INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

The chromosomes of Escherichia coli and Bacillus subtilis, and some plasmids like R6K, initiate replication from specific origins and the replication forks moving either bi- or unidirectionally, are arrested at sequence-specific replication termini (1-3). Sequence-specific replication fork barriers have also been observed in the nontranscribed spacer regions of ribosomal DNAs of Saccharomyces cerevisiae (4, 5), plants (6), Xenopus (7), and human (8).

The replication termini are sequence-specific and bind to cognate terminator proteins, and the protein-DNA complex arrests replication forks in an orientation-dependent manner (9, 10). The terminator protein of E. coli, called Tus, is a 36-kDa monomer that binds to the ~22-bp¹ Ter () sequence and arrests replication fork by antagonizing helicase-catalyzed DNA unwinding in a polar mode (11-13). The replication terminator protein (RTP) of B. subtilis, in contrast, is a homodimer of 14.5-kDa monomers that binds cooperatively as two interacting dimers to the cognate Ter site of ~30 bp that comprises an overlapping core and an auxiliary sequence (14-17). The most frequently used terminus of B. subtilis is Ter I (also called BS3; Ref. 18), which binds to two interacting dimers of RTP and arrests replication forks in a polar mode (14-16).

Since a usable in vitro replication system for B. subtilis is yet to be worked out, we have developed an E. coli-based in vitro system to analyze the mechanism of action of RTP (19). The RTP-BS3 complex of B. subtilis has been independently shown to block replication fork of E. coli in vivo in an orientation-dependent manner (20). Both Tus and RTP block the replication fork of E. coli by impeding DNA unwinding by the replicative helicase DnaB (11, 12, 19). Detailed biochemical and biophysical analyses have been made possible in both the systems by the determination of the crystal structures of RTP at 2.6 Å (21) and of Tus-Ter () complex at 2.7 Å (22). The DNA binding (23), dimer-dimer interaction (24), and DnaB interaction domains (25) of RTP have been determined by genetic and biochemical analyses that used the crystal structure as a guide. Similarly, the DNA binding domain of Tus has been determined with the help of crystal structure and genetic analysis (22).

The impact of transcription on the initiation (26, 27) and elongation stages of DNA replication have been reported (28, 29). We have reported previously that Tus and RTP can block RNA chain elongation catalyzed by several prokaryotic RNA polymerases in a polar mode and that the passage of an RNA transcript that invades the terminus from the permissive direction causes functional inactivation of the replication terminus (30). We have suggested that the polar arrest of transcription protects the terminus from functional inactivation by transcriptional invasion (30, 31). The interaction between transcription and termination of DNA replication is of further interest because the replication check points of B. subtilis that are conditionally active under stringent conditions (32) appear to be controlled by transcription.²

In the present work, we endeavored to address four important questions regarding the interplay between transcription and replication termination. (i) What is the eventual fate of RNA polymerase (RNAP) when it reaches the Tus-Ter () complex that is positioned in blocking orientation? Is the RNAP arrested or does it dissociate from the DNA leading to authentic termination of transcription? (ii) Does the passage of transcription through the replication terminus cause the dissociation of the Tus protein from the DNA-protein complex thus abrogating replication fork arrest or is it caused by a transcription-mediated conformational change of the Tus-Ter () complex that does not involve removal of Tus? (iii) Are the DNA and RNA chains blocked at the same site of the terminus? (iv) Are eukaryotic RNAPs blocked by prokaryotic replication termination proteins, and, conversely, can eukaryotic transcription termination proteins block replicative helicases and a prokaryotic RNAP? In this report, we present a series of in vivo and in vitro experiments that address these questions. The results show that replication terminator proteins arrested RNAPs but did not cause transcription termination. Transcriptional invasion caused the terminator proteins to dissociate from the terminus, thus explaining the mechanism of abrogation of replication fork arrest. The transcription and replication arrest sites are different. Although the replication terminator proteins blocked eukaryotic RNAPs, a eukaryotic transcription terminator protein failed to block T7 RNAP and E. coli DnaB helicase.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

DNA-- For experiments on arrest of T7 RNAP by Tus, the plasmids pET22b- and pET22b- rev. were constructed by cloning a 177-bp HindIII fragment of DNA containing the Ter () site in either orientation at the HindIII site of pET22b(+) plasmid (Novagen). pET22b-BS3 and pET22b-BS3 rev. plasmids (used in experiments to investigate T7 RNAP blockage by RTP) and T7A1-BS3 and T7A1-BS3 rev. DNA substrates (used in experiments to investigate E. coli RNAP blockage by RTP) have been described previously (30). For experiments on blockage of E. coli RNAP by Tus, the Ter () site was cloned in either orientation downstream of a T7A1 promoter to generate the DNA substrates T7A1- and T7A1- rev. DNA fragment containing the T7A1- was obtained by polymerase chain reaction (PCR) with a primer upstream of the T7A1 promoter (A1 primer, 5-AGGAGAGACTTAAAGAG-3) and the M13 universal primer or the A1 primer and a primer 5-TATTAACCACTTTAGTTACAACATACTTATTTTA-3 that overlaps the Ter () site. Similarly, a DNA fragment containing the T7A1- rev. was obtained by PCR with the A1 primer and M13 universal primer or the A1 primer and a primer 5-TAAAATAAGTATGTTGTAACTAAAGTGGTTA ATA-3 that overlaps the Ter () site. PCR was carried out for 30 cycles in each case with each cycle consisting of three steps: 94 °C for 30 s, 55 °C for 60 s, and 72 °C for 120 s. For experiments with Reb1p, the Reb1 binding site (33) from the plasmid p6.50 (from Dr. Walter Lang, University of Washington, Seattle, WA) was cloned as a HindIII-SacI or HindIII-NotI fragment into pET22b plasmid to generate templates with the blocking and the nonblocking orientations of Reb1p binding site, respectively.

