Alignment of the B“ Subunit of RNA Polymerase III Transcription Factor IIIB in Its Promoter Complex
1999; Elsevier BV; Volume: 274; Issue: 40 Linguagem: Inglês
10.1074/jbc.274.40.28736
ISSN1083-351X
AutoresSheila M.A. Shah, Ashok Kumar, E. Peter Geiduschek, George A. Kassavetis,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoTFIIIB, the central transcription initiation factor of the eukaryotic nuclear RNA polymerase (pol) III is composed of three subunits: the TATA-binding protein; Brf, the TFIIB-related subunit; and B", the Saccharomyces cerevisiae,TFC5 gene product. The orientation of the B" subunit within the TFIIIB-DNA complex has been analyzed at two promoters by two approaches that involve site-specific photochemical protein-DNA cross-linking: a collection of B" internal and external deletion proteins has been surveyed for those deletions that alter the interaction of B" with DNA or change the orientation of B" relative to DNA; a method for regionally mapping cross-links between specific DNA sites and 32P-end-labeled protein has also been applied. The results map an N-proximal segment of B" to the upstream end of the TFIIIB-DNA complex and amino acids 299–315 to the principal DNA-contact site, approximately 8 base pairs upstream of the TATA box. The analysis also indicates that a segment comprising amino acids 316–434 loops away from DNA, and locates the C-proximal 170 amino acids of B" downstream of the TATA box. Examination of two-cross-link products formed by DNA with adjacent and nearby photoactive nucleotides supports the conclusion that Brf and B" share an extended interface along the length of the TFIIIB-DNA complex. TFIIIB, the central transcription initiation factor of the eukaryotic nuclear RNA polymerase (pol) III is composed of three subunits: the TATA-binding protein; Brf, the TFIIB-related subunit; and B", the Saccharomyces cerevisiae,TFC5 gene product. The orientation of the B" subunit within the TFIIIB-DNA complex has been analyzed at two promoters by two approaches that involve site-specific photochemical protein-DNA cross-linking: a collection of B" internal and external deletion proteins has been surveyed for those deletions that alter the interaction of B" with DNA or change the orientation of B" relative to DNA; a method for regionally mapping cross-links between specific DNA sites and 32P-end-labeled protein has also been applied. The results map an N-proximal segment of B" to the upstream end of the TFIIIB-DNA complex and amino acids 299–315 to the principal DNA-contact site, approximately 8 base pairs upstream of the TATA box. The analysis also indicates that a segment comprising amino acids 316–434 loops away from DNA, and locates the C-proximal 170 amino acids of B" downstream of the TATA box. Examination of two-cross-link products formed by DNA with adjacent and nearby photoactive nucleotides supports the conclusion that Brf and B" share an extended interface along the length of the TFIIIB-DNA complex. The eukaryotic (nuclear) RNA polymerases are brought to their promoters by relatively complex core transcription apparati. In the RNA polymerase (pol) 1The abbreviations used are:polpolymeraseTFtranscription factorTBPTATA-binding proteinbpbase pair(s)PAGEpolyacrylamide gel electrophoresisNTCB2-nitro-5-thiocyanobenzoic acidABdUMP5-[N-(p-azidobenzoyl)-3-aminoallyl]-deoxyuridine monophosphate III transcription system of Saccharomyces cerevisiae, which is the focus of this work, the polymerase recruitment function is executed by the core transcription factor (TF) IIIB. TFIIIC and TFIIIA, the other components of the core transcription apparatus, bind DNA and serve as assembly factors for TFIIIB; the six-subunit TFIIIC interacts directly with TFIIIB, and TFIIIA serves as a 5 S rRNA gene-specific platform for TFIIIC. TFIIIB is composed of three subunits: the TATA-binding protein (TBP), Brf, and B" (Tfc5). TBP co-directs binding of TFIIIB to very strong TATA boxes. At promoters that lack these intrinsic TBP-binding sites, TFIIIC functions as an assembly factor that deposits TFIIIB on its upstream DNA site (reviewed in Ref. 1White R.J. RNA Polymerase III Transcription. 