TATA-Binding Protein Mutants That Are Lethal in the Absence of the Nhp6 High-Mobility-Group Protein

Abstract

The Saccharomyces cerevisiae Nhp6 protein is related to the high-mobility-group B family of architectural DNA-binding proteins that bind DNA nonspecifically but bend DNA sharply. Nhp6 is involved in transcriptional activation by both RNA polymerase II (Pol II) and Pol III. Our previous genetic studies have implicated Nhp6 in facilitating TATA-binding protein (TBP) binding to some Pol II promoters in vivo, and we have used a novel genetic screen to isolate 32 new mutations in TBP that are viable in wild-type cells but lethal in the absence of Nhp6. The TBP mutations that are lethal in the absence of Nhp6 cluster in three regions: on the upper surface of TBP that may have a regulatory role, near residues that contact Spt3, or near residues known to contact either TFIIA or Brf1 (in TFIIIB). The latter set of mutations suggests that Nhp6 becomes essential when a TBP mutant compromises its ability to interact with either TFIIA or Brf1. Importantly, the synthetic lethality for some of the TBP mutations is suppressed by a multicopy plasmid with SNR6 or by an spt3 mutation. It has been previously shown that nhp6ab mutants are defective in expressing SNR6, a Pol III-transcribed gene encoding the U6 splicing RNA. Chromatin immunoprecipitation experiments show that TBP binding to SNR6 is reduced in an nhp6ab mutant. Nhp6 interacts with Spt16/Pob3, the yeast equivalent of the FACT elongation complex, consistent with nhp6ab cells being extremely sensitive to 6-azauracil (6-AU). However, this 6-AU sensitivity can be suppressed by multicopy SNR6 or BRF1. Additionally, strains with SNR6 promoter mutations are sensitive to 6-AU, suggesting that decreased SNR6 RNA levels contribute to 6-AU sensitivity. These results challenge the widely held belief that 6-AU sensitivity results from a defect in transcriptional elongation.

The TATA-binding protein (TBP) is required for all eukaryotic transcription, whether the genes are transcribed by RNA polymerase I (Pol I), Pol II, or Pol III (28). TBP binding to promoters of genes transcribed by Pol II correlates with transcriptional activity; therefore, regulation of TBP binding may be the critical event in determining transcriptional activation (36, 40). Full-length TBP binds DNA slowly in a multistep process (29, 69), and crystallographic studies demonstrate that TBP bends DNA sharply upon binding (46, 47). Transcriptional activation by Pol II requires the assembly of a complex of general transcription factors at a promoter (18, 27). It is believed that transcriptional coactivators function by stimulating DNA binding by TBP and facilitating formation of a complex between TBP and the general transcription factors TFIIA and TFIIB.

The Spt3 factor can regulate TBP binding to promoters. Spt3, which is part of the SAGA histone acetyltransferase complex (63), physically interacts with TBP and plays an important role in transcriptional start site selection for RNA Pol II (19). Spt3 can act as a positive or negative transcriptional regulator, depending on the promoter (5, 17, 67). Although Spt3 acts to promote TBP binding at the GAL1 promoter (17), we have shown that Spt3 inhibits TBP binding to HO (67). Genetic interactions have been seen between SPT3 and other regulators of TBP binding (4, 13, 43).

The Saccharomyces cerevisiae Nhp6 protein is similar to the high-mobility-group B class of architectural DNA-binding proteins (1). Nhp6 is an abundant protein (50), and it has multiple roles in transcription, including transcriptional initiation and elongation by Pol II and promoting transcription by Pol III. For its role in elongation, Nhp6 interacts genetically and biochemically with Spt16/Pob3 (10, 22), the yeast equivalent of FACT that promotes elongation through chromatin templates (49). Nhp6 is required for Spt16/Pob3 to bind to nucleosomes (22, 54).

Nhp6 is encoded by two genes, NHP6A and NHP6B, that are functionally redundant, as only the nhp6ab double mutants show any observable phenotype. nhp6ab mutants are unable to grow at 37°C, but this growth defect can be suppressed by a multicopy plasmid with either SNR6 or BRF1 (35). SNR6 encodes the U6 RNA required for mRNA splicing, and it is suggested that a deficiency in SNR6 RNA contributes to the temperature-sensitive growth defect seen in nhp6ab mutants. Brf1 is the limiting component in TFIIIB, a factor required for Pol III transcription (56); therefore, BRF1 overexpression could increase SNR6 expression and facilitate growth of the nhp6ab mutant at 37°C. Overexpression of TBP also suppresses the temperature-sensitive growth defect of an nhp6ab mutant (67). In addition to its well-documented role in Pol II transcription, TBP is also a component of the RNA Pol III factor TFIIIB and is required for Pol III transcription (33). Thus, TBP overexpression could suppress the nhp6ab growth defect by affecting either Pol II or Pol III transcription. Data from Paull et al. (50) suggest that Nhp6 stimulates transcription by promoting formation of preinitiation complexes.

Our genetic studies suggest that Nhp6 and the Gcn5 histone acetyltransferase function in parallel to activate expression of the yeast HO gene (67, 68). A gcn5 nhp6a nhp6b triple mutant is extremely sick, but this growth defect can be suppressed by mutations in SIN4, SPT3, or SPT8. The data suggested that TBP was the critical target, as overexpression of TBP suppresses the temperature-sensitive growth defect of nhp6ab mutants and partially restores HO expression in the absence of either Nhp6 or Gcn5. In this study, we continue the genetic analysis examining the relationship between Nhp6 and TBP. We performed a novel synthetic lethal screen to identify strains with TBP mutations that are unable to grow in the absence of Nhp6. We also find that a multicopy plasmid with SNR6 can suppress some of the TBP nhp6ab synthetic lethalities. Our results suggest that Nhp6 and TBP work together to facilitate transcriptional activation by both Pol II and Pol III.

MATERIALS AND METHODS

Strains and media.

All S. cerevisiae strains used are listed in Table 1 and are isogenic in the W303 background (65). Standard genetic methods were used for strain construction (53, 59). W303 strains with disruptions in gcn5, nhp6a, nhp6b, and swi2 have been described previously (67, 68). The TBP-Myc-tagged allele (51) was provided by Rick Young and crossed to generate the strains used here. Cells were grown in yeast extract-peptone-dextrose (YEPD) medium (59) at 30°C, except where the use of other temperatures is noted, or where synthetic complete medium with 2% glucose supplemented with adenine, uracil, and amino acids, as appropriate, but lacking essential components was used to select for plasmids. Medium containing 5-fluoroorotic acid (5-FOA) was prepared as described previously (8).

