immunocytochemistry. HeLa (3 × 105/mL) were grown on cover slips in a 6-well plate for 1 day, and transfected with various expression plasmids using Lipofectamine 2000 (Invitrogen). Fixation and staining were done as described previously.(27) Fluorescence signals were detected using confocal microscopy (Radiance 2000; Bio-Rad, Hercules, CA, USA). HUT102 and MT-2 cells (1 × 106/mL) were cultured with RPMI-1640 containing 10% FCS. Fixation and staining were done as described previously.(27) Data were obtained using an immunofluorescence microscope (BX50; Olympus, Tokyo, Japan).
Reporter gene assays. HEK293 cells (1.6 × 105/mL) were transfected with reporter plasmids and expression plasmids using Lipofectamine 2000. After 24 h incubation, luciferase activity was measured with a luciferase assay kit (Promega). A control plasmid, RSV-Renilla, was prepared by inserting the RSV LTR to the Renilla reporter plasmid (Promega). The measured activities were standardized by the activities of Renilla luciferase, and transactivation was expressed as fold activation compared with the basal activity of p6κB-Luc without effectors. Another reporter plasmid, LTR-Luc, was described previously.(27)
shRNA-mediated SMYD3 knockdown. SMYD3 knockdown was established according to the manufacturer’s protocol (Retrovirus Packaging kit Ampho; Takara Bio, Shiga, Japan) using the sense and antisense DNA oligomers as already described.(28,31) shRNA-mediated knockdown was done as described previously.(31)
SMYD3 expresses in T cells and interacts with HTLV-1 Tax. To determine whether Tax can interact with SMYD3, we first carried out co-immunoprecipitation assays by transient expression for these proteins. The results clearly showed that immunoprecipitates of anti-Flag antibody contained Tax protein (Fig. 1a). Conversely, when the cell lysates were immunoprecipitated with Lt-4, the immunocomplex was shown to contain Flag-tagged SMYD3 (Fig. 1b). Taken together, these results indicated that Tax interacts with SMYD3 in cultured cells. We also showed that Tax protein directly interacts with GST-SMYD3 in vitro (Fig. 1c).
SMYD3 expression was originally reported in solid tumors such as hepatoma and breast cancer.(28,29) Thus, to examine whether SMYD3 is expressed in other cells, we carried out quantitative RT-PCR analysis and quantified SMYD3 gene expression in primary T cells, human fibroblast, and other cell lines. As shown in Figure 1(d), the level of SMYD3 transcription was higher in T-cell lines with or without HTLV-1 infection than those in others. Next we examined whether endogenous SMYD3 and Tax can interact with each other in HTLV-1-infected cell lines. Immunoprecipitates with anti-SMYD3 or Lt-4 were shown to contain Tax or SMYD3, respectively (Fig. 1e, top panels). This result indicated that endogenous Tax and SMYD3 interact with each other in T cell lines.
Analyses of domains responsible for interaction between SMYD3 and Tax. To define the domains within SMYD3 and Tax that are responsible for the interaction, we next carried out in vitro binding assays. First, we constructed various GST-fusion mutants of SMYD3 according to the domain structure(28) (Fig. 2a) and examined binding to the in vitro translated and S35-labeled wild-type Tax. When C-terminal deleted series of SMYD3 were examined, the SMYD3-A showed a marked decrease in the intensity of the bound Tax. The SMYD3-B showed no binding activity to Tax. When N-terminal deletion series of SMYD3-mutants were tested, both the SMYD3-C and SMYD3-D showed binding activity to Tax. However, the SMYD3-E did not show any binding activity (Fig. 2b, upper panel). The relative levels of bound Tax compared with that of the wild-type Tax are shown in Figure 2(c). These results suggested that the region of amino acids from 248 to 428 of SMYD3 appears to be enough to show a significant level of affinity against Tax similar to that of the wild-type SMYD3.
We next analyzed the domains of Tax responsible for the interaction with SMYD3. In addition to the wild-type Tax, we used three kinds of mutants, TaxN180, TaxΔN108 and ΔCBP-B (Fig. 2d).(27) The results showed that the wild-type Tax and all of these mutants could bind to SMYD3 (Fig. 2e, upper panel). The relative levels of bound Tax compared with that of the wild-type Tax are shown in Figure 2(f). Generally the bands of pulled-down Tax mutants were significantly weaker in intensity than that of the wild-type Tax, with the weakest being that of TaxN180 (Fig. 2e,f). These results suggested that amino acids 109–353 of Tax are more important to bind with SMYD3 than N-terminal 108 amino acids.
