The Bacillus subtilis replication terminator system functions in Escherichia coli.

The Bacillus subtilis DNA terminators, IRI + IRII, were inserted into the Escherichia coli plasmid pACYC184 such that the IRI terminator would be in its active orientation with respect to the approaching unidirectionally moving replication fork. When this new plasmid was transferred into E. coli, harbouring an expression plasmid producing the B. subtilis terminator protein RTP, fork arrest was observed to occur at the position of the inserted terminator region. Thus, the B. subtilis replication terminator system can function in E. coli. It was shown that the B. subtilis system operated with approximately 30% of the efficiency of the E. coli system utilizing the R6K TerR2 DNA terminator and the E. coli Tus terminator protein. Assuming that RTP and Tus have quite different folded structures these results suggest that fork arrest in B. subtilis is not dependent upon a highly specific recognition and interaction between RTP positioned on the DNA terminator and a component(s) of the approaching replisome.

Definition and polarity of action of DNA replication terminators in Bacillus subtilis.

The first stage in termination of chromosome replication in Bacillus subtilis involves arrest of the clockwise fork at the inverted repeat region (IRR), comprising the opposed IRI and IRII sequences, adjacent to the upstream region of the rtp gene, which encodes the replication terminator protein RTP. RTP binds to IRI and IRII. The ability of the IRR and its components to function as terminators, in conjunction with RTP, and their polarity of action have now been tested by the use of plasmids replicating in B. subtilis as unidirectional theta structures and into which potential terminator sequences were inserted in alternate orientations relative to fork movement. When the complete IRR was inserted into such plasmids and the new plasmids transferred into a B. subtilis strain overproducing RTP, it was able to block movement of a replication fork approaching from either direction. IRI and IRII were shown to function as polar terminators, each blocking movement of a fork when it approached from one particular direction but not the other. Furthermore, the polarity of action was in accordance with the IRR being able to operate as a replication fork trap. Thus, a fork approaching the IRR would pass through the first terminator encountered (IRI or IRII) and be halted by the second. The previously observed nonfunctioning of a particular orientation of chromosomal IRR as a fork arrest site probably reflects a limiting level of RTP in the cell. Interestingly, a 21 base-pair core sequence spanning a single RTP binding site within IRI (the 47 base-pair IRI contains 2 binding sites) was unable to arrest a fork approaching from either direction in the plasmid system. This suggests that both binding sites within an IR must be filled in order to function as an arrest site. It is possible that co-operative interaction between adjacent dimers within IRI or IRII provides the necessary conformation for causing fork arrest.

The normal replication terminus of the Bacillus subtilis chromosome, terC, is dispensable for vegetative growth and sporulation.

The Bacillus subtilis strains CU1693, CU1694 and CU1695 were shown by hybridization analysis to carry large deletions of the terminus region that originated within discrete fragments of the SP beta prophage genome. The absence of terC in CU1693 was demonstrated definitively by the identification of a novel junction fragment comprising SP beta DNA and DNA that lies on the other side of terC in the parent strain. This represented the deletion of approximately 230 kb of CU1693 DNA, with the removal of approximately 150 kb to the left of terC and approximately 80 kb to the right of terC. The lack of hybridization of CU1694 and CU1695 DNA to cloned DNA carrying the terC sequence and to cloned DNAs flanking terC suggested that terC is absent from the chromosome of each of these strains also, and that the deletions in CU1694 and CU1695 extend beyond the segment of the terminus region that has been mapped and cloned. The normal growth rate and morphology of CU1693, CU1694 and CU1695 relative to the parent strain when grown in complex medium indicated dispensability of terC for vegetative growth and division. B. subtilis SU153 was constructed using a specific deletion-insertion vector that was designed to effect the deletion of 11.2kb of DNA spanning terC, with the removal of approximately 9.7kb to the left of terC and approximately 1.kb to the right of terC. This manipulation did not introduce any readily detectable auxotrophic requirement. Physiological characterization of SU153 confirmed the dispensability of terC for vegetative growth and cell division, and also established the lack of requirement of terC for the specialized cell division that is associated with formation of the bacterial endospore.

