A Replication Terminus Located at or Near a Replication Checkpoint of Bacillus subtilis Functions Independently of Stringent Control

Abstract

We have examined a replication terminus (ψL1) located on the left arm of the chromosome of Bacillus subtilis and within the yxcC gene and at or near the left replication checkpoint that is activated under stringent conditions. The ψL1 sequence appears to bind to two dimers of the replication terminator protein (RTP) rather weakly and seems to possess overlapping core and auxiliary sites that have some sequence similarities with normal Ter sites. Surprisingly, the asymmetrical, isolated ψL1 site arrested replication forks in vivo in both orientations and independent of stringent control.In vitro, the sequence arrested DnaB helicase in both orientations, albeit more weakly than the normal Ter1 terminus. The key points of mechanistic interest that emerge from the present work are: (i) strong binding of a Ter (ψL1) sequence to RTP did not appear to be essential for fork arrest and (ii) polarity of fork arrest could not be correlated in this case with just symmetrical protein-DNA interaction at the core and auxiliary sites of ψL1. On the basis of the result it would appear that the weak RTP-L1Ter interaction cannot by itself account for fork arrest, thus suggesting a role for DnaB-RTP interaction. We have examined a replication terminus (ψL1) located on the left arm of the chromosome of Bacillus subtilis and within the yxcC gene and at or near the left replication checkpoint that is activated under stringent conditions. The ψL1 sequence appears to bind to two dimers of the replication terminator protein (RTP) rather weakly and seems to possess overlapping core and auxiliary sites that have some sequence similarities with normal Ter sites. Surprisingly, the asymmetrical, isolated ψL1 site arrested replication forks in vivo in both orientations and independent of stringent control.In vitro, the sequence arrested DnaB helicase in both orientations, albeit more weakly than the normal Ter1 terminus. The key points of mechanistic interest that emerge from the present work are: (i) strong binding of a Ter (ψL1) sequence to RTP did not appear to be essential for fork arrest and (ii) polarity of fork arrest could not be correlated in this case with just symmetrical protein-DNA interaction at the core and auxiliary sites of ψL1. On the basis of the result it would appear that the weak RTP-L1Ter interaction cannot by itself account for fork arrest, thus suggesting a role for DnaB-RTP interaction. guanosine tetraphosphate replication terminator protein base pair(s) polymerase chain reaction Escherichia coli, Bacillus subtilis, and possibly other bacteria respond to adverse metabolic conditions, such as amino acid starvation by synthesizing and accumulating the alarmone, guanosine tetraphosphate (ppGpp),1 the so called magic spot 1. The alarmone shuts down the synthesis of several RNAs including rRNA. It is an adaptive response to metabolic stress and it promotes energy conservation by turning off some of the synthetic processes (1Cashel M. Rudd K.E. Neidhart F.C. Ingraham J.L. Low K.B. Magasanik B. Schachter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium, Cellular and Molecular Biology. 2nd Ed. 2. American Society of Microbiology, Washington, D. C.1997: 1410-1438Google Scholar). In E. coli, ppGpp synthesis is mediated by two genes called relA and spoT that encode ppGpp synthetase I and a gene product that catalyzes both the synthesis and degradation of ppGpp, respectively (2Gentry D.R. Cashel M. Mol. Microbiol. 1996; 19: 1373-1384Crossref PubMed Scopus (141) Google Scholar). In B. subtilis, a relA homolog has been found but no homolog for spoT has been discovered so far (see Ref. 1Cashel M. Rudd K.E. Neidhart F.C. Ingraham J.L. Low K.B. Magasanik B. Schachter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium, Cellular and Molecular Biology. 