Role of Mediator in transcription of the methionine biosynthesis genes. (A) Phenotypic analyses. Wild-type, med1Δ, med2Δ, med3Δ, med9Δ, med5Δ (nut1Δ), med20Δ (srb2Δ), med18Δ (srb5Δ), med14-100 (rgr1-100), med15Δ (gal11Δ), med16Δ (sin4Δ), and met4Δ strains were grown to exponential phase, collected, washed, and suspended in water. A serial fivefold dilution was spotted on B medium containing 0.05 mM ammonium sulfate (SO₄²⁻) or 0.05 mM l-methionine (Met). The most concentrated spot corresponds to a dilution at a density of 0.2 × 10⁷ cells/ml. Plates were incubated at 28°C for 3 days. (B) Analysis of mRNA levels for selected Met4 target genes in strains used in the experiment shown in panel A. Cells were grown in YPD medium, collected, washed, and incubated for 60 min into B medium. RNA levels were quantified by reverse transcription, followed by real-time PCR, and normalized with U4 snRNA. Values represent the average of two independent experiments, and error bars indicate standard deviations. (C) Analysis of mRNA levels in cells containing temperature-sensitive mutations in Med6, Med17 (Srb4), and Med11. Strains containing wild-type or temperature-sensitive alleles of MED6 (40), MED17 (SRB4) (23), and MED11 (21) were grown in YPD medium at a permissive temperature (25°C), shifted at a restrictive temperature (38°C) for 90 min, collected, washed, and incubated in B medium at 38°C for 30 min. RNA levels were quantified as described in panel B, except that 25S rRNA was used for normalization. Values represent the average of two independent experiments. Error bars indicate standard deviations. (D) Summary of the effects of Mediator mutations on MET gene transcription. The model of topological organization of Mediator is adapted from reference 20. The Mediator subunits analyzed in this analysis are marked in boldface type. Subunits essential for viability are underlined.

Association of Mediator with selected Met4-activated genes. (A) Med14 (Rgr1), TBP, and Pol II association. ChIP was performed on a strain expressing TAP-tagged Med14 and HA₃TBP using IgG to immunoprecipitate Med14-TAP and antibodies to HA and Pol II CTD. Cells were grown in B medium containing 0.05 mM l-methionine until early exponential phase, filtered, washed, and transferred into B medium. After 1-h incubation (activating conditions [A]), half of the culture was cross-linked with formaldehyde and the other half was supplemented with 1 mM l-methionine and incubated an additional 40 min (repressing conditions [R]) prior to formaldehyde addition. IPs were quantified with PCR primers for the indicated MET promoters and the IME2 or POL1 ORF as a control. Values represent the average of two independent experiments. Error bars indicate standard deviations. (B and C) Med5 (Nut1) and Med18 (Srb5) association, respectively. ChIP was performed on strains expressing TAP-tagged Med5 or Med18 and HA₃TBP as described in panel A. The results for TBP and Pol II were essentially the same as those shown in panel A (data not shown). (D) Med2, Med14 (Rgr1), Med10, Med21 (Srb7), Med6, and Med19 (Rox3) association. ChIP was performed on strains expressing TAP-tagged Mediator subunits and HA₃TBP as described in the legend to panel A, except that the entire culture was cross-linked after a 1-h incubation in B medium. IPs were quantified using PCR primers for the CYS3 and MET2 promoters and the POL1 ORF as a control. Values represent the average of two independent experiments. Error bars indicate standard deviations.

Role of SAGA in transcription of Met4-activated genes. (A) Analysis of mRNA levels in cells lacking Gcn5, Spt3, and Spt20. Wild-type, gcn5Δ, spt3Δ, and spt20Δ strains were grown in YPD medium, collected, washed, and incubated for 60 min into B medium. RNA levels were quantified by reverse transcription-PCR and normalized using 25S rRNA. Values represent the average of two independent experiments, and error bars indicate standard deviations. (B) Gcn5, Spt3, and Spt20 association with Met4 target promoters under activating and repressing conditions. ChIP was performed on strains expressing HA-tagged Gcn5, Spt3, or Spt20 as described in the legend to Fig. 3A. IPs were quantified using PCR primers for the indicated MET promoters and the IME2 ORF, as well as the GAL1 promoter as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations. (C) Gcn5 association with Met4 target promoters in WT and met4Δ strains. ChIP was carried out on WT and met4Δ strains expressing HA-tagged Gcn5 using antibodies to HA and PCR primers for the indicated Met4 target promoters and the POL1 ORF as a control. Cells were grown in YPD medium, collected, washed, and incubated for 60 min into B medium prior to cross-linking. Values represent the average of two independent experiments, and error bars indicate standard deviations.

