Features of the Chromosomal Terminus Region
Chapter
100
THOMAS M. HILL
The convergence of two replication forks at the end of the DNA replication cycle sets in motion a series of events that are necessary for complete separation of the two newly replicated chromosomes prior to cell division. The first event is termination of DNA replication, which is defined for the purposes of this chapter as the meeting of the two replication forks and the completion of the daughter chromosomes. Termination is followed by decatenation, which removes the catenated links that result from replication termination and link the daughter chromosomes (90). In some cases, decatenation is insufficient to physically separate the daughter chromosomes, because the chromosomes are covalently joined as dimers. In these instances, site-specific or homologous recombination resolves the dimers into monomers. Following complete separation of the daughter chromosomes, the final event takes place: the active segregation of the chromosomes into daughter cells. That these events occur in the terminus region, located 180° around the circular chromosome from the origin, is not just happenstance. The terminus regions of the chromosomes of Escherichia coli and Salmonella typhimurium (official designation, Salmonella enterica serovar Typhimurium) contain sites which trap replication forks, ensuring that these events are always confined to the chromosomal terminus. As our knowledge has increased, it has become clear that the terminus region contains features that distinguish it from other parts of the chromosome and that facilitate the terminal events of chromosome separation.
This chapter describes our current understanding of known features of the terminus region, including the replication arrest sites, which limit the end of the DNA replication cycle to the terminus region; the dif-XerCD resolvase system, which promotes site-specific recombination to resolve dimeric chromosomes; and hyperrecombination in the terminus region. The issues of decatenation and chromosome segregation are dealt with in chapter 105 of this volume.
Masters and Broda (71) and Bird et al. (10) introduced the concept of a replication terminus by demonstrating that the two replication forks initiated from oriC (min 84) meet on the opposite side of the E. coli chromosome between the trp (min 28) and his (min 44) loci. Kuempel et al. (57, 58, 59) and Louarn et al. (66, 67) showed that the region in which the converging replication forks meet contains an impediment to replication fork progression such that replication forks traveling in clockwise or counterclockwise are inhibited between trp (min 28) and manA (min 36). The identification of a barrier to replication fork progression suggested that termination events are always confined to this section of the chromosome, called the terminus region, and that this region might contain features which facilitate the decatenation and partitioning of the daughter chromosomes.
Attempts to identify genetic elements in the terminus region that are involved in specific cellular processes initially met with limited success, despite the construction of cotransduction (11, 25) and restriction (14, 15) maps. In fact, the terminus region of the E. coli chromosome still remains something of an enigma. With the exception of the dif locus (discussed below), the entire E. coli terminus region can be deleted without significantly affecting cell growth, an indication that terminus region genes, including the barrier to replication forks, are not essential (35). Relatively few genetic loci have been identified, even though the number of expressed genes per unit length in the terminus region is equivalent to that of other regions of the chromosome (72). Furthermore, several of the identified genes, such as dicAB (6) and recE (93), are apparently of bacteriophage origin and reflect the presence of several defective prophage found scattered throughout the region.
The paucity of known genes, the presence of phage remnants, and the apparent dispensability of the terminus region have given the region the reputation of a "genetic graveyard" (a phrase coined by Jean-Pierre Bouché). Nonetheless, several phenomena associated with the terminus region have now been described in molecular terms and have reaffirmed that this region of the chromosome contains unique genetic and structural elements that distinguish it from other chromosomal domains. The first of these elements to be described is the dif locus.
Site-specific recombination in the terminus region plays an important role in separating covalently linked chromosomes. The dif (for deletion-induced filamentation) locus is a DNA sequence located at kb 1600 (Fig. 1) that acts as a substrate for the site-specific recombinases XerC and XerD. dif shares both sequence homology and functional similarity with sites found on plasmids (Table 1), such as the cer, parB, and psi sites of plasmids ColE1, CloDF13, and pSC101, respectively (21, 34, 89). The function of dif is to resolve dimeric chromosomes that result from an uneven number of recombination events between sister chromosomes during the replication cycle. Loss of dif leads to a filamentous phenotype in a subpopulation of cells, presumably because cells containing dimeric chromosomes are unable to partition the chromosomes. The dif locus was discovered by three laboratories, each working independently of the other, each starting from a different point, and all with results pointing to the same spot on the E. coli chromosome.
Table 1Comparison of dif-like recombination sitesa |
Clerget (19) found dif while examining the self-maintenance properties of a transposable element, Tn2350, which is derived from the plasmid R1. Tn2350 can be maintained in circular form in E. coli but is unstable and is lost rapidly in the absence of selective pressure. The minimal region required for Tn2350 self-maintenance has two activities: the ability to promote dimerization of plasmids in a recA mutant background by site-specific recombination and the ability to rescue replication-defective plasmids. The latter activity was thought to be due to a replication origin in the sequence. However, Clerget showed that this property of the self-maintenance region was due not to autonomous replication but rather to a very efficient integration and excision of the self-maintenance sequence with a similar site in the E. coli chromosome. The chromosomal integration site was subsequently mapped to the terminus region, and sequence similarity between the chromosomal site and the Tn2350 site was demonstrated. Furthermore, the self-maintenance properties of Tn2350 were dependent on the xerC gene, which encodes the XerC recombinase.
Kuempel et al. (60) were examining the effects of deletions in the terminus region when they noticed that removal of the region from kb 1552 to 1621 produced cultures with a filamentous phenotype in a subpopulation of the cells. The SOS repair system was also induced in cells with this region of the chromosome deleted, and the filamentous cells displayed abnormal nucleoids. By using targeted deletions, dif was localized further to kb 1600. Introduction of a dif-containing plasmid into a dif mutant strain relieved the filamentous phenotype only when the plasmid was able to recombine into the chromosome, suggesting that dif functioned as a cis-acting site. Furthermore, dif-containing plasmids formed dimers at a high frequency, even in a recA mutant background, suggesting that recombination at the dif locus was recA independent.
