A Yeast Transcriptional Stimulatory Protein Similar to Human PC4*

A yeast protein has been identified that stimulates basal transcription by RNA polymerase II, binds both single- and double-stranded DNA, and interacts with both a general transcription factor and a transcriptional activator. Phosphorylation appears to regulate these interactions. The gene for the transcriptional stimulatory protein, termed TSP1 , was cloned and found to be dispensable for yeast cell viability. The deduced amino acid sequence is similar to that of mamma- lian coactivator protein PC4. A set of general transcription factors required for initiation at most RNA polymerase II promoters has been isolated from yeast and mammalian cells (reviewed in Refs. 1 and 2). All 15 polypeptides of this minimal set of protein factors exhibit sig-nificant homology between yeast and human systems. Functional attributes of the protein factors so far determined are similar in the two systems as well. Additional proteins are required for positive regulation (ac-tivation) of transcription by enhancer-binding proteins (re-viewed in Refs. 3 and 4). TAF 1 complexes from several systems support transcriptional activation in vitro (reviewed in Ref. 5). Mediator complex from yeast enables a response to activator proteins and also stimulates basal (unregulated) transcription (6, 7). A human cofactor of activated transcription, termed USA, comprises several components, including PC4, which stimulates basal transcription as well (8, 9). In the course of purifying one of the general transcription factors (TF), TFIIF, from yeast, we resolved a polypeptide stimulatory to basal transcription. We report here on the isolation and characterization of this protein and the cloning and expression of the gene that encodes it. Functional and sequence sim-ilarities to PC4 column with 8 ml of buffer F and 8 ml of buffer G-0.01, and eluted with buffer G-0.25. Surface Plasmon Resonance— Measurements of protein-protein in- teractions were performed with a BIAcore Biosensor (Pharmacia Biotech Inc.). All measurements were performed at 25 °C in running buffer (40 m M Hepes, pH 7.6, 7.5 m M MgCl 2 , 120 m M potassium acetate, 0.005% Surfactant P-20) with a flow rate of 15 (cid:109) l/min. Proteins were immobilized on CM5 research grade sensor chips using the amine coupling kit (Pharmacia). TFIIB was coupled in 100 m M sodium acetate at pH 6.0 at 0.1 mg/ml, Gal4(1–147) was coupled in running buffer at 0.11 mg/ml, and Gal4-VP16 was coupled in 100 m M sodium formate at pH 3.0 at 0.11 mg/ml. All injected proteins were diluted in running buffer from concentrated stocks prior to injection. After the injection, the chip was washed with running buffer containing 1 M potassium acetate to remove the bound protein. Protein-DNA interactions were tested on SA5 research grade sensor chips (Pharmacia) with surface-bound strepavidin. To test binding to single-stranded DNA, a 30-mer random oligonucleotide was synthesized with biotin at the 3 (cid:57) -end. To test binding to double-stranded DNA, complementary 33-base pair oligonucleotides were synthesized based on the adenovirus major late promoter sequence from (cid:50) 43 to (cid:50) 14 (23) (5 (cid:57) -TTTCTGAAGGGGGGCTATAAAAGGGGGTGGGGG-3 (cid:57) ). The oligonucleotide corresponding to the coding strand had biotin at the 3 (cid:57) -end. The single-stranded oligonucleotides were annealed form double-stranded DNA, which immobilized chip 1 (cid:109) l/min concentrations determined method of Bradford bovine serum albumin standard. SDS-polyacrylamide gel electrophoresis precipitation trichloroacetic acid sodium deoxy-cholate

A yeast protein has been identified that stimulates basal transcription by RNA polymerase II, binds both single-and double-stranded DNA, and interacts with both a general transcription factor and a transcriptional activator. Phosphorylation appears to regulate these interactions. The gene for the transcriptional stimulatory protein, termed TSP1, was cloned and found to be dispensable for yeast cell viability. The deduced amino acid sequence is similar to that of mammalian coactivator protein PC4.
A set of general transcription factors required for initiation at most RNA polymerase II promoters has been isolated from yeast and mammalian cells (reviewed in Refs. 1 and 2). All 15 polypeptides of this minimal set of protein factors exhibit significant homology between yeast and human systems. Functional attributes of the protein factors so far determined are similar in the two systems as well.
