An RNA Polymerase III-defective Mutation in TATA-binding Protein Disrupts Its Interaction with a Transcription Factor IIIB Subunit inDrosophila Cells*

A subunit of the Drosophila RNA polymerase III transcription factor IIIB (TFIIIB) complex has been identified using antibodies directed against the analogous human protein, hIIIB90. This protein has an apparent molecular mass of 105 kDa and has been designated dTAFIII105.Drosophila S-2 cell extracts that were immunodepleted of dTAFIII105 were substantially reduced in their capacity to support tRNA gene transcription. A protein (far Western) blot analysis revealed that dTAFIII105, present in a TFIIIB fraction, directly interacts with TATA-binding protein (TBP). Coimmunoprecipitation assays demonstrated that this protein associates with TBP in S-2 cell extracts. Our previous studies have identified a mutation at position 332 within Drosophila TBP that changes a highly conserved arginine residue to a histidine residue, which renders it specifically defective in its ability to support RNA polymerase III transcription in S-2 cells (Trivedi, A., Vilalta, A., Gopalan, S., and Johnson, D. L. (1996) Mol. Cell. Biol. 16, 6909–6916). We further demonstrate that extracts prepared from a stable cell line expressing epitope-tagged wild-type TBP exhibit an increase in tRNA gene transcription, whereas extracts derived from cells expressing the mutant TBP protein do not. Coimmunoprecipitation assays and far Western blot analysis demonstrate that this mutation in TBP abolishes its ability to stably interact with dTAFIII105. Thus, we have identified both aDrosophila protein that is directly associated with TBP in the TFIIIB complex, dTAFIII105, and an amino acid residue within the highly conserved carboxyl-terminal region of TBP that is critical for dTAFIII105-TBP interactions.

The transcription of small untranslated RNAs by RNA polymerase (pol) 1 III requires several ancillary factors (for reviews, see Refs. 2 and 3). There are three major classes of RNA pol III promoters, typified by the genes encoding tRNA, 5 S RNA, and U6 RNA. The tRNA class of promoters utilizes transcription factor IIIB (TFIIIB) and TFIIIC. The 5 S RNA gene additionally requires TFIIIA. The U6 RNA class of promoter, in vertebrates, requires a multiprotein complex referred to as SNAPc-PTF-PBP, one of the two TFIIIC subcomplexes, the TATA-binding protein (TBP), and possibly other components of TFIIIB (4 -6). Thus, in addition to RNA pol III, the only other component that is required for the transcription of all classes of RNA pol III promoters is TFIIIB.
TFIIIB is a multisubunit protein complex that contains TBP and at least two TBP-associated factors (TAFs). TBP has been shown to be essential for the transcription of all cellular genes (for a review, see Ref. 7). It is associated with at least 13 distinct TAFs that form the TFIID complex, which is specific for the transcription of RNA pol II genes (8 -11). Another complex, B-TFIID, consisting of TBP and a single large TAF, has also been implicated in RNA pol II transcription (11,12). For the transcription of RNA pol I promoters, TBP is associated with three unique TAFs that compose the SL1 complex (13,14). In addition, another complex, SNAPc-PTF-PBP, which is thought to associate with TBP (4) and is used to transcribe specific small nuclear RNA genes, is composed of four polypeptides (4,5). Thus, the ability of TBP to form these distinct TBP⅐TAF complexes dictates its role in the transcription of specific cellular promoters.
The structure of TFIIIB is best understood in Saccharomyces cerevisiae. Yeast TFIIIB is composed of TBP, a 70-kDa polypeptide (BRF1/PCF4/TDS4; yTFIIIB70) (15)(16)(17), and a 90-kDa polypeptide (BЉ; yTFIIIB90) (18 -20). yTFIIIB70 has been shown to directly interact with TBP, whereas yTFIIIB90 is thought to associate with TBP primarily through its interaction with yTFIIIB70 (19,20). Suppression of RNA pol IIIspecific TBP mutations by overexpression of yTFIIIB70 (15,16) first suggested a direct interaction between yTFIIIB70 and TBP. This interaction has been confirmed by in vitro experiments using recombinant TBP and yTFIIIB70 (21,22). In addition, yTFIIIB70 has been shown to interact with the 135-kDa subunit of TFIIIC (23,24) and the 34-kDa subunit of RNA pol III (25). These studies are consistent with the observation that TFIIIC bound to the DNA serves to recruit TFIIIB to the template, and TFIIIB then functions as an initiation factor to recruit RNA pol III (26).
