Comparing transcriptional initiation by RNA polymerases
E Peter Geiduschek and George A Kassavetis University of California at San Diego, La Jolla, USA We comment on the current understanding of transcriptional initiation by RNA polymerases I and Ill, and look for common modes of operation of these enzymes, emphasizing selected recent developments. These include definitive experiments on the constitution of the human RNA polymerasel transcription factor SLl/TIF-IB, the development of a genetic system for analyzing the fun~ion of RNA ~lymera~l in yeast, the elucidation of the structure of the human snRNA gene transcription factor SNAP,, and initial stages of mapping the protein-protein interactions involved in the assembly of transcriptional initiation complexes.
Current Opinion in Cell Biology 1995, 7:344-351
Introduction Important strides have been taken recently in enumerating the composition of the core components of the eukaryotic nuclear transcription machinery, and in understanding their individual roles in the initiation of RNA synthesis. The most rapid early progress in this endeavor, it is fair to say, has come from the analysis of RNA polymerase (pal) III. In this short article, we summarize the statns of certain aspects of current work on pol I and pol III. We comment on early experiments to identify and map the key protein-protein interactions that are required in order to assemble pol I and pol III at their respective promoters, relate how a combined genetic and biochemical approach to the pol I transcription apparatus of yeast is progressing, and discuss certain new results in the composition of pol I and pol III transcription factors. Our side-by-side presentation of these two RNA polymerases serves as a comparison of sorts, but we also comment on functional homologies in pol I and pol III transcription in a closing section.
The basal transcription apparatus of RNA pol ymerase I I I The RNA polymerase III enzymes are the most complex of the nuclear polymerases in the sense of comprising the largest number of subunits; their basal transcription machinery includes the core transcription factors (TF)IIIA, IIIB and IIIC, and promoters that are, with
a few prominent exceptions, located within transcribed DNA (reviewed in [l-4]). In discussing transcription by pol III, it is convenient to separate out the vertebrate U6 snRNA genes and deal first with everything else.
TFIIIA, the 5s rRNA gene-specific transcription &tor, serves as the pladorm on which that gene’s transcription initiation complex is assembled. The TFIIIAs contain nine zinc fingers; Xenopus TFIIIA, the prototype and most studied of these proteins, forms an irregularly spaced complex with its extensive (-45 bp), punctuated promoter, a manifestation of the structural adaptability of zinc finger arrays. An even more striking manifestation of that adaptability is the use of an entirely separate set of zinc finger interactions with the 55 rRNA itself to form the 7s ribonucleoprotein oocyte storage particle [5*].
The general transcription factor TFIIIC binds to the 6oxA and boxB promoter elements of tRNA genes (and of certain other genes with the same promoter organization), and to the TFIIIA-DNA complexes at 5s rRNA genes. It is TFIIIC that directs TFIIIB to its position in the transcription initiation complex, upstream of the transcriptional start site; when polII1 is positioned over the transcriptional start site, it is nestled between TFIIIB and TFIIIC (Fig. 1). Experiments with the yeast (Saccharomycescerwisiae) transcription factors show that TFIIIB alone is the central transcription factor of pol III,
Abbreviations h-human; t-MC--high
mobility group; pal-polymerase; PSE-essential proximal element; PTF-PSE-binding transcription factor; SNAPCsnRNA gene activation protein complex; snRNA-small nuclear RNA; TBP-TATA-binding protein; TF-transcription factor; TIF-l-transcriptional initiation factor for RNA pol I; UBF-u~tream-binding factor; x-Xenopus.
0 Current Biology Ltd ISSN 0955-0674
Comparing transcriptional initiation by RNA polymerases I and III Geiduschek and Kassavetis 345 in the sense that it alone specifically recruits this polymerase to the transcriptional start site. TFIIIC serves as the assembly factor for TFIIIB, and TFIIIA serves as a 5S rlKNA gene-specific adaptor for TFIIIC . It is widely believed that what has been demonstrated for yeast holds true generally (e.g. [1,4]). Proof of generality would be desirable but there has been little progress in that direction. In fact, (indirect) evidence that Xenopus and yeast TFIIIB bind at approximately the same place, relative to the transcriptional start site, has only been obtained recently . IfTFIIIB is generally the central transcription factor ofpol III, that is consistent with its being the target ofphosphorylation (and thereby inactivation) by mitosispromoting factor . The TFIIICs are multisubunit assemblies: the yeast factor (-500 kDa) has six subunits and the genes encoding its three largest subunits are essential [9,10]. The functionally homologous human protein (named TFIIIC2) is comparably complex. Remarkably, the human 230 kDa subunit that binds to the box B promoter element and the box B binding subunit of yeast TFIIIC have no recognizable amino acid sequence identity [11,12]. The general disposition of subunits of yeast TFIIIC on a tlKNA gene and on the 5S rlKNA gene has been analyzed by site-specific DNA-protein photo-crosslinking (reviewed in ; ). Most important for our purpose is the finding that the second largest TFIIIC subunit (120 kDa; designated TFC120) projects upstream of the transcriptional start site and is therefore the prime candidate for interacting directly with TFIIIB.