2nd Ed. Springer-Verlag/Landes Bioscience, New York1998Crossref Google Scholar). In both kinds of situations, TBP is located close to DNA (2Persinger J. Sengupta S.M. Bartholomew B. Mol. Cell. Biol. 1999; 18: 5218-5234Crossref Scopus (26) Google Scholar). polymerase transcription factor TATA-binding protein base pair(s) polyacrylamide gel electrophoresis 2-nitro-5-thiocyanobenzoic acid 5-[N-(p-azidobenzoyl)-3-aminoallyl]-deoxyuridine monophosphate The 596-amino acid Brf is joined from two evolutionarily distinct parts. Its designation as the TFIIB-relatedfactor derives from the homology of its N-proximal half to TFIIB (3Buratowski S. Zhou H. Cell. 1992; 71: 221-230Abstract Full Text PDF PubMed Scopus (109) Google Scholar, 4Colbert T. Hahn S. Genes Dev. 1992; 6: 1940-1949Crossref PubMed Scopus (128) Google Scholar, 5López-De-León A. Librizzi M. Puglia K. Willis I.M. Cell. 1992; 71: 211-220Abstract Full Text PDF PubMed Scopus (108) Google Scholar). Just as TFIIB is able to recruit pol II to the transcriptional start site, so the principal polymerase recruitment capacity of TFIIIB resides in the corresponding, N-proximal, half of Brf. The C-proximal half of Brf is pol III-specific (6Khoo B. Brophy B. Jackson S.P. Genes Dev. 1994; 8: 2879-2890Crossref PubMed Scopus (108) Google Scholar), and has no sequence-homologous counterparts in the pol I and pol II transcription apparati. The principal TBP and B" affinities of Brf, and also its TFIIIC affinity, reside in this C-proximal half (7Kassavetis G.A. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1998; 18: 5587-5599Crossref PubMed Scopus (60) Google Scholar). However, the N-proximal half of Brf alone has sufficient affinity for TBP and B" to form a TFIIIB-DNA complex at a strong TATA box that is able to recruit pol III to accurately initiated transcription (7Kassavetis G.A. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1998; 18: 5587-5599Crossref PubMed Scopus (60) Google Scholar). Since TFIIB and the N-proximal half of Brf sit in corresponding locations in their respective DNA complexes (7Kassavetis G.A. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1998; 18: 5587-5599Crossref PubMed Scopus (60) Google Scholar, 8Nikolov D.B. Chen H. Halay E.D. Usheva A.A. Hisatake K. Lee D.K. Roeder R.G. Burley S.K. Nature. 1995; 377: 119-128Crossref PubMed Scopus (484) Google Scholar), and exercise similar functions, it is surprising that Brf and TBP are not by themselves competent to direct transcriptional initiation by pol III. In fact, the 594-amino acid B" is absolutely required for transcription by pol III, in duplex DNA or chromatin, at TATA-containing and TATA-less promoters, in vivo as well as in vitro. It is also B" that makes the TFIIIB-DNA complex extraordinarily stable (9Kassavetis G.A. Braun B.R. Nguyen L.H. Geiduschek E.P. Cell. 1990; 60: 235-245Abstract Full Text PDF PubMed Scopus (360) Google Scholar). In the experiments that are reported here, we have mapped B" along its DNA site in the TFIIIB-DNA complex; B" external and internal deletion proteins that form stable TFIIIB-DNA complexes have been examined by photochemical protein-DNA cross-linking along the extended DNA site in order to find out which of these deletions alter the interaction of B" with DNA or change the orientation of B" relative to DNA. A relatively simple and highly sensitive method has also been developed to map cross-links from specific DNA sites to their targeted region on32P-end-labeled protein. TheSUP4 tRNA gene was polymerase chain reaction-amplified from plasmid pTZ1 (10Kassavetis G.A. Riggs D.L. Negri R. Nguyen L.H. Geiduschek E.P. Mol. Cell. Biol. 1989; 9: 2551-2566Crossref PubMed Scopus (185) Google Scholar) with primers creating a BspE1 site beginning at bp −64 (+1 designating the start site of transcription) and an NsiI site beginning at bp +124. The DNA was purified on native 5% polyacrylamide gel, passively eluted (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar), restricted with BspEI andNsiI, and 5′ end-labeled with T4 polynucleotide kinase. The 191-nucleotide non-transcribed and 183-nucleotide transcribed strands were separated on 5% polyacrylamide, 8 m urea gel, excised, and recovered by passive elution (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). The transcribed strand was used to construct DNA with site-specifically placed photoactive nucleotides, as described (12Bartholomew B. Tinker R.L. Kassavetis G.A. Geiduschek E.P. Methods Enzymol. 