|Home | About | Journals | Submit | Contact Us | Français|
The human T-cell lymphotropic virus type 1 (HTLV-1) infects and transforms CD4+ lymphocytes and causes adult T-cell leukemia/lymphoma (ATLL), an aggressive lymphoproliferative disease that is often fatal. Here, we demonstrate that the HTLV-1 pX splice-variant p30II markedly enhances the transforming potential of Myc and transcriptionally activates the human cyclin D2 promoter, dependent upon its conserved Myc-responsive E-box enhancer elements, which are associated with increased S-phase entry and multinucleation. Enhancement of c-Myc transforming activity by HTLV-1 p30II is dependent upon the transcriptional coactivators, transforming transcriptional activator protein/p434 and TIP60, and it requires TIP60 histone acetyltransferase (HAT) activity and correlates with the stabilization of HTLV-1 p30II/Myc-TIP60 chromatin-remodeling complexes. The p30II oncoprotein colocalizes and coimmunoprecipitates with Myc-TIP60 complexes in cultured HTLV-1-infected ATLL patient lymphocytes. Amino acid residues 99 to 154 within HTLV-1 p30II interact with the TIP60 HAT, and p30II transcriptionally activates numerous cellular genes in a TIP60-dependent or TIP60-independent manner, as determined by microarray gene expression analyses. Importantly, these results suggest that p30II functions as a novel retroviral modulator of Myc-TIP60-transforming interactions that may contribute to adult T-cell leukemogenesis.
The human T-cell lymphotropic virus type-1 (HTLV-1) infects CD4+ T cells and causes adult T-cell leukemia/lymphoma (ATLL), an aggressive lymphoproliferative disease that is often fatal (59, 61, 65, 83). HTLV-1-infected leukemic lymphocytes exhibit deregulated cell cycle progression and characteristic multinucleation or polyploidy (evidenced by the appearance of flower-shaped or lobulated nuclei). A conserved sequence, known as pX, located within the 3′ terminus of the HTLV-1 genome, encodes at least five nonstructural regulatory factors, including the viral transactivator Tax and an alternative splice-variant, p30II (or Tax open reading frame II [ORF II], Tof), which was shown to possess a functional transactivation domain (6, 13, 15, 29, 34, 35, 66, 86, 87). The pX sequence is generally retained in the majority of ATLL patient isolates, even those containing partially deleted proviruses (33, 68), indicative of its importance for pathogenesis.
The viral Tax protein transcriptionally activates numerous lymphoproliferative pathways (NF-κB, CREB/ATF, and p67SRF) (29, 72, 73, 74, 75, 80, 84, 88) and has been shown to inhibit transcription functions associated with the tumor suppressor p53, which likely contributes to a loss of G1/S-phase checkpoint control in HTLV-1-infected T cells (8, 46, 58). Many of the pleiotropic effects of Tax upon cellular signaling may derive from its aberrant recruitment of the transcriptional coactivators, p300/CREB-binding protein (p300/CBP) and p300/CBP-associated factor (P/CAF) (9, 22, 23, 27, 36, 37, 49, 50, 77, 78). Further, Tax interacts with cell cycle modulators, including D-type cylin-cdk4/6 complexes, retinoblastoma (Rb) protein, and the human mitotic arrest deficiency type 1 (hMAD-1) protein (21, 28, 31, 32, 39, 47, 52, 76). Although HTLV-1 Tax expression markedly promotes G1/S transition (38, 40, 64), Tax has been demonstrated to inhibit Myc-dependent transactivation and prevent Myc-associated anchorage-independent cell growth (67). As ATLL patient-derived lymphocytes and tumors from HTLV-1 pX transgenic mice are known to possess deregulated Myc functions, these findings collectively suggest that other pX-encoded factors may influence Myc to promote cellular transformation by HTLV-1 (20, 43, 63).
The Myc transcription factor promotes S-phase cell cycle entry, induces apoptosis or programmed cell death, and causes neoplastic cellular transformation (2, 3, 7, 12, 19, 41, 51). The expression of the Myc protooncogene is deregulated in many solid tumors and hematological malignancies, including ATLL, diffuse large-cell lymphomas, CD30+ anaplastic large-cell lymphomas, and Burkitt's B-cell lymphomas (18, 24, 26, 43, 55, 60). The transforming viruses, HTLV-1 and Epstein Barr virus, deregulate Myc functions associated with development of ATLL and Burkitt's lymphomas, respectively (11, 18, 26, 43, 63, 67). Our preliminary studies indicated that the HTLV-1 accessory protein p30II markedly increases S-phase cell cycle progression and induces significant polyploidy. As relatively little is known with respect to the roles of pX-encoded factors (e.g., p30II, p13II, p12I, and Rexp27) in HTLV-1-associated pathogenesis (6, 29, 34, 35), we sought to characterize the molecular mechanism by which p30II promotes Myc-dependent S-phase progression and multinucleation. While others have proposed that p30II's transcriptional functions are targeted against the viral LTR to repress HTLV-1 gene expression (1, 86, 87), the physiological role of p30II in ATLL-development remains unclear. Using microarray analyses, we now demonstrate that numerous cellular genes are transcriptionally activated by HTLV-1 p30II in a 60-kDa Tat-interacting protein (TIP60)-dependent or TIP60-independent manner. Nicot et al. (48) and Younis et al. (85) have shown that p30II binds and inhibits nuclear export of the doubly spliced Tax/Rex HTLV-1 mRNA, and it is intriguing that p30II might perform diverse functions to regulate viral gene expression and promote altered cellular growth, as has been noted for Tax, which drives LTR transactivation and deregulates host lymphoproliferative-signaling pathways (13, 21, 28, 29, 38, 40, 47, 52, 64, 72-76, 84). Robek et al. (62) have previously demonstrated that p30II is dispensable for immortalization and transformation of human peripheral blood mononuclear cells by an infectious HTLV-1 molecular clone, ACH.p30II, which is defective for p30II production; however, the ACH.p30II mutant exhibited an approximately 20 to 50% reduction in transformation efficiency compared to the wild-type ACH.wt (62), suggesting that p30II is required for the full transforming potential of HTLV-1. Importantly, our findings indicate that HTLV-1 p30II is a novel retroviral modulator of Myc transcriptional and of transforming activities that may significantly contribute to adult T-cell leukemogenesis through stabilization of Myc-TIP60 transcriptional interactions.
