Human mediator enhances activator-facilitated recruitment of RNA polymerase II and promoter recognition by TATA-binding protein (TBP) independently of TBP-associated factors

Abstract

Mediator is a general cofactor implicated in the functions of many transcriptional activators. Although Mediator with different protein compositions has been isolated, it remains unclear how Mediator facilitates activator-dependent transcription, independent of its general stimulation of basal transcription. To define the mechanisms of Mediator function, we isolated two forms of human Mediator complexes (Mediator-P.5 and Mediator-P.85) and demonstrated that Mediator-P.5 clearly functions by enhancing activator-mediated recruitment of RNA polymerase II (pol II), whereas Mediator-P.85 works mainly by stimulating overall basal transcription. The coactivator function of Mediator-P.5 was not impaired when TATA-binding protein (TBP) was used in place of TFIID, but it was abolished when another general cofactor, PC4, was omitted from the reaction or when Mediator-P.5 was added after pol II entry into the preinitiation complex. Moreover, Mediator- P.5 is able to enhance TBP binding to the TATA box in an activator-dependent manner. Our data provides biochemical evidence that Mediator functions by facilitating activator-mediated recruitment of pol II and also promoter recognition by TBP, both of which can occur in the absence of TBP-associated factors in TFIID.

FIG. 1.

Isolation of two human Mediator complexes. (A) Detection of epitope-tagged human Med7 in P11 fractions. Nuclear extracts (NE) from HeLa-derived hMED7-7 cells, which express FLAG-tagged human Med7 (f:hMed7), were fractionated over a phosphocellulose P11 ion-exchange column and step eluted with BC buffer (12) containing different concentrations of KCl as indicated. Western blotting was performed with anti-FLAG M2 monoclonal antibody (Sigma). (B) Silver staining of purified human Mediator complexes. Mediator-P.5 and Mediator-P.85 were purified, respectively, from P11 0.5 M (P.5) and 0.85 M (P.85) KCl fractions of hMED7-7 nuclear extracts. Control purification was conducted in parallel with similar fractions derived from HeLa cells. Protein size markers (in kilodaltons) are indicated on the left. The assignment of Mediator components was based on comparison of various Mediator complexes purified from different laboratories with a newly proposed unifying nomenclature (77) and further confirmed by Western blotting with available antibodies against TRAP240/Med240, TRAP230/Med230, TRAP220/Med220, TRAP170/CRSP150/Med150, CRSP130/Med130, ARC105/Med105, TRAP100/Med100, TRAP80/Med78, cdk8, and cyclin C (data not shown). Polypeptides confirmed by Western blotting are indicated by solid arrows, whereas proteins corresponding to previously characterized Mediator components based on the migration patterns are pointed by open arrows with dashed lines. Positions of cyclin C, cdk8, and Med70 are depicted on the gel by open arrowheads.

FIG. 2.

The coactivator function of Mediator depends on the presence of pol II, TFIIH, and PC4. (A) Immunodepletion of pol II from Mediator-P.5 and Mediator-P.85. Anti-pol II CTD antibodies (8WG16) were used to remove pol II from purified Mediator complexes. The amounts of pol II and Mediator remaining after two rounds (1x and 2x) of immunodepletion (depl) were revealed by Western blotting with anti-CTD 8WG16 (for RPB1 detection) and anti-FLAG M2 (for f:hMed7 detection) antibodies, respectively. Purified FLAG-tagged human pol II (9.6 nM, in which 0.75 μl is normally used for one transcription reaction) was loaded in parallel for evaluating the efficiency of immunodepletion. (B) Mediator activity is dependent on exogenous pol II, TFIIH, and PC4. In vitro transcription was reconstituted with recombinant TFIIB, TFIIE, TFIIF, PC4, Gal4-VP16 (VP16), and epitope-tagged multiprotein complexes (TFIID, TFIIH, and pol II) in the presence (+) of Mediator-P.5 or Mediator-P.85. Individual components (pol II, TFIIH, or PC4) were then left out (−) from the reactions as indicated. The pG₅MLT template contains five Gal4-binding sites linked to the adenovirus major late core promoter preceding a G-less cassette of approximately 380 nucleotides. The pMLΔ53 template is devoid of activator-binding sites but contains the same adenovirus major late core promoter linked to a shorter G-less cassette (∼280 nucleotides).

FIG. 3.

Mediator-P.5 enhances activation mediated by Gal4-VP16, whereas Mediator-P.85 stimulates overall basal transcription. In vitro transcription was performed with individually purified transcription components (TFIIB, TFIID, TFIIE, TFIIF, TFIIH, pol II, and PC4) in the absence (−) or presence of Gal4-VP16 (VP16), E2, Mediator-P.5, or Mediator-P.85, as indicated. Fold activation is defined as the signal intensity, quantified by PhosphorImager (Amersham Biosciences), from each activator-binding site-containing template relative to that from the same DNA template performed in the absence of activator and Mediator (i.e., lane 1). The p2E2(IR)Δ53 template contains two human papillomavirus E2-binding sites linked to the same core promoter with a shorter G-less cassette (∼280 nucleotides).

