Activation of Rho-dependent Transcription Termination by NusG
1999; Elsevier BV; Volume: 274; Issue: 8 Linguagem: Inglês
10.1074/jbc.274.8.5245
ISSN1083-351X
AutoresChristopher M. Burns, William Nowatzke, John P. Richardson,
Tópico(s)RNA and protein synthesis mechanisms
ResumoThere is a kinetic limitation to Rho function at the first intragenic terminator in the lacZ gene (tiZ1) which can be overcome by NusG: Rho can terminate transcription with slowly moving, but not rapidly moving, RNA polymerase unless NusG is also present. Here we report further studies with two other Rho-dependent terminators that are not kinetically limited (tiZ2 and λ tR1) which show that the requirement for NusG depends on the properties of the terminator and its location in the transcription unit. NusG is also shown to increase the rate of Rho-mediated dissociation of transcription complexes arrested at a specific termination stop point in the tiZ1 region and the rates of dissociation with three different Rho factors and two different terminators correlated with their sensitivity to RNA polymerase elongation kinetics. These results suggest a model of NusG function which involves an alteration in the susceptibility of the transcription complex to Rho action which allows termination to occur within the short kinetic window when RNA polymerase is traversing the termination region. There is a kinetic limitation to Rho function at the first intragenic terminator in the lacZ gene (tiZ1) which can be overcome by NusG: Rho can terminate transcription with slowly moving, but not rapidly moving, RNA polymerase unless NusG is also present. Here we report further studies with two other Rho-dependent terminators that are not kinetically limited (tiZ2 and λ tR1) which show that the requirement for NusG depends on the properties of the terminator and its location in the transcription unit. NusG is also shown to increase the rate of Rho-mediated dissociation of transcription complexes arrested at a specific termination stop point in the tiZ1 region and the rates of dissociation with three different Rho factors and two different terminators correlated with their sensitivity to RNA polymerase elongation kinetics. These results suggest a model of NusG function which involves an alteration in the susceptibility of the transcription complex to Rho action which allows termination to occur within the short kinetic window when RNA polymerase is traversing the termination region. Termination of RNA transcription in Escherichia coli is mediated by two distinct mechanisms. Intrinsic termination requires only RNA polymerase and specific signals in the DNA template. Factor-dependent termination is mediated by the action of Rho protein (1Richardson J.P. Greenblatt J.L. Neidhardt F.C. Curtiss III, R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology Press, Washington, D. C.1996: 822-848Google Scholar) with the assistance of NusG (2Sullivan S.L. Gottesman M.E. Cell. 1992; 68: 989-994Abstract Full Text PDF PubMed Scopus (150) Google Scholar). Although Rho can act by itself to terminate transcription efficiently in vitro at a number of terminators, an in vivo requirement for NusG has been shown to exist for several terminators. The nature of this intracellular requirement for NusG and the mechanism by which NusG acts to enhance Rho function must be understood to present a clear picture of factor-dependent termination in E. coli. The action of NusG has been examined for several Rho-dependent terminators in vitro (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar, 4Li J. Mason S.W. Greenblatt J. Genes Dev. 1993; 7: 161-172Crossref PubMed Scopus (110) Google Scholar, 5Nehrke K.W. Zalatan F. Platt T. Gene Expr. 1993; 3: 119-133PubMed Google Scholar). When reactions are carried out under conditions that are optimized for Rho function, NusG does relatively little. It can enhance the efficiency of termination somewhat and also activate promoter proximal termination stop points that are not normally used by Rho. However, a requirement for NusG in Rho function was identified for the first intragenic terminator in lacZ (tiZ1) when transcription reactions were performed using in vivo concentrations