Restriction map of DNA spanning the replication terminus of the Bacillus subtilis chromosome.

The Bacillus subtilis 168 dna-1 chromosome was labelled during sporulation with [3H]thymine for five minutes immediately before termination of replication. The isolated radioactive DNA was cleaved with BamHI (or SalI) and the resulting restriction fragments separated by agarose gel electrophoresis. The individual fragments, fractionated into a series of slices cut from the gel, were then cleaved with SalI (or BamHI) and the double-digest fragments identified by electrophoresis and fluorography. All major fragments and most minor ones present in a whole double-digest were assigned to BamHI and SalI parents. Such information enabled the construction of an unambiguous restriction map of 150 X 10(3) bases of the approximately 250 X 10(3) bases of DNA labelled in the five minutes. In conjunction with published data on the order of replication of restriction fragments as termination is approached, it was clear that most (105 X 10(3) bases) of the mapped DNA was replicated by a major fork moving in one direction towards a BamHI 24.8 X 10(3) base fragment. The 45 X 10(3) bases extending to the other side of this region were labelled only slightly, and presumably was replicated by a fork that approached the other in an opposite direction until its progress was blocked or severely impeded within this region at a site, referred to as terC, sometime (less than 5 min) earlier. The regions of the map replicated in the final 2.5 and 1.0 minute by the major fork were also identified.

Impediment to replication fork movement in the terminus region of the Bacillus subtilis chromosome.

The terminus regions of the chromosomes of three strains of Bacillus subtilis 168 were radioactively labelled by supplying [3H]thymine towards the end of a round of replication. These strains lacked or contained the prophage SP beta c2. Following restriction endonuclease digestion of the purified DNA and fluorography, an SP beta c2-related perturbation of the terminus-labelling profile was observed, which was completely consistent with the previously suggested existence of an impediment to replication fork movement (terC) within a BamHI 24.8 X 10(3) base fragment (Weiss & Wake, 1983). The present data suggest that terC is located within the 11.4 X 10(3) base BamHI + SalI double-digest portion of this BamHI fragment.

An immobilized fork as a termination of replication intermediate in Bacillus subtilis.

The structure of a DNA intermediate associated with termination of chromosome replication in Bacillus subtilis and derived from a unique BamHI 24.8 X 10(3) base-pair (bp) region of the chromosome has been investigated. The intermediate has properties expected for a forked structure. Gel electrophoresis followed by Southern transfer and hybridization to cloned DNA has shown it to comprise single strands of 15.4 X 10(3) bp and 24.8 X 10(3) bp, in approximately equimolar amounts. After purification away from the bulk of chromosomal DNA, electron microscopy of the intermediate established that 15% of the DNA was present as branched molecules and a significant proportion (11 of 31) of these contained two arms of matching length. The average dimensions (best estimates) of this unique class of Y-shaped molecule were 9.5(+/- 0.3) X 10(3), 15.1(+/- 0.4) X 10(3) and 24.6 24.6(+/- 0.6) X 10(3) bp for the stem, arms and end-to-end length, respectively. These values are consistent with the single strand composition of the intermediate as found. Furthermore, hybridization of the single strands to DNA from known locations within the BamHI 24.8 X 10(3) bp region has established the orientation of the forked intermediate relative to the genetic map. The intermediate presumably reflects the immobilization of the clockwise replication fork within the 24.8 X 10(3) bp region, at a location approximately 15.4 X 10(3) bp from the right end.

The Bacillus subtilis DNA replication terminator.

