Venters Molecular and Cellular Biology 2011 2253-2261

MOLECULAR AND CELLULAR BIOLOGY, June 2011, p. 2253–2261 Vol. 31, No. 110270-7306/11/$12.00 doi:10.1128/MCB.01464-10Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Genome-Wide Transcriptional Dependence on Conserved Regions of Mot1�

Bryan J. Venters, Jordan D. Irvin,† Paul Gramlich,‡ and B. Franklin Pugh*Center for Eukaryotic Gene Regulation, Department of Biochemistry and Molecular Biology,

Pennsylvania State University, University Park, Pennsylvania 16802

Received 22 December 2010/Returned for modification 25 January 2011/Accepted 21 March 2011

TATA binding protein (TBP) plays a central role in transcription complex assembly and is regulated by avariety of transcription factors, including Mot1. Mot1 is an essential protein in Saccharomyces cerevisiae thatexerts both negative and positive effects on transcription via interactions with TBP. It contains two conservedregions important for TBP interactions, another conserved region that hydrolyzes ATP to remove TBP fromDNA, and a fourth conserved region with unknown function. Whether these regions contribute equally totranscriptional regulation genome-wide is unknown. Here, we employ a transient-replacement assay usingdeletion derivatives in the conserved regions of Mot1 to investigate their contributions to gene regulationthroughout the S. cerevisiae genome. These four regions of Mot1 are essential for growth and are generallyrequired for all Mot1-regulated genes. Loss of the ATPase region, but not other conserved regions, caused TBPto redistribute away from a subset of Mot1-inhibited genes, leading to decreased expression of those genes. Acorresponding increase in TBP occupancy and expression occurred at another set of genes that are normallyMot1 independent. The data suggest that Mot1 uses ATP hydrolysis to redistribute accessible TBP away fromintrinsically preferred sites to other sites of intrinsically low preference.

From Saccharomyces cerevisiae to human, the TATA bindingprotein (TBP) provides an indispensable role in nearly allRNA polymerase I, II, and III transcription events (29). TBPis the central component of a complex regulatory networkgoverning transcription complex assembly (33). Consequently,TBP is subjected to an extraordinary level of regulation bynumerous transcription factors, one of which is Mot1 (6). Mot1is a conserved Snf2/Swi2-related ATPase (21) that regulatesthe dynamics of TBP-promoter interactions by removing TBPfrom DNA using the energy of ATP hydrolysis (6, 45, 47). Thefirst 800 amino-terminal residues of Mot1 are both necessaryand sufficient for TBP binding (2, 7). The Snf2-related ATPasedomain resides within the last 600 carboxy-terminal residues(7). Genome-wide expression studies using temperature-sensi-tive mot1 alleles indicate that Mot1 regulates between 3 and15% of the yeast genome, some negatively and others posi-tively (4, 17, 24).

In vitro biochemical experiments have largely defined mech-anisms by which Mot1 regulates TBP-DNA interactions. Mot1can bind and stabilize TBP-DNA interactions, but in the pres-ence of ATP, it dissociates TBP from DNA and, to someextent, Mot1 from TBP, thereby recycling both (2, 7, 12, 26).This reaction is important for two reasons. First, dynamic as-sembly and disassembly of the transcription machinery impartprecise control over gene expression. Therefore, regulated re-cruitment of TBP to promoters must be accompanied by reg-

ulated removal of TBP, which Mot1 is well suited to do. In thiscontext, Mot1 is a negative regulator.

Since TBP binds to the minor groove of DNA, which haslimited sequence specificity, TBP has relative high affinity fornonspecific DNA (14). If bound inappropriately, this mightlead to aberrant or nonproductive assembly of the transcrip-tion machinery. Biochemical experiments have demonstratedthat Mot1 can remove nonspecifically bound TBP (41), per-haps acting as a chaperone allowing TBP to rebind in a pro-ductive mode. For example, at the URA1 gene, Mot1 canpromote transcription by removing a nonproductive TBPbound in the reverse orientation (46). In this context, Mot1operates as a positive regulator.

The mechanism by which Mot1 acts on TBP is well definedbiochemically, and this provides a basis for interpreting lessdefined in vivo experiments. Because Mot1 is essential forgrowth in S. cerevisiae (19), in vivo functional analysis of im-portant regions of Mot1 is not straightforward in that loss offunction is lethal. Thus, an investigation into the genome-widefunctions of essential proteins like Mot1 is hampered by thepractical limitation that mutations that eliminate functioncause cell death. Temperature-sensitive mutations might alle-viate this problem to some extent, but they are difficult totarget to specific regions of the protein and often vary inseverity. To circumvent this limitation, we utilized a transient-replacement strategy (34) to investigate the contributions ofconserved Mot1 domains to transcription and TBP recruitmentgenome-wide. Our study revealed that Mot1-regulated tran-scription is dependent on nearly all conserved regions of Mot1.Strikingly, transcriptional dependence for a subset of genes isspecifically altered when the Mot1 ATPase domain is deleted.Genome-wide location analysis of TBP in a strain that lacks theMot1 ATPase region corroborates the expression-profiling ex-periments, suggesting a direct effect on TBP. Furthermore,coimmunoprecipitation of TBP and the Mot1 ATPase deletionmutant demonstrates that the two directly interact. These find-

* Corresponding author. Mailing address: Department of Biochem-istry and Molecular Biology, Penn State University, 456 N. Frear Lab-oratory, University Park, PA 16802. Phone: (814) 863-8252. Fax: (814)863-8595. E-mail: [email protected].

† Present address: Gene Regulation and Chromosome Biology Lab-oratory, National Cancer Institute, National Institutes of Health, Fred-erick, MD 21702-1201.

‡ Present address: Schering-Plough Research Institute, Union, NJ07083.

� Published ahead of print on 28 March 2011.

2253

ings reveal that Mot1-regulated genes are generally regulatedby all parts of Mot1 and that a TBP-binding portion of Mot1can alter the selectivity of TBP for promoters.

MATERIALS AND METHODS

Plasmids. pCALF-T(PGK) (36) was converted to pCALF-FHT-T(PGK) 2.2by inserting a 66-bp HIS-TEV oligonucleotide into the NdeI site. pUG6-FHT-P(4,170 bp) was made by PCR amplifying 259 bp containing the FHT (Flu-His-TEV) sequence from the pCALFHT-T(PGK) 2.2 plasmid. The PCR product wasdigested with SalI, and 161 bp was ligated into the SalI-digested pUG6 plasmid(4,009 bp) so that the orientation was FHT-loxP-kanMX-loxP.