Sherratt’s group (12) found dif by way of studies of the site-specific recombination systems of plasmids. Many low-copy-number plasmids carry site-specific recombination systems that resolve dimeric plasmids into monomers, thereby ensuring equipartitioning of the daughter plasmids at cell division. High-copy-number plasmids, which segregate randomly, also use recombination systems to resolve multimeric plasmids, thereby increasing the number of monomeric plasmids and the probability that daughter cells will receive a plasmid at cell division. Sherratt and colleagues had previously shown that ColE1 plasmids contain a site, called cer, that promotes plasmid monomerization and that function of the cer site depends on several host genes (12, 13, 20). One of these genes, xerC, encodes a recombinase of the λ integrase family. When mutants of xerC were examined under the microscope, many of the cells showed a filamentous phenotype and aberrant nucleoid distribution similar to that observed with the Dif phenotype. The sequence homology between cer and dif (Table 1) suggested that xerC might also act at dif. This possibility was confirmed by demonstrating that dif promoted xerC-mediated recombination in plasmids and that XerC specifically bound to short oligomers containing either the cer or the dif sequence.
The chromosomal dif system varies from that of plasmids in that the plasmid systems are more specific and require accessory proteins. Four genes (argR, pepA, xerC, and xerD) are required for the cer site in ColE1 plasmids, but only xerC and xerD are required for dif functioning (13). The ArgR and PepA accessory proteins presumably impart resolution selectivity to cer function in a manner similar to that observed for other site-specific recombination systems, such as the res-resolvase system of Tn3 (88). The resolution specificity provided by the accessory proteins promotes preferential recombination between linked cer sites on multimeric plasmids but not between sites on separate monomeric plasmids. This resolution selectivity does not appear to exist for dif, thereby permitting interconversion between monomers and dimers (12).
The xerC gene maps to kb 4024 (min 85) and is part of a four-gene operon that includes the dapF (diaminopimelate epimerase) gene and two other open reading frames (ORFs). All four genes are probably transcribed from a promoter upstream of dapF, and all are probably expressed to the same level. The xerC gene encodes a protein of 298 amino acids (molecular mass, 33.8 kDa) that contains two regions with high homology to the lambda integrase family of recombinases.
The xerD gene was identified by Blakely et al. on the basis of its homology to the xerC gene (13). XerD shows 37% amino acid identity to XerC and also contains 298 amino acids, suggesting that the genes are closely related. xerD maps to kb 3055 (min 62) and appears to be part of the same transcriptional unit as recJ and xprA. Inactivation of xerD by insertional mutagenesis, as with xerC mutants, abolishes site-specific recombination at dif sites, indicating that XerD is required for dif functioning.
By using an oligomer containing a dif site, it was shown that either XerC or XerD alone is capable of binding specifically to free dif DNA but that each protein binds with a higher affinity to a preformed complex containing dif and the other protein. This suggests that binding of the two recombinases is cooperative (13). Footprinting studies of XerC-dif or XerD-dif complexes also demonstrated that each protein bound specifically to a different half of the palindromic dif site: XerC binds to the left half and XerD binds to the right half of a dif site (Table 1). However, artificial dif sites containing two left ot two right halves of the dif site were inactive, indicating that binding of both XerC and XerD was absolutely required for normal dif functioning.
The dif and xer genes play an essential role in facilitating chromosome partitioning that is distinct from the function of topoisomerases. The catenated chromosomes that form at the end of the DNA replication cycle can be physically separated by topoisomerase IV as long as the sister chromosomes undergo an even number of recombination events during replication and the chromosomes exist as monomers (see chapter 105). However, if an uneven number of recombination events occurs during the replication cycle, removal of the catenates by topoisomerase IV is insufficient to physically separate the resulting dimeric chromosome (Fig. 2). The action of XerCD at the juxtaposed dif loci serves to monomerize the linked chromosomes. Thus, in a dif or xer mutant, the filamentous phenotype observed in a significant fraction of the population is assumed to result from cells containing linked chromosomes that cannot be partitioned.
The lack of resolution specificity at the dif site is somewhat curious, as there appears to be a great advantage in providing a directional bias to this process. Without resolution specificity, the action of XerCD at dif would be expected to dimerize monomeric chromosomes with the same efficiency as dimeric chromosomes are resolved. Thus, continued recombination at dif would be expected to occur regardless of whether the daughter chromosomes were monomers or dimers. What, then, provides directionality to the dif-XerCD system? Blakely et al. (12) proposed that the active segregation of topologically unlinked daughter chromosomes may impart directionality, for as the chromosomes pull apart, the two dif sites become spatially separated and can no longer function as recombination sites. As long as recombination events occur frequently and rapidly at the dif site, dimeric chromosomes will be converted to the monomeric form at some point during segregation and will be sufficiently separated to prevent further recombinational events.
As pointed out by Baker (3), this mechanism for dif resolution specificity works fine as long as dif is located in the last domain of the chromosome to be pulled apart. If reorganization of nucleoid structure accompanies DNA replication, then distinct domains of the nucleoid may be spatially separated quite rapidly following passage of a replication fork. Domains close to oriC would be expected to be segregated faster than terminal domains. Thus, placement of dif anywhere other than the terminal domain of the nucleoid would allow dimeric chromosomes to become trapped in that configuration, preventing chromosome segregation. This suggests that the location of the dif site in the terminus is not a result of happenstance but is a requirement for its function.