FIG. 4.

Mediator-P.5 acts by enhancing activator-facilitated recruitment of pol II to the preinitiation complex. (A) Diagram of order-of-addition and template challenge experiments. (B) Mediator-P.5 enhances pol II recruitment by Gal4-VP16. In vitro transcription was performed as outlined in panel A by preincubating pG₅MLT and p2E2(IR)Δ53 DNA templates with TFIID (D), together with TFIIB (B) and pol II (II), or the rest (All) of general transcription factors (GTFs), in the absence (−) or presence (+) of PC4, activator, or Mediator-P.5, at 30°C for 20 min. The remaining transcription components and ribonucleoside triphosphates (NTPs) were then added, together with a 10-fold excess of pΔMLP challenge template, which contains only the adenovirus major late core promoter linked to a G-less cassette of approximately 200 nucleotides, to initiate transcription. The reaction was continued at 30°C for 60 min and processed as described in Materials and Methods. Fold activation is defined as the signal intensity relative to that detected in lane 1 for reactions performed in the absence of Mediator-P.5 and in lane 10 for reactions performed in the presence of Mediator-P.5.

FIG. 5.

The coactivator function of Mediator-P.5 was not impaired following the removal of cdk8-associated polypeptides. (A) Immunodepletion of cdk8-associated polypeptides from Mediator-P.5. Anti-cdk8 (α-cdk8) antibodies were used to remove cdk8-associated polypeptides from Mediator-P.5. The input (InP), flowthrough (FT), and bound Mediator complexes were analyzed by silver staining (top panels) and Western blotting (bottom panels) with the indicated antibodies. Positions of cyclin C and cdk8 are indicated by solid arrowheads. Cross-linked heavy- and light-chain immunoglobulin bands are depicted by asterisks. (B) Coactivator function of Mediator-P.5 before and after cdk8 depletion. In vitro transcription was performed as outlined in the legend for Fig. 4A, except that pMLΔ53 template was used in the absence (−) or presence (+) of Mediator-P.5 without or with (α-cdk8) immunodepletion with anti-cdk8 antibodies. Fold activation in each pair of reactions is defined as the signal increase from pG₅MLT performed in the presence of Gal4-VP16 relative to that in its absence.

FIG. 6.

The coactivator function of Mediator-P.5 can occur in the presence or absence of TAFs. In vitro transcription was performed as outlined in the legend for Fig. 4A, with TFIID (D) or TBP (T) as the TATA-binding factor, without (−) or with Mediator-P.5 added either during (+) or after (aft) preincubation. Fold activation in each set of reactions is defined as the signal intensity from pG₅MLT relative to that performed in the absence of Gal4-VP16 and Mediator-P.5 (i.e., the first lane of each reaction set).

FIG. 7.

Activator-facilitated recruitment of pol II to the promoter region is enhanced in the presence of Mediator-P.5. (A) Activator-dependent recruitment of Mediator to the promoter region. In vitro transcription and Western blotting were conducted with immobilized pG₅MLT template in the absence (−) or presence (+) of Mediator-P.5 and Gal4-VP16, as described in Materials and Methods. Antibodies used for Western blotting analysis are indicated on the right of individual strips. (B) Enhancement of activator-facilitated recruitment of pol II by Mediator can occur in the presence or absence of TAFs. In vitro transcription and Western blotting were conducted with immobilized pG₅MLT template in the absence (−) or presence (+) of Mediator-P.5 and Gal4-VP16, as described in the legend for panel A. (C) Assembly of transcription complexes can occur in the absence of a functional TATA box. Immobilized template assays were performed with pG₅MLT (TATAwt) or pG₅MLT-mutTATA (TATAmt) as described in the legend for panel A.

FIG. 8.

The coactivator function of Mediator-P.5 and PC4 was eliminated after preincubation of DNA templates with TFIID, but not with TBP. In vitro transcription was performed as outlined, without (−) or with Mediator-P.5 and PC4 added either during (+) or after (aft) preincubation.

FIG. 9.

Mediator-P.5 enhances promoter recognition by TBP in an activator-dependent manner. (A) DNase I footprinting performed with a high concentration of TBP. DNase I footprinting was performed with a ³²P-labeled DNA fragment containing five Gal4-binding sites linked to the adenovirus major late promoter preceding a 380-nucleotide G-less cassette in the absence (−) or presence (+) of Mediator- P.5, TBP, and Gal4-VP16 (VP16), as described in Materials and Methods. The protected regions surrounding the TATA box and the Gal4-binding site are bracketed with thick and thin lines, respectively. (B) DNase I footprinting performed with a low concentration of TBP. The footprinting assay was performed as described in the legend for panel A, except using 0.18 nM DNA, 2.8 nM TBP, 10 nM Gal4-VP16, and 0.005 U of DNase I.