TFIIIB is the pollII transcription factor that contains the TATA-binding protein (TBP). S. cerevisiae TFIIIB also comprises two other proteins: Bff (molecular mass 67 kDa; encoded by B R F 1 / P C F 4 / T D S 4 ) and an -90kDa protein . The cloning and sequencing of BRF1 genes from the evolutionarily diverged yeasts Kluyveromyces lactis and Candida albicans exposes four regions of particularly conserved sequence , including the already known amino-proximal homology to TFIIB . The C. albicans gene substitutes functionally for S. cerevisiae BRF1 . TBP and Brf associate relatively tightly but are separable from the -90 kDa protein under mild conditions [15,16]; it is not certain that yeast TFIIIB is a single stably associated entity.
Protein-protein interactions Current experiments on the specific protein-protein interactions in the TFIIIC--~TFIIIB--)poIIII reaction sequence that places yeast pol III over the transcriptional start site (Fig. 1) show that Bffand TFC120 interact primarily through the amino-proximal half of Brf (the half that shares homology with TFIIB) and the amino-terminal 165 amino acids of TFC120, and that TBP binds to Brf primarily through the latter's carboxy-terminal half (N Chaussivert, C Conesa, S Shaaban, A Sentenac, personal communication). (According to a partly conflicting recent report, however, these interactions are dispersed along the lengths of Brf and TFC120 .) Two-hybrid
TFC ( 70
rest of TFIIIC
l 111128 111160 ~ t 9!)5 Current Opiitioll in Ce[[ Biology
Fig. 1. Protein-protein interactions in the formation of pollll transcription initiation complexes. TFIIIC, TFIIIB and pollll are oriented in accordance with their alignment along DNA , as indicated at the top of the figure. The transcriptional start site (+1), boxA, and distances in bp are marked. Recently identified protein-protein interactions are indicated by black bands. Relevant components of TFIIIC (TFC), TFIIIB (TFB) and pol I11 (t11)are indicated by their known or estimated (-) molecular masses. Our expectation that TFB-90 will be found to interact with pol III is indicated by the black band and question mark. The amino-proximal parts of Brf and TFC120 are indicated (-N).
Nutleusand gene expression analyses also show that Brfis able to bind to the 34 kDa subunit of pol III [17*], which is located at the furthest upstream end of promoter-bound RNA polymerase, as judged by photo-crosslinking [18*]. The 34 kDa subunit is part of a complex of three interacting subunits that are specific to polII1, in the sense that neither they nor &m&ally related proteins are present in poll or ~0111 . Thus, the interaction of the 34kDa polII1 protein with Brf may account for the topography and for the polymer~e-spec~ci~ of the recrui~ent function of TFIIIB.
Yeast TFIIIB binds autonomously to an extended TATA box, such as is present in the yeast U6 snRNA gene, and can serve, in this instance, as the sole essential transcription factor of ~01111 in vitro . However, it functions primarily at promoters with weakly TATA-like or TATA-less sequences, at which it must be correctly placed by TFIIIC. Either mode of presentation unmasks an avidity of TFIIIB for DNA of widely diverse sequence that would, if manifested outside of the assembly pathway, dissipate yeast TFIIIB to randomly located and unproductive sites. How that is prevented is one of the interesting puzzles in understanding pol III promoters; structural transitions that expose cryptic ~NA-binding sites may be involved . Two examples of such transitions come to mind. First, the ~70 subunit endows the Escherichia coli RNA polymerase holoenzyme with the ability to recognize a very large set of conjugate promoters, but does not bind to DNA on its own, primarily because its twin DNA-binding sites are occluded by an amino-terminal flap  that is displaced when 070 associates with the RNA polymerase core. Second, acidic activators induce a conformational change in TFIIB, exposing binding sites for other basal transcription factors by disrupting an internal interaction . A two-hybrid analysis suggests, along similar lines, that internal interactions within TFC120 might shield a B&binding site (N Chaussivert, C Conesa, S Shaaban, A Sentenac, personal co~unication). Conceivably, the TFIIIC-box A interaction might trigger a structure change in TFC120 that exposes its Brf-binding site.
