1 Endoplasmic Reticulum Stress-Responsive Transcription

JBC Papers in Press. Published on May 10, 2012 as Manuscript M112.369504 The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M112.369504

Endoplasmic Reticulum Stress-Responsive Transcription Factor ATF6α Directs Recruitment of the Mediator of RNA Polymerase II Transcription and Multiple Histone Acetyltransferase Complexes

Dotan Sela1, Lu Chen 1,2, Skylar Martin-Brown1, Michael P. Washburn1,3, Laurence Florens1, Joan Weliky Conaway1,2, and Ronald C. Conaway1,2

1From the Stowers Institute for Medical Research, Kansas City, MO 64110; 2 the Department of Biochemistry and Molecular Biology, Kansas University Medical Center, Kansas City, KS 66160; and 3 the Department of Pathology & Laboratory Medicine, Kansas University Medical Center, Kansas City, KS 66160

To whom correspondence should be addressed: Joan W. Conaway and Ronald C. Conaway, Stowers Institute for Medical Research, 1000 E 50th St, Kansas City, MO 64110, Tel: 816-926-4092, E-mail, [email protected] or [email protected].

Background: Transcription factor ATF6 is a acetyltransferase complexes, among which are master regulator of genes induced by endoplasmic the Spt-Ada-Gcn5 Acetyl-transferase (SAGA) Downloaded from reticulum stress. and Ada-Two-A-Containing (ATAC) Results: ATF6 can recruit Mediator and histone complexes. Our findings shed new light on the acetyltransferase complexes to promoter DNA via mechanism of action of ATF6α, and they interactions with overlapping sites in ATF6's outline a straightforward strategy for applying activation domain. MudPIT mass spectrometry to determine www.jbc.org Conclusion: ATF6 sequences essential for gene which Pol II transcription factors and activation recruit Mediator and histone coregulators are recruited to promoters and

acetyltransferases. other regulatory elements to control by guest, on June 25, 2012 Significance: Learning how coregulators transcription. communicate with DNA binding transcription factors is important for understanding gene Endoplasmic reticulum (ER)1 stress regulation. activates signal transduction pathways implicated in a variety of human diseases, including SUMMARY atherosclerosis, diabetes, and neurodegeneration The basic-leucine zipper transcription (1-3). A major signaling pathway activated by ER factor ATF6α functions as a master regulator stress is the “unfolded protein response” (UPR), of endoplasmic reticulum (ER) stress response which is triggered by accumulation of misfolded genes. Previous studies have established that, proteins in the ER and serves to protect cells from in response to ER stress, ATF6α translocates to ER stress in part by down-regulating synthesis of the nucleus and activates transcription of ER some ER-destined proteins and by up-regulating stress response genes upon binding sequence- expression of proteins involved in ER protein specifically to ER stress response enhancer folding. elements (ERSEs) in their promoters. In this A critical downstream event in the UPR report, we investigate the biochemical signal transduction pathway is activation of mechanism by which ATF6α activates transcription of a collection of ER stress response transcription. By exploiting a combination of genes by a mechanism that depends on the basic- biochemical and MudPIT-based mass leucine zipper (b-ZIP) transcription factor ATF6α spectrometry approaches, we have obtained (1-3). In unstressed cells, ATF6α resides in the ER evidence that ATF6α functions at least in part as a type II transmembrane protein . In response by recruiting to the ERSEs of ER stress to ER stress, ATF6α is transported to the Golgi response genes a collection of RNA polymerase apparatus, proteolytically processed by the site 1 II (Pol II) coregulatory complexes, including and site 2 proteases, S1P and S2P (4-8), and then the Mediator and multiple histone released into the nucleus, where it binds together

Copyright 2012 by The American Society for Biochemistry and Molecular Biology, Inc. with constitutively expressed transcription factors provide a relatively straightforward strategy for including NF-Y and YY1 to ERSEs in the exploiting the sensitivity of MudPIT mass promoters of ER stress response genes (4-11). spectrometry to determine how particular DNA Ectopic expression of the proteolytically processed binding transcription factors recruit Pol II form of ATF6α is sufficient to activate the ER coregulators to genes. stress-induced transcriptional program in unstressed cells, arguing that ATF6α acts as a EXPERIMENTAL PROCEDURES master regulator of ER stress response genes Materials. pET28a-Gal4-VP16 AD, which (4,9,12,13). encodes a fusion protein containing residues 1-94 Although cell biological studies have of the Gal4 DNA binding domain fused to residues provided significant insight into the mechanism by 412-490 of the VP16 activation domain, was a gift which ATF6α is proteolytically processed and from Jerry Workman (Stowers Institute). pGEX- transported to the nucleus, comparatively little is TRLBD, which encodes GST fused to amino acids understood about how it activates transcription of 90-380 of the thyroid receptor (19), was a gift ER stress response genes. ATF6α has an N- from Kristina Johnson and Michael Carey terminal transcription activation domain and a C- (UCLA). The ATF6α VN8-like peptide, which terminal DNA binding domain, whose affinity for contains 4 repeats of DFDLDLMP, was obtained its binding sites in ERSEs can be enhanced from Biosynthesis. Anti-ADA2b (ab57953), anti- Downloaded from through interactions with NF-Y (11,12,14,15). TBP (ab818), anti-TAF6 (ab51026), and anti-Pol Upon ER stress, multiple Pol II coregulators II (ab817) were from Abcam; anti-MBIP (10685- including the histone acetyltransferases (HATs) 1-AP) was from Proteintech, anti-α-Tubulin (sc- CBP/p300 and SAGA, the protein arginine 58666), anti-Med6 (sc-9433), anti-GST (sc-459),

methyltransferase PRMT1, and the INO80 anti CDK7 (sc-7344), anti-TFIIF RAP30 (sc- www.jbc.org chromatin remodeling complex have been shown 136408), TFIIEα (sc-28715) and TFIIB (sc-23875) to be recruited to ERSEs in the promoter of were from Santa Cruz Biotechnology; anti-GST

HSPA5 and other stress response genes (10,16-18). (A190-123A) was from Bethyl, anti-Med25 by guest, on June 25, 2012 Results of prior studies suggest that PRMT1and antiserum was a gift from Michael Carey, and the INO80 complex are recruited through YY1 anti-Med21 (H00009412-M05) was from Novus (10,17), but it is not known whether the ATF6α biologicals. Thyroid hormone T3 (T7650) was activation domain also contributes to recruitment from Sigma. All PCR primers were obtained from of these or other coregulators. IDT and are listed in Supplemental Table 1. In this report, we apply a combination of Immobilized HSPA5 promoter recruitment assay. biochemical and multi-dimensional protein To generate biotinylated DNA containing the wild identification technology (MudPIT)-based mass type HSPA5 promoter, a fragment containing 429 spectrometry approaches to investigate the role of bp of the human HSPA5 promoter (-282 to + 147 ATF6α in recruitment of Pol II coregulatory relative to the transcriptional start site) was proteins to the ERSEs of the HSPA5 gene. Below amplified from human genomic DNA by PCR and we present our findings, which are consistent with cloned into pGEM-T Easy Vector (Promega) to the model that ATF6α activates transcription at generate pGEM-HSPA5P. A biotinylated 673 bp least in part by orchestrating the recruitment of a fragment was amplified from pGEM-HSPA5P collection of Pol II coregulators, including the using a biotinylated primer derived from the T Mediator, SAGA, and ATAC complexes, to the vector and the reverse HSPA5 primer. A ERSEs of its target genes. Dissection of the biotinylated HSPA5 fragment lacking the ERSEs mechanism by which ATF6α recruits Pol II (HSPA5ΔERSE) was constructed using a two step coregulators argues that Mediator and HAT PCR procedure. In the first step, PCR products complexes are recruited through interactions with containing sequences immediately upstream and non-identical, but overlapping, regions of the downstream of the ERSEs were amplified and ATF6α transcription activation domain. Taken purified on an agarose gel. In the second step, together, our findings shed new light on the these PCR products were used as templates for the biochemical mechanisms underlying ATF6α- final PCR reaction. Biotinylated DNA fragments dependent transcription activation, and they were agarose gel purified and immobilized on