of NTPs which allow RNA polymerase to elongate RNA chains at the in vivo rate (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar). In this case, NusG is required to overcome a kinetic limitation to Rho function which is likely to exist in the intracellular environment. However, under these same conditions a second terminator downstream from tiZ1 on that same template, tiZ2, functioned as well in the absence of NusG as in its presence, suggesting that either it is a NusG-independent terminator or its position with respect to the start point of transcription can influence the relative requirement for NusG as a cofactor. Several individual activities of Rho factor are required to achieve termination of RNA transcription (6Platt T. Richardson J.P. McKnight S.L. Yamamoto K.R. Transcriptional Regulation. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1992: 365-388Google Scholar). Rho must first bind to the RNA. This interaction then activates an ATPase activity that allows the Rho hexamer to track along the RNA in the 5′ to 3′ direction (7Brennan C.A. Dombroski A.J. Platt T. Cell. 1987; 48: 945-952Abstract Full Text PDF PubMed Scopus (186) Google Scholar, 8Howard B. de Crombrugghe B. J. Biol. Chem. 1976; 251: 2520-2524Abstract Full Text PDF PubMed Google Scholar, 9Lowery-Goldhammer C. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1974; 71: 2003-2007Crossref PubMed Scopus (93) Google Scholar) and to dissociate the RNA from the transcription complex at a pause site (10Richardson J.P. Conaway R. Biochemistry. 1980; 19: 4293-4299Crossref PubMed Scopus (46) Google Scholar,11Shigesada K. Wu C.-W. Nucleic Acids Res. 1980; 8: 3355-3369Crossref PubMed Scopus (35) Google Scholar). Each of these steps has a kinetic component that could be affected by NusG. The effects of NusG on many of these steps of Rho action have been measured. NusG does not change the K d of the Rho-RNA binding interaction, the K m for RNA to activate ATPase, the V max of ATP hydrolysis, the concentration of Rho necessary for half-maximal termination, or the rate with which it can dissociate a single-stranded DNA base pairing to the 3′-end of the transcript (the helicase activity) (4Li J. Mason S.W. Greenblatt J. Genes Dev. 1993; 7: 161-172Crossref PubMed Scopus (110) Google Scholar, 5Nehrke K.W. Zalatan F. Platt T. Gene Expr. 1993; 3: 119-133PubMed Google Scholar, 12.Burns, C. M., Nus Factor Modulation of RNA Chain Elongation and Rho-dependent Transcription Termination in Escherichia coli Ph.D. Dissertation, 1996, 109, 110, Indiana University.Google Scholar). It has been shown to slow the dissociation rate of Rho from nascent RNA from 2 to 5 min (13Nehrke K.W. Platt T. J. Mol. Biol. 1994; 243: 830-839Crossref PubMed Scopus (47) Google Scholar). However, because RNA polymerase passes through the tiZ1 terminator in about 5 s, it is not clear how such a stabilization of the Rho-RNA interaction might enable termination of rapidly elongating polymerases at tiZ1. The final component activity is the dissociation of RNA from the transcription complex. This has been examined for transcription complexes randomly stopped in the trp t′ terminator region (5Nehrke K.W. Zalatan F. Platt T. Gene Expr. 1993; 3: 119-133PubMed Google Scholar). NusG was shown to increase the rate of dissociation of RNA from complexes that were stopped at the promoter proximal termination points, sites at which Rho-dependent termination occurs very slowly, if at all, in the absence of NusG. Thus, this result did not distinguish whether NusG enhances the rate of Rho-mediated RNA release or was simply allowing Rho to act at positions where it normally does not. In this report we investigate further the generality of the requirement for NusG to allow Rho to cause termination of rapidly moving RNA polymerase and the effects of NusG on the rate of transcript release. Specifically, we test the effects of NusG on two terminators,lac tiZ2 and λ tR1, placed at two different positions in transcription units. Our results indicate that both the location of the terminator sequence and the nature of the terminator itself influence greatly the kinetic restraints on Rho function and the requirement for NusG to