TABLE 1.

S. cerevisiae strains used in this study

Strain	Relevant genotype
DY150	MATaade2 can1 his3 leu2 trp1 ura3
DY151	MATα ade2 can1 his3 leu2 trp1 ura3
DY2381	MATα nhp6a::URA3 nhp6b::HIS3 ade2 can1 his3 leu2 trp1 ura3
DY3398	MATaade2 can1 his3 leu2 trp1
DY6439	MATα nhp6a::URA3 nhp6b::ADE2 ade2 can1 his3 leu2 trp1 ura3
DY7242	MATaspt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 ade3 can1 his3 leu2 trp1 ura3
DY7244	MATα nhp6a::KanMX nhp6b::HIS3 spt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 ade3 can1 his3 leu2 lys2 trp1 ura3
DY7472	MATaspt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 can1 his3 leu2 trp1 ura3
DY7474	MATaspt3::ADE2 spt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 can1 his3 leu2 trp1 ura3
DY7588	MATanhp6a::URA3 nhp6b::ADE2 ade2 can1 his3 leu2 lys2 trp1 ura3
DY7723	MATanhp6a::KanMX nhp6b::HIS3 spt3::ADE2 spt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 can1 his3 leu2 trp1 ura3
DY8162	MATα nhp6a::KanMX nhp6b::HIS3 spt15::ADE2 + SPT15(YCp-URA3-ADE3) ade2 ade3 can1 his3 leu2 lys2 trp1 ura3
DY8356	MATanhp6a::KanMX nhp6b::HIS3 spt15::LEU2 + SPT15(YCp-URA3-ADE3) ade2 can1 his3 leu2 lys2 trp1 ura3
DY8360	MATaSPT15-MYC(18)::TRP1 nhp6a::KanMX nhp6b::HIS3 ade2 can1 his3 leu2 trp1 ura3
DY8408	MATaSPT15-MYC(18)::TRP1 ade2 can1 his3 leu2 lys2 trp1 ura3
DY8409	MATα SPT15-MYC(18)::TRP1 ade2 can1 his3 leu2 trp1 ura3
DY8855	MATanhp6a::KanMX nhp6b::ADE2 spt3::TRP1 ade2 can1 his3 leu2 trp1
DY8858	MATα snr6::LEU2 + SNR6-wild type(YCp-TRP1) ade2 can1 his3 leu2 lys2 met2 trp1
DY8859	MATα snr6::LEU2 + SNR6-Δ42(YCp-TRP1) ade2 can1 his3 leu2 lys2 met2 trp1
DY8860	MATα snr6::LEU2 + SNR6-T14A(YCp-TRP1) ade2 can1 his3 leu2 lys2 met2 trp1
DY8861	MATα snr6::LEU2 + SNR6-T5-flip(YCp-TRP1) ade2 can1 his3 leu2 lys2 met2 trp1
DY8980	MATaspt3::TRP1 ade2 can1 his3 leu2 lys2 met15 trp1

Plasmids.

All plasmids used are listed in Table 2, except for the TBP mutant plasmids, which were all YCp-TRP1 plasmids. A 1.2-kb XhoI-PstI fragment with SNR6 from pRS314-SNR6, provided by Dave Brow, was cloned into pRS425 and pRS422, constructing M4479 and M4488, respectively. A 3.8-kb BamHI fragment with ADE3 from plasmid pTF96, provided by Tim Formosa, was cloned into BamHI-cleaved plasmid pDE28-6 (19), generating M4373.

TABLE 2.

Plasmids used in this study^a

Plasmid	Description	Reference or source
pRS422	YEp-ADE2 vector	9
pRS423	YEp-HIS3 vector	60
pRS425	YEp-LEU2 vector	60
pRS314	YCp-TRP1 vector	60
YEplac195	YEp-URA3 vector	25
pRS426-SNR6	SNR6 in YEp-URA3 plasmid	Dave Brow
M4479	SNR6 in YEp-LEU2 plasmid	This work
M4488	SNR6 in YEp-ADE2 plasmid	This work
pRS423/B70	BRF1 in YEp-HIS3 plasmid	56
pDE28-6	TBP (SPT15) in YCp-URA3 plasmid	19
pTM8	TBP (SPT15) in YCp-TRP1 plasmid	34
M4373	TBP (SPT15) in YCp-URA3-ADE3 plasmid	This work
pSH346	TFIIA (TOA1 and TOA2) in YEp-LEU2 plasmid	Steve Hahn

^aAll plasmids used in this study except for the TBP mutant plasmids which were all YCp-TRP1.

Screen for TBP mutants.

Strain DY7244 was transformed with a library of mutagenized TBP genes on YCp-TRP1 plasmids (2), generously provided by Karen Arndt, and we screened for solid red (nonsectoring) colonies that could not lose the wild-type TBP gene (the YCp-URA3-ADE3-SPT15 plasmid). The solid red colonies were plated onto fresh plates to verify the nonsectoring phenotype, and growth on medium containing FOA confirmed the inability of these strains to lose the YCp-ADE3-URA3-TBP plasmid. The majority of these nonsectoring colonies contained lethal TBP mutations, and these mutations were eliminated by mating the strain to strain DY7472. Plasmids which produced a nonsectoring and 5-FOA-sensitive phenotype in the NHP6/nhp6ab diploid were discarded, while plasmids with a sectored, 5-FOA-resistant phenotype were retained. Plasmids were purified after passage through Escherichia coli, retransformed into nhp6ab and NHP6⁺ strains to verify the phenotypes, and then sequenced.

ChIP.