Dynabeads M-280 (Invitrogen) streptavidin were removed and beads were equilibrated in magnetic beads (1 pmol biotinylated DNA : 24 μl transcription buffer by washing twice with 10 mM 1% (w/v) slurry of beads) according to the Hepes pH 7.9, 10% glycerol, 0.05 M KCl, 6 mM manufacturer’s instructions. MgCl2, 0.1 mM EDTA, 0.5 mM DTT, 0.25 mM Nuclear extracts were prepared from PMSF. To initiate transcription, beads were suspension cultures of HeLa-S3 cells as described resuspended in 25 µl of the same buffer containing (20). 10-20 μl nuclear extract, with or without 6 40 units RNasin® (Promega) and 0.4 mM each of pmol of GST-ATF6α, were incubated for 10 min ATP, CTP, GTP, and UTP. After 90 min at 30oC, at 30oC in buffer PB (20 mM Hepes pH 7.9, 0.1 M reaction products were analyzed by primer NaCl, 15% Glycerol, 2 mM MgCl2, 0.15 mM extension as described (21) using an EDTA, 1 mM DTT, 0.02% NP-40) containing 0.1 oligonucleotide with the sequence GCT TCC CTC mg/ml bovine serum albumin (BSA) and 150 TCA CAC TCG CG. μg/ml dI-dC in a total volume of 80 µl. This GST-ATF6α binding assays. GST-ATF6α mixture was then added to ~450 fmol immobilized expression plasmids were introduced into pET41 HSPA5 promoter fragment on Dynabeads that had and transformed into BL-21 DE3 cells. Cells from been pre-equilibrated with PB containing 0.1 a single freshly transformed colony were grown to mg/ml BSA. The mixture was incubated a further an A600 of about 0.5, and protein expression was 30 min at 30oC with occasional mixing. The beads induced by adding IPTG to 1mM. Cells were Downloaded from were washed twice with 100 μl PB containing 0.1 grown overnight at 16 oC with shaking at 130 rpm. mg/ml BSA and once with PB lacking BSA. Cells were harvested at 6000 xg and resuspended Bound proteins were eluted with SDS sample in 20 ml of 50 mM Tris-Cl pH 7.9, 300 mM NaCl, buffer (2% SDS, 63 mM Tris-Cl pH 6.8, 10% 10% Glycerol, 0.5 mM AEBSF, 1 mM DTT. After Glycerol, 0.0025% bromophenol blue, 1.25% β- lysis with a French Press, the cell suspension was www.jbc.org mercaptoethanol) at 99 oC for 5 min and analyzed clarified by centrifugation for 30 min at 100,000xg by Western blotting. For Mudpit analysis, 800 µl at 4 oC, and the resulting supernatant was brought

binding reactions contained 100-200 µl nuclear to 0.1% Triton X-100. GST-tagged proteins were by guest, on June 25, 2012 extracts, 25 pmol of GST-ATF6α , and ~4.5 pmol purified from cell lysates using Glutathione DNA fragment on Dynabeads. Beads were Sepharose™ 4B beads (GE Healthcare Life washed twice with 100 μl PB containing 0.1 Sciences) using standard methods. Purified mg/ml BSA and once with PB lacking BSA. proteins were exchanged into buffer containing Bound proteins were then eluted by incubating 40mM Hepes-Cl pH 7.9, 0.05% Triton x-100, with 2%SDS, 50 mM Tris-Cl pH 8.8, 1.25 mM β- 1.5mM MgCl2, 0.1M NaCl, 1mM DTT (GB mercaptoethanol at 70 oC for 10 min. buffer) using Zeba desalt spin columns (Thermo In vitro transcription. Biotinylated DNA Scientific). fragments containing the wild type HSPA5 ~6 pmol GST-ATF6α or GST-ATF6α promoter (-282 to + 147) or HSPA5 promoter deletion mutants were mixed with 10-15 μl of lacking the TATA box sequence TATAAAG were HeLa-S3 nuclear extract in GB Buffer with generated by PCR, gel purified, and immobilized 0.5mg/ml BSA in a total volume of 80 μl, on Dynabeads. Hela cell nuclear extracts were incubated at 30 oC for 30 min, and then added to prepared as described (20) except that nuclei were 20 µl Glutathione Sepharose™ 4B equilibrated in extracted with 0.24 M KCl and dialyzed into 20 GB Buffer with 0.5mg/ml BSA. The bead-protein mM Hepes pH 7.9, 20% glycerol, 0.2 mM EDTA, slurry was incubated at 4 oC for 2 hrs on a Nutator 100 mM KCl, 1 mM DTT, and 0.5 mM PMSF. mixer (BD Diagnostics), washed 3 times with Transcription reactions were performed in two 100μl GB, and eluted with 3 bed volumes of 20 stages. First, 25 µl binding reactions containing mM Glutathione in GB. The same procedure was 10 mM Hepes pH 7.9, 10% glycerol, 0.1 M KCl, 2 used to purify proteins for MudPIT analysis, mM MgCl2, 0.1 mM EDTA, 0.5 mM DTT, 0.25 except that binding reactions contained ~25 pmol mM PMSF, 0.1 mg/ml BSA, 12 µl dialyzed GST- ATF6α and either 100 µl or 360 µl HeLa nuclear extract, and ~450 fmol immobilized DNA nuclear extracts, in total reaction volumes of 420 fragment, with or without GST-ATF6α, were μl or 1500 μl, respectively, and the concentration incubated for 30 min at 30oC. Unbound proteins of BSA in GB Buffer was reduced to 0.1 mg/ml.

Mass spectrometry. Proteins were identified using dSAF = i a modification of the multidimensional protein dNSAFi N identification technology (MudPIT) procedure ∑ dSAFi (22,23). TCA-precipitated proteins were urea- i=1 denatured, reduced, alkylated and digested with endoproteinase Lys-C (Roche) and modified uSpC trypsin (Roche) as described (22). Peptide i uSpCi + M × sSpCi mixtures were loaded onto 100 µm fused silica ∑m=1 uSpCm microcapillary columns packed with 5-μm C with dSAFi = 18 L reverse phase (Aqua, Phenomenex), strong cation i , exchange particles (Partisphere SCX, Whatman), and reverse phase (24). Loaded microcapillary in which shared spectral counts (sSpC) are columns were placed in-line with an LTQ ion trap distributed based on spectral counts unique to each mass spectrometer equipped with a nano-LC protein i (uSpC) divided by the sum of all unique electrospray ionization source (Thermo Scientific). spectral counts for the M protein isoforms that Fully automated MudPIT runs were carried out on shared peptide j with protein i. the electrosprayed peptides as described (23). Tandem mass (MS/MS) spectra were interpreted Downloaded from using SEQUEST (25) against a database of a RESULTS AND DISCUSSION database of 37394 human proteins (downloaded ATF6α-dependent recruitment of Mediator from NCBI on 2009-07-02), 4228 proteins from and HAT complexes to the HSPA5 ERSEs in Escherichia coli BL21 strain (downloaded from vitro. ATF6α activates transcription of ER stress

