• Embed Doc
  • Copy Link
  • Readcast
  • Collections
  • CommentGo Back
 
TIBS 21 -
SEPTEMBER 1996
REVIEWSThe role of general initiationfactors in transcription byRNA polymerase II
Robert G. Roeder
Transcription initiation on protein-encoding genes represents a major con-trol point for gene expression in eukaryotes, and is mediated by RNApolymerase II and a surprisingly complex array of general initiation factors(TFliA, -B, -D, E, -F and -H) that are highly conserved from yeast to man.Elucidation of structural and functional features of these factors on modelpromoters has revealed insights into biochemical mechanisms and pro-vides a basis for understanding their regulation on diverse promoters bygene- and cell-specific activators.
GENES TRANSCRIBED BY RNA polym-erase II (Pol II) typically contain (1) com-mon core-promoter elements that arerecognized by general transcription ini-tiation factors, and (2) gene-specific DNAelements that are recognized by regu-latory factors, which in turn modulatethe function of the general initiationfactors. While natural constraints (suchas limiting-factor concentrations, weakpromoter-binding sites, chromatin struc-ture, negative co-factors, etc.) restrictthe function of general initiation factors
in vivo,
RNA Pol If and cognate generalinitiation factors alone have an intrinsicability to effect low levels of accuratetranscription from core promoters
invitro
(defined as 'basal transcription') 1.This property has made it possible tostudy the fundamental aspects of tran-scription initiation mechanisms thatare prerequisite for understanding thesuperimposed regulatory mechanisms.Analogous to the situation in prokary-otes, eukaryotic transcription can be de-scribed in terms of pre-initiation complex(PIC) assembly, PIC activation (DNA melt-ing), initiation, promoter clearance, elon-gation and termination steps. However,a major distinction is the remarkablygreater complexity of the general tran-scription machinery in eukaryotes, anobvious consequence of which is a morediverse set of mechanistic steps andtargets for regulatory factor interactions.
R. G. Roeder s at the Laboratory ofBiochemistry and Molecular Biology,The Rockefeller University, New York,NY 10021, USA.
9 1996, ElsevierScience Ltd
Core-promoter elements and generaltranscription initiation factors
Core-promoter elements are defined as'minimal DNA elements that are necessaryand sufficient for accurate transcriptioninitiation by RNA Pol II in reconstitutedcell-free systems'. The most common ofthese elements, which can function in-dependently or synergistically, are theTATA box (consensus TATAa/tAa/t), lo-cated near position -30 to -25, and apyrimidine-rich initiator (Inr, consensusYYANt/aYY) located near the transcrip-tion start site. Despite the complexity(12 evolutionarily conserved subunits)of RNA Pol II, it nonetheless requiresadditional factors for accurate tran-scription from even the strongest corepromoters. This requirement was firstdemonstrated with purified RNA Pol IIand a crude fraction from human cellsin 1979 (Ref. 2), and subsequent frac-tionation studies led to the first demon-stration of multiple RNA Pol II-specificfactors in 1980 (Ref. 3). At present,there are six well-characterized generalinitiation factors that have now been iso-lated from human, rat,
Drosophila
andyeast4-7; and cDNAs encoding almost allof the component polypeptides havebeen isolated. The designations, poly-peptide compositions and general func-tions of these factors, along with thoseof RNA Pol II, are summarized in Table I.Although these factors are surprisinglycomplex - together they are formallyequivalent to a single (r factor in prokary-otes - it is gratifying to note that thereis a remarkable conservation (based oncorresponding polypeptide sequences)PIl: s0968-0004(96) 10050-5of these components from yeast to man.In relation to the fact that prokaryotic(~ factors are directly responsible forpromoter selection by prokaryotic RNApolymerases, it is worth noting thatTFIID is the only general initiation factorwith components capable of sequence-specific binding to eukaryotic promoterDNA. These components include TBP,which directly recognizes the TATA el-ement 7, and certain TBP-associatedfactors (TAFs), which have been impli-cated directly7 or indirectly in sequence-specific binding.In the context of this review the gen-eral transcription initiation factors aredefined as 'those factors that are necess-ary, in naturally occuring forms, for accu-rate transcription initiation from a broadgroup of core promoters' - and not justthe potent TATA- and INR-containingAdenovirus Major Late (AdML) promoterused for their initial characterization.Importantly, as described below, thespecific factor or factor subunit require-ments might vary somewhat, dependingupon specific core-promoter sequences,template topology, the integrity of thegeneral initiation factors (notablyTFIID) and the presence of TBP-bindingnegative co-factors. The broader groupof general transcription factors (GTFs)includes those involved in initiation,elongation (reviewed in D. Reines
et al.,
this issue) and termination.