The structure of TFIIIB in higher eucaryotes is less well defined. Human TFIIIB preparations have been reported to contain, in addition to TBP, polypeptides of (i) 190, 96, 87, and 60 kDa (27); (ii) 150, 82, and 54 kDa (28); and (iii) 172 kDa and a TBP-associated factor L of undefined size (29). However, the largest polypeptide in these preparations appears to result from contamination of TFIIIA with the B-TFIID complex (30). In Xenopus, TFIIIB was reported to be composed of TBP and at least two additional polypeptides of 92 and 75 kDa (31). A cDNA encoding a non-TBP component of the TFIIIB complex has been cloned from human cells (30,32). The amino-terminal region of hTFIIIB90 exhibits high sequence similarity to yT- FIIIB70, while the carboxyl-terminal region contains a high  mobility group protein 2-related domain. A recombinant protein containing 394 amino acids of the carboxyl-terminal region  of TFIIIB90 strongly associates with TBP, while an aminoterminal 300-amino acid fragment of the protein interacts relatively weakly. TFIIIB has been shown to be a key target for the regulation of RNA pol III gene activity in several diverse processes. Alterations in TFIIIB activity have been shown to occur during different stages of cell differentiation (33), cell growth (34,35), and the cell cycle (36,37). In addition, increases in TFIIIB activity have been shown to mediate the induction of RNA pol III genes in response to certain viral proteins such as the hepatitis B virus X protein (38) and the human T-cell leukemia virus type 1 Tax protein (39). The activation of cellular protein kinases also increases TFIIIB activity (40). Thus, the determination of the structure of TFIIIB and its interactions with other transcription components is fundamental to our understanding of RNA pol III gene regulation.
We have previously shown that TBP is a limiting factor for RNA pol III transcription in Drosophila S-2 cells (1). Both stable and transient transfection of epitope-tagged TBP enhanced transcription of transfected tRNA and U6 RNA gene promoters. We further determined whether the stimulation in transcription was a direct result of increased TBP complexes at these promoters. A mutant TBP protein that changed a highly conserved arginine residue to a histidine at position 332 within the carboxyl-terminal region was expressed and found to be specifically defective in its ability to support transcription of either of the RNA pol III promoters. However, the mutant TBP protein completely maintained its ability to support transcription of two RNA pol II promoters. These results demonstrated that the TBP-dependent increase in RNA pol II gene activity is not sufficient for enhanced RNA pol III transcription; rather, a direct effect on RNA pol III promoters is required. Although other transcription components may also be limiting for the transcription of RNA pol III genes, these studies demonstrate that TBP is one component that limits the expression of these genes.
In this study, we have investigated why the mutation at position 332 within Drosophila TBP renders it defective for RNA pol III transcription. To do so, we have identified a polypeptide within the Drosophila TFIIIB complex. This represents the first identification of a Drosophila RNA pol III transcription factor. This subunit, TAF III 105, associates with epitope-tagged TBP that is expressed in a Drosophila S-2 stable cell line. In contrast, TAF III 105 fails to associate with the epitope-tagged mutant TBP332 protein. Far Western blot analysis revealed that TBP, but not TBP332, binds to TAF III 105 present in a TFIIIB fraction. Our results provide the first report of a TBP mutation that eliminates its ability to support RNA pol III transcription by disrupting its direct association with a subunit of the TFIIIB complex in Drosophila cells. Furthermore, these studies provide new physical evidence that the carboxyl-terminal direct repeat within TBP directly interacts with the TFIIIB subunit.