NCBI on 2010-01-10), 177 sequences from usual response genes by binding sequence-specifically www.jbc.org contaminants (human keratins, IgGs, proteolytic to its binding sites in the ERSEs of their enzymes), and complemented with the sequences promoters. The promoter of the HSPA5 gene

of each of the GST-ATF6 constructs used in this contains three ERSEs located ~60 basepairs by guest, on June 25, 2012 study. In addition, to estimate false positive upstream of the transcription start site (9) (Fig. discovery rates, each non-redundant sequence was 1A). In an attempt to identify in an unbiased way randomized (keeping amino acid composition and Pol II transcription regulatory proteins recruited by length the same), and the resulting "shuffled" ATF6α to the HSPA5 promoter, we coupled DNA sequences were added to the "normal" database affinity purification with MudPIT mass (doubling its size) and searched at the same time. spectrometry. MudPIT has proven to be a highly Peptide/spectrum matches were sorted and sensitive and reproducible means of identifying selected using DTASelect (26) with the following proteins in complex mixtures. In a MudPIT criteria set: spectra/peptide matches were only experiment, a mixture of proteins is first digested retained if they had a DeltaCn of at least 0.08, and into peptides, which are then fractionated by minimum XCorr of 1.8 for singly-, 2.0 for doubly- multidimensional HPLC and analyzed by tandem , and 3.0 for triply-charged spectra. In addition, mass spectrometry without first isolating peptides had to be fully-tryptic and at least 7 individual proteins from gels (23). Previous amino acids long. Combining all runs, proteins studies have shown that for many proteins in a had to be detected by at least 2 such peptides, or 1 MudPIT dataset, the number of spectra from peptide with 2 independent spectra. Peptide hits peptides of that protein is a function of its length from multiple runs were compared using and abundance. Consequently, the relative amount CONTRAST (26). To estimate relative protein of a particular protein in different samples can levels, distributed Normalized Abundance Factors often be estimated from a normalized spectral (dNSAFs) were calculated for each non-redundant abundance factor, or dNSAF (27-30). protein or protein group using the following To begin to investigate the function of formula as described by Zhang et al (27): ATF6α in regulation of ER stress response genes, we immobilized on streptavidin Sepharose beads a biotinylated DNA fragment containing HSPA5 gene sequences from –282 to +147 of the

transcription start site (Fig. 1A); this portion of the catalyze acetylation of nucleosomal histones, and HSPA5 gene includes the ERSEs, core promoter, all contribute to transcriptional regulation at least and early transcribed region. In preliminary in part by regulating histone acetylation in experiments, we confirmed that the immobilized chromatin (35-38). HSPA5 promoter supports ATF6α-activated The heat map shows the ratio of dNSAFs of transcription. To do so, we incubated the bead- Mediator and HAT subunits bound to immobilized bound HSPA5 DNA fragment with Hela cell HSPA5 promoter DNA in the presence (+) or nuclear extracts, with or without addition of absence (-) of ATF6α . The MudPIT results can be recombinant ATF6α, to allow preinitiation summarized as follows. First, all of the 30 complex formation. After washing to remove Mediator subunits detected were enriched in the unbound proteins, beads were incubated with presence of ATF6α. Twelve Mediator subunits ribonucleoside triphosphates, and transcription were detected only when binding reactions was measured using primer extension. As shown included exogenous ATF6α , and an additional 8 in Fig. 1B, a fragment containing the wild type Mediator subunits were enriched by more than 10- HSPA5 promoter supported ATF6α-activated fold. Similarly, most of the Pol II subunits transcription, whereas a fragment from which the detected were seen only in the presence of ATF6α. TATA-box was deleted did not. In addition to subunits of Mediator and Pol II, we To identify proteins recruited to the HSPA5 also observed increased binding to the HSPA5 Downloaded from promoter in an ATF6α-dependent way, we DNA fragment of subunits of the SAGA, ATAC, incubated the bead-bound, wild type HSPA5 DNA and TRRAP/TIP60 complexes. Finally, all three fragment with Hela cell nuclear extracts, with or subunits of the ERSE binding transcription factor without addition of recombinant ATF6α. After NF-Y were detected with similar dNSAF values washing, bound proteins were eluted with buffer when binding reactions were performed with or www.jbc.org containing SDS and analyzed by MudPIT. As without added ATF6α, consistent with previous illustrated in the heat map of Fig. 1C, MudPIT studies indicating that NF-Y can be detected by

identified subunits of the Mediator and several chromatin immunoprecipitation at the HSPA5 by guest, on June 25, 2012 HAT complexes, including the SAGA, ATAC, ERSEs in the absence of ER stress (10,11). and TRRAP/TIP60 complexes as proteins The ATF6α-dependent recruitment of recruited to the HSPA5 DNA fragment by ATF6α . Mediator, SAGA, and ATAC to the HSPA5 The Mediator complex is a very large, promoter was investigated further in assays using multisubunit complex that is required for proper Western blotting to monitor recruitment of regulation of the majority of Pol II genes and is representative subunits of each complex. thought to contribute to transcriptional regulation Consistent with results of MudPIT experiments, by acting directly on Pol II and other components recruitment of the Mediator subunit Med6 was of the general transcription apparatus. Although almost completely dependent on exogeneously its detailed mechanism(s) of action are poorly added ATF6α, while binding of SAGA subunit understood, it has critical roles in multiple stages ADA2b and ATAC subunit MBIP was enhanced of transcription, from assembly and function of the upon addition of ATF6α (Fig. 2A, lanes 1-4, 9). Pol II initiation complex to control of transcript Although not detected in our MudPIT datasets, we elongation (reviewed in (31-34). The human also observed that subunits of Pol II general SAGA, ATAC, and TRRAP/TIP60 HATs are also initiation factors, including TFIID (TAF6), TFIIB, large multisubunit complexes. SAGA and ATAC TFIIE, and TFIIF, were recruited to the HSPA5 share a common catalytic core that includes either promoter in an ATF6α-dependent manner (Fig. of two closely related acetyltransferases, GCN5 or 2B), consistent with previous studies indicating PCAF; each also includes a collection of that recruitment to promoters of Mediator (19) and additional subunits specific to one complex or the in some cases SAGA (42,43) is accompanied by other (reviewed in (35-38). Among the SAGA recruitment of TFIID and other components of the subunits is the transformation/transcription Pol II preinitiation complex. domain-associated protein (TRRAP), which is in It was shown previously that ATF6α- turn shared by the TRRAP/TIP60 HAT complex dependent transcription in cells requires a (39-41). SAGA, ATAC, and TRRAP/TIP60 functional ERSE (9,12). Importantly, recruitment