TATA-directed nitiation with minimal basalfactors
The clearest picture of the funda-mentals of PIC assembly, structure andfunction has come from studies of aminimal set of factors, or componentsthereof, that suffice for transcriptionfrom the consensus TATAelement of theAdML promoter. These so-called 'basal'factors include, under normal conditionswith relaxed DNA templates, the TBP(TATA-binding polypeptide) subunit ofTFIID, TFIIB, TFIIE, TFIIE TFIIH and RNAPol II. The ability of the small TBP sub-unit to substitute for the intact TFIIDcomplex in this reaction, with the con-comitant elimination of any TFIIA re-quirement 4,5, has greatly simplified theanalysis of core-promoter (basal) tran-scription mechanisms. This was also ofhistorical importance, in that it allowedthe initial identification and cloning of thegene encoding TBP from yeast (reviewedin Ref. 4), which in turn facilitated thedirect cloning of TBP species from higherorganisms and the development of cor-responding affinity reagents for the iso-lation and characterization of the other
327
 
REVIEWS
IIII
TIBS 21 -
SEPTEMBER 1996
-30
TATA
1 , I
. ; TAFs:~ ::
:!+1 +30
9 9
INRI I DNA
~
TBP or TFIID + TFIIAt TFIIB
~
TFIIF-RNAPol IIA + ............~ TFIIE~ TFIIH-- DNA
LsllJ
PICclose~
ATP ~ DNA melting
[ PIG open ]
NTPs ~ Initiation
~ CTD phosphorylation
\\
:"" ::'TFIIB ~ Pr~176learance"" "TFIIH'TFIIE Elongation........
~'
TAF~ ~i
:J
/,R~l^Residual promoter complexTFIIF-RNA Pol IIADephosphorylation
~
RNA
TerminationElongation complex
Figure 1
Model for pre-initiation complex(PIC) assembly and function on aTATA-containing core promoter, inthe simple situation [with TATA-binding protein (TBP)] a minimalset of basal factors (TFIIB, -E, -F,-H and RNA Pol II) suffice, inpurified form, for the stepwiseassembly
in vitro
of a functionalPIC through a consensus TATA el-ement, and stable intermediatescorresponding to each step ofassembly are demonstrable. Shortsolid bars indicate documentedprotein-protein interactions be-tween the minimal basal factors.Although not essential for for-mation of a functional PIC withminimal factors, TFIIA can enterthe PIC at any point after TBPto form corresponding intermedi-ates; and in situations with weakTBP-TATA box interactions, TFIIAmight be essential for stablecomplexes. In the more physio-logical situation, with TFIID inplace of the derived TBP, ad-ditional components [TFIIA andTBP-associated factors (TAFs)]modulate PIC assembly by stabi-lizing interactions (short openbars) with TBP and, most likely,with each other and with otherinitiation factors. DownstreamTAF-DNA interactions (see Fig. 3)might also contribute to complexstability. In this situation, and ondifferent core promoters, theassembly pathway might vary withrespect to the number of stableintermediates and requirementsfor other essential factors. In thenatural situation, a number of gen-eral factors (in addition to TFIIF)might enter the PIC in
associ-ation
with RNA Pol II in the formof a holoenzyme complex that
also
contains various coactivators.Following formation, the stable(closed) form of the PIC is acti-vated in an ATP-dependent man-ner to form an unstable (open)complex. The presence of NTPsthen results in initiation, promoterclearance, elongation and recy-cling of PIC components, as indi-cated and as detailed in the text. DNA melting and phosphorylation of the carboxy-terminal domain (CTD) of RNA Pol II are dependent uponTFIIH activities, although there is still some uncertainty about the timing and exact function (e.g. in promoter clearance) of the CTD phosphoryl-ation events. Phosphorylation of the CTD is represented by the presence of green dots.