HeLa cells (ATCC CCL-2) were grown in Dulbecco's modified Eagle's medium (DMEM; ATCC) supplemented with 10% fetal bovine serum (FBS; Atlanta Biologicals), 100 U/ml penicillin, and 100 μg/ml streptomycin sulfate (Invitrogen-Life Technologies) and cultured at 37°C and 5% CO2. 293A fibroblasts (Quantum Biotechnology) were cultured in ATCC 46-X medium supplemented with sodium bicarbonate (Invitrogen-Life Technologies), 10% FBS, and 100 U/ml penicillin and 100 μg/ml streptomycin sulfate. Molt-4 (ATCC CRL-1582), Jurkat E6.1 (ATCC TIB-152) and HTLV-1-infected MJ[G11] (ATCC CRL-8294) and HuT-102 lymphocytes (ATCC TIB-162) were grown in RPMI medium (ATCC) supplemented with 20% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin sulfate, and 20 μg/ml gentamicin sulfate (Sigma Chemical Corp.) and cultured at 10% CO2. Primary HTLV-1-infected lymphocytes were obtained after informed consent from three ATLL patients (ATL-1, ATL-2, ATL-3) and were cultured in RPMI medium supplemented with 20% FBS, 50 U/ml hIL-2 (Invitrogen-Life Technologies), 100 U/ml penicillin, 100 μg/ml streptomycin sulfate, and 20 μg/ml gentamicin sulfate. The cytomegalovirus (CMV)-HTLV-1 p30II (hemagglutinin [HA]) expression construct was kindly provided by G. Franchini (NCI, NIH) and has been reported by Koralnik et al. (34). pSG5-HTLV-1 p13II (10), which expresses a protein corresponding to amino acid residues 155 to 241 of HTLV-1 p30II, was provided by V. Ciminale (University of Padua, Italy) and CMV-HTLV-1 p13II (HA) was provided by C. Nicot (University of Kansas). In order to generate the human cyclin D2 promoter-luciferase reporter construct, sequences encompassing the human cyclin D2 promoter were located in the clone with GenBank accession number U47284; according to these sequences, a PCR product that contains 1,622 nucleotides upstream of the ATG start codon was generated. Two closely spaced E-boxes (5′-CACGTG) are localized within the promoter region which binds Myc/Max/Mad network components (7). This fragment was cloned into the pGL3-luciferase vector. Both E-box sequences were mutated to 5′-CTCGAG using the quick change method. The M4-tk-luciferase (M4-tk-luc) reporter plasmid was reported by Bouchard et al. (7) and Vervoorts et al. (79). The CβF-FLAG-Myc, CβF-FLAG-TRRAP1261-1579, CβS-TRRAPantisense, and CβS constructs were described by McMahon et al. (41). The pOZ-wildtype-TIP60 and pOZ-TIP60ΔHAT expression constructs were reported by Ikura et al. (25), and the CMV-TIP60L497A expression plasmid was reported by Gaughan et al. (17). All transfections were performed using Lipofectamine (Invitrogen-Life Technologies) or Superfect (QIAGEN) reagents as recommended by the manufacturers.
Molt4 and Jurkat E6.1 lymphocytes were seeded in 100 mm2 tissue culture dishes and transfected with CMV-HTLV-1 p30II (HA) or an empty CβS vector. After 48 h, cultures were split and either labeled for 4 h by adding BrdU (BD-Pharmingen) to the medium or immediately stained using annexin V-(fluorescein isothiocyanate [FITC])/propidium iodide (BD-Pharmingen). For cell cycle analyses, transfected BrdU-labeled cells were permeabilized and stained with a FITC-conjugated anti-BrdU antibody, and total genomic DNA was stained using 7-AAD (BD-Pharmingen). Flow cytometry was performed and data were analyzed using ModFit LT 3.0 software.
Immortalized Werner's Syndrome (WRN−/−) fibroblasts (45) were seeded at 6 × 105 cells in 60 mm2 tissue-culture dishes in DMEM supplemented with 10% FBS and cultured at 37°C under 5% CO2. Cells were transfected with an empty CβS vector, CMV-HTLV-1 p30II (HA), CβF-FLAG-Myc, and combinations of CMV-HTLV-1 p30II (HA)/CβF-FLAG-Myc or CβS/CβF-FLAG-Myc using Superfect reagent. Foci were observed within 2 weeks and quantified by direct counting. Expression of HTLV-1 p30II (HA) was detected by fixing plates with 0.2% glutaraldehyde-1% formaldehyde in PBS and immunostaining using a monoclonal antibody against the HA epitope tag (CA5; Roche Molecular Biochemicals), diluted 1:1,000 in BLOTTO buffer (50 mM Tris-HCl [pH 8.0], 2 mM CaCl2, 80 mM NaCl, 0.2% [vol/vol] NP-40, 0.02% [wt/vol] sodium azide, and 5% [wt/vol] nonfat dry milk). HTLV-1 p30II (HA) was visualized by immunofluorescence microscopy. p30II-expressing fibroblast colonies were isolated and expanded in six-well tissue culture plates in DMEM supplemented with 10% FBS, 100 U penicillin, and 100 μg/ml streptomycin sulfate.
