Figure 1. Conserved regions of s.
Brian A. Young 1, Tanja M. Gruber 2, and Carol A. Gross 2, @
1 Department of Biochemistry and Biophysics, University
of California, San Francisco, San Francisco, CA 94143 USA
2 Departments of Stomatology and Microbiology and Immunology,
University of California, San Francisco, San Francisco, CA 94143 USA
@ Correspondence: Carol A. Gross cgross@cgl.ucsf.edu
Initiation of transcription is the first step in gene expression and a major point of regulation. Recent structural studies reveal the nature of the initiating complex and suggest new ways of accomplishing the processes required for initiation.
RNA polymerase (RNAP), the enzyme that carries out transcription, is a remarkable molecular machine. During initiation, it must recognize promoter DNA from the vast excess of non-promoter DNA, separate the duplex to expose the template strand, and initiate RNA synthesis using only mononucleotides. Before beginning processive elongation, it must transition to a non-sequence-specific DNA binding protein that moves forward (and in some cases backward) along the DNA. Four recent reports illuminate these processes. Three use crystallography to provide structural information about the prokaryotic initiation factor s (Campbell et al., 2002), and the initiating form of prokaryotic RNAP without (Murakami et al., 2002b) and with promoter DNA (Murakami et al., 2002a). A fourth provides a distance constrained model of initiating RNAP and its interaction with promoter DNA based upon systematic measurements of fluorescence resonance energy transfer (FRET) of probes located throughout the initiating RNAP and in the DNA (Mekler et al., 2002). We discuss these results and the insights and speculations they provoke about how this machine accomplishes these complex processes.
Transcription initiation in prokaryotes is carried out by holoenzyme (holo), comprising core RNAP (core) plus the initiation specific subunit, s. Core is an ~400 kDa complex of five subunits (a2bb'w), which shares considerable sequence and even more structural homology with its eukaryotic counterparts (e.g., RNAP II), whereas s has little sequence homology to its eukaryotic counterparts, the general transcription factors. Holo first recognizes the two conserved hexamers in the promoter, located at -10 and -35 relative to the transcription startpoint of +1, then melts the DNA from -11 to +4 to form the open complex, and then begins synthesizing the nascent RNA. The three sections below summarize the structures of s, holo, and the open complex.
sSubunit.
All bacteria have one primary s factor, which
directs the majority of transcription. ss have
four conserved regions, which mediate binding to core and to DNA (Figure
1).
Figure 1. Conserved regions of s.
sA has three stably folded domains, s2, s3 and s4, connected by flexible linkers. Each domain is predicted to bind both core and DNA (Figure 1). s2 is essentially identical to E. coli s2, with an exposed region 2.2 helix predicted to form a primary interface with core and the region 2.3–2.4 helix, which recognizes the -10 element and contains aromatic residues important for melting and recognition of the non-template strand of the -10 element. Both s3 and s4 are comprised of three helices. One helix in s3 is responsible for recognizing two conserved bases located upstream of the -10 region, present in ''extended -10 promoters,'' which do not need a -35 promoter element. Two helices in s4 form an HTH motif; one of these helices recognizes the -35 region of the promoter. Campbell et al. were also able to obtain the structure of s4 complexed with a -35 element, allowing the first high-resolution view of promoter recognition. This pivotal work defined the domain structure of s, provided assurance that the genetic inferences about how s recognized the -35 element were generally correct, and produced high-resolution structures that allowed definitive placement of s on the holo structure.
Promoter Recognition and Spacer Accommodation. The work presented here goes some way toward explaining how binding to core relieves region 1.1 autoinhibition of DNA binding in free s. In holo, region 1.1 is removed from its location in s and placed in the active site channel of polymerase, possibly because its high negative charge allows it to act as a downstream DNA mimic. This same idea could explain autoinhibition: region 1.1 might bind DNA recognition determinants in free s because of its negative charge, thereby out-competing promoter DNA.
The structures also suggest a mechanism by which RNAP binds promoters with as few as 16 or as many as 18 nucleotides in the spacer region between the -35 and -10 elements, which can change the distance between these elements by as much as 10 Å. s4 (-35 recognition), is perched on the end of the flexible flap of b. Shifting the angle at which this flexible flap juts out of the core enzyme alters the distance between s4, and s2 (-10 recognition) somewhat. Larger lengths of DNA can be accommodated by stretching (or ''kinking'') them over a bulge in b' that intervenes between the domains in s that recognize the -10 and -35 hexamers. The enhanced DNaseI hypersensitivity in the spacer region of promoters with longer spacers is consistent with this explanation, as such ''kinked'' DNA would be expected to be more susceptible to DNase I cleavage (Murakami et al., 2002a).