The recent discovery of the Bacillus subtilis plasmid terminator TerLS20 with bidirectional fork arrest activity has provided the opportunity to probe further the structural and functional features of B. subtilis replication terminators in general. The minimal TerI and TerLS20 terminators each comprise two 13 nt segments flanking a central trinucleotide, which is almost completely conserved in all terminators. It corresponds to the region of overlap of the two RTP binding sites (A and B) on the DNA. It has been shown that, despite this conservation, considerable variation in this trinucleotide region still allows fork arrest activity. Thus, the productive interaction of the RTP dimers, which presumably occurs in the vicinity of this trinucleotide region, is not dependent upon stringently defined contacts with the bases in this region. A completely synthetic and highly symmetrical terminator was constructed by replacing the 13 nt segment of the A site of TerI with an opposed segment identical to that in the B site. The efficient bidirectional activity of this new terminator, TerSymB, established more firmly the need for two opposed RTP binding sites in a functional terminator. TerSymB was used to investigate the effect of sequence deviation in one of the 13 nt segments, from that in the B site, on bidirectionality of the terminator. It was found that the deviations introduced converted the terminator significantly towards polarity of action. The partial symmetry within each of the 13 nt segments of TerSymB, and the presumed recognition of this symmetry in the binding of a symmetrical dimer of RTP to each overlapping site, suggest that the bound dimers are centred over positions in the DNA sequence separated by 15 nt. This separation distance has been used in conjunction with the mode of binding of RTP to DNA proposed by Bussiere et al., based on their crystal structure for RTP, to model the interaction of the two dimers of RTP with unbent B-form DNA. Increased separation of the two binding sites of TerSymB was performed by inserting an extra three, seven or ten nucleotides centrally within the TerSymB sequence. The effects of these insertions on RTP binding and fork arrest activity were consistent with the proposed positioning of the RTP dimers within the terminator sequence, and interaction between the dimers bound to TerSymB. A model to account for the generation of RTP-terminator complexes with bidirectional or polar fork arrest activity utilising TerSymB or TerI-VI is presented.

The minimal sequence needed to define a functional DNA terminator in Bacillus subtilis.

The 47 bp DNA replication terminator (IRI) of Bacillus subtilis, contains two binding sites, A and B, for the replication terminator protein (RTP). Each site binds a dimer of RTP. Removal of the first two base-pairs (bp 1-2) from IRI completely destroyed in vivo terminator (fork arrest) function and was accompanied by loss of RTP binding to the A site, which is distal to the approaching fork that is arrested. Removal of base-pairs 34 to 47 from the other end, proximal to the approaching fork, lowered in vivo function to approximately 50% of the complete IRI. RTP binding appeared to be largely unaffected. Terminator function remained at the approximately 50% level with further deletions that proceeded as far as to include base-pair 28; and RTP binding remained largely unaffected. Removal of more of the sequence beyond base-pair 27 and into the region that makes extensive contact with RTP resulted in a further impairment to in vivo function, and caused altered RTP binding. The base-pairs 1 to 24 segment retained only 16% fork arrest activity and the effect on RTP binding was largely evidenced by an elimination of the ability of this extensively truncated sequence to fill the B site alone. The behaviour of the various terminator deletions emphasize the importance of the previously defined RTP-DNA contacts which allow the binding of RTP to the two overlapping sites, A and B, of IRI for terminator function. A comparison of the affinities of selected truncated terminators for RTP raises the possibility that the overall affinity of RTP for its DNA terminator is not the sole determinant of terminator function.

Normal terC-region of the Bacillus subtilis chromosome acts in a polar manner to arrest the clockwise replication fork.

A procedure is described for relocating a functional terC-region to various sites on the Bacillus subtilis chromosome, and in alternative orientations. The relocated terC-region comprised the IRR-rtp portion of the chromosome contained within a 1.75 x 10(3) base-pair segment of DNA. This segment was first cloned into the Tn 917 vector pTV20 in both orientations, and the two new plasmids used for inserting the terC-region into chromosomal copies of Tn 917. When relocated to the pyr and metD loci (139 degrees and 100 degrees positions on the 360 degrees map) it was found that clockwise replication fork arrest occurred only when the IRR-rtp (or terC-) region was oriented, in relation to the direction of approach of the fork, in the same way as in the wild-type strain. Thus, the complete IRR when located in the chromosome, and apparently made up of opposing terminators which might enable it to function in both orientations, is polar in its action. Of the two inverted repeats present in the IRR, it appears that IRI is functional in the chromosome, but not IRII.