Proteins-- Purification of Tus, RTP, and T7 RNAP has been described previously (30, 31). E. coli RNA polymerase containing 6-histidine-tagged -subunit was purified from the strain AG2005/-6His (kindly provided by Dr. Alex Goldfarb, Public Health Research Institute, New York, NY) according to Kashlev et al. (34), with some modifications. The fractions from Ni⁺-agarose column that contained the RNA polymerase were pooled and fractionated further on a 1-ml MonoQ (Pharmacia Biotech Inc.) column (35). The RNA polymerase was eluted with a 0- 0.6 M NaCl gradient in buffer containing 40 mM Tris·HCl, pH 7.9, 5% glycerol, 0.1 mM EDTA, 1 mM -mercaptoethanol). Peak fractions were pooled and used for transcription. For some experiments, we have also used E. coli RNA polymerase that was purified as described (30). Reb1 protein was purified according to published procedures (33) with some modifications. Yeast RNA polymerase I and II were kind gifts from Dr. Walter Lang. S1 nuclease was purchased from U. S. Biochemicals. EcoRI E111Q mutant protein was a kind gift from Dr. Paul Modrich (Duke University, Durham, NC).

Transcription Impedance-- Multiple-round transcription reactions were carried out as described previously (30). Single-round transcription reactions in the presence of rifampin were carried out as described below in the appropriate sections.

Fate of E. coli RNAP Blocked by E111Q Mutant Form of EcoRI Restriction Endonuclease, Tus, and RTP-- To study the fate of E. coli RNAP when it is blocked by E111Q mutant form of EcoRI (that binds to DNA site with a very long half-life but does not cleave DNA), Tus or RTP single-round transcription reaction was carried out essentially as described below. 50 fmol of the DNA fragment were incubated with 5-10-fold molar excess of E111Q protein or 10-20-fold RTP or Tus at 37 °C for 5 min in transcription buffer containing 30 mM Tris·HCl, pH 7.6, 40 mM KCl, 0.1 mM EDTA, 1 mM DTT, 50 µg/ml bovine serum albumin, 100 µM each of the four NTPs, and 10 µCi of [-³²P]GTP. 100 fmol of E. coli RNA polymerase (2-fold molar excess over DNA) was added and incubated at 37 °C for 5 min. Transcription was initiated by adding MgCl₂ to 2 mM, followed by rifampin to 100 µg/ml and heparin to 100 µg/ml. The reaction was carried out at 37 °C for 15 min. The transcription mix was then divided into two equal halves. To one half, KCl was added (in 1 × transcription buffer containing 2 mM MgCl₂) to a final concentration of 0.5 or 0.7 M KCl. To the other half, 1 × buffer with 2 mM MgCl₂ was added. The reaction was continued for another 30 min at 37 °C. 2 units of RNase-free DNase I (Pharmacia) was added, and incubation was continued at 37 °C for 15 min. The RNA was precipitated and run in a 6% polyacrylamide, 7 M urea gel.

Transcription and Gel Shift Assay to Determine the Fate of RNAP-- To determine the fate of E. coli RNAP when it is blocked by Tus, a gel mobility shift assay was carried out. A single-round transcription reaction was set up in the presence of Tus as described above. After the addition of RNAP, MgCl₂, rifampin, and heparin, incubation was carried out at 37 °C for 30 min, followed by addition of 5 µl of the loading dye containing 25% glycerol with bromphenol blue and xylene cyanol dyes (to give a final concentration of 5% glycerol) and was resolved in a 5% polyacrylamide gel.

Analysis of Tus-Ter () Complex by Gel Shift-- The binding of Tus to ³²P-labeled transcription template DNA and the effect of salt on the binding were analyzed by nondenaturing polyacrylamide gel electrophoresis. The steps of a standard single-round transcription reaction were followed except that RNAP dilution buffer was added instead of RNAP. After incubation of the DNA with Tus followed by incubation with RNAP dilution buffer, MgCl₂, rifampin, and heparin were added and further incubation was carried out for 15 min at 37 °C. Then, KCl in 1 × transcription buffer or 1 × buffer alone was added to the appropriate tube, and incubation was continued for another 30 min at 37 °C. Sample dye was added to give a final concentration of 5% glycerol, and the samples were run in a 5% polyacrylamide gel.

Analysis of Fate of the Terminator Protein in the Nonblocking Orientation-- 100 fmol of the pET22b- or pET22b- rev. plasmid linearized at the BlpI site was incubated with 65 fmol of Tus protein at room temperature for 15 min. 0.5-1.0 pmol of T7 RNAP was added and incubated for 1 min, following which 12 fmol of an end-labeled (by [-³²P]ATP) 177-bp Ter () fragment was added and incubation was continued for 30 min at 37 °C. The reaction mix was run in a 5% polyacrylamide gel to monitor trapping of the Tus protein displaced from transcription template by the ³²P-labeled Ter fragment.