Myc-interacting complexes were immunoprecipitated from transfected Jurkat E6.1 or HTLV-1-infected MJ[G11] and HuT-102 lymphocytes expressing HTLV-1 p30II (HA) using a monoclonal anti-HA tag antibody. Immunoprecipitation of endogenous p30II from cultured HTLV-1-infected ATLL patient-derived lymphocytes was performed using a rabbit polyclonal antibody against the COOH terminus of p30II (anti-HTLV-1 p30II antibody was generously provided by G. Franchini, NCI, NIH ). Briefly, 3 × 106 cells were harvested by centrifugation and lysed in RIPA buffer (1× PBS, 1% [vol/vol] IGEPAL CA-630, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate [SDS]) containing the protease inhibitors bestatin, pepstatin, antipain dihydrochloride, chymostatin, and leupeptin (50 ng/ml each; Roche Molecular Biochemicals) followed by passage through a 27.5-gauge tuberculin syringe. Immunoprecipitations were carried out by incubating precleared extracts with primary antibodies. Ten microliters of recombinant protein G-agarose (Invitrogen-Life Technologies) was added, and reactions were incubated with agitation at 4°C overnight. Matrices were pelleted by centrifugation at 6,500 rpm for 5 min and washed twice with RIPA buffer. Samples were resuspended in 40 μl 2× SDS-polyacrylamide gel electrophoresis loading buffer, and bound proteins were resolved by electrophoresis through 4 to 15% gradient or 12.5% Tris-glycine SDS-polyacrylamide gels. Chromatin-immunoprecipitations were performed using a kit from Upstate Biotechnology. Nucleoprotein complexes were cross-linked in vivo by adding 270 μl formaldehyde to approximately 3 × 106 Molt-4 or HTLV-1-infected MJ[G11] and HuT-102 lymphocytes in 100 mm2 tissue-culture dishes for 10 min. Cells were pelleted by centrifugation and resuspended in 200 μl SDS lysis buffer. Chromatin DNA was fragmented by sonication, and oligonucleosomal-protein complexes were immunoprecipitated using primary antibodies and 60 μl salmon sperm DNA/protein A agarose. Precipitated oligonucleosomal-protein complexes were washed, cross-links were reversed, and bound DNA fragments were amplified by PCR using specific oligonucleotide primer pairs that flank conserved E-box elements within the human cyclin D2 gene promoter (PRM, 5′-CCCCTTCCTCCTGGAGTGAAATAC-3′ and 5′-CGTGCTCTAACGCATCCTTGAGTC-3′) or anneal within an untranslated region (UTR, 5′-ATCAGACCCTATTCTCGGCTCAGG-3′ and 5′-CAGTCAGTAAGGCACTTTATTTCCCC-3′), as described by Vervoorts et al. (79). PCR products were electrophoresed through a 2% Tris-acetate-EDTA agarose gel and visualized by ethidium bromide staining.
HTLV-1 p30II increases S-phase progression and promotes polyploidy. The conserved pX domain of HTLV-1 encodes at least five nonstructural regulatory factors, including the viral transactivator Tax and an alternative splice variant, p30II (Fig. (Fig.1A).1A). The HTLV-1 p30II protein is comprised of 241 amino acid residues and contains Arg- and Ser/Thr-rich domains (1, 34, 35). RasMol structural prediction analyses (Brookhaven protein databank) indicate that p30II possesses 4 alpha-helices and 19 beta-sheet regions (Fig. (Fig.1B).1B). The alpha-helices likely serve as interacting or docking sites for cellular factors, whereas the Ser/Thr-rich domains may provide targets for phosphorylation by kinases that modulate p30II's functions or interactions. As relatively little is known with respect to the functions of HTLV-1 pX accessory factors, such as p30II, we investigated whether the p30II protein contributes to lymphoproliferation in HTLV-1-infected T cells by altering cell cycle regulation. To determine whether HTLV-1 p30II influences cell cycle progression and/or apoptosis, Molt-4 and Jurkat E6.1 lymphocytes were transfected with a CMV-HTLV-1 p30II (HA) expression construct or an empty CβS vector control, and transfected cultures were assayed for bromodeoxyuridine (BrdU)-incorporation/cell cycle progression or programmed cell death using flow cytometric analyses (Fig. 1C and D and data not shown). HTLV-1 p30II-expressing cells exhibit markedly increased S-phase progression and significant polyploidy as determined by BrdU incorporation and 7-AAD staining of total genomic DNA (Fig. 1C and D, top left panels). However, p30II did not induce apoptosis in transfected cells, as determined by annexin V-FITC/propidium iodide-staining and FACS (Fig. 1C and D, top right panels). These results suggest that p30II may contribute to lymphoproliferation and genomic instability in HTLV-1-infected cells during ATLL by affecting S-phase regulatory factors, such as Myc and/or E2F (2, 29, 43).
The p30II protein was detected in cultured HTLV-1-infected lymphocytes, derived from three different ATLL patients (ATL-1, ATL-2, ATL-3) diagnosed with clinical acute-stage leukemias, by immunofluorescence laser confocal microscopy and immunoblotting (Fig. 2A and B). Three-dimensional Z-stack composite images for ATL-3 demonstrate that p30II/Myc proteins colocalize in the nucleus in all focal planes in HTLV-1-infected cells (Fig. (Fig.2A,2A, right panels). Relative fluorescence intensities for p30II/Myc-specific signals and DAPI (4′,6′-diamidino-2-phenylindole) nuclear staining are shown for reference (Fig. (Fig.2A,2A, right panels). HTLV-1 p30II is present in Myc-containing immunoprecipitated complexes in ATLL patient lymphocytes (Fig. (Fig.2B).2B). Intriguingly, immunoprecipitation of Myc revealed that TIP49 (RUVBL1), TIP48 (RUVBL2) (81), and Max are present and are bound to Myc, but the TIP60 histone acetyltransferase (HAT) was not detected in Myc-containing coimmune complexes in uninfected Jurkat E6.1 lymphocytes (Fig. (Fig.2B).2B). The NH2 terminus of Myc is essential for Myc-dependent transformation and apoptosis-inducing functions and contains two conserved Myc homology domains (Myc box I [MBI] and MBII, respectively) that interact with cellular factors (2, 3, 7, 41, 42, 51, 81). The transcriptional coactivator, TRRAP/p434, and the ATPases/helicases, TIP49 (RUVBL1) and TIP48 (RUVBL2), interact with amino acids within MBII (41, 81). To determine if HTLV-1 p30II interacts with known Myc-binding partners, we transfected Jurkat E6.1 lymphocytes or HTLV-1-infected Hut-102 and MJ[G11] lymphocytes with CMV-HTLV-1 p30II (HA) or an empty CβS vector control and performed coimmunoprecipitations using a monoclonal anti-HA antibody (CA5; Roche Molecular Diagnostics). As shown in Fig. Fig.2C,2C, HTLV-1 p30II (HA) coimmunoprecipitates with Myc, TRRAP, TIP60, and TIP49 (RUVBL1). However, TIP48 (RUVBL2) and RNA polymerase II were not detected in anti-HA immunoprecipitates, although both proteins were detected in control immunoprecipitations using antibodies against known interacting proteins (Fig. (Fig.2C).2C). To further confirm these interactions, we reimmunoprecipitated HTLV-1 p30II (HA) from extracts prepared from transfected Jurkat T cells using antibodies against Myc, TRRAP, TIP60, TIP48, and TIP49 (Fig. (Fig.2C,2C, lower panels). A nonspecific antibody (rabbit preimmune serum) was included as a negative control. Interestingly, the ATPase/helicase, TIP48, was detected in p30II-complexes immunoprecipitated with an anti-TIP48 (RUVBL2) polyclonal antibody (Fig. (Fig.2C,2C, lower panels) (81). These data suggest that HTLV-1 p30II may modulate Myc functions through interactions with Myc-associated transcriptional coactivators on promoters of responsive genes (14).