Recently, Tecklenburg et al. showed that the function of dif is in fact position dependent. When the normal dif locus is deleted and replaced by a new dif sequence located some 120 kb counterclockwise from the usual site, the new dif locus is unable to suppress the filamentous phenotype, even though the new site still functions as a site for XerCD-mediated recombination (91). Additional attempts to relocate the site to other points in the chromosome by using a dif-containing transposon failed; the only transposon insertions that relieve the dif phenotype are those that are located within 10 kb on either side of the original dif site. Thus, dif appears to be functional only when it is located in the terminal domain of the nucleoid, where decatenation presumably occurs. The catenates present in this region bring the dif sites on daughter chromosomes into juxtaposition and facilitate the monomerization of the linked genomes.
Recently, it was shown that the frequency of recombination in the terminus region is increased relative to that in other regions of the chromosome. It appears that this phenomenon can be attributed in part to the unique position of the terminus region with respect to chromosome replication. However, another contributing factor to hyperrecombination may be Tus-mediated arrest of DNA replication (discussed below), which produces stalled replication forks that may be highly recombinogenic. In this section, the current evidence and possible models for hyperrecombination in the terminus region are discussed.
Increased levels of both homologous and illegitimate recombination occur near Ter sites (reviewed in reference 9). Horiuchi and coworkers found eight recombinational hotspots (called Hot DNA) while searching for the replication origins utilized by E. coli rnh mutants during stable DNA replication. Seven of these Hot DNAs are located in the terminus region between TerA and TerB (75; Fig. 1). These sites do not function as autonomous replicons; rather, they can be isolated as plasmid-like forms because they integrate into and are excised from the chromosome at a high frequency by homologous recombination. All of the Hot DNAs possess chi activity (75). Three of the Hot DNAs are adjacent to Ter sites, and hyperrecombination activity at these sites depends on the presence of a functional tus gene (49). The other four Hot DNAs function independently of tus. In one case (the HotA site located near TerB), mutation of the chi sequence diminishes Hot activity; however, insertion of a new, properly oriented chi site between HotA and the TerB site restores hyperrecombination. It is postulated that arrested replication forks at Ter sites produce double-stranded breaks that act as an entry point for the RecBCD enzyme (49). The RecBCD enzyme then utilizes the chi site to repair the chromosome.
Replication arrest at a Ter site can also promote illegitimate recombination (7). In hybrid plasmids containing both pBR322 and M13 origins of replication, a TerB site acts as a deletion hotspot when it is oriented to arrest DNA replication initiated from either plasmid or phage origin. Approximately 80% of the deletions originating at M13 have endpoints that map within several nucleotides of the start of the TerB sequence, which is the precise point of arrest of leading-strand synthesis (46, 61). Furthermore, the deletion hotspot activity is dependent on the tus gene. Bierne et al. (7) suggest that the arrested replication fork, in conjunction with the nicked M13 origin of replication, produces structures that promote deletion formation.
Louarn et al. (68) found that the terminus has an increased frequency of RecA-mediated homologous recombination. In particular, a region near kb 1600 had a frequency of recombination 1,000-fold higher than that of other parts of the chromosome. This hyperrecombination is not sequence specific but appears to be connected to events that follow termination of DNA replication. Also, this form of recombination is not tus dependent and therefore does not rely on replication arrest (65, 68).
By measuring the excision of prophage from many different points around the genome, the frequency of homologous recombination at many locations on the E. coli chromosome was accurately determined (68). The excision frequency is uniform (10–5 per generation) for most of the chromosome (from min 44 clockwise to min 23), increases significantly (10- to 100-fold) at the edges of the terminus region near TerA and TerB, and reaches a maximum of 10–2 at min 33.8 near the TerC locus. The region of maximal hyperrecombination is called the terminus recombination zone (TRZ) and is located at kb 1600 to 1640 (65). Inactivation of the recA gene abolishes hyperrecombination, as does loss of RecBCD function. This loss of RecBCD function suggests that this region may, in part, coincide with some of the tus-independent Hot DNA sites (49).
Hyperrecombination in the TRZ does not appear to be associated with specific sequence elements, such as chi sites. Deletions that remove the region encompassing the TRZ (and therefore some of the tus-independent Hot DNA sites) do not eliminate terminal hyperrecombination; the zone of maximum recombination is simply shifted in the direction of the deletion. Furthermore, hyperrecombination activity in the TRZ is independent of the dif locus. Deletion of dif or introduction of xerC mutations has no effect on recombination frequency in the TRZ. These observations suggest that hyperrecombination in the TRZ is due to the position of the TRZ in the nucleoid rather than to a special recombination system (65).
Hyperrecombination in the TRZ is not linked to termination of DNA replication, even though the TRZ is proximal to TerC, where termination is believed to occur most frequently (discussed below). The separation of replication termination and TRZ activity was demonstrated by introducing an extra Ter site at kb 1400, thereby shifting the position of replication termination from kb 1600 to kb 1400. Even though the likely meeting point for replication forks was shifted by 200 kb, hyperrecombination still occurred at the same frequency at kb 1600. This indicates that hyperrecombination at TRZ is independent of replication termination and probably takes place after replication has been completed (65).
To explain the phenomenon of the TRZ, it is proposed that after the terminal stages of replication, catenation links between progeny chromosomes are "chased" to the last region of the nucleoid to be organized, which presumably includes the TRZ (65). This would happen regardless of where replication termination occurs. The driving force for chasing the catenation links to the TRZ may be the restructuring of the nucleoid into distinct domains after passage of a replication fork. The reorganization of the nucleoid would physically separate newly replicated domains of the daughter chromosomes, and if replication termination occurred outside of the terminus, the catenation links would be driven into the terminus region as the nucleoid was reorganized. Furthermore, the noninvertable zones that flank the terminus region (27, 28) may contain polarized organizing sequences for nucleoid structure, which could provide a topographical impetus to nucleoid reordering and help force the catenation links to the terminal domain of the nucleoid. As a result of these activities, the catenated progeny chromosomes would be highly susceptible to recombination in the TRZ during the final stages of nucleoid assembly and decatenation.