Breakdown and quantitation of the forked termination of replication intermediate of Bacillus subtilis.

Using a procedure that minimizes shear forces, the BamHI-derived forked termination of replication intermediate of Bacillus subtilis, called band I DNA, can be extracted with little or no accompanying band II DNA. It has been shown that band II DNA is a product of band I breakdown. Nuclease P1-mediated breakdown of the forked band I DNA proceeds in two steps. The first causes the release of one of the arms as band II DNA; in the second step, the remaining arm is cleaved away to yield the free stem. It is concluded that band I represents the primary termination of replication intermediate. A quantitative assessment of the level of band I in DNA from cells of the merodiploid strain, GSY1127, growing at different rates has been made. For cells grown in a minimal medium, at least, the experimentally measured level of band I is of the order (approx. 60%) of that predicted for a complete block to movement of the clockwise fork at the replication terminus, terC.

Identification of the replication terminator protein binding sites in the terminus region of the Bacillus subtilis chromosome and stoichiometry of the binding.

DNase I footprinting of the interaction between the replication terminator protein (RTP) of Bacillus subtilis and the inverted repeat region (IRR) at the chromosome terminus, to which it binds to block the clockwise replication fork, showed that two major regions of 41 base pairs (bp) were protected from cleavage. These regions corresponded approximately to the imperfect inverted repeats (IRI and IRII) identified previously. Band retardation analyses of the interaction between RTP and portions of the IRR established that each inverted repeat (IRI or IRII) contained two RTP binding sites. By sedimentation equilibrium in the ultracentrifuge, RTP was found to exist as a dimer of 29 kDa at neutral pH and concentrations above 0.2 g/l. Quantitative studies of the RTP-IRR interaction using [3H]RTP and [32P]IRR showed that the fully saturated complex contained eight RTP monomers per IRR. It is concluded that a dimer of RTP binds to each of the four sites in IRR. The apparent dissociation constant for the interaction was estimated (in the presence of 50% glycerol) to be 1.2 x 10(-11) M (dimer of RTP). Glycerol was found to have a marked effect on the affinity of RTP for the IRR and on the relative amounts of the interaction complexes formed; in the absence of glycerol the dissociation constant was approximately 50-fold higher and there was pronounced co-operative binding of RTP dimers to adjacent sites in each inverted repeat. Examination of the DNA sequence in IRI and IRII identified two 8 bp direct repeats in each. The regions protected from DNase I cleavage in each inverted repeat and the protection afforded by a core sequence spanning just one of the 8 bp direct repeats were consistent with each 8 bp repeat representing a recognition sequence for the RTP dimer. A model describing the binding of RTP to the IRR is presented.

Site-directed mutants of RTP of Bacillus subtilis and the mechanism of replication fork arrest.