2nd Ed. 2. American Society of Microbiology, Washington, D. C.1997: 1410-1438Google Scholar). Apparently, inhibition of protein synthesis by amino acid analogs, such as arginine hydroxamate causes intracellular accumulation of uncharged tRNA, that in turn activates the relA gene. Synthesis of ppGpp is thus induced and the alarmone binds to the β subunit of RNA polymerase, causing turning off of transcription from certain promoters (3Chatterji D. Fujita N. Ishihama A. Genes to Cells. 1998; 3: 279-287Crossref PubMed Scopus (139) Google Scholar). The effect of stringent response on DNA replication has been investigated in bacteria and in bacteriophage λ. In B. subtilis, induction of stringent response by addition of arginine hydroxamate causes replication forks initiated from oriC to be arrested at check points, that we have labeled as ψL and ψR (Fig. 1), that are located ∼200 kilobases on either sides oforiC (4Serore S.J. Levine S. Vanniere V. Hencke G. Nature. 1986; 321: 709-710Crossref Scopus (23) Google Scholar, 5Levine A. Vanniere F. Dehbi M. Henckes G. Serore S.J. J. Mol. Biol. 1991; 219: 605-613Crossref PubMed Scopus (85) Google Scholar). Apparently, arrest at both of the checkpoints requires the replication terminator protein (RTP), suggesting the existence of Ter-like sites about the checkpoints. Upon return to relaxed conditions, the forks are released from the checkpoints and continue the duplication of the chromosome until arrested and terminated at the terminus region called TerC (Fig. 1, top; see Ref. 16Levine A. Autret S. Seror S.J. Mol. Microbiol. 1995; 15: 287-295Crossref PubMed Scopus (36) Google Scholar). In bacteriophage λ, stringent response causes the turning off of the ori-O-P transcript, thereby shutting down the synthesis of O and P proteins and possibly also abrogating the transcriptional activation of the λ ori. λ Replication is thereby turned off at the initiation step (6Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO, J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar). Initiation from oriC of E. coli appears to be terminated at the origin under stringent conditions (5Levine A. Vanniere F. Dehbi M. Henckes G. Serore S.J. J. Mol. Biol. 1991; 219: 605-613Crossref PubMed Scopus (85) Google Scholar). Interaction of ppGpp with RNA polymerase is believed to promote fork passage by dislodging the RNA polymerase stalled at damaged sites on DNA and thus, in this context, stringent conditions promote rather than block fork progression (7McGlynn P. Lloyd R.G. Cell. 2000; 101: 35-45Abstract Full Text Full Text PDF PubMed Scopus (246) Google Scholar). A recent article reported the existence of a short sequence of 17 bp that was located at or near the left checkpoint and had some homology with the core sequence of the Ter sites of B. subtilis. The sequence bound to a single dimer of RTP and was reported to have arrested replication in a polar fashion under stringent but not under relaxed conditions (8Autret S. Levine A. Vanniere F. Serore S.J. Mol. Microbiol. 1999; 31: 1665-1679Crossref PubMed Scopus (23) Google Scholar). This result was surprising because two interacting dimers of RTP, bound to the core and auxiliary site are known to be essential for replication termination (9Carrigan C.M. Peck R.A. Smith M.T. Wake R.G. J. Mol. Biol. 1991; 222: 197-207Crossref PubMed Scopus (25) Google Scholar, 10Sahoo T. Mohanty B.K. Bastia D. EMBO, J. 1995; 14: 619-628Crossref PubMed Scopus (26) Google Scholar, 11Manna A.C. Pai K.S. Bussiere D.E. White S.W.A. Bastia D. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 3253-3258Crossref PubMed Scopus (21) Google Scholar) and in fact polarity is generated from symmetrical dimers of RTP (12Bussiere D.E. Bastia D. White S.W. Cell. 1995; 80: 651-660Abstract Full Text PDF PubMed Scopus (76) Google Scholar) by differential interaction with the core and auxiliary sites of Ter (13Bussiere D.E. Bastia D. Mol. Microbiol. 