1995; 262: 476-494Crossref PubMed Scopus (34) Google Scholar). Cross-linking probes were synthesized by annealing specific oligonucleotides to the transcribed strand. Primer extension to add ABdUMP (5-[N-(p-azidobenzoyl)-3-aminoallyl]-deoxyuridine monophosphate) and [32P]dNMP utilized 1 unit of exonuclease-free Klenow fragment (incubation for 5 min at 37 °C with 10 μm ABdUTP and 0.5 μm appropriate [α-32P]dNTP). Extension of the photoactive DNA strand was completed by an additional 10-min incubation with 500 μm unlabeled dNTPs. Probes were generated with ABdUMP at bp −38/−37,-33/−32, −30, −26, −22/−21, −19/−17, −14/−12, and −3/−2 of the non-transcribed strand (Fig. 1 A). For probes −19/−17, −14/−12, and −3/−2, this process left an upstream, 22-nucleotide 3′ overhang. An upstream oligonucleotide (the same as the primer for making the −38/−37 probe) was annealed and ligated to these constructs as the final step in preparing the corresponding photoactive 183-bp DNA. SNR6 photoprobes −39/−38, −33, −28, −13/−12, and −5 were made as just described, using the appropriate non-transcribed strand oligonucleotides annealed to an 88-mer transcribed strand spanning bp −56 to bp +32. SNR6 photoprobes −42 and −23/−22 were made using the appropriate non-transcribed strand oligonucleotide primer annealed to a 60-mer transcribed strand template spanning bp −58 to bp +2. Expression plasmids for B"-(1–370) and B"-(371–594) were constructed as described and referenced for other B" deletions (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). The corresponding proteins were overproduced inEscherichia coli BL21(DE3), (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). TFIIIC, wild-type TBP, TBPm3, Brf, Brf-(1–282), Brf-(284–596), truncated and full-length B" were purified as described (7Kassavetis G.A. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1998; 18: 5587-5599Crossref PubMed Scopus (60) Google Scholar, 13Kassavetis G.A. Nguyen S.T. Kobayashi R. Kumar A. Geiduschek E.P. Pisano M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9786-9790Crossref PubMed Scopus (91) Google Scholar, 14Kassavetis G.A. Bardeleben C. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1997; 17: 5299-5306Crossref PubMed Scopus (47) Google Scholar, 15Whitehall S.K. Kassavetis G.A. Geiduschek E.P. Genes Dev. 1995; 9: 2974-2985Crossref PubMed Scopus (59) Google Scholar). Concentrations of TBP, full-length B" and B"-(138–594) were determined by their predicted extinction coefficients. Concentrations of TBPm3, Brf, and other B" deletion proteins were estimated by Coomassie stain on SDS-polyacrylamide gels, standardized to bovine serum albumin. TFIIIC concentration was measured as described (9Kassavetis G.A. Braun B.R. Nguyen L.H. Geiduschek E.P. Cell. 1990; 60: 235-245Abstract Full Text PDF PubMed Scopus (360) Google Scholar). Ten pmol of B"-(138–594) was incubated for 30 min at 21 °C with 25 units of bovine heart muscle kinase and 5 μm [γ-32P]ATP (6,000 Ci/mmol) in buffer containing 20 mm Tris-Cl (pH 7.5), 10 mm MgCl2, 100 mm NaCl, and 2.5 mm dithiothreitol (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar) for 32P labeling of its natural phosphorylation site at Ser-164. Unincorporated [γ-32P]ATP was removed, and efficiency of phosphorylation determined as described (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). Protein-SUP4gene complexes were allowed to form for 60 min at 21 °C in 20 μl of pol III buffer (40 mm Tris-Cl (pH 8.0), 100 mm NaCl, 7 mm MgCl2, 3 mm β-mercaptoethanol, 4–6% (v/v) glycerol, and 100 μg/ml bovine serum albumin) containing 0.2–2 fmol of SUP4photoprobe, 18.7 fmol of TFIIIC, 