FHT-Mot1 mutant strains. A list of the yeast strains used in this study isprovided in Table 1. S. cerevisiae strain BY4743 (9) (Invitrogen) was used as theparental strain. Initially, the strain was transformed with pSH47 (URA3) (25)encoding galactose-inducible Cre recombinase. Oligonucleotides (70-mer) wereused to PCR amplify 1,991 bp of pFA6a-His3MX6-PGAL1 containing the HIS3gene and GAL1 promoter (38). The PCR product was transformed into BY4743using a high-efficiency lithium acetate method (25) to replace 550 bp of theendogenous MOT1 promoter with the GAL1 promoter, creating strain yLAC1.HIS� homologous-recombination transformants were selected on complete syn-thetic medium lacking histidine and uracil (CSM-HIS-URA medium) and veri-fied by colony PCR.

Regions of MOT1 were deleted by replacing coding sequences with an FHTtag. The FHT tag encodes three hemagglutinin (HA) (Flu) repeats, a decahis-tidine (H), sequence, and the TEV protease sequence (T). The kanamycinresistance region of pUG6-FHT-p was PCR amplified with 68-mer oligonucle-otides with 50-bp homology to distinct regions of MOT1. The PCR products weretransformed into yLAC1 and selected on CSM-HIS-URA (dextrose) plates con-taining 500 mg/ml G418 (Invitrogen). The kanamycin resistance cassette flankedby loxP sites was removed by induction of Cre recombinase with 2% galactose for4 h, leaving the FHT tag coding sequence upstream of the mutation in MOT1.Kanamycin-sensitive colonies were identified by replica plating them on mediacontaining and lacking G418. Additionally, mutations were verified by colonyPCR with primers specific to each mutation.

Haploids. Kanamycin-sensitive FHT-Mot1 strains (Table 1) were plated onCSM-HIS plus 5-fluoroorotic acid (5-FOA) to select cells that had lost pSH47and verified by replica plating on CSM-HIS and CSM-HIS-URA. The strainswere then transformed with pMR13 (MOT1 wild type [WT] and URA3), andtransformants were selected on CSM-HIS-URA medium. The strains wereplated on presporulation medium (1% yeast extract, 2% peptone, and 10%dextrose) for 2 days at 30°C. Cells were cultured in sporulation medium (0.3%potassium acetate, 0.02% raffinose) for 3 days at 30°C. Two hundred microlitersof the culture was pelleted; resuspended in 1.2 M sorbitol, 10 mM Tris, pH 7.4;and treated with 20 units of (1 mg/ml) zymolyase (MP Biomedicals) at roomtemperature for 20 min. Tetrads were dissected according to standard yeasttechniques on YPD (yeast-peptone-dextrose) plates. Spores were replica platedonto CSM-HIS-URA medium to select for the HIS3 gene (and therefore theGAL1 promoter). Mating types of the mot1 strains were confirmed with MATaand MATa sex tester strains. MATa leu� HIS� LYS� tetrads were selected. Thestrains were then transformed with pMOT221 (mot1-42 LEU2) or pAV20(MOT1 WT LEU2) and selected on CSM-LEU medium. Cells that lost pMR13(MOT1 WT URA3) were selected by plating them on CSM-LEU plus 5-FOA.

Colony PCR. For colony PCR, 1� 25 mM MgCl2 buffer (Gene Choice), 2.5 UTaq polymerase (Gene Choice), 0.0002 U Pfu polymerase (Stratagene), 0.4 mM

deoxynucleoside triphosphates (dNTPs), and 0.2 mM each primer were used per50-ml reaction for 32 cycles.

Cell growth assays. MATa haploid FHT-Mot1 mutant strains carryingpMOT221 (mot1-42 LEU2) were grown at 25°C in YPR (yeast-peptone-3%raffinose) to mid-log phase. Cells (A600 � 0.5) were removed, and 2.5 �l of10-fold serial dilutions was spotted on three sets of YPD (2% dextrose) andYPG (2% galactose) plates and incubated at 25°C, 30°C, or 37°C. Photo-graphs were taken after 48 h.

Microarray analysis. Microarrays were performed essentially as describedpreviously (13, 32). Briefly, cultures were grown to an A600 of �0.6, induced with2% galactose in YPR medium for 60 min at 25°C, and shifted to 37°C for 45 minto inactivate the temperature-sensitive copy of Mot1 encoded by the mot1-42allele. Cells were harvested by centrifugation at room temperature, washed inRNase-free DEPC (diethyl pyrocarbonate)-treated double-distilled H2O, andfrozen in liquid nitrogen.

Total RNA was isolated as described previously (32), and poly(A) tail mRNAwas purified using oligo(dT) cellulose (Ambion) according to the manufacturer’sinstructions. Reverse transcription, labeling with fluorescent dyes (Cy3 and Cy5;Amersham), hybridization, and scanning were all performed as described previ-ously (13, 32). Four micrograms of mRNA was used for hybridizations instead ofthe conventional 2 �g. Slides were treated with Dye Saver2 (Genisphere) ac-cording to the manufacturer’s instructions to preserve signal intensity. Data setswere mode normalized by using R software (Bioconductor) to mode-centerreplicates (dye swaps) (data are available at http://atlas.bx.psu.edu/).

Genes were filtered by several criteria to minimize false positives. (i) Geneswere eliminated if their signals on the array were greater than 25% saturated. (ii)The mean foreground signal minus the median background signal had to begreater than the standard deviation of background signal. (iii) Quality data wereneeded from both replicates of the dye swap. (iv) The directional change of themutant’s signal (relative to the reference) had to be equivalent in the replicates.False-discovery rates (FDRs) were determined using a modified version of amethod described previously (35). The false-discovery rate is reported as apercentage of the number of expression changes above and below a given thresh-old (�0.59; log2 scale) in the homotypic control (yjdi420) compared to eachFHT-Mot1 mutant expression experiment (yjdi410 to -416). The FDRs are 4.3%,2.0%, 2.4%, 3.0%, and 2.9% for TBP1, TBP2, UK1, ATP1, and Null, respec-tively.

K-means clustering was performed using Cluster (22) on 515 genes that con-tained data in 80% of the experiments and that had a change of at least 1.5-fold(log2 ratio � 0.59) in one of the mutants. K was chosen to equal 6 clusters (K �6). Clusters 5 and 6 were merged because they were visually indistinguishable.Clustering information was visualized using Treeview (22).