Additional proteins are required for positive regulation (activation) of transcription by enhancer-binding proteins (reviewed in Refs. 3 and 4). TAF 1 complexes from several systems support transcriptional activation in vitro (reviewed in Ref. 5). Mediator complex from yeast enables a response to activator proteins and also stimulates basal (unregulated) transcription (6,7). A human cofactor of activated transcription, termed USA, comprises several components, including PC4, which stimulates basal transcription as well (8,9).
In the course of purifying one of the general transcription factors (TF), TFIIF, from yeast, we resolved a polypeptide stimulatory to basal transcription. We report here on the isolation and characterization of this protein and the cloning and expression of the gene that encodes it. Functional and sequence similarities to PC4 and other mammalian proteins are noted.
Transcription Assays-The reconstituted transcription system was as described (11), except that the reactions contained 250 ng of plasmid DNA template, 100 ng of purified RNA polymerase II (12), and 1 l of nearly homogeneous TFIIH (13), and the total salt concentration was 115 mM potassium acetate. The template contained the yeast CYC1 promoter fused to a G-less cassette (14).
Homology Searches-The protein data bases Swiss-Prot (release 31 plus updates), PIR (release 45), and Genpept (release 91 plus updates) were searched with the amino acid sequence of Tsp1 using the BLAST program (15). To determine the degree of homology to individual proteins, percentiles were calculated by the BESTFIT program (Genetics Computer Group, Madison, WI; Ref. 16), comparing the amino acid sequences of the proteins to the corresponding homologous proteins of other organisms. The parameters were gap weight 2.0 and gap length weight 0.1. Multiple sequence alignments were performed using the CLUSTAL V computer application (17).
Null Mutation-A fragment of TSP1 containing the coding region and 757 base pairs of upstream sequence was amplified by polymerase chain reaction from genomic DNA. The fragment was cloned into pBluescript SKϩ (Stratagene), which lacked the region from EcoRV to XbaI (plasmid pPL293) with the use of XhoI to create plasmid pPL297. This plasmid was digested with EcoRV and BamHI, end-filled with Klenow enzyme, and the AatII-NaeI fragment from pRS303 bearing the HIS3 gene, end-filled with Klenow enzyme, was inserted to create pPL301.
The plasmid was digested with XhoI, which cut in the polylinker but not in the cloned gene, and transformed into the Saccharomyces cerevisiae strain CRY3 (MAT a/␣ ade 2-1 can 1-100 his 3-11 15 leu 2-3 112 trp 1-1 ura 3-1; Ref. 18). Yeast genomic DNA was isolated, and Southern blot analysis was performed to verify that the HIS3 gene had recombined correctly into the gene. The strain harboring the null mutation was induced to undergo meiosis. Tetrads were dissected by micromanipulation on YPD agar (19) and replica plated to synthetic medium lacking histidine.
Antibody Production-For expression of a C-terminal fragment of Tsp1 (amino acids 106 -292) in bacteria, the coding region of the gene was amplified by polymerase chain reaction using the primers 5Ј-ATATCATATGGGATCCAAGAGACCAA-3Ј and 5Ј-ATATGAATTCTC-GAGTTCTTCTTCACTTATGT-3Ј (NdeI and XhoI sites underlined). The * This research was supported by National Institutes of Health Grant GM36659 (to R. D. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) Z48502.
‡ Howard Hughes Medical Institute Predoctoral Fellow. § Supported by National Institutes of Health Training Grant GM07276-20.
¶ To whom correspondence should be addressed. Tel.: 415-723-6988; Fax: 415-723-8464. 1 The abbreviations used are: TAF, TATA-binding protein-associated factor; TF, transcription factor. resulting fragment was cloned into the pET20b expression plasmid (Novagen), which incorporated a six-histidine tag at the C terminus of the protein to create pPL299.
