Faculty and Research Interests
Sergei Borukhov, PhD
2 Medical Center Drive
Moscow Institute of Genetics and Selection of Microorganisms, PhD, 1987.
Molecular mechanisms of transcription and its regulation. Structure and function of bacterial transcription factors GreA, GreB, Gfh1 and DksA modulating the catalytic activities of RNA polymerase
DNA-dependent RNA polymerase (RNAP) is the key enzyme of transcription process, and plays a central role in gene expression in all cellular organisms. The transcription cycle of RNAP includes initiation, elongation, and termination, and in each of these steps, RNAP is targeted by various regulatory factors. Knowing how these factors modify and regulate basic biochemical reactions of RNAP is essential to our understanding of gene regulation. Prokaryotic transcript cleavage factors GreA and GreB (1-4), their homolog Gfh1 of the Deinococcus-Thermus phylum described recently by us and others (6), and the bacterial stringent response regulator DksA (5) constitute a novel class of transcription factors that regulate transcription initiation and elongation by directly modulating the catalytic properties of RNAP by acting through the RNAP secondary channel (4-6). Although these factors have been extensively studied in the recent past, details of their molecular mechanism of action and their in vivo functions remain poorly understood.
Functional activities of transcript cleavage factors Gre.
GreA, GreB, and their eukaryotic analog TFIIS (SII) suppress or prevent transcription pause or arrest during elongation, enhance transcription fidelity, facilitate promoter escape by RNAP, and assist in transcription-coupled DNA repair (3, 6). These functions require stimulation by Gre/SII factors of an intrinsic nucleolytic activity of RNAP (3), which is an evolutionarily conserved function of all multisubunit RNAPs of bacterial, achaeal, viral and eukaryotic origin. During transcription in vitro and in vivo, RNAP can stall or become arrested upon encountering a roadblock such as a DNA binding protein, or a special DNA sequence, or after misincorporation. Some roadblocks induce RNAP to slide backwards along the template (“backtrack”) (7, 8), forcing RNA 3’-terminus to extrude through RNAP secondary channel, typically by 2-18 nt (3, 4). The endonucleolytic cleavage of such extruded RNA by RNAP catalytic center generates a viable 3’ terminus which can be extended in the presence of rNTPs, giving RNAP a second chance to read through the roadblock (7, 8). RNA cleavage by RNAP alone is weak, but is markedly stimulated by Gre/SII factors or by alkaline pH in the absence of factors (9, 3). GreA induces hydrolysis of di- and trinucleotides (type "A" cleavage activity), while GreB and SII induce cleavage of 2-18 nucleotide-long RNA fragments (type "B" cleavage activity) depending on the state of RNAP backtracking (1-3, 6-8).
Transition from initiation to elongation is also facilitated by Gre-induced cleavage reaction (10, 11). During abortive initiation DNA undergoes "scrunching", accounting for long abortive products synthesized by RNAP while still bound to promoter. The stress accumulated from scrunching (12) is relieved through disruption of RNAP-promoter contacts, leading to RNAP promoter escape, or through reversal of scrunching, which forces abortive products to dissociate through RNAP secondary channel. Alternatively, Gre may induce endonucleolytic cleavage of RNA, leaving the 5’ terminal fragment intact and poised for re-extension. By inducing cleavage, Gre factors prevent 5’-terminal RNA fragment from dissociating, thus increasing RNA occupancy and RNAP’s chance of promoter escape (13).
The cleavage activity of Gre factors is also implicated in cellular mechanisms that resolve conflicts between DNA repair/replication and transcription processes. GreA and GreB may help remove stalled or backed-up RNAPs from sites of chromosome lesion caused by UV-irradiation, oxidation or chemicals to provide access to DNA repair/recombination apparatus (14). Similarly, Gre factors may help resolve collisions between replication forks and stalled RNAPs at transcription terminators during co-directional replication and transcription processes (15).
Besides above mentioned functional activities requiring transcript cleavage, GreB (and to a much less extent, GreA) also exerts inhibitory effect on certain promoters by disrupting the binary RNAP-DNA closed (RPc) or open (RPo) promoter complexes (16). This activity of GreB, which is reminiscent of DksA activity (see below), was observed in vitro and in vivo but is yet poorly characterized.
Structures of GreA/B and their proposed mechanism of action.
All members of Gre family are homologous polypeptides of ~160 amino acids. The structures of GreA and GreB revealed by X-ray crystallography show similar topology comprising two domains, an N-terminal extended coiled-coil domain (NTD) and a C-terminal globular domain (CTD) (17-20) (Fig. 1A). The spatial organization of the two domains of Gre factors and most of the structural features of E. coli GreA and GreB are highly conserved. Their surface charge distribution is similar except for the small and large clusters of basic residues (basic patch) on the surface of GreA- and GreB-NTD, respectively. Basic patch residues are involved in Gre-RNA interaction in backtracked TCs which are important for induction of transcript cleavage and antiarrest activities (21). Our previous studies of individual domains of GreA (20) as well as chimeric GreA/GreB proteins (18) showed that NTD confers GreA- or GreB-type specific activity, whereas CTD anchors Gre to RNAP.