overcome these restraints. To examine the effects of NusG on the rate of RNA release caused by Rho factor we made use of homogeneously stopped transcription complexes that were positioned at termination stop points within tiZ1 and λ tR1 which were normally used by Rho alone. The results from these assays reveal that Rho can mediate release of RNA from the complex at tR1 much more rapidly than from that at tiZ1 and that NusG enhances the rate of release for both. This action of NusG suggests that its effects on release are what allows it to overcome the kinetic limitations of Rho to function alone at certain terminators. Wild-type Rho, purified as described previously (14Nowatzke W.L. Richardson L.V. Richardson J.P. Methods Enzymol. 1996; 274: 353-363Crossref PubMed Scopus (27) Google Scholar), was provided by Lislott Richardson. NusG was from Barbara Stitt (Temple University) (15Washburn R.S. Jin D.J. Stitt B.L. J. Mol. Biol. 1996; 260: 347-358Crossref PubMed Scopus (14) Google Scholar), and NusA was from Richard Burgess (University of Wisconsin) (16Olins P.O. Erickson B.D. Burgess R.R. Gene (Amst.). 1983; 26: 11-18Crossref PubMed Scopus (15) Google Scholar). EcoRI-Gln111 mutant endonuclease was from Paul Modrich (Duke University) (17Wright D.J. King K. Modrich P. J. Biol. Chem. 1989; 264: 11816-11821Abstract Full Text PDF PubMed Google Scholar). F62S Rho (18Martinez A. Burns C. Richardson J.P. J. Mol. Biol. 1996; 257: 909-918Crossref PubMed Scopus (32) Google Scholar) was provided by Chon Martinez. The variant form ofMicrococcus luteus Rho, Mlu des(60–300) Rho lacks most of the 256-residue RNA-binding domain insertion segment that is a characteristic of M. luteus Rho (19Nowatzke W.L. Richardson J.P. J. Biol. Chem. 1996; 271: 742-747Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar) and also contains a hexahistidine amino-terminal extension. Mlu des(60–300) Rho was expressed in E. coli and purified using standard procedures (14Nowatzke W.L. Richardson L.V. Richardson J.P. Methods Enzymol. 1996; 274: 353-363Crossref PubMed Scopus (27) Google Scholar) plus chromatography on a Ni2+-nitrilotriacetic acid agarose (20Nowatzke W.L. Burns C.M. Richardson J.P. J. Biol. Chem. 1997; 272: 2207-2211Crossref PubMed Scopus (20) Google Scholar). E. coli RNA polymerase was purchased from Epicentre Technologies (Madison, WI). Enzymes used for DNA manipulations were from New England Biolabs. E. coli MRE600 tRNA was from Boehringer Mannheim. DNA oligonucleotides were from the Indiana Institute for Molecular Biology or Life Technologies, Inc. Gelman BioTrace NT filters were from Baxter Scientific Products Division. Other enzymes and reagents were from Sigma and J. T. Baker. pCBZ6, an 8,318-bp 1The abbreviations bpbase pair(s) plasmid, was constructed from pTL61T (21Linn T. St. Pierre R. J. Bacteriol. 1990; 172: 1077-1084Crossref PubMed Google Scholar) and pCBZ4 (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar). pTL61T was digested withBsu36I, the ends blunted with Klenow DNA polymerase, and then digested with EcoRI. The 8,073-bp fragment of pTL61T is missing the lac DNA sequences from −171 to + 277. The region from −171 to +75 was prepared by amplification of pCBZ4 from −178 to +75 with Vent DNA polymerase followed by cleavage withEcoRI, which reduces this segment to −171 to +75. The two fragments were then joined with DNA ligase. pCBZ6 has a deletion of 202 bp in the lacZ DNA from position +76 to +277. base pair(s) pCBC1, a 4,086-bp plasmid, was constructed by insertion of the 751-bpSau3A fragment of pDE13 (22Erie D.A. Hajiseyedjavadi O. Young M.C. von Hippel P.H. Science. 1993; 262: 867-873Crossref PubMed Scopus (266) Google Scholar), which contains a C to G change at position 6 of the cro gene, into pIF2 (23Faus I. Richardson J.P. Biochemistry. 1989; 28: 3510-3517Crossref PubMed Scopus (44) Google Scholar) cut withBglII. This generates a complete cro gene driven by its natural pR promoter. The C6G change allows preparation of stable ternary transcription complexes (A24 complexes) by transcription in the absence of CTP. pCBC2, a 3,979-bp plasmid, was constructed from pCBC1 by digestion withBglII and AvaI, filling in the ends with Klenow DNA polymerase and recircularization with DNA ligase. It contains a 107-bp deletion in the early region of the cro gene from position +86 to +192. The integrity of the DNA sequence of each plasmid in the transcription template region was confirmed by sequencing the DNA (24Sanger F. Nicklen S. Coulson A.R. Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467Crossref PubMed Scopus (52769) Google Scholar). The DNA templates used for in vitrotranscription from pCBZ4 and pCBZ6 were prepared by amplification of a segment of the lacZ DNA from (−167 to +839) with Vent DNA polymerase using primers designed to add EcoRI sites to both ends of the product DNA for cloning purposes. Both templates contain the C10T mutation necessary for A16 complex formation as well as the UV5 and L8 promoter mutations. The transcription templates derived from the plasmids pCBC1 and pCBC2 were prepared by digestion of the plasmids with PvuI. This enzyme cuts the DNA downstream from the cro gene (at position 649) and allows examination of all stop points in tR1. Both templates contain the C6G mutation necessary for A24 complex formation. The DNA templates used for the release assays were prepared by amplification of pCBZ4 or pCBC2 with Vent DNA polymerase. The pCBZ4 template is from position −167 to +189 of the lacZ DNA, and the pCBC2 template is from position −188 to +219. In both templates, primer-directed EcoRI sites were added immediately downstream from the gene segments as is required for this experiment. In vitro transcription reactions were carried out using ternary transcription complexes that were stopped by omission of CTP. A16 complexes on lacZtemplates from pCBZ4 and pCBZ6 were prepared as described (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar). A24 complex formation on cro templates from pCBC1 and pCBC2 was carried out by a modification of this procedure. 3 pmol of RNA polymerase was mixed with approximately equal molar amounts of DNA template in transcription buffer (150 mm potassium glutamate, pH 7.8, 40 mm Tris-HOAc, pH 7.8, 4 mm Mg(OAc)2, 1 mm dithiothreitol, 0.02% Nonidet P-40, 0.002% acetylated bovine serum albumin, 1% glycerol) and preincubated for 5 min at 37 °C. GTP, UTP, and [α-32P]ATP (350 nCi/pmol) were then added to 4 μm each and the reactions incubated at 16 °C for 10 min. The transcription complexes were isolated by gel filtration chromatography on Sephacryl S-300HR columns as described (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar). All enzymes used for transcription studies were diluted in transcription buffer containing 0.012% acetylated bovine serum albumin and 0.12% Nonidet P-40. Termination reactions (20 μl) were done using approximately 5 fmol of isolated complex in transcription buffer. Transcription factors were mixed with transcription complexes on ice and preincubated for 3 min at 37 °C. NTP mixtures were then added to either low (0.2 mm GTP, 0.2 mm ATP, 0.02 mm UTP, 0.2 mm CTP, 4 mm Mg(OAc)2) or high (1.1 mm GTP, 2.7 mm ATP, 1.4 mmUTP, 0.7 mm CTP, 10 mm Mg(OAc)2) concentrations and the reactions incubated for 3 min at 37 °C. The reactions were stopped and analyzed by 6% polyacrylamide and 7m urea gel electrophoresis as described (3Burns C.M. Richardson J.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4738-4742Crossref PubMed Scopus (64) Google Scholar). Quantitation of autoradiograms was done using a Molecular Dynamics scanning densitometer. Ternary transcription complexes were prepared with lacZ (A16 complexes) or cro (A24 complexes) DNA templates containing EcoRI sites. After complex formation, but prior to isolation, EcoRI-Gln111 mutant endonuclease (17Wright D.J. King K. Modrich P. J. Biol. Chem. 1989; 264: 11816-11821Abstract Full Text PDF PubMed Google Scholar) was added to 400 nm and the reactions incubated for 3 min at 37 °C. NTPs were then added to the low level with 10 μg/ml rifampicin and the reactions incubated an additional 3 min at 37 °C. The EcoRI-stopped complexes were then isolated by gel filtration chromatography.EcoRI-Gln111 acts as a roadblock to stop RNA polymerase 14 bp upstream from the EcoRI binding site (25Pavco P.A. Steege D.A. J. Biol. Chem. 1990; 265: 9960-9969Abstract Full Text