Escherhichia coli BL21 (DE3) pLysS strain (Stratagene) was transformed with pPL299, grown in Luria broth (2 liters) supplemented with 100 g/ml ampicillin and 25 g/ml chloramphenicol at 30°C to an A 600 of 0.4, and induced with 0.1 mM isopropyl-1-thio-␤-D-galactopyranoside for 2 h. The cells were harvested by centrifugation, suspended in 40 ml of buffer, and centrifuged for removal of cellular debris. The soluble fraction was incubated for 2 h at 4°C with 5 ml of Ni 2ϩ -NTA-agarose resin (Qiagen) equilibrated in buffer B, followed by washing in a column with buffer B, buffer C, buffer D-0.02 (buffer D containing 0.02 M imidazole), and elution with buffer D-0.1. The peak fraction was injected into a rabbit (Berkeley Antibody Co.), and polyclonal serum was collected 10 -14 days after immunization.
The C-terminal fragment of Tsp1 was also coupled to cyanogen bromide-activated Sepharose (Sigma) and used to affinity purify anti-Tsp1 polyclonal sera as described (20). Western blots were performed according to Chasman and Kornberg (21). The dilutions of the affinitypurified anti-Tsp1 antibody and the goat anti-rabbit IgG (conjugated to alkaline phosphatase; Bio-Rad) secondary antibody were both 1/1000.
Baculovirus Expression and Recombinant Protein Purification-To express full-length Tsp1 in baculovirus, the coding region of the gene was amplified by polymerase chain reaction using the primers 5Ј-ATATCTGCAGATGCACCATCATCACCATCACTCATATTACAACA-GGTAT-3Ј and 5Ј-ATATGGATCCTTATTCTTCTTCACTTAT-3Ј (PstI and BamHI sites underlined). The resulting fragment was end-filled with T4 DNA polymerase, digested with PstI, and cloned into the pVL1392 expression plasmid (Invitrogen) digested with PstI and SmaI, which incorporated a hexahistidine tag at the amino terminus of the protein, to create pPL321. To place the first ATG of the TSP1 coding region out-of-frame with respect to the polyhedrin start codon, the plasmid was digested with PstI, end-filled with Klenow, and religated to form plasmid pPL324.
The expression plasmid was cotransfected with BaculoGold-linearized baculovirus DNA (Pharmingen) into Sf9 cells according to manufacturer's instructions. The virus in the transfection supernatant was amplified on 3 ϫ 10 6 Sf9 cells in a T25 flask for 4 days and then in suspension in 200 ml of Sf9 cells (Life Technologies, Inc.) at a concen-tration of 2 ϫ 10 6 cells/ml for 3 days. Half of the final supernatant was added to Sf21 cells at a concentration of 2 ϫ 10 6 cells/ml grown in Sf900-II medium (1 liter) supplemented as described. The cells were grown in suspension at 27°C for 39 h and harvested by centrifugation, and all subsequent manipulations were performed at 4°C. Following resuspension of the cells (10.5 g) in buffer E (2 ml/g of cells), they were lysed with 15 strokes of a tight fitting pestle in a Dounce homogenizer. Sucrose was added to a final concentration of 10% and sodium chloride to a concentration of 0.5 M. The extract was rotated slowly for 30 min and centrifuged in a 70Ti rotor (Beckman) for 90 min at 31,000 rpm. Tween 20 was added to the supernatant to a concentration of 0.2%, followed by incubation for 2.5 h with 4 ml of TALON resin (Clontech) equilibrated in buffer F and washing in a column at 60 ml/h with 60 ml of buffer F, 60 ml of buffer G-0.01 (buffer G containing 0.01 M imidazole), and 60 ml of buffer G-0.02. Some Tsp1 eluted with buffer G-0.02, but a majority of the protein was recovered by elution with buffer G-0.1. A fraction of the buffer G-0.1 eluate was applied at 0.5 ml/min to a Bio-Gel SP-5-PW high performance liquid chromatography column (75 ϫ 7.5 mm; Bio-Rad) equilibrated in buffer H-0.2 (buffer H containing 0.2 M potassium acetate). The column was washed with buffer H-0.4, and developed with a linear gradient (24 ml) to buffer H-0.8. Tsp1 eluted in two peaks, a minor one at 530 mM potassium acetate and a major one at 600 mM potassium acetate.
