In vivo requirement of activator-specific binding targets of mediator

Abstract

There has been no unequivocal demonstration that the activator binding targets identified in vitro play a key role in transcriptional activation in vivo. To examine whether activator-Mediator interactions are required for gene transcription under physiological conditions, we performed functional analyses with Mediator components that interact specifically with natural yeast activators. Different activators interact with Mediator via distinct binding targets. Deletion of a distinct activator binding region of Mediator completely compromised gene activation in vivo by some, but not all, transcriptional activators. These demonstrate that the activator-specific targets in Mediator are essential for transcriptional activation in living cells, but their requirement was affected by the nature of the activator-DNA interaction and the existence of a postrecruitment activation process.

FIG. 1

Gal11 module of Mediator is required for activator binding and transcriptional activation. (A) GST pulldown assay of wild-type hpol II (lanes 1 to 6) and hrs1Δ hpol II (lanes 7 to 12). Purified hpol II in the Mono-Q fraction (19) was incubated with GST beads containing the GST-activator fusion proteins indicated at the top of each lane. As controls, GST alone (lanes 2 and 8) and GST fused to an activation-defective version of VP16 (VP16^Δ456FP442; lanes 3 and 9) were used. Bead-bound proteins, along with the proteins used in the binding reactions (input; lanes 1 and 7), were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequent immunoblotting with the Mediator antibodies indicated at the left of the panel. (B) The transcriptional activity of wild-type and hrs1 null (hrs1Δ) hpol II were analyzed in an in vitro transcription system reconstituted with purified transcription factors. The presence (+) or absence (−) of activator proteins is indicated at the top of the panel. The transcripts from the G-less templates containing either the Gal4 (Gal4:G−) or Gcn4 (Gcn4:G−) DNA-binding sites are indicated by arrows at the right of the panel. The fold activation of hpol II by each activator is shown at the bottom of the panel. (C) Far-Western blot analysis of activator-binding targets. Affinity-purified hpol II blotted on nitrocellulose membrane after resolution by SDS-PAGE was probed with radiolabeled VP16 (V) or Gcn4 (C) protein (left panel). The labeled protein bands were visualized by reprobing the same blot with anti-Gal11 and anti-Hrs1 antibodies (right panel). The apparent molecular sizes (in kilodaltons) are shown in the middle. (D) GST pulldown assay of Gal11 module proteins. Purified HA-Gal11 (G), HA-Med2 (M), and HA-Hrs1 (H) proteins were incubated with the bead-bound GST-activator fusion proteins indicated at the top of the panel. After washing away the unbound proteins, the proteins that remained bound to the beads were visualized by immunoblotting with anti-HA antibody. The protein bands are indicated with arrows at the left of the panel. The largest band in the G lanes is the full-length HA-Gal11 fusion protein, and the smaller bands are degradation products of the parent protein.

FIG. 2

GST-Gal11 pulldown assay with activator proteins Gal4, Gcn4, and VP16. (A) Radiolabeled activators indicated at the left-hand side of the panel were incubated with GST beads containing the Gal11 fragments indicated at the top of each lane. Fragments of Gal11 polypeptide fused to GST are represented as solid bars, and the Gal11 amino acid residues contained in each fragment are indicated. The proteins bound to the beads were visualized by SDS-PAGE and autoradiography. The regions identified to interact with activator proteins are labeled G11-A to G11-D. (B) GST pulldown assays with GST beads fused to a series of Gal11 derivatives that contained various internal deletions or with GST beads fused to the Gal11 fragments indicated at the top of each lane as in panel A. The parts of the amino acid residues deleted from the Gal11(1-864) fragment (lanes 3 to 6) are indicated by numbers in parentheses following the Δ.

FIG. 3

Requirement for the VP16-interacting region of Gal11 for transcriptional activation in vivo. (A) In vivo analysis of gal11 mutants for VP16 activation. The GAL11 derivatives introduced into the gal11 null strain are labeled [wild type, null, and Δ(176-262)]. Transcriptional activation of the lacZ reporter gene containing LexA-binding sites by LexA-VP16 in each strain is shown at the right of each allele. In this and the other figures in this paper, reporter gene expression level is given in units of β-galactosidase activity. The percent transcriptional activation in mutants compared to that in the wild type is shown in parentheses. (B) Composition of hpol II purified from the gal11Δ(176-262) mutant. Shown are the immunoblot analyses of hpol II immunopurified from wild-type and gal11Δ(176-262) mutant cells with the antibodies indicated at the left.