TFIID subunits (TAFs)7. As implied above,the TAFs, like TFIIA, are not essential for
basal
transcription with purified
factors
from conventional TATA-containingpro-moters, but do have either coactivatorfunctions or basal factor functions onother promoters that are discussed later.TFIIA is also discussed here, in conjunc-tion with minimal basal factors, becauseof its well-documented interactions with
328
TBP and its effect on basal transcrip-tion under specific conditions.
Preinitiation complex assembly and structure.
Consistent with earlier studies 9 of PICassembly with TFIID and other factors,later studies with purified minimalcomponents (many recombinant) haveestablished a stepwise
in vitro
assemblypathway on the AdML promoter, andidentified corresponding intermediatesthat remain stable to various nuclease-protection, electrophoretic mobilityshift and template-challenge assays4,s,1~Further information regarding the under-lying intermolecular interactions, and theprotein domains involved (not detailedhere), has been provided by binaryinteraction assays coupled with
in vitro
mutagenesis techniques. More recently,and beginning with the structure of TBP
 
REVIEWS
IBS 21 -
SEPTEMBER 1996
in 1992, these studies have been com-plemented with high resolution X-raystructural studies of early intermedi-ates in PIC assembly with minimal fac-tors (reviewed in Refs 7, 11). The path-way Figs 1, 2; Table I) includes thefollowing steps:(1) Binding of TBP to the TATA el-ement, through minor-groove contacts,forms a stable complex with an un-precedented DNA distortion and a re-sulting bend that brings sequences up-stream and downstream of the TATAelement into closer apposition 7,u. Whilenot important for PIC assembly in theminimal system, TFIIA nonetheless canbind stably to this complex~,1~ hroughdirect contacts both with TBP and withupstream DNA sequences 12A3 (Fig. 2).When weakened by altered ionic condi-tions 14 or mutations in the DNA-bindingsurface of TBP ~ef. t5), TBP-DNA inter-actions can be greatly stabilized byTFIIA - of potential importance for basaltranscription on other promoters withweak TATA elements.(2) Binding of TFIIB, through directinteractions with TBP and with DNAsequences both upstream and down-stream of the TATA element 16, results ina complex in which the amino-terminaldirect-repeat domain of core TFIIB isoriented towards the (downstream) in-itiation site (Fig. 2). This is consistentwith an essential role for more amino-terminal sequences (not present in thecrystal structure) in TFIIF-Pd~/APol II re-cruitment, but not for TBP-TFIIB com-plex formation (see below), and with theestablished function of TFIIB in start-siteselection by RNA Pol II (reviewed inRefs 6, 17). A possible physical basisfor the latter is suggested by a two-dimensional crystallographic analysis ofa TFHB-RNAPol II complex, which indi-cates that the minimal distance betweenTFIIB and the presumptive active site ofthe enzyme corresponds roughly to thatbetween the TBP-TATA complex andthe initiation site on the DNA (Ref. 17).Like TFIIA, TFIIB can stabilize TBP-TATA interactions that are weakened bysuboptimal binding conditions ~4,18. THIBcan bind simultaneously with TFIIA to aTBP-TATA complex5,~~ As revealed bythe combined X-ray structures of theDNA-TBP-TFIIB and DNA-TBP-TFIIAcomplexes (]Zig. 2), TFIIB and TFIIA showno overlapping contacts with TBP or withDNA (the upstream DNA contacts resid-ing on opposite faces of the DNA helix)and no direct contacts with each other.This is consistent with distinct, but inter-dependent, functions of TFIIA and TFIIB.