Dephosphorylated Tsp1-Tsp1 (500 g), which eluted from the SP-5-PW column, was treated with calf intestine alkaline phosphatase (Boehringer Mannheim) in 0.1 mM Tris, pH 7.6, 2 mM MgCl 2 , and 0.1 mM ZnCl 2 at 37°C for 30 min. Imidazole was added to 2 mM, and Tween 20 was added to 0.2%. The mixture was incubated with 400 l of TALON resin equilibrated in buffer F at 4°C for 2 h, washed in a Surface Plasmon Resonance-Measurements of protein-protein interactions were performed with a BIAcore Biosensor (Pharmacia Biotech Inc.). All measurements were performed at 25°C in running buffer (40 mM Hepes, pH 7.6, 7.5 mM MgCl 2 , 120 mM potassium acetate, 0.005% Surfactant P-20) with a flow rate of 15 l/min. Proteins were immobilized on CM5 research grade sensor chips using the amine coupling kit (Pharmacia). TFIIB was coupled in 100 mM sodium acetate at pH 6.0 at 0.1 mg/ml, Gal4(1-147) was coupled in running buffer at 0.11 mg/ml, and Gal4-VP16 was coupled in 100 mM sodium formate at pH 3.0 at 0.11 mg/ml. All injected proteins were diluted in running buffer from concentrated stocks prior to injection. After the injection, the chip was washed with running buffer containing 1 M potassium acetate to remove the bound protein.
Protein-DNA interactions were tested on SA5 research grade sensor chips (Pharmacia) with surface-bound strepavidin. To test binding to single-stranded DNA, a 30-mer random oligonucleotide was synthesized with biotin at the 3Ј-end. To test binding to double-stranded DNA, complementary 33-base pair oligonucleotides were synthesized based on the adenovirus major late promoter sequence from Ϫ43 to Ϫ14 (23) (5Ј-TTTCTGAAGGGGGGCTATAAAAGGGGGTGGGGG-3Ј). The oligonucleotide corresponding to the coding strand had biotin at the 3Ј-end. The single-stranded oligonucleotides were annealed to form doublestranded DNA, which was immobilized to the chip by flowing over the surface at a rate of 1 l/min at a concentration of 0.1 mg/ml.
Other Methods-Protein concentrations were determined by the method of Bradford (24) using bovine serum albumin as standard. Proteins were concentrated for SDS-polyacrylamide gel electrophoresis (25) by precipitation with 10% trichloroacetic acid using sodium deoxycholate as carrier and were visualized with Coomassie Blue R-250. Growth of yeast and E. coli and standard DNA manipulations were as described (26,27).

Isolation of Transcriptional Stimulatory
Protein-After six steps of fractionation from a crude yeast extract, general transcription factor TFIIF was nearly homogeneous, with a single major contaminant of apparent molecular mass 43 kDa (10). Resolution of these proteins was achieved by further chromatography on SP-5-PW, with TFIIF and the contaminant (p43) eluting at potassium acetate concentrations of 425 and 600 mM, respectively. The specific activity of TFIIF was diminished upon separation from p43, and addition of pure p43 (Fig. 1A) to transcription mixtures reconstituted with homogeneous TFIIF gave a 3-5-fold stimulation of the reaction (Fig. 1B). In the presence of mediator complex and the activator protein Gal4-VP16, a further stimulation was observed (data not shown). By contrast, addition of TFIIA, stimulatory in some systems, was without effect.
Cloning and Characterization of Transcriptional Stimulatory Protein Gene-Amino acid sequences from two tryptic peptides of p43 were present in an open reading frame on S. cerevisiae chromosome XIII, which encodes a protein of 292 amino acids, with a molecular mass of 33.1 kDa and pI of 5.31 (Fig. 2). Half of the amino acids are charged, divided almost equally between acidic and basic residues. Major sequence motifs include a single-stranded DNA binding domain and a number of potential phosphorylation sites for casein kinase II and protein kinase C. Blot hybridization of total yeast DNA with a fragment of the coding region was consistent with a single copy in the yeast genome (data not shown). In view of the effect of p43 upon transcription in vitro, we refer to the gene as TSP1 (for transcriptional stimulatory protein 1).