Chromatin immunoprecipitation (ChIP) was performed as described previously (7) using either 9E11 monoclonal antibody to the Myc epitope or polyclonal anti-TBP serum provided by Laurie Stargell. Multiplex PCR was performed with oligonucleotides specific to the HO and PGK1 promoter and the TRA1 open reading frame, and products were visualized on ethidium bromide-stained 2.6% MetaPhor agarose (BioWhittaker) gels. Quantitative ChIP was performed by real-time PCR with a LightCycler (Roche) and primers specific to SNR6 or TRA1. The amounts of specific target DNA regions amplified in immunoprecipitated samples were determined with LightCycler software (version 3.5; Roche) by comparing the PCR logarithmic amplification threshold (crossing point) values for ChIP DNAs versus a standard dilution series of input samples prior to IP. Each PCR was performed in triplicate, and the normalized mean and standard deviation of the ratio of SNR6 to TRA1 values was calculated according to equation 7 of van Kempen and van Vliet (66) to determine the relative enrichment of the specific target versus the nontarget control. Results of at least two independent ChIP reactions are reported here. The sequences of the PCR primers are as follows: for SNR6, CTGGCATGAACAGTGGTAAA and GGGGAACTGCTGATCATCTC; for TRA1, CTGAGTATGCACTATGGGAA and CTTGATCTCCTTTTCTGCTT; and for PGK1, CCAGAGCAAAGTTCGTTCGA and GCTTGTCCTTCAAGTCCAAA.

Other methods.

For the SNR6 Northern blot, RNA was separated on an 8% polyacrylamide gel containing 8 M urea, transferred to a Biodyne B nylon membrane (Pall), and probed with labeled oligonucleotides specific for SNR6 (CCTTATGCAGGGGAACTGCTGATC) and SNR128 (CCGAGAGTACTAACGATGGGTTCGTAAGCGTACTCC) RNAs. The Northern blots were quantitated using ImageQuant software and a Molecular Dynamics PhosphorImager. The Hi Lithium method (31) was used for yeast transformations. For dilution plating assays, cells were grown to saturation in either rich or selective medium (depending on the plasmid) and washed with water, and then aliquots of diluted samples (10-fold dilutions) were plated on appropriate medium.

RESULTS

Reduced TBP binding to the SNR6 promoter in nhp6ab mutant cells.

To investigate whether Nhp6 affects TBP binding to promoters in vivo, we performed ChIP assays to compare TBP binding in NHP6 and nhp6ab cells. We used PCR to examine the amount of DNA from a number of promoters transcribed by Pol II, including HO, HIS3, TRP3, URA1, ELP3, STE3, ADH2, MET14, MFA2, SER3, VID28, VPS38, and XRS2. However, none of these promoters showed a strong difference in TBP binding when immunoprecipitated material from wild-type and nhp6ab cells were compared (data not shown). It is possible that the nhp6ab mutation affects TBP binding but only modestly or that TBP binding to other Pol II promoters that we have not tested is affected.

We next examined TBP binding to the SNR6 promoter in the nhp6ab mutant. Reduced expression of SNR6, transcribed by RNA Pol III, in an nhp6ab mutant is at least partially responsible for the 37°C growth defect, as a multicopy plasmid with SNR6 can at least partially suppress this growth defect (35, 42). We constructed isogenic strains with a TBP-Myc-tagged allele, either NHP6 or nhp6ab, and performed ChIP experiments. The experiment was done in duplicate, using cells grown at 30°C. Real-time PCR was used to quantitate the amount of SNR6 DNA in the immunoprecipitated chromatin. As shown in Fig. 1A, TBP-Myc shows strong binding to SNR6, compared to the untagged control strains. Importantly, binding of TBP-Myc to SNR6 was significantly reduced in the nhp6ab mutant.

FIG. 1.

TBP binding to SNR6 is reduced in nhp6ab mutant. (A) ChIP was performed with untagged strains, TBP-Myc-tagged NHP6 strains, and TBP-Myc-tagged nhp6ab strains. Real-time PCR quantitation of TBP-Myc binding to the SNR6 gene shows reduced binding in the nhp6ab mutant. The units are arbitrary units after normalization to a TRA1 internal control. Strains DY150, DY151, DY8408, DY8409, DY8360, and DY8361 were used. (B) After an aliquot of NHP6 and nhp6ab strains grown at 30°C were harvested, the remaining cells were shifted to 37°C for 60 min and then harvested. Chromatin samples were prepared and immunoprecipitated with polyclonal anti-TBP antibody. Serial twofold dilutions were used to assess binding of native TBP to PGK1 (positive control), TRA1 (negative control), and SNR6 by multiplex PCR. Lanes 1 to 3, input controls; lanes 4 to 6, ChIP from 30°C samples; lane 7, mock IP from 30°C sample; lanes 8 to 10, ChIP from 37°C samples; lane 11, mock IP from 37°C sample. Strains DY151 and DY2381 were used.

The TBP-Myc-tagged nhp6ab strain shows a marked growth defect, and we have previously reported synthetic lethality between nhp6ab and hemagglutinin-tagged TBP alleles (67). Therefore, we wanted to confirm that TBP binding to SNR6 is reduced in the nhp6ab mutant, but using a polyclonal TBP antibody instead of an epitope-tagged strain. nhp6ab mutants do not grow at 37°C, and Kruppa et al. (35) showed that shifting an nhp6ab strain to 37°C results in reduced SNR6 RNA levels. Thus, we grew wild-type and nhp6ab cells at 30°C and then shifted them to 37°C for 60 min. More than 85% of nhp6ab cells retain viability after 60 min at 37°C. Samples were taken for ChIP analysis both before and after the temperature shift. In addition to testing for TBP binding to SNR6, TBP binding to the PGK1 promoter and TRA1 open reading frame were assessed as positive and negative controls, respectively. The ChIP material was analyzed by multiplex PCR, with products separated on gels. The results show clearly that the occupancy of the SNR6 TATA element by TBP is Nhp6 dependent (Fig. 1B). Quantitation of the SNR6 ChIP signal, normalized to the PGK1 control, shows TBP binding to SNR6 of 3 to 5.5 binding units (arbitrary units) of that in wild-type cells, while TBP binding to SNR6 is reduced in the nhp6ab mutant to about 1 binding unit, the SNR6/PGK1 ratio seen in the input DNA. We conclude that TBP binding to SNR6 is reduced in an nhp6ab mutant.

Multicopy SNR6 and BRF1 suppress the 6-AU sensitivity of nhp6ab mutants.