To investigate the possibility that HTLV-1 p30II might affect Myc-dependent transcription, we next cotransfected HeLa cells with a human cyclin D2 promoter-luciferase reporter construct, containing two conserved Myc-responsive E-box enhancer elements (CACGTG), in the presence of increasing amounts of CMV-HTLV-1 p30II (HA). Results in Fig. Fig.3A3A demonstrate that HTLV-1 p30II significantly transactivates the human cyclin D2 promoter. A mutant cyclin D2 promoter, lacking Myc-responsive E-box elements (79), was not transcriptionally activated by p30II, indicating that p30II-mediated transactivation from the human cyclin D2 promoter requires the conserved Myc-responsive E-box enhancer elements (Fig. 3A and B). The HTLV-1 p30II (HA)-tagged protein was detected in transfected cells by immunoblotting using a monoclonal anti-HA antibody (CA5; Roche Molecular Biochemicals) (Fig. (Fig.3A).3A). Intracellular levels of Myc were not altered by HTLV-1 p30II expression (Fig. (Fig.3A,3A, lower panels). HTLV-1 p30II also transcriptionally activates the human cyclin D2 promoter in transfected 293A fibroblasts in a dose-dependent manner (Fig. (Fig.3C).3C). To confirm that HTLV-1 p30II promotes Myc-dependent transcription from E-box enhancer elements, we cotransfected 293A fibroblasts and HeLa cells with a synthetic tk minimal promoter-luciferase reporter construct (M4-tk-luc) that contains four tandem E-boxes (79). As shown in Fig. Fig.3D,3D, HTLV-1 p30II transactivates E-box enhancer elements within M4-tk-luc, suggesting that p30II promotes S-phase progression through Myc-dependent transcriptional interactions. Interestingly, we observed that p30II, at the lowest concentration used, induced approximately 13-fold transactivation from the synthetic M4-tk-luc promoter, whereas higher concentrations induced lower (5- to 7-fold) levels of transcriptional activation (Fig. (Fig.3D).3D). These observations are consistent with findings by Zhang et al. (87) demonstrating that p30II-dependent transactivation from the HTLV-1 promoter (Tax-responsive elements) occurs maximally at low p30II concentrations and diminishes with increased p30II expression (87).
Frank et al. reported that Myc interacts with the transcriptional coactivator/HAT, TIP60 (16), and Patel et al. have recently shown that c-Myc is a substrate for lysine acetylation by TIP60 and hGCN5 (56). Myc has also previously been demonstrated to interact in chromatin-remodeling complexes with the ATM-related TRRAP/p434 protein (41, 42, 51). Therefore, we tested whether HTLV-1 p30II-mediated transactivation requires TIP60 and TRRAP/p434 functions. HeLa cells were cotransfected with a human cyclin D2 promoter-luciferase reporter construct and CMV-HTLV-1 p30II (HA) in the presence of increasing amounts of CMV-TIP60, CMV-TIP60ΔHAT (a trans-dominant-negative HAT-inactive mutant ), or CMV-TIP60L497A, a COOH-terminal mutant impaired for interactions with cellular factors, including the androgen receptor (17). Ectopic expression of TIP60 alone did not significantly transactivate the human cyclin D2 promoter; however, TIP60 overexpression enhanced HTLV-1 p30II-mediated transactivation in a dose-dependent manner (Fig. (Fig.4A).4A). The trans-dominant-negative TIP60ΔHAT mutant potently inhibited p30II-mediated transcriptional activation (Fig. (Fig.4A),4A), suggesting that HTLV-1 p30II transactivation requires TIP60-associated HAT activity (25). The TIP60L497A mutant also weakly enhanced p30II-mediated transactivation (Fig. (Fig.4A).4A). Overexpression of wild-type TIP60 or the trans-dominant-negative TIP60ΔHAT mutant did not alter expression of the HTLV-1 p30II (HA) protein in transfected HeLa cells (Fig. (Fig.4A,4A, lower panels). Inhibition of TRRAP/p434, as a result of coexpressing either TRRAPantisense RNA or a trans-dominant-negative TRRAP mutant, TRRAP1261-1579 (FLAG-epitope-tagged ), prevented HTLV-1 p30II-mediated transcriptional activation from the human cyclin D2 promoter (Fig. (Fig.4B).4B). The trans-dominant-negative, FLAG-tagged TRRAP1261-1579 protein did not alter the expression of HTLV-1 p30II (HA) (Fig. (Fig.4B,4B, lower panels). We then performed immunofluorescence microscopy, using a monoclonal anti-FLAG M2 antibody (Sigma Chemical Corp.) and a rabbit polyclonal anti-TIP60 antibody (Upstate Biotechnology), to visualize expression of the FLAG-tagged wild-type TIP60 or TIP60ΔHAT proteins relative to endogenous TIP60 (25). Results shown in Fig. Fig.4C4C demonstrate that the FLAG-tagged TIP60 proteins were drastically overexpressed relative to endogenous TIP60 in transfected cells. To demonstrate the specificity of transcriptional inhibition due to TRRAPantisense RNA in panel B, we repeated these experiments using a pSPORT-lacZ control plasmid which expresses β-galactosidase mRNA. Results shown in Fig. Fig.4D4D demonstrate that increased β-galactosidase mRNA expression did not influence HTLV-1 p30II-dependent transactivation from the cyclin D2 promoter, whereas TRRAPantisense inhibited p30II transcriptional activation in a dose-dependent manner. These data collectively indicate that HTLV-1 p30II synergizes with the TIP60 HAT to transactivate Myc-responsive E-box elements within the human cyclin D2 promoter, requiring the transcriptional coactivator TRRAP/p434 (7, 25, 41, 79).