of ATF6α and Pol II coregulatory proteins to an The ATF6α transcription activation HSPA5 promoter fragment that lacked the ERSEs domain is necessary and sufficient for Mediator (HSPA5ΔERSE) was substantially reduced (Fig. and HAT binding. To begin to address the 2A, compare lanes 1-4 to 5-8). We note, however, mechanism of ATF6α-dependent recruitment of that binding of the ATAC subunit MBIP to HSPA5 the Pol II transcription machinery to the HSPA5 was less dependent than MED6 or ADA2b on promoter in our assays, we sought to determine added ATF6α or an intact ERSE, suggesting that it whether ATF6α is capable of interacting with any has more non-specific DNA binding activity of these Pol II transcription regulatory proteins in and/or that additional factor(s) present in the the absence of promoter DNA. To accomplish nuclear extracts also contribute to its binding to this, Hela cell nuclear extracts were incubated with the HSPA5 promoter. GST-ATF6α or one of several GST-ATF6α ATF6α-dependent recruitment of mutants (Fig. 3A). GST-ATF6α and associated Mediator and HATS to the HSPA5 promoter proteins were enriched using glutathione depends on both its DNA binding and Sepharose and analyzed by MudPIT mass transcription activation domains. Previous spectrometry and Western blotting. As shown in studies have shown that the ATF6α transcription Figs. 3B and 3C, these experiments revealed that activation domain resides within its first 150 GST-fusion proteins containing the ATF6α amino acids (12,15), while its b-ZIP family DNA transcription activation domain can interact with Downloaded from binding domain lies between residues 308 and 373 Mediator, Pol II, and each of the various HAT (4). To determine which ATF6α regions are complexes recruited by ATF6α to the HSPA5 required for recruitment of Mediator and HATs to promoter in the experiments of Figs. 1 and 2, as the immobilized HSPA5 promoter, binding well as with p300, which has been shown

reactions were performed using various ATF6α previously to be recruited to the HSPA5 ERSEs in www.jbc.org deletion mutants (Fig. 2C). Consistent with the response to ER stress (10). Binding of these Pol II earlier findings, deletion of a portion of the ATF6α coregulators to ATF6α was dependent on the

DNA binding domain prevented it from binding ATF6α transcription activation domain, since (i) by guest, on June 25, 2012 the HSPA5 promoter, while deletion of the GST-ATF6α (1-150) bound as much or more of transcription activation domain had no effect on the coregulators as did GST fused to full-length DNA binding. Arguing that the previously ATF6α and (ii) neither Mediator nor HAT mapped ATF6α activation domain plays a key role complexes bound to GST-ATF6α (151-326), in recruitment of Mediator and SAGA to the which lacks the transcription activation domain. HSPA5 promoter in our assays, deletion of either Finally, whereas we were able to detect binding of the DNA binding domain or the transcription Mediator and HATs to the immobilized ATF6α activation domain led to a dramatic decrease in transcription activation domain in the absence of Med6 and ADA2b recruitment to the promoter. In HSPA5 promoter DNA, we did not detect any of addition, although there was a significant the Pol II general initiation factors in these background of ATAC subunit MBIP even in the experiments (data not shown), suggesting that absence of ATF6α, the modest but reproducible recruitment of these proteins by ATF6α occurs in ATF6α-dependent increase in the amount of MBIP the context of preinitiation complexes assembled recruited to the immobilized promoter required on DNA. both the ATF6α DNA binding and transcription Mediator and HAT complexes bind to non- activation domains. We also observed a modest identical but overlapping regions of the ATF6α increase in NF-Y recruitment in the presence of transcription activation domain. We next sought full-length ATF6α, consistent with previous to define in more detail portions of the ATF6α evidence that ATF6α activation can lead to an activation domain required for its interaction with approximately two-fold increase in NF-Y detected Mediator and the various HAT complexes using at the HSPA5 ERSE by chromatin the series of GST-ATF6α fusion proteins immunoprecipitation (11) and suggesting that co- diagrammed in Fig. 4A. As shown in Figs. 4B and occupancy of ERSEs by ATF6α and NF-Y may 4C, there was little or no interaction between the lead to an increase in NF-Y's DNA binding Mediator complex and GST fusion proteins affinity. containing ATF6α (1-43) or ATF6α (20-60).

ATF6α fragments containing residues 20-80 and We next considered the possibility that the 44-150 bound less well to Mediator than ATF6α ATF6α VN8-like sequence might contribute to (1-150), whereas very similar amounts of Mediator and/or HAT recruitment. Indeed, we Mediator were recovered after glutathione observed that mutating the ATF6α VN8-like Sepharose chromatography of proteins that bound sequence to DADALLP led to a dramatic decrease GST fusion proteins containing ATF6α (20-100) in the interaction of ATF6α with Mediator (Figs and the full-length ATF6α activation domain 5C and 5D), reducing it to a level similar to that (residues 1-150). In contrast, maximal binding of seen when VP16 was included as competitor. the HAT complexes to ATF6α was observed only Adding Gal4-VP16 to binding reactions to GST-ATF6α (1-150), suggesting (i) that containing the ATF6α VN8-like mutant caused Mediator and the HAT complexes interact with little or no further reduction in the amount of ATF6α through overlapping but non-identical Mediator bound. In addition, mutation of the surfaces and (ii) that optimal binding of the HAT ATF6α VN8-like sequence substantially reduced complexes to ATF6α requires a larger region of the interaction of ATF6α with SAGA and ATAC, the transcription activation domain. even though Gal4-VP16 does not prevent binding Thureauf et al (15) previously noted that of ATF6α to SAGA and ATAC. the first 100 amino acids of the ATF6α In complementary experiments, we transcription activation domain exhibits some observed that micromolar concentrations of a 32 Downloaded from sequence similarity to the VP16 transcription amino acid peptide containing four repeats of the activation domain. Most similar within this region ATF6α VN8-like sequence blocked binding of is an 8 amino sequence, DFDLDLMP, that closely GST-ATF6α to Mediator, whereas higher resembles a VP16 sequence, DFDLDMLG, concentrations of the peptide were needed to block referred to as VN8 (15). In VP16, the VN8 binding to SAGA (Fig. 6A). Furthermore, we www.jbc.org sequence is necessary for transcription activation observed that even at millimolar concentrations (44,45). In addition, mutations in the VN8 the VN8-like peptide had little effect on thyroid

sequence have been reported to abolish interaction hormone (T3)-dependent binding of Mediator to by guest, on June 25, 2012 of VP16 with the Mediator complex (46), and the activation domain of thyroid receptor (TR), mutations in the VN8-like domain of ATF6α led supporting the specificity of inhibition by the to a 5-fold reduction in ATF6α-dependent VN8-like peptide and suggesting that ATF6α activation of a luciferase reporter driven by the interacts with a different binding site on Mediator HSPA5 ERSEs (15). than does TR. Taken together, the observations To explore in more detail the potential described thus far suggest that the VN8-like functional relationship between the ATF6α and sequence is important for the binding of both VP16 transcription activation domains, we first Mediator and the HAT complexes to the ATF6α asked whether the VP16 activation domain transcription activation domain, but that additional competes with ATF6α for interaction with sequences outside of the VN8-like region make a Mediator and HAT complexes. Remarkably, more significant contribution to the ATF6α-HAT adding increasing amounts of a Gal4-VP16 interaction. transcription activation domain fusion protein to Finally, we compared the effect of the Hela cell nuclear extracts prevented binding of VN8-like peptide on ATF6α-dependent GST-ATF6α (1-150) to Mediator and Pol II but recruitment of Mediator, SAGA, and the TFIID not to the HAT complexes, as revealed by component of the Pol II preinitiation complex to MudPIT and confirmed by Western blotting (Figs the HSPA5 promoter in vitro. As shown in Fig. 5A and 5B); that Pol II was depleted along with 6B, concentrations of peptide that had little effect Mediator is consistent with the possibility that on SAGA binding strongly reduced binding of ATF6α-dependent recruitment of Pol II is via a Mediator to the HSPA5 promoter. Notably, the Mediator-Pol II holoenzyme complex. VN8-like peptide reduced TFIID binding to the Importantly, the observation that the HAT HSPA5 promoter in parallel with its reduction of complexes remain associated with ATF6α in the Mediator binding, consistent with the possibility presence of Gal4-VP16 indicates they can that Mediator contributes to ATF6α-dependent associate with ATF6α independent of Mediator. TFIID recruitment to the promoter.