EXPERIMENTAL PROCEDURES
Cell Cultures-Early passage Drosophila Schneider S-2 cells were grown in Schneider medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum (Gemini). Medium for F-TBP and F-TBP332 cell lines additionally contains hygromycin B (Boehringer Mannheim) at 250 g/ml. F-TBP and F-TBP332 cell lines were obtained by stable transfection of Drosophila Schneider S-2 cells as described previously (1).
Extracts and Transcription Factor Preparation-Cytoplasmic (S100) extracts derived from Drosophila Schneider S-2 cells, F-TBP cells, or F-TBP332 cells were prepared according to the method of Dingermann (41). F-TBP and F-TBP332 cells were treated with 500 M CuSO 4 48 h prior to harvesting the cells. The Drosophila TFIIIB and TFIIIC/TFIID fractions were obtained by fractionation of S-2 cell S100 extracts on phosphocellulose (Whatman P-11) as described previously (40,42). The TFIIIB fraction was eluted stepwise at 0.35 M KCl, and the TFIIIC/ TFIID fraction was eluted at 1 M KCl.
Plasmid DNAs-The plasmid, pArg-maxi, used as template for in vitro transcription reactions, contains a derivative of a Drosophila tRNA Arg gene that has an additional 12 nucleotides inserted between the internal promoter regions (43). pF-TBP denotes an expression plasmid that produces an amino-terminally Flag-tagged (MDYKDDDDK) Drosophila TBP protein. pF-TBP332 contains a mutation within the TBP cDNA corresponding to amino acid position 332 that changes an arginine residue to a histidine residue. The expression of both TBP proteins is under the control of the metallothionein promoter. These plasmids were used to produce Drosophila Schneider S-2 stably transfected cell lines as described previously (1). Plasmids pCITE-FTBP and pCITE-FTBP332, used as templates for in vitro transcription/translation, were obtained by subcloning the epitope-tagged wild-type and mutant TBP cDNAs into plasmid pCITE-2b (Novagen) using NcoI and SalI restriction sites.
Transcription Assays-Transcription assay mixtures contained 0.3 g of DNA template; 20 mM HEPES (pH 7.9); 5 mM MgCl 2 ; 3 mM dithiothreitol; 100 mM KCl; 10% glycerol; 0.5 mM each ATP, CTP, and UTP; and 0.1 mM [␣-32 P]GTP (6 Ci/mmol) in a 60-l final volume. Equal amounts of protein (160 g) from each extract were used in the experiment shown in Fig. 1. The amount of extract used for each reaction was previously determined to be within the linear range of the assay. Reactions were incubated for 1 h at room temperature and stopped by the addition of 0.1% SDS and 400 mg/ml proteinase K. After 15 min at 37°C, RNA was purified by phenol extraction and ethanol precipitation and analyzed by electrophoresis on 8 M urea, 8% polyacrylamide gels. Transcription products were visualized by autoradiography and quantitated by densitometry. Western Blot Analysis-Proteins from the cell extracts were separated by SDS-polyacrylamide gel electrophoresis (7.5%) followed by blotting onto nitrocellulose (Schleicher & Schuell) by the semidry method described by Kyhse-Andersen (44). Total Drosophila TBP was detected with affinity-purified polyclonal antibodies raised against Drosophila TBP (kindly provided by J. Kadonaga, University of California, San Diego). The M5 monoclonal antibody (Eastman Kodak Co.) specific for the Flag epitope was used to specifically detect the Flag-tagged F-TBP and F-TBP332 proteins (45). Drosophila TAF III 105 was detected using anti-human IIIB90 rabbit polyclonal antibodies that were affinity-purified (30). Horseradish peroxidase-linked antibodies (Vectastain) and enhanced chemiluminescence reagents (HRPL kit; National Diagnostics) were used to detect bound antibodies.