U6 snRNA genes
The (human) genes coding for the small nuclear (sn)RNAs are tr~scriptiona~y diverse, some being transcribed by pol II and others by pol III, yet their promoters are structur~y very similar, with an enhancerlike distal element (DSE) situated -200 bp upstream (-ZOO), and an essential proximal element (PSE) situated -50 bp upstream (-50) of the transcriptional start. The promoters of a set of human U6 snRNA genes that is transcribed by pol III differ from pol II-transcribed
Ul and U2 snRNA genes only in having an additional TATA box located near -25. Removal of the TATA box converts the U6 gene to ~0111 transcription; appropriate placement of a TATA box in a Ul gene promoter converts it to polII1 transcription . The PSE-binding protein complex that participates in ~0111 and pol III transcription , SNAP, (the small nuclear RNA gene activation protein complex), has now been purified and shown to contain TBP together with at least three other proteins (-50, -45 and -43 kDa) and possibly a fourth (-200 kDa); the gene for the smallest of these proteins, SNAP43, has been cloned and sequenced [24**]. Antibodies to SNAP43 coprecipitate TBP Transcription of the U6 snRNA gene by polII1 requires SNAP, and, in addition, two components of, but not all of, TFIIIB : TBP and a separable sub&action of TFIIIB called 0.48 ~-TFIIIB. The latter is not yet characterized as to composition, but conceivably is f~ction~y comparable with the 90kDa component of yeast TFIIIB. The most interesting speculation that might be prompted by these new findings concerns TBI? For the past three years or so, it has been appreciated that TBP is common to all nuclear ~anscription, and that it associates with polymerase-specific accessory proteins, TAFs (TBPassociated factors), to form the pol I-, II- and III-specific assemblies SLl/TIF-IB, TFIID and TFIIIB, respectively. It appears that we can now add the snRNA-specific, pol II- and pol III-compatible SNAP, to this fleetingly canonical set, breaking the comforting symmetry of one place-one role for TBP in transcription. TBP is evidently involved in two capacities in transcription of the human U6 gene by pol III: as a part of SNAP,, bound to a TATA-less promoter element, and as the TATA box bound anchor for the transcription factor that resolves the bivalence of SNAP, against pol II 13,231. In view of its placement relatively far upstream of the transcriptional start, SNAP, probably communicates with pol II and pol III indirectly, through their respective core transcription hctors. When it assembles a pol II-specific initiation complex, SNAP, must also direct the co-assembly of an snRNA-specific termination or 3’ end-generating factor [31There is a catch to all this: the essentially concurrent report of purification of a PSE-binding transcription factor, PTF, that contains four proteins of similar molecular weight to, and probably identical with, the constituents of SNAP,, but that lacks TBP in the most purified transcriptionally active material 125.1. The discrepancy regarding TBP is important (particularly with respect to the preceding argument!), but is more a matter of interpretation than irreducible fact at this stage of purification of all the required components. That TBP interacts with the other constituents of SNAP, is highly probable. How tightly it does so remains to be seen. A determination of the absolutely required constituents of SNAPJPTF will require cloning and reconstitution.
of a role for chr~tin
initiation by RNA polymerases I and III Geiduschek and Kassavetis
transcription of the yeast U6 gene The yeast U6 gene (SNR6) is also transcribed by pol III, as already stated, but clearly does not follow SNAP, rules: it does not have a functional PSE, requires its 6oxA and ano~lously placed box B promoter elements for fimction in viva [26,27,28’], and yet can use its strong TATA box to assemble a properly placed and stable transcription initiation complex in vitro, using TFIIIB and pol III alone [20,27,28’,29*]. What, besides the already enumerated components of the core transcription apparatus, is involved in SNR6 gene transcription in vivo, and what is responsible for generating the grossly contrasting properties of its in vitro and in viva transcription? Three lines of analysis suggest a role for chromatin. First, box B, and by implication TFIIIC, is necessary for transcription in vitro in a crude extract that is rich in non-specific DNA-binding proteins. TFIIIC is also able to displace preassembled nucleosomes in such a way as to activate transcription of the SNR6 gene in chromatin . Second, probing the structure of the SNR6 gene in chromatin by nuclease digestion reveals an ordering or boundary effect of active SNR6 transcription on the positioning of upstream and downstream flanking nucleosomes [31*]. Chromatin footprints of the SNR6 gene itself, generated using DNase I and micrococcal nuclease, suggest a folded, though probably not nucleosomal, structure [28*]. Third, promoter-debilitating SNR6 mutations are suppressed in mutant cells that contain, on average, approximately 50% of the normal nucleosome complement, but the activity of the wild-type promoter is not affected by this degree of nucleosome depletion. Certain amino-terming deletions and mutations in histone H4, which strongly affect transcriptional activity and regulation of certain pol II genes, are without effect on SNR6 transcription [3 1’1. These results suggest a competition between chromatin assembly and transcription complex assembly on the wild-type SNR6 gene that strongly favors the transcription factors [31*]. How essential promoter elements that are identical to those of tRNA genes (the box A and box B elements) but poorly placed  generate extremely active transcription in vivo remains unresolved. If a nucleosome were involved in positioning TFIIIB in these transcription initiation complexes, as has been suggested , it would be subsequently dislodged by active transcription.
The core transcription
apparatus of RNA
polymerase I Pol I transcriptional initiation complexes assemble at promoters that extend from slightly downstream to -150 bp upstream of the transcription start site (reviewed in [32,33]). Two DNA-binding transcription
factors, UBF (the umber-binding factor) and SLl, also called TIF-IB (for transcription initiation factor for pol I), recruit pol I to these promoters. The potential for extremely high transcriptional activity at ribosomal RNA gene promoters  is determined by systems of multiple enhancer-like elements that are located in the intergenic spacer of ribosomal transcription units, and include binding sites for core transcription factors, as well as auxiliary poll promoters. These so-called spacer promoters are especially prominent in the organization of ribosomal RNA genes in X~~~s and ~os~~i~a [32,33]. We comment next on UBF and SLl/TIF-IB, and on the recruitment of ~011 to the inscriptions start site and initiation of transcription, aspects of transcriptional initiation by ~011 that lend themselves well to comparisons with ~01111. Two polymerase-binding factors that are also required for effective transcription have been resolved. The activity of one of these factors (TIF-IA, also designated as TFlC) is ~wth-replated, and the other factor (TIF-IC) resembles TFIIF in having a dual role in transcriptions i~tiation and elongation ([35,36]; reviewed in ). We regret the omission of this aspect of transcription by pol I from further discussion, principa~y for want of space.