Co-localization of Tax and SMYD3 in vivo. When SMYD3 alone was expressed in HeLa cells, SMYD3 localized in the cytoplasm or nucleus, whereas singly expressed Tax localized in the nuclei in most of the transfected cells (Fig. 3a). However, when Tax and SMYD3 were simultaneously expressed in HeLa cells, Tax was colocalized with SMYD3 in the cytoplasm in a part of the cotransfected cells (Fig. 3b, lower panel). To evaluate the effects of co-expression quantitatively, we counted the cell numbers that show Tax and SMYD3 localization in the nucleus and cytoplasm. The frequency of the cells where Tax is localized in cytoplasm increased with co-expression of SMYD3, when compared with those where Tax is expressed alone (Fig. 3c), suggested a tethering of Tax by SMYD3 to the cytoplasm in these cells. To confirm whether the same results were obtained in T cell lines, we analyzed the subcellular localization of endogeneous SMYD3 and Tax in MT-2 and HUT102 cells. The results showed that Tax colocalized with SMYD3 in most of the cells (Fig. 3d). Taken together, these results suggested that Tax was tethered to the cytoplasm by SMYD3.
To examine this possibility, we used two Tax mutants, neither of which enter the nuclei because of deletion of the N-terminal region containing the nuclear localization signal (TaxΔN108)(27) or a point mutation in the amino acids of the nuclear localization signal (Tax-C29A).(32) TaxΔN108 and Tax-C29A interacted with SMYD3 (Fig. 4a), and a clear cytoplasmic localization when expressed alone (Fig. 4b,c, top panels). However, TaxΔN108 and Tax-C29A were found in the nuclei in a part of the transfected cells, and in the other cells they were colocalized with SMYD3 in the cytoplasm when co-expressed (Fig. 4b,c, lower panels).
Next we studied effects of SMYD3 mutants on subcellular localization of Tax, using two mutants that are localized in the cytoplasm but cannot enter the nuclei. One of the mutants, SMYD3-B, does not interact with Tax and the other, SMYD3-D, can interact with Tax (Fig. 4d). When transfected alone, SMYD3-B and SMYD3-D are mostly localized in the cytoplasm (Fig. 4e, left upper and lower panels). When SMYD3-B was co-expressed with Tax, it was localized in the cytoplasm, whereas Tax was localized in the nucleus (Fig. 4e, right upper panels). However, Tax was localized in the cytoplasm when co-expressed with SMYD3-D (Fig. 4e, right lower panels).
SMYD3 enhances Tax-mediated NF-κB activation. To examine whether NF-κB activation by Tax is enhanced in the presence of SMYD3, we carried out reporter gene assays using the p6κB-Luciferase assay system to measure the activity of NF-κB. The results showed that expression of Tax induced 15-fold activation of NF-κB in HEK293 cells. The luciferase activity was enhanced up to 25-fold in these cells when co-expressed with SMYD3 (Fig. 5a, upper panel). In contrast, expression of SMYD3 alone showed no significant activation of p6κB-Luc (Fig. 5a, lower panel). To study the effects of co-expression of SMYD3 and Tax in another system, we next studied the phosphorylation of IκBα as an indicator of IKK complex activity. Immunoblot analysis of the cell lysates from Tax-transfected HEK293 cells showed increased levels of phosphorylation of IκBα Ser32/36 and the level appeared to be further enhanced when co-expressed with SMYD3 (Fig. 5b, upper panel). However, no change was observed in the levels of IκBα Ser32/36 phosphorylation when SMYD3 alone was expressed. Intensities of the detected bands are shown in the graphs (Fig. 5b, lower panel).
To examine the effect on LTR promoter activity, we carried out another luciferase assay using pLTR-Luc as a reporter. Tax-mediated transactivation of LTR-Luc was suppressed by co-expression of SMYD3 down to approximately half of the activity without SMYD3 (Fig. 5c, left panel). However, SMYD3 alone did not show any significant effects on LTR activity (Fig. 5c, right panel).