As we have shown that HTLV-1 p30II transcriptionally activates the conserved Myc-responsive E-box enhancer elements within the human cyclin D2 promoter (Fig. 3A and C) (7), we sought to determine whether p30II is present in Myc-containing chromatin-remodeling complexes using the ChIP procedure as described by Vervoorts et al. (79). Formaldehyde cross-linked genomic DNA complexes in uninfected Molt-4 lymphocytes or HTLV-1-infected MJ[G11] and HuT-102 lymphocytes were fragmented by sonication, and oligonucleosomal-protein complexes were precipitated using antibodies against candidate Myc-binding factors. Cross-links were reversed, and specific oligonucleotide DNA primer pairs were used in PCRs to amplify immunoprecipitated DNA regions spanning conserved E-box elements (PRM) or an untranslated sequence (UTR) as negative control (79). Results in Fig. Fig.5A5A (top panels) demonstrate that HTLV-1 p30II was detected only bound to E-box enhancer elements in HTLV-1-infected lymphocytes. Myc, TRRAP, TIP49 (RUVBL1), TIP48 (RUVBL2), and the acetyltransferase hGCN5 (42) were present in chromatin-remodeling complexes in uninfected Molt-4 cells and in HTLV-1-infected MJ[G11] and HuT-102 lymphocytes (Fig. (Fig.5A,5A, top panels). Surprisingly, TIP60 was detected only in Myc-containing transcription complexes that contained p30II in HTLV-1-infected T cells (Fig. (Fig.5A,5A, top panels), consistent with coimmunoprecipitation results and observed effects of ectopic TIP60 in transactivation assays (see Fig. Fig.2B2B and and4A).4A). The diminished recruitment of TIP49 to Myc-containing transcription complexes on the cyclin D2 promoter in HTLV-1-infected MJ[G11] cells was not attributable to apparent differences in p30II/Myc/TIP60 interactions (Fig. (Fig.5A).5A). Histone H3 acetylation surrounding the E-box enhancer elements within the human cyclin D2 promoter, consistent with transcriptional activation, was detected in all cell types with the exception of H3, which appeared to be differentially acetylated on Lys-9 and Lys-14 residues in HTLV-1-infected MJ[G11] and HuT-102 cells, respectively (Fig. (Fig.5A,5A, lower panels). Differences in histone H3 acetylation, however, did not correlate with the stabilization of p30II/Myc/TIP60 transcriptional interactions in HTLV-1-infected T-cell lines.
To identify residues within HTLV-1 p30II that interact with Myc/TIP60 complexes in vivo, we generated a panel of pGEX 4T.1-glutathione S-transferase (GST)-HTLV-1 p30II constructs, expressing full-length GST-HTLV-1 p30II or various truncation mutants, GST-p30II (residues 1 to 98), GST-p30II (residues 99 to 154), GST-p30II (residues 155 to 241) spanning the entire coding region of HTLV-1 p30II (Fig. (Fig.5B,5B, see diagram). These proteins were expressed in Escherichia coli BL21 bacteria, and purified recombinant GST-HTLV-1 p30II fusion proteins were used in GST pull-down experiments as described by Harrod et al. (23). GST proteins were incubated with HeLa nuclear extracts at 4°C overnight, and complexes were precipitated with glutathione-Sepharose 4B (Amersham-Pharmacia Biotech). The matrices were washed, and bound factors were eluted using 10 mM reduced glutathione buffer. Input levels of purified recombinant GST or GST-HTLV-1 p30II proteins, Myc, and TIP60 are shown in Fig. Fig.5B.5B. Results shown in Fig. Fig.5B5B (right panels) demonstrate that full-length GST-HTLV-1 p30II interacts with both Myc and TIP60 in HeLa nuclear extracts. Deletion of amino acid residues from either the NH2 terminus or COOH terminus of p30II disrupts Myc binding; however, the TIP60-interacting region of HTLV-1 p30II was mapped to residues between positions 99 and 154 (Fig. (Fig.5B).5B). Our future efforts will biochemically characterize specific amino acid contacts responsible for the stabilization of HTLV-1 p30II/Myc/TIP60 transcriptional interactions.
We next examined recruitment of HTLV-1 p30II/Myc/TIP60 chromatin remodeling complexes to conserved, Myc-responsive E-box enhancer elements within the cyclin D2 promoter in cultured HTLV-1-infected ATLL patient lymphocytes (ATL-1). Chromatin-immunoprecipitations were performed using antibodies that recognize endogenous HTLV-1 p30II (34), Myc, and known Myc-interacting factors as described previously. Polymerase chain-reaction amplification of ChIP products was performed using the PRM and UTR oligonucleotide DNA primer pairs (79). Results shown in Fig. Fig.5C5C demonstrate that p30II is present in Myc/TIP60 transcription complexes assembled on E-box enhancer elements within the cyclin D2 promoter in HTLV-1 ATLL patient lymphocytes. The transcriptional coactivators, TRRAP/p434, TIP48, TIP49, and hGCN5 were also detected in p30II/Myc/TIP60/cyclin D2 promoter complexes (Fig. (Fig.5C5C).
We next investigated whether HTLV-1 p30II interacts similarly in Myc/TIP60 transcription complexes in 293A fibroblasts. Nicot et al. (48) have demonstrated that an HTLV-1 p30II-green fluorescent protein (GFP) is functionally identical to HTLV-1 p30II (HA) (48). We therefore cotransfected 293A cells with CMV-HTLV-1 p30II-GFP (kindly provided by G. Franchini, NCI, NIH ) or a pcDNA3.1-GFP vector control and performed ChIP analyses. Nucleoprotein complexes were cross-linked by treatment with formaldehyde, and oligonucleosomal fragments were generated by brief sonication of extracted genomic DNA. Chromatin immunoprecipitations were performed as described above, and ChIP products were amplified by PCR using the PRM and UTR oligonucleotide DNA primer pairs (79). Similar expression of HTLV-1 p30II-GFP and GFP proteins was visualized with transfected 293A fibroblasts by fluorescence microscopy (Fig. 6A and B). The HTLV-1 p30II-GFP protein was immunoprecipitated and bound to Myc-containing transcription complexes on conserved E-box elements within the cyclin D2 promoter in transfected 293A fibroblasts, using an anti-GFP antibody (Fig. (Fig.6A).6A). No ChIP product was detected for the anti-GFP immunoprecipitation in 293A cells transfected with the pcDNA3.1-GFP control (Fig. (Fig.6B).6B). While the transcriptional coactivators TRRAP/p434, TIP48, TIP49, and hGCN5 were present in Myc-containing complexes in both HTLV-1 p30II-GFP and GFP-expressing cells, the TIP60 HAT was detected predominantly in HTLV-1 p30II-GFP/Myc/TIP60 complexes (compare Fig. 6A and B). However, TIP60 was weakly present in Myc-containing ChIP complexes in GFP-expressing cells, consistent with the demonstration of pre-existing Myc-TIP60 interactions by Frank et al. (16) and Patel et al. (56) (Fig. (Fig.6B6B).