DNA replication fork arrest during the termination phase of chromosome replication in Bacillus subtilis is brought about by the replication terminator protein (RTP) bound to specific DNA terminator sequences (Ter sites) distributed throughout the terminus region. An attractive suggestion by others was that crucial to the functioning of the RTP-Ter complex is a specific interaction between RTP positioned on the DNA and the helicase associated with the approaching replication fork. In support of this was the behaviour of two site-directed mutants of RTP. They appeared to bind Ter DNA normally but were ineffective in fork arrest as ascertained by in vitro Escherichia coli DnaB helicase and replication assays. We describe here a system for assessing the fork-arrest behaviour of RTP mutants in a bona fide in vivo assay in B. subtilis. One of the previously studied mutants, RTP.Y33N, was non-functional in fork arrest in vivo, as predicted. But through extensive analyses, this RTP mutant was shown to be severely defective in binding to Ter DNA, contrary to expectation. Taken in conjunction with recent findings on the other mutant (RTP.E30K), it is concluded that there is as yet no substantive evidence from the behaviour of RTP mutants to support the RTP-helicase interaction model for fork arrest. In an extension of the present work on RTP.Y33N, we determined the dissociation rates of complexes formed by wild-type (wt) RTP and another RTP mutant with various terminator sequences. The functional wtRTP-TerI complex was quite stable (half-life of 182 minutes), reminiscent of the great stability of the E. coli Tus-Ter complex. More significant were the exceptional stabilities of complexes comprising wtRTP and an RTP double-mutant (E39K.R42Q) bound to some particular terminator sequences. From the measurement of in vivo fork-arrest activities of the various complexes, it is concluded that the stability (half-life) of the whole RTP-Ter complex is not the overriding determinant of arrest, and that the RTP-Ter complex must be actively disrupted, or RTP removed, by the action of the approaching replication fork.

Variations and coding features of the sequence spanning the replication terminus of Bacillus subtilis 168 and W23 chromosomes.

In a comparative study of the sequences of the 3-kb regions of DNA spanning the replication terminus, terC, of Bacillus subtilis strains 168 and W23, it was found that the latter contained an insertion of a large open reading frame (ORF405) whose translated protein product is a member of the cytochrome P-450 family. The insertion was about 34 nucleotides upstream from a putative promoter for the rtp gene. The sequenced regions contained a number of other ORFs. The translation product of one (ORF238) is a member of a previously identified oxidoreductase superfamily. The translation product of another (ORF257) is significantly similar to the proC product of Escherichia coli, but this ORF does not code for a functional proC product of B. subtilis.

Expression of the rtp gene of Bacillus subtilis is required for replication fork arrest at the chromosome terminus.

It was earlier proposed that clockwise replication fork arrest at the chromosome terminus in Bacillus subtilis is dependent upon expression of the rtp gene adjacent to the site of arrest, terC [Smith and Wake, J. Bacteriol. 170 (1988) 4083-4090]. A merodiploid strain of B. subtilis, in which rtp was placed under the control of the IPTG-inducible spac-1 promoter, was constructed. Replication fork arrest at terC, as monitored by the level of a forked DNA molecule of predicted dimensions, was shown to be dependent upon IPTG-induced expression of rtp in this strain. The very low concentration of IPTG needed to induce a substantial level of fork arrest suggests that relatively little RTP, the protein product of rtp, is needed for fork arrest at terC.

The Bacillus subtilis cell-division 135-137 degrees region contains an essential orf with significant similarity to murB and a dispensable sbp gene.

Sequence similarity analysis has revealed that orf2, in the cell division 135-137 degrees region of the Bacillus subtilis (Bs) chromosome, is the probable homolog of Escherichia coli murB (encoding a reductase involved in peptidoglycan synthesis). The amino-acid sequences of the two protein products show 24% identity (47% overall similarity), with several regions of higher similarity which may represent functional domains of the proteins. Attempts to insertionally inactivate orf2 were unsuccessful, strongly suggesting that it is an essential Bs gene. A small gene found in the same region as orf2, sbp (encoding the 'small basic protein'), was shown to be non-essential in Bs.

Cloning and localization of the Bacillus subtilis chromosome replication terminus, terC.

A 10.9-kb segment of the Bacillus subtilis 168 chromosome has been cloned in an Escherichia coli plasmid and shown to contain terC (the replication terminus of the chromosome). The terC-containing portion of this plasmid has been subcloned within each of two overlapping fragments of DNA, 1.75 and 1.95 kb, again in E. coli plasmids. These have afforded a more precise definition of the location of terC in the B. subtilis chromosome and provided material for a detailed analysis of the structure and functioning of this site.

Relocation of the replication terminus, terC, of Bacillus subtilis to a new chromosomal site.