ChIP-chip. ChIP-chip was performed essentially as described previously (48),with minor changes noted below. Briefly, cultures were grown to an A600 of �0.6,induced with 2% galactose in YPR medium for 60 min at 25°C, and shifted to37°C for 45 min to inactivate the temperature-sensitive copy of Mot1 encoded bythe mot1-42 allele. The cells were then fixed by adding formaldehyde to a finalconcentration of 1% for 2 h at 25°C (instead of a typical 15-min cross-link time)and quenched for 5 min with glycine. The cultures were diluted 2-fold with thesame volume of temperature-adjusted distilled water just prior to addition offormaldehyde to achieve a medium temperature of 25°C. The harvested cellswere lysed with glass beads, and the chromatin pellet was washed and sonicated.Sheared chromatin was immunoprecipitated with IgG-Sepharose. This ChIP-enriched DNA was amplified by ligation-mediated PCR (LMPCR) as describedelsewhere (27), and 100- to 250-bp LMPCR-amplified fragments were gel puri-

TABLE 1. MOT1 yeast strains used in this study

Strain Promoter MOT1 allele Deletion (aa)a Plasmid MAT Reference

BY4743 MOT1 WT Diploid 9yLAC1 GAL1 WT pSH47 Diploid 34yjdi408 GAL1 FHT-WT1 pMOT221 (mot1-42) LEU2 Alpha This studyyjdi410 GAL1 FHT-TBP1 05–80 pMOT221 (mot1-42) LEU2 Alpha This studyyjdi412 GAL1 FHT-TBP2 307–432 pMOT221 (mot1-42) LEU2 Alpha This studyyjdi414 GAL1 FHT-UK1 1090–1259 pMOT221 (mot1-42) LEU2 Alpha This studyyjdi416 GAL1 FHT-ATP1 1403–1867 pMOT221 (mot1-42) LEU2 Alpha This studyyjdi418 GAL1 Null MOT1::kanMX pMOT221 (mot1-42) LEU2 Alpha This studyyjdi420 GAL1 WT pMOT221 (mot1-42) LEU2 Alpha This study

a aa, amino acids.

2254 VENTERS ET AL. MOL. CELL. BIOL.

fied according to the manufacturer’s protocol (Qiagen) and subsequently hybrid-ized to low-density tiled spotted microarrays containing �21,000 oligonucleotideprobes as described previously (48). Briefly, each yeast gene is interrogated witha set of at least 3 oligonucleotide probes, which survey the relative occupancylevels for a given transcription factor at the �250 and �60 sites relative to thetranslational start site and the downstream portion of the open reading frame(ORF). Data were filtered and analyzed as previously described (49) (data areavailable at http://atlas.bx.psu.edu/).

Coimmunoprecipitation and immunoblotting. Briefly, FHT-Mot1 mutantstrains were grown as described above for ChIP-chip, except the cells were notcross-linked with formaldehyde. The harvested cells were then pelleted and flashfrozen in liquid nitrogen. The cells were lysed with glass beads in NP-S buffer (10mM Tris-Cl, pH 7.5, 0.5 mM Spermidine, 0.075% Igepal (Sigma), 50 mM NaCl,5 mM MgCl2, 1 mM CaCl2). Chromatin pellets were washed in NP-S buffer, thechromatin was micrococcal nuclease digested (15 units) in a volume of 300 �l for20 min, and then the chromatin was solubilized by washing the spun pellet withFA lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.1%Na deoxycholate, 2 mM EDTA). This native nuclear extract was immunopre-cipitated with anti-TBP rabbit polyclonal antibody serum, and the eluate wasassayed by Western blot analysis for interacting FHT-Mot1 mutants.

For Western blotting, FHT-Mot1 mutant coimmunoprecipitation eluates wereelectrophoresed in 6% SDS-PAGE and transferred to a polyvinylidene difluoride(PVDF) membrane (Pall Gelman Laboratory) in Western transfer buffer for 60min at 1.5 A. FHT-Mot1 mutants were detected with 1:3,300 anti-HA (HA.11;Babco) and 1:5,000 anti-mouse–horseradish peroxidase (HRP) antibodies (Am-ersham) and exposed to Hyperfilm (Amersham) with enhanced chemilumines-cence (ECL) (Amersham).

Comparisons with public microarray data. The relationships to the top andbottom 10th percentiles of the expression and ChIP-chip data were calculated inExcel with the data downloaded from the referenced laboratory or journal’s web-sites. The percent rank of the distribution was calculated with the PERCENTRANKfunction. Next, the numbers of genes that appear in the top 10% (�0.9 inPERCENTRANK) or the bottom 10% (0.1 in PERCENTRANK) and also ap-pear in each cluster were calculated. The CHITEST function of Excel was then usedto calculate P values from the observed and expected values.

RESULTS

Replacement strategy for essential Mot1 functions. Mot1 isdemarcated by several regions that are highly conserved fromyeast to human (12, 16). To test their in vivo importance, wetargeted four conserved regions for deletion, creating mutantsnamed TBP1, TBP2, UK1, and ATP1 (Fig. 1A to C). Thenames reflect the associated functional regions (e.g., the TBP1deletion removes one of two TBP interaction domains, UK1 isunknown, and ATP1 removes the ATPase domain). Deletionof each region of MOT1 was achieved by homologous recom-bination using a PCR-amplified cassette containing kanMXflanked by loxP Cre recombination sites (34). The cassette alsocontained coding sequences that allowed the deleted region tobe replaced by an FHT epitope tag that encodes a triple-HAtag, a decahistidine tag, and a TEV protease cleavage site.After selection for recombinants on G418 plates and subse-quent excision of kanMX with the Cre recombinase, the de-leted region was replaced with the FHT tag and a single loxPsite, both of which maintained an open reading frame throughthe replaced region. Cell viability was maintained in a residentmot1-42 temperature-sensitive allele, which allowed subse-quent temperature inactivation of the mot1-42 allele at 37°C(Fig. 2A). The location of each deletion mutation was verifiedby PCR across the deletion borders, with the appropriate-sizeproducts detected (not shown). Expression of the deletionmutants was placed under the control of the GAL1 promoter.Immunoblot analysis demonstrated the presence of an appro-priate-size band that reacted with anti-HA antibodies and waspresent only after the addition of galactose to the cells (Fig.

2B). The untagged Mot1 (WT) and null controls were notdetected because both lack the FHT tag and thus are notrecognized by the HA antibody. All mutants except UK1 wereexpressed at levels roughly equivalent to that of the wild-typeMot1 containing an FHT tag at the N terminus (WT1). UK1was expressed at about 50% of the WT1 level. Importantly, theexpression levels of the mutants were not diminished afterinactivation of the mot1-42 allele at 37°C for 45 min.

Four conserved regions of Mot1 contribute essential func-tions to Mot1. Haploid strains carrying a galactose-induciblechromosomal copy of one of the mot1 deletion mutants and aplasmid-borne temperature-sensitive mot1-42 (16) were testedfor the ability of the mot1 domain deletion mutants to supportviability. The mot1-42 supporting cell viability was inactivatedat 37°C and replaced with the domain deletion mutants byadding galactose to the medium. At 25°C and 30°C, mot1-42remained functional, allowing cell growth, as expected. In thepresence or absence of the galactose-induced domain deletionmutants (Fig. 3, 25°C and 30°C), viability was unaffected,indicating that the domain deletion mutants did not have adominant-negative effect on cell growth. At 37°C, wild-typeMot1 (WT and WT1) and all mutants failed to supportviability in dextrose medium, where these proteins are notexpressed, verifying the temperature-sensitive nature of themot1-42 allele (16). As positive controls, both the untagged(WT) and tagged (WT1) inducible Mot1 supported growthin galactose medium at 37°C. However, none of the Mot1domain deletion mutants support growth, indicating thateach of the four conserved domains is essential in S. cerevi-siae. These findings are consistent with related studies pre-sented elsewhere (1, 7, 16).