To determine whether the HTLV-1 p30II-GFP protein also transcriptionally activates the human cyclin D2 promoter in a TIP60-dependent manner, we cotransfected 293A fibroblasts with a tk promoter-Renilla luciferase plasmid, a human cyclin D2 promoter-luciferase reporter plasmid, and CMV-HTLV-1 p30II-GFP in the presence of increasing amounts of either CMV-TIP60 (wild-type) or CMV-TIP60ΔHAT, which expresses a trans-dominant-negative TIP60 mutant (7, 25, 48). Results shown in Fig. Fig.6C6C demonstrate that HTLV-1 p30II-GFP transcriptionally activates the human cyclin D2 promoter approximately 14-fold in transfected 293A fibroblasts compared to an empty pcDNA3.1-GFP control. Overexpression of wild-type TIP60, in the presence of HTLV-1 p30II-GFP, significantly increased p30II-GFP-dependent transcriptional activity in a dose-dependent manner (Fig. (Fig.6C).6C). Coexpression of the trans-dominant-negative TIP60ΔHAT mutant (25) repressed p30II-GFP-dependent transactivation from the human cyclin D2 promoter (Fig. (Fig.6C),6C), consistent with the results shown in Fig. Fig.4A4A and with an essential role for the TIP60 HAT in HTLV-1 p30II transcriptional activation. Relative Renilla luciferase activities for each sample are shown in Fig. Fig.6D6D for comparisons of similar transfection efficiencies.
To comprehensively identify cellular gene sequences whose expressions are altered by HTLV-1 p30II-TIP60 transcriptional interactions, we cotransfected 293A fibroblasts with a CβS empty vector control, CMV-HTLV-1 p30II (HA), or CMV-HTLV-1 p30II (HA) and TIP60ΔHAT, which expresses a trans-dominant-negative mutant that interferes with endogenous TIP60 functions (25). Total cellular RNAs were extracted, and microarray gene expression analyses were performed using Affymetrix Human U133Plus 2.0 full-genomic chips. Transcriptional activation of cellular target genes is expressed as activation (n-fold) relative to the empty CβS vector control, and the lower limit for transactivation was set at 2.5-fold. Figure Figure7A7A shows a graphical representation of cellular target genes transcriptionally activated by HTLV-1 p30II (HA) (red lines). TIP60-dependent gene sequences were identified based upon their transcriptional repression in the presence of the TIP60ΔHAT mutant (25) and are indicated by green lines (Fig. (Fig.7A).7A). In general, the fold transactivation by HTLV-1 p30II (HA) ranged between 2.5-fold to 393-fold for specific target genes (Fig. (Fig.7A).7A). Michael et al. (44) have demonstrated that numerous cellular genes are also transcriptionally repressed as a result of HTLV-1 p30II expression (44). Results shown in Fig. Fig.7B7B graphically represent cellular target genes transcriptionally repressed (with levels ranging between 2.5-fold to 125-fold transrepression) by HTLV-1 p30II (HA) (red lines). Effects of the trans-dominant-negative TIP60ΔHAT mutant upon transcriptional repression by HTLV-1 p30II (HA) are indicated by green lines (Fig. (Fig.7B7B).
In Fig. Fig.7C,7C, we provide a representative list of the major target gene sequences that are transcriptionally activated by HTLV-1 p30II (HA) as determined by Affymetrix microarray gene expression analyses. TIP60-dependent gene sequences are shown in boxes. Transcriptional activation is expressed as activation (n-fold) relative to the empty CβS vector control. Numerous cellular genes were transcriptionally induced by HTLV-1 p30II (HA) in a TIP60-dependent or TIP60-independent manner, suggesting that p30II may participate in multiple, distinct transcription complexes (Fig. (Fig.7C).7C). With respect to the potential role of HTLV-1 p30II in adult T-cell leukemogenesis, transcriptional activation of the following genes is of significant interest: myeloid cell nuclear differentiation 1 antigen (31.1-fold; TIP60 dependent), protocadherin 15 (26.1-fold; TIP60 dependent), human protein tyrosine phosphatase delta precursor (23.3-fold; TIP60 dependent), cadherin 11-like precursor (20.2-fold; TIP60 dependent), colony-stimulating factor 2 receptor beta (19.6-fold; TIP60 independent), human protein tyrosine phosphatase receptor type Z polypeptide (16.4-fold; TIP60 dependent), Schizosaccharomyces pombe RAD21-like protein (16-fold; TIP60 independent), human transmembrane phosphatase with tensin homology (15.5-fold; TIP60 independent), H2B histone family member N (15.1-fold; TIP60 independent), major histocompatibility complex class II DR beta 3 (14.0-fold; TIP60 dependent), human CD84 leukocyte antigen (14.0-fold; TIP60 independent), prostate-specific G protein-coupled receptor (14.0-fold; TIP60 independent), fibroblast growth factor 20 (13.8-fold; TIP60 dependent), protein kinase C alpha-binding protein (13.2-fold; TIP60 independent), regulator of G-protein-signaling 1 (13.1-fold; TIP60 dependent), cytoplasmic linker associated protein 2 (13.0-fold; TIP60 independent), POU domain 4 transcription factor 2 (12.8-fold; TIP60 independent), RNA-binding motif protein (RBMY2B) (12.6-fold; TIP60 independent). Robek et al. (62) have demonstrated that an infectious HTLV-1 molecular clone, ACH.p30II, exhibits an approximately 20 to 50% reduction in transformation efficiency compared to the wild-type ACH.wt (62), suggesting that p30II is required for the full transforming potential of HTLV-1. Our microarray analyses indicate that numerous cellular genes are transcriptionally activated by p30II, and proteins encoded by these genes may contribute to HTLV-1 leukemic transformation and development of ATLL.