The physiological advantage of having elevated levels of recombination associated with the terminus region is hard to fathom, because it is difficult to imagine that the cell would benefit from persistently unstable regions of the genome. Nonetheless, there is evidence for genome instability in the terminus, as Masters and Oliver (M. Masters and I. R. Oliver, J. Cell. Biochem. Abstr. Suppl., 17E:300, 1993) showed that homology between the terminus regions of E. coli and other enterobacterial species is less than that at other regions of the chromosome. Also, extensive rearrangements must have occurred, because regions containing several kilobases of DNA that are present in E. coli appear to be missing in other enterobacteria. This suggests that insertions and deletions occur with relative frequency on an evolutionary time scale. Perhaps this is one of the reasons for the apparent dispensability of the terminus region in E. coli; the constant shuffling, deletion, and acquisition of genetic material by both homologous and illegitimate recombination may exclude essential genes from this region.
Perhaps hyperrecombination at the terminus is simply a price bacteria pay for having circular chromosomes. The catenated daughter chromosomes that are formed at the completion of replication serve as willing substrates for homologous recombination. Thus, the high level of terminal recombination at the completion of chromosome replication is probably not the result of design but rather is a consequence of the recombinational machinery being provided with a suitable substrate for significant periods of time.
The terminus regions of the chromosomes of E. coli and S. typhimurium contain specific sites that block the progression of DNA replication forks. These sites, called Ter sites (2), constitute the recognition sequence for the Tus protein, which mediates arrest of the replisome. The binding of Tus to a Ter site forms an asymmetric complex that impedes replication forks approaching from one direction but not the other. Thus, the Ter sites are oriented in the bacterial chromosome to allow completion of the DNA replication cycle, but they prevent replication forks from traveling in the terminus-to-origin direction.
It is important to emphasize that termination of DNA and replication arrest are not synonymous terms, even though the former is often used in place of the latter. Termination of DNA replication is defined as the events that occur when the two replication forks meet and complete the daughter chromosomes. Arrest of DNA replication is defined as the events that occur when a replication fork is stopped by a Tus-Ter complex. That termination of DNA replication and replication arrest are not equivalent has been demonstrated most clearly in strains lacking a tus gene. In these cells, termination of DNA apparently occurs in the absence of a functional replication arrest system, indicating that the events required for completion of the daughter chromosomes can proceed without replication forks being blocked at a Ter site.
The Ter sites in the E. coli chromosome were identified by using strains in which DNA replication was initiated from plasmid or phage origins integrated near the terminus region (57, 58, 59, 66, 67). This approach avoided problems associated with initiation of replication from oriC, in which both replication forks arrive at the terminus region at approximately the same time. In the initial studies, the barrier to replication was localized to a broad area of the terminus region. Clockwise traveling replication forks were halted between rac (min 30) and aroD (min 37), and counterclockwise replication forks were arrested between rac and trp (min 28) (58, 59, 66, 67).
The impediment to DNA replication in the terminus region was separated into two distinct loci in Louarn’s (24) and Kuempel’s (44) laboratories. Each of the two sites was specific for the arrest of either clockwise or counterclockwise replication forks, and the two sites were well separated at the edges of the terminus region. The sites functioned in a polar fashion; that is, the Ter sites halted replication forks approaching from one direction but not the other. Both groups suggested that replication forks are arrested only transiently at the Ter sites; the arrested replication forks eventually continue past the arrest sites. The TerA site, which halts counterclockwise replication forks only, is located near pyrF at 28.5 min. A second site, which halts only clockwise replication forks, was mapped differently by the two groups; deMassy et al. (24) placed TerB at min 33.5, and Hill et al. (44) placed it between min 34.5 and 35.7, near manA. It was shown later that this apparent discrepancy reflects the presence of two distinct arrest sites for clockwise replication (TerC and TerB, respectively) on this side of the terminus region. The polarity of the Ter sites and their positions at the edges of the terminus region suggest that they function as a trap, allowing replication forks to enter but not exit this region.
Additional arrest sites were subsequently found in and around the terminus region (26, 39, 40, 84). Six sites are asymmetrically distributed over approximately 25% of the total chromosome and are oriented to halt DNA replication moving in the origin-to-terminus direction, with TerB, TerC, and TerF positioned to arrest clockwise replication and TerA, TerD, and TerE positioned to arrest counterclockwise replication (Fig. 3). Recently, we identified a putative seventh Ter site on the basis of sequence similarity. TerG is located at min 49, within one of the genes for the biosynthesis of menaquinine (vitamin K2) (78) and is also oriented to arrest DNA replication forks moving in the terminus-to-origin direction.
Empirical evidence for additional Ter sites exists. François et al. (26) identified 10 different chromosomal fragments containing homology to a oligomeric probe containing the TerA sequence. Most likely, these putative Ter sites are situated outside of the traditional terminus region, as is already the case for TerE, TerF, and TerG, and may be used rarely, if at all, during normal replication cycles.
Ter sites have also been found in E. coli plasmids, most notably in the R6K plasmid (30, 50, 56) and in members of the IncFII compatibility group, such as R100 and R1 (47, 50). In virtually all cases, a pair of Ter sites is present, and these sites are oriented to form a replication fork trap. As in the E. coli chromosome, these Ter sites also depend on the Tus protein for function.
A list of Ter sequences from the E. coli chromosome, S. typhimurium chromosome, and plasmids is given in Table 2. The invariant nucleotides are restricted to the G residue at position 6 and the 11-bp sequence from residues 9 through 19 (39, 47). The one exception to this rule is TerF, which has a G substitution at position 18. The four nucleotides upstream of the core sequence, from positions 2 through 5, are neither invariant nor protected by Tus binding but are generally A/T rich, suggesting that sequence context may influence either Tus binding or activity. A good example of the effect of context can be seen by comparing the binding affinities of Tus for the TerB and R6KTerR2 sites, since both of these sites contain the consensus core sequence yet they show a 30-fold difference in the equilibrium binding constant (31; discussed below). Adjoining sequences outside the standard 23-bp Ter sites may also affect replication arrest activity (8).