The uracil analog 6-azauracil (6-AU) inhibits transcriptional elongation by causing imbalances in the pools of ribonucleotide triphosphates (20, 57), and some believe that mutants sensitive to 6-AU are defective in elongation. An nhp6ab mutant is very sensitive to 6-AU (10, 22), and a concentration of 25 μg of 6-AU per ml is sufficient to effectively inhibit growth (Fig. 2A). We wanted to know whether specific genes on multicopy plasmids will suppress this 6-AU sensitivity.

FIG. 2.

Multicopy BRF1 or SNR6 suppresses the 6-AU sensitivity of nhp6ab mutants. (A) Dilutions of strain DY6439 transformed with the indicated plasmid were plated on plates containing complete medium or medium containing 6-AU (25 μg/ml) and incubated at 30°C for 2 days (complete medium) or 6 days (medium containing 6-AU). Plasmids pRS423 (YEp-HIS3 vector), pRS423/B70 (YEp-BRF1), pRS425 (YEp-LEU2 vector), M4479 (YEp-SNR6), and M4480 (YEp-TBP) were used. (B) Dilutions of strains DY8858 (SNR6 wild type), DY8859 (SNR6-Δ42), DY8860 (SNR6-T14A), and DY8861 (SNR6-T5-flip), were plated on plates containing complete medium or medium containing 6-AU (75 μg/ml) and incubated at 30°C for 2 days (complete medium) or 6 days (medium containing 6-AU).

Nhp6 is required for efficient transcription of the SNR6 gene by RNA Pol III (35, 42, 44). SNR6, the gene for U6 RNA, and BRF1, encoding a Pol III transcription factor, act as multicopy suppressors of the temperature-sensitive phenotype of nhp6ab mutants (35). A multicopy plasmid with TBP also suppresses the temperature-sensitive growth defect of an nhp6ab mutant (67). We determined whether SNR6, BRF1, or TBP, when present on a multicopy plasmid, could suppress 6-AU sensitivity of the nhp6ab mutant (Fig. 2). The results show strong suppression from either overexpression of the BRF1 Pol III factor or increased copies of the SNR6 structural gene. This is a surprising result, suggesting that decreased SNR6 expression contributes to the 6-AU sensitivity. Interestingly, overexpression of TBP does not suppress the 6-AU sensitivity.

It is possible that reduced SNR6 RNA levels contribute to the sensitivity to 6-AU, as the steady-state levels of SNR6 RNA are lower in an nhp6ab mutant than in wild-type cells (42). To test this idea, we constructed strains with mutations in the SNR6 promoter that reduce expression, and these mutations result in a 6-AU-sensitive phenotype (Fig. 2B). Interestingly, all three of these SNR6 promoter mutations are lethal in the absence of Nhp6 (44). It is possible that the decrease in SNR6 RNA affects mRNA splicing, and this results in 6-AU sensitivity (see Discussion). Nonetheless, these experiments clearly show that sensitivity to 6-AU can result from a mutation in a gene not having a direct role in transcriptional elongation.

Screen for TBP mutations that are lethal in the absence of Nhp6.

Cells containing TBP with a hemagglutinin epitope tag, whether the tag is at the N or C terminus, are viable and healthy, but the tagged versions of TBP are lethal in an nhp6ab mutant (67). Similarly, TBP(K138T Y139A), TBP(G174E), and TBP(F237D) mutants are viable in an NHP6⁺ strain but lethal in an nhp6ab mutant (67). We wondered what other types of TBP mutations would not be tolerated in the absence of Nhp6, and we have taken two approaches. First, we set up a screen to look for TBP point mutations that are specifically lethal in an nhp6ab strain. Second, we tested a variety of TBP mutations that had been characterized in other labs.

We used a red and white sectoring assay (6) to identify specific TBP mutations. We constructed an ade2 ade3 spt15::LEU2 strain containing a YCp plasmid with three genes, the wild-type TBP gene and two nutritional markers, URA3 and ADE3. The SPT15 gene, which encodes TBP, is essential for viability, so this strain cannot lose the YCp-ADE3-URA3-SPT15 plasmid, and the strains are red in color and sensitive to 5-FOA. This strain was transformed with a PCR-mutagenized TBP library on a YCp-TRP1 vector (2), and we screened for solid red (nonsectoring) colonies (Fig. 3). The majority of these nonsectoring colonies contain lethal TBP mutations, and these mutants were eliminated by mating to an NHP6⁺ strain and testing the resulting NHP6⁺/nhp6ab diploids for sectoring and 5-FOA sensitivity. Finally, all of the candidate TRP1 plasmids with TBP mutations that were synthetic lethal with nhp6ab were purified and retransformed into nhp6ab spt15 and NHP6⁺ spt15 strains to verify that the mutations are specifically lethal with nhp6ab.

FIG. 3.

Screen for TBP mutants specifically lethal in an nhp6ab mutant. Strain DY7244 (MATα nhp6a nhp6b spt15 ade2 ade3 ura3 with a YCp-URA3-ADE3-SPT15 plasmid) was transformed with a YCp-TRP1 library of TBP mutants. Nonsectoring colonies were identified, and then mated to strain DY7244 (MATa NHP6A NHP6B spt15 ade2 ade3 ura3 with a YCp-URA3-ADE3-SPT15 plasmid), and the resulting diploids were tested for sectoring and 5-FOA growth phenotypes. Diploid strains giving sectoring and 5-FOA-resistant phenotypes contain a TBP plasmid specifically lethal in an nhp6ab mutant.

We screened 8,500 colonies transformed with mutagenized TBP plasmids and recovered 32 TBP mutations that are lethal in the absence of Nhp6 (Fig. 4). These mutants were sequenced, and all but five had a single amino acid substitution (Table 3). We also screened 33 TBP mutants previously characterized by other investigators and identified 11 additional TBP mutations that are lethal in the absence of TBP (Table 3; examples in Fig. 5A). As described below in the Discussion, these mutations cluster in three regions, near where TBP interacts with TFIIA (clusters 1, 2, and 3), where TBP interacts with Spt3 (clusters 5), and on the upper surface where the TBP core may interact with TBP-associated factors or the N-terminal region of TBP (cluster 4 [Fig. 4]). The TBP mutations that are tolerated in the nhp6ab mutant are interesting, and they include substitutions that eliminate interaction with TFIIB (E186L, E186 M, E188A, and L189A [38; S. Buratowski, unpublished data]) or TFIIA (K133L and K133L/K138L [11]), substitutions that cause defective DNA binding (V71A, S118L, F148H, N159D, N159L, and V161A [2, 38, 61]), substitutions that cause defective transcriptional activation (V71A, S118L, F148H, T153I, N159D, N159L, V161A, and E236P [2, 37]), and substitutions that are lethal in the presence of a Taf1 truncation (S118L, N159D, N159L, V161A, and E336P [34]).