Mapping of the in Vivo Leading Strand Block Site-- E. coli Tus⁺ strain TH423 (a kind gift of Dr. T. M. Hill, University of North Dakota School of Medicine and Health Sciences, Grand Forks, ND) containing the plasmid pUC18- with the Ter () site in replication arresting orientation was grown and replication intermediates were isolated from it as described previously (30). The plasmid was run in a 0.8% agarose gel without ethidium bromide. The replication intermediate, which ran slower than the fully replicated plasmid was eluted from the gel, ethanol-precipitated, and dissolved in TE (10 mM Tris·HCl, 1 mM EDTA, pH 8.0). The DNA was digested with HindIII, dephosphorylated with shrimp alkaline phosphatase (U.S. Biochemical Corp.), phenol-extracted, precipitated with ethanol, and dissolved in TE. The dephosphorylated DNA was end-labeled with [-³²P]ATP by T4 polynucleotide kinase, passed through Sephadex G25 spin column, precipitated with ethanol, and dissolved in TE. The labeled DNA was run in a 6% polyacrylamide, 7 M urea sequencing gel alongside the pUC18 sequencing ladder.

Mapping of Lagging Strand Block Site-- A unidirectional multiple-round primer extension by PCR was performed with the replication intermediate to determine the lagging strand blockage in vivo. The M13/pUC reverse sequencing primer was annealed to the replication intermediates in which it would anneal to the lagging strand and one parental strand. Extension was carried out by PCR using Vent DNA polymerase (New England Biolabs) for 30 cycles with each cycle having three steps: 94 °C for 30 s, 55 °C for 60 s, and 72 °C for 120 s. The primer extension products were run in a 6% polyacrylamide, 7 M urea gel alongside pUC18 sequencing ladder.

Mapping of Transcription Arrest/Block Sites-- For mapping the 3-ends of the transcripts blocked at the terminator protein binding sites, scaled-up reactions were carried out in 100-µl volumes. After assembly, to 10 µl of the reaction mix [-³²P]GTP was added to monitor transcription impedance by the terminator proteins. The remainder of the sample (90 µl) was used for mapping. Samples with and without [-³²P]GTP were incubated at 37 °C for 1 h, extracted with phenol:chloroform:isoamyl alcohol and chloroform:isoamyl alcohol, and precipitated with ammonium acetate and ethanol as described previously (30). The ³²P-labeled sample was run in a 6% polyacrylamide, 7 M urea gel to confirm transcription blockage. For mapping the 3-end of RNA generated in the experiments with RTP, the DNA probe was generated as below; an NdeI-SacI fragment containing the BS3 site obtained from the plasmid pET22b-BS3 was end-filled by T4 DNA polymerase in the presence of TTP and [-³²P]dATP so that only the NdeI end was labeled. For mapping experiments with Tus, the DNA probe was generated as described below; pET22b- plasmid was digested with NcoI, end-filling was done by T4 DNA polymerase in the presence of dCTP and [-³²P] dATP, the labeled DNA was digested with XhoI, and finally the labeled NcoI-XhoI fragment containing the Ter () site was recovered from a 6% polyacrylamide gel.

Yeast and T7 RNAP Transcription Blockage by Reb1 Protein-- Templates used for run-off transcription were BglII-BlpI fragments from pET22b plasmids containing RebIp binding site in the blocking (at HindIII-SacI sites) or the nonblocking (at HindIII-NotI site) orientation. The ~400-bp BglII-BlpI fragments were modified by ligating a 14-nt oligonucleotide (5-GATCAAAAAACCA-3) to the BglII end by T4 DNA ligase to create a 3 overhang to serve as RNAP entry site. Typical reactions were carried out in a 40-µl volume containing 40 fmol of template DNA, 50 ng of Reb1 protein (500 fmol) in reaction buffer containing 10 mM HEPES-KOH pH 7.9, 10 mM MgCl₂, 25 mM KCl, 2.5 mM EDTA, 20 mM DTT, 10 units of RNAguard (Pharmacia), 1 µl of 20 mM dinucleotide UpG (Sigma), 0.5 mM each of ATP, CTP, and UTP, 0.1 mM GTP, and 20 µCi of [-³²P]GTP. Yeast RNA polymerase I (12 units) or RNAP II (6 units) were added after dilution in the buffer containing 20 mM HEPES-KOH pH 7.9, 50 mM KCl, 5 mM EGTA, 50 mM EDTA, 2.5 mM DTT, 1 mM phenylmethylsulfonyl fluoride, and 0.5 mg/ml leuopeptin. After binding Reb1p or RTP to the corresponding template in the reaction buffer for 15 min at room temperature, yeast RNAP was added and incubation was carried out at 30 °C for 30 min, followed by Dnase I treatment at 37 °C for 30 min. Transcription products were extracted with phenol:chloroform:isoamyl alcohol and precipitated with ethanol in the presence of 10 µg of yeast tRNA. RNA was dissolved in loading buffer (30) and resolved in a 6% polyacrylamide, 7 M urea gel. For transcription by T7 RNA polymerase, all the conditions were same as above except that the transcription buffer was as described earlier for T7 RNA polymerase (30) and nucleotides used were 0.5 mM each of ATP, CTP, and UTP, 12 µM GTP, 20 µCi of [-³²P]GTP, and 100 ng of purified T7 RNAP.