1999; 31: 1611-1618Crossref PubMed Scopus (71) Google Scholar). In view of these apparently puzzling differences between the behavior of a normal Ter and the one present at or near the left checkpoint of B. subtilis, we decided to reinvestigate the characteristics of the site in vivo and in vitro. We identified a 28-bp long terminator site that we call ψL1 that bound to two dimers of RTP and arrested replication in a bipolar modein vivo, when present in a plasmid context. The 17-bp sequence reported by Autret et al. (8Autret S. Levine A. Vanniere F. Serore S.J. Mol. Microbiol. 1999; 31: 1665-1679Crossref PubMed Scopus (23) Google Scholar) is included in the 28-bp sequence. The isolated 17-bp core site of ψL1 (L1), on the other hand, neither arrested replication in vivo nor DnaB helicase in vitro. The RTP-L1 complex had a half-maximal stability of just a few seconds whereas the Ter1-RTP complex under identical conditions had a half-life of ∼180 min. The observation that a very weak terminator-RTP complex still showed appreciable replication arrest in vivo tends to support the conclusion that strong protein-DNA interaction by itself is not obligatory for replication fork arrest at the Ter sites ofB. subtilis. The E. coli strains JM109 (sup E44, rel A1,recA, endA1,gyrA96,hsdR17D, Δ[lac,proAB],[F′tra D36,lacIq Δ (lacZ)M15,proA+,rk−,mk−B+) was used for making M13 DNA, DH5α [F′sup E44,lacU169 (φ80 lacZ Δ M15), hsdR17,recA1, endA1, gyrA96,thi-1, relA1} was used for cloning purposes. Wild type B. subtilis IS58 (trpC2lys3), isogenic relA− mutant ofBacillus IS 56 were used for RNA mapping and genomic DNA preparation. The RTP overproducer strain SU230 of B. subtilis (20Smith M.T. Wake R.G. J. Mol. Biol. 1992; 227: 648-657Crossref PubMed Scopus (41) Google Scholar) was used for transforming shuttle vector clones and preparation of replication intermediates. Total RNA was prepared from wild type B. subtilis IS 58 RelA and IS 56 rel A− strains which were grown in relaxed and stringent conditions. To 25 ml of cell culture pellet, 1 ml of TRIZOL reagent (Life Technologies) was added, mixed, and incubated at 30 °C for 5 min to lyse the cells. Chloroform (200 μl) was added, mixed, and centrifuged at 12,000 × g for 15 min at 4 °C. The upper aqueous layer having RNA was precipitated using equal volume of isopropyl alcohol. The RNA pellet was washed with 70% ethanol and dissolved in RNase-free water. The RNA samples were resolved on formaldehyde-agarose gels and blotted onto nitrocellulose filters. YxcC gene of B. subtilis was amplified from genomic DNA using the primer set 5′-GATCAGGTTAGCCGCCGATAACACCAAAGTCG-3′ and 5′-AAATGTCTTAATCAAACCTTACTCCGCGCGGG-3′. The PCR product was radiolabeled using the random priming NEBlot kit (New England Biolabs) and used for Northern hybridization. The assays were carried out as described (10Sahoo T. Mohanty B.K. Bastia D. EMBO, J. 1995; 14: 619-628Crossref PubMed Scopus (26) Google Scholar). Briefly, PCR products of Ter 1, core, L1 and L1 core-binding sites were amplified from pUC18BS3 (Ter1 RTP-binding site cloned as EcoRI-HindIII fragment), pUC18core (Ter1 core-binding site cloned asEcoRI-HindIII fragment), pBAG 62 (LI RTP-binding site cloned as EcoRI-BamHI fragment in pWS64-1 vector (20Smith M.T. Wake R.G. J. Mol. Biol. 1992; 227: 648-657Crossref PubMed Scopus (41) Google Scholar) and pBAG60 (L1 core RTP-binding site cloned asEcoRI-BamHI fragment in pWS 64-1) plasmids, respectively, using universal M13 forward and reverse primers. These fragments were end-labeled with [γ-32P]ATP and T4 polynucleotide kinase. 