Table 2Comparison of known Ter sitesa |
It is worth noting that the Ter sites identified to date are primarily in intergenic regions, where transcriptional activity is lowest. This is probably because Ter sites are able to inhibit transcription (82). The orientation of a Ter site that blocks replication forks also halts RNA polymerase; the opposite orientation is permissive for both. Thus, a Tus-Ter complex located inside a transcriptional unit can interfere with RNA polymerase function if the Ter site is oriented to prevent transcription. By the same token, if the orientation of the Ter site is permissive for transcription, as occurs with the TerF site in the rcsC gene (84), RNA polymerase function would not be affected and the polymerase would displace Tus, possibly reducing the effectiveness of the Ter site. In fact, Tus-Ter complexes in intergenic regions appear to be protected from transcriptional interference by the presence of Rho-independent transcriptional terminators on the permissive side of the sites (39). It is also interesting to note that the TerF site, which is located within the rcsC gene and is oriented in the permissive direction with respect to transcription, is still functional (84). This indicates that transcriptional interference does not necessarily obviate the function of Ter site but may reduce its efficiency.
The possibility that a cell-encoded trans-acting factor is required for replication arrest was first raised in studies examining in vitro replication of hybrid R6K/ColE1 plasmids (4, 30). Deletion analysis of the terminus of the E. coli chromosome subsequently identified a region that contained a gene necessary for function of the Ter sites (45). Deletions that removed the TerB site also inactivated TerA, even though these two sites are separated by approximately 350 kb. Introduction of a plasmid carrying the deleted region near TerB restored TerA activity, suggesting that a trans-acting factor encoded by a gene in the vicinity of TerB is required for replication arrest. The gene is named tus (terminus utilization substance) or, alternatively, tau (53). The region encoding tus has two large ORFs separated by only 75 bp, with the TerB site situated between the two ORFs and just upstream of the second ORF. Insertional inactivation of the second ORF showed it to be the tus gene (42, 48), since loss of this gene was correlated in vitro with loss of a Ter-specific DNA-binding activity and in vivo with loss of the replication arrest function of Ter sites (48, 53).
Fewer than a hundred copies of Tus protein are normally present in a cell (73). This relatively low level of expression is due to several features of the tus gene region. First, the primary promoter for transcription of the tus gene, which is located immediately upstream of the gene, shows only weak homology to the consensus E. coli promoter sequence (82) and is probably transcribed at a low level. Second, the TerB site is located within the promoter region of the tus gene, suggesting that tus expression is autoregulated by binding of Tus to the TerB site (42, 48). Both in vivo (42, 82, 83) and in vitro (74) studies have verified this second supposition.
Upstream of tus and TerB are two genes, called urpT (unidentified regulatory protein/Terminus) and uspT (unidentified sensory protein/Terminus), that appear, on the basis of sequence information (83), to code for potential sensor-regulator proteins. urpT and uspT are not involved in replication arrest, since insertional inactivation of uspT has no effect on the function of the Ter sites (48). Both genes are transcribed in the same direction as tus, and the 3' end of uspT is located only 75 bp from the 5' end of tus. No potential transcriptional terminators are apparent in the intragenic region, and readthrough transcription of tus from uspT occurs in a tus mutant strain (82), albeit at a much lower level than transcription originating from the major tus promoter. However, in a tus + strain, only a very small fraction of the transcripts originating from uspT traverse the tus gene, presumably because of the barrier posed by Tus bound to the TerB site. Thus, the Tus-TerB complex regulates tus expression by two methods: by occluding the major promoter and by impeding the progression of an actively transcribing RNA polymerase originating from upstream promoters.
Just downstream of and transcribed convergently with tus is fumC, the gene for a class II fumarase (33, 95). The tus and fumC genes overlap at their stop codons, and a potential Rho-independent terminator for fumC is positioned within tus (94). This suggests that transcription of fumC may interfere with tus transcription, but the effect of fumC on the expression of tus has not been examined.
The relatively long in vitro half-life of the Tus-TerB complex (31) suggests that expression of tus may be coordinated with the cell cycle, since displacement of Tus from the TerB site would be expected to occur only when a replication fork had passed through the tus gene region. However, a recent examination of tus expression during synchronized chromosomal replication did not reveal periodicity in the level of tus transcripts (C. E. Helmstetter, J. A. Bogan, P. Zhou, P. W. Theisen, H.-J. Wang, K. Welch, and J. Grimwade, Abstracts: EMBO Workshop on the Bacterial Cell Cycle, p. 41–43).
The sequence of the tus gene predicts a highly basic protein of 309 amino acids (35,780 molecular mass). No significant homology to the binding motifs of other known DNA-binding proteins is found in Tus (48). Thanks to the advent of overexpression vectors and the relative ease of Tus purification (42, 46, 86), many of the biochemical and physical properties of Tus have been determined (Table 3). Two of these properties deserve comment. First, in combination with the lack of dyad symmetry in the Ter sequence, the fact that Tus binds as a monomer (22, 85) ensures the polar function of the protein-DNA complex, because a different face of the Tus protein will be exposed to the replication fork depending on the direction from which the replication apparatus approaches. Second, the observed isoelectric point of the native protein is 7.5 (22), which is considerably lower than the pI of 10.1 calculated from the nucleotide sequence of the tus gene (48). The difference between the observed and predicted pIs is presumably due to the folding of the native protein, since no posttranslational modifications of the protein that could account for this difference have been reported.