Yeast RNAP I and II Blockage by RTP-- Templates used were ~400-bp BglII-BlpI fragments from the plasmids pET22b-BS3 and pET22b-BS3 rev. to which the 14-nt oligonucleotide described in the previous section was ligated to provide an entry site for the RNAP. The transcription reaction contained 40 fmol of template DNA and 400 or 800 fmol (10-20-fold molar excess) of RTP. Reaction conditions were same as described for transcription blockage experiments with Reb1p.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Orientation-dependent Blockage of T7 and E. coli RNAP by Tus-Ter () Complex-- We have shown previously that RTP impedes RNA chain elongation catalyzed by E. coli and T7 RNAPs in an orientation-dependent manner (30). The same orientation of the RTP-TerI (BS3) complex impedes both RNA and DNA chain elongation. We wished to generalize and extend our previous observation by investigating the effect of Tus-Ter () complex on chain elongation catalyzed by both T7 and the E. coli RNA polymerases. We linearized pET22b- DNA carrying the TerB (see Fig. 1) site of E. coli in the orientation that impedes replication forks (11, 12), and initiated transcription from a T7 promoter by T7 RNAP in the presence of 0-, 2-, 4-, or 8-fold molar excess of Tus over DNA substrate (100 fmol of DNA and 0, ~7, 14, and 28 ng of Tus). A full-length transcript of ~450 nt (arrow showing F) and two very closely spaced truncated transcripts of ~290 nt (arrow showing T) were generated (Fig. 2A, lanes 1-4). Transcription of the template pET22b- rev. that contained the Ter () site in the reverse orientation under identical conditions did not show impedance of the transcription, and thus only full-length run-off transcripts were observed even if an 8-fold molar excess of Tus was used (Fig. 2A, lanes 6-8). Transcription from a control DNA fragment that also had a T7 promoter but no downstream Ter () site generated only full-length run-off transcripts (marked C in Fig. 2A). Thus, the Tus-Ter () complex blocked up to a maximum of 60% of the transcripts in the replication blocking orientation.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Replication terminus region of E. coli and B. subtilis chromosomes. Each chromosome has six terminator sites with three sites facing each other. Each site in E. coli called Ter () binds a monomer of Tus protein. In B. subtilis, the sites are called Ter, IR, or BS3 and bind two interacting dimers of RTP each. The replication fork coming from the origin (shown by arrow) will pass through the first set and can be blocked by any site of the second set. In B. subtilis, the fork is generally blocked at IRI or TerI. Although the distance between TerA and TerC is 270 kilobase pairs in E. coli, that between TerI and TerII in B. subtilis is only 59 base pairs.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Blockage of T7 and E. coli RNA polymerases by E. coli replication terminator protein Tus. Top, map of the BglII-BlpI part of the pET22b- and pET22b- rev. plasmids used for transcription in bottom (A). Transcription from the T7 promoter in either substrate in the absence of Tus will generate a transcript of ~450 nt. In the presence of Tus, a truncated transcript of ~290 nt will be generated from pET22b-, whereas from pET22b- rev. no truncated transcript will be formed. Bottom, A, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing orientation-dependent blockage of T7 RNA polymerase by Tus protein. Lanes 1-4, transcription of pET22b- DNA substrate (linearized with BlpI) containing Ter () site in the blocking orientation in the presence of 0-, 2-, 4-, and 8-fold molar excess of Tus protein (over 100 fmol of DNA); lanes 5-8, transcription of pET22b- rev. DNA (linearized with BlpI) containing the Ter () site in the nonblocking orientation in the presence of 0-, 2-, 4-, and 8-fold molar excess of Tus protein (over 100 fmol of DNA). All lanes contain an internal control of 100 fmol of the plasmid pET22b linearized at EcoRI site. Note that Tus blocks T7 RNA polymerase transcription only in the replication blocking orientation (lanes 2-4). Bottom, B, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing blockage of E. coli RNA polymerase by Tus protein. Lanes 1-4, transcription of T7A1- template containing the Ter () site in the blocking orientation in the presence of 0-, 10-, 20-, and 40-fold molar excess of Tus protein; lanes 5-8, transcription of T7A1- rev. template in the presence of 0-, 10-, 20-, and 40-fold molar excess of Tus protein. Note that Tus blocks E. coli RNA polymerase only in the replication blocking orientation (lanes 2-4).