Next, we tested whether the suppression of SMYD3 expression may influence the Tax localization and NF-κB activation in MT-2 and HUT102 cell lines (Fig. 6). Quantitative RT-PCR and Western blot analysis showed a significant reduction of endogenous SMYD3 expression in SMYD3-shRNA expressing cells compared with cells infected with a mock retrovirus (Fig. 6a,b). Tax was less localized in cytoplasm when SMYD3-shRNA was expressed in MT2 and in HUT102 (Fig. 6c). Furthermore, 6κB-Luc activity was significantly suppressed in SMYD3-shRNA expressing MT-2 (Fig. 6d). These results suggested that SMYD3-mediated cytoplasmic localization of Tax results in efficient activation of the NF-κB signaling pathway.
Our results of in vitro analyses indicated that the major interacting domain of SMYD3 with Tax was the C-terminal region of approximately 180 amino acids that does not contain the SET domain of SMYD3 (Fig. 2). The results were different from our previous observation that SUV39H1 interacts with Tax by the SET domain,(27) suggesting another mechanism. Although we were not able to clarify the binding domain in Tax, determination of the amino acids of SMYD3 and Tax that are involved in the interaction between these proteins will provide additional information for understanding the mechanisms of Tax interaction with host cellular proteins. Although we have not examined the effect of Tax interaction on the histone methyltransferase activity of SMYD3, the complex formation of Tax and SMYD3 may lead to changes in SMYD3 function. Another possibility will be that complex formation of Tax–SMYD3 may result in a new target specificity of SMYD3, because Tax may recruit proteins that are not original targets of SMYD3. These possibilities remain to be studied in the future.
As for the intracellular distribution of Tax, initial studies indicated nuclear localization with little or no Tax localized to the cytoplasm, using laboratory cell lines including HeLa, COS, and 293T.(3,4,6,33–35) However, subsequent studies have reported the localization of Tax to the cytoplasm in a number of cell types including Tax-transfected and HTLV-1-infected cells of both cell line and primary cell origin.(7,8,36) Thus, the regulatory mechanisms of Tax subcellular localization have been a focus of research interest. A number of studies have reported nucleocytoplasmic shuttling of Tax and functional effects of subcellular localization of Tax, although the regulatory mechanisms of subcellular localization remain to be clarified.(37) The result of our study clearly showed that the interaction between Tax and SMYD3 regulates subcellular localization of Tax, depending on the shuttling of SMYD3 (Figs 3,4). Tax was localized in the cytoplasm and the nucleus (Fig. 3). The shuttling of Tax was observed in HTLV-1-infected T cell lines as well as in the cell lines transiently transfected with the SMYD3 expression vector (Fig. 3). Accordingly, it is concluded that the subcellular localization of Tax can be changed in the presence of SMYD3, but the detail of the mechanism is not clear and remains to be studied.
A number of published studies have indicated that Tax functions both in the cytoplasm and nucleus. The results showed nucleocytoplasmic shuttling of Tax, at least in some of the Tax-expressing cells, which was associated with changes in functions of Tax. We observed enhanced activation of Tax-mediated NF-κB activation with SMYD3 co-expression (Fig. 5), which was eliminated by SMYD3 knockdown, probably due to the decreased amount of Tax in the cytoplasm as observed by immunocytochemistry (Fig. 6). These observations are consistent with the idea that many aspects of Tax function depend on its subcellular localization. Our study revealed for the first time the interaction between Tax and SMYD3 and the apparent tethering of Tax by SMYD3 between the nucleus and cytoplasm. The results suggest that SMYD3-mediated nucleo-cytoplasmic shuttling of Tax provide bases for pleiotropic effects of Tax, which are mediated by interaction of cellular proteins localized in the cytoplasm or nucleus. Continued studies are required to understand how Tax–SMYD3 interaction contributes to the unique phenotype of HTLV-1-infected T-cells and the pathogenesis of HTLV-1-related diseases.
We thank Professor Jun-ichi Fujisawa (Kansai Medical University, Moriguchi, Japan) for pCG-Tax and pET3d/Tax plasmids. This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, to TW (No. 20390267), and by Grants-in-Aid from the Ministry of Health, Labor, and Welfare for the 3rd-term Comprehensive 10-year Strategy for Cancer Control and for Cancer Research, to TW (H21-G-002).