Region 1.1 and Open Complex Formation. Region 1.1 also plays a positive role in transcription initiation: holo having s70 lacking region 1.1 forms open complexes very slowly at several promoters (Gross et al., 1998 and references therein). The jaws are closed in the holo structure (which was obtained with sA lacking region 1.1) but must be open in wild-type holo to permit downstream DNA to enter. Thus, Murakami et al. (2002b) propose that opening the jaws is a rate-limiting step at some promoters, and that region 1.1 accelerates this opening by binding between them, thereby accelerating open complex formation. At one weak promoter, holo lacking region 1.1 forms melted complexes more readily than intact holo (Vuthoori et al., 2001); perhaps, as suggested by Mekler et al. (2002), this is because the rate limiting step at this promoter is ejecting region 1.1 from the jaws.
A great deal about the relationship between s and jaw opening remains to be worked out. Region 1.1 is conserved only in primary or housekeeping s factors; how are the jaws of the polymerase opened during initiation with alternate s factors, which lack region 1.1? Additionally, the notion that s opens the jaws of polymerase currently lacks experimental backing. The most recent analysis of E. coli core RNAP indicates that its jaws are wide open (Darst et al., 2002). Although this could be artifactual, it is worthwhile recalling that s (and TFIIF in eukaryotes) decreases binding to non-promoter DNA. Do they do this by partially closing the jaws? Even after solution measurements of the placement of the jaws in core and holo resolve this particular question, additional work is clearly needed to understand how downstream DNA is efficiently placed in the jaws during open complex formation.
De Novo RNA Synthesis. In contrast to DNA polymerases, which can only extend existing nucleic acid chains, RNAP is able to create a nucleic acid polymer de novo—using only mononucleotides. The initiation step is difficult: both incoming nucleotide and the initiating nucleotide, which attacks the incoming NTP, must be stabilized in the correct geometry. This is especially difficult because RNAP prefers to begin RNA chains with an ATP (which base pairs with the template strand more weakly than would CTP or GTP).
How does RNAP accomplish de novo synthesis? Specific interactions with the initiating nucleotide must hold it rigidly in place, facilitating chemical attack on the incoming nucleotide. Such specific interactions would explain why polymerase prefers to start transcripts with ATP (followed by GTP, UTP, and then CTP). Indeed, a subcomplex of core polymerase, a2b, and possibly even the isolated b subunit has a site for the initiating nucleotide (Naryshkina et al., 2001). Darst now suggests that a disordered loop of s near the beginning of the s3–s4 linker, pointing toward the active site, assists in binding the initiating nucleotide. They tested this idea, using an extended -10 promoter, which does not use -35 recognition determinants and can therefore be transcribed by holo ending at s3, which lacks the disordered loop. Holo with this truncated s requires a much higher concentration of initiating dinucleotide to reach maximal levels of transcription than does holo with full-length s (Campbell et al., 2002). Using s to provide specificity is appealing. A marked preference for a particular nucleotide in the attacking site might have disagreeable side effects when elongating a transcript. If s performed this function, the selectivity required for de novo synthesis would be present only at initiation.
We suggest another possible role for this disordered loop of s-stabilization of the melted state of the promoter by binding to the template strand near the start site in single stranded form (thus keeping it from reannealing). This idea comes from a consideration of the ribosomal RNA (rRNA) promoters. rRNA promoters cannot form stable open complexes; they require high levels of initiating nucleotide to stabilize the melted state required for efficient transcription. This requirement is an important regulation mechanism in the cell, and is mediated in part by the presence of a stretch of CG bases, called the discriminator region, near the start site of rRNA promoters. The unique nucleotides in the discriminator may prevent the stabilizing interactions between the disordered loop in s and the template strand, thereby preventing stable open complex formation. In this context, an alternate single-strand-specific interaction may be necessary to achieve a stable open complex. The bridge created between the template strand and RNAP by the initiating nucleotide may provide this interaction—thus the requirement for high initiating nucleotides. Likewise, removal of the disordered loop in s should prevent stable open complex formation and create a requirement for high initiating nucleotides.
Promoter Clearance and Abortive Initiation. All RNAPs reiteratively synthesize and release short RNA transcripts called abortives, (~2 to 9 nucleotides in length). Based on the structure of elongating eukaroytic RNAP II, Kornberg proposed that shorter nascent RNAs dissociate because they make fewer contacts with polymerase than do longer RNAs (Gnatt et al., 2001). Another explanation for these transcripts has emerged from the Murakami et al. structure (2002b). Region 3.2 occupies the RNA exit channel, leading to the speculation that nascent RNAs must successfully compete with region 3.2 to be retained in elongating polymerase. When RNA transcripts lose the competition, they are ejected as abortive transcripts; when they win, region 3.2 is ejected and the transcript is successfully elongated. Consistent with this idea, holo having s lacking region 3.2 produced fewer abortives relative to full-length transcript than does holo with wt s (Murakami et al., 2002b). If this is a universal explanation for abortive transcription, a general transcription factor must play this role in eukaryotic initiation.