200 fmol of B" (full-length or mutant), 50 fmol of TBP, 64 fmol of Brf, and 200 ng of pLNG56 (nonspecific carrier) DNA (9Kassavetis G.A. Braun B.R. Nguyen L.H. Geiduschek E.P. Cell. 1990; 60: 235-245Abstract Full Text PDF PubMed Scopus (360) Google Scholar). After incubation with heparin (200 μg/ml), samples were UV-irradiated for 5 min, as described (16Bartholomew B. Kassavetis G.A. Geiduschek E.P. Mol. Cell. Biol. 1991; 11: 5181-5189Crossref PubMed Scopus (125) Google Scholar). A 6-μl aliquot of each sample was loaded on non-denaturing 4% polyacrylamide gel containing 20 mm Tris-HCl (pH 8.0), 2 mm EDTA, and 4% (v/v) glycerol (with 20 mmTris, 2 mm EDTA running buffer). The remaining material of each sample was treated with DNase I (7.5 units) and S1 nuclease (80 units), and resolved by 8% SDS-PAGE. 32P-Tagged (cross-linked) proteins were visualized and quantified by phosphoimaging using software provided with the imager. TFIIIB complexes were assembled on SNR6 photoprobes with 400 fmol of TBPm3, 64 fmol of Brf, 200 fmol of B" (full-length or mutant), and 5 fmol of an SNR6 photoprobe, incubated for 60 min at 21 °C in 20 μl of pol III buffer containing 50–70 mm (instead of 100 mm) NaCl and 100 ng of poly(dG-dC)·poly(dG-dC) (instead of pLNG56). Reaction mixtures were treated with 200 ng of poly(dA-dT)·poly(dA-dT) (in place of heparin), UV-irradiated, and processed as described above and elsewhere (12Bartholomew B. Tinker R.L. Kassavetis G.A. Geiduschek E.P. Methods Enzymol. 1995; 262: 476-494Crossref PubMed Scopus (34) Google Scholar). Photochemically cross-linked reaction mixtures were prepared with 800 fmol of TBPm3, 64 fmol of Brf, 80 fmol of32P-labeled B"-(138–594), and 50–100 fmol of the appropriate unlabeled SNR6 photoprobe in 20 μl of pol III buffer (with 40–50 mm in place of 100 mmNaCl). For protein complexes with the SUP4 gene, 30 fmol of the appropriate unlabeled photoprobe was incubated with 18.7 fmol of TFIIIC, 100 fmol of TBP, 64 fmol of Brf, and 33 fmol of32P-labeled B"-(138–594) in 20 μl of pol III buffer (with 100 mm NaCl). After UV irradiation, 0.6 pmol of unlabeled B" was added (to reduce the background due to free32P-labeled B") and proteins were resolved by 6% SDS-PAGE to separate DNA-cross-linked 32P-labeled B" from free B". The corresponding bands were visualized by phosphoimaging and excised from the gel. Proteins were passively eluted overnight into buffer containing 10 mm Tris-Cl (pH 8.0), 0.2% (w/v) SDS, 0.1 mm dithiothreitol, and 25 μg/ml bovine serum albumin, and concentrated by ultrafiltration. Eluted proteins were subjected to partial cleavage with 100 mm CNBr or 10 mm NTCB. For CNBr partial cleavage, samples were brought to low pH with 1% SDS and 0.05n HCl. CNBr (100 mm) was then added for 20 min at 21 °C. Further reaction was quenched by adding 0.25 volumes of 100 mg/ml dithiothreitol in 1 m NaHepes (pH 7.8). For NTCB partial cleavage, samples were treated with 10 mm NTCB for 30 min at 37 °C for cyanylation of cysteine residues. The pH was then adjusted to 9.0 with Tris base, and samples were incubated overnight at 37 °C for the subsequent proteolytic cleavage. The32P-labeled fragments generated by NTCB or CNBr cleavage were resolved on 11–15% SDS-polyacrylamide gels. Comparisons of cleavage patterns of free and DNA-cross-linked 32P-labeled B" were made on phosphoimage profiles. Previous examination of the internal structure of a TFIIIB-DNA complex by photochemical cross-linking revealed that B" cross-links to DNA between 43 and 2 bp upstream of the transcriptional start site (16Bartholomew B. Kassavetis G.A. Geiduschek E.P. Mol. Cell. Biol. 1991; 11: 5181-5189Crossref PubMed Scopus (125) Google Scholar, 17Kassavetis G.A. Kumar A. Letts G.A. Geiduschek E.P. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 9196-9201Crossref PubMed Scopus (54) Google Scholar). In order to map the locations of individual domains of B" relative to DNA, we have analyzed the photochemical cross-linking patterns of B" with internal and external deletions, using ABdUMP incorporated at specific sites along the TFIIIB-binding sites of the SUP4 andSNR6 genes (Fig. 1). Each DNA photoprobe also has a radioactive nucleotide incorporated next to, or in close vicinity to, its photoactive nucleotide(s). TFIIIB complexes containing either full-length or truncated B" were assembled on each of these DNA photoprobes; reaction mixtures were then UV-irradiated, digested with nucleases, and analyzed by SDS-PAGE. The efficiency of TFIIIB-DNA complex formation was monitored in parallel by electrophoretic mobility-shift analysis of an UV-irradiated, but not nuclease-treated, portion of each reaction mixture. If deletion of a specific segment of B" results in the loss of cross-linking at a particular DNA site, this suggests localization of that peptide segment in the vicinity of that DNA site (so long as complex formation is not correspondingly diminished). In the case of the SNR6 gene promoter, unidirectional assembly of TFIIIB-DNA complexes is specified by a modified TATA-box (TGTAAATA) in conjunction with the mutant TBPm3 (15Whitehall S.K. Kassavetis G.A. Geiduschek E.P. Genes Dev. 1995; 9: 2974-2985Crossref PubMed Scopus (59) Google Scholar). At the SUP4 tRNA promoter, with its very weak TATA box, TFIIIB-DNA complex assembly requires the transcription factor TFIIIC, and is unidirectional (15Whitehall S.K. Kassavetis G.A. Geiduschek E.P. Genes Dev. 1995; 9: 2974-2985Crossref PubMed Scopus (59) Google Scholar). Efficiencies of forming heparin-stable TFIIIB-DNA complexes and photochemical cross-linking of each B" deletion mutant and wild type B" were compared for the photoactive DNA probes shown in Fig. 1 A. B" lacking its N-terminal 262 amino acids or its C-terminal 130 amino acids remains competent to assemble into a heparin-resistant TFIIIB-DNA complex via the TFIIIC-dependent assembly pathway on the SUP4gene (but does so relatively inefficiently; Ref. 11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). Full-length B" cross-linked most efficiently to ABdUMP positioned between bp −38 and bp −32 on the SUP4 gene (Fig.2 A; cf. Ref. 16Bartholomew B. Kassavetis G.A. Geiduschek E.P. Mol. Cell. Biol. 1991; 11: 5181-5189Crossref PubMed Scopus (125) Google Scholar). Removal of 223 or 262 amino acids from the N terminus of B" increased the efficiency of cross-linking between bp −38 and bp −30 more than 2-fold compared with full-length B" (adjusted for differences in formation of heparin-resistant complexes; TableI, part A, and Fig. 2 A). Further downstream, between bp −26 and −2, the cross-linking efficiencies of B"-(224–594) and full-length B" did not differ significantly.Table IB" deletions: effects on patterns of cross-linking to the SUP4 and SRN6 genesA. B" fragments and internal deletionsaCross-linking efficiency normalized to formation of the heparin-resistant TFIIIB complex, which ranged between 0.5 and 1.5 times full-length B", except as otherwise noted in footnotesb and c.SUP4 photoprobe−38/−37−33/−32−30−26−22/−21−19/−17−14/−12−3/−2224–594+++++++++++263–594++++++bHeparin-resistant TFIIIB complex formation ranged between 0.1 and 0.4 times full-length B", preventing quantification of cross-linking downstream of bp −30.bHeparin-resistant TFIIIB complex formation ranged between 0.1 and 0.4 times full-length B", preventing quantification of cross-linking downstream of bp −30.bHeparin-resistant TFIIIB complex formation ranged between 0.1 and 0.4 times full-length B", preventing quantification of cross-linking downstream of bp −30.bHeparin-resistant TFIIIB complex formation ranged between 0.1 and 0.4 times full-length B", preventing quantification of cross-linking downstream of bp −30.bHeparin-resistant TFIIIB complex formation ranged between 0.1 and 0.4 times full-length B", preventing quantification of cross-linking downstream of bp −30.1–487+++cClose proximity to Brf on the separation gel prevented quantification for DNA sites that weakly cross-link to full-length B".cClose proximity to Brf on the separation gel prevented quantification for DNA sites that weakly cross-link to full-length B".cClose proximity to Brf on the separation gel prevented quantification