FIG. 4

Requirement for the Gal4-interacting regions of Gal11 for transcriptional activation in vivo. (A) Growth of wild-type and gal11 mutant yeast strains on galactose medium. Wild-type (GAL11), gal11 null (gal11Δ), and two internal deletion [Δ(48-618) and Δ(176-262)] mutant yeast strains grown on YP-galactose at 30°C for 72 h are shown. (B) Transcriptional activity of gal11 mutants under Gal4 induction conditions. The GAL11 derivatives introduced into the gal11 null strain are labeled, and the transcriptional activation of the lacZ reporter gene containing Gal4 binding sites in each strain is shown as in Fig. 3A. (C) Comparison of the transcriptional activation abilities and activator-binding efficiencies of wild-type GAL11 (1 to 1081) and a series of internal deletion mutants [Δ(176-262), Δ(48-326), and Δ(48-618)]. (D) Composition of wild-type hpol II and hpol II purified from the gal11Δ(48-618) mutant yeast strains. hpol II was immunopurified from wild-type and gal11Δ(48-618) mutant yeast strains, and their compositions were compared by immunoblot analysis with the antibodies indicated at the left. The internal deletion mutant of gal11 produced a smaller Gal11 protein derivative than did wild-type GAL11. Sizes are shown in kilodaltons.

FIG. 5

Mediator-binding specificity of Gcn4. (A) Mapping of Gcn4-binding regions within Hrs1. Radiolabeled Gcn4 or Gal4 proteins were incubated with GST beads containing the GST-Hrs1 fragments indicated at the top of each lane. Fragments of Hrs1 polypeptide fused to GST are represented as solid bars, and the Hrs1 amino acid residues contained in each fragment are indicated. The proteins bound to the beads were visualized by SDS-PAGE and autoradiography. The regions identified to interact with the transcriptional activator proteins Gcn4 and Gal4 are labeled H1-A and H1-B. (B) Correlation between Mediator binding affinity and transcriptional activation potency of Gcn4 derivatives. Purified hpol II or recombinant HA-Gal11 or HA-Hrs1 proteins were analyzed by GST pulldown assays with GST beads fused to a series of Gcn4 mutant protein derivatives that contained single (lanes 3 to 5), double (lanes 6 to 8), or triple (lane 9) mutations in the hydrophobic clusters (HC) (16). Immunoblot analysis of the bound proteins with the antibodies indicated at the right of the panel is shown.

FIG. 6

Requirement for Gcn4-interacting regions of Mediator for transcriptional activation in vivo. (A) In vivo analysis of the gal11 mutants for Gcn4 activation. Transcriptional activation of the lacZ reporter gene controlled by a natural HIS4 promoter or a synthetic promoter containing Gal4 binding sites under the respective activation conditions is shown; amino acid starvation or expression of exogenous Gcn4 protein fused to a Gal4 DNA-binding domain (Gal4DBD-Gcn4) is used for transcriptional activation of each reporter gene. The percent transcriptional activation in mutants compared to the wild type is shown in parentheses. (B) In vivo analysis of the effects of hrs1 mutations on transcriptional activation by Gcn4. The HRS1 derivatives and their transcriptional activities are shown as described for panel A except that a synthetic promoter containing Gcn4-binding sites was used instead of the natural HIS4 promoter. (C) Composition of hpol II purified from hrs1 deletion mutants. Immunoblot analysis of hpol II purified from wild-type, hrs1 null mutant (hrs1Δ), hrs1Δ(2-82) (deletion of H1-A), and hrs1Δ(180-343) (deletion of H1-B) strains was carried out with antibodies against the hpol II subunits indicated at the left.

FIG. 7

Chromatin immunoprecipitation analysis of the Gal11 module-deficient mutant. The occupancy of wild-type (wt) and Gal11 module-deficient (Δ) hpol II at active or inactive promoters is shown by assay with antibodies to Rgr1. Growth of cells on rich medium containing galactose (YPGal) induces GAL1 transcription but represses the HIS4 promoter. On the other hand, growth of yeast cells on synthetic medium containing dextrose and limited amino acids (SDex) activates HIS4 transcription but represses the GAL1 promoter. As a control, hpol II occupancy at the constitutively active actin promoter was analyzed in parallel. PCR amplification of the promoter regions before (Total; lanes 1, 3, 5, and 7) and after (Precipitate; lanes 2, 4, 6, and 8) the immunoprecipitation is shown.