Table 3Physical and biochemical characteristics of Tus protein |
Tus has a high affinity for the TerB site in vitro, with an observed equilibrium binding constant of 3.4 × 10–13 M and a half-life of approximately 550 min for the Tus-TerB complex in optimal buffer conditions (50 mM Tris-Cl at pH 7.5, 150 mM potassium glutamate) (31). This half-life exceeds by a factor of 100 that of the Lac repressor-operator complex (52), which is considered a relatively tightly binding protein. Studies of the interactions of Tus with another chromosomal site, TerF, and with the plasmid R6KTerR2 site (a less efficient replication arrest site) (39, 80) demonstrated a 10- to 30-fold-lower affinity for these sites, with observed Kd s of 10–12 M (84) and 10–11 M (31), respectively. The reduced affinities of Tus for these two sites is primarily due to a faster dissociation rate of the Tus-Ter complexes, with half-lives of 46 and 43 min for TerF and R6KTerR2, respectively.
Footprinting studies of Tus bound to the chromosomal TerB site show an asymmetric distribution of protein-DNA contacts (31), which is consistent with the observed polarity of function (Fig. 4). The pattern of protected residues suggests that Tus spans the major and minor grooves of the Ter site, contacting primarily only one face of the double helix, and that a flexible arm of the protein extends around to the back of the helix in the major groove. Also, the region in which Tus appears to contact both strands of the double helix is on the side of the complex that impedes replication forks, whereas on the other side, which allows free passage of replication forks, only a single strand is in contact with Tus. As discussed below, this feature of Tus binding in conjunction with the stability of the protein-DNA complex may account for the observed orientation-dependent inhibition of a broad range of helicases in in vitro assays.
A protein that appears to specifically abrogate the arrest of DNA replication by Tus was reported by Natarajan et al. (73). This protein, called the anti-ter, was purified from E. coli cell extracts on the basis of its ability to "supershift" a Tus-Ter complex during gel retardation electrophoresis. When anti-ter was added to an in vitro replication system derived from crude cell extracts, the ability of Tus to halt replication of a plasmid containing a functional Ter site was significantly reduced. Likewise, the anti-ter protein prevented Tus-mediated inhibition of DnaB helicase activity in a helicase assay that used a short oligomeric substrate. The mode of action of the anti-ter is not known, but the protein does not appear to function by displacing Tus from the Ter site, suggesting that the anti-ter may interfere with the interaction between Tus and the replication fork. The in vivo role of the anti-ter is also unknown, but possibilities for its function might include bypassing of Tus-Ter complexes during recombination, DNA repair, or conjugation.
Over the last few years, the molecular mechanism by which Tus exerts its unique ability to block the progression of replication forks has been the primary focus of studies on the Tus-Ter interaction. These studies have employed in vitro systems to attempt to identify the component of the replication apparatus Tus interacts with and to understand how Tus blocks DNA replication. The results from these studies support the conclusion that the target of Tus is the replicative helicase, DnaB, and that Tus prevents DnaB from unwinding the DNA strands ahead of the replication fork. In addition, Tus appears to exert its effect through protein-protein interactions with DnaB. Although the accumulated evidence clearly favors these views, conclusions about Tus function should be viewed with a modicum of skepticism, as no genetic evidence supporting the biochemical results exists.
Tus acts alone in arresting DNA replication, as shown by its ability to do so in an in vitro replication system composed entirely of purified proteins (46, 62). In such a system, Tus function depends on the orientation of the Ter site, faithfully mimicking its in vivo activity. These studies also demonstrated that the primary arrest sites of leading-strand replication are the first and second nucleotides of the Ter site (46, 61), which suggests that the DNA polymerase III holoenzyme is able to replicate very close to the Tus protein before being inhibited. However, these studies were unable to pinpoint the target of Tus; for this purpose, a simplified assay was required.
Because the replicative helicase is a likely target for Tus action, a strand displacement assay has been used to measure the effect of a Tus-Ter complex on helicase function. In this assay, a short, radiolabeled oligomer (30 to 60 bases) containing a Ter site is hybridized to a circular single-stranded DNA molecule to form a partial DNA heteroduplex. This helicase substrate is then incubated with Tus to form the Tus-Ter complex, and a purified DNA helicase is added to the mixture. The intrinsic unwinding activity of the helicase displaces the oligomer, which can be easily separated from the heteroduplex substrate by electrophoresis. The advantages of this assay are that (i) the direct action of Tus on the helicase can be determined because no other replication proteins are necessary and (ii) the effect of Tus on the activity of a wide variety of helicases can be tested with this simple protocol.
This assay showed that Tus acts as a polar barrier to helicases of both prokaryotic and eukaryotic origins and to helicases that unwind DNA in either the 5'→3' or the 3'→5' direction (Table 4). These observations suggest that Tus acts as an antihelicase: (i) it prevents strand separation by these helicases, and (ii) Tus function depends on the orientation of the Ter site, which is consistent with the observed polarity of Tus in vivo. The helicase assay also showed that a Tus-Ter complex inhibits strand displacement by DNA polymerases from T5 and T7 phages and the E. coli polymerase I large fragment (61). However, in these cases, inhibition is independent of the orientation of the Ter site. The endpoints of T7 DNA polymerization coincide almost exactly with the limits of the Tus footprint at the TerB site, suggesting that the tight binding of Tus to theTerB site is sufficient to halt these polymerase activities.
Table 4DNA-translocating proteins inhibited by Tus-Ter complexes |
If Tus functions as an antihelicase, how does it inhibit helicase activity? The variety of DNA-translocating proteins blocked by Tus suggests that Tus functions as a general barrier to the progression of these proteins, a notion first advanced by Lee et al. (62). However, conflicting results regarding the activity of Tus against certain helicases, such as Rep and UvrD, confused the issue (Table 4). Also, some helicases, including the Dda helicase of T4 phage (5) and, with the appropriate helicase substrates, DnaB, are not inhibited at all by Tus (37). These observations suggest that Tus may function via specific protein-protein interactions with its target (51).