Far Western Blot Analysis-Far Western blot analysis was performed as described previously (21) with the following modifications. Twenty micrograms of protein from the designated phosphocellulose P-11 fractions was subjected to SDS-PAGE and blotted onto nitrocellulose as described for Western blot analysis. After the transfer, the filters were rinsed briefly in distilled water followed by incubation at 4°C for 1 h in buffer E (20 mM HEPES (pH 7.9), 0.5 mM dithiothreitol, 0.2 mM EDTA, 20% glycerol, 0.1% Nonidet P-40) containing 5% nonfat milk and 100 mM KCl. Filters were rinsed in buffer E containing either 100 mM KCl (for the "low stringency" wash) or 350 mM KCl (for the "high stringency" wash) and then incubated with 35 S-labeled F-TBP or F-TBP332 (100,000 cpm/ml) overnight at 4°C. The filters were washed twice for 15 min at 4°C with buffer E containing either 100 mM KCl or 350 mM KCl and then once with the same buffer without glycerol. The filters were then washed with 15% trichloroacetic acid followed by a wash with 50% ethanol. The bound F-TBP or F-TBP332 was detected by autoradiography, and bands were quantified using a Bioimage scanner.
Immunoprecipitations and Immunodepletions-For the immunodepletions and immunoprecipitation assays, equal amounts of protein, typically 300 -600 g, from the cytoplasmic extracts were incubated for 1 h at 4°C with 5 l of the antibodies specified. Protein A-Sepharose (Santa Cruz) was pretreated by incubating equal volume amounts of the resin and the cytoplasmic extract. After a 1-h incubation on ice, the sample was centrifuged to pellet the resin, the supernatant was discarded, and the resin pellet was resuspended in buffer E to the original volume. Protein A-Sepharose was added (15 l each) to the cytoplasmic extracts that had been incubated with antibody, and the samples were incubated at 4°C overnight with gentle mixing. Samples were centrifuged, and the immunodepleted supernatants were used for in vitro transcription assays. For analysis of the immunoprecipitated proteins, the pellets were washed extensively with buffer containing 20 mM HEPES (pH 7.9), 0.5 mM dithiothreitol, 0.2 mM EDTA, 20% glycerol, 350 mM KCl, and 0.1% Nonidet P-40. The pellets were then resuspended in SDS-PAGE sample buffer and boiled for 5 min. Equal volume amounts of the samples were used for Western blot analysis.

Mutation at Position 332 within Drosophila TBP Renders It
Inactive for tRNA Gene Transcription in Vitro-Drosophila S-2 cells were previously used to generate stable cell lines that express epitope-tagged wild-type and mutant TBP proteins (1). TBP expression plasmids were constructed to contain a nineamino acid Flag epitope coding region fused at the amino terminus of a Drosophila TBP cDNA, and the genes were placed under the control of the metallothionein promoter, which is inducible with copper. The S-2 cell line generated by stable transfection of the expression plasmid containing the wild type TBP cDNA is designated F-TBP. The stable cell line generated by transfection of the expression plasmid harboring a mutation within the TBP cDNA that results in an arginine to histidine change at position 332 in the protein is designated F-TBP322. We have previously shown that when the F-TBP stable line is induced with copper, transient expression of a tRNA gene promoter is enhanced. However, when the F-TBP332 cell line is induced with copper to express the mutant protein, no increase in the expression of the RNA pol III promoter is observed (1).
To further determine if the differential abilities of the overexpressed wild type and mutant proteins to enhance RNA pol III transcription could be reproduced in vitro, we prepared extracts from the two stable cell lines and determined their transcriptional activities. As shown in Fig. 1, cytoplasmic extracts derived from the F-TBP cells consistently exhibited at least a 3-fold increase in the transcriptional activity of a tRNA gene template compared with the parental S-2 cell extracts. In contrast, extracts derived from the cells expressing F-TBP332 actually exhibited an approximate 3-fold decrease in their ability to support tRNA gene transcription compared with the S-2 cell extracts. The 10-fold difference in the transcriptional activities of extracts derived from the F-TBP and F-TBP332 stable lines can not be attributed to differences in the amount of expressed F-TBP and F-TBP332 proteins, since we have previously shown that similar levels of these proteins are produced (1). Thus, these results provide further evidence that the mutation at position 332 in the TBP protein renders it defective in RNA pol III transcription and that TBP is a limiting transcription component in Drosophila.