FIG. 4.

TBP mutations on the TBP core structure. The structures of TBP (purple), TFIIB (light blue), TFIIA (green), and DNA (black) are shown. Mutations in TBP residues shown in red, yellow, and green are lethal in an nhp6ab mutant. The red residues have been previously shown to affect TBP interaction with TFIIA, and the G174 residue that contacts Spt3 is shown in green. Site-directed mutations created on the TBP surface that eliminate interaction with TFIIB are shown in orange, and these mutations are viable in an nhp6ab mutant. The clusters of mutations are described in the text. SwissPDB Viewer version 3.7 (26) was used to merge the TFIIB-TBP-DNA 1VOL (48) and TFIIA-TBP-DNA 1YTF (64) crystal structures into a single PDB file, and the positions of mutations on the resulting structure were visualized with RasMac version 2.6 (55).

TABLE 3.

TBP mutations lethal in the absence of Nhp6

Substitution(s)	Cluster location(s) of substitution	Lethality in nhp6ab mutant suppressed by:		Source of mutant
		spt3Δ	YEp-SNR6
Single mutations
P65S	Cluster 4	No effect	ND	34
L67Q	Cluster 1	NDa	ND	This work
K83E	Cluster 3	ND	ND	This work
E93G	Cluster 3	No effect	No effect	This work
Y94C	Cluster 3	ND	ND	This work
P109A	Cluster 2	ND	ND	2
L114F	Cluster 5	Suppression	ND	3
E129G	Cluster 1	ND	Suppression	This work
E129V	Cluster 1	ND	ND	This work
K133R	Cluster 1	ND	Suppression	This work
K138L	Cluster 2	ND	ND	11
I142N	Cluster 2	ND	ND	This work
Q144K	Cluster 2	ND	ND	This work
K145L	Cluster 2	ND	ND	11
G147W	Cluster 2	Suppression	No effect	This work
K151E	Cluster 1	ND	ND	This work
F152S	Cluster 1	ND	ND	This work
F155K	Cluster 1	ND	ND	This work
I157T	Cluster 1	ND	ND	This work
C164W	Cluster 4	No effect	No effect	This work
C164Y	Cluster 4	ND	ND	This work
L172P	Cluster 5	Suppression	Suppression	This work
R173G	Cluster 5	No effect	ND	This work
G174E	Cluster 5	Partial suppression	Partial suppression	19
K211Eb		Suppression	ND	This work
R220H	Cluster 1	No effect	ND	34
Q225P	Cluster 4	ND	ND	This work
F227L	Cluster 4	Suppression	ND	This work
F227S	Cluster 4	ND	ND	This work
E228K	Cluster 4	ND	ND	This work
Y231A	Cluster 4	No effect	ND	34
F237D	Cluster 5	No effect	ND	34
F237L	Cluster 5	Suppression	ND	This work
K239E	Cluster 5	No effect	ND	This work
K239T	Cluster 5	ND	Suppression	This work
K239Stop	Cluster 5	ND	ND	This work
Double mutationsc
F18L K110Kd	Cluster 2	ND	ND	This work
K97R L193S	Cluster 3, 5	Suppression	Suppression	This work
I103T K239Stop	Cluster 3, 5	ND	Partial suppression	This work
E108G Y231H	Cluster 2, 5	ND	ND	This work
K120I C164Y	Cluster 4, 4	No effect	ND	This work
K133L K145L	Cluster 1, 2	Suppression	Partial suppression	11
K138T Y139A	Cluster 2, 2	Partial suppression	No effect	62

^aND, not determined.

^bK211 does not fall within any of the clusters.

^cFor the double mutations, the cluster locations of the substitutions are shown in the same order as the substitutions.

^dF18 is not present in the TBP crystal structure.

FIG. 5.

TBP mutants synthetic lethal with nhp6ab can be suppressed by YEp-SNR6 or by spt3. (A) Examples of synthetic lethality between TBP mutants and nhp6ab. Dilutions of strain DY7244 (nhp6ab spt15 with a YCp-URA3-TBP plasmid [with the wild-type {wt} TBP gene]) transformed with the indicated TBP mutant plasmids were plated on complete medium or medium containing 5-FOA and incubated at 30°C for 3 days. Plasmids pRS314 (vector), pTM8 (with the wild-type TBP gene), YCp-TBP(F237D), and YCp-TBP(R220H) were used. (B) Strain DY7244 (nhp6ab spt15) was transformed with two plasmids, a TRP1 plasmid with the indicated TBP mutant and either pRS422 (YEp-HIS3 vector) or M4488 (YEp-SNR6), and dilutions of transformed strains were plated on medium containing 5-FOA and incubated at 30°C for 5 days. (C) An SPT3 gene disruption suppresses the TBP nhp6ab synthetic lethality for TBP(F227L) and TBP(F237L). Strain DY7723 (nhp6ab spt3 spt15) was transformed with the indicated plasmids, and dilutions were plated on medium containing 5-FOA and incubated at 30°C for 2 days (complete) or 4 days (5-FOA). (D) An SPT3 gene disruption suppresses the temperature sensitivity of TBP mutants. Dilutions of strains DY7474 (NHP6 spt3 spt15) and DY7472 (NHP6 spt15) transformed with plasmids YCp-TBP(L114F), YCp-TBP(N159D), or YCp-TBP(R220H) were plated on medium containing 5-FOA and incubated at 30°C for 3 days or at 36 or 37°C for 6 days.

Interestingly, two of this group of TBP point mutations that are tolerated in the absence of Nhp6 affect interaction with TFIIA, while other TBP mutations affecting TFIIA binding are lethal in the nhp6ab mutant (Table 3) (67). For example, the K133L K145L double mutation is lethal in the nhp6ab mutant, but cells with the K133L single mutation in TBP are viable. The K133L K145L double mutation causes defective interaction with TFIIA in vitro, and this mutant shows a much stronger growth defect than the K133L single mutant (11). We suggest that cells with the K133L mutation have only a modest defect in interaction with TFIIA and thus tolerate the lack of Nhp6. Interestingly, while cells with the K133L mutation are viable in the absence of Nhp6, the K133R mutation recovered in the screen is lethal in the nhp6ab strain. It may be that replacement of lysine with the larger arginine residue interferes with TFIIA interaction.