Recently, compelling evidence that supports the protein-protein interaction model has been presented. Hiasa and Marians (37) demonstrated a differential ability of Tus to inhibit DnaB depending on the type of helicase substrate used in the assay. They compared Tus function on helicase substrates with either a short or a long oligomer hybridized to the single-stranded circle. With helicase substrates containing a short (60-base) TerB oligomer, inhibition of DnaB was orientation dependent, which is consistent with previous reports. However, when the helicase substrate contained a TerB oligomer of approximately 250 bp, DnaB passed through the Tus-TerB complex in either orientation, indicating that the barrier posed by the Tus-Ter complex is insufficient to stop the progression of the helicase. To account for this, Hiasa and Marians postulated that assays using short oligomeric substrates measure only strand displacement activity rather than true helicase activity. On longer duplex substrates, strand displacement activity gives way to true DNA unwinding, a transition that occurs somewhere between turns 2 and 10 in the duplex DNA. Thus, the characteristics of Tus binding to TerB account for the orientation-dependent, generalized inhibition of helicase strand displacement on short oligomeric substrates. However, the Tus-TerB complex cannot halt DnaB once true duplex unwinding has started, suggesting that protein-protein interactions must be an integral step in Tus-mediated arrest of DNA replication.
The model that Tus mediates replication arrest through protein-protein interactions has been strengthened further by the isolation of two classes of mutant Tus proteins. One class has a reduced affinity for the TerB site yet can still arrest DNA replication in vivo with a relatively high efficiency (87). One of these mutants, TusA173V, has a much shorter half-life for the Tus-Ter complex (t 1/2 = 4.5 min versus 150 min for wild-type Tus in buffer containing 200 mM potassium glutamate), yet arrests DNA replication at approximately 75% of the efficiency of wild-type tus. A second class of Tus mutants show more normal binding affinities yet demonstrate a markedly reduced ability to arrest DNA replication in vivo. The most interesting of this class of mutants is TusE49K, which shows reduced arrest function, yet has a half-life that is even greater than that of the wild type (t 1/2 = 180 min). This strongly suggests that binding to DNA alone cannot account for the ability of the Tus-Ter complex to halt replication forks.
If specific protein-protein interactions are required for Tus function, how is Tus able to inhibit such a broad array of proteins in the helicase assay? The answer probably lies in the tight binding of Tus and suggests that the inhibition of many of these helicases may be an artifact of the helicase assay. Using the stability of the Tus-Ter complex and the asymmetry of the pattern of contacts of Tus to the Ter site, Gottlieb et al. proposed a model for how Tus could function as a polar barrier to helicase unwinding (31). Although this model does not reflect the true mechanism by which Tus halts replication forks, it may explain the ability of Tus to nonspecifically inhibit unwinding activity in the helicase assay.
Even though our understanding of the mechanism of Tus-mediated termination of DNA replication has advanced considerably, little progress has been made in understanding the true biological function of the Tus-Ter complex. Other than the simple explanation that Ter sites serve as one-way gates to prevent replication in the origin-to-terminus region, no other major physiological function has been ascribed to the replication arrest system. In large part, this ignorance stems from the lack of biological consequences following loss of tus gene function, even though several attempts have been made to determine whether tus + cells have a selective advantage. For instance, Roecklein et al. (83) compared the growth characteristics of tus + and tus::Kanr strains under a variety of conditions, including growth in rich and minimal media, medium shifts, anaerobic conditions, UV irradiation, and exposure to agents that inhibit DNA synthesis. None of these conditions allowed tus + cells to significantly outgrow tus null mutants.
Another unknown is the frequency of replication cycles that terminate with replication forks arrested at any one of the known Ter sites, nor is it known how often the two replication forks simply collide without the use of a Ter site. In vitro replication studies using oriC plasmids containing a pair of Ter sites arranged as on the chromosome showed that virtually all replication cycles end with replication forks stalled at one or the other Ter site (38). However, because of the small size of the plasmid (5 kb), the frequency of Ter site usage may be unusually high. In the chromosome, replication cycles may not show the same bias for Ter site usage, as replication forks must travel over 2 Mb before arriving at the terminus.
Of the known Ter sites, TerC is probably used most often because it is almost equidistant from oriC (Fig. 3), and the region of the chromosome just upstream of TerC is replicated predominantly in the clockwise direction (65, 68). The frequent use of TerC may be significant, as the dif locus (12, 19, 60) and the TRZ (65) are both located adjacent to it. However, it is also clear that other Ter sites, such as TerA and TerB, are utilized in a substantial fraction of replication cycles. In exponentially growing cells, a significant population of the cells have replication forks arrested at either TerA and TerB, indicating that these sites are used frequently (79). Also, strains containing chromosomal inversions that double the distance from oriC to TerB relative to the distance from oriC to TerA use TerA and TerB at almost the same frequency. This was a surprise, because it should take twice as long to replicate the clockwise arm of the chromosome in this strain as the counterclockwise arm. This observation has not been satisfactorily explained.
Replication arrest sites located well outside of the traditional terminus region, such as TerE and TerF, are probably not used in normal replication cycles because the replication fork would have to pass through at least two other Ter sites located closer to the terminus region before reaching either of these sites (Fig. 3). For example, in the studies that identified TerF, replication arrest at TerF could be demonstrated only when the region containing the TerC and TerB sites had been deleted, indicating that replication forks only rarely escape from these sites (84). The anticipated discovery of additional Ter sites outside the terminus region poses the question of the purpose of these nonterminus sites. Perhaps they are utilized when replication forks are initiated at origins other than oriC. Under these circumstances, these Ter sites would prevent replication in the terminus-to-origin direction but allow replication to proceed normally along the origin-to-terminus axis. Potential replication origins other than oriC are the sdr origins (1), prophage origins, or replication origins of integrated plasmids (see chapter 99 in this volume).