E. coli RNAP Is Arrested but Not Terminated by Tus-Ter () and RTP-BS3 Complexes-- The observed impedance of RNAP by the Tus-Ter () complex could be explained as (i) a transient pause, (ii) arrest of the transcripts for a longer period without the dissociation of the RNAP from the DNA, or (iii) an authentic termination of transcription caused by the dissociation of RNAP from the DNA and release of the transcripts. We wished to distinguish among these possibilities as follows, using two different approaches. The first approach was based on the assumption that, if RNA chain elongation by RNAP is arrested but not terminated at the terminator protein-DNA complex, a truncated transcript should be generated and removal of the terminator protein from its binding site should allow the RNAP to resume transcription leading to the extensions of the truncated products into full-length transcripts. We decided to use E111Q mutant form of EcoRI restriction endonuclease as a positive control, with which we compared the behavior of Tus. Other investigators have shown that E111Q binds to DNA with a relative K_d of 10¹⁵ mol/liter without cleaving the DNA (36). The bound E111Q protein arrests but does not terminate transcription initiated from an upstream promoter (37). It has been shown that high salt concentrations do not disrupt a DNA-RNAP-RNA ternary complex or inhibit an elongating E. coli RNA polymerase from template DNA (38). The experimental strategy consisted of initiation of transcription from the designated promoter by E. coli RNAP and blockage of the transcripts by E111Q, Tus, or RTP. The blocking protein was then dissociated from the template by adding KCl to 0.5 or 0.7 M, and the RNA transcript was resolved on denaturing polyacrylamide gels. Fig. 3B shows the fate of E. coli RNAP blocked by E111Q and RTP. Transcription was initiated from the T7A1 promoter (that binds E. coli RNAP) downstream of which was positioned a BS3 sequence (TerI site of B. subtilis; see Fig. 1) flanked by two EcoRI sites. Fifteen minutes after transcription was initiated in the absence or the presence of E111Q or RTP proteins bound to the cognate sites, the reaction mixture was divided into two equal halves and KCl was added to 0.5 or 0.7 M to one half of the reaction mixture, whereas to the other half only buffer was added and the reaction was continued for another 30 min. Transcription by E. coli RNAP in the presence of E111Q protein generated two truncated transcripts in addition to the full-length transcript (Fig. 3B, lane 3). In the presence of RTP bound to the single BS3 site, a truncated and a full-length transcript were generated (lanes 5 and 7). Addition of 0.5 or 0.7 M KCl caused the truncated transcripts to be almost quantitatively extended into full-length transcripts (lanes 4, 6, and 8). Presence of high salt did not diminish the amount of transcription by RNA polymerase (Fig. 3B, compare lane 1 with lane 2).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Fate of E. coli RNA polymerase blocked by E111Q, Tus and RTP. A, experimental strategy to determine the fate of RNA polymerase. B, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing arrest of E. coli RNAP blocked by 10-fold molar excess of E111Q or 20-fold molar excess RTP (over DNA substrate). The substrate had the T7A1 promoter and a downstream BS3 site. The two EcoRI sites downstream of the promoter flank the BS3 site. Lane 1, DNA + RNAP; lane 2, DNA + RNAP + 0.7 M KCl; lane 3, DNA + E111Q + RNAP; lane 4, DNA + E111Q + RNAP + 0.7 M KCl; lane 5, DNA + RTP + RNAP; lane 6, DNA + RTP + RNAP + 0.5 M KCl; lane 7, DNA + RTP + RNAP; lane 8, DNA + RTP + RNAP + 0.7 M KCl. C, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing arrest of E. coli RNA polymerase by E111Q and Tus proteins. The substrate T7A1- has the T7A1 promoter downstream of which are two EcoRI sites followed by the Ter () site. Lane 1, DNA + RNAP; lane 2, DNA + RNAP + 0.7 M KCl; lane 3, DNA + E111Q + RNAP; lane 4, DNA + E111Q + RNAP +0.7 M KCl; lane 5, DNA + Tus + RNAP; lane 6, DNA + Tus + RNAP + 0.5 M KCl; lane 7, DNA + Tus + RNAP; lane 8, DNA + Tus + RNAP + 0.7 M KCl. D, autoradiogram of a 5% polyacrylamide gel showing that salt removes Tus from Ter () site. Gel shift was done with 32P-labeled T7A1- DNA fragment used in the transcription experiments. Lane 1, DNA alone; lane 2, DNA + Tus; lane 3, DNA + Tus + 0.7 M KCl.

Arrest of E. coli RNA Polymerase at the Replication Terminus as Shown by Gel Mobility Shift Assay-- To confirm the arrest of E. coli RNAP by Tus-Ter () complex as described above, we performed a gel mobility shift assay of the transcribing RNA polymerase (Fig. 4). The experimental strategy is shown in Fig. 4 (top). A single-round transcription of the templates containing the T7A1 promoter and the downstream Ter () site, present in either orientation, was carried out and analyzed as described under experimental procedures. In the template having the Ter () site in the blocking orientation, the transcribing RNAP arrested by Tus showed a mobility shift of the DNA-Tus-RNAP-RNA complex (shown by arrow in lane 5, Fig. 4, bottom panel). However, in the absence of Tus, there was no corresponding gel shift (lane 6, Fig. 4, bottom panel), suggesting that all RNAPs had completed transcription and had fallen off the template. Rebinding of the enzyme to the template was prevented because of the presence of rifampin. In comparison, the template with the Ter () site in the nonblocking orientation did not show any gel shift of the transcribing RNA polymerase in the presence or absence of Tus (compare lanes 5 and 6 with lanes 11 and 12, respectively, Fig. 4, bottom panel). The control experiments included gel shifts performed in the absence of Tus (lanes 1 and 7), in the presence of Tus (lanes 2 and 8), in the presence RNA polymerase (lanes 3 and 9), or in the presence of both Tus and RNA polymerase (lanes 4 and 10). Gel shifts with RNA polymerase alone or with RNA polymerase and Tus in the control lanes were done without MgCl₂, NTPs, and rifampin to prevent the occurrence of transcription (lanes 3 and 4 and lanes 9 and 10). The gel shift caused by RNA polymerase without transcription was negligible probably because of the instability of the RNA polymerase-promoter complex. Thus, these experiments confirmed the observation that RNA polymerase is arrested at the Tus-Ter () complex but does not dissociate from the DNA.

View larger version (49K): [in this window] [in a new window]   Fig. 4.   Fate of E. coli RNA polymerase blocked by Tus-Ter () complex as shown by gel shift. Top, experimental strategy. Bottom, autoradiogram of a 5% polyacrylamide gel showing results of gel shift assay to determine the fate of E. coli RNA polymerase blocked by Tus protein. After single-round transcription, reactions were carried out for 30 min, the samples were loaded on a the polyacrylamide gel. Lanes 1-6, experiments with T7A1- DNA ( or Ter site in the blocking orientation). Lane 1, DNA; lane 2, DNA + Tus; lane 3, DNA + RNAP without transcription; lane 4, DNA + Tus + RNAP without transcription; lane 5, DNA + Tus + RNAP with transcription; lane 6, DNA +RNAP with transcription. Lanes 7-12, experiments with T7A1- rev. DNA ( or Ter site in the nonblocking orientation). Lane 7, DNA; lane 8, DNA + Tus; lane 9, DNA + RNAP without transcription; lane 10, DNA + Tus + RNAP without transcription; lane 11, DNA + Tus + RNAP with transcription; lane 12, DNA + RNAP with transcription. Note that lane 5 shows arrest of RNAP by Tus (the thick extra band shown by arrow).