As the c-Myc oncogene is known to cause cellular transformation (7, 41, 51), we next investigated whether HTLV-1 p30II might influence Myc-associated transforming activity in focus formation assays using immortalized human WRN−/− fibroblasts, which lack Werner's syndrome helicase functions (45). This cellular background was chosen because ATLL is an aging-related malignancy requiring clinical latency periods of 25 to 40 years prior to disease onset (29), which suggests that genetic mutations linked to the aging process likely contribute to leukemogenesis. Werner's syndrome is a premature aging disorder (45) that mimics or recapitulates many of the clinical and cellular features of normal aging, and WRN locus (8p11-12) mutations have been found in HTLV-1-infected ATLL patient lymphocytes and in HTLV-1-infected mycosis fungoides/Sezary syndrome cells (4, 30, 53, 69, 82). Neither c-Myc nor HTLV-1 p30II (HA) alone significantly induces focus formation in immortalized human WRN−/− fibroblasts (Fig. (Fig.8A).8A). Surprisingly, in combination, HTLV-1 p30II (HA)-Myc coexpression reproducibly induces between 35 and 58 foci in different assays (Fig. 8A and B). The expression of HTLV-1 p30II (HA) and c-Myc (FLAG) was detected in transformed colonies by immunofluorescence microscopy (Fig. 8D and E), and the p30II protein appeared to be distributed throughout the nucleoplasm (Fig. (Fig.8C).8C). We also observed a high incidence of multinucleated giant cells in isolated HTLV-1 p30II (HA) Myc-transformed fibroblasts that were expanded in culture, consistent with HTLV-1 p30II-induced polyploidy observed during BrdU-FACS analyses (Fig. (Fig.8F;8F; compare to control cells in Fig. Fig.8D).8D). The expression of HTLV-1 p30II (HA) in transformed fibroblasts was confirmed by immunoblotting using a monoclonal anti-HA antibody (Fig. (Fig.8E).8E). As expected, the majority of expanded HTLV-1 p30II (HA)-expressing colonies showed increased levels of intracellular Myc protein by immunoblotting (Fig. (Fig.8F).8F). Indeed, these findings indicate that HTLV-1 p30II markedly enhances the transforming potential of c-Myc and may promote genomic instability, resulting in polyploidy.
Our transcriptional activation data suggested that enhancement of Myc functions by HTLV-1 p30II requires the coactivators TIP60 and TRRAP/p434. Therefore, we tested whether focus formation induced by coexpressing HTLV-1 p30II (HA)-Myc might be affected by overexpressing wild-type TIP60 or TIP60ΔHAT and TIP60L497A mutant proteins (17, 25). Results from two independent experiments in Fig. Fig.9A9A indicate that none of the TIP60 expression constructs, either alone or in combination with c-Myc, significantly induces focus formation in immortalized human WRN−/− fibroblasts. However, ectopic TIP60 markedly increases focus formation induced by HTLV-1 p30II (HA)-Myc coexpression (Fig. (Fig.9A).9A). The trans-dominant-negative TIP60ΔHAT mutant completely abrogated colony formation by HTLV-1 p30II (HA)-Myc, and the TIP60L497A mutant partially inhibited focus formation (Fig. (Fig.9A).9A). Increased colony formation by HTLV-1 p30II (HA)/Myc/TIP60, compared to inhibition of focus formation by the trans-dominant-negative TIP60ΔHAT mutant, is shown in Fig. Fig.9B.9B. Inhibition of TRRAP/p434, as a result of coexpressing increasing amounts of TRRAPantisense RNA (41), also significantly decreased focus formation by HTLV-1 p30II (HA)-Myc (Fig. (Fig.9C).9C). These findings collectively agree with our transcriptional activation data and suggest that HTLV-1 p30II enhances Myc transcriptional and transforming activities in a TIP60 HAT- and TRRAP-dependent manner.
As we have mapped the TIP60-interacting domain of HTLV-1 p30II to amino acid residues 99 to 154 through biochemical GST pull-down experiments (see Fig. Fig.5B),5B), we next analyzed a naturally occurring truncation mutant of p30II, HTLV-1 p13II, which expresses the carboxyl terminus of p30II, spanning from residue 155 to 241 (Fig. 10A) (1, 10, 34, 70). The p13II mutant lacks the TIP60-interacting region of p30II but contains the nuclear localization sequence as reported in references 1 and 34. Molt-4 lymphocytes were transfected with CMV-HTLV-1 p30II (HA), CMV-HTLV-1 p13II (HA), or a CβS empty vector control, and immunofluorescence microscopy was performed using an anti-HA (CA5) primary antibody and rhodamine red-conjugated fluorescent secondary antibody to visualize protein expression in transfected cells. The p30II (HA) and p13II (HA) proteins were observed in approximately 20 to 30% of transfected Molt-4 lymphocytes (Fig. 10B). We then analyzed BrdU incorporation and S-phase cell cycle progression in HTLV-1 p30II (HA)- or p13II (HA)-expressing transfected lymphoid cultures, compared to the CβS control. Results shown in Fig. 10C demonstrate that p30II (HA) expression markedly increased S-phase progression and polyploidy as noted in previous experiments (see Fig. Fig.1D),1D), whereas neither p13II (HA) nor the CβS control resulted in altered cell cycle progression (Fig. 10C).
To determine whether the TIP60-interacting domain (residues 99 to 154) of HTLV-1 p30II (HA) is essential for its oncogenic function, we compared the ability of p30II (HA) and p13II (HA) (corresponding to amino acids 155 to 241 of HTLV-1 p30II) to promote focus formation in immortalized human WRN−/− fibroblasts in combination with c-Myc, as shown in Fig. Fig.8A.8A. These results demonstrate that the p13II (HA) mutant, lacking residues 1 to 154 of p30II, is significantly defective for cellular transformation and focus formation compared to wild-type p30II (HA) (Fig. 10D), suggesting that TIP60 recruitment is required for p30II-associated oncogenic activity. Finally, we tested the capacity of HTLV-1 p30II (HA) and p13II (HA) to transcriptionally activate the human cyclin D2 promoter-luciferase reporter construct in transfected 293A fibroblasts. Results shown in Fig. 10E demonstrate that p13II (HA), lacking the TIP60-interacting domain, is impaired for transcriptional-activating functions compared to p30II (HA), which transactivates the cyclin D2 promoter approximately eight- to ninefold. Indeed, p13II (HA) exhibited a trans-dominant-negative effect upon Myc-dependent transactivation from the cyclin D2 promoter and slightly repressed transcription below the basal level (Fig. 10E). Chromatin-immunoprecipitation analyses were performed with 293A fibroblasts expressing either HTLV-1 p30II (HA) or p13II (HA), by using antibodies against HTLV-1 p30II (the anti-HTLV-1 p30II or TofII antibody recognizes a peptide epitope within the COOH terminus of p30II that is also present in HTLV-1 p13II ), Myc, TIP60, TRRAP, TIP48, TIP49, and hGCN5. Immunoprecipitation products were amplified using the PRM primer pair, which anneals to nucleotide sequences flanking the conserved Myc-responsive E-box elements within the human cyclin D2 gene promoter (79). The p30II (HA) protein was precipitated in Myc-containing chromatin-remodeling complexes that contain TIP60, TRRAP, TIP48, TIP49, and hGCN5 (Fig. 10F, top panel). However, the p13II (HA) protein was not detected bound to Myc-responsive E-box elements within the cyclin D2 promoter and, consistent with p13II's transcriptional impairment, the TIP60 HAT was not present in Myc-containing cyclin D2 promoter complexes in the absence of p30II (HA) (Fig. 10F, lower panel). Our data suggest that the HTLV-1 p30II oncoprotein enhances Myc-dependent transcriptional and transforming activities through the stabilization of Myc-TIP60 interactions on promoters of Myc-responsive genes, which may also influence the acetylation of Myc protein by the TIP60 coactivator (Fig. (Fig.11)11) (56).