Mot1-regulated genes display strong transcriptional depen-dence on all regions of Mot1. Inasmuch as Mot1 regulatesgenes both negatively and positively, we tested whether thefour conserved regions of Mot1 make distinct gene-specificcontributions to gene expression on a genome-wide scale. Ex-pression of the Mot1 mutants was induced with galactose for60 min, and then the resident functional mot1-42 allele wasshut down by an abrupt temperature shift from 25°C to 37°Cfor 45 min (Fig. 2A). During this time, we expect heat shock-regulated genes to change in expression within 15 min of thetemperature shift and then return to near-normal expressionlevels by 45 min (11, 23).

To place any changes in gene expression into the appropri-ate context, each expression profile conducted on a mutant wasalso conducted in parallel using a galactose-inducible untaggedwild-type MOT1 allele (WT; yjdi420) (Table 1). Thus, if amutant is as functional as wild-type Mot1, then no changes ingene expression are expected. Changes in gene expressionwere mode centered, meaning that the most frequent binnedratio (mutant/WT) corresponded to no change in gene expres-sion. This centering is valid, since most genes are not appre-ciably regulated by Mot1 (4, 17, 24). Log2-transformed changesin gene expression are presented as a cluster plot (Fig. 4A),where red and green denote increased and decreased expres-sion, respectively. Black denotes no change. Each row corre-sponds to a protein-coding gene, and each column correspondsto the expression profile for a Mot1 mutant. The 515 genes thatmet a specified cutoff (Fig. 4A) for changes in expression in atleast one set of experiments are shown. The data were clus-

VOL. 31, 2011 IN VIVO STRUCTURE-FUNCTION STUDY OF Mot1 2255

tered by K-means (22) into five clusters, representing the max-imum number of visually nonredundant clusters. The homo-typic control expression profile, reflecting comparisonsbetween two independent biological replicates of galactose-induced wild-type Mot1, produced no appreciable changes inexpression (Fig. 4A, WT), as expected.

The homotypic control also provided a measure of intrinsicvariability in the data. The mean log2 ratio was 0.045 (1.03- �0.12-fold [standard deviation] change). To ensure that theFHT tag was not perturbing expression, an FHT-tagged wild-type Mot1 control (Fig. 4A, WT1) was tested. Changes in gene

expression were very modest compared to the untagged refer-ence (log2 average � 0.033, or 1.02-fold � 0.12-fold change),indicating that the FHT tag had little or no effect on expressiongenome-wide. In particular, the standard deviation in geneexpression was not significantly different than that observedwith the homotypic control.

Changes in gene expression of the galactose-induced nullmutant (empty cassette in the mot1-42 strain) at 37°C provideda measure of the maximal level of expected change in expres-sion and correlated well with changes in gene expression re-ported for the mot1-1 or mot1-14 allele (4, 17) (Fig. 4B). Thus,

FIG. 1. Mutagenesis of conserved regions in the MOT1 gene. (A) Schematic of the MOT1 gene showing the approximate locations of fourconserved regions. (B) Schematic of the constructed Mot1 mutants. The null control represents WT1 in which the kanMX selection gene was notremoved. (C) Conservation of the four regions of Mot1 targeted for mutagenesis in this study. S. cerevisiae (Sc) and Homo sapiens (Hs) Mot1protein sequences were aligned using the BLAST algorithm (3).


this transient-replacement system appears to provide an ade-quate reflection of Mot1 dependency.

From these expression-profiling experiments, we find genesthat are negatively regulated by Mot1 (clusters 1 to 3 in Fig.4A) and those that are positively regulated by Mot1 (cluster 5)each require the TBP1, TBP2, and UK1 conserved domains ofMot1 in that changes in gene expression were similar to that ofthe null mutant (Fig. 4A, compare columns 3 to 6). Thus, ingeneral, Mot1 uses the same conserved domains to positivelyand negatively regulate transcription. Genes in cluster 1 arecharacterized as being stress induced, TATA containing,SAGA dominated, and negatively regulated by a wide range ofTBP and chromatin regulators (Table 2). Cluster 2 genes havethe same characteristics as those of cluster 1, except that theyare not stress induced and are not inhibited by histones. Clus-ter 3 appears to be a mixture of cluster 1 and 2 genes. Genesin cluster 5 are characterized as being stress repressed, TATA-

less, and TFIID dominated. In addition, the ribosomal proteingenes dominate this group. The positive contribution of Mot1to cluster 5 expression is in line with reports proposing thatMot1 positively regulates transcription by redistributing TBPthroughout the genome (15, 41) and/or by dismantling tran-scriptionally inactive TBP (18, 46). Taken together, Mot1 typ-ically regulates 6% of all TFIID-dominated genes in a positivemanner, whereas 28% of all SAGA-dominated genes tend tobe negatively controlled by Mot1.

Deletion of the Mot1 ATPase domain causes ectopic geneexpression. The Mot1 ATP1 mutant, corresponding to a dele-tion of the ATPase domain, had mixed behavior. At mostanalyzed genes, reflected by clusters 2 and 5 in Fig. 4A, nega-tive and positive regulation by Mot1 required the ATPasedomain as much as it required the other conserved regions.Cluster 1 expression displayed less dependence on the ATPasedomain of Mot1. The difference between clusters 1 and 2 may

FIG. 2. Transient-replacement strategy. (A) The transient-replacement approach simultaneously expresses the mot1 test allele from aninducible galactose-driven promoter while inactivating the temperature-sensitive (ts) version expressed from the mot1-42 allele supporting viability.Consequently, at the time of harvest, the Mot1 mutants expressed from the galactose promoter have replaced the ts copy of Mot1. (B) Galactoseinduction of Mot1 derivatives. An immunoblot probed with monoclonal anti-HA antibodies is shown, revealing the FHT epitope-tagged Mot1derivatives. Cells were taken just prior to addition of 2% galactose to raffinose media (�), after 2% galactose induction for 60 min at 25°C (�),or after galactose induction and another 45 min of incubation at 37°C (�) to inactivate the endogenous mot1-42 allele (not detectable in this assay).The uniformity of the background bands just above the 121- and 54-kDa markers confirms approximately equal loading and transfer of samples.