Identification of a Drosophila TFIIIB Subunit, dTAF III 105-Previous studies have shown that transient expression of epitope-tagged TBP enhanced transcription from both TATA-containing and TATA-lacking RNA pol III promoters, whereas expression of epitope-tagged TBP332 failed to stimulate either promoter class (1). These results suggested that the arginine residue at position 332 is important for TBP interactions with other proteins in both the tRNA and U6 RNA gene transcription complexes. TBP has been found to be associated with at least one other polypeptide in the TFIIIB complex in the yeast and human systems. In the human system, a polypeptide with an apparent molecular mass of 90 kDa, hIIIB90, has been identified that interacts with TBP in the TFIIIB complex (30). We therefore investigated whether the mutation at position 332 in Drosophila TBP affected its ability to associate with the corresponding Drosophila TAF. To do so, antibodies directed against hIIIB90 were first used to identify the analogous TAF in a partially purified TFIIIB fraction derived from Drosophila S-2 cell extracts. Western blot analysis revealed that the antibodies cross-reacted with a Drosophila polypeptide with an apparent molecular mass of 105 kDa, compared with the 90-kDa cross-reacting polypeptide present in HeLa extracts ( Fig.  2A). A faint signal was also observed in both lanes in a region corresponding to approximately 65 kDa. Since this signal was also observed in lanes where no protein was present, this represents a nonspecific interaction. To further determine the specificity of the interaction of the 105-kDa protein with the antibodies, this polypeptide was immunodepleted from S-2 extracts using the anti-hIIIB90 antibodies, and the ability of the resultant extracts to transcribe a tRNA gene template was assessed. As shown in Fig. 2B, immunodepletion with the hIIIB90 antibodies substantially reduced the transcriptional activity of the extract (approximately 15-fold) compared with mock-depleted extracts. Together, these results suggest that the 105-kDa polypeptide is the Drosophila equivalent to the hIIIB90 protein.
To further assess whether the 105-kDa protein was a subunit within the TFIIIB complex, we determined whether this polypeptide was associated with TBP. Extracts were prepared from S-2, F-TBP, and F-TBP332 cell lines that had been induced with copper sulfate, and immunoprecipitation studies were performed to examine protein-protein interactions (Fig.   FIG. 1. Effect of expression  3). The extracts were immunoprecipitated with Drosophila TBP antibodies (␣TBP), the Flag epitope antibodies (␣Flag), or the hIIIB90 antibodies (␣hIIIB90). The resultant immunoprecipitates were then analyzed by Western blot to determine the presence or absence of F-TBP, F-TBP332, or the 105-kDa protein. We first determined whether the 105-kDa protein was associated with TBP present in all extracts. When extracts were immunoprecipitated with ␣hIIIB90 and the resultant precipitates were probed with ␣TBP, we found that TBP coprecipitated with the 105-kDa protein in all three extracts (Fig. 3A). Likewise, immunoprecipitation with ␣TBP resulted in the coprecipitation of the 105-kDa protein (Fig. 3B). These results demonstrate that the 105-kDa polypeptide is physically associated with TBP in Drosophila cells.
To further determine whether the 105-kDa polypeptide is directly associated with TBP, we used a protein blot (far Western) analysis to examine the interaction of F-TBP with polypeptides within a phosphocellulose-derived TFIIIB (Fig. 4A). Pro-teins within the TFIIIB fraction were separated on SDS-PAGE, transferred to nitrocellulose, and probed with 35 S-labeled F-TBP. A single polypeptide within the TFIIIB fraction with an apparent molecular mass of 105 kDa was found to bind to F-TBP. Immunodepletion of the TFIIIB fraction with ␣hIIIB90 significantly reduced the amount of F-TBP binding to this polypeptide (data not shown). This polypeptide-TBP interaction was not observed in a phosphocellulose fraction containing TFIIIC and TFIID activities. However, several polypeptides in the TFIIIC/TFIID fraction were found to bind to TBP, most notably those with apparent molecular masses of 250, 140, and FIG. 3. Drosophila TAF III 105 is associated with TBP but not mutant TBP332 in Drosophila S-2 cells. Immunoprecipitation of extracts derived from the F-TBP cell line, the F-TBP332 cell line, or the parental S-2 cell line, as designated, were carried out as described under "Experimental Procedures." After immunoprecipitation of the extracts, the resultant precipitated proteins were electrophoresed on SDS-PAGE, and Western blot analysis was carried out as described under "Experimental Procedures" using antibodies directed against either human IIIB90 (␣hIIIB90), Drosophila TBP (␣TBP), or the Flag epitope (␣FLAG) as indicated.