for DNA sites that weakly cross-link to full-length B".cClose proximity to Brf on the separation gel prevented quantification for DNA sites that weakly cross-link to full-length B".cClose proximity to Brf on the separation gel prevented quantification for DNA sites that weakly cross-link to full-length B".40–487+++NDND+++Δ291–310−+/−+/−+++++Δ253–269, Δ312–325,++++++++Δ329–338, Δ338–353,Δ355–372, Δ370–387,Δ388–409B. B" fragments and internal deletionsdCross-linking efficiency normalized by comparing B"/Brf cross-linking ratios to those obtained with full-length B".SNR6 photoprobe−39/−38−33−28−13/−12−5186–594+++++++224–594+++++/−+/−263–594++ND+/−+/−40–487+++++Δ272–292+/−++++Δ409–421+ND+++/−Δ424–438+ND+/−+/−+Δ253–269, Δ291–310, Δ312–325, Δ327–338, Δ340–355, Δ355–372, Δ372–387, Δ388–409, Δ438–449+ND+++++, >2× cross-linking relative to full-length B"; +, 0.5–2× full-length B"; +/−, 0.1–0.4× full-length B"; −, 2× cross-linking relative to full-length B"; +, 0.5–2× full-length B"; +/−, 0.1–0.4× full-length B"; −, <0.1× full-length B"; ND, not done. The combination of low level formation of heparin-resistant TFIIIB-DNA complexes with B"-(263–594) and low efficiency cross-linking of B" between bp −26 and bp −2, resulted in cross-linking signals too close to background to be reliably quantified, but the observed levels were consistent with unimpaired cross-linking of B"-(263–594) at these positions (Table I, part A). C-terminally truncated B"-(1–464) and B"-(1–487) were not sufficiently well resolved from Brf in SDS-PAGE to quantify reliably. It was, however, possible to assess that B"-(1–487) was not significantly impaired in cross-linking to bp −38/−37, −33/−32, and −30, where B" cross-links efficiently (Table I, part A). N-and-C-terminally truncated B"-(40–487) did resolve well from Brf, and its cross-linking to all probes tested was unimpaired (TableI, part A, and Fig. 2 B). These results indicate that the cross-links of B" to the SUP4 gene (more specifically to the bp −38/−2 segment of this gene) primarily involve its amino acid 224–487 segment; this is also the active core of B" forSUP4 transcription (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). Internal deletion mutants of B", which span much of the region of B" not covered by the N- and C-terminal truncations, and which are capable of being assembled into heparin-resistant DNA complexes, were also examined. Only B" with amino acids 291–310 deleted displayed a defect in cross-linking, predominantly with probe −38/−37 for which cross-linking efficiency is reduced more than 10-fold (Fig. 2 C, lanes 3 and 5) and, to a lesser extent, with probes −33/−32 and −30 (Table I, part A). Fig. 2 C shows two other aspects of this defect. 1) Heparin did not diminish cross-linking of intact B" (lanes 2 and 3), but substantially reduced cross-linking of B"Δ291–310 without substantially reducing the cross-linking of Brf (lanes 4 and 5). 2) Full-length B" displaced the 120-kDa subunit of TFIIIC from the vicinity of the photoactive nucleotide in the −39/−38 photoprobe (comparelanes 1 and 2), but B"Δ291–310 did not (lane 4). On the SNR6 gene (18Brow D.A. Guthrie C. Genes Dev. 1990; 4: 1345-1356Crossref PubMed Scopus (110) Google Scholar, 19Margottin F. Dujardin G. Gerard M. Egly J.M. Huet J. Sentenac A. Science. 1991; 251: 424-426Crossref PubMed Scopus (117) Google Scholar), B" truncated by N-terminal deletion to amino acid 186 or by C-terminal deletion to amino acid 487 was not deficient in cross-linking (Table I, part B). N-terminal truncation to amino acid 224 or 263 somewhat lowered cross-linking at bp −13/−12 and −5 (20–40% of that obtained with intact B"; Table I, part B), but a comparable depression of cross-linking was not observed when TFIIIB complexes with theSUP4 gene were probed at the close-by −14/−12 and −3/−2 sites (Table I, part A). Since no single region of B" is required for TFIIIB-SNR6 DNA complex formation, a more complete set of internal deletions, spanning regions not covered by N- and C-terminal truncations, could be examined at this promoter. B" cross-links most efficiently