Transcriptional intermediary factor 1alpha mediates physical interaction and functional synergy between the coactivator-associated arginine methyltransferase 1 and glucocorticoid receptor-interacting protein 1 nuclear receptor coactivators

Abstract

In previous studies transcriptional intermediary factor 1alpha (TIF1alpha) was identified as a direct binding partner and potential transcriptional coactivator for nuclear receptors (NRs) but its overexpression inhibited, rather than enhanced, transcriptional activation by NRs. Here we show that TIF1alpha bound to and enhanced the function of the C-terminal activation domain (AD) of coactivator associated arginine methyltransferase 1 (CARM1) and the N-terminal AD of glucocorticoid receptor-interacting protein 1 (GRIP1). Furthermore, although TIF1alpha had little or no NR coactivator activity by itself, it cooperated synergistically with GRIP1 and CARM1 to enhance NR-mediated transcription. Inhibition of endogenous TIF1alpha expression reduced transcriptional activation by the GRIP1 N-terminal domain but not by the CARM1 C-terminal domain, suggesting that TIF1alpha may be more important for mediating the activity of the former than the latter. Reduction of endogenous TIF1alpha levels also compromised the androgen-dependent induction of an endogenous target gene of the androgen receptor. Finally, TIF1alpha formed a ternary complex with the GRIP1 N-terminal and CARM1 C-terminal domains. Thus, we conclude that TIF1alpha cooperates with NR coactivators GRIP1 and CARM1 by forming a stable ternary complex with them and enhancing the AD function of one or both of them.

Fig. 1. TIF1α stimulates transcriptional activation by the CARM1-C activation domain

A, Domains of TIF1α are shown, along with the fragment identified by yeast two-hybrid (Y2H) screen (amino acids 193–607). B, CV-1 cells were transfected with 250 ng of GK1 reporter plasmid; 125 ng of pM vector encoding Gal4 DBD or Gal4 DBD fused to CARM1 full length (C1), CARM1-N (C1N, amino acids 3–460), or CARM1-C (C1C, amino acids 461–608); and 400 ng of pSG5 empty vector (white bars) or pSG5.TIF1α (black bars). Luciferase activity was measured 48 h after transfection. Each data point represents the mean and range of variation of two transfected cell cultures. Results shown are from a single experiment, which is representative of four independent experiments.

Fig. 2. Transcriptional synergy between the three coactivators: GRIP1, CARM1, and TIF1α

CV-1 cells were transiently transfected with 250 ng of luciferase reporter plasmid controlled by an appropriate hormone response element; expression vector for GR (0.1 ng), ER (0.1 ng), TR (0.1 ng), or AR (0.5 ng); and expression vectors for GRIP1 (50 ng), CARM1 (250 ng), and TIF1α (250 ng), as indicated. Transfected cells were grown without hormone (white bars in panel D) or with (black bars in panels A-D) 20 nM dex for GR, E2 for ER, T3 for TR, or DHT for AR; after 48 h cell extracts were assayed for luciferase activity. Each data point represents the mean and range of variation of two transfected cell cultures. Results shown are from a single experiment, which is representative of four separate experiments for GR, six experiments for ER, three experiments for TR, and two experiments for AR.

Fig. 3. Coactivator domains required for GRIP1-CARM1-TIF1α synergy

A, CV-1 cells were transiently transfected with MMTV(ERE)-LUC reporter plasmid (250 ng) and expression vectors encoding ER (0.01 ng), GRIP1 (50 ng), CARM1 (100 ng), and TIF1α wild type or L/A mutant (100 ng), as indicated. B, CV-1 cells were transiently transfected with MMTV-LUC reporter plasmid (250 ng) and expression vectors encoding GR (0.01 ng), CARM1 (200 ng), TIF1α (200 ng), and GRIP1 wild type or mutants GRIP1ΔN (deletion of amino acids 5–563) or GRIP1ΔC (lacking amino acids 1161–1462) (25 ng). C, CV-1 cells were transiently transfected with MMTV-LUC reporter plasmid (250 ng) and expression vectors encoding GR (0.01 ng), GRIP1 (25 ng), TIF1α (200 ng), and CARM1 wild type, CARM1-N (amino acids 3–460), or CARM1-C (amino acids 461–608) (200 ng). D, CV-1 cells were transiently transfected with MMTV(ERE)-LUC reporter plasmid (250 ng) and expression vectors encoding ER (0.01 ng), GRIP1 (125 ng), TIF1α (250 ng), and CARM1 wild type or CARM1 methyltransferase deficient mutants E267Q or VLD (250 ng). In panels A–D, cells were grown for 48 h with dex or E2. Each data point represents the mean and range of variation of two transfected cell cultures. The results presented are from a single experiment representative of three independent experiments.