10 Fmol of labeled DNA was used in each reaction, that were carried out in 20-μl volumes of 40 mmTris-Cl (pH 7.5), 4 mm MgCl2, 50 μg/ml bovine serum albumin, 50 mm potassium glutamate, 5 μg of calf thymus DNA, and increasing amounts of RTP protein (0, 20, and The were carried out at for 30 min and resolved on was using 10 of end-labeled Ter1 and L1 PCR products and of having or sites. L1 and Ter1 were incubated at for 15 min with and 40 of RTP, respectively, in a 20-μl reaction volume 40 mm (pH 7.5), 4 mm MgCl2, 50 μg/ml bovine serum albumin, 50 mm potassium of was for increasing from to The reaction was with 4 of × DNA and samples resolved on assays were as described T. P. Bastia D. Cell. Full Text PDF PubMed Scopus Google Scholar). Briefly, 10 of or L1 core-binding sites were using [γ-32P]ATP and T4 polynucleotide and to 1 of (Ter1 RTP-binding site cloned in as EcoRI-HindIII or M13 (Ter1 RTP-binding site cloned in as EcoRI-HindIII fragment), M13 L1 (L1 RTP-binding site cloned in asEcoRI-BamHI fragment), M13 L1 (L1 RTP-binding site cloned in asEcoRI-BamHI and M13 L1 core (L1 core RTP-binding site cloned in asEcoRI-BamHI DNA, These helicase was and 10 of the used in each reaction of volumes 40 mm (pH 7.5), 4 mm MgCl2, 50 μg/ml bovine serum albumin, mm 50 mm potassium glutamate, 5 mm and increasing amounts of RTP (0, and 15 min at of DnaB helicase was to the reaction and for 30 min at °C. The reaction was with and resolved on Ter1 and L1 were cloned as EcoRI-BamHI fragments in both orientations in the shuttle vector pWS (20Smith M.T. Wake R.G. J. Mol. Biol. 1992; 227: 648-657Crossref PubMed Scopus (41) Google Scholar) and RTP overproducer B. subtilis strain These clones were grown in 50 ml of potassium potassium or in having μg/ml at 30 °C to and RTP was induced by for and μg/ml arginine hydroxamate for to stringent 10 mm was and on at × g and the cell pellet was as Briefly, the were in 4 ml of 50 mm (pH 10 mm and μg/ml and on for 15 min. 1 ml of was very and incubated on for 15 min and ml of 5 was and left on The was by at at 4 °C for min and plasmid DNA precipitated with volume of isopropyl alcohol. The pellet washed in 70% ethanol was dissolved in 10 mm (pH 1 mm The samples were with precipitated with volumes of and dissolved in The samples were and used for with or and to and samples were resolved on a and used for to nitrocellulose filters. The were under at °C for and for 4 at °C in × shuttle vector DNA was radiolabeled with random priming using the NEBlot kit (New England Biolabs) and used as a for at °C The were washed in × and to The samples were used for and was to Cell. Full Text PDF PubMed Scopus Google Scholar). and replication samples were to on long at 30 for and the in 1 The were and in 1 μg/ml for the at for The resolved samples were to a nitrocellulose and with labeled vector 64-1) DNA, and to an work A. Autret S. Seror S.J. Mol. Microbiol. 1995; 15: 287-295Crossref PubMed Scopus (36) Google Scholar) had the left replication checkpoint of B. subtilis between the (Fig. We to for that had homology with the core sequence of Ter and with a sequence as ψL1 (Fig. and was to for identified by a of the homology with core sequence of Ter that a binding to We were by a of on that had that a single dimer of RTP bound to DNA was not to arrest replication forks and that two interacting dimers bound to the sequence were essential for replication fork arrest A.C. Pai K.S. Bussiere D.E. White S.W.A. Bastia D. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 3253-3258Crossref PubMed Scopus (21) Google Scholar, T. Mohanty B.K. Bastia D. J. Biol. 1995; Full Text Full Text PDF PubMed Scopus Google Scholar). was that the a terminus sequence in this region of B. subtilis. in the region we found a long sequence that the when bound to RTP as described The of ψL1 L1 is in 1, B. A with the core, with L1 and the core sequence of L1 are (Fig. 1, A and The that were identified by are (10Sahoo T. Mohanty B.K. Bastia D. EMBO, J. 1995; 14: 619-628Crossref PubMed Scopus (26) Google Scholar). The of L1 that to the points on Ter1 are also (Fig. 1, Ter1 RTP-binding core site with the core L1 28-bp long