Characterizing the Interaction between the HTLV-1 Transactivator Tax-1 with Transcription Elongation Factor ELL2 and Its Impact on Viral Transactivation

Abstract

The human T-cell leukemia virus type 1 (HTLV-1)-encoded transactivator and oncoprotein Tax-1 is essential for HTLV-1 replication. We recently found that Tax-1 interacts with transcription elongation factor for RNA polymerase II 2, ELL2, which enhances Tax-1-mediated transactivation of the HTLV-1 promotor. Here, we characterize the Tax-1:ELL2 interaction and its impact on viral transactivation by confocal imaging, co-immunoprecipitation, and luciferase assays. We found that Tax-1 and ELL2 not only co-precipitate, but also co-localize in dot-like structures in the nucleus. Tax-1:ELL2 complex formation occurred independently of Tax-1 point mutations, which are crucial for post translational modifications (PTMs) of Tax-1, suggesting that these PTMs are irrelevant for Tax-1:ELL2 interaction. In contrast, Tax-1 deletion mutants lacking either N-terminal (aa 1-37) or C-terminal regions (aa 150-353) of Tax-1 were impaired in interacting with ELL2. Contrary to Tax-1, the related, non-oncogenic Tax-2B from HTLV-2B did not interact with ELL2. Finally, we found that ELL2-R1 (aa 1-353), which carries an RNA polymerase II binding domain, and ELL2-R3 (aa 515-640) are sufficient to interact with Tax-1; however, only ELL2-truncations expressing R1 could enhance Tax-1-mediated transactivation of the HTLV-1 promoter. Together, this study identifies domains in Tax-1 and ELL2 being required for Tax-1:ELL2 complex formation and for viral transactivation.

Keywords: ELL2; HTLV-1; Tax-1; Tax-2; human T-cell leukemia virus type 1; transcription elongation factor for RNA polymerase II; viral oncoprotein; viral transactivation.

Figure A1

Tax point mutants differentially activate NF-κB signaling. The NF-κB reporter plasmid (NF-κB-Luc; 0.1 μg) was transfected together with Tax-1 wildtype (pSG-Tax-WT) or point mutants of Tax-1 (pSG-Tax-PQ, pSG-Tax-K1-10R, pSG-Tax-K4-8R, pSG-Tax-K6-8R, and pSG-TaxR7-8K; 0.01 μg each) in 293T cells. Co-transfection of the dominant-negative mutant IκBα (pIκBα-DN; 0.1 μg) and empty vector (pSG5M) served as negative controls. All transfection mixtures were replenished with pSG5M to 1 μg of total DNA. After 48 h, cells were lysed and analyzed via immunoblot (upper panel) and luciferase assay (lower panel). Relative light units (RLUs) were normalized on protein concentration and on cells co-transfected with pSG5M (mock). The mean values ± SE of four independent experiments, each performed in triplicate, are shown.

Figure 1

Tax-1 and ELL2 co-localize in dot-like structures in the nucleus. (a–p) Indirect immunofluorescence analysis of ELL2, Tax-1 and the nucleus was performed. 293T cells were transfected with expression plasmids pEF-ELL2-myc, pEF-Tax-1 (1 µg each), both plasmids or the empty vector pEF-1α (mock). After 24 h, cells were stained with primary mouse anti-Tax and rabbit anti-ELL2 antibodies, followed by anti-mouse Alexa Fluor 647 and anti-rabbit Alexa Fluor 555 antibodies, respectively. Staining of nuclei was performed using Prolong Gold reagent with DAPI (4,6-diamidino-2-phenylindole). Images were generated using confocal laser scanning microscope Leica TCS SP5. Images of ELL2 (red), Tax-1 (green), the nucleus (blue), and the merge of all three stains are shown. Scale bars indicate 10 µm. Yellow dots indicate partial co-localizations of Tax-1 and ELL2. (p) A region of interest (ROI) is highlighted and (q) the graph shows the fluorescence intensities of Tax-1- and ELL2-specific fluorescence along the ROI.