Suppression by multicopy SNR6.

TBP interacts with TFIIA and TFIIB, and transcriptional activation requires formation of a TBP-TFIIA-TFIIB-DNA complex. We therefore asked whether a multicopy plasmid with either SUA7, encoding TFIIB, or the two genes encoding the TFIIA subunits, TOA1 and TOA2, could suppress the synthetic lethality of TBP alleles in the nhp6ab mutants. We tested more than half of the TBP mutants for multicopy suppression, but none of them were suppressed by either YEp-TFIIA or YEp-TFIIB. Since the 6-AU sensitivity of nhp6ab mutants could be suppressed by multicopy SNR6 (Fig. 2), we determined whether YEp-SNR6 could suppress the synthetic lethality from combining nhp6ab with TBP point mutations. We tested 12 TBP alleles and found that for 8 alleles a multicopy plasmid with the SNR6 gene allowed cells to grow on 5-FOA (Table 3; examples in Fig. 5B). There is no obvious correlation between the location of the TBP substitution on the structure (Fig. 4) and the ability to be suppressed by YEp-SNR6.

Suppression by spt3 mutation.

Spt3 physically interacts with TBP, and Spt3 acts to promote or inhibit TBP binding, depending on the promoter (17, 19, 67). Additionally, an spt3 mutation can suppress growth defects in both nhp6a nhp6b and gcn5 nhp6a nhp6b strains (67). On the basis of these results, we examined whether an spt3 gene disruption can suppress the synthetic lethality observed between nhp6 and mutant TBP alleles, using an spt15 nhp6a nhp6b spt3 strain with wild-type TBP on a YCp-URA3 plasmid. This strain was transformed with the TRP1 plasmids carrying various mutant versions of TBP, and the transformed strains were then plated on medium containing 5-FOA to determine whether the cells were viable after loss of the YCp-URA3 plasmid with wild-type TBP. Although we did not test all of the TBP alleles that are lethal in the nhp6ab mutant, we determined that for 10 of the 19 TBP alleles tested, the spt3 mutation suppressed the synthetic lethality with nhp6ab (Table 3). As shown in Fig. 5C, the synthetic lethality of the F227L mutant with nhp6ab is strongly suppressed by spt3, and the F237L mutant is suppressed to a lesser extent. Again, there is no correlation between suppression by spt3 and the position of the TBP substitution on the structure (Fig. 4).

A number of the TBP mutants show temperature-sensitive growth in an otherwise wild-type (e.g., NHP6⁺) strain. We decided to determine whether an spt3 mutation would affect the growth characteristics of these TBP mutants. Two isogenic spt15Δ strains containing the YCp-ADE3-URA3-SPT15 plasmid were constructed, one SPT3⁺ and the other spt3⁻. The two strains were transformed with 15 plasmids with TBP alleles and plated at various temperatures on medium containing 5-FOA to assess the ability of these strains to grow in the absence of the plasmid with wild-type TBP. As shown in Fig. 5D, the L114F, N159D, and R220H mutants grow on 5-FOA at 30°C in both SPT3 and spt3 strains, although colony size is reduced in the spt3 mutants. All three of these TBP mutants are unable to grow on 5-FOA at an elevated temperature when the SPT3 gene is present. Interestingly, an spt3 mutation allows these TBP mutants to grow at the higher temperature. Similar suppression of temperature-sensitive growth was also seen for L172P and R173G mutants and the K97R L193S and K120I C164Y double mutants (data not shown). These suppression results are consistent with the suggestion that Nhp6 and Spt3 affect TBP function in opposing ways (67).

Since an spt3 gene disruption can suppress many nhp6ab phenotypes, we also asked whether this suppression extends to the 6-AU sensitivity of nhp6ab mutants. As shown in Fig. 6A, the nhp6ab strain is very sensitive to 6-AU, but this effect is largely suppressed in the nhp6ab spt3 strain. This result is surprising, because the data in Fig. 2 suggest that the 6-AU sensitivity in the nhp6ab strain is largely due to the decreased expression of SNR6, a Pol III-transcribed gene. However, there is no known role of SPT3 in Pol III transcription, although the function of SPT3 in Pol II transcription is well documented. To examine this question, we measured SNR6 RNA levels by Northern hybridization in isogenic strains (Fig. 6B). SNR6 RNA levels are reduced to 23% in the nhp6ab mutant, and this value is unchanged in the nhp6ab spt3 strain. Additionally, SNR6 RNA levels are not increased in the spt3 single mutant, compared to the wild type. In wild-type strains, growth in the presence of 6-AU results in a modest decrease in SNR6 RNA levels (Fig. 6C). In summary, the decrease in SNR6 RNA levels in the nhp6ab strain may contribute to the strong 6-AU sensitivity. However, while the spt3 mutation restores growth of the nhp6ab mutant on 6-AU, spt3 does not suppress the nhp6ab defect in SNR6 RNA levels.

FIG. 6.

The 6-azauracil-sensitive phenotype in nhp6ab strains can be suppressed by spt3. (A) Dilutions of strains DY3398 (wild type), DY8980 (spt3), DY7588 (nhp6ab), and DY8855 (nhp6ab spt3) were plated on complete medium or medium containing 6-AU (25 μg/ml) and incubated at 30°C for 3 days (complete medium) or 4 days (medium containing 6-AU). (B) SPT3 does not regulate SNR6 expression. RNA levels for SNR6 and SNR128 (internal control) were determined by Northern blot hybridization. RNA was isolated from strains DY3398 (wild type), DY7588 (nhp6ab), DY8980 (spt3), and DY8855 (nhp6ab spt3). (C) 6-AU causes a modest reduction in SNR6 levels. RNA levels for SNR6 and SNR128 (internal control) were determined by Northern blot hybridization. RNA was isolated from strain DY3398 (wild type) grown for 3 h at 30°C in synthetic complete medium lacking uracil containing the indicated amount of 6-AU (in micrograms per milliliter).