RTP-binding and L1 core long L1 core were used for It be that the DNA were cloned a plasmid vector and the DNA fragments bp the and vector were amplified by PCR and of RTP were to these DNA fragments and on of and the site by the two and of of the DNA the of RTP sequence showed one because it bound one dimer of RTP (Fig. L1 DNA also showed a single when amounts of RTP were at of RTP, a of binding of two dimers was (Fig. were differences between the of Ter1 and L1 in that L1 was not a very weak but showed In fact we could not binding and of a of the L1 DNA to the of the under the that included some The binding of L1 to RTP was and is in E. L1 core DNA showed a single the of RTP that a single dimer of RTP bound to the DNA (Fig. A of the binding of L1 with L1 core the conclusion that the L1 sequence bound to two dimers The binding was rather weak and showed or no in with the Ter1 sequence. We the binding described with with an of Ter1 and L1 were with an of RTP and with a of the DNA and were on gels to and the released DNA from the of the for and L1 are in A and The was from from of gels and are in It is from the that whereas the Ter1-RTP complex to a half-maximal in the complex in a few seconds the addition of the In the of the DNA, both were min (Fig. A and The L1 sequence is located within the yxcC gene that is in a to that of replication fork We to this gene was under the stringent for two because the L1 site is located within the of the we were in of transcription have some effect on replication termination at this site and we to the transcriptional of this gene addition of arginine hydroxamate as a that the were under stringent conditions. response relA− and isogenic wild type B. subtilis strains were for the of yxcC in vivo normal and stringent conditions. The Northern showed that the that to the yxcC DNA was present of arginine hydroxamate the RelA the one RelA In in isogenic wild type with arginine hydroxamate the to be in with the the relaxed stringent in the support the conclusion that the transcription of gene was under stringent and the arginine hydroxamate as used in to stringent The shuttle vector pWS was used for cloning L1 is a of and the B. subtilis plasmid (20Smith M.T. Wake R.G. J. Mol. Biol. 1992; 227: 648-657Crossref PubMed Scopus (41) Google Scholar). The vector had and that were in E. coli and B. subtilis, The L1 and Ter1 were cloned between the and sites in both orientations and the RTP overproducer strain SU230 of B. subtilis. The replication from grown under stringent and relaxed conditions were and by and with either or The plasmid DNA was by but (Fig. Ter1 a replication with than the DNA that were of an arrested The arrested fork was in the but not in the by we had more of the DNA in the gels the two In L1 was to block replication fork in both relaxed and stringent conditions in both orientations, the of was in to that of 1 L1 to be bipolar and did not have a or We to that the is by in A were replication rather than or to by in gels Cell. Full Text PDF PubMed Scopus Google Scholar). the stalled would a spot on the replication and at the in the The of Ter1 with but not the had an of stalled (Fig. arrested forks were in the The arrest of the fork is so strong that a replication is not at this of and the spot and the arrested at this of of the the the Ter1 and the and generated by the few forks that the terminus are not in the We have reported the of an in gels when the forks are arrested at a Ter site T. P. Bastia D. Cell. Full Text PDF PubMed Scopus Google Scholar). The present is with that The L1 DNA, in the the core generated a replication that was to and just the spot to a stalled replication fork (see in The generated to the from a T. P. Bastia D. Cell. Full Text PDF PubMed Scopus Google Scholar). the of the that would be generated by the few forks that the L1 site single and are not in this long was in the the auxiliary