FIG. 3. Effects of Mot1 mutants on cell growth. Strains carrying pMOT221 (mot1-42 LEU2) and mot1 mutant derivative alleles were grown at25°C in YPR (3% raffinose) to mid-log phase. Then, 0.5 A600 unit of cells were removed, and 2.5 �l of 10-fold serial dilutions were spotted on threesets of YPD (2% dextrose) and YPG (2% galactose) plates and incubated at 25°C, 30°C, or 37°C. The mot1 mutant strains are indicated at theleft of the plate images. The photographs were taken after 48 h.


be subtle, such as a greater rate-limiting dependency on TBPbinding versus ATP hydrolysis. Alternatively, since cluster 1genes were generally induced and then partially repressed bythe temperature regime used here, it is possible that whileMot1 binding is needed for this shutdown, the Mot1 ATPaseactivity may be partially dispensable. In other words, Mot1binding in the absence of ATPase activity is sufficient to elicitsome repression. This would be consistent with current modelsof Mot1 function in which simply binding to TBP would besufficient to preclude binding to certain general transcriptionfactors.

Cluster 4 genes appeared to be largely Mot1 independent,since expression of the null or any of the Mot1 domain dele-

tions (except ATP1) had little effect on transcription (Fig. 4A,columns 3 to 6 in cluster 4). The ATP1 mutant caused anincrease in transcription. Since the primary TBP binding re-gions of this protein are intact, conceivably the ATP1 mutantmight promote DNA binding of TBP at these genes but isunable to dissociate TBP without the ATPase domain. In thiscase, the binding of the Mot1 ATP1 mutant to TBP wouldseem not to completely interfere with subsequent transcriptioncomplex assembly. A plausible rationale for this lies in ourobservation that cluster 4 genes tend to be repressed by theSSN6-TUP1 complex (P � 10�39) (Table 2). In this context,any enhancement of TBP binding by Mot1(ATP1), even in aweakened state, would provide some increase in expression.One implication is that stabilization of TBP binding at SSN6-TUP1-repressed genes circumvents to some extent SSN6-TUP1 repression.

Cluster 3 genes are generally inhibited by wild-type Mot1but showed decreased expression when the ATPase domainwas removed (Fig. 4A, column 2, cluster 3), paradoxically sug-gesting that the Mot1 ATPase region plays an apparently pos-itive role at these genes while Mot1 as a whole plays a negativerole. Two alternative explanations might account for the ap-parent paradox. First, as in cluster 4, the ATP1 mutant mightstabilize TBP binding at the promoter of cluster 3 genes. How-ever, this TBP-Mot1(ATP1) mutant complex seems to inter-fere with productive transcription complex assembly at thesecluster 3 genes. Alternatively, the ATP1 mutant might promotenonspecific binding of TBP to the genome, which would reducethe amount of TBP that could be recruited to cluster 3 genes,resulting in decreased expression of cluster 3 genes. This wouldbe consistent with the notion that Mot1 normally removes TBPfrom nontargeted regions of the genome (41). We explorethese possibilities below.

Mot1(ATP1) directly interacts with TBP. Potential interpre-tations of the Mot1(ATP1) expression profile are predicatedupon the Mot1(ATP1) mutant maintaining the ability to inter-act with TBP. Thus, to test whether Mot1(ATP1) directly in-teracts with TBP in vivo, we performed coimmunoprecipitationassays for TBP in the FHT-Mot1 mutant strains. Growth andMot1 replacement were performed as in the expression stud-ies. Importantly, to maintain native protein-protein interac-tions, the cells were not cross-linked throughout the coimmu-noprecipitation procedure. TBP was immunoprecipitated fromdigested, soluble chromatin derived from Mot1 mutant cells.The ability of Mot1 mutants to interact with TBP on chromatinwas revealed by immunoblot analysis (Fig. 5A). The positivecontrol, the tagged (WT1) inducible Mot1, showed an interac-tion with TBP (Fig. 5A, bottom). As in Fig. 2B, the untaggedMot1 (WT) and null controls were not detected, because bothlack the FHT tag and thus are not recognized by the HAantibody. Among the Mot1 mutant derivatives, theMot1(ATP1) mutant showed the strongest interaction in vivowith TBP, while the other mutants (TBP1, TBP2, and UK1)displayed either weak or no interaction. Therefore, the inter-action between Mot1(ATP1) and TBP revealed by their coim-munoprecipitation supports one assertion of the interpretationthat expression changes in clusters 3 and 4 may be a directresult of the Mot1(ATP1) mutant altering the DNA-bindingstatus of TBP at these genes.

FIG. 4. Genome-wide transcription profiling of Mot1 mutants.(A) Cluster analysis of Mot1 mutant expression profiles. Strains har-boring each of the indicated Mot1 mutants were harvested accordingto the transient-replacement strategy (Fig. 2A). mRNA was isolatedand cohybridized, along with an untagged wild-type reference (WT), tospotted full-length PCR-generated ORF microarrays. Changes in geneexpression (log2 scale) were clustered using Cluster software and vi-sualized with Treeview (22). Membership required quality data in sixof the seven experiments and a log2 absolute value of 0.59 (1.5-foldchange) in at least one experiment and data in six of the seven clus-tered experiments. A total of 515 ORFs met these criteria, with FDRsacross all experiments of �4.3%. Rows were clustered by K means intofive clusters (n � 120, 120, 21, 49, and 209). The rows representindividual genes, and fold changes in gene expression are reflected inthe color intensity. Columns representing an average of at least twodye-swapped experiments were clustered hierarchically. The dendro-gram relating them is depicted above the column labels. The tablebelow the cluster plot provides the median log2 ratio in each cluster ineach experiment. (B) Moving-average comparison of genome-widechanges in expression for mot1-42 (null) with mot1-1 (4) and mot1-14(17). Fold changes in gene expression (log2 scale) for mot1-42 weresorted. The average changes in expression for mot1-42 in each sliding200-gene window was plotted against an equivalent moving averageobtained from mot1-1 and mot1-14 strains.


Deletion of the Mot1 ATPase domain causes TBP redistri-bution in the genome. To further understand how theMot1(ATP1) mutant might affect TBP recruitment to promot-ers, we used genome-wide location analysis (ChIP-chip) tomonitor the changes in TBP, Mot1, and TFIID (Taf1 and Taf4subunits) occupancy at every yeast gene in the Mot1(ATP1)mutant. Growth and Mot1 replacement were performed as inthe expression studies. Each factor was immunoprecipitatedfrom sheared soluble chromatin derived from formaldehydecross-linked wild-type (WT1) and Mot1(ATP1) mutant cells.Bound DNA was differentially labeled and cohybridized tomicroarrays containing all intergenic regions. Median log2

changes in TBP and Mot1 occupancy for each of the clustersdefined in Fig. 4A are plotted in Fig. 5B and compared tochanges in gene expression for the Mot1(ATP1) and null mu-tants.