Summary and perspectives. In this report, residues in the ATF6α VN8-like sequence, which we present findings that shed new light on the has been shown to be important for ATF6α mechanism of action of ATF6α. By exploiting a activation domain function (15), led to a dramatic combination of biochemical and MudPIT-based decrese in binding of ATF6α to Mediator, SAGA, mass spectrometry approaches, we have obtained and ATAC. Thus, our findings are consistent with evidence that the ATF6α transcription activation the model that ATF6α regulates transcription of domain can recruit a collection of Pol II ER stress response genes at least in part by coregulators to the ERSEs in the promoter of the recruiting these Pol II coregulators to ERSEs. ER stress response gene HSPA5. These Pol II In the course of these experiments, we coregulators include the Mediator complex as well observed that the well-characterized VP16 as several HATs, including the SAGA, ATAC, transcription activation domain competes with and TRRAP-Tip60 complexes and p300. These ATF6α for binding to the Mediator but not to the findings are consistent with previous studies HAT complexes, suggesting that the ATF6α and reporting that p300 and the SAGA complex VP16 transcription activation domains might bind exhibit increased occupancy at the HSPA5 ERSEs to the same or overlapping surfaces on Mediator. following induction of the ER stress response with In light of previous studies indicating that VP16 thapsigargen (10,18). Similarly, in chromatin binds to Mediator through its Med25 subunit (46- immunoprecipitation experiments, we have 51), we are now investigating the possibility that Downloaded from observed increased binding of the Mediator ATF6α might also bind to Mediator through subunit Med26 at the HSPA5 promoter in cells Med25. In the future, these studies should provide subjected to ER stress (data not shown). a deeper understanding of the mechanism by In structure-function experiments which ATF6α facilitates recruitment of Mediator dissecting the mechanism by which ATF6α and other Pol II coregulators to the genes it www.jbc.org recruits coregulators, we observe that its binding regulates. Finally, our success in combining DNA to Mediator and HAT complexes depends on non- affinity chromatography and MudPIT mass

identical but overlapping regions of its spectrometry to identify proteins recruited by by guest, on June 25, 2012 transcription activation domain. Importantly, the ATF6α to the HSPA5 promoter suggests that this interactions between ATF6α, Mediator, SAGA, approach might be generally applicable in future and other HATs described in this report depend studies aimed at determining how particular DNA strictly on ATF6α domains shown previously to be binding transcription factors recruit Pol II essential for its transcription activity in cells coregulators to genes to regulate transcription. (12,15). Of particular note, mutation of three

Acknowledgements. We thank Chieri Tomomori-Sato, Shigeo Sato, and other members of the Conaway lab for helpful discussions and advice and Michael Carey (UCLA) for discussions and for a generous gift of the anti-Med25 antiserum used in this study. We are particularly grateful to Ariel Paulson for preparation of the heat maps; Maria Katt, Valerie Neubauer, and Tari Parmely for assistance with tissue culture; and Kyle Weaver for help generating expression constructs.

FOOTNOTES

*This work was supported in part by National Institute of General Medical Sciences grant GM41628 to JWC and RCC and by a grant to the Stowers Institute from the Helen Nelson Medical Research Fund at the Greater Kansas City Community Foundation.

1 Abbreviations: ATAC, Ada-Two-A-Containing; BSA, bovine serum albumin; b-ZIP, basic leucine zipper; dNSAF, distributed normalized spectral abundance factor; ER, endoplasmic reticulum; ERSE, ER stress response element; HAT, histone acetyl transferase; HPLC, high pressure liquid chromatography; MudPIT, multidimensional protein identification technology; NFY, nuclear factor Y; Pol II, RNA

polymerase II; S1P, site 1 protease; S2P, site 2 protease; SAGA, Spt-Ada-Gcn5 Acetyl-transferase; UPR, unfolded protein response; YY1, Yin-Yang 1

FIGURE LEGENDS

Fig. 1. ATF6α recruits a collection of Pol II coregulators to the HSPA5 promoter in vitro. (A) Diagram showing avidin bead-bound, biotinylated HSPA5 promoter fragment used in assays for ATF6α - dependent recruitment of Pol II coregulators to promoter DNA. Circle indicates position of avidin bead. (B) Assembly of functional preinitiation complexes on immobilized HSPA5 promoter DNA. Preinitiation complexes were assembled with or without ATF6α on bead-bound wild type or ΔTATA HSPA5 promoter fragments, washed, and assayed for their ability to support promoter-specific transcription as described in Experimental Procedures. (C) Heat map showing the ratio of dNSAFs of proteins bound to immobilized HSPA5 promoter in the presence (+) or absence (-) of ATF6α. Immobilized HSPA5 promoter binding assays were performed with or without addition of purified ATF6α , and bound proteins were detected by MudPIT mass spectrometry. See Supplemental Table 2 for supporting data.

Fig. 2. ATF6α-dependent recruitment of Mediator and HAT complexes to the HSPA5 ERSEs Downloaded from requires ATF6α DNA binding and activation domains. (A) ERSE dependence of ATF6α and Pol II coregulator binding to the HSPA5 promoter. Immobilized template recruitment assays were performed with either wild type (WT) or ∆ERSE HSPA5 promoter DNA and varying amounts of ATF6α . Bound proteins were detected by Western blotting using the designated antibodies. (B) ATF6α recruits components of the Pol II preinitiation complex to the HSPA5 promoter. (C) Diagram of GST- ATF6α www.jbc.org fusion proteins used in (E). WT, wild type; AD, transcription activation domain; DBD, DNA binding domain. (D) Effect of deleting ATF6α DNA binding or transcription activation domains on HSPA5

promoter binding. Immobilized HSPA5 promoter recruitment assays were performed with WT HSPA5 by guest, on June 25, 2012 and the indicated GST- ATF6α fusion proteins. Input (10% of total) and bound proteins were detected by Western blotting with the indicated antibodies.

Fig. 3. Binding of Mediator and HATs to the ATF6α transcription activation domain does not require HSPA5 promoter DNA. GST ATF6α-binding proteins were enriched from Hela cell nuclear extracts on glutathione Sepharose, using GST-fusion proteins containing the regions of ATF6α shown in panel (A). Black box indicated by the asterisk corresponds to the VN8-like sequence. (B) Heat map summarizing Mediator and HAT subunits bound by GST-ATF6α fusion proteins, detected by MudPIT. See Supplemental Table 2 for supporting data. (C) Mediator and HAT subunits bound by GST- ATF6α fusion proteins. The indicated proteins were detected by Western blotting.

Fig. 4. Mediator and HAT complexes bind to non-identical but overlapping regions of the ATF6α transcription activation domain. GST-ATF6α-binding proteins were purified from Hela cell nuclear extracts on glutathione Sepharose, using GST-fusion proteins containing the ATF6α transcription activation domain fragments shown in panel (A). (B) Heat map summarizing Mediator and HAT subunits bound by GST-ATF6α fusion proteins, detected by MudPIT. See Supplemental Table 2 for supporting data. (C) Mediator and HAT subunits bound by GST-ATF6α fusion proteins, detected by Western blotting with antibodies against the proteins indicated in parentheses. The data shown in Figs. 3C and 4C are derived from the same original gel; lane 6 of Fig. 4C is the same as lane 3 of Fig.3C.