DISCUSSION

Nhp6 is an abundant protein of the high-mobility-group B family of transcription factors that has been implicated in transcriptional activation by both RNA Pol II and Pol III. TBP is required for transcription by all three eukaryotic RNA polymerases. In this study, we used a novel genetic screen to identify substitutions in TBP that are viable, but lethal in the absence of Nhp6. The TBP nhp6ab synthetic lethality for a number of the TBP mutants can be suppressed by deletion of the SPT3 gene. Spt3 is known to interact directly with TBP, and spt3 mutations affect Pol II transcription start site selection (19), suggesting that Nhp6 contributes to the activation of Pol II transcription by TBP. Nhp6 also plays an important role in transcription of the SNR6 gene by RNA Pol III, and TBP binding to SNR6 is reduced in an nhp6ab mutant. A multicopy plasmid with SNR6 suppresses the 6-AU sensitivity seen in nhp6ab mutants, as well as the lethality of some TBP substitutions in nhp6ab strains. Promoter mutations that reduce SNR6 expression also cause 6-AU sensitivity, suggesting that decreased levels of the SNR6 gene product contribute to sensitivity to this postulated inhibitor of transcriptional elongation. Our results suggest that Nhp6 and TBP work together to promote both Pol II and Pol III transcription.

TBP substitution mutants.

The 43 TBP mutations that are synthetic and lethal in combination with nhp6ab cluster in interesting regions of TBP. Figure 4 shows the structure of the core of TBP, along with TFIIB, TFIIA, and DNA, merged from two separate structure determinations (48, 64) as a top view, a front view, and a 90° rotated view. New substitutions isolated in our screen that are lethal in the absence of Nhp6 are shown in red. The residues that have been previously shown to affect TBP interaction with TFIIA (K133, K138, Y139, and K145) are shown in yellow. In the TBP-TFIIA-DNA crystallographic structures, none of these four TBP residues make direct contact with TFIIA (24, 64). Only a small part of TFIIA was successfully crystallized with TBP and DNA, and it is assumed that full-length TFIIA does indeed touch these contact residues in TBP. We show that all of these previously characterized TBP double mutants (K133L K138L, K133L K145L, and K138T Y139A) that affect interaction with TFIIA are lethal in an nhp6ab mutant but viable in an NHP6 strain. Residue K133 in TBP makes important contacts with TFIIA, as binding to TFIIA is lost in either the K133L K138L or K143 K145L double mutant (11, 62). We also recovered a single substitution at this position, K133R, that is lethal in the absence of Nhp6. On the top surface of TBP, K133 is part of a continuous line of substitutions, E129, K151, F152,F155, I157, and R220, that are synthetic and lethal with nhp6ab, and we describe these mutations as cluster 1 (top view in Fig. 4). Cluster 2 is a group of substitutions surrounding the other residues required for TFIIA interaction, K145 at the right edge of TBP, and slightly to the left are adjacent residues K138 and K139 (front view in Fig. 4). New substitutions at Q144 and G147 are just above K145, and the E108, P109, K110, and I142 mutations surround K138 and K139. The cluster 3 group of mutations, K83, E93, Y94, K97, and I103, are at the interface between TBP and TFIIA defined by the crystallographic studies. Clusters 1, 2, and 3 represent regions that probably interact with TFIIA, and some interesting patterns emerge for substitutions here. An spt3 deletion suppresses the nhp6ab TBP synthetic lethality for two of four mutants tested, and three of five mutants tested can be suppressed by YEp-SNR6. These results show the important role of SPT3 and SNR6 in Nhp6 and TBP function, as discussed below.

Cluster 4 is a group of substitutions lethal in nhp6ab, including Q225, F227, E228, and Y231, that are on the upper left surface of TBP (top view in Fig. 4). Cluster 4 substitutions are adjacent to the P232 and V233 mutations that were identified in a screen for TBP mutations that affect NC2 binding (12) and adjacent to A226 where three mutations that specifically affect Pol II transcription were identified (14). It is not clear what this region of TBP interacts with, but one possibility is this region interacts with factors that regulate TBP binding, such as NC2, Mot1, the Not-Ccr4 complex, and TBP-associated factors (39, 52).

Cluster 5 is a collection of substitutions near G174, a residue that contacts Spt3 (shown in green in Fig. 4). Surrounding G174, we recovered clustered mutations at L172, R173, F237, and K239. Interestingly, we recovered three independent substitutions at K239. There are some interesting phenotypes common to most of the mutations in cluster 5. The synthetic lethality with nhp6ab for most of these TBP mutations can be suppressed by a spt3 deletion. Interestingly, F237L is suppressed by spt3, but not the F237D substitution at the same position. L172P and G174E can be suppressed by spt3, but not the R173G mutant at the intervening position. YEp-SNR6 suppresses the nhp6ab TBP synthetic lethality for all of the cluster 5 substitutions tested. Below cluster 5 are three residues shown in orange. These are positions where site-directed mutations were created on the TBP surface that eliminate interaction with TFIIB (38), based on the TFIIB-TBP-DNA cocrystal. These substitutions had no additional growth defect in an nhp6ab strain.

Nhp6 and Pol III transcription.

Experiments suggest that Nhp6 is required for efficient transcription of the SNR6 gene by Pol III (35, 42, 44). SNR6 encodes the U6 RNA required for mRNA splicing, and it is suggested that a deficiency in SNR6 RNA contributes to the temperature-sensitive growth defect seen in nhp6ab mutants. Brf1 is the limiting component in TFIIIB, a factor required for Pol III transcription (56); therefore, BRF1 overexpression could increase SNR6 expression and facilitate growth of the nhp6ab mutant at 37°C. Overexpression of TBP also suppresses the temperature-sensitive growth defect of an nhp6ab mutant (67). Additionally, TBP overexpression suppresses the defect in HO transcription by RNA Pol II in an nhp6ab strain (67). In addition to its well-documented role in Pol II transcription, TBP is also a component of the RNA Pol III factor TFIIIB and is required for Pol III transcription (33). Thus, TBP overexpression could suppress the nhp6ab growth defect by affecting Pol II or Pol III transcription.