TBP and Mot1 occupancy changes at cluster 3 and 4 genesmirrored the expression output for these genes in the ATP1mutant (a decrease at cluster 3 and an increase at cluster 4),suggesting that the expression change is a direct result of theMot1(ATP1) mutant altering the DNA-binding status of TBPat these genes. The loss of TBP at cluster 3 genes suggests thatwhen Mot1 lacks its ATPase domain, the corresponding loss intranscription is not due to stabilization of an inactive form ofTBP at these promoters. Instead, the results are more consis-tent with the loss of TBP possibly being due to stabilization ofTBP bound to other sites in the genome (e.g., cluster 4 genes)by the Mot1(ATP1) mutant. Indeed, TBP and Mot1 occupancy

increased in the Mot1(ATP1) mutant at cluster 4 genes. Be-cause cluster 3 genes normally have more TBP than cluster 4genes (Fig. 5C), they have more to lose if TBP is distributednonspecifically throughout the genome in the Mot1(ATP1)mutant.

One interpretation of the positive activity of Mot1 on ex-pression at cluster 5 genes is that it removes an inactive formof TBP, allowing productive binding of TFIID. Accordingly aMot1(ATP1) mutant might “lock down” TBP at cluster 5genes, thereby preventing TFIID from binding. To test thishypothesis, we conducted ChIP-chip on the Taf1 and Taf4subunits of TFIID, comparing its occupancy at cluster 5 genesin the wild type versus a Mot1(ATP1) mutant. Consistent withthis hypothesis, Taf1 and Taf4 occupancy levels decreased atcluster 5 genes in the Mot1(ATP1) mutant, whereas TBP levelsremained largely unchanged (Fig. 5D). A constant level of TBPis consistent with the hypothesis, in that one type of TBP (TAFfree) replaces another type of TBP (TAF bound; TFIID).

DISCUSSION

The binding of TBP to DNA is generally considered to be aprimary nucleating event in transcription complex assembly.TBP-DNA binding is therefore subjected to substantial posi-tive and negative regulation. TBP not only binds to the TATAbox located in promoters, it also binds to TATA-less promoterregions, and it binds to nonspecific DNA with fairly high af-finity. Since nonspecific DNA binding by TBP can nevertheless

TABLE 2. Relationship of Mot1 expression profiles with published microarray data

Category Protein/characteristic % rankP value for clustera:

Reference1 2 3 4 5

Group/property TATA box 52 26 10 9 – 8SAGA dominated 73 33 13 – – 32TATA box and SAGA dominated 85 26 18 – – 32ESRb upregulated 69 – 7 – – 23ESR downregulated – – – – 52 23Rap1 ChIP and ribosomal protein – – – – 101 37High transcription frequency – – – – 40 30

ExpressionStress response Heat shock Top 10 68 6 7 – – 51

Salt stress Top 10 35 – – – – 11Alkali stress Top 10 36 12 – – – 11Diauxic shift Top 10 29 – – – – 20Oxidative stress Top 10 25 – – – – 11Amino acid starvation Top 10 23 11 – – – 28

Regulatory factors mot1-14 Top 10 59 45 19 – – 17mot1-14 Bottom 10 – – – – 81 17mot1-1 Top 10 33 19 – – – 4mot1-1 Bottom 10 – – – – 14 4bur6-1 (NC2) Top 10 39 17 17 – – 10tup1 (SSN6-TUP1) Top 10 14 – 7 39 – 31TBP K145E (Mot1 binding) Top 10 24 35 – – – 33TBP F182V (NC2 binding) Top 10 76 30 – – – 33

Nucleosome H31-28 Top 10 8 – 6 8 – 42H42-26 Top 10 31 – – – – 43htz1 Top 10 26 – – – – 40

a The �log10(P value) for overlap between membership in each of the five Mot1 expression clusters and the top/bottom 10th percentile expression change orgroup/property membership from published genomic data sets. The P value was calculated using the chi-square test and returns the probability that the overlap betweenthe two data sets occurs by chance. Thus, the most statistically significant relationships have the largest �log10(P value). For clarity �log10 (P value) values of less thanor equal to 5 were replaced with dashes.

b ESR, environmental stress response.


nucleate transcription (14), robust mechanisms exist in vivo toprevent promiscuous assembly.

Mot1 plays an important role in removing TBP from inap-propriate genomic sites (18, 41), which would free up TBPand/or the underlying DNA to engage in productive interac-tions (46). In this way, Mot1 would play a positive role atpromoters that are rate limited either by the availability of TBPor by dissociation of an inactive TBP-promoter complex. Suchan arrangement might predominate at ribosomal proteingenes, where Mot1 plays a positive role (cluster 5 in Fig. 4A).Conceivably, a Mot1-inaccessible form of TBP (i.e., TFIID)binds to ribosomal promoters (12, 39). Nonproductive bindingof a TFIID-independent form of TBP might antagonize TFIIDrecruitment or any other stage in transcription initiation. Re-moval of this nonproductive TBP would therefore positivelyimpact transcription at these genes. Indeed, we find that sta-bilizing a TBP-Mot1 promoter interaction at Mot1-upregu-lated genes has the effect of displacing TFIID.

Stress-induced promoters often are highly transcribed even

under nonstress conditions (11, 23). These promoters tend torely more on a free form of TBP that does not involve TFIID.This form of TBP may be directed to the appropriate promot-ers via the SAGA complex, and since it is not TFIID, it may bemore accessible to Mot1. Indeed, the abundance of accessibleTBP at such promoters attracts Mot1, where it downregulatesexpression (cluster 1 in Fig. 4A). A moderately high level ofexpression is achieved through this balance of positive SAGA-TBP action (among other factors) and negative Mot1-TBPaction (among other factors). Indeed, by using multidimen-sional chromatography coupled to mass spectrometry (5), arecent study found that Mot1 interacts with a variety of acti-vators and transcriptional coregulators, such as Msn2, Hsf1,and RSC.

Since Mot1 has the ability to bind both TBP and DNA (44),it potentially can stabilize TBP-DNA interactions. However,this is not realized in general because Mot1 uses the energy ofATP hydrolysis to remove TBP from DNA. A form of Mot1that lacks the ATPase domain provides a window into howpotential stabilization of TBP-DNA interactions through Mot1affects the distribution of TBP genome-wide. If theMot1(ATP1) mutant were to increase TBP-DNA stability in anundirected way, then genomic loci that normally lack TBPshould see an increase in TBP in the mutant. Those loci thatnormally have higher TBP levels should suffer a decrease inoccupancy as TBP is sequestered at the vast number of non-specific loci. Indeed we identified a set of genes (cluster 4) thathad lower levels of TBP and whose levels of TBP, Mot1, andtranscription increased in the Mot1(ATP1) mutant. We iden-tified another set of genes (cluster 3) that had higher levels ofTBP, which decreased in the Mot1(ATP1) mutant.