Fig. 5. The VP16 transcription activation domain competes with ATF6α for binding to Mediator. (A, B) GST-ATF6α-binding proteins were purified from Hela cell nuclear extracts on glutathione Sepharose, using the GST-ATF6α transcription activation domain (amino acids 1-150) as bait. Binding reactions included the indicated amounts of GST-ATF6α (1-150) and Gal4-VP16. Bound proteins were analyzed by MudPIT (A, See Supplemental Table 2 for supporting data) and by Western blotting (B). (C)

Wild type and mutant GST-ATF6α fusion proteins used in the experiment of panel (D). The asterisk corresponds to the VN8-like sequence in the ATF6α transcription activation domain. Mutated residues are shown in red. (D) Glutathione Sepharose purification of GST-ATF6α binding proteins was performed with 12 pmol of wild type (WT) or mutant GST-ATF6α (1-373) as bait. Where indicated, binding reactions included 12 pmol of Gal4-VP16 as competitor; bound fractions were analyzed by Western blotting.

Fig. 6. The ATF6α VN8-like peptide blocks the interaction of ATF6α with Mediator and prevents TFIID recruitment to the immobilized HSPA5 promoter. (A) Glutathione Sepharose purifications were performed using 6 pmols GST-TR ligand binding domain, GST-ATF6α transcription activation domain (residues 1-150, AD), or GST-ATF6α DNA binding domain (residues 151-373, DBD) as bait, with (+) or without (-) 1 µM thyroid hormone. Where indicated, binding reactions included 1 nM, 1 µM, or 1 mM of a peptide containing 4 repeats of the VN8-like sequence from the ATF6α transcription activation domain (ATF6 VN8-like peptide). Bound fractions were analyzed by Western blotting with the antibodies shown in parentheses. (B) Immobilized template recruitment assays performed using biotinylated HSPA5 promoter fragment, the indicated amounts of the indicated VN8-like peptide, with or without ATF6α (1-373). Downloaded from

REFERENCES

1. Zhang, K. and Kaufman, R. J. (2008) From endoplasmic-reticulum stress to the inflammatory

response. Nature 454, 455-462 www.jbc.org 2. Ron, D. and Walter, P. (2007) Signal integration in the endoplasmic reticulum unfolded protein response. Nat. Rev. Mol. Cell Biol. 8, 519-529

3. Schroder, M. and Kaufman, R. J. (2005) The mammalian unfolded protein response. Annu. Rev. by guest, on June 25, 2012 Biochem. 74, 739-789 4. Haze, K., Yoshida, H., Yanagi, H., Yura, T., and Mori, K. (1999) Mammalian transcription factor ATF6 is synthesized as a transmembrane protein and activated by proteolysis in response to endoplasmic reticulum stress. Mol. Biol. Cell 10, 3787-3799 5. Ye, J., Rawson, R. B., Komuro, R., Chen, X., Dave, U. P., Prywes, R., Brown, M. S., and Goldstein, J. L. (2000) ER stress induces cleavage of membrane-bound ATF6 by the same proteases that process SREBPs. Mol. Cell 6, 1355-1364 6. Chen, X., Shen, J., and Prywes, R. (2002) The luminal domain of ATF6 senses endoplasmic reticulum (ER) stress and causes translocation of ATF6 from the ER to the Golgi. J. Biol. Chem. 277, 13045-13052 7. Shen, J., Chen, X., Hendershot, L., and Prywes, R. (2002) ER stress regulation of ATF6 localization by dissociation of BiP/GRP78 binding and unmasking of Golgi localization signals. Dev. Cell 3, 99-111 8. Shen, J. and Prywes, R. (2004) Dependence of site-2 protease cleavage of ATF6 on prior site-1 protease digestion is determined by the size of the luminal domain of ATF6. J. Biol. Chem. 279, 43046-43051 9. Yoshida, H., Haze, K., Yanagi, H., Yura, T., and Mori, K. (1998) Identification of the cis-acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose-regulated proteins. Involvement of basic leucine zipper transcription factors. J. Biol. Chem. 273, 33741-33749 10. Baumeister, P., Luo, S., Skarnes, W. C., Sui, G., Seto, E., Shi, Y., and Lee, A. S. (2005) Endoplasmic reticulum stress induction of the Grp78/BiP promoter: activating mechanisms mediated by YY1 and its interactive chromatin modifiers. Mol. Cell. Biol. 25, 4529-4540

11. Luo, R., Lu, J. F., Hu, Q., and Maity, S. N. (2008) CBF/NF-Y controls endoplasmic reticulum stress induced transcription through recruitment of both ATF6(N) and TBP. J. Cell Biochem. 104, 1708-1723 12. Yoshida, H., Okada, T., Haze, K., Yanagi, H., Yura, T., Negishi, M., and Mori, K. (2000) ATF6 activated by proteolysis binds in the presence of NF-Y (CBF) directly to the cis-acting element responsible for the mammalian unfolded protein response. Mol. Cell Biol. 20, 6755-6767 13. Yamamoto, K., Sato, T., Matsui, T., Sato, M., Okada, T., Yoshida, H., Harada, A., and Mori, K. (2007) Transcriptional induction of mammalian ER quality control proteins is mediated by single or combined action of ATF6alpha and XBP1. Dev. Cell 13, 365-376 14. Thuerauf, D. J., Arnold, N. D., Zechner, D., Hanford, D. S., DeMartin, K. M., McDonough, P. M., Prywes, R., and Glembotski, C. C. (1998) p38 Mitogen-activated protein kinase mediates the transcriptional induction of the atrial natriuretic factor gene through a serum response element. A potential role for the transcription factor ATF6. J. Biol. Chem. 273, 20636-20643 15. Thuerauf, D. J., Morrison, L. E., Hoover, H., and Glembotski, C. C. (2002) Coordination of ATF6- mediated transcription and ATF6 degradation by a domain that is shared with the viral transcription factor, VP16. J. Biol. Chem. 277, 20734-20739 16. Donati, G., Imbriano, C., and Mantovani, R. (2006) Dynamic recruitment of transcription factors and epigenetic changes on the ER stress response gene promoters. Nucleic Acids Res. 34, 3116- Downloaded from 3127 17. Cai, Y., Jin, J., Yao, T., Gottschalk, A. J., Swanson, S. K., Wu, S., Shi, Y., Washburn, M. P., Florens, L., Conaway, R. C., and Conaway, J. W. (2007) YY1 functions with INO80 to activate transcription. Nat. Struct. Mol. Biol. 14, 872-874

18. Nagy, Z., Riss, A., Romier, C., le, G., X, Dongre, A. R., Orpinell, M., Han, J., Stunnenberg, H., and www.jbc.org Tora, L. (2009) The human SPT20-containing SAGA complex plays a direct role in the regulation of endoplasmic reticulum stress-induced genes. Mol. Cell Biol. 29, 1649-1660