Genes coding for human (h), rat, mouse and frog (x) UBF have been cloned and sequenced. They encode -67&770 amino acid proteins with highly conserved sequences. Each UBF contains an ~no-pixie dimerization domain, five HMG boxes (80-90 amino acid segments with significant sequence homology to high mobility group [HMG] proteins), and an acidic and serine-rich tail. The dimerization region of UBF and the amino-proximal HMG box confer effective and specific DNA-binding [37,38,39*] and truncated xUBF containing the three amino-proximal HMG boxes has essentially the same affinity for a linear DNA site spanning the transcriptional start of the Xenopw ribosomal promoter as does the full length protein [37,40*]. HMG proteins bind preferentially to crossed DNA helices 139.1, structures that are presented by nucleosomeexiting DNA in chromatin, by cruciform DNA and by plectone~ca~y supercoiled (braided) DNA. UBF also binds to cruciform DNA (probably through interactions with the minor groove), prefers supercoiled over linear DNA, and induces sharp bends in DNA [38,39*,41*]. Its extended footprints and the fact that UBF binding foreshortens linear DNA by nearly 200 (k 80) bp are consistent with DNA wrapping around a UBF dimer (; see also Grosschedl, this issue, pp 362-370). How the DNA site selectivity of UBF binding is generated remains unclear: UBF binds to DNA of diverse sequence and binding is competed by diverse nucleic acids [38,41*]. Quite probably, bendability and associated variations in the width of the DNA minor groove are important determinants of site selectivity. Given its ability to bind to enhancers as well as promoters  and to wrap DNA in
Nucleus and gene expression
so doing, UBF must be capable of conferring complex structures on a ribosomal gene transcription control region. We should note that UBF is subject to regulatory modifications and interactions (e.g. ).
The constitution of SLl/TlF-IB
Perhaps the most important recent advance in dissecting pol I transcription is the elucidation of the composition of human and mouse SLl/TIF-IB, which contains TBP and three associated proteins (TAFts) with molecular masses of 95 (-95), 64 (-68), and 53 (-48) kDa (the figures in brackets are the apparent molecular weights of the mouse proteins) [45,46,47*,48*]. The genes for the three human TAFts have been cloned and sequenced and active human SLl/TIF-IB has been reconstituted from the four recombinant proteins [48**,49**]. The three human TAFts are not recognizably homologous to TAFII or core ~0111 transcription factors (nothing resembling the TFIIB-yeast Brf [TAFttI] relationship). The two smaller mouse and human subunits in SLl/TIF-IB can be photo-crosslinked to DNA containing bromodeoxyuridine, signifying close approach to the DNA major groove, but the TBP in SLl/TIF-IB is not crosslinked to these TATA-less promoters. These experiments yield up important insights into the internal organization of SLl/TIF-IB and into the assembly of multiple, alternative polymerase-specific complexes by TBI? They show that there is no hierarchy in the assembly of human SLl /TIF-IB, as there is in the assembly of TFIID : every subunit of hSLl/TIF-IB interacts with all the others. The ability of each TAFt-TBP pair to exclude TBP-binding TAFI1, and vice versa, explains how the formation of non-functions, chimeric transcription factor complexes is prevented (Fig. 2). TIF-18
1995 Current Opinion in Cell Biology
Fig. 2. Polymerase-specificity of a TBP-containing transcription factor. Ail the other constituents of human TIF-IB/SLl (TAFs95, TAF164 and TAF153) interact with hTBP and with each other (a), and the complex of each with TBP prevents it from interacting with the two TBP-binding TAF,,s of the RNA polymerase II transcription factor hTFllD (TAF11250 and TAF,,lSO) (b).