Figure 2

Tax-1 and ELL2 co-precipitate independent of established Tax-1 point mutations affecting Tax-1 post translational modifications. (A) Schematic overview of Tax-WT (wildtype) and Tax-1 point mutants. The N-terminal nuclear localization signal (NLS, amino acids (aa) 1–50), the Tax-1 speckled structure localization signal (TSLS, aa 50–75) and the potential TRAF-binding motif (PTQRT) and its substitution to an ATART motif in the construct Tax-PQ (Tax P79A Q81A) are indicated. Lysine (K) residues K1-K10 are depicted in black, mutations to arginines (R) in red, and mutations from R to K (based on K1-10R) in green. Lysine residues being critical for ubiquitination and SUMOylation are marked on top. (B) 293T cells were transfected with expression plasmids pSG-Tax-WT (Tax-WT), pEF-ELL2-myc (ELL2-myc), pSG-Tax-His (Tax-His; 1 µg each) or the respective empty control vectors pSG5M and pEF (mock; 1 µg each). Each sample was supplemented with the respective empty control vector to a total amount of 2 µg of plasmid DNA if necessary. After 48 h, cells were lysed and 10% of the lysates were taken as input (IN) control. Co-immunoprecipitation (IP) was performed using anti-Tax antibodies or IgG control antibodies. One representative immunoblot out of five independent experiments is shown using antibodies targeting ELL2, Tax-1, β-Actin, and α-Tubulin. (C,D) 293T cells were transfected with expression plasmids pSG-Tax-WT, pSG-Tax-PQ (Tax-PQ), pSG-Tax-K1-10R (Tax-K1-10R), pSG-Tax-K4-8R (Tax-K4-8R), pSG-Tax-K6-8R (Tax-K6-8R), pSG-Tax-R7-8K (Tax-R7-8K), pSG-Tax-His (Tax-His), and pEF-ELL2-myc (ELL2-myc; 1 µg each). IPs were performed using (C) anti-ELL2 or IgG control antibodies or (D) anti-Tax antibodies. One representative out of (C) four or (D) five independent experiments is shown.

Figure 3

N-terminal (aa 1–37) and C-terminal (aa 150–353) domains of Tax-1 are critical for Tax-1:ELL2 complex formation. (A) Schematic representation of Tax-1-wildtype (WT) and six truncation mutants harboring deletions beginning from the N-terminus (TD1) to the C-terminus (TD319) of the Tax sequence. Numbers indicate amino acids of Tax-1. (B) 293T cells were transfected with 1 µg of the FLAG-tagged expression plasmids pCAG-FLAG-Tax-WT (Tax-WT), Tax truncations pCAG-FLAG-Tax-TD1 (Tax-TD1), pCAG-FLAG-Tax-TD55 (Tax-TD55), pCAG-FLAG-Tax-TD99 (Tax-TD99), pCAG-FLAG-Tax-TD150 (Tax-TD150), pCAG-FLAG-Tax-TD254 (Tax-TD254), and pCAG-FLAG-Tax-TD319 (Tax-TD319) together with pEF-ELL2-myc (ELL2-myc; 1 µg) or the respective empty control vectors pCAG-FLAG and pEF (1 µg each). After 48 h, cells were lysed, and 10% of the lysates were taken as input (IN) control. Co-immunoprecipitations (IPs) were performed using anti-ELL2 or IgG control antibodies. Representative immunoblots out of four independent experiments are shown using antibodies targeting ELL2, FLAG (for detection of Tax), and α-Tubulin. Densitometry was performed to quantitate the amount of co-precipitated FLAG-Tax (Tax) after precipitation of ELL2, and values were normalized on the respective ELL2 expression in the input (normalized on α-Tubulin). Binding of Tax-WT to ELL2 was set to 100%. Bars indicate the means of three independent experiments ± SE and values were compared to Tax-WT using Student’s t-test. * p < 0.05; ** p < 0.01; *** p < 0.001; n.d., not determined.

Figure 4

ELL2 interacts with Tax-1, but not with Tax-2B. (A) Schematic comparison of structural and functional domains of HTLV-1 Tax-1 and HTLV-2 Tax-2B. Numbers indicate amino acids of Tax. (B) 293T cells were transfected with expression plasmids pEF-ELL2-myc (ELL2-myc), pCAG-FLAG-Tax (FLAG-Tax-WT), and Tax-2B-FLAG (1 µg each) and supplemented with the respective empty control vectors pEF or pCAG-FLAG to a total amount of 2 μg of DNA, if necessary. After 48 h, cells were lysed, and 5% of the lysates were taken as input (IN) control. The co-immunoprecipitation (IP) was performed using anti-FLAG or IgG control antibodies. One representative immunoblot out of four independent experiments is shown using antibodies targeting ELL2, FLAG (for detection of Tax-1 and Tax-2), and GAPDH.

Figure 5

Characterization of ELL2 truncations and bioinformatic predictions of post translational modifications. (A) Schematic representation of ELL2 wildtype (ELL2-WT) and prediction of globular domains based on bioinformatic predictions using HHpred and IUPred. R1, R2, R3, conserved regions R1, R2, and R3; RNAPII, RNA polymerase II binding; ZO-1, homology to zonula occludens-1; Met186, alternative start codon. P Ser, phosphorylated serine residues. HHpred: Red bars indicate functionally related globular regions predicted through sequence comparisons of homologous, evolutionary related proteins of known 3D structure. IUPred: the prediction profile of protein disorder (green) tendency over the residue position. (B) Schematic representation of myc-his-tagged ELL2 truncations together with the respective predicted molecular weight (MW) values for proteins translated from the standard start codon (Met1, first value) and the alternative start codon at Met186 (second value), if applicable. (C) Expression analysis of ELL2 truncations. 293T cells were transfected with 1 µg of pEF-1α-driven expression plasmids ELL2-myc, N-myc, Met186-myc, C-myc, R1-myc, R2-myc, and R3-myc. After 48 h, cells were analyzed by Western blot using antibodies specific for Myc and the housekeeping gene GAPDH. * indicates bands corresponding to the predicted sizes (plus post translational modifications) of proteins translated from the first start codon being present.