Fate of Tus Protein When RNAP Invades Tus-Ter () Complex from the Nonblocking End-- We have reported previously that passage of an RNA transcript through the replication terminus causes the release of the arrested replicative helicase and the replication fork (30, 31). There are at least two possible mechanisms that could explain the functional inactivation of the replication terminus by transcriptional invasion: (i) the passage of the transcript could alter the conformation of the DNA-protein complex at the replication terminus without dissociating the bound form of Tus (RTP) or (ii) it could dissociate the Tus (RTP) from the cognate DNA sequences. We wished to distinguish between these two possibilities by performing the following experiments. The experimental design is shown in Fig. 5 (top panel). A ³²P-labeled, 177-bp-long DNA fragment containing the Ter () site was used to trap the Tus protein that might be dislodged from the pET22b- and pET22b- rev. templates during transcription by T7 RNAP. Templates containing Tus-Ter () complex (with protein:DNA ratio of 0.65:1.0, so that there was little or no unbound protein) were transcribed by T7 RNAP. When pET22b- rev. was transcribed, the bound Tus dissociated from the complex and trapped by the 177-bp labeled fragment showing a gel shift (Fig. 5, bottom panel, lane 8). Transcription of pET22b- did not show a gel shift, thus suggesting that RNAP was arrested at the Tus-Ter () complex and thus could not dissociate the bound Tus from the Ter () site. Addition of only RNAP or NTPs alone did not show gel shifts either with pET22b- rev. (Fig. 5, bottom, lanes 6 and 7) or with pET22b- (Fig. 5, bottom, lane 1 and 2) templates. The 177-bp fragment bound to Tus alone showed gel shift (Fig. 5, bottom, lanes 4 and 5). No gel shift was observed when the same 177-bp fragment (that lacked a promoter) was incubated with T7 RNAP (Fig. 5, bottom, lane 9). These experiments support the conclusion that passage of an RNA transcript through the Tus-Ter () complex from the permissive end causes the dissociation of the Tus protein from the complex, thus explaining the mechanism of transcriptional inactivation of replication terminus.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Fate of Tus when T7 RNA polymerase transcribes through the Tus-Ter () complex from the permissive end. Top, experimental strategy. Bottom, autoradiogram of a 5% polyacrylamide gel showing gel shift of a 32P-labeled Ter () fragment (hereafter called  DNA) when Tus is dislodged by T7 RNAP transcribing through Tus-Ter () complex from the nonblocking orientation. Lane 1, pET- + Tus + NTPs +  DNA; lane 2, pET22b- + Tus + RNAP +  DNA; lane 3, pET22b- + Tus + NTPs + RNAP +  DNA; lane 4,  DNA; lane 5;  DNA + Tus; lane 6, pET22b- rev. + Tus + NTPs +  DNA; lane 7, pET22b- rev. + Tus + RNAP +  DNA; lane 8, pET22b- rev. + Tus + NTPs + RNAP +  DNA; lane 9,  DNA + RNAP. Note the gel shift of the Ter () fragment only when transcription goes through the Ter () site in the nonblocking orientation (lane 8) or when only  DNA plus Tus are present (lane 5).

Mapping of in Vivo Arrest Sites for Leading and Lagging DNA Strand in E. coli-- Since the replication terminator proteins arrest both replication and transcription, we wanted to compare the sites of replication and transcription arrest with the hope that this information might contribute to the understanding of the mechanism of arrest. The experimental strategy for mapping of the arrest sites of the leading and the lagging strand is schematically shown in Fig. 6A. We reasoned that the prolonged arrest of the replication fork would allow the processing and ligation of the Okazaki pieces, thus generating a continuous stretch of the lagging strand of the unidirectionally replicating pUC18- template. The location of the point of arrest of the lagging strand can be determined by unidirectional primer extension by PCR using an appropriately chosen end-labeled primer.

View larger version (42K): [in this window] [in a new window]   Fig. 6.   Determination of leading strand and lagging strand block sites in vivo. A, partial map of pUC18- plasmid showing experimental strategy for mapping the leading and lagging strand block sites. Replication intermediates of pUC18- have DNA replication blocked at the Ter () site. The DNA digested with HindIII releases a 177-bp fragment of the parental DNA substrate. Leading strand and lagging strand downstream of the left HindIII site up to the stop site will be less than 177 bases. A primer complementary to the lagging strand was designed to perform a unidirectional PCR to determine the lagging strand stop site. The 5 end of this stop site was expected to be about 60-100 bp before the leading strand stop site (39). B, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing the lagging strand stop site. A, C, G, and T represent the pUC18 sequencing ladder. Lane 1, pUC18- from Tus+ strain; lane 2, pUC18- from Tus strain; lane 3, pUC18- rev. from Tus+ strain; lane 4; pUC18- rev. from Tus strain.  represents pUC18-, and  rev. represents pUC18- rev. clones. C, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing the leading strand stop site. A, C, G, and T represent pUC18 sequencing ladder. Lanes 1 and 2 show the 177-base HindIII Ter () fragment and the major and minor block sites. The top arrow represents the leading strand major block site, and bottom arrow represents the leading strand minor block site. The rest of the plasmid and the leading and lagging strands from origin to the left HindIII site, which ran almost at the top of the gel, are not shown. The portion of the lagging strand from the left HindIII site to the block site, being small, ran at the very bottom of the gel and is not seen.

Mapping of Transcription Arrest/Block Sites and Comparison with the Sites of Arrest of DNA Strands-- The 3 ends of the RNAs generated by transcription blockage experiments were mapped by S1 nuclease mapping technique using ³²P-labeled DNA probes. Fig. 7 shows S1 nuclease mapping data on transcription stop sites of E. coli RNAP and T7 RNAP at Tus-Ter () and RTP-BS3 complexes, respectively. A comparison of the E. coli and T7 RNAP block sites and leading and lagging strand block sites is shown in Fig. 7E (top panel). The data showed that both the RNA polymerases and DNA polymerase (for leading strand) stop at different positions from the terminator site though not very far from one another. As expected, the 5 end of the last Okazaki fragment was 63-65 nt upstream of leading strand arrest site (39).