FIG. 4. The TBP332 mutation disrupts its direct association with Drosophila TAF III 105. A, TAF III 105 within a TFIIIB fraction interacts directly with TBP. A phosphocellulose fraction eluted at 0.35 M KCl containing TFIIIB and a fraction eluted at 1 M KCl containing TFIIIC and TFIID were subjected to SDS-PAGE and transferred to nitrocellulose, and far Western blot analysis was performed as described under "Experimental Procedures" using 35 S-labeled F-TBP and low stringency conditions. The arrow designates the TBP-interacting protein with an apparent molecular mass of 105 kDa. The bands labeled with an asterisk denote TBP-interacting proteins in the TFIIIC/TFIID fraction with apparent molecular masses of 250, 140, and 110 kDa. B, the TBP332 mutation disrupts its interaction with TAF III 105. Far Western blot analysis was performed using high stringency conditions as described under "Experimental Procedures" using either 35 S-labeled F-TBP or 35 S-labeled F-TBP332 as designated. The bound F-TBP or F-TBP332 proteins to the 105-kDa polypeptide in the TFIIIB fraction and the 110-kDa polypeptide in the TFIIIC/TFIID fraction were quantified from the resulting autoradiographs by densitometry. 110 kDa. The 250-and 110-kDa polypeptides probably represent subunits of Drosophila TFIID that have been previously shown to bind directly to TBP (46). Thus, using two different approaches, we find that the 105-kDa polypeptide associates with TBP in vitro and in vivo, further supporting the possibility that it is a subunit of the TFIIIB complex. We have, therefore, designated this protein dTAF III 105.
A Single Amino Acid Change in TBP Disrupts Its Ability to Stably Interact with dTAF III 105-We further determined whether the F-TBP332 polypeptide, which was unable to support tRNA gene transcription (Fig. 1) was defective in its ability to associate with dTAF III 105. To determine if stable complexes were formed between the mutant TBP protein and dTAF III 105 in Drosophila cells, extracts were prepared from F-TBP, F-TBP332, and S-2 cell lines. Immunoprecipitation of the extracts with ␣Flag revealed that, as expected, the epitopetagged TBP was present in both of the F-TBP and F-TBP332 cell extracts but not in the S-2 extracts (Fig. 3C). When these immunoprecipitates were then tested for the presence of dTAF III 105, only the F-TBP protein was shown to coprecipitate dTAF III 105 (Fig. 3D). The inability to detect the presence of dTAF III 105 in the F-TBP332 precipitate was not due to gross differences in the amounts of expressed epitope-tagged TBP proteins in these cell lines, since Western blot analysis revealed that the relative amount of F-TBP332 expressed in the cells was approximately equal to the amount of F-TBP protein expressed (1). In addition, the F-TBP332 polypeptide was able to be immunoprecipitated more efficiently compared with F-TBP with the ␣Flag antibodies (Fig. 3C). Immunoprecipitation of the extracts with ␣hIIIB90 and detection of the epitope-tagged TBP proteins demonstrated that only the F-TBP protein was coprecipitated with dTAF III 105 (Fig. 3E), although approximately equal amounts of the dTAF III 105 protein were immunoprecipitated from each of the extracts (Fig. 3F).