Taken together, our findings suggest that the four conservedregions of Mot1 are essential for viability and required forproper regulation of most Mot1-regulated genes. Loss of theATPase domain, however, imparts some unexpected regula-tion on certain genes. Although only a fraction of all yeastgenes are overtly regulated by Mot1, essentially all yeast genesrequire TBP and thus are potential targets for Mot1. Genesthat are insensitive to loss of Mot1 likely reflect those that arerelatively quiescent and thus lack TBP or those that involveTFIID, which is a Mot1-insensitive form of TBP. AllowingTBP to redistribute genome-wide in a potentially more non-specific manner results in a net gain of TBP at TBP-deficientgenes and a net loss at TBP-enriched genes. Thus, one conse-quence of Mot1 using ATP hydrolysis to remove TBP fromDNA may be an increase in promoter selectivity.

ACKNOWLEDGMENTS

We are grateful to David Auble for plasmids pMOT221 (mot1-42LEU2), pMR13 (MOT1 WT URA3), and MOT1 WT (pAV20 LEU2).We thank Joe Reese for kindly providing the Taf1 and Taf4 antibodies.

This work was supported by National Institutes of Health grantGM059055.

REFERENCES

1. Adamkewicz, J. I., K. E. Hansen, W. A. Prud’homme, J. L. Davis, andJ. Thorner. 2001. High affinity interaction of yeast transcriptional regulator,Mot1, with TATA box-binding protein (TBP). J. Biol. Chem. 276:11883–11894.

2. Adamkewicz, J. I., C. G. Mueller, K. E. Hansen, W. A. Prud’homme, andJ. Thorner. 2000. Purification and enzymic properties of Mot1 ATPase, aregulator of basal transcription in the yeast Saccharomyces cerevisiae.J. Biol. Chem. 275:21158–21168.

FIG. 5. TBP coimmunoprecipitation (CoIP) and location profilingof TBP, Mot1, and TFIID subunits in an ATPase-defective Mot1strain. (A) TBP was immunoprecipitated from chromatin extracts andthen probed by Western blotting for both TBP (top) and FHT-Mot1(anti-HA antibodies) (bottom). (B) ChIP-chip was performed in anATPase-defective Mot1 mutant (ATP1) and wild-type Mot1 strain(WT1) background. The median log2 changes in occupancy for each ofthe five clusters in Fig. 4A were calculated. For comparison, the me-dian log2 expression changes for the null and ATP1 expression profilesfrom Fig. 4A are also shown. (C) The median TBP occupancy at 25°Cfrom a previous study (50) for each cluster is shown as a histogram.Log2 values are relative to nonpromoter intergenic regions (�1,800),which are regions located between two convergently transcribed genes.(D) ChIP-chip was performed on two subunits of TFIID, Taf1 andTaf4. Shown is a histogram of the median log2 changes in occupancyfor each of the five clusters in Fig. 4A for Taf1 and Taf4 occupancychanges in an ATPase-defective Mot1 strain relative to wild-type Mot1(TBP is also shown for comparison).


3. Altschul, S. F., et al. 1997. Gapped BLAST and PSI-BLAST: a new gener-ation of protein database search programs. Nucleic Acids Res. 25:3389–3402.

4. Andrau, J. C., et al. 2002. Mot1p is essential for TBP recruitment to selectedpromoters during in vivo gene activation. EMBO J. 21:5173–5183.

5. Arnett, D. R., J. L. Jennings, D. L. Tabb, A. J. Link, and P. A. Weil. 2008. Aproteomics analysis of yeast Mot1p protein-protein associations: insights intomechanism. Mol. Cell Proteomics 7:2090–2106.

6. Auble, D. T., et al. 1994. Mot1, a global repressor of RNA polymerase IItranscription, inhibits TBP binding to DNA by an ATP-dependent mecha-nism. Genes Dev. 8:1920–1934.

7. Auble, D. T., D. Wang, K. W. Post, and S. Hahn. 1997. Molecular analysis ofthe SNF2/SWI2 protein family member MOT1, an ATP-driven enzyme thatdissociates TATA-binding protein from DNA. Mol. Cell. Biol. 17:4842–4851.

8. Basehoar, A. D., S. J. Zanton, and B. F. Pugh. 2004. Identification anddistinct regulation of yeast TATA box-containing genes. Cell 116:699–709.

9. Brachmann, C. B., et al. 1998. Designer deletion strains derived from Sac-charomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14:115–132.

10. Cang, Y., and G. Prelich. 2002. Direct stimulation of transcription by nega-tive cofactor 2 (NC2) through TATA-binding protein (TBP). Proc. Natl.Acad. Sci. U. S. A. 99:12727–12732.

11. Causton, H. C., et al. 2001. Remodeling of yeast genome expression inresponse to environmental changes. Mol. Biol. Cell 12:323–337.

12. Chicca, J. J., II, D. T. Auble, and B. F. Pugh. 1998. Cloning and biochemicalcharacterization of TAF-172, a human homolog of yeast Mot1. Mol. Cell.Biol. 18:1701–1710.

13. Chitikila, C., K. L. Huisinga, J. D. Irvin, A. D. Basehoar, and B. F. Pugh.2002. Interplay of TBP inhibitors in global transcriptional control. Mol. Cell10:871–882.

14. Coleman, R. A., and B. F. Pugh. 1995. Evidence for functional binding andstable sliding of the TATA binding protein on nonspecific DNA. J. Biol.Chem. 270:13850–13859.

15. Collart, M. A. 1996. The NOT, SPT3, and MOT1 genes functionally interactto regulate transcription at core promoters. Mol. Cell. Biol. 16:6668–6676.

16. Darst, R. P., et al. 2003. Mot1 regulates the DNA binding activity of freeTATA-binding protein in an ATP-dependent manner. J. Biol. Chem. 278:13216–13226.

17. Dasgupta, A., R. P. Darst, K. J. Martin, C. A. Afshari, and D. T. Auble. 2002.Mot1 activates and represses transcription by direct, ATPase-dependentmechanisms. Proc. Natl. Acad. Sci. U. S. A. 99:2666–2671.

18. Dasgupta, A., S. A. Juedes, R. O. Sprouse, and D. T. Auble. 2005. Mot1-mediated control of transcription complex assembly and activity. EMBO J.24:1717–1729.

19. Davis, J. L., R. Kunisawa, and J. Thorner. 1992. A presumptive helicase(MOT1 gene product) affects gene expression and is required for viability inthe yeast Saccharomyces cerevisiae. Mol. Cell. Biol. 12:1879–1892.