PDF PubMed Google Scholar). The position of RNA polymerase stopped on the lacZ template was +175, and on the cro template it was +205 (which corresponds to position +312 of wild-type cro). These positions were verified by gel electrophoretic analysis of the RNA from the complex. Release assays were done in transcription buffer at 37 °C. Transcription factors were mixed with the transcription complexes on ice, preincubated for 3 min at 37 °C, and the reactions started by addition of 1 mm ATP. The 50-μl reactions were stopped after different times by the addition of 500 μl of 0.5 mKCl, 10 μg/ml poly(C), and 10 μg/ml heparin. The reactions were then incubated on ice for 5 min and filtered through BioTrace NT filters. The filters were washed three times with 2 volumes of stop solution. The amount of RNA that remained associated with the transcription complexes was determined by measuring radioactivity with a liquid scintillation spectrometer. To determine the importance of terminator position within a transcription unit on the ability of Rho to act on rapidly moving RNA polymerase, we prepared a derivative of the lacZ DNA template with the segment from bp 76 to bp 276 deleted (Fig. 1). The deleted segment contains the termination stop point region for tiZ1 (bp 140 to bp 229) plus its likely rut sequences. In this derivative, called tiZ2d (in the plasmid pCBZ6), the stop point region of the tiZ2, which normally runs between bp 360 and bp 480, is now between bp 160 and 280 or nearly in the same position as the stop point region of tiZ1 in the normal template (pCBZ4). This was verified by examination of the size distributions of the RNA molecules produced from transcription of the two templates with low NTPs in the presence of Rho (shown in Fig. 2 A). As expected, the addition of NusG under these conditions caused some RNA polymerase molecules to terminate transcription at earlier stop points in the tiZ1 region of pCBZ4 and in similar places in the tiZ2d region of pCBZ6. The overall efficiencies of termination at each terminator, determined by scanning densitometry of the autoradiogram shown in Fig.2 and others like it, are presented in Fig.3 A. The results demonstrate that tiZ2 functions in the absence of tiZ1 and thus confirms that tiZ1 and tiZ2 are distinct terminators.Figure 2Rho-dependent termination on wild-type and deletion templates. RNA transcription reactions were carried out with 50 nm Rho (hexamers) with and without 25 nm NusG, using both the low and high NTP levels as indicated in the figure. The lanes designated Uand C were generated by RNA chain terminating sequence reactions using 3′-dUTP or 3′-dCTP. The sizes of the transcripts were determined from the mobility of RNAs in these sequence lanes.Panel A, transcription of the lacZ DNA templates. pCBZ4 is the wild type, and pCBZ6 is a deletion variant that repositions tiZ2 202 bp closer to the promoter. These reactions were analyzed on a single polyacrylamide gel, but some lanes have been deleted from this figure. Panel B, transcription of thecro DNA templates. pCBC1 is the wild type, and pCBC2 is a deletion variant that repositions tR1 107 bp closer to the promoter.View Large Image Figure ViewerDownload (PPT)Figure 3Efficiencies of termination with different terminator constructs. Transcription reactions were done without any addition (open bars), with 50 nm Rho (stippled bars), or with 50 nm Rho plus 25 nm NusG (black bars). RNAs were separated by polyacrylamide gel electrophoresis and quantitated by scanning densitometry. The efficiency of termination is the percent of RNA polymerase molecules that encounter a terminator which terminate at that terminator. tiZ1 and tiZ2 were measured on the pCBZ4 template and tiZ2d from the deletion construct pCBZ6. tR1 was measured on a template from pCBC1 and tR1d from the deletion construct in pCBC2. Panel A, termination with the low NTP condition. Panel B, termination with the high NTP condition.View Large Image Figure ViewerDownload (PPT) The critical test is whether tiZ2 can still function in this new position when transcription is done with high NTPs in the absence of NusG. The results (Figs. 2 A and 3 B) show that virtually no termination occurred at tiZ2d in the pCBZ6 template in the absence of NusG and that only low levels of termination occurred with NusG present. As was found previously, however, Rho by itself terminated transcription in tiZ2 when it was in its normal position (pCBZ4; see Figs. 2 A and 3 B). Hence, deletion of 202 bp upstream of tiZ2 introduced a kinetic limitation to Rho function which was similar to that observed for tiZ1 in the normal template. This suggests that the amount of upstream RNA is an important determinant of the ability of Rho to act alone with rapidly moving RNA polymerase. The results also show that the weak intrinsic terminator located within tiZ2 (26Ruteshouser E.C. Richardson J.P. J. Mol. Biol. 1989; 208: 23-43Crossref PubMed Scopus (55) Google Scholar) is partially affected by the deletion and by the level of NTPs. This confirms similar observations on the importance of the overall sequence context (27Goliger J.A. Yang X. Guo H.-C. Roberts J.W. J. Mol. Biol. 1989; 205: 331-341Crossref PubMed Scopus (52) Google Scholar) and the elongation rate (28McDowell J.C. Roberts J.W. Jin D.J. Gross C. Science. 1994; 266: 822-825Crossref PubMed Scopus (105) Google Scholar, 29Reynolds R. Bermœdez-Cruz R.M. Chamberlin M.J. J. Mol. Biol. 1992; 224: 31-35Crossref PubMed Scopus (144) Google Scholar) in the Rho-independent termination process. The λ tR1 terminator is the only terminator to be studied in vitro which has been shown to be NusG-dependentin vivo (2Sullivan S.L. Gottesman M.E. Cell. 1992; 68: 989-994Abstract Full Text PDF PubMed Scopus (150) Google Scholar). It has been examined at the low NTP level and also at an intermediate NTP level (200 μm each NTP). In neither case was Rho function greatly dependent on NusG (4Li J. Mason S.W. Greenblatt J. Genes Dev. 1993; 7: 161-172Crossref PubMed Scopus (110) Google Scholar, 15Washburn R.S. Jin D.J. Stitt B.L. J. Mol. Biol. 1996; 260: 347-358Crossref PubMed Scopus (14) Google Scholar, 18Martinez A. Burns C. Richardson J.P. J. Mol. Biol. 1996; 257: 909-918Crossref PubMed Scopus (32) Google Scholar,19Nowatzke W.L. Richardson J.P. J. Biol. Chem. 1996; 271: 742-747Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). We examined Rho function at tR1 during transcription with the low and high concentrations of NTPs which revealed the kinetic limitation to Rho function at tiZ1. With both NTP levels, Rho functioned very efficiently on its own, and the addition of NusG did not significantly improve the termination efficiency (Fig. 2 B, pCBC1 template, and Fig. 3). Like tiZ2, tR1 has a longer upstream sequence than tiZ1 does (Fig. 1). To determine if this extra upstream DNA was responsible for the ability of Rho to act alone at tR1 with the high NTP level, a deletion variant (referred to as tR1d) lacking residues +86 to +192 was made in which the tR1 stop points are positioned at a distance from the promoter similar to those in tiZ1, while leaving the rutregion (residues 224–380; (30Chen C.-Y.A. Richardson J.P. J. Biol. Chem. 1987; 262: 11292-11299Abstract Full Text PDF PubMed Google Scholar)) intact (Fig. 1). The ability of Rho to act was not greatly affected by this deletion. The distribution of stop points closely resembled those of the natural tR1 terminator but shifted closer to the promoter by the expected 107 nucleotides (Fig. 2 B). With the low NTP level, termination at tR1 and tR1d was indistinguishable with and without NusG (Fig. 3). With the high NTP level, termination at tR1d was reduced slightly compared with tR1 and was not affected by NusG (Fig. 3). Thus, although position within a transcription unit with its consequent effect on the size of the nascent transcript can influence whether Rho action depends on NusG, other aspects of the terminator also contribute to this requirement. We have shown previously that NusG can partially compensate for a defective Rho factor. F62S Rho is partially defective in vivo but almost completely defective in vitro at tR1, giving only 8% termination with low NTPs. NusG was able to restorein vitro termination to 65%, nearly the in vivoefficiency (18Martinez A. Burns C. Richardson J.P. J. Mol. Biol. 1996; 257: 909-918Crossref PubMed Scopus (32) Google Scholar). We test here the ability of F62S Rho to terminate transcription at the lacZ intragenic terminators (Figs.4 and 5). The results with the low NTP level closely parallel those for the tR1 terminator in that F62S Rho caused very little termination on its own, but it was able to terminate transcription with moderate efficiency when NusG was also present (Figs. 4 and 5).Figure 5Efficiencies of termination by different Rho factors at tiZ1 and tiZ2. The amounts of Rho and NusG used were as described in Fig. 4. Efficiencies were determined as described in Fig.3. Panel A, with low NTP level; panel B, with high NTP level. Open bars, at tiZ1 with Rho alone;gray bars, at tiZ1 with Rho and NusG; stippled bars, at tiZ2 with Rho alone; solid bars, at tiZ2 with Rho and NusG. The extents of termination in the absence of Rho were the same as shown in Fig. 3. For tiZ2, which has a weak intrinsic terminator, those values were the same as with the F62S Rho in the absence of NusG.View Large Image Figure ViewerDownload (PPT) F62S Rho did not cause significant termination of transcription at tiZ1 with the high NTP level, even when NusG was present (Figs. 4 and 5). The inability of F62S Rho to terminate at tiZ1 with the high NTP level suggests that the kinetic limitation that exists for wild-type Rho factor is more substantial for F62S Rho because NusG can no longer overcome this limitation. Further, F62S Rho was unable to cause any increase of termination above the intrinsic levels at tiZ2 (Figs. 4 and5) or any termination at all at tR1 (not shown) with the high NTP level, even with NusG present. This was not expected as wild-type Rho was able to act at these terminators as efficiently with high NTPs as with low NTPs. This suggests that the rate of chain elongation imposes limitations on Rho action at all terminators but that wild-type Rho can normally overcome these limitations on its own for many terminators. The Rho factor from M. luteus is unusual. It terminates with higher efficiency and at more promoter proximal positions than E. coli Rho at tR1 (19Nowatzke W.L. Richardson J.P. J. Biol. Chem. 1996; 271: 742-747Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). We have examined the ability of this Rho and a deletion variant that behaves more like E. coli Rho (20Nowatzke W.L. Burns C.M. Richardson J.P. J. Biol. Chem. 1997; 272: 2207-2211Crossref PubMed Scopus (20) Google Scholar) to terminate at the lacZ intragenic terminators. The wild-type M. luteus Rho terminated transcription at tiZ1 with near 100% efficiency at both the low and high NTP levels and was only slightly affected by NusG (not shown). The deletion variantM. luteus des(60–300) Rho factor behaved very similarly to the E. coli Rho protein during transcription of thelacZ template with E. coli RNA polymerase. The positions of the stop points were similar to those caused by E. coli Rho, although some earlier points were also observed (Fig.4). Using the low NTP level, this Rho factor terminated transcription with very high efficiency (>90%) at both tiZ1 and tiZ2 (Fig. 5). With the high NTP level, the efficiency of termination was somewhat lower at both terminators. Termination efficiency at tiZ2 was similar to that with E. coli Rho (about 60%). However, the efficiency of termination at tiZ1 was very different from that caused by E. coli Rho (Fig. 5). At this terminator, the deletion variant of M. luteus Rho terminated transcription on its own with high efficiency (about 50%). This efficiency was enhanced by NusG to about 70%. These results suggest that the kinetic limitation of E. coli Rho to cause termination at tiZ1 does not exist for M. luteus des(60–300) Rho and also that E. coli NusG can activate the termination activity of a heterologous Rho factor. Because NusG allows Rho to function within a kinetic window that is too short for the action of Rho alone, NusG likely affects the rate of Rho action to allow termination of rapidly elongating polymerase. To determine if NusG changes the rate of Rho-mediated dissociation of RNA from transcription complexes, RNA release assays were done using transcription complexes in which RNA polymerase had been elongated to position 175 on th
Referência(s)