Fig. 4. TIF1α interacts with the GRIP1 N-terminal activation domain AD3

A, CV-1 cells were transfected with 250 ng of GK1 reporter plasmid; 125 ng of pM vector encoding Gal4 DBD or Gal4 DBD fused to GRIP1 full length (G1), GRIP1-N (G1N, amino acids 5–765), or GRIP1-C (G1C, amino acids 1121–1462); and 400 ng of pSG5 empty vector (white bars) or pSG5.TIF1α (black bars). Luciferase results shown are the mean and range of variation of two transfected cell cultures and are from a single experiment representative of three independent experiments. B, COS7 cells were transfected with expression vectors for HA-tagged GRIP1-N (amino acids 5–765) (2 μg) and TIF1α (3 μg), as indicated. Coimmunoprecipitation (IP) and immunoblots (WB) were conducted with the indicated antibodies.

Fig. 5. TIF1α stabilizes a complex containing GRIP1-N and CARM1-C

A, CV-1 cells were transfected with 250 ng of GK1 reporter plasmid, 125 ng of pM vector encoding Gal4DBD or Gal4DBD fused to GRIP1-N (amino acids 5–765), 125 ng of pVP16 vector encoding VP16 AD or VP16-CARM1-C (amino acids 461–608), and 400 ng of pSG5 empty vector (white bars) or pSG5.TIF1α (black bars). Each data point represents the mean and range of variation of two transfected cell cultures. The results presented are from a single experiment representative of three independent experiments. B, COS7 cells were transfected with expression vectors for HA-tagged GRIP1-N (amino acids 5–479) (1 μg), HA-tagged CARM1-C (amino acids 461–608) (1 μg), and TIF1α (3 μg), as indicated. Coimmunoprecipitation (IP) and immunoblots (WB) were conducted with the indicated antibodies. C, Model for a functional ternary complex between GRIP1, CARM1 and TIF1α. N and C within protein diagrams represent N-terminal and C-terminal domains. NRs bind to a hormone response element (HRE) and recruit coactivators. GRIP1, CARM1, and TIF1α form a ternary complex. GRIP1 binds to NRs with its LXXLL motifs, to TIF1α with its N-terminal AD3, and to CARM1 with its C-terminal AD2. CARM1 binds to GRIP1 with its central methyltransferase domain and to TIF1α with its C-terminal AD. TIF1α binds to CARM1-C and to GRIP1-N using unknown domains, and it may interact with NRs through its LXXLL motif. Some coactivators contribute to the assembly of the RNA polymerase II transcription initiation complex (Pol II TIC) by catalyzing post-translational modifications of histones and other proteins, using cofactors S-adenosylmethionine (SAM) or acetyl CoA (AcCoA). Other coactivators contribute through direct or indirect protein-protein interactions with the transcription machinery.

Fig. 6. Requirement of endogenous TIF1α for mediating transcriptional activation by GRIP1-N, CARM1-C, and AR

A, COS7 cells in 24-well plates were transfected with no siRNA (white bars), 90 pmol of the TIF1α-specific siRNA duplex (black bars), or 90 pmol of scrambled-sequence control siRNA duplex (gray bars). After one day, cells were transfected with pG5-Luc reporter plasmid (200 ng) and expression vector encoding Gal4DBD-GRIP1-N (400 ng) or Gal4DBD-CARM1-C (100 ng). Luciferase activity was quantified 24 h after DNA transfection. Each data point represents the mean and range of variation of three transfected cell cultures. The results presented are from a single experiment representative of three independent experiments. Top panel. Immunoblots of extracts from COS7 cells subjected to the same siRNAs treatment and analyzed with anti-TIF1α and anti-β-actin antibodies. B, LNCaP cells were transfected with the indicated amounts of siRNA against TIF1α or control scrambled sequence siRNA. After 72 h cells were treated with DHT and then harvested after an additional 24 h. Total RNA was used for reverse transcription, and the resulting cDNA was analyzed by real-time PCR to measure levels of β-actin, TIF1α, and PSA mRNA. Each sample was run in duplicate, and average CT values (with range of variation) for TIF1α and PSA mRNA were normalized to that of β-actin. Results shown are from a single experiment, which is representative of four independent experiments.