Gre-RNAP interaction requires a direct contact between Gre-CTD and the coiled-coil element of RNAP ß’ subunit that forms the rim of RNAP secondary channel (Fig. 1B) (19-23). The channel is an entry port for NTPs during synthesis, and is an exit port for RNA 3'-termunus during backtracking, in prokaryotic and eukaryotic RNAPs (6). The ß’ coiled-coil element serves as a docking platform for two other factors that act through the secondary channel, Gfh1 and DksA (see below). It would be instructive to learn how CTDs of GreA, GreB and Gfh1, and the RNAP binding domain of DksA, are differentiated by RNAP. Based on our biochemical, genetic and structure modeling data, we found that Gre-NTD occupies the secondary channel, and places 2 conserved acidic residues, D41 and E44, in the immediate vicinity of RNAP catalytic center where they coordinate the second catalytic Mg2+ required for nucleolytic reaction (22). A similar mechanism was proposed for SII based on the 3D crystal structure of SII-Pol II complex (24). Thus, Gre and SII represent the first known examples of transcription factors that modify RNAP activity by directly participating in the catalytic process.
Functional and structural properties of other Gre-like factors, Gfh1 and DksA.
Other factors similarly acting through the secondary channel have recently been described. Gfh1, a Gre-like protein from Thermus genus, is a potent inhibitor of transcription both in vivo and in vitro (25-28). It shares significant amino acid sequence and structural homology with GreA (Fig. 1), but lacks basic patch residues, does not interact with RNA, and is devoid of transcript cleavage activity (25-28). Gfh1 competes with GreA for binding to RNAP, and inhibits GreA-induced cleavage reaction as well as all other known catalytic functions of RNAP: RNA synthesis, pyrophosphorolysis, and exo-/endonucleolysis (61-64). It does so by directly competing with NTP’s or pyrophosphates for binding to the active site of RNAP (28). Four conserved acidic residues at the tip of Gfh1-NTD are implicated in metal ion coordination. Gfh1 activity is pH-dependent; at pH below 7, it exists in functionally “active” conformation, while at pH 7 and above, its CTD assumes an alternative conformation that renders Gfh1 incapable of RNAP binding, and thus “inactive” as inhibitor (28).
Another Gre-like protein, DksA, was recently identified as co-activator of (p)ppGpp-mediated bacterial stringent response (29, 30). While it shares no sequence homology with GreA, its X-ray crystal structure revealed a domain structure that bears a striking similarity to the coiled-coil structure of GreA-NTD (30). Based on biochemical, genetic, and structure modeling data, it was proposed that DksA, like Gre and Gfh1, reaches the RNAP catalytic center through its secondary channel, and partially co-opts metal binding function of RNAP using two conserved acidic residues at the tip of its coiled-coil domain (30). DksA is not a transcript cleavage factor, and it mostly affects the stability of RPo, either stabilizing or weakening it depending on the promoter. The activity of DksA is best characterized on the promoters of ribosomal and amino acid biosynthetic operons, where it has inhibitory and stimulatory effects, respectively (29),
The GreA/B-DksA, GreA-Gfh1, and possibly even GreA-GreB pairs display functional antagonism by stimulating and inhibiting (respectively) transcription from certain promoters in vitro and in vivo (4, 5, 16, 26, 28, 31). Intriguingly, DksA and GreB have dual functionality, and each can both inhibit and stimulate transcription depending on the promoter (16). Since all four factors bind RNAP similarly (22, 26-28; 30), they are expected to compete for RNAP in vivo and antagonize one another functionally according to their individual affinities and active intracellular concentrations. However, our recent in vitro studies show that these factors can discriminate between open promoter, initiation, elongation, and paused/arrested complexes. This implies that they target different classes of transcription complexes in vivo, and thus do not act as functional competitors/inhibitors. The molecular basis for such target specificity is yet unknown.
Understand the molecular mechanism of GreA, GreB, and Gfh1 action
Three experimental avenues will be explored:
- Identify the minimum discrete structural elements in GreA, Gre, and Gfh1 responsible for their specific stimulatory, inhibitory or pH-sensing functions using site directed mutagenesis, biochemical and in vitro transcription assays of modified Gre.
- Determine whether functional Gre-RNAP interaction is static or dynamic in nature using Gre proteins crosslinked to RNAP in the secondary channel at multiple sites, and testing their activity in transcription assays.
- Investigate the role of the ß’ G-loop, a conserved mobile element of secondary channel, in enabling Gre factors to discriminate between free RNAP and RNAP in open promoter complex (RPo), active elongation complex (EC) and backtracked complex (BC) using engineered “open” and “closed” conformations of G-loop, and testing the modified RNAP’s activity towards GreA and GreB.
Characterize the intermolecular interface for Gre factors in RNAP.
Several sites of Gre-RNAP interactions were recently identified by us and others in the C-terminal domain of Gre and on the rim of RNAP secondary channel. However, interactions of Gre N-teminal domain (NTD) with the elements of the secondary channel are still unknown. Genetic experiments will be used to identify residues in RNAP ß and ß’ subunits that interface the Gre-NTD and are essential for mediating their functional activities. We will apply site-directed mutagenesis to change specific residues in GreA-NTD and in the regions of ß and ß’ that are implicated in interaction with Gre based on biochemical, genetic and structural modeling data. We will also use random mutagenesis of ß and ß’ to look for suppressors of dominant lethal phenotypes of Gre mutants.
Establish the role of Gre/Gfh1 on a genome-wide scale
The majority of genes regulated by Gre/Gfh1 in vivo are yet unidentified. Moreover, the role of Gre/Gfh1 as regulators of cellular response to environmental stress remains unclear. We will perform microarray and phenotypic array analyses of E.coli and Tth using a collection of our mutant strains to determine the effect of Gre/Gfh1 on global gene expression under various growth conditions. We will also apply chromatin immunoprecipitation coupled to microarray analysis (ChIP-chip) to chracterize distribution/localizataion of Gre/Gfh1 on the chromosome in response to change in growth conditions. The potential targets of Gre/Gfh1 identified through microarray and ChIP-chip analyses will be further characterized by in vivo and in vitro assays.