A power&d yeast genetic system for identifying and analyzing the components of the pol I transcription apparatus has been devised, based on making &lures of rRNA synthesis conditionally, rather than absolutely, lethal . Mu~nts in twelve complementation groups that become selectively defective in rRNA synthesis under restrictive
conditions have been isolated thus far. The recent analysis of the RRNd and RRN7 genes is exceptionally interesting. These genes code for the 102 and 60 kDa subunits of an assembly that also includes a genetically unidentified -66 kDa protein. In vitro transcription experiments with crude fractions from mutant cells establish that the Rrn6-Rrn7 complex can associate stably with DNA, committing that template to transcription, in the presence of an additional unspecified component [52**]. This property appears to identify the Rrn6-Rrn7 complex as being part of the core ~011 transcription apparatus, clearly resembling SLl/TIF-IB rather than UBE Doubts that the Rrn6-Rrn7 complex is the yeast SLl/TIF-IB remain, but they are not decisive. First, the Rrn6-Rrn7 complex may interact with TBP but does not copurify with it, as SLl/TIF-IB does ; on the other hand, components of yeast TFIIIB also separate under mild conditions. Second, the Rrn6 and Rrn7 proteins are not detectably homologous to any hSL1 /TIF-IB protein; on the other hand, no homology has been detected between the box B-binding subunits of yeast and human TFIIIC either. Third, the Rrn6-Rrn7 complex does not stably bind to a ribosomal promoter when supplemented with TBP; that may not be fundamentally different I%om the properties of hSLl/TIF-IB (or yeast TFIIIB).
Concluding comments Is SLl/TIF-IB the TFMB homulogue in pal I transcription? Initiation of human rRNA synthesis in vitro requires UBF, and the same is evidently true in Acunthamoeba (cited in ). Nevertheless, multiple rounds of rRNA synthesis can be initiated with mouse and rat SLl/TIF-IB in the absence of UBF, although UBF generally stabilizes even these SLl/TIF-IB-DNA interactions and stimulates transcription. It is therefore plausible to regard rodent SLl/TIF-IB as the DNA-binding transcription factor that recruits ~011 to the promoter, that is, as the polI-specific counterpart of TFIIIB. Assigning UBF to be the poll counterpart of TFIIIC is less appropriate for the following reasons. Firstly, UBF probably interacts with the ribosomaI transcription unit in a much more complex way than does TFIIIC with its boxA and boxB binding sites. The massive changes in the accessibility of DNA to DNase I that occur when hSLl/TIF-IB joins hUBF at the human ribosomal promoter ( and earlier work cited therein) also suggest (to us, at any rate) that the outcome of this interaction is structurally complex. Secondly, hUBF interacts with hSLl/TIF-IB at multiple sites . It would be most informative to have, for the ~011 transcription initiation complex, the kind of low-resolution information about internal structure and topography that has been very helpful in unders~nding ~01111 transcription (21. (In fact, the materials for achieving very much higher resolution with the pol I system by mapping DNA-protein contacts on the polypeptide chain as well as the DNA strand are now at hand.) We hazard the guess
jnitiation by RNA polymerases I and Ill Geiduschek and Kassavetis
that, when such a structural map is available for pol I transcription fktor-promoter complexes, they will be found not to closely resemble the corresponding pol III complexes. Finally, direct interactions with pol I  suggest that mouse UFSF may not fknction exclusively as an assembly factor, as yeast TFIIIC is thought to do.
Initiation of nucleotide polymerization without ATP hydrolysis Once yeast ~01111 is bound at a tRNA gene promoter, it can open the latter in a temperature-dependent and reversible manner in the absence of nucleotides . In this instance, the resemblance between pol III and E. coli RNA polymerase is impressive. Acanthamoeba ~011 has been shown to polymerize non-hydrolyzable nucleotide analogs at a ribosomal promoter [SS]. In contrast, it has been thought for some time  that ATP hydrolysis is required for initiation of transcription by pol II. Finding that an ATPase activity is associated with TFIIH, and that certain promoters do not require TFIIH for basal DNA , activity [57,58] in negatively supercoiled is compatible with the notion that TFIIH-associated ATPase activity might be required for formation of an open promoter complex . It has been shown instead that abortive initiation by ~0111 (the process of repetitively polymerizing the initial nucleotides of an RNA chain) can be independent of ATP hydrolysis in linear as well as in negatively supercoiled DNA, suggesting a role for TFIIH and ATP hydrolysis in promoter clearance rather than promoter opening . The ability to detect this ATP hydrolysis requirement in the first instance  depends on its having been rate-limiting. No such rate-limiting step exists in RNA synthesis by pol III.
The Xenopus TFIIIA uses different sets of zinc fingers for its principal DNA- and RNA-binding sites. 6.
Kassavetis GA, Braun ER, Nguyen LH, Geiduschek EP: 5. cerevifiae TFIIIB is the transcription initiation factor proper
of RNA polymerase Ill, while TFIIIA and TFIIIC are assembly f&on. Celi 1990, 60:235-245. 7.
McBtyant SJ, Kassavetis GA, Gottesfeld JM: Repression of vertebrate RNA polymerase Ill transcription by DNA binding proteins located upstream from the transcription start site. 1 MO/ Siof 1995, in press,
Gottesfeld ]M, Wolf
VJ, Dang T, Forbes DJ, Hard P: Mitotic
repression of RNA polymerase Ill transcription in vitro mediated by phosphorylation of a TFIIIB component. Science 1994, 263:81-84. 9.
Conesa C, Swanson RN, Schultz
P, Oudet P, Sentenac A: On
the subunit composition, stoichiometry, and phosphorylation of the yeast transcription factor TFlllC/~u. f Biof Chem 1993, 268: 18047-l 8052. 10.
Marck C, LefGbvre 0, Carles C, Riva M, Chaussivert N, Ruet A, !%ntenac A: The TFIIIB-assembling subunit of yeast transcrip-
tion factor TFIIIC has both tetratricopeptide repeats and basic helix-loop-helix motifs. Proc Nat/ Acad Sci USA 1993, 90:4027-4031.
Shen Y, Aebersold R, Berk Al:
Human transcription factor IIIC boxB binding subunit. Proc Nat/ Acad Sci USA 1994, 91:1652-1656. 12.
Lagna C, Kovelman R, Sukegawa J, Roeder RG: Cloning and characterization of an evolutionarily divergent DNAbinding subunit of mammalian TFIIIC. MO/ Cell Viol 1994, 14:3053-3064.
The topo&hy of tra~~pt~n factor c&nplexes on the Saccharomyces cerevisiae 55 RNA gene. J Mel Viol 1992, 228:1063-1077.
Khoo B, Brophy B, Jackson SP: Conserved functional
of the RNA polymerase Ill general transcription factor BRF. Genes Llev 1994, 8:2879-2890.
Huet J, Conesa C, Manaud N, Chaussivert N, Sentenac A: lnteractions between yeast TFIIIB components. Nucleic Acids Res 1994, 22:2282-228%.
Kassavetis GA, Geiduschek EP:
Two essential components of the Sac&aromyces cerevisiae transcription factor 1118: transcription and DNA-binding properties. Proc Nat/ Acad Sci USA 1991,
References and recommended
Papers of particular interest, published within the annual period of review, have been highlighted of special inter&t of outstanding interest
. .. 1.
IM: RNA polymerase Ill. Genes, factors and transcrip tional specificity. Eur 1 Biochem 1993, 212:1-11.
KassavetisCA. Bardeleben C, Braun 6, Joazeiro CAP, Pisano M, Ceiduschdk EP: Transcription by RNA polymerase III. In Transcription Mechanisms and Regulation, vol 3. Edited by Conaway RC, Conaway JW. New York: Raven Press; 1994:107-l 26.
Lobe SM, Hernandez NT: Transcription of snRNA genes by RNA polymerases II and Ill. In Transcription Mechanisms and
White RJ: RNA polymerase 111transcription. Molecular Biology Intelligence Unit, Austin, TX: RC Landes Company; 1994.
Clemens KR, Wolf V, McBryant SJ, Zhang P, Liao X, Wright PE, Cottesfeld JM: Molecular basis for specific recognition of both RNA and DNA by a zinc finger protein. Science 1993, 260:53CL553.
M, Chaussivert N, Willis IM, Sentenac A: Interaction between a complex of RNA potymerase Ill subunits and the 70kDa component of TFIIIB. / 5iol Chem 1993,
268:20721-20724. A two-hybrid analysis of protein-protein interactions: the Brf subunit of yeast TFIIIB interacts with the most upstream-lying subunit of RNA polymeraselll. Ba~holo~w B, Durkovich D, Kassavetis GA, Geiduschek EP: Orientation and topography of RNA polymerase Ill in transcription complexes. MO/ Cell Biol 1993, 13:942-952. One of a series of papers that analyze the internal structure of RNA polymerase Ill transcription initiation complexes by photo-crosslinking.
Sentenac A, Riva M, Thuriaux P, Buhler J-M, Treich I, Carles C. Werner M. Ruet A, Huet I, Mann C er a/.: Yeast RNA rxrlvierase subunits and .genes.. In Transcriptional Regulation; vbl 1. Edited by McKnight SL, Yamamoto KR. Cold Spring Harbor: Cold Spring Harbor Laboratory Press; 1992:27-54.
Margottin F, Dujardin G, Gerard M, Egly J-M, Huet J, Sentenac A: Participation of the TATA factor in transcription of the yeast U6 gene by RNA polymerase C. Science 1991, 251:424-426.
Dombroski AJ, Walter WA, Record MT Jr, Siegele DA, Gross CA: Polypeptides containing hiily conserved regions of transcription initiation factor 070 exhibit specificity of binding to promoter DNA. Cell 1992, 70:501-512.
Regulation, vol 3. Edited by Conaway RC, Conaway JW. New York: Raven Press; 1994127-159.
Nucleus and gene expression 22.
Roberts SGE, Green MR: Activator-induced conformational change in general transcription factor TFIIB. Nature 1994, 371:717-720.
Sadowski CL, Henry RW, Lob0 SM, Wernandez N: Targeting TBP to a non-TATA box cis-regulatory element: a TBP-cabining complex activates ~~~ from snRNA promoters through the PSE. Genes Dev 1993, 1:1535-1548.
Henry RW, Sadowski CL, Kobayashi R, Hernandez N: A TBPTAF complex rquired for transcription of human MRNA genes by RNA polymerases II and Ill. Nature 1995, in press.
The transcription factor that binds to the proximal sequence element (PSE) site of snRNA gene promoters has been purified. It contains TATA-binding protein and three or possibly four other proteins. The gene encoding the smallest of these proteins has been cloned and sequenced. 25.