19. Johnson, K. M., Wang, J., Smallwood, A., Arayata, C., and Carey, M. (2002) TFIID and human by guest, on June 25, 2012 mediator coactivator complexes assemble cooperatively on promoter DNA. Genes Dev. 16, 1852- 1863 20. Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian cell nuclei. Nucleic. Acids. Res. 11, 1475-1489 21. Carey, M. F., Peterson, C. L., and Smale, S. T. (2009) Transcription and Preinitiation Complex Assembly In Vitro. Transcriptional Regulation in Eukaryotes: Concepts, Strategies, and Techniques, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 22. Washburn, M. P., Wolters, D., and Yates, J. R., III (2001) Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 19, 242-247 23. Florens, L. and Washburn, M. P. (2006) Proteomic analysis by multidimensional protein identification technology. Methods Mol. Biol. 328, 159-175 24. McDonald, W. H., Tabb, D. L., Sadygov, R. G., MacCoss, M. J., Venable, J., Graumann, J., Johnson, J. R., Cociorva, D., and Yates, J. R., III (2004) MS1, MS2, and SQT-three unified, compact, and easily parsed file formats for the storage of shotgun proteomic spectra and identifications. Rapid Commun. Mass Spectrom. 18, 2162-2168 25. Eng, J. K., McCormack, A. L., and Yates III, J. R. (1994) An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J. Am. Soc. Mass. Spec. 5, 976-989 26. Tabb, D. L., McDonald, W. H., and Yates III, J. R. (2002) DTASelect and Contrast: tools for assembling and comparing protein identifications from shotgun proteomics. J. Proteome Res. 1, 21- 26 27. Zhang, Y., Wen, Z., Washburn, M. P., and Florens, L. (2010) Refinements to label free proteome quantitation: how to deal with peptides shared by multiple proteins. Anal. Chem. 82, 2272-2281

28. Florens, L., Carozza, M. J., Swanson, S. K., Fournier, M., Coleman, M. K., Workman, J. L., and Washburn, M. P. (2006) Analyzing chromatin remodeling complexes using shotgun proteomics and normalized spectral abundance factors. Methods 40, 303-311 29. Paoletti, A. C., Parmely, T. J., Tomomori-Sato, C., Sato, S., Zhu, D., Conaway, R. C., Conaway, J. W., Florens, L., and Washburn, M. P. (2006) Quantitative proteomic analysis of distinct mammalian Mediator complexes using normalized spectral abundance factors. Proc. Natl. Acad. Sci. U. S. A. 103, 18928-18933 30. Zybailov, B., Mosley, A. L., Sardiu, M. E., Coleman, M. K., Florens, L., and Washburn, M. P. (2006) Statistical Analysis of Membrane Proteome Expression Changes in Saccharomyces cerevisiae. J. Proteome Res. 5, 2339-2347 31. Sato, S., Tomomori-Sato, C., Parmely, T. J., Florens, L., Zybailov, B., Swanson, S. K., Banks, C. A., Jin, J., Cai, Y., Washburn, M. P., Conaway, J. W., and Conaway, R. C. (2004) A set of consensus mammalian mediator subunits identified by multidimensional protein identification technology. Mol. Cell 14, 685-691 32. Conaway, R. C. and Conaway, J. W. (2011) Function and regulation of the Mediator complex. Curr. Opin. Genet. Dev. 21, 225-230 33. Taatjes, D. J. (2010) The human Mediator complex: a versatile, genome-wide regulator of transcription. Trends Biochem. Sci. 35, 315-322 Downloaded from 34. Malik, S. and Roeder, R. G. (2010) The metazoan Mediator co-activator complex as an integrative hub for transcriptional regulation. Nat. Rev. Genet. 11, 761-772 35. Rodriguez-Navarro, S. (2009) Insights into SAGA function during gene expression. EMBO Rep. 10, 843-850

36. Weake, V. M. and Workman, J. L. (2010) Inducible gene expression: diverse regulatory www.jbc.org mechanisms. Nat. Rev. Genet. 11, 426-437 37. Pijnappel, W. W. and Timmers, H. T. (2008) Dubbing SAGA unveils new epigenetic crosstalk.

Mol. Cell 29, 152-154 by guest, on June 25, 2012 38. Spedale, G., Timmers, H. T., and Pijnappel, W. W. (2012) ATAC-king the complexity of SAGA during evolution. Genes Dev. 26, 527-541 39. Cai, Y., Jin, J., Tomomori-Sato, C., Sato, S., Sorokina, I., Parmely, T. J., Conaway, R. C., and Conaway, J. W. (2003) Identification of new subunits of the multiprotein mammalian TRRAP/TIP60-containing histone acetyltransferase complex. J. Biol. Chem. 278, 42733-42736 40. Doyon, Y., Selleck, W., Lane, W. S., and Cote, J. (2004) Structural and functional conservation of the NuA4 histone acetyltransferase complex from yeast to humans. Mol. Cell. Biol. 24, 1884-1896 41. Ikura, T., Ogryzko, V., Gigoriev, M., Groisman, R., Wang, J., Horikoshi, M., Scully, R., Qin, J., and Nakatani, Y. (2000) Involvement of the TIP60 histone acetylase complex in DNA repair and apoptosis. Cell 102, 463-473 42. Bhaumik, S. R. and Green, M. R. (2002) Differential requirement of SAGA components for recruitment of TATA-Box-Binding protein to promoters in vivo. Mol. Cell. Biol. 22, 7365-7371 43. Dudley, A. M., Rougeulle, C., and Winston, F. (1999) The Spt components of SAGA facilitate TBP binding to a promoter at a post-activator-binding step in vivo. Genes Dev. 13, 2940-2945 44. Molinari, E., Gilman, M., and Natesan, S. (1999) Proteasome-mediated degradation of transcriptional activators correlates with activation domain potency in vivo. EMBO J. 18, 6439- 6447 45. Ikeda, K., Stuehler, T., and Meisterernst, M. (2002) The H1 and H2 regions of the activation domain of herpes simplex virion protein 16 stimulate transcription through distinct molecular mechanisms. Genes Cells 7, 49-58 46. Mittler, G., Stuhler, T., Santolin, L., Uhlmann, T., Kremmer, E., Lottspeich, F., Berti, L., and Meisterernst, M. (2003) A novel docking site on Mediator is critical for activation by VP16 in mammalian cells. EMBO J. 22, 6494-6504

47. Yang, F., DeBeaumont, R., Zhou, S., and Naar, A. M. (2004) The activator-recruited cofactor/Mediator coactivator subunit ARC92 is a functionally important target of the VP16 transcriptional activator. Proc. Natl. Acad. Sci. U. S. A. 101, 2339-2344 48. Milbradt, A. G., Kulkarni, M., Yi, T., Takeuchi, K., Sun, Z. Y., Luna, R. E., Selenko, P., Naar, A. M., and Wagner, G. (2011) Structure of the VP16 transactivator target in the Mediator. Nat. Struct. Mol. Biol. 18, 410-415 49. Vojnic, E., Mourao, A., Seizl, M., Simon, B., Wenzeck, L., Lariviere, L., Baumli, S., Baumgart, K., Meisterernst, M., Sattler, M., and Cramer, P. (2011) Structure and VP16 binding of the Mediator Med25 activator interaction domain. Nat. Struct. Mol. Biol. 18, 404-409 50. Bontems, F., Verger, A., Dewitte, F., Lens, Z., Baert, J. L., Ferreira, E., de, L. Y., Sizun, C., Guittet, E., Villeret, V., and Monte, D. (2011) NMR structure of the human Mediator MED25 ACID domain. J. Struct. Biol. 174, 245-251 51. Eletsky, A., Ruyechan, W. T., Xiao, R., Acton, T. B., Montelione, G. T., and Szyperski, T. (2011) Solution NMR structure of MED25(391-543) comprising the activator-interacting domain (ACID) of human mediator subunit 25. J. Struct. Funct. Genomics 12, 166

Downloaded from www.jbc.org by guest, on June 25, 2012

Fig. 1

A. C. Med29 Med13 Med28 Med31 ERSEs Med13L Med30 Med6 WT Med7 Med9 Med20 CDC2L6 CDK8

Mediator Med27 Med21 Med11 Med23 Med17 Med22 Med24 Med15 Med4 Med16 B. Med25 HSPA5 WT HSPA5 TATA Med14 Med26 Med8 Med19 ATF6 Med1 0 1.5 6 0 1.5 6 Med10 (pmol) MED12 POLR2C POLR2D POLR2A HSPA5 POLR2B Downloaded from