to bp −39/−38 on the SNR6 gene (7Kassavetis G.A. Kumar A. Ramirez E. Geiduschek E.P. Mol. Cell. Biol. 1998; 18: 5587-5599Crossref PubMed Scopus (60) Google Scholar); only at bp −33 and −13/−12 do cross-linking efficiencies exceed 10% of cross-linking to bp −39/−38. Only B"Δ272–292 was deficient in cross-linking to probe −39/−38 (20% of the efficiency of cross-linking to intact B"; Table I, part B, and Fig. 2 D). Curiously, B"Δ291–310, which was deficient for cross-linking at a similar location on the SUP4 gene, was not significantly impaired for cross-linking on the SNR6 gene (Table I and Fig. 2; B"Δ272–292 was not examined on the SUP4 gene because it does not assemble into a stable, heparin-resistant complex; Ref. 11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). B"Δ424–438 displayed somewhat lower cross-linking to bp −28 and −13/−12 (20–40% of the cross-linking efficiency with intact B") and B"Δ409–421 was somewhat impaired in cross-linking to bp −5 (40% efficiency relative to intact B"). Comparisons of cross-linking to the SUP4 and SNR6genes are complicated by the fact that TFIIIC positions TFIIIB somewhat heterogeneously onto the AT-rich upstream sequence of theSUP4 gene (20Joazeiro C.A.P. Kassavetis G.A. Geiduschek E.P. Genes Dev. 1996; 10: 725-739Crossref PubMed Scopus (75) Google Scholar), whereas assembly on the SNR6 gene directed by TBPm3 is fixed at the TGTA box. One would therefore expect a "blurring" in the B" cross-linking pattern to the SUP4gene relative to SNR6 (i.e. efficient cross-linking between bp −38 and −30 on the SUP4 gene, but predominantly at bp −39/−38 on the SNR6 gene). Conversely, even if as few as 2% of TFIIIB complexes on the SNR6 gene were bound in the reverse orientation (cf. Ref. 15Whitehall S.K. Kassavetis G.A. Geiduschek E.P. Genes Dev. 1995; 9: 2974-2985Crossref PubMed Scopus (59) Google Scholar), they might make a significant contribution to cross-linking at bp −13/−12 (through the B" segment that cross-links efficiently to bp −39/−38 in the opposite orientation of TFIIIB). B" assembly into a TFIIIB-DNA complex buries two regions that are more surface-exposed in the free protein (as judged by hydroxyl radical footprinting): region I, covering amino acids ∼390–470; and region II, covering amino acids ∼270–305 (11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). The above experiments clearly suggest that region II lies close to the major sites of B" cross-linking on the SNR6 andSUP4 genes (since B"Δ291–310 depressed cross-linking to the SUP4 gene and B"Δ272–292 to the SNR6gene). In order to gain further information about the location of region II relative to DNA, B" was split at amino acid 370 (which is highly accessible to hydroxyl radical cleavage in TFIIIB-DNA complexes; Ref. 11Kumar A. Kassavetis G.A. Geiduschek E.P. Hambalko M. Brent C.J. Mol. Cell. Biol. 1997; 17: 1868-1880Crossref PubMed Scopus (45) Google Scholar). Heparin-resistant TFIIIB-DNA complexes can be formed on both the SUP4 and SNR6 genes only when both parts of this split B" (amino acids 1–370 and 371–594) are provided together. The transcriptional activity of these TFIIIB-DNA complexes is, however, diminished (data not shown). Fig. 3compares the cross-linking profiles of wild type and split B" on theSUP4 and SNR6 genes. Cross-linking to theSUP4 gene at bp −38/−37, −33/−32, and −30 is contributed almost entirely by amino acids 1–370 of B", but there is some cross-linking of the C-terminal amino acids 371–594 to bp −33/−32 (Fig. 3 A). The weak cross-linking of intact B" to sites downstream of bp −30 on the SUP4 gene appears to involve solely its C-terminal half. Similarly, B"-(1–370) dominates cross-linking at bp −39/−38 of the SNR6 gene (Fig.3 B), but some cross-linking of the C-terminal segment is also apparent at this site. In contrast to the cross-linking pattern at bp −33/−32 of the SUP4 promoter, most of the cross-linking of B" at bp −33 of SNR6 is contributed by its C-terminal segment. As on the SUP4 gene, most of the C-terminal B" segment cross-links also to bp −28, −13/−12, and −5, but l
Referência(s)