Several studies have defined how TBP interacts with Pol III transcription factors. Cormack and Struhl (14) mutagenized TBP, isolated substitutions with a temperature-sensitive phenotype, and identified 65 TBP mutants specifically defective in Pol III transcription. Although some of these residues also showed up in our screen for TBP nhp6ab synthetic lethality, including P65, E129, E133, Q144, F152, F155, I157, R220, Y231, and F237, not all substitutions confer the same phenotypes. For example, while F237P and F237L substitutions specifically affected Pol III transcription, F237R specifically affected Pol II, and the defects of the F237D substitution were not specific to any polymerase. Additional TBP residues important for Pol III transcription were identified through studies examining the interaction of TBP mutants with the Brf1 Pol III transcription factor (58) by the TBP-Brf1 cocrystal (32). Interestingly, all of the residues in TFIIA that are required for TFIIA-TBP interaction (K133, K138, Y139, and K145) are also required for TBP interaction with Brf1. In summary, the regions defined by clusters 1, 2, and 3 are involved in TBP binding to both TFIIA and Brf1. The fact that these substitutions are lethal in an nhp6ab mutant suggests that Nhp6 becomes essential when a substitution reduces that ability of TBP to interact with TFIIA or Brf1.

Several of our experiments reinforce the role for Nhp6 in Pol III transcription. ChIP experiments show strong binding of TBP to the SNR6 gene, transcribed by Pol III, and TBP binding to SNR6 is markedly reduced in the absence of Nhp6. nhp6ab mutants are extremely sensitive to 6-AU, and this could be suppressed by multicopy plasmids with either the Pol III factor BRF1 or the Pol III target gene SNR6. Finally, multicopy plasmids with SNR6 suppress the nhp6ab TBP synthetic lethality for several mutants, suggesting that a defect in Pol III transcription is partially responsible for the nhp6ab TBP synthetic lethality.

The suppression of the 6-AU sensitivity in nhp6ab mutants is particularly interesting. The 6-AU inhibitor affects the ribonucleotide triphosphate pools, and many transcriptional elongation mutants are sensitive to 6-AU (20, 57). Nhp6 interacts with the transcriptional elongation factor yFACT (Spt16/Pob3); therefore, the 6-AU sensitivity of nhp6ab mutants was attributed to a defect in transcriptional elongation (10, 22). We find that this 6-AU sensitivity can be suppressed by either SNR6 or BRF1 on a multicopy plasmid (Fig. 2A) or by disruption of the SPT3 gene (Fig. 6A). Although an spt3 mutation suppresses the 6-AU sensitivity caused by nhp6ab, spt3 does not suppress the nhp6ab defect in SNR6 expression (Fig. 6B).

It is not clear why an spt3 gene deletion can suppress nhp6ab, in terms of sensitivity to 6-AU, but it does not suppress the defect in SNR6 expression. An spt3 mutation does not cause an increase in SNR6 RNA levels, and an spt3 strain is resistant to 6-AU (Fig. 6). However, an spt3 mutation, by itself, has been reported to cause sensitivity to 6-AU in a different strain background (16). Previously, SPT3 has never been shown to have a role in Pol III transcription, and spt3 mutations have been observed to affect only RNA Pol II transcription. Thus, it is possible that the spt3 suppression of the nhp6ab 6-AU sensitivity is an indirect effect of altered gene expression by RNA Pol II.

Nonetheless, our data show clearly that 6-AU sensitivity results from decreased SNR6 expression, whether because of a SNR6 promoter mutation or an nhp6ab mutation. This challenges the widely held belief that mutations affecting transcriptional elongation result in sensitivity to 6-AU. One speculative explanation is that the reduced levels of the U6 splicing factor, produced from the SNR6 gene, affect the efficiency of mRNA splicing, and this in turn has an effect on transcriptional elongation. Several reports have recently suggested a strong link between splicing and elongation. Fong and Zhou (21) report that spliceosomal small nuclear ribonucleoproteins interact with the transcription elongation factor TAT-SF1 and stimulate Pol II elongation. A mutation in the largest subunit of RNA Pol II that slows elongation by RNA Pol II has effects on splicing, in both human and Drosophila cells (15). Additionally, mutations in the SPT4 and SPT5 yeast genes that encode elongation factors result in splicing defects (41), and growth in the presence of 6-AU also affects splicing (30). Additionally, growth of wild-type cells in the presence of 6-AU causes a slight reduction in SNR6 RNA levels. We suggest that the nhp6ab mutation reduces the levels of the U6 splicing factor, produced from the SNR6 gene, and that the reduced levels of splicing factors affect transcriptional elongation, resulting in 6-AU sensitivity. In any case, our studies show that one cannot assume that 6-AU sensitivity results from a defect in transcriptional elongation.

Nhp6 and Pol II transcription.

While the lethality of TBP substitution mutants in an nhp6ab strain could be because Nhp6 is required for efficient SNR6 expression by RNA Pol III, we believe Nhp6 plays a role in TBP binding at genes transcribed by RNA Pol II. In vitro binding experiments show that Nhp6 stimulates the formation of a TBP-TFIIA-TFIIB-DNA complex, and in vivo experiments with chimeric promoter constructs suggest that Nhp6 acts at core promoters (50). Nhp6 acts positively to promote activation of the HO gene (68) and negatively to repress FRE2 transcription (23). Nhp6 acts both positively and negatively at the CHA1 gene, because induced levels are reduced and basal levels are increased in an nhp6ab mutant (45). In support of the idea that Nhp6 promotes DNA binding by TBP at genes transcribed by RNA Pol II, in vitro binding experiments show that Nhp6 stimulates formation of a TBP-TFIIA-DNA complex (D. Biswas, A. N. Imbalzano, P. Eriksson, Y. Yu, and D. J. Stillman, submitted for publication). TFIIA is encoded by the TOA1 and TOA2 genes, and a toa2 mutation that eliminates TFIIA interaction with TBP is viable in wild-type cells but lethal in an nhp6ab mutant strain (Biswas et al., submitted). These results suggest that Nhp6 plays a role in formation of the TBP-TFIIA-DNA complex. Additionally, we have found genetic interactions between Nhp6 and factors that regulate TBP binding (D. Biswas and D. J. Stillman, unpublished data). These strong genetic interactions between NHP6 and basal factors exclusively used for Pol II transcription provide strong support for the idea that Nhp6 has an important role in RNA Pol II transcription. Thus, Nhp6 facilitates expression of genes transcribed by both Pol II and Pol III.