of L1 with the We the DNA samples at both sites (Fig. to the termination spot still the the spot were an by of plasmid DNA by a single it would not on the replication and would not be of the DNA at two from under both relaxed and stringent conditions the termination spot (Fig. It be in that L1 has no The and the core or the auxiliary site the it be that the spot and the stalled and the was not at this of The be in very long of the that the termination spot to It be that the of replication with reported for that generated from an a Ter when the replication are within the near the Ter site (10Sahoo T. Mohanty B.K. Bastia D. EMBO, J. 1995; 14: 619-628Crossref PubMed Scopus (26) Google Scholar, T. P. Bastia D. Cell. Full Text PDF PubMed Scopus Google Scholar). We have also examined the replication from the plasmid the L1 core sequence by gels and found no of replication arrest, with S. Levine A. Vanniere F. Serore S.J. Mol. Microbiol. 1999; 31: 1665-1679Crossref PubMed Scopus (23) Google and not The was to the bipolar arrest of L1 forks vivo by an in We that L1 arrest DnaB helicase in both orientations used as a polar arrest of helicase under identical conditions. The these The of helicase assays showed that arrested DnaB in an L1 arrested the helicase in an fashion A and The of the from assays showed that both the core site of L1 and the as very of helicase arrest at RTP orientations of L1 were more in helicase arrest than both the but than the of Ter1 (Fig. the in helicase arrest appear to be with the in vivo replication arrest The work reported a very weak binding of RTP to a site and the fact that the complex was and in a few seconds with an of DNA, the complex was still to arrest replication forks in vivo and DnaB helicase in vitro. The chromosome not as DNA but is with proteins PubMed Scopus Google some of which are strong DNA but the replication forks apparently are to the bound proteins arrest until the TerC region are This would that protein bound to DNA, a weak complex such as the by would not be to be to arrest replication forks the complex had some such as an to with the helicase and The observation reported with A.C. Pai K.S. Bussiere D.E. White S.W.A. Bastia D. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 3253-3258Crossref PubMed Scopus (21) Google would that in addition to interaction between RTP and the helicase is in fork It is surprising that L1 was to arrest forks from both the arrest that Ter sites with the of a bipolar Ter in a plasmid of B. subtilis Smith M.T. Wake R.G. G. S. Mol. Microbiol. 1996; 19: PubMed Scopus Google Scholar). the bipolar the L1 has a symmetrical sequence and it symmetrical base to amino acid with D. E. T. S. J. J. S. and D. Biol. in this we not have a for the bipolar of L1 with to replication fork would that the complex a that promotes with the helicase from either and in of DNA It be to as to have to that the L1 sequence bound a single dimer of the for the is that the 17-bp core sequence of L1 was used by for RTP binding and the binding behavior of the core and auxiliary sequence was not examined (8Autret S. Levine A. Vanniere F. Serore S.J. Mol. Microbiol. 1999; 31: 1665-1679Crossref PubMed Scopus (23) Google Scholar). It more to as to stringent arrest of replication was by the other in a genomic DNA fragment that included L1 (8Autret S. Levine A. Vanniere F. Serore S.J. Mol. Microbiol. 1999; 31: 1665-1679Crossref PubMed Scopus (23) Google Scholar). We no that replication arrest at L1 was by stringent L1 at in stringent fork It is that it be one of the fork and that other RTP-binding that with L1 to an to fork of this site in the It be in that L1 by itself was not as strong an of replication forks as the Ter1 terminus. In the genomic out of the L1 more mapping of the site of arrest of the forks in under stringent conditions, and more to the stalled fork from the chromosome are to be to the