Yoon J-B, Murphy 5, Bai L, Wang Z, Roeder RG: Proximal sequence element-binding transcription hctor (PTF) is a mul&&unit complex required for trances of both RNA polymerase II- and RNA polymerase Ill-dependent small nuclear RNA genes. MO/ Cell Biol 1995, 15:201 g-2027.
Covers similar ground to 124”], but concludes that the protein complex does not contain TATA-binding protein. 26.
Eschenlauer 18, Kaiser MW, Cerlach VL, Brow DA: Architecture of a yeast U6 RNA gene promoter. MO/ Cell Biol 1993, 13:3015-3026. Burnol AF, Margottin F, Schultz P, Marsolier MC, Oudet P, Sentenac A: Basal promoter and enhancer element of yeast U6 snRNA gene. I Mel Biol 1993, 233644-658.
Gerlach VL, Whitehall SK, Ceiduschek EP, Brow DA: TFIIIB placement on a yeast U6 RNA gene in viva is directed primarily by TFIIIC rather than by sequence-specific DNA contacts. Mel Cell Biol 1995, 15:1455-l 466.
Dissects the contrasting influences for placement of yeast TFtllB at a yeast U6 snRNA gene promoter in vivo and in vitro. Analyzes the structure and protein occupancy of the U6 transcription unit in chromatin. 29.
Kassavetis GA, Ceiduschek EP: Identical components of yeast transcription factor IIIB are required and sufficient for transcriotion of TATA box-containini! and TATAless genes. MO/ Ce//‘Bio/ 1994, 14:279&2808. ”
Specifies that the same three components of yeast TFIIIB, including TATAbinding protein, are essential and sufficient for transcription of tRNA and yeast U6 snRNA genes. 30.
Burnol AF, Margottin F, Huet J, Almouzni G, Prioleau MN, Mechali M, Sentenac A: TFIIIC relieves repression of U6
snRNA t~n~~ption 31. l
by cbromatin. Nature 1993, 362~475-477.
Marsolier M-C, Tanaka 5, Livingstone-ZatchejM, Grunstein M, Thoma F, Sentenac A: Reciprocal int~er~c~ between nucleosomal organization and transcriptional activity of the yeast
SNR6 gene. Genes Dev 1995, 9:410-422. Examines chromatin structure surrounding a gene actively transcribed by RNA polymerase Ill. Specifies the consequences of histone octamer depletion and histone mutations on transcriptional activity of the yeast U6 snRNA gene. 32.
Reeder RI-I: Raulation of transcriotion bv RNA oolvmerase I. In Transcrip~onal regulation, v&l 3. E&d bi ticKnight SL, Yamamoto KR. Cold Spring Harbor: Laboratory Press; 19923 151347.
Genes Dev 1992, 6:195&1963. 38.
Cold Spring . ” Harbor
Paule MR: Transcription of ribosomal RNA bv eukarvotic RNA polymerase I. in T&scription mechanisms and re&lation, vol 3. Edited by Conaway RC, Conaway jW. Press; 1994:83-l 06.
New York: Raven
Miller OL, Beany BR: Visualization of nucleolar genes. Science 1969, 164~955-957.
Schnapp A, Schnapp G, Erny B, Grummt I: Function of the growth-regulated transcription initiation factor TIF-IA in initiation complex formation at tbe murine ribosomal gene promoter. Moi Cell Biol 1993, 13:6723-6732.
Copenhaver CP, Putnam CD, Demon ML, Pikaard CS: The RNA polymerase I transcription factor UBF is a sequence-tokrant HMC-box protein that can recognize structured nucleic acids. Nucleic Acids Res 1994, 22:2651-2657.
Hu Cti, McStay 8, Jeong S-W, Reeder RH: xUBF, an RNA polymerase I transcription factor, binds crossover DNA with low sequence spe&icity. MO/ Ceil Biof 1994 14:2871-2882. Presentsevidence that Xenopus UBF binds crossed DNA helices and generates loops in linear DNA. 39.
Leblanc B, Read C, Moss T: Recognition of the Xenqus rii somal core promoter by the transcription factor xUBF involves rn~ti~e HMG box domaii and leads to an xUBF stein interaction. EM80 I 1993, 12513-525.
Maps the binding of Xenopus UBF to a rRNA gene core promoter and suggests extensive folding of DNA within its UBF complex. 41.
Putnam CD, Copenhaver GP, Denton ML, Pikaard CS: The RNA
polymerase I transactivator upstream binding factor requires its dimerization domain and high-mobility-group (HMG) box 1 to bend, wrap, and positively supercoil enhancer DNA. MO/ Cell Bio/ 1994, 1416476-6488. DNA bending and wrapping by Xenopus (x)UBF are analyzed; assigns
Schnapp G, Schnapp A, Rosenbauer H, Grummt I: TIF-IC, a factor involved in both transcription initiation and elongation of RNA polymerase I. EMEO I 1994, 13:4028-1035.
M, Moss T: Short-range DNA looping by the Xenopus HMG-box transcription factor, xUBF. Science 1994, 264:113+&1137.