POLII POLR2E ATXN7 USP22 ATXN7L3 TAF12 KAT2A ATXN7L2 SUPT7L SF3B3 TRRAP TADA3L TADA2B SUPT3H SAGA TAF5L

TADA1L www.jbc.org ENY2 TAF6L ATF6 CCDC101 TAF9 [+/-] TAF10 DR1 TADA2L KAT2A TADA3L CSRP2BP by guest, on June 25, 2012 MBIP ATAC ZZZ3 CCDC101 25 YEATS2 WDR5 C1orf149 EPC1 MORF4L2 BRD8 VPS72 DMAP1 YEATS4 KAT5 TRRAP EP400 C20orf20

TRRAP/Tip60 MORF4L1 ACTL6A RUVBL1 RUVBL2 NFYA NFYB

1 NFY NFYC Fig. 2

A. C. HSPA5 promoter WT ERSE WT ATF6 AD DBD GST-ATF6 (pmol) 0 0.75 1.5 3 6 0.75 1.5 3 6 1 150 308 373 ATF6 (GST) GST - WT

151 373 Downloaded from Mediator (Med6) GST - AD 1 150 326 SAGA (ADA2b) GST - DBD

ATAC (MBIP) D. GST-ATF6

12345678 9 www.jbc.org WT AD DBD

B. HSPA5

promoter ATF6 (GST) by guest, on June 25, 2012 input (WT)

GST-ATF6 -+ -+ Mediator (Med6) Bound ATF6 SAGA (ADA2b)

Mediator (Med25) ATAC (MBIP) Pol II (RPB1) NFY (NFYa) TFIIH (CDK7) TFIIF (RAP 30) TFIIE (α) ATF6 (GST) Input TFIIB TFIID (TAF6) Mediator (Med6) 12 34 12345678 Fig. 3

GST-ATF6

151-326

1-373 1-326 1-150 A. B. dNSAF Med12 GST-ATF6 Med13 -2 CDC2L6 CDK8 10

Kinase Med13L AD DBD CCNC Med11 1 150 308 373 Med8 Med22 Med28 Med27 Med6 -3 1 150 326 Med30 10 Head Med17 Med29

Med20 Downloaded from Med19 1 150 Med21 Med4

Mediator Med10 Med9 Med1 -4

151 326 Middle Med31 10 * Med26 Med7 Med14 Med24

Med23 www.jbc.org

Tail Med16 Med15 Med25 POLR2C -5 POLR2D 10 POLR2E POLR2I

POLR2B by guest, on June 25, 2012 POLII GST-ATF6 POLR2A TADA3L C. 151-326 TRRAP TAF6L 1-373 1-326 1-150 SF3B3 ATXN7L3 TAF10 TAF5L CCDC101 KAT2A ENY2 TADA2B Mediator (Med6) SAGA SUPT7L TADA1L TAF9 USP22 SUPT3H SAGA (ADA2b) TAF12 ATXN7L2 RUVBL1 RUVBL2 ACTL6A ATAC (MBIP) EP400 DMAP1 C20orf20 YEATS4 BRD8 MORF4L1 KAT5 C1orf149 VPS72

TRRAP/Tip60 MORF4L2 EPC1 MYST2 ATF6 (GST) TADA3L CCDC101 KAT2A YEATS2 WDR5 ZZZ3

ATAC DR1 MBIP CSRP2BP TADA2L 123 4 p300 EP300 Fig. 4

GST-ATF6

GST-ATF6 44-150 20-100 A. 20-60 20-80 1-150 B. 1-43 1 43 Med12 Med13 CDC2L6 20 60 CDK8 dNSAF Kinase Med13L CCNC Med11 20 80 Med8 -2 Med22 10 Med28 Med27 100 Med6

20 Med30 Head Med17 Downloaded from Med29 Med20 44 150 Med19 -3 Med21 10 Med4

Mediator Med10 1 150 Med9 Med1

Middle Med31 Med26 Med7

Med14 -4 www.jbc.org * Med24 10 Med23

Tail Med16 Med15 Med25 POLR2C C. POLR2D GST-ATF6 POLR2E by guest, on June 25, 2012 POLR2I -5 POLR2B 10 POLII 20-100 44-150 POLR2A 20-60 20-80 1-150 TADA3L 1-43 TRRAP TAF6L SF3B3 ATXN7L3 TAF10 TAF5L Mediator (Med6) CCDC101 KAT2A ENY2 TADA2B

SAGA SUPT7L SAGA (ADA2b) TADA1L TAF9 USP22 SUPT3H TAF12 ATAC (MBIP) ATXN7L2 RUVBL1 RUVBL2 12345 6 ACTL6A EP400 DMAP1 C20orf20 YEATS4 BRD8 MORF4L1 KAT5 C1orf149 VPS72

TRRAP/Tip60 MORF4L2 EPC1 ATF6 (GST) MYST2 TADA3L CCDC101 KAT2A YEATS2 WDR5 ZZZ3 DR1 ATAC MBIP CSRP2BP TADA2L 12345 6 p300 EP300 Fig. 5

A. B. GST-ATF6 (pmol) 2424 24 - GST-ATF6 (pmol) 2424 24 24 - G4-VP16 (pmol) - 1224 24 G4-VP16 (pmol) - 1224 96 24 Med12 Med13 Mediator (Med6) CDC2L6 CDK8 SAGA (ADA2b)

Med13L Kinase CCNC Med11 dNSAF ATAC (MBIP) Med8 Med22 -2 Med28 10 CBP/p300 Med27 Med6 Med30

Head Med17 Med29 Med20 Med19 -3 Downloaded from Med21 C. Med4 10 1 150 308 373

Mediator Med10 Med9 ATF6 * Med1

Middle Med31 Med26 Med7 Med14 -4 WT DFDLDLMP Med24 10 Med23 mutant DADADLLP www.jbc.org Tail Med16 Med15 Med25 POLR2C POLR2D POLR2E POLR2I -5 D. by guest, on June 25, 2012 POLR2B 10 GST-ATF6 WT mutant

POLII POLR2A TADA3L G4-VP16 - - + + - - + + TRRAP TAF6L SF3B3 Mediator (Med6) ATXN7L3 TAF10 TAF5L CCDC101 KAT2A SAGA (ADA2b) ENY2 TADA2B

SAGA SUPT7L TADA1L ATAC (MBIP) TAF9 USP22 SUPT3H TAF12 ATXN7L2 ATF6 (GST) RUVBL1 RUVBL2 ACTL6A EP400 12345678 DMAP1 C20orf20 YEATS4 BRD8 MORF4L1 KAT5 C1orf149 VPS72 MORF4L2 TRRAP/Tip60 EPC1 MYST2 TADA3L CCDC101 KAT2A YEATS2 WDR5 ZZZ3 DR1 ATAC MBIP CSRP2BP TADA2L CBP/p300 EP300 Fig. 6 Downloaded from A. B. ATF6 VN8-like peptide [μM] 000.1 1 2.5 10 GST - TR GST - ATF6 ATF6+++++ - Bound to immobilized

AD DBD ATF6 www.jbc.org ATF6 VN8-like

peptide --nM M mM - nM M mM - SAGA (ADA2b)

T3 - ++++----- by guest, on June 25, 2012 (Med25) Mediator (Med6) (Med6) SAGA (ADA2b) HSPA5 promoter Mediator (Med18) GST (TBP) 1 2 345678910

TFIID (TAF6) input TFIID (TAF6)