Far Western blot analysis was further used to examine whether F-TBP332 could bind to dTAF III 105 present in the TFIIIB fraction. When a low salt (100 mM KCl) buffer was used for the incubation and wash steps, the F-TBP332 still was capable of effectively binding to dTAF III 105 (data not shown). Therefore, higher salt conditions (350 mM KCl) were used in the wash steps to examine if potential differences in the interactions between dTAF III 105 and the F-TBP and F-TBP332 proteins could be detected. A representative analysis is shown in Fig. 4B. Under these more stringent conditions, the ratio of 35 S-labeled F-TBP bound to dTAF III 105 and the 110-kDa protein was the same as that previously observed under the less stringent conditions (Fig. 4, compare panels A and B), yet we observed a significant reduction in overall signal (data not shown). When 35 S-labeled F-TBP332 was incubated under these conditions, binding of the 250-, 140-, and 110-kDa polypeptides in the TFIIIC/TFIID fraction to the F-TBP332 was maintained. However, in comparison with the 110-kDa polypeptide, a substantial decrease in binding of F-TBP332 to dTAF III 105 was observed (Fig. 4B). The ability of F-TBP332 to interact with dTAF II 250 and dTAF II 110 and form functional TFIID complexes is consistent with our previous studies, which showed that F-TBP332 was able to support transcription of RNA pol II promoters in Drosophila cells (1). Together, these results demonstrate that the mutation at position 332 in the Drosophila TBP protein disrupts the ability of dTAF III 105 to associate with TBP. DISCUSSION These studies have identified a polypeptide with an apparent molecular mass of 105 kDa that is a subunit of the Drosophila TFIIIB complex, which directly interacts with TBP. Previous studies have identified yeast and human polypeptides that directly associate with TBP and are involved in RNA pol III transcription, and the corresponding genes from several yeast species (BRF/TDS4/PCF4) (15)(16)(17)22) and human (IIIB90) (30,32) have been cloned. Based on the function of dTAF III 105, and its cross-reactivity with the anti-human IIIB90 antibodies, it is likely that this protein represents the Drosophila homolog. These results further argue that this protein does not correspond to the B-TFIID TAF previously identified in mammalian cells (12). In yeast, the yIIIB70 polypeptide directly interacts with TBP in the TFIIIB complex; the yIIIB90 subunit associates with TBP primarily through yIIIB70, although weak TBP-IIIB90 interactions have been detected (19,20). Consistent with these results, we find that dTAF III 105 is the only protein in the TFIIIB fraction that we detect that stably interacts with TBP by far Western blot analysis.
It has been previously shown that Drosophila, unlike human and Xenopus systems, possesses a pronounced dependence on the 5Ј-flanking sequence of tRNA and 5 S RNA genes for their transcription in vitro (47) and in vivo. 2 This observed difference in DNA sequence requirements for RNA pol III transcription between insect and other eucaryotic systems appears to be a result of structural differences between the TFIIIB components. This is reflected by the fact that human and Drosophila components cannot be exchanged in reconstitution assays (48). Thus, detailed comparisons of the subunit structures of the TFIIIB complex from these systems will be necessary to elucidate differences in protein-protein and protein-DNA interactions that are responsible for dictating the differential DNA sequence requirements observed. The identification of a Drosophila TFIIIB subunit will now allow the gene encoding this polypeptide to be cloned and its function to be assessed.
We have constructed and analyzed a mutant Drosophila TBP protein that changes a highly conserved arginine to histidine at residue 332 within the second basic repeat region. This mutation is analogous to the S. cerevisiae TBP mutant R220H reported by Cormack and Struhl (49), which was shown to be defective for tRNA and 5 S RNA synthesis. When Drosophila TBP is overexpressed in Schneider S-2 cells, a significant increase in the expression of transiently transfected tRNA or U6 RNA genes is observed; however, overexpression of the TBP332 protein fails to stimulate either RNA pol III promoter (1). Both TBP proteins are able to enhance the transcription of certain RNA pol II promoters. Thus, these results identified an amino acid residue in Drosophila TBP that is critical either for the formation or function of RNA pol III transcription complexes in vivo. Our present studies demonstrate that extracts derived from the F-TBP stable cell line exhibit an increase in RNA pol III transcription activity, while the F-TBP332 stable line, which expresses the mutant TBP, does not. This provides further evidence that TBP is a limiting component for tRNA gene transcription in Drosophila S-2 cells. This is consistent with our previous studies that have shown that increasing the cellular level of TBP by the activation of cellular kinases produces an increase in TFIIIB activity that is limiting for transcription in vitro (38,40). Using immunoprecipitation assays, we have shown that this TBP mutation disrupts its interaction with a TFIIIB subunit, dTAF III 105. It is therefore likely that this alteration in dTAF III 105-TBP interactions renders the mutant TBP incapable of forming stable TFIIIB complexes and results in its inability to support RNA pol III transcription in vivo and in vitro.