BLAST searches of National Institutes of Health data bases identified homologies with three mammalian proteins, a rat pancreatic B-cell protein (28), mouse p9 (29), and human PC4  Gal4(1-147) was bound to the surface, and Tsp1 proteins were injected at a concentration of 20 g/ml. All curves have been corrected for bulk refractive index changes by subtracting injections made over blank surfaces. (30,31). The three mammalian proteins were 61% similar and 38% identical to the amino-terminal region of Tsp1 (Fig. 3) and contained single-stranded DNA binding motifs at their carboxyl termini (31) but were considerably smaller than the yeast protein (9 -15 kDa, compared with 33 kDa for Tsp1).
A yeast strain was constructed in which a portion of TSP1 was deleted and replaced with HIS3 (see legend for Fig. 2). The consequences of this gene disruption were determined by sporulation and tetrad dissection (data not shown). All four spores of each tetrad formed colonies on rich medium, indicating that TSP1 is not required for yeast cell viability. An additional knockout strain was constructed in which the entire coding region of TSP1 and 370 base pairs upstream was replaced with LEU2, and dissection of this strain confirmed that the TSP1 gene is not essential (data not shown). Wild-type and tsp1 null strains grew at similar rates at 30°C on various carbon sources, including glucose, galactose, and glycerol, and at temperatures up to 37°C (data not shown).
Recombinant Tsp1 with a hexahistidine tag at the amino terminus was expressed in a baculovirus system and purified to homogeneity (Fig. 4A). Two forms of the protein were resolved in the final chromatographic step, one of which was less abundant and phosphorylated to a greater extent than the other (data not shown). The carboxyl-terminal portion of Tsp1 (residues 106 -292), which lacks homology with mammalian proteins, was also expressed in bacteria and used for production of polyclonal antibodies. Affinity-purified anti-Tsp1 antibodies reacted both with p43 purified from yeast and with recombinant Tsp1 from baculovirus (Fig. 4B), whereas preimmune sera failed to react with either protein (data not shown). The apparent molecular weight of the recombinant protein was slightly greater than that of p43 from yeast, perhaps due to the addition of the hexahistidine tag or to post-translational modifications. We conclude that p43 is the product of the TSP1 gene.
Tsp1 Phosphorylation and Interaction with Other Transcription Proteins-In view of previous studies of phosphorylation of PC4, we investigated the effects of phosphorylation on Tsp1. Recombinant Tsp1 from baculovirus was treated with alkaline phosphatase and purified by immobilized metal affinity chromatography, resulting in a shift to a faster migrating species in SDS-polyacrylamide gel electrophoresis (Fig. 4C). Dephosphorylation had little effect on the transcriptional stimulatory activity of Tsp1, either in the presence (not shown) or in the absence (Fig. 4D) of mediator and Gal4-VP16.
Interactions of Tsp1 with other transcription proteins and with DNA were investigated by surface plasmon resonance. No interaction with TFIIA was detected (data not shown), in contrast to the behavior of human PC4 (9). No direct interaction between Tsp1 and TFIIF was observed (data not shown), indicating that the co-chromatography of the proteins was coincidental. Dephosphorylated recombinant Tsp1 interacted with TFIIB and with the activator protein Gal4-VP16, whereas the phosphorylated form of Tsp1 did not bind to TFIIB and interacted with Gal4-VP16 almost 10-fold less tightly (Fig. 5, A and  B). Conversely, phosphorylated but not dephosphorylated Tsp1 interacted with the Gal4 DNA binding domain of Gal4-VP16 (Fig. 5C). Similar results were obtained with p43 isolated from yeast, which bound Gal4-VP16 about half as well as dephosphorylated recombinant Tsp1, perhaps reflecting partial phos-FIG. 6. Interaction of Tsp1 with DNA. A, dephosphorylated Tsp1 (dotted line) interacts with single-stranded DNA more strongly than does phosphorylated Tsp1 (solid line). Random 30-nucleotide oligonu-cleotides were bound to the surface, and Tsp1 proteins were injected at a concentration of 20 g/ml. B, dephosphorylated Tsp1 (dotted line) interacts with double-stranded DNA more strongly than does phosphorylated Tsp1 (solid line). A 33-base pair oligonucleotide corresponding to the TATA region of the adenovirus major late promoter was bound to the surface, and Tsp1 proteins were injected at a concentration of 20 g/ml. phorylation, and which failed to bind the Gal4 DNA binding domain (data not shown). While further studies of the ionic strength dependence of these interactions are needed to assess the contribution of electrostatic effects and the question of specificity, binding to TFIIB is likely to be nonionic, since the positively charged TFIIB interacted preferentially with the less electronegative, dephosphorylated form of Tsp1. Binding to Gal4-VP16 appears to involve the VP16 activation domain, since the dephosphorylated form of Tsp1 interacted preferentially, and this form did not bind the Gal4 domain. Binding to the VP16 domain could be electrostatic in view of the preference for the less electronegative form of Tsp1 for interaction with the highly negatively charged VP16.