The terC-deletion strain of Bacillus subtilis 168, SU153 [Iismaa and Wake, J. Mol. Biol. 195 (1987) 299-310] was used for the re-insertion of a 1.75-kb segment of DNA containing terC at a site approx. 25 kb from its original position. The relocated terC in the new strain, SU160, was oriented normally with respect to the approaching clockwise replication fork, and was positioned such that this fork was the first to reach it. The relocated terC was effective in causing arrest of the clockwise fork, as evidenced by the appearance of a unique DNA species with a characteristic mobility in agarose gel electrophoresis and with a predicted single-strand composition. Thus, the previously cloned 1.75-kb terC-containing segment [Smith et al., Gene 38 (1985) 9-17] has not been altered with respect to TerC function and contains sufficient sequence for this function. The findings reported here provide the opportunity for establishing the minimal and essential sequence features of terC, and for examining its possible polarity of action in causing fork arrest.

Physical map of the Bacillus subtilis replication terminus region: its confirmation, extension and genetic orientation.

The library of Bacillus subtilis DNA previously cloned in the cosmid vector pHC79 has been screened for the replication terminus region using a higher level of terminus probe. 24 of 48 recombinant cosmids which gave a positive response had restriction fragment compositions consistent with their inserts originating from or extending into the terminus region for which a 150-kb restriction map has already been constructed (Weiss and Wake, 1983). DNA spanning terC, the site of termination, appears to be missing from the library, although DNA to either side of terC has been cloned. A detailed analysis of four of the newly identified recombinant cosmids has confirmed most of the previous 150-kb map and allowed it to be extended to 180 kb. Physical linkage of the two cosmid inserts that most closely approach terC on each side has been demonstrated. The location of the genetic marker gltA and the orientation of the restriction map relative to the genetic map of the B. subtilis chromosome have also been established.

Conservation of the 168 divIB gene in Bacillus subtilis W23 and B. licheniformis, and evidence for homology to ftsQ of Escherichia coli.

The chromosomal regions of Bacillus subtilis (Bs) W23 and Bacillus licheniformis (Bl), which span the sequence encoding the homolog of the division initiation gene, divIB, of Bs168 were cloned and sequenced. The high level of conservation of the amino acid (aa) sequence of the DivIB protein (99 and 68% identity for BsW23 and Bl, respectively) was consistent with a significant role for this protein in the cell cycle of the two species. The hydropathy profile for DivIB of Bl was almost identical to that of Bs168 and consistent with a membrane location, as previously established for the latter. The higher than average level of identity (87%) of the 31-aa N-terminal cytoplasmic domain of DivIB between Bs168 and Bl raised the possibility of a special role for this domain. Database analyses using the Bl DivIB sequence and similarity analyses also strongly suggested that DivIB, of Bl and Bs, is a homolog of FtsQ of Escherichia coli. The flanking sequences extending into the unidentified orfs both upstream and downstream from divIB were highly conserved between Bs168 and Bl at both the nucleotide and aa levels. It was confirmed that orf4 of Bs168 is dispensable.

Autoregulation of the gene encoding the replication terminator protein of Bacillus subtilis.

One of two putative sigma A promoters identified previously in the region immediately upstream from the rtp gene (encoding the replication terminator protein) [Smith and Wake, J. Bacteriol. 170 (1988) 4083-4090] has been shown by transcription start point (tsp) mapping to be the functional rtp promoter. In these tsp mapping experiments, it was observed that the level of mRNA from this promoter, Prtp, was increased by a factor of 30 in the absence of the replication terminator protein (RTP), consistent with the autoregulation of rtp at the level of transcription. In vitro transcription from Prtp by sigma A RNA polymerase has been shown to be specifically repressed by RTP. A Prtp-spoVG-lacZ fusion was inserted into the chromosome of a strain in which RTP production was inducible by IPTG. Addition of IPTG to cultures of the new strain lowered beta Gal production by a factor of at least four. It is concluded that rtp is autoregulated in vivo at the level of transcription.