View larger version (56K): [in this window] [in a new window]   Fig. 7.   RNA polymerase arrest/block sites at Tus-Ter () and RTP-BS3 complexes and their comparison with replication fork arrest sites. A-D, autoradiogram of 6% polyacrylamide, 7 M urea gels showing S1 nuclease mapping results for RNA polymerase stop sites (arrows in all panels). A, E. coli RNAP stop site at Tus-Ter () complex. A, C, G, and T represent pUC18 sequencing ladder. Lane 1, DNA probe used for mapping; lanes 2 and 3 S1 mapped product. B, T7 RNA polymerase stop site at Tus-Ter () complex. Lane 1, DNA probe used for mapping; lane 2, S1 mapped product. A, C, G, and T represent pUC18 sequencing ladder. C, E. coli RNA polymerase stop site at RTP-BS3 complex. A, C, G, and T represent sequencing ladder; lane 1, DNA probe used for mapping; lane 2, S1 nuclease mapping product. D, T7 RNA polymerase stop site at RTP-BS3 complex. A, C, G, and T represent sequencing ladder; lane 1, DNA probe used for mapping; lane 2, S1 mapping product. E, sequence of DNA fragments used, showing the replication and transcription block sites at Tus-Ter () (top) and RTP-BS3 complex (bottom).

RTP-BS3 Complex Blocks Yeast RNAP I and II, but Reb1p Does Not Block T7 RNA Polymerase and DnaB-- Replication terminator proteins are known to arrest several replicative helicases but not those involved in rolling circle replication, conjugational transfer, and DNA repair (11, 12, 19, 31). RNAPs of yeast and mammalian cells are blocked by transcription terminator proteins that bind to specific DNA sequences (33, 40). We were curious to test the specificity of blocks mediated by the yeast Reb1 protein since the topic is relevant to the possible mechanism of polar arrest of DNA and RNA chain elongation. We wished to test if the yeast transcription terminator Reb1p, which blocks all the three RNAPs of yeast (33), could also block a prokaryotic phage RNAP, and if RTP, which impedes prokaryotic and phage RNAP, could also block eukaryotic RNAP. Fig. 8A shows that, whereas Reb1p, as reported previously (33), blocked yeast RNA polymerase II in an orientation-dependent manner (Fig. 8A, lanes 1-4), it could not block T7 RNAP in any orientation (Fig. 8A, lanes 5-8). We have also observed that Reb1p did not block DnaB replicative helicase of E. coli (data not shown). Fig. 8B shows that yeast RNA polymerase II was blocked by RTP in an orientation-dependent manner (Fig. 8B, compare lanes 1-3 with lanes 4-6). Similarly, RTP was also found to block yeast RNA polymerase I in an orientation-dependent manner (Fig. 8C, compare lanes 1 and 2 with lanes 3 and 4). In summary, although Reb1p could not block either T7 RNA polymerase or DnaB helicase, RTP was capable of blocking both yeast RNAP I and II.

View larger version (33K): [in this window] [in a new window]   Fig. 8.   Blockage of yeast RNA polymerases by Reb1p and RTP. A, autoradiogram of 6% polyacrylamide, 7 M urea gel showing blockage of yeast RNA polymerase II and lack of blockage of T7 RNA polymerase by Reb1p. Lanes 1-4, blockage of yeast RNAP II by Reb1p; lanes 5-8, lack of blockage of T7 RNA polymerase by RebIp. Lanes 1 and 2, transcription of template with Reb1p binding site in the nonblocking orientation in the absence (lane 1) or presence of 50 ng of Reb1p (lane 2); lanes 3 and 4, transcription of template with Reb1p binding site in the blocking orientation in the absence (lane 3) or presence of 50 ng of Reb1p (lane 4). Note appearance of truncated transcript in lane 4 only. Lanes 5 and 6, transcription of the template with Reb1p binding site in the nonblocking orientation in the absence (lane 5) or presence of 50 ng of Reb1p (lane 6); lanes 7 and 8, transcription of template with Reb1p binding site in the blocking orientation in the absence (lane 7) or presence of 50 ng of Reb1p (lane 8). Note there is no arrest of transcription by T7 RNAP in any lane. B, autoradiogram of 6% polyacrylamide, 7 M urea gel showing blockage of yeast RNA polymerase II by RTP. Lanes 1-3, transcription of the template with BS3 site in the blocking orientation in the absence (lane 1) or presence of 400 fmol (lane 2) and 800 fmol (lane 3) of RTP; lanes 4-6, transcription of template with BS3 site in the nonblocking orientation in the absence (lane 4) or presence of 400 fmol (lane 5) and 800 fmol (lane 6) of RTP. Note blockage of transcription in lanes 2 and 3. C, autoradiogram of a 6% polyacrylamide, 7 M urea gel showing blockage of yeast RNA polymerase I by RTP. Lanes 1 and 2, transcription of template with BS3 site in the blocking orientation in the absence (lane 1) or presence of 400 fmol (lane 2) of RTP; lanes 3 and 4, transcription of template with BS3 site in the nonblocking orientation in the absence (lane 3) or presence of 400 fmol (lane 4) of RTP. Note blockage of transcription in lane 2.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The spatial and temporal overlap of replication and transcription in the chromosome has stimulated investigations not only of the possible mechanistic impact of transcription on the three steps of replication but also of the impact of replication on transcription (26-30, 41, 42). Transcriptional activation of replication origin of bacteriophage  and of yeast have been reported (26, 27). The impact of transcription on the elongation step of replication has been elegantly investigated, and the results show that transcription does not have much impact on fork movement other than causing a transient arrest of the fork when the replication fork and the transcriptional apparatus encounter each other while approaching from opposite directions. Conversely, passage of a replication fork did not dislodge RNAP from a promoter (28, 29).