Functional relationship between Mediator and SAGA in Met4-activated transcription. (A) Med14 (Rgr1), Pol II, and Met4 association with MET promoters in cells lacking Gcn5 and Spt20. ChIP of wild-type, gcn5Δ, and spt20Δ strains expressing Med14-TAP using IgG to immunoprecipitate Med14-TAP and antibodies to Pol II CTD and Met4 was carried out. Cells were grown in YPD medium, collected, washed, and incubated for 60 min into B medium prior to cross-linking. IPs were quantified using PCR primers for MET2 and MET17 promoters and the POL1 ORF as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations. (B) Gcn5, Med14 (Rgr1), Pol II, and Met4 association with MET promoters in cells lacking Med2 and Med3. ChIP of wild-type, med2Δ, and med3Δ cells expressing Med14-TAP and Gcn5-HA₃ using IgG to immunoprecipitate Med14-TAP, antibodies to HA to immunoprecipitate Gcn5-HA₃, and antibodies to Pol II CTD and Met4 was carried out. Cells were grown in YPD medium, collected, washed, and incubated for 60 min into B medium prior to cross-linking. IPs were quantified using PCR primers for MET2 and MET17 promoters and the POL1 ORF as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations. (C) Spt3 association with MET promoters in cells lacking Med2 and Med3. ChIP of wild-type, med2Δ, and med3Δ cells expressing Spt3-HA₃ using antibodies to HA was carried out. Cells were grown in YPD medium, collected, washed, and incubated for 60 min into B medium prior to cross-linking. IPs were quantified using PCR primers for MET2 and MET17 promoters and the POL1 ORF as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations.

Model for Mediator and SAGA interplay in transcriptional activation. Like Met4, Gal4 and Gcn4 are able to direct Mediator and SAGA recruitment to their target promoters (see the text for references). Recruitments of SAGA and Mediator are independent at Met4 target promoters and interdependent at Gcn4 target promoters. In contrast, at Gal4 target genes, optimal Mediator recruitment requires SAGA but SAGA recruitment does not require Mediator (see the text for details and references). The function of SAGA involves acetylation of histone H3 at Gcn4 (30) and Met4 target promoters and recruitment of TBP via interactions with Spt3/Spt8 at Gal4 target promoters (16, 38, 54). Mediator functions primarily by promoting recruitment of Pol II to the promoter, but roles at subsequent steps are not excluded.

Association of Pol II and Met4 with MET genes in Mediator mutants. ChIP was performed on wild-type, med3Δ, med14-100 (rgr1-100), and med15Δ (gal11Δ) using antibodies against Pol II CTD or Met4 and PCR primers for the indicated Met4 target promoters and the IME2 open reading frame (ORF) as a control. Cells were grown as described in the legend to Fig. 1B. Values represent the average of two independent experiments. Error bars indicate standard deviations.

Mediator is recruited through Met4 activation domain. (A) Localization of Med14 (Rgr1), Met4, and TBP along the MET2 promoter. (Top) Schematic of MET2 representing Cbf1- and Met31/32-binding sites (gray boxes), TATA element (black box), and open reading frame (open rectangle). (Bottom) ChIP on a strain expressing TAP-tagged Med14 and HA₃TBP using IgG to immunoprecipitate Med14 and antibodies to Met4 and HA. Cells were grown as described in the legend to Fig. 1B. IPs were quantified using PCR primers for fragment A, fragment B, and the POL1 ORF as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations. (B) Association of Med14 (Rgr1) with selected Met4 target genes in wild-type (WT) and met4Δ strains. ChIP was performed as described in the legend to panel A on WT and met4Δ strains expressing TAP-tagged Med14. IPs were quantified using PCR primers for the indicated MET promoters and the POL1 ORF as a control. Values represent the average of two independent experiments, and error bars indicate standard deviations. (C) A met4Δ strain expressing TAP-tagged Med14 (Rgr1) and bearing a GAL1-lexAop-lacZ chimeric gene (which contains four LexA operators in place of GAL1 UAS) integrated at the URA3 locus was transformed with a plasmid expressing the LexA DNA-binding domain fused to either wild-type Met4 or a derivative lacking the activation domain (Met4Δ12, deleted for residues 79 to 180). As a control, the strain was also transformed with an empty plasmid (pRS313). Cells were grown in YNB medium, collected, washed, incubated for 60 min into B medium, and subjected to ChIP with IgG to immunoprecipitate Med14 and antibodies against Met4. IPs were quantified using PCR primers for the GAL1-lexAop-lacZ promoter. Values represent the average of two independent experiments, and error bars indicate standard deviations.

Levels of histone acetylation at Met4-activated genes. (A) Effect of Gcn5 inactivation on histone H3 acetylation at MET17 and MET2. ChIP of wild-type and gcn5Δ strains using antibodies specific for histone H3 acetylated at K9 or K14 and PCR primers for the 5′ end of MET17 and MET2 ORF was carried out. Cells were grown in YPD medium, filtered, and suspended into B medium. Aliquots were cross-linked at different times. Values represent the average of two independent experiments, and error bars indicate standard deviations. (B) Status of histone H3 at MET genes upon induction. Conditions were the same as described in the legend to panel A, except that an antibody specific for the carboxy-terminal end of histone H3 was used. (C) Recruitment of Pol II to MET17 and MET2 in a Gcn5 HAT mutant. ChIP of strains containing wild-type GCN5 or gcn5^hat possessing the KQL mutation (24) using antibodies for Pol II CTD and PCR primers for MET17 and MET2 promoters was performed. Cells were grown and cross-linked as described in the legend to panel A. Values represent the average of two independent experiments, and error bars indicate standard deviations.

Growth phenotype on plates. Wild-type, met4Δ, med2Δ, gcn5Δ, and med2Δ gcn5Δ strains were grown to exponential phase, collected, washed, and suspended in water. Serial fivefold dilution was spotted on YNB minimal medium with or without 0.05 mM l-methionine. The most concentrated spot corresponds to a dilution at a density of 0.2 × 10⁷ cells/ml. Plates were incubated at 28°C for 3 and 4 days.