We find that the Drosophila TBP332 mutation disrupts dTAF III 105-TBP interactions in S-2 cells. The fact that this mutation also renders it incapable of supporting both tRNA and U6 RNA transcription (1) provides new evidence to support the idea that the TFIIIB subunit, dTAF III 105, may be necessary for the transcription of both classes of promoters in Drosophila. In S. cerevisiae, the analogous TFIIIB subunit is required for U6 RNA transcription (50,51). This is in contrast to the human system, where the evidence suggests that TFIIIB is differentially used by tRNA and U6 RNA promoters (28,30,32,52,53) and that the hIIIB90 subunit identified is not required for U6 RNA promoters (32). However, we cannot exclude the possibility that the TBP332 mutation affects interactions both with dTAF III 105 and with a modified or variant form of dTAF III 105, which has a selective function in U6 RNA gene transcription. An alternative explanation for our results is that this residue within TBP may also be necessary for its interaction with a component of a distinct complex required for the Drosophila U6 RNA gene promoter. The Drosophila U6 promoter does contain a PSE, and a protein complex that binds to the PSE in the Drosophila U1 and U6 RNA gene promoters has been isolated that appears to be functionally equivalent to the SNAPc-PTF complex in mammalian cells. 3 Both human SNAPc-PTF subunits SNAPc43-PTF␥ (54,55) and SNAPc45-PTF␦ (54,56) directly interact with TBP. Thus, we cannot rule out the possibility that this TBP mutation, which disrupts the ability of TBP to form TFIIIB complexes, also abolishes its ability to form functional SNAPc-PTF complexes in Drosophila. The identification of a TBP mutation that fails to support both tRNA and U6 RNA gene transcription will now direct studies to address the function of dTAF III 105 in the different RNA pol III gene promoter complexes and to further elucidate the nature of the U6 RNA gene transcription complex. Cormack and Struhl (49) previously generated and identified a large number of single amino acid mutations within the highly conserved carboxyl-terminal domain of yeast TBP that rendered it specifically defective for the synthesis of tRNA and 5 S RNA. These mutations map to a large interaction surface of TBP that may associate with one or more RNA pol III transcription components and is composed of regions from both the amino-and carboxyl-terminal direct repeats. Overexpression of yIIIB70 in yeast strains harboring some of these TBP mutations was shown to partially restore tRNA synthesis. Subsequent studies in vitro revealed that two of these mutations within the amino-terminal direct repeat of yeast TBP, F155S and K138L, impaired its ability to interact with a recombinant carboxyl-terminal domain of yIIIB70 but not with the fulllength yIIIB70 protein (22). Kim and Roeder (57) have also shown that the yeast TBP K138L mutation is defective for tRNA and 5 S RNA transcription. In contrast, and as a complement to these studies, we provide the first direct evidence that a mutation within the carboxyl-terminal direct repeat region of TBP abolishes TAF-TBP interactions. Importantly, this disruption of TAF-TBP interactions is observed both in the context of the native TFIIIB complex and in a direct interaction assay. Together with the previous analysis in yeast, these results demonstrate that this TFIIIB subunit can interact with domains contributed by both the amino-and carboxyl-terminal direct repeats of TBP. These regions are probably part of a pocket that interacts with dTAF III 105 or its human or yeast homologs.