In eukaryotes, genes that are transcribed early in S phase also replicate early (41) and chromosomal segments that are heterochromatic tend to replicate late (43). Orc proteins, which have been implicated in initiation of chromosome replication, also contribute to silencing of transcription of yeast mating type cassettes (42). Thus, transcription and DNA replication seem to have a significant influence on each other.

We have investigated the impact of transcription on the termination step of DNA replication and vice versa and have noted that replication terminator proteins not only arrest replication forks and replicative helicases but also transcription elongation catalyzed by several prokaryotic RNA polymerases in a polar mode. When transcription is allowed to pass through a replication terminus from the permissive direction, the ability of the replication terminus to arrest replication forks is abrogated (30). Since transcription, in some cases, can pass through a protein-DNA complex without dissociating the bound protein (44-46), it was necessary to work out the mechanism of transcriptional inactivation of replication termination. The evidence presented in this article showed that the transcriptional passage dissociated the replication terminator proteins from the replication termini. Thus, in principle, inducible transcription should provide a mechanism for regulation of replication termination.

In B. subtilis, replication forks under stringent conditions are arrested at conditional termini ( sites) located approximately 200 kilobase pairs on either side of the replication origin and the arrest requires RTP. Under relaxed conditions, the forks are released and proceed to the normal replication termini before being terminated (32). Our recent work shows that transcription modulated by the alarmone ppGpp seems to be responsible for the conditional derepression of the  sites.³ The arrest of transcriptional elongation has also been implicated in the autoregulation of Tus protein of E. coli (47).

The elongation phase of transcription in both prokaryotes and eukaryotes can be subject to regulation of three types: pausing, arrest, and authentic termination (48). The pausing is due to the interaction of RNA polymerase with specific DNA sequences and is transitory in duration, and the RNA polymerase continues to elongate the RNA chain after the brief period of pausing. Arrest of transcription is of longer duration and is caused by the encounter of RNAP with proteins bound to DNA, although not all DNA-binding proteins arrest RNAPs (37, 48). The arrested RNAP remains bound to the DNA and can continue to elongate the RNA chain once the bound protein is removed by some factor. Authentic termination of transcription can be induced by DNA sequences and transcription terminator proteins, resulting in the dissociation of RNAP and release of the RNA chain. The evidence presented in this article showed that replication terminator proteins of E. coli and B. subtilis arrested but did not terminate RNA chains. In this regard, the replication terminator proteins behaved very much like the EcoRI E111Q mutant protein, which binds very strongly to EcoRI sites with a K_d of 10¹⁵ mol/liter and arrests (or pauses) but does not terminate transcription (36, 37, 48). The DnaA initiator protein is also known to arrest transcription in a somewhat polar mode (49).

Does the arrest of RNAPs by the replication terminator proteins involve a nonspecific barrier created by the interaction of the proteins with the replication terminus, or does it also involve some protein-protein interaction between the arresting and the arrested proteins? On one hand, the diversity of primary sequences of the arrested proteins that include several replicative helicases, prokaryotic RNAPs and, as reported here, RNAPs I and II of yeast by RTP (and Tus) might suggest a lack of specific protein-protein interaction (2, 30, 31, 37). On the other hand, replication terminator proteins do not arrest all helicases and, therefore, show some specificity. For example, helicases involved in rolling circle replication are not arrested either in vitro or in vivo (19, 31, 50). It is likely that there may be protein-protein interactions between the arrested and the arresting proteins. Even if such interactions turn out to be relatively nonspecific, these interactions probably are critical for the polar arrest of DNA and RNA chains by replication terminator proteins.

Such essential but nonspecific protein-protein interactions have been reported to occur between transcriptional activator proteins and RNAPs in both eukaryotes and prokaryotes (51). Transcriptional activator proteins contact different subunits of E. coli RNAP. The single strand binding protein of phage N4 called N4SSB contacts the subunit (52), whereas  repressor contacts the carboxyl terminus of the  subunit of RNAP. Although mutations at the contact points of either the activator or RNA polymerase can cause loss of transcription activation, the loss of one contact can be relieved by the establishment of a different contact with a different subunit of RNAP (51-54). Thus, these interactions, although biologically important, need not be very specific.

The observation reported here that replication forks and RNAPs are arrested at different locations of the Ter () sequence may be due to protein-protein interaction in addition to the interaction of Tus and RTP with their cognate sites on DNA. The differences might suggest different locations of the active sites of DNA and RNA polymerases from the point of arrest or from the point of contact between these enzymes and the terminator protein-replication terminus complex. The observation that the Reb1 protein of yeast while arresting RNAP I of yeast was incapable of arresting either T7 RNAP or DnaB helicase, as reported here, would suggest that the eukaryotic transcription terminator protein probably also interacts with the RNAPs that it arrests (40).

Finally, future work on the interaction between transcription and replication termination will be directed toward the isolation of mutants of Tus and RTP that bind normally to DNA but do not arrest RNAPs. Preliminary work shows that the Y33A mutant of RTP fails to arrest RNAPs.⁴ If protein-protein interaction is involved, future work will also be directed toward mapping physically the contact points between RNAPs and Tus and RTP by photo-cross-linking procedures (52, 54).