Figure 6

Multiple domains in ELL2 except ELL2-R2 are important for Tax-1:ELL2 complex formation. Co-immunoprecipitations (IP) of (A) Tax-1 and ELL2 and (B) Tax-1 and ELL2 truncations using Tax-specific precipitation antibodies. 293T cells were transfected with expression plasmids pEF-1α-Tax-1 (Tax), pEF-1α-ELL2-myc (ELL2-myc) and ELL2-truncations together with the respective empty control vector. After 48 h, cells were lysed and 10% of the lysates were taken as input (IN) control. IP was performed using anti-Tax antibodies. (A,B) One representative immunoblot out of five independent experiments is shown using antibodies targeting ELL2, Tax-1, and α-Tubulin. (C) Densitometric analysis of three independent experiments +/− SE of co-precipitated ELL2 truncations relative to Tax-1 of the input, normalized on ELL2-WT. Mean values were compared using two-tailed Student’s t-test (** indicates p < 0.01; *** p < 0.001) (D) Schematic representation of ΔR2-myc together with the respective predicted molecular weight (MW) values for proteins translated from the standard start codon (Met1, first value) and the alternative start codon at Met186 (second value). (E) Co-immunoprecipitation of Tax-1, ELL2, and ΔR2 using Tax-specific precipitation antibodies. One representative out of three independent experiments is shown. IgG, isotype control antibody; HC, heavy chain antibodies; LC, light chain antibodies.

Figure 6

Multiple domains in ELL2 except ELL2-R2 are important for Tax-1:ELL2 complex formation. Co-immunoprecipitations (IP) of (A) Tax-1 and ELL2 and (B) Tax-1 and ELL2 truncations using Tax-specific precipitation antibodies. 293T cells were transfected with expression plasmids pEF-1α-Tax-1 (Tax), pEF-1α-ELL2-myc (ELL2-myc) and ELL2-truncations together with the respective empty control vector. After 48 h, cells were lysed and 10% of the lysates were taken as input (IN) control. IP was performed using anti-Tax antibodies. (A,B) One representative immunoblot out of five independent experiments is shown using antibodies targeting ELL2, Tax-1, and α-Tubulin. (C) Densitometric analysis of three independent experiments +/− SE of co-precipitated ELL2 truncations relative to Tax-1 of the input, normalized on ELL2-WT. Mean values were compared using two-tailed Student’s t-test (** indicates p < 0.01; *** p < 0.001) (D) Schematic representation of ΔR2-myc together with the respective predicted molecular weight (MW) values for proteins translated from the standard start codon (Met1, first value) and the alternative start codon at Met186 (second value). (E) Co-immunoprecipitation of Tax-1, ELL2, and ΔR2 using Tax-specific precipitation antibodies. One representative out of three independent experiments is shown. IgG, isotype control antibody; HC, heavy chain antibodies; LC, light chain antibodies.

Figure 7

ELL2-R1 is sufficient for enhancing Tax-1-mediated transactivation of the HTLV-1 promoter. (A) Luciferase reporter gene assays (upper part) and corresponding Western blots (lower part). 293T cells were transfected with expression constructs pEFneo-Tax-1 (Tax; 10 ng), pEF1α-ELL2-myc wt (ELL2-myc) or truncations (890 ng) together with the luciferase reporter construct pGL3-U3R of the HTLV-1 promotor (100 ng each) as indicated. Cells transfected with pGL3-Basic together with Tax and ELL2-myc served as negative control. At 48 h post transfection, luciferase activities were measured in triplicates and relative light units (RLUs) were normalized on protein content of the respective sample. Mean values ±SE are shown and were analyzed by Shapiro–Wilk test, and data pairs were compared using two-tailed Student’s t-test (n = 3; * p < 0.05; *** p < 0.001). Western blot to detect expression of ELL2 truncations (anti-myc), Tax-1 and α-Tubulin. Blots were cut due to technical resasons. (B) Luciferase reporter assays and Western blots were performed as described in (A). 293T cells were transfected with expression constructs pEFneo-Tax-1 (Tax; 10 ng), pEF1α-ELL2-myc wt (ELL2-myc; 890 ng) or decreasing amounts of R2-myc (890 ng, 400 ng, 200 ng, and 100 ng) together with the luciferase reporter construct pGL3-U3R of the HTLV-1 promotor (100 ng each). Mean values ± SE are shown and were analyzed by Shapiro–Wilk test, and data pairs were compared using two-tailed Student’s t-test (n = 4; * p < 0.05; ** p < 0.01; n.s., not significant).