Because TSP1 encodes a possible single-stranded DNA binding domain, interactions with both single-and double-stranded DNA were investigated. Approximately equal amounts of either a random 30-mer oligodeoxyribonucleotide or a 33-base pair DNA including the adenovirus major late promoter were coupled to a plasmon resonance sensor surface through biotinavidin interaction. Dephosphorylated recombinant Tsp1 bound strongly to single-stranded DNA and much less well to doublestranded DNA (compare amplitudes of surface plasmon resonance signals in Fig. 6, A and B). The phosphorylated form of Tsp1 bound more weakly, and again, p43 isolated from yeast behaved similarly to the dephosphorylated form of the recombinant protein (data not shown).

DISCUSSION
Tsp1 exhibits both sequence homology and functional similarities to human PC4. Functional studies indicate a division of PC4 in two parts, an amino-terminal region rich in serine residues, required for transcriptional activity, and a carboxylterminal region, which includes a DNA binding domain but which is dispensable for function in vitro (31). The aminoterminal region shares sequence homology with viral immediate early proteins involved in transcriptional regulation, such as IE62, ICP4, and IE180 of Varicella-Zoster virus, herpes simplex virus type 1, and pseudorabies virus, respectively (discussed in Ref. 31). Phosphorylation of serine residues in the amino-terminal region negatively regulates transcriptional activity and affects interactions with the VP16 activation domain and with double-stranded DNA (30 -32). The sequence similarity between Tsp1 and PC4 includes the entirety of PC4, and Tsp1 shares, in addition, the capacity to stimulate basal transcription severalfold, interaction with the activation domain of Gal4-VP16 and with DNA, and modulation of these interactions by phosphorylation.
Tsp1 differs from PC4 in three regards. First, it is more than twice the size, containing a carboxyl-terminal extension of unknown significance, and further differs in containing fewer amino-terminal serine residues (5 serines in 17 residues of the yeast protein, compared with 9 of 16 and 8 of 13 in the two SEAC domains of the human protein; see Ref. 31). Second, Tsp1 fails to bind TFIIA, in contrast with PC4 (9). Finally, Tsp1 had no effect on the level of activated transcription in the yeast system, whereas PC4 enables as much as 90-fold transcriptional enhancement in the human system (9). The difference in regard to transcriptional activation may relate to the use of mediator complex but not TAFs for studies of activation in the yeast system and TAFs but not mediator in the human system. While TAFs are also present in yeast, their actions may be promoter specific, and TAF-dependent promoters remain to be identified, so investigation of Tsp1 in regard to TAF-dependent transcriptional activation is not yet possible.
Following completion of this manuscript, an independent characterization of Tsp1 was reported (33). The protein, termed Sub1, was identified as a suppressor of a cold-sensitive TFIIB mutation, and overexpression of Sub1 was shown to stimulate transcriptional activation in vivo. It was further demonstrated that the protein interacted with TFIIB in vitro and inhibited the formation of TBP-TFIIB-promoter complexes. The genetic characterization of Sub1 complements our biochemical analysis and further suggests that this protein may play a role in transcriptional activation.