20. DeRisi, J. L., V. R. Iyer, and P. O. Brown. 1997. Exploring the metabolic andgenetic control of gene expression on a genomic scale. Science 278:680–686.

21. Eisen, J. A., K. S. Sweder, and P. C. Hanawalt. 1995. Evolution of the SNF2family of proteins: subfamilies with distinct sequences and functions. NucleicAcids Res. 23:2715–2723.

22. Eisen, M. B., P. T. Spellman, P. O. Brown, and D. Botstein. 1998. Clusteranalysis and display of genome-wide expression patterns. Proc. Natl. Acad.Sci. U. S. A. 95:14863–14868.

23. Gasch, A. P., et al. 2000. Genomic expression programs in the response ofyeast cells to environmental changes. Mol. Biol. Cell 11:4241–4257.

24. Geisberg, J. V., Z. Moqtaderi, L. Kuras, and K. Struhl. 2002. Mot1 associ-ates with transcriptionally active promoters and inhibits association of NC2in Saccharomyces cerevisiae. Mol. Cell. Biol. 22:8122–8134.

25. Guldener, U., S. Heck, T. Fielder, J. Beinhauer, and J. H. Hegemann. 1996.A new efficient gene disruption cassette for repeated use in budding yeast.Nucleic Acids Res. 24:2519–2524.

26. Gumbs, O. H., A. M. Campbell, and P. A. Weil. 2003. High-affinity DNAbinding by a Mot1p-TBP complex: implications for TAF-independent tran-scription. EMBO J. 22:3131–3141.

27. Harbison, C. T., et al. 2004. Transcriptional regulatory code of a eukaryoticgenome. Nature 431:99–104.

28. Hardwick, J. S., F. G. Kuruvilla, J. K. Tong, A. F. Shamji, and S. L.

Schreiber. 1999. Rapamycin-modulated transcription defines the subset ofnutrient-sensitive signaling pathways directly controlled by the Tor proteins.Proc. Natl. Acad. Sci. U. S. A. 96:14866–14870.

29. Hernandez, N. 1993. TBP, a universal eukaryotic transcription factor? GenesDev. 7:1291–1308.

30. Holstege, F. C., et al. 1998. Dissecting the regulatory circuitry of a eukaryoticgenome. Cell 95:717–728.

31. Hughes, T. R., et al. 2000. Functional discovery via a compendium of ex-pression profiles. Cell 102:109–126.

32. Huisinga, K. L., and B. F. Pugh. 2004. A genome-wide housekeeping role forTFIID and a highly regulated stress-related role for SAGA in Saccharomy-ces cerevisiae. Mol. Cell 13:573–585.

33. Huisinga, K. L., and B. F. Pugh. 2007. A TATA binding protein regulatorynetwork that governs transcription complex assembly. Genome Biol. 8:R46.

34. Irvin, J. D., and B. F. Pugh. 2006. Genome-wide transcriptional dependenceon TAF1 functional domains. J. Biol. Chem. 281:6404–6412.

35. Johnson, W. E., et al. 2006. Model-based analysis of tiling-arrays for ChIP-chip. Proc. Natl. Acad. Sci. U. S. A. 103:12457–12462.

36. Kou, H., J. D. Irvin, K. L. Huisinga, M. Mitra, and B. F. Pugh. 2003.Structural and functional analysis of mutations along the crystallographicdimer interface of the yeast TATA binding protein. Mol. Cell. Biol. 23:3186–3201.

37. Lieb, J. D., X. Liu, D. Botstein, and P. O. Brown. 2001. Promoter-specificbinding of Rap1 revealed by genome-wide maps of protein-DNA association.Nat. Genet. 28:327–334.

38. Longtine, M. S., et al. 1998. Additional modules for versatile and economicalPCR-based gene deletion and modification in Saccharomyces cerevisiae.Yeast 14:953–961.

39. Mencía, M., Z. Moqtaderi, J. V. Geisberg, L. Kuras, and K. Struhl. 2002.Activator-specific recruitment of TFIID and regulation of ribosomal proteingenes in yeast. Mol. Cell 9:823–833.

40. Meneghini, M. D., M. Wu, and H. D. Madhani. 2003. Conserved histonevariant H2A.Z protects euchromatin from the ectopic spread of silent het-erochromatin. Cell 112:725–736.

41. Muldrow, T. A., A. M. Campbell, P. A. Weil, and D. T. Auble. 1999. MOT1can activate basal transcription in vitro by regulating the distribution ofTATA binding protein between promoter and nonpromoter sites. Mol. Cell.Biol. 19:2835–2845.

42. Sabet, N., et al. 2003. Global and specific transcriptional repression by thehistone H3 amino terminus in yeast. Proc. Natl. Acad. Sci. U. S. A. 100:4084–4089.

43. Sabet, N., S. Volo, C. Yu, J. P. Madigan, and R. H. Morse. 2004. Genome-wide analysis of the relationship between transcriptional regulation byRpd3p and the histone H3 and H4 amino termini in budding yeast. Mol.Cell. Biol. 24:8823–8833.

44. Sprouse, R. O., M. Brenowitz, and D. T. Auble. 2006. Snf2/Swi2-relatedATPase Mot1 drives displacement of TATA-binding protein by grippingDNA. EMBO J. 25:1492–1504.

45. Sprouse, R. O., et al. 2008. Regulation of TATA-binding protein dynamics inliving yeast cells. Proc. Natl. Acad. Sci. U. S. A. 105:13304–13308.

46. Sprouse, R. O., et al. 2008. Function and structural organization of Mot1bound to a natural target promoter. J. Biol. Chem. 283:24935–24948.

47. van Werven, F. J., H. A. van Teeffelen, F. C. Holstege, and H. T. Timmers.2009. Distinct promoter dynamics of the basal transcription factor TBPacross the yeast genome. Nat. Struct. Mol. Biol. 16:1043–1048.

48. Venters, B. J., and B. F. Pugh. 2009. A canonical promoter organization ofthe transcription machinery and its regulators in the Saccharomyces genome.Genome Res. 19:360–371.

49. Zanton, S. J., and B. F. Pugh. 2004. Changes in genomewide occupancy ofcore transcriptional regulators during heat stress. Proc. Natl. Acad. Sci.U. S. A. 101:16843–16848.

50. Zanton, S. J., and B. F. Pugh. 2006. Full and partial genome-wide assemblyand disassembly of the yeast transcription machinery in response to heatshock. Genes Dev. 20:2250–2265.

51. Zhang, H., D. N. Roberts, and B. R. Cairns. 2005. Genome-wide dynamics ofHtz1, a histone H2A variant that poises repressed/basal promoters for acti-vation through histone loss. Cell 123:219–231.


Documents

Venters Molecular and Cellular Biology 2011 2253-2261