HTLV-1 infects CD4+ T cells and promotes deregulated cell growth and lymphoproliferation associated with the development of ATLL. While numerous studies have demonstrated that the viral Tax protein transcriptionally activates growth/proliferative-signaling pathways, it has become increasingly evident that other pX-encoded regulatory factors (p12I, p13II, p30II, Rex) are likely to perform essential functions during adult T-cell leukemogenesis (1, 6, 29, 34, 35, 48, 85). Indeed, the majority of partially deleted HTLV-1 proviruses in ATLL patient isolates contain intact pX sequences (33, 68), and alternatively spliced ORF I and ORF II mRNAs in HTLV-1-infected transformed T-cell lines and ATLL patient samples have been detected (6, 35). Cytotoxic T-lymphocytes specifically targeted against ORF I and ORF II peptides have been obtained from ATLL patients, suggesting that these proteins are present during in vivo HTLV-1 infections (57). Zhang et al. (86) reported that p30II interacts with p300/CREB-binding protein and represses Tax-mediated transactivation from the HTLV-1 LTR (86) and differentially modulates CREB-dependent transcription (87). Nicot et al. (48) and Younis et al. (85) have demonstrated that p30II prevents nuclear export of the doubly spliced Tax/Rex mRNA, and others have shown that p30II is required for maintenance of high viral titers in a rabbit model of ATLL using an infectious HTLV-1 molecular clone, ACH.30II, which is defective for p30II production (5, 71). Interestingly, Robek et al. (62) have previously demonstrated that p30II is dispensable for immortalization and transformation of human peripheral blood mononuclear cells by ACH.p30II; however, this mutant exhibited an approximately 20 to 50% reduction in transformation efficiency compared to the wild-type ACH.wt (62), suggesting that p30II is required for the full transforming potential of HTLV-1. The physiological role of p30II in HTLV-1 pathogenesis remains unclear, and it is intriguing that, similar to Tax, p30II may perform multiple functions to control viral gene expression and promote deregulation of CD4+ T-cell growth/proliferative pathways.
With this study, we have demonstrated that HTLV-1 p30II markedly enhances Myc-associated transcriptional and transforming activities and increases S-phase progression and polyploidy through interactions with the coactivator/HAT, TIP60 (Fig. (Fig.11).11). HTLV-1 p30II transactivates conserved E-box enhancer elements within promoters of Myc-responsive genes, requiring TIP60 HAT activity and the transcriptional coactivator TRRAP/p434. Frank et al. (16) have shown that pre-existing Myc-TIP60 interactions contribute to Myc-dependent transcriptional activation and chromatin-remodeling associated with histone H4 acetylation on a subset of Myc-responsive genes in rodent and human fibroblasts, although their data suggest that Myc-TIP60 interactions may be relatively unstable on certain promoters. Patel et al. also recently demonstrated that c-Myc is a substrate for lysine acetylation by the TIP60 and hGCN5 acetyltransferases (56). Indeed, Myc and the TIP60 HAT likely exist in multiple distinct nuclear complexes, and Park et al. have demonstrated that TIP60 is not present in Myc/BAF53-containing transcription complexes (54). Our data indicate that, in absence of HTLV-1 p30II-interactions, ectopic TIP60 overexpression does not significantly alter Myc transcriptional and transforming activities in functional assays (see Fig. Fig.4A,4A, ,6C,6C, and and9A).9A). Further, we have shown that TIP60 is not detectably present in Myc-containing chromatin-remodeling complexes on the human cyclin D2 promoter (7, 79), in the absence of HTLV-1 p30II, in uninfected Molt-4 lymphocytes (Fig. (Fig.5A).5A). However, we did detect weak recruitment of TIP60 to Myc transcription complexes on the cyclin D2 promoter in pcDNA3.1-GFP-transfected 293A fibroblasts by ChIPs (Fig. (Fig.6B),6B), consistent with the notion that Myc-TIP60 interactions may be relatively unstable on certain gene promoters. Thus, aberrant stabilization of Myc-TIP60 interactions, as a result of HTLV-1 p30II or other stabilizing factors, may contribute prominently to neoplastic transformation in hematological malignancies and solid tumors where Myc functions are deregulated or where myc locus mutations are present (18, 24, 26, 43, 55, 60).
The GST-HTLV-1 p30II protein interacts with both Myc and TIP60, and amino acid residues located between positions 99 and 154 of p30II interact with the TIP60 HAT in vivo. Recruitment of TIP60 is essential for p30II-dependent effects upon cell cycle progression and focus formation/transformation. Affymetrix microarray gene expression analyses indicate that numerous cellular genes are transcriptionally activated by HTLV-1 p30II in a TIP60-dependent or TIP60-independent manner. These gene products could play important roles in HTLV-1-associated neoplastic disease. Our results indicate that HTLV-1 p30II is a novel retroviral enhancer of Myc-TIP60 transcriptional and transforming activities that may contribute to adult T-cell leukemogenesis.
This work was supported by the Department of Biological Sciences, Southern Methodist University, Dallas, TX 75275-0376. B.L. acknowledges grant support from the Deutsche Forschungsgemeinschaft.
We thank G. Franchini (NCI, NIH) for generously providing CMV-HTLV-1 p30II (HA), CMV-HTLV-1 p30II-GFP, and the anti-HTLV-1 p30II polyclonal antibody. We thank V. Ciminale (Department of Oncology and Surgical Sciences, University of Padua, Italy) for providing pSG-HTLV-1 p13II and C. Nicot (Department of Microbiology, Immunology and Molecular Genetics, University of Kansas) for providing CMV-HTLV-1 p13II (HA). We also thank J. K. Nyborg (Department of Biochemistry and Molecular Biology, Colorado State University) and R. S. Jones (Department of Biological Sciences, Southern Methodist University) for helpful comments and Carolyn K. Harrod for assistance in preparing the manuscript. Other members of the Harrod lab are thanked for their discussions and for critically reading the manuscript.