Why might polymerase place region 3.2 of s right where the RNA must go, thus wasting valuable NTP energy synthesizing abortive RNAs? Maybe the competition between s and RNA is an important part of the promoter clearance process. During promoter clearance, the polymerase must extricate itself from promoter-specific contacts so it can processively elongate trancripts. Interestingly, promoter clearance tends to coincide with the end of abortive synthesis and could be set in motion by release of region 3.2 from core by the successfully elongating RNA chain. Release of region 3.2 could cause promoter clearance because it weakens the s/core interface, thus allowing core to dissociate from s. Alternatively, either movement of region 3.2 out of the channel or of RNA into the channel could alter the position of s4 (which is perched on the flap surrounding the channel), making correct interactions with the -35 element impossible and promoting promoter dissociation. In either case, the NTP energy utilized during abortive initiation may be a small price to pay in order to switch from specific interaction with the promoter to processive elongation.
These first glimpses of the initiation competent polymerase provide extraordinary insight into the functions the machine performs. Although these structures do not provide final answers for how these processes are accomplished, they do allow us to conceptualize concrete models. No doubt these first peeks into the structure of initiating RNAP will motivate an enormous number of future experiments to test these ideas.
Campbell E.A., Muzzin O., Chlenov M., Sun J.L., Olson C.A., Weinman O., Trester-Zedlitz M.L. and Darst S.A. (2002) Mol. Cell, 9:527-539.
Darst S.A., Opalka N., Chacon P., Polyakov A., Richter C., Zhang G. and Wriggers W. (2002) Proc. Natl. Acad. Sci. USA, 99:4296-4301.
Gnatt A.L., Cramer P., Fu J., Bushnell D.A. and Kornberg R.D. (2001) Science, 292:1876-1882.
Gross C.A., Chan C., Dombroski A., Gruber T., Sharp M., Tupy J. and Young B. (1998) Cold Spring Harb. Symp. Quant. Biol., 63:141-155.
Korzheva N., Mustaev A., Kozlov M., Malhotra A., Nikiforov V., Goldfarb A. and Darst S.A. (2000) Science, 289:619-625.
Malhotra A., Severinova E. and Darst S.A. (1996) Cell, 87:127-136.
Mekler V., Kortkhonjia E., Mukhopadhyay J., Knight J., Revyakin A., Kapanidis A.N., Niu W., Ebright Y.W., Levy R. and Ebright R.H. (2002) Cell, 108:599-614.
Murakami, K.S., Masuda, S., Campbell, E.A., Muzzin, O., and Darst, S.A. (2002a). Science, in press.
Murakami, K.S., Masuda, S., and Darst, S.A. (2002b). Science, in press.
Naryshkin N., Revyakin A., Kim Y., Mekler V. and Ebright R.H. (2000) Cell, 101:601-611.
Naryshkina T., Mustaev A., Darst S.A. and Severinov K. (2001) J. Biol. Chem., 276:13308-13313.
Vuthoori S., Bowers C.W., McCracken A., Dombroski A.J. and Hinton D.M. (2001) J. Mol. Biol., 309:561-572.
Zhang G., Campbell E.A., Minakhin L., Richter C., Severinov K. and Darst S.A. (1999) Cell, 98:811-824.
1. Grundy FJ, Winkler WC, and Henkin TM, "tRNA-Mediated Transcription Antitermination in vitro: Codon-Anticodon Pairing Independent of the Ribosome", Proc. Natl. Acad. Sci. USA, vol. 99, no. 17, pp. 11121-11126 (August 20, 2002).
2. Frenster JH, and Hovsepian JA, "RNA Feedback Mechanisms during Eukaryotic Gene Regulation".
3. Blumenthal T, Evans D, Link CD, Guffanti A, Lawson D, Thierry-Mieg
J, Thierry-Mieg D, Chiu WL,
Duke K, Kiraly M, and Kim SK, "A Global
Analysis of Caenorhabditis elegans Operons".
4. Frenster JH, "Nuclear Polyanions as De-repressors of Synthesis of Ribonucleic Acid".
5. Frenster JH, "A Model of Specific De-repression within Interphase Chromatin".
6. Britten RJ, and Davidson EH, "Gene Regulation for Higher Cells: A Theory".
7. Herstein PR, and Frenster JH, "Mated Models of Gene Regulation in Eukaryotes".
8. Frenster JH, "Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA".