RNA Polymerase Pausing and Transcript Release at the AtRl ...

T H E J O U R N A L OF BIOLOGICAL CHEMISTRY Vol. 25R, No. 15, Issue of August 10, pp. 9391-9397, 1983 Prmted in U.S.A.

RNA Polymerase Pausing and Transcript Release at the AtRl Terminator in Vitro*

(Received for publication, December 27, 1982)

Lester F. Laus, Jeffrey W. Roberts§, and Ray WuB From the Section of Biochemistry, Molecular and Cell Biology, Cornell University, Ithaca, New York 14853

Transcription of the X rightward operon encounters multiple tandem p-dependent termination sites at a region known as tR1, yielding the mRNA for the lytic repressor cro. We show that in the absence of p factor, RNA polymerase pauses at each of three p-dependent termination sites (I, 11, and 111) at tR1 during in vitro transcription. W e demonstrate that p factor itself does not affect transcriptional pausing at these sites. Rela- tive occupancies of the pause sites during initial transcription have been measured. The Xcin-lcnc-1 mutations, which abolish or vastly reduce termination at site I1 of tRlr nevertheless elicit transcriptional pausing at that site. The NusA protein (L factor) prolongs these transcriptional pauses to varying degrees. Transcrip- tional pausing at termination site I1 is dramatically enhanced by NusA in X wild type, but not in Xcin-lcnc- 1. NusA enhances pausing at site I but inhibits termination at that site, suggesting that it may also affect transcript release. We show that NusA does alter the kinetics of transcript release from isolated ternary complexes. We suggest that the ability of the NusA protein to increase or decrease termination efficiencies at various sites is attributed to its ability to modulate transcriptional pausing as well as transcript release.

Correct termination is important to the regulation of transcription in prokaryotes (for reviews, see Roberts, 1976; Adhya and Gottesman, 1978; Pribnow, 1979; Rosenberg and Court, 1979; Greenblatt, 1981; Platt, 1981; Yanofsky, 1981; Ward and Gottesman, 1982). In uitro, some DNA sequences function as terminators without additional factors, whereas others require interposition by the termination factor p (Roberts, 1969). While a large number of studies has provided a rela- tively detailed description of the p-independent termination process (see Platt, 1981), the mechanism of p-dependent termination is not clearly understood.

Transcription termination necessarily entails at least two separate events: cessation of transcription elongation and release of nascent RNA from the DNA-RNA-polymerase transcription ternary complex. p-independent termination sites characteristically consist of a 3’-terminal stretch of uridine residues in the RNA transcript, following a stem and loop structure rich in GC residues. It is thought that formation

* T h e costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address, Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine, Baltimore, MD 21205.

§ Recipient of National Institutes of Health Grant GM 21941. 7 Recipient of National Science Foundation Grant DAR-79-

171310.

of the stem in the nascent RNA causes RNA polymerase to pause, while the stretch of uridine residues forms an unstable (U-dA), heteroduplex with the DNA template that is thought to promote dissociation of the RNA, resulting in termination (Martin and Tinoco, 1980). Among p-dependent termination sites, however, no significant stretches of uridine residues can be found. The role of p factor in these cases may be to release the nascent transcript in an ATP-dependent reaction (Adhya and Gottesman, 1978; Richardson and Conaway, 1980). Ac- cording to this view, transcript release must occur at high probability during the pause for efficient termination to take place.

Transcription of the X rightward operon encounters p- dependent termination sites at the tR1 region, located downstream of the structural gene for the lytic repressor cro (see Fig. 1A). Termination at A t R l is prevented by the X N gene product, thus allowing transcription of distal genes. Regula- tion of transcription termination at tH1 and at other X transcription terminators is crucial to the regulated expression of genes essential for lytic growth or lysogeny (Herskowitz and Hagen, 1980). We have recently characterized three tandem p-dependent termination sites (I , 11, and 111) at A t R l (Lau et al., 1982; Fig. Z), extending earlier studies of this region (Roberts, 1969; Rosenberg et al., 1978; Court et al., 1980). To further understand the termination process, we now examine RNA polymerase pausing and transcript release at the three termination sites of A t R I . Several questions motivated this investigation. Although Rosenberg et al. (1978) noted a transcriptional pause at tRl, it is not clear how this pause relates to the three termination sites. Does the extent of transcriptional pausing at a site correlate with the termination efficiency at that site? How does NusA protein (L factor), which affects the efficiency of termination (Greenblatt et al., 1981), influence pausing and transcript release? We approach these questions by characterizing the transcriptional pausing behavior a t A t R l using wild type and mutant templates, and by investigating the roles of p factor and NusA in pausing and transcript release. We show that RNA polymerase pauses a t all three termination sites at tR,, and that pausing is not quantitatively correlated with termination efficiency. p factor only affects the release of transcripts from pause sites and does not influence transcriptional pausing. By contrast, NusA protein modulates both pausing and release, We argue that these activities contribute to its ability either to increase or to decrease termination efficiencies at different sites.

EXPERIMENTAL PROCEDURES

Materials-[a-”’PJUTP and [a-”’P]CrTP a t about 400 Ci/mmol were purchased from Amersham. Unlabeled nucleoside triphosphates were purchased from P-L Biochemicals. Restriction enzymes were obtained from New England Biolabs. Heparin was purchased from Sigma and poly(C) was from Collaborative Research.

D N A Preparations-XcI857Nam7Nam53Sam7 DNA was a gener-

9391

by guest on March 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

9392 RNA Polymerase Pausing and Transcript Release

A. PRE

* PRM P 2 3 F % CJ ORPR CIO y cn 0

/-

, e - o , 9 s - t R , -"""""""""" --/+CtR2

B

t t f ? w pf 1 I

ORPR cro r urn ' R i

t . Readthrough 372 Nucleotides

FIG. 1. Partial genetic and physical maps of phage A. A, partial genetic and transcription map of phage X. Horizontal arrows indicate directions of transcription. A p-dependent transcription terminator, XtRl, is located downstream of cro. B, the HinfISGOrestriction fragment used for most transcription experiment described, which yields a read-through transcript of 372 nucleotides, is shown on an expanded scale. The locations of some restriction sites within this

AuoI; 0, MboII. fragment are also indicated: 0, HinfI; A, AluI; 0, BgLII; V, HaeIII; x,

ous gift from E. J. Grayhack, Cornel1 University. Strains XcZ857cin- Icy42Sam7 (Wulff, 1976) and Xc1857cin-lcnc-Icy42Sam7 (McDermit et al., 1976) were obtained from D. L. Wulff, State University of New York a t Albany. X DNA was purified as described by Schleif and Wensink (1981). Restriction fragments were isolated from 1% low melting agarose gels (Bethesda Research Laboratories) after electrophoresis. A gel slice containing the desired restriction fragment was melted a t 80 "C, followed by the addition of NaCl to 0.5 M and extracting twice with phenol. The resulting DNA was twice precipitated with ethanol before use in transcription reactions.

Purification of Proteins-RNA polymerase holoenzyme, 0 factor, and NusA were all purified from RNase I- strain MRE600 (Grain Processing, Muscatine, IA). RNA polymerase was purified according to the procedure of Burgess and Jendrisak (1975) as modified by Lowe et al. (1979). The final holoenzyme preparation was at least 95% pure and 95% saturated with u factor. p factor was purified as described by Finger and Richardson (1981), and the final preparation was homogeneous as judged by sodium dodecyl sulfate gel electrophoresis. The NusA protein was purified using a modification of the procedure described by Kung et al. (1975). We used the transcription depression assay of Greenblatt et al. (1981) and sodium dodecyl sulfate gel electrophoresis to monitor the NusA preparation. After the second DE52 column (Kung et ~ l . , 19751, fractions containing NusA were pooled and dialyzed against a column buffer (10 mM KP,, pH 7.0, 1 mM DTT,' 1 mM EDTA, 10% glycerol, 100 mM NaC1). The dialysate was precipitated a t 60% ammonium sulfate saturation, and taken up in 2 ml of the above buffer. The protein (3 mg) was then passed through a 50-ml AcA 44 (LKB) gel filtration column equilibrated in the same buffer. Peak fractions containing NusA were combined and dialyzed against a loading buffer (20 mM KP,, pH 7.5, 0.1 mM DTT, 0.1 mM EDTA, and 25% ammonium sulfate) in which a 4-ml phenyl- Sepharose (Sigma) column was equilibrated. The protein was loaded onto the phenyl-Sepharose column, washed with 2 column volumes of loading buffer, and eluted with a 40-ml gradient from the loading buffer to 20 mM KP,, pH 7.5, 0.1 mM DTT, 0.1 mM EDTA, and 60% ethylene glycol. NusA was eluted a t about 50% ethylene glycol and 4% ammonium sulfate. The resulting NusA preparation was 95% pure as judged by sodium dodecyl sulfate gel electrophoresis, and was dialyzed against 10 mM KP,, pH 7.0, 100 mM NaCl, 1 mM EDTA, 1 mM DTT, 50% glycerol for storage.

In Vitro Transcription-Single round in vitro transcription reactions were carried out as follows. A reaction mixture containing 150 p~ ATP, 150 p~ GTP, 150 p M CTP, 20 mM Tris-HC1, pH 7.9 (as measured a t 23 "C), 0.1 mM EDTA, 0.1 mM DTT, 50 mM KCI, 200 pCi/ml of [R-:"P]UTP at 400 Ci/mmol, 10 pmol/ml of DNA restriction fragment as template, 3 pg/ml of RNA polymerase, and 5% glycerol was pre-incubated a t 37 "C for 10 min. Protein factors were added as indicated. The concentration of p factor used (6 pg/ml) saturates the assay without being in large excess. Templates were in excess of active RNA polymerase under these conditions, which we verified by determining DNA dose response. Reactions were initiated

~~ _ _ ~ ~ _ _ ~- ' The abbreviation used is: DTT, dithiothreitol.

synchronously by the simultaneous addition of MgC12 to 5 mM, UTP to 25 p ~ , and rifampicin to 10 pg/ml. Reactions were incubated a t 37 "C, and 50-pl aliquots were withdrawn at indicated times and immediately added to Eppendorf tubes containing 50 p1 of 50 mM EDTA and 100 p1 of phenol equilibrated with Tris-HCI, pH 8.0, to terminate the reactions. Terminated reaction mixtures were extracted with phenol, followed by extraction with chloroform/isoamyl alcohol, 24:l (v/v), and ethanol precipitated with Escherichia coli tRNA (10 pg/sample) as carrier. The samples were then dissolved in a loading dye mixture containing 90% formamide, 50 mM Tris-HCI, pH 8.0, 0.05% bromphenol blue, and 0.05% xylene cyanol. Transcripts were analyzed by electrophoresis on 8% polyacrylamide, 8 M urea slab gels in a buffer containg 50 mM Tris-borate, pH 8.3, and 10 mM EDTA followed by autoradiography.

Isolation of Ternary Complexes and Assay for Transcript Release- Transcription ternary complexes were isolated and assayed for RNA release using the procedures of Richardson and Conaway (1980) with minor modifications. After in vitro transcription for 60-120 s (as indicated in text), the reactions were arrested by the addition of EDTA to 10 mM and chilling on ice. RNA transcripts were labeled using [w:"P]UTP as described above. Ternary complexes were isolated by chromatography a t 4 "C on a Bio-Gel A-5m column (1 X 15 cm) (column buffer: 40 mM Tris-HC1, pH 8.0, 50 mM KC], 0.1 mM MgC12, 0.1 mM DTT). Reaction mixtures for release assays contained 20 mM Tris-HC1, pH 8.0, 50 mM KCI, 1 mM MgC12, 1 mM ATP, 0.1 mM EDTA, 0.1 mM DTT, and isolated ternary complexes. p factor was added to 6 pg/ml where indicated, and samples were incubated at 37 "C for 2 min prior to the addition of p . Aliquots (120 pl) were withdrawn a t indicated times and added to Eppendorf tubes containing 500 p1 of filtration buffer (0.5 M KCl, 10 pg/ml of poly(C), and 10 pg/ml of heparin) to stop the reaction. Such a reaction mixture was filtered through a 2.5-cm nitrocellulose membrane (0.45 p ; Schleicher and Schuell), and washed with 1 ml of the filtration buffer. The filter was dried and counted in a toluene-based scintillation fluid.

RESULTS AND DISCUSSION

RNA Polymerase Pauses at AtRl during in Vitro Transcrip- tion-We examined RNA polymerase pausing in transcription of the X rightward operon using a restriction fragment produced by t,he enzyme Hinfr (560 nucleotide pairs) as template, shown in Fig. 1B. This restriction fragment contains the X rightward promoter P R , and extends through the structural gene for cro and the three tandem termination sites (I, 11, and 111) a t tR1. A single round of transcription was carried out in the absence of p factor as described under "Experimental Procedures." Transcription of the X wild type template for various periods of time showed paused RNAs at several dis- crete sites that were chased into longer RNA upon prolonged incubation (Fig. 3). Notably, RNA polymerase pauses at all three termination sites in the absence of p factor. These transcriptional pause sites are identical with the p-dependent termination sites within the resolution of our gel electrophoresis analysis (+3 nucleotides), although they could differ from the termination sites in the fine distribution of 3'- termini within each termination cluster. These transcriptional pauses do not depend upon low concentrations of UTP in the transcription reaction mixture. The pausing pattern was similar when transcription reactions were carried out at 150 PM UTP, while GTP was lowered to 40 FM and transcripts were labeled with [cv'''PP]GTP (data not shown). Under these conditions, pausing at site 111 was slightly enhanced and pausing at site I1 somewhat diminished. It has been reported that guanosine tetraphosphate can enhance transcription termination and pausing (Kingston and Chamberlin, 1981; Kingston et al., 1981). We do not observe an effect of 0.2 mM guanosine tetraphosphate on termination a t htR, (data not shown), and thus have not examined its effect on transcriptional pausing.

The fact that RNA polymerase pauses at each of the p- dependent termination sites indicates that these sites are recognized by polymerase in the absence of p factor. However,


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

FIG. 2. Nucleotide sequence of the transcript made from the X rightward promoter (Schwarz et al., 1978; Rosenberg et af., 1978). Indi- cated on the sequence are the transcription termination sites tlllr (Calva and Burgess, 1980), and I. 11, and 111 at till (Lau ct a/.. 198'2). Also shown are the nucleotide changes of the cin-1, cnc-1, and the cy42 mutations, as well as the NutR sequence (Kosenherg (7t a/., 1978). The approximate locations of pause sites upstream of' tIl1. P S I and PS2. are indicated.

i 2

RT m II I

PS2

PSI

3 4 5 6 7

PS2

8 9 to 1 1 12 -

SECONDS 600 5 Rho + -

15 30 45 60 75 " " _ 90 905 135 165 600 " " _ FIG. 3. Transcriptional pausing at XtR,. Wild type X DNA

(Hinflncu, fragment) was transcrihed for the times indicated in the absence of' factor (lanc,s 2-12). The read-through transcript (RT) and transcripts terminated at sites I , 11, and 111 were made during a IO-min incuhation in a reaction containing 6 pg/ml of p factor (lane 1) . Transcripts of molecular weights higher than the read-through transcript result f'rom initiation at the ends of the restriction fragment: this initiation is inhibited by /J factor (Lau e t nl., 1982).

not all pause sites are [dependent transcript release sites (Kassavetis and Chamberlin, 1981). We detected other pause sites upstream of till that are not p-dependent termination sites, including two major pause sites to which we refer a s PSI and PS2 (Figs. 2 and 8) . We mapped the approximate locations of these pause sites by comparing the sizes of these paused RNAs determined by polyacrylamide gel electrophoresis to marker transcripts. These marker transcripts were made from several templates containing the same promoter but truncated by restriction cuts. These templates included

-1 I l l

g 4 2

restriction fragments produced by the following enzymes: Hinfl-AvaI, Hinfl-MboII, Hinfl-HaeIII, Hinfl-RglII, and AluI (see Fig. l R ) , which yielded transcripts expected to be 192, 188, 115,81, and 76 nucleotides in length, respectively. RNAs paused a t PS1 and PS2 were found to be about 105 and 190 nucleotides long, respectively (data not shown), although their precise 3"terrnini were not determined. I t should be noted that PS1 is located close to tH,,, found to be a p-dependent termination site functioning in vitro to produce a 5 S RNA (Calva and Burgess, 1980). However, we have not detected such a 5 S RNA in the presence of p factor.

We excised gel slices containing the RNAs paused at termination sites I, 11, and 111 at various times and counted the radioactivity in the transcripts. The counts were normalized to the lengths of the transcripts, and scaled to the maximum amount of read-through transcript in that experiment. The results are plotted in Fig. 4A. T h e relative occupancies of these pause sites can be determined by measuring the area under each peak, and are shown in Table I.

Similar experiments were carried out on two mutants thought to affect termination at the till region, kin-1 and kin-lcnc-1 (Rosenberg et al., 1978; see Fig. 2). k in -1 was isolated to allow cll-independent lysogeny (Wulff, 1976), which it does for reasons probably unrelated to termination at tl{l; k in- lcnc-1 is a second site revertant of kin-I (Mc- Dermit et al., 1976). Rosenberg et al. (1978) found that cin-I increases termination efficiency at tl{l, whereas the cin-lcnc- I mutations decrease termination efficiency at trll. We did not detect an increase in termination efficiency due to cin-I; however, we showed that cin-Icnc-1 specifically abolished termination at site 11, and only site 11, of till (Lau et al., 1982). Since the cnc mutation has not been separated from cin, the effect of cincnc may be due to cnc alone or both cin and cnc. We examined the transcriptional pausing behavior of these mutants and the results are shown in Fig. 4, R and C, and Table I. We observed transcriptional pausing at all three till termination sites in both mutants in the absence of p factor.

T h e cin-1 mutation, either alone or in combination with cnc- 1 , reduces the occupancy of site I from 36 to about 25 mol min in 500 mol min of initial transcription. Other than this minor effect, k i n - 1 exhibits pausing behavior nearly indistin- guishable from wild type. The most striking observation is that k i n - l c n c - I , which abolishes termination at site 11, does not eliminate the pause at that site. kin-lcnc-I still elicits a substantial pause at site 11, although the occupancy is reduced by about 40% (Table I ) . A similar reduction in pausing at site I imposed by the cin-I mutation, however, does not affect termination efficiency at that site. Our interpretation of these results is that cnc-1 is primarily a mut,ation in p-dependent


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

9394 R N A Polymerase Pausing and Transcript Release

FIG. 4. Time course of appearance and decay of paused RNA spe- cies at A h . Transcription was carried out in the absence of p factor using the HinfIbW restriction fragments from the following X strains as templates: A, wild type; B, cin-1; C, cin-lcnc-1. Transcrip- tion products were analyzed by gel electrophoresis and bands corresponding to the read-through transcript (0) and RNAs paused at termination sites I (A), I1 (0), and I11 (A) were excised and counted. The counts were normalized to the lengths of the transcripts and scaled to the maximum accumulation of read- through transcript in that experiment. Data in A were taken from the gel shown in Fig. 3 .

TABLE I Relative occupancies of transcriptional pause sites

Data were taken from Figs. 4 and 5. The area under each peak within the initial 5 min of transcription is measured and the inte- grated occupancy is expressed in terms of mole minutes. The total amount of transcription measured is 500 mol min. Thus, a pause occupancy of 200 mol min implies that within the initial 5 min of transcription, RNA polymerase molecules spend 40% of the time paused at that site.

Without NusA With NusA Site X wild Xcin-l Xcin- X wild Xcin-l kin-

lcnc- 1

mol rnin twe lcnc-l type _ _ ~

I 36 25 23 52 44 24 I1 27 23 16 200 177 36

111 19 25 19 35 39 31

transcript release and affects pausing only in a minor way. Our result is in conflict with that of Rosenberg et al. (1978), who reported that pausing at tR1 is abolished by the cincnc mutations. We note that their results are more comparable to the pausing pattern we observe in the presence of the NusA protein (see below), although they did not include NusA protein.

Our results are consistent with an earlier proposal that RNA polymerase pausing at a site is required for p-dependent termination (Rosenberg et al., 1978). However, not all pause sites are p-dependent release sites, nor is the propensity for RNA polymerase to pause at a site quantitatively proportional to the termination efficiency at that site. For instance, the relative occupancy of sites I, 11, and I11 are 36, 27, and 19 mol min, respectively, whereas the termination efficiencies are 44, 62, and 67%, respectively (Lau et al., 1982). The discrepancy between pausing and termination efficiency becomes more evident as we study the effects of NusA, as detailed below.

NusA Prolongs Transcriptional Pauses-The NusA protein (L factor) is required for N protein-mediated antitermination (Friedman, 1971; Greenblatt et al., 1980) and affects efficiencies of transcription termination (Kung et al., 1975; Green- blatt et aL, 1981; Farnham et al., 1982). NusA has been shown to bind directly to core RNA polymerase and thus might be a subunit of RNA polymerase during elongation (Greenblatt and Li, 1981). Others have found that NusA enhances transcriptional pauses in transcription of the E. coli rrnB operon (Kingston and Chamberlin, 1981) and phage T7 (Kassavetis and Chamberlin, 1981), a t ktKn (Greenblatt et al., 1981), and in transcription of the trp operon (Farnham et al., 1982).

We examined transcriptional pausing as before but in the presence of NusA, shown in Fig. 5. The relative occupancy of

S E C O N D S

each pause site is shown in Table I. In both X wild type and kcin-I, NusA increases the occupancies of sites I and 111 up to 80% (Table I). The occupancy of site I1 is dramatically increased in the presence of NusA by more than 7-fold. During the initial 5 min of transcription, RNA polymerase molecules spend nearly 40% of the time pausing at tRI site I1 on the wild type or cin-I template in the presence of NusA. Strikingly, there is no enhancement of pausing at site I1 by NusA on the kcin-Icnc-1 template.

Thus, the single nucleotide change of cnc-I is sufficient to eliminate the NusA-dependent enhancement of transcriptional pausing at site 11. NusA must therefore respond to a signal that dictates an enhancement of pausing at site 11, either by recognizing a sequence directly or responding to the effect of a sequence on the polymerase. Moreover, since CRC

appears to be defective in p-dependent transcript release, the signal or condition for p-dependent release and NusA enhancement of pausing may overlap or in fact be identical.

The peaks of RNAs paused at termination sites I, 11, and I11 are distinctly sequential (Fig. 5 ) . However, we cannot determine what fraction of polymerase molecules pause at each site prior to read-through transcription. Some fraction of RNA polymerase molecules may continue RNA synthesis through tR1 without pausing, while other molecules may pause at one, two, or all three sites before transcription elongation continues.

Effect of p Factor on Transcriptional Pauses-Since termination at tH1 requires p factor, it is important to examine if p factor itself affects transcriptional pausing. We approached this question as follows. First, we followed the kinetics of appearance and decay of paused RNAs at PS1 and PS2, which are not p-dependent release sites, in the presence and absence of p factor, as shown in Fig. 6. Under both conditions, the peaks for both paused RNAs essentially superimpose, indi- cating that pausing is insensitive to p factor at these sites.

Since p factor does not appear to alter the kinetics of transcription elongation upstream of t R l , we can monitor the effects of p factor on pausing at termination sites I, 11, and I11 by following the appearance of the read-through transcript. I t is possible to quantitate the read-through transcript in the presence of p factor since tRl is not 100% efficient in vitro (Rosenberg et al., 1978; Lau et al., 1982). If p factor affects pausing at tRl of transcripts that do not actually terminate, then the rate of accumulation of the read-through transcript should be altered. We examined the time course of transcription with k wild type HinfT5b.o fragment as template, with 6 pg/ml of p factor included in the reaction mixture prior to pre-incubation. Aliquots (50 ~ 1 ) were withdrawn at various times and analyzed as before (Fig. 7A). The time required to


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

RNA Polymerase Pausing and Transcript Release 9395

FIG. 5. Effects of NusA on transcriptional pausing at AtR1. Tran- scription was carried out in the presence of NusA (5 pg/ml) using the H i n f k restriction fragments from the following X strains as templates: A , wild type; B, cin-I; C, cin-lcnc-I. The data were analyzed as described in Fig. 4. The time course of appearance of the read-through transcript (0) and appearance and decay of RNAs paused a t termination sites I (A), I1 (O), and 111 (A) is shown.

i 2 3 4 5 ( 2 3 4 5 1 2 3 4 5

M I N U T E S

i 2 4

SECONDS

FIG. 6. Time course of appearance and decay of paused RNAs upstream of tR1. Transcription was carried out using the X wild type H i f l f I a a o fragment as template. RNAs paused at PS1 in the presence (0) and absence (0) of p factor, and at PS2 in the presence (A) and absence (A) of p factor were determined by gel electrophoresis. p factor was included where indicated at 6 pg/ml. Bands containing the PS1 and PS2 RNAs were excised and counted, and the counts were normalized to the lengths of the transcripts. Multi- plication of the ordinate values by 2 X lo-” gives actual molar amounts.

FIG. 7. Accumulation of transcripts in the presence of p factor. Transcription was carried out with the X wild type HiflfIseo fragment in the absence ( A ) and presence ( B ) of NusA (5 pg/ml). The accumulation of the read-through transcript (O), and transcripts terminating a t site I (A), I1 (O), and 111 (A) are shown. Multiplication of the ordinate by 2 X lo-” gives actual molar amounts.

reach half-maximal accumulation of the read-through transcript is 70 s either in the presence or absence of p factor (compare Fig. 4A to Fig. 7A). A similar experiment was carried out in the presence of NusA. The rate of accumulation of each transcript was measured after transcription of the HinflSsO fragment in the presence of p factor (6 pg/ml) and NusA (5 pg/ml), as shown in Fig. 7B. The time of half- maximal accumulation of the read-through transcript is 160 s, identical with the time required in the absence of p factor (Fig. 5A). The simplest interpretation of these results is that transcripts that read-through in the presence of p factor nonetheless pause as do the bulk of transcripts in the absence of p factor, so that p factor does not affect transcriptional pausing across tR,. According to this interpretation, the failure of some transcripts to terminate is thus not due to a failure to pause but instead to a lack of quantitative transcript release. These conclusions are further substantiated by the pattern of t.ranscript accumulation at site I. We observe significant accumulation of transcripts at site I between 60 and 150 s in the presence of both p factor and NusA (Fig. 7B), similar to the pausing pattern exhibited in the presence of NusA alone (Fig. 5). In time, a large portion of this transcript decays, with only a fraction remaining as terminated transcripts. This result clearly shows that p factor does not affect transcriptional pausing at site I in the presence of NusA. Moreover, the inhibition of termination at site I by NusA is not due to curtailed pausing in the presence of both p factor and NusA. This result demonstrates that pausing is not quantitatively proportional to termination. Furthermore, it is likely that the reduction in termination at site I is a consequence of reduced transcript release imposed by NusA.

Effect of NusA on Transcript Release-We investigated the possibility that NusA may affect transcript release directly by observing its effects on transcript release from isolated ternary complexes. We transcribed the X wild type HinfIseo fragment (Fig. 1B) in vitro for 60 s, a time which corresponded to the peak of transcriptional pausing at tR, (Fig. 4). The reaction was arrested by the addition of EDTA to 10 mM, and ternary complexes of polymerase, DNA, and RNA were isolated by chromatography as described under “Experimental Procedures.” To a portion of the isolated ternary complexes, NusA was added to 5 pg/mI and incubation was continued a t 37 “C for 2 min. After the addition of p factor, release of transcript was monitored using a membrane filter binding assay. Another sample was subjected to similar analysis with the 2 min of pre-incubation at 37 “C but without added NusA. Control experiments were carried out in the same way except that p storage buffer was added instead of p factor. Only 10- 20% of the isolated complexes made on the restriction frag-


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

9396 RNA Polymerase Pausing and Transcript Release

ment were retained on membrane filters without the addition of p factor, compared to 70-90% retention when transcripts were made using whole T7 DNA as template (Richardson and Conaway, 1980; Shigesada and Wu, 1980). We assume that the fraction retained on membrane filters contains a repre- sentative distribution of extents of transcription.

Fig. 8 shows that NusA alters the kinetics of p-dependent transcript release a t t ~ ] . The most substantial effect occurs within the first 60 s, during which transcript release in the presence of NusA is slower than in its absence. Over a longer period of time, NusA does not alter the total amount of transcript released; a total of 70-80% of transcripts is released with or without NusA. This result is sufficient to explain the NusA inhibition of termination at site I. In the presence of NusA, pausing at site I occurs with an apparent half-life of 0.7 min (Fig. 5). During this period, transcript release is slow and, thus, the total terminated transcript at site I is curtailed. Since the apparent half-lives of pauses a t sites I1 and 111 are considerably longer than 60 s, the lowered initial rate of release does not affect the total amount of transcripts terminated at these sites.

We observed kinetics of transcript release very similar to that shown in Fig. 8B using ternary complexes isolated from transcription reaction mixtures which contained NusA during RNA synthesis (data not shown). Thus, it appears that NusA has the same effect on transcript release whether it can associate with the ternary complex only at the pause or is present from the beginning of transcription. We have not determined if transcripts from certain sites are selectively released a t different times in the presence or absence of NusA.

Further Discussion-We have shown that RNA polymerase pauses at the three p-dependent termination sites at X t K l in the absence of p factor. However, termination efficiency is not proportional to the duration of the pause; the probability of transcript release at a site must also play a role. The efficiency of a terminator may thus depend upon both the kinetics of transcript release at that site and the duration of the transcriptional pause.

We thus envision that polymerase pausing and p-dependent transcript release are two separate sequence-dependent events. It is clear that not all pause sites are also transcript release sites. On the other hand, since the average rate of transcription elongation across a distance corresponding to release site (usually about 5 nucleotides) is considerably faster than the rate of transcript release (see Fig. 81, p-dependent

SECONDS

FIG. 8. Kinetics of transcript release from isolated ternary complexes. Transcription was carried out using the X wild type HinfISm fragment as template. Transcription was arrested after 60 s a t 37 "C, and terhary complexes were isolated. The time course of transcript release from isolated ternary complexes was monitored by a membrane filter assay in the presence ( A ) or absence ( B ) of NusA protein (5 pg/ml). 0, RNA released in the presence of p factor; 0, RNA released in the absence of p factor.

transcript release sites could be manifested only if they are also pause sites. An efficient termination site must signal polymerase to pause for a duration compatible with the kinetics of transcript release at that site. The sequences signal- ing pausing and transcript release are not understood, although these two functions are separated by the cnc mutation.

Besides their varying abilities to induced pausing and p- dependent transcript release, the several wild type and mutant termination sites at A t H l also respond differently to the NusA protein. NusA generally both prolongs transcriptional pausing and lowers the initial rate of transcript release from pause sites. Therefore, depending on the extent of the pause and the rate of release dictated by NusA, it can potentially increase or decrease termination efficiency. Although p-dependent termination does not require NusA, it may have an important regulatory role in the control of termination.

We have shown that the cnc mutation abolishes the NusA- dependent enhancement in pausing as well as p-dependent transcript release at site I1 of t R 1 . This observation might provide an important clue to the biochemical functions of both NusA and p factor. An attractive interpretation is that cnc changes the response of polymerase to this region of DNA in a way that simultaneously prevents these different activities of p factor and NusA.

Acknowledgments-We thank J. Tso for patient help in taking time points, and Fran Ormsbee for typing the manuscript.

REFERENCES

Adhya, S., and Gottesman, M. (1978) Annu. Reu. Biochem. 47, 967-

Burgess, R. R., and Jendrisak, J. J. (1975) Biochemistry 14, 4634-

Calva, E., and Burgess, R. R. (1980) J . Biol. Chem. 255,11017-11022 Court, D., Brady, C., Rosenberg, M., Wulff, D. L., Behr, M., Mahoney,

Farnham, P. J., Greenblatt, J., and Platt, T. (1982) Cell 29, 945-951 Finger, L. R., and Richardson, J. P. (1982) Biochemistry 20, 1640-

1645 Friedman, D. I. (1971) in The Bacteriophage Lambda (Hershey, A.

D., ed) pp. 733-738, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

996

4638

M., and Izumi, S. (1980) J. Mol. Biol. 138, 231-254

Greenblatt, .I. (1981) Cell 24, 8-9 Greenblatt, J . , and Li, J. (1981) Cell 24, 421-428 Greenblatt, J., Li, J., Adhya, S., Friedman, D. I., Baron, L. S.,

Redfield, B., Kung, H.-F., and Weissbach, H. (1980) Proc. Natl. Acad. Sci. U. S . A. 77, 1991-1994

Greenhlatt, J., McLimont, M., and Hanly, S. (1981) Nature (Lond.)

Herskowitz, I., and Hagen, D. (1980) Annu. Reo. Genet. 14,399-445 Kassavetis, G. A,, and Chamberlin, M. J. (1981) J . Biol. Chem. 256,

Kingston, R. E., and Chamberlin, M. J. (1981) Cell 27, 523-531 Kingston, R. E., Nierman, W. C., and Chamberlin, M. J. (1981) J.

Kung, H., Spears, C., and Weissbach, H. (1975) J. Bid. Chem. 250,

Lau, L. F., Roberts, J. W., and Wu, R. (1982) Proc. Natl. Acad. Sci.

Lowe. P. A,, Hazer, D. A., and Burgess, R. R. (1979) Biochemistry

292, 215-220

2777-2786

Biol. Chem. 256, 2787-2797

1556-1562

U, S. A . 79, 6171-6175

18; 1344-1352 Martin. F.. and Tinoco. I. (1980) Nucleic Acids Res. 8, 2295-2300 McDermit; M., Pierce,' M., Staley, D., Shimaji, M., Shaw, R., and

Platt, T. (1981) Cell 24, 10-23 Pribnow, D. (1979) in Biological Regulation and Deuelopment (Gold-

berger, R. F., ed) Vol. 1, pp. 219-277, Plenum Press, New York Richardson, J . P., and Conaway, R. (1980) Biochemistry 19, 4293-

4299 Roberts, ,J. W. (1969) Nature (Lond.) 224, 1168-1174 Roberts, d . W. (1976) in R N A Polymerase (Losick, R., and Chamber-

lin, M., eds) pp. 247-271, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Wulff, D. L. (1976) Genetics 82, 417-422


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

RNA Polymerase Pausing and Transcript Release 9397

Rosenberg, M., and Court, D. (1979) Annu. Rev. Genet. 13,319-353 (Lond.) 272, 410-414 Rosenberg, M., Court, D., Shimatake, H., Brady, C., and Wulff, D. L. Shigesada, K., and Wu, C.-W. (1980) Nucleic Acids Res. 6,3543-3557

(1978) Nature (Lond.) 272, 414-422 Ward, D. F., and Gottesman, M. E. (1982) Science (Wash. D. C.) Schleif, R. F., and Wensink, P. C. (1981) Practical Methods in 216, 946-951

Molecular Biology, pp. 27-34, Springer-Verlag, New York Wulff, D. L. (1976) Genetics 82, 401-416 Schwarz, E., Scherer, G., Hobom, G., and Kossel, H. (1978) Nature Yanofsky, C. (1981) Nature (Lond.) 289, 751-758


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

L F Lau, J W Roberts and R Wuvitro.

RNA polymerase pausing and transcript release at the lambda tR1 terminator in

1983, 258:9391-9397.J. Biol. Chem.

http://www.jbc.org/content/258/15/9391Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/258/15/9391.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/258/15/9391

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;258/15/9391&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/258/15/9391

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=258/15/9391&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/258/15/9391

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/258/15/9391.full.html#ref-list-1

http://www.jbc.org/

RNA Polymerase Pausing and Transcript Release at the AtRl ...

Documents

Transcript of RNA Polymerase Pausing and Transcript Release at the AtRl ...