Pikaard CS: Ribosomal gme promoter domains can function as artificial enhancers of RNA polymerase I transcription, sup. porting a promoter origin for natural enhancers in X-5.
Cavanaugh AH, Hempel WM, Taylor LJ, Rogalsky V, Todorov G, Rothblum LI: Activity of RNA polymerase I transcription factor UBF blocked by Rb gene product. Nature 1995, 374:177-l 80.
Comai L, Tanese N, Tjian R: The TATA-binding protein and associated factors are integral components of tbe RNA polymerase I transcription factor, SLl. Cell 1992, 58:965-976.
Eberhard D, Tora L, Egly J-M, G~mmt I: A TBP-c~n~g multiprotein complex (TIF-IB) mediates transcription specificity of murine RNA polymerase I. Nucleic Acids Res 1993,
Bazett-Jones DP, Leblanc 6, Herfort
Proc Nat/ Acad Sci USA 1994, 91:464-468.
TBP-asso&ted factors interact with DNA aa g&m species specificity of RNA polymerase I transcription. FMBO 1 1994, 13:2611-2616.
Eberhard D. Tora L. StunnenberP
Reports the purification and constitution of mouse SLl/nF-18. 48. l.
Comai L, Zomerdijk JCBM, Beckmann H, Zhou 5, Admon A, Tjian R: Rec~tituti~ of transcription factor Stl: exclusive binding of TBP by 511 or TFIID subunits. Science 1994, 266:196&1972. Reports the cloning of the genes coding for the three TAF, proteins of human SLl (TIF-IB). Shows that each of these three proteins interacts with TATA-biting protein (TBP). Demonstrates that mhtually exciusive TBP interactions prevent the formation of mixed TAFl-TAFII complexes. 49.
lantzen H-M, Chow AM, Kina OS, Tiian R: Multiple domains of the RNA polymerase I a&vator ‘LUBF inten’a witb the TATA-binding protein complex hSL1 to mediite transcription.
the principal affini~-determinants
Zomerdijk jCBM, Beckmann H, Comai L, Tjian R: Assembly of transcriptionally active RNA polymerase I initiation factor 511 from recombinant subunits. Science 1994 266:201 S-201 8.
~onstrates that the four subunits of human SLl are necessary and sufficient for transcription factor activity. Reconstitutes active SLl (TIF-16) from recombinant constituents. 50.
ROJ, Dynlacht BD, Tjian R: largest subunit of Drosop/tila transcription faaor IID directs assembly of a complex c~taini~ TBP and a coactivator. Nahrre 1993, 362:511-517.
Nogi Y, Vu 1, Nomura M: An approach fur isolation of mutants defective in 355 ribosomal RNA synthesis in Saccharomyces cerevisiae. Proc Nat/ Acad Sci USA 1991, B&7026-7030.
Keys DA, Vu L, Steffan JS, Dodd JA, Yamamoto RT, Nogi Y, Nomura M: RRN6 and RRN7 encode subunits of a multiprotein
complex essential for the initiation of rDNA tmnscription by RNA poiymerase I in Saccharomyces cerevisiae. Genes L&v 1994, 8:2349-2362. Applies a powerful genetic system, described in [Sll, to identify two essential genes coding for constituents of a protein complex that commits a rRNA gene to transcription. The complex shares certain properties with the veitebrate tran~ription factor TiF-IB/SLl. 53.
Schnapp C, Santori F, Caries C, Riva M, Grummt I: The HMG
82:439H398. Parvin JD, Sharp PA: DNA topology and a minimal set of basal factors for transcription by RNA polymerase II. Cell 1993,
Tyree CM, George CP, Lira-DeVito LM, Wampler SL, Dahmus ME, Zawel L, Kadonaga IT: identification of a minimal set of proteins that is sufficient for accurate initiation of transcription by RNA polymerase II. Genes Dev 1993, 7:1254-l 265.
Goodrich JA, Tjian R: Transcription factors IIE and IIH and ATP hydrolysis direct promoter clearance by RNA polymerase II. Cell 1994, 77:145-l 56.
13:190-199. Kassavetis CA, Blanc0 JA, Johnson TE, Ceiduschek
tion of open and elongating transcription complexes by RNA polymerase III. j Mof Biol 1992, 226:47-58. 55.
Lofquist AK, Li l-t, lmboden MA, Paule MR: Promoter opening (melting) and transcription initiation by RNA polymerase I requires neither nucleotide @,y hydrolysis nor protein phosphoryhtion. Nucleic Acids Res 1993, 21:3233-3238.
Sawadogo M, Roeder RG: Factors involved in specific transcription by human RNA polymerase II: analysis by a rapid and quantitative in vitro awy. froc Nat/ Acad Sci USA 1985,
box-containing nucleolar transcription factor UBF interacts with a specific subunit of RNA polymerase I. EM50 1 1994,
polymerases I and Ill Geiduschek and Kassavetis
EP Geiduschek and GA Kassavetis, Department of Biology and Center for Molecular Genetics, University of California at San Diego, La Jolla, CA 92093-0634, USA.