Factor
Table I. General transcription initiation factors from human cellsa,b
Number ofsubunits Mw (kDa) Function
/ TBP 1 38TFIID
%
-- TAFs 12 15-250TFflA 3 12, 19, 35TFIIB 1 35TFIIF 2 30, 74RNA Pol II 12 10-220TFIIE 2 34, 57TFIIH 9 35-89Core promoter recognition (TATA);TFIIB recruitmentCore promoter recognition (non-TATA elements);positive and negative regulatory unctionsStabilization of TBP binding; stabilization ofTAF-DNA interactions; anti-repression functionsRNA Pol II-TFIIF recruitment; start-site selectionby RNA Pol IIPromoter targeting of Pol II; destabilization ofnon-specific RNA Pol II-DNA interactionsCatalytic functions in RNA synthesis;recruitment of TFIIETFIIH recruitment; modulation of TFIIH helicase,ATPase and kinase activities; directenhancement of promoter melting (?)Promoter melting using helicase activity;promoter clearance (?) by CTD kinase activityaThe subunit compositions and polypeptide sizes are those described for the human factors, buthomologues for virtually all have also been identified in rat,
Drosophila
and yeast. References to papersdescribing the isolation and characterization of these factors, and corresponding cDNAs, can be found inresearch papers (especially Ref. 17 for yeast) and reviews (Refs 4, 5, 7 and J. Q. Svejstrup
et al.
in thisissue) cited in the text. tn human and
Drosophila,
the two largest TFIIA subunits are derived from acommon precursor, which corresponds to the single large subunit of yeast TFIIA (reviewed in Ref. 12).bAbbreviations used: CTD, carboxy-terminal domain; RNA Pol II, RNA polymerase II; TAFs, TATA-bindingprotein-associated factors; TBP, TATA-binding protein.
Thus, TFIIB is involved directly in P,NAPol II-TFIIF recruitment (below), whereasTFI~ has both an anti-repression func-tion, involving dissociation of TBP-boundnegative co-factors that prevent TFIIBbinding to the PIC (reviewed byK. Kaiser and M. Meisterernst, this issue),and a separable activation function inTFIID-dependent transcription 19,2~(3) Binding of a pre-formed TFIIF-Pd~IA Pol lI complex, through directinteractions of TFIIB with both TFIIFand RNA Pol II, follows the binding ofTFIIB (Refs 6, 17). Although TFIIF clearlyplays a direct role in promoter targetingof PdNAPol II through these interactions,it also plays an indirect role by reduc-ing PdNA Pol II binding to non-specificsites in DNA ~efs 4, 5). RNA Pol II(form IIA) that is not phosphorylatedon the carboxT-terminal domain (CTD)of its largest subunit is recruited to thePIC in preference to the phosphorylatedform (1IO)5. Although the small subunit~P30) of TFIIF appears sufficient forthis recruitment 5, genetic and biochemi-cal studies suggest functions for boththe small (RAP30) and large 0lAP74)subunits in stabilization of the PIC inter-mediate, through TFIIB interactions,and in transcription initiation 2~,~4.~.Interestingly, THIF also serves as a tran-scription elongation factor, although thelarge and small subunit domains impli-cated in elongation are largely differentfrom those implicated in initiation (re-viewed by D. Reines
et aL
in this issue).Various nuclease protection studieshave indicated TFIIF and/or RNA Pol IIcontacts with DNA sequences (in thePIC intermediate) that lie between thedownstream THIB contact sites (pos-itions -23 to-14) and position +17 (Refs4, 5, 10). Consistent with these results,photo-crosslinking studies have localizedRAP30 (and the two large RNA Pol IIsubunits) to a position near a down-stream TFIIB-DNA contact site (position-19) and RAP74 to an adjacent 3' region(near positions -15 to -5) on the sameface of the DNA helix 22.(4) Binding of TFIIE, through directinteractions with RNA Pol II (Refs 5, 17,23), and potentially THIF and TBP 0tefs5, 23), is the next step in the pathway.Within the PIC, the small subunit ofTFIIE has been localized, by photo-crosslinking, to a region just upstreamof the start site~z, consistent with theproposed functions (below) of TFIIE "inpromoter melting.(5) Binding of TFIIH completes as-sembly of the PIC ~efs 4-6) and is me-diated by direct contacts with TFIIE~efs 23, 24). Consistent with this find-ing, TFIIH has been reported to stabilize
329
of 00

Leave a Comment

You must be to leave a comment.
Submit
Characters: ...
You must be to leave a comment.
Submit
Characters: ...