To date, the only examples of loss of tus being correlated with a phenotype are in E. coli strains in which completion of DNA replication is delayed or blocked by Ter sites. In one case, the DNA replication cycle was delayed because the chromosome was replicated predominantly in one direction (76); in the other case, inverted Ter sites, oriented to prevent completion of replication, were artificially introduced into the terminus region (B. Sharma and T. M. Hill, Mol. Microbiol., in press). In the first instance, oriC was inactivated by insertion of the origin from plasmid R1, which then substituted as the chromosomal origin. In strains whose R1 origin is oriented to replicate in the counterclockwise direction (called intR1CC), the chromosome is replicated primarily in one direction; in strains with the opposite orientation of the R1 plasmid (intR1CW), replication is bidirectional, presumably because the replication fork passing through oriC initiates the counterclockwise fork (23). These intR1CC strains form long filaments, which subsequently pinch off small, anucleate cells (76). Presumably, the phenotype of the intR1CC strains results from the inability of the unidirectional replication fork to finish replication in coordination with cell division, since the replication fork has to pass through a series of active termination sites before the chromosome is completed. In agreement with this model, inactivation of the tus gene reduces filamentation (23).
In the second example, the DNA replication cycle of E. coli was disrupted by inserting inverted Ter sites into the terminus region of the chromosome to prematurely arrest DNA replication (Sharma and Hill, submitted). A pair of inverted Ter sites was inserted into the chromosome of a Δ tus strain at approximately min 31, between TerA and TerC. The resulting strain construction was called the InvTer strain. A functional tus gene was then introduced by transforming the InvTer strain with a plasmid carrying a copy of the tus gene under the control of an arabinose-inducible promoter. Induction of tus gene expression in the InvTer strain activates the inverted Ter sites and arrests replication forks after they complete all but the last 2 kb of the chromosome. In the absence of arabinose, tus expression is repressed and cell morphology is normal, but in its presence, the InvTer cells form filaments. This suggests that activation of the inverted Ter sites by induction of tus gene expression arrests DNA replication prior to completion of the chromosome and delays the onset of cell division.
Filamentation of InvTer cells is due to induction of the sfiA pathway normally associated with SOS induction and the sfi-independent pathway (18) as well. Thus, in replication cycles that are asymmetric, with one replication fork arriving at the terminus in advance of the other, arrest of the first fork may induce cell division inhibitors to prevent initiation of cell division until the second replication fork arrives and completes the chromosome. This function would primarily be a fine-tuning mechanism for coordinating DNA replication and cell division, and as a result, loss of tus gene expression would produce an uncoordinated cell cycle in only the small fraction of the population that had highly asymmetric replication cycles.
An intriguing hypothesis that links RecA activity to the function of the Tus-Ter system was presented by Zyskind et al. (97), who proposed that RecA may be activated by arrested replication forks at the Ter sites. Their hypothesis was based on their own observations, which suggested that RecA was involved with chromosome partitioning, and the observations of others, who suggested that the RecA dependence of certain integratively suppressed strains may be connected to replication arrest at the Ter sites (69; S. Dasgupta, R. Bernander, and K. Nordström, unpublished results). Zyskind et al. speculated that RecA is required for completion of the replication cycle when replication forks are blocked by a Tus-Ter complex; loss of RecA function leads to the production of anucleate cells. While this hypothesis is intriguing, examining it directly has not been possible. Also, a recent study questions the importance of RecA function following arrest of replication at a Ter site. Kogoma et al. (55) studied the RecA dependence of constitutive stable DNA replication and were unable to establish a link between RecA dependence and Tus activity. Thus, the relationship of RecA and replication arrest remains an interesting but uncertain proposition.
Despite our inability to identify a clear tus phenotype, the importance of arrest of DNA replication at sequence-specific sites is attested to not only by the high degree of similarity between the DNA sequences and the Tus proteins of E. coli and S. typhimurium, but also by the ever-increasing number of similar systems that have been identified in other organisms. The gram-positive bacterium Bacillus subtilis contains a functionally analogous system, in terms of both replication arrest components and their organization in the chromosome (for a review, see reference 43 or 96) In Saccharomyces cerevisiae, replication pausing has been associated with specific parts of the rRNA gene regions (16, 54, 63) and the centromeres (32) and is believed to depend on specific protein-DNA complexes (17, 32). Replication arrest is also associated with sequences in the rRNA genes of peas (36) and humans (64) and has been observed in bovine mitochondrial DNA at conserved sequences called termination-associated sequences, which appear to function by binding a trans-acting factor (70).
If the replication arrest system does not confer an evolutionary advantage, significant differences between the replication arrest proteins and the Ter sites of S. typhimurium and E. coli might have developed in the 160 to 180 million years of evolution that separate the two species (77). To examine the functional similarity of the replication arrest systems of the two bacteria, Roecklein et al. (83) introduced a plasmid containing a functional E. coli TerA site into S. typhimurium and showed that plasmid replication was halted at the Ter site. Furthermore, when a similar plasmid containing a TerA site in the nonfunctional orientation was introduced into S. typhimurium, plasmid replication was not halted, indicating that the polarity of Ter site function was also conserved. Hybridization with DNA probes containing the E. coli TerB site or part of the tus gene revealed a minimum of three Ter sites in the S. typhimurium genome as well as a region that hybridized strongly to the E. coli tus gene. As with E. coli, Ter sequences can be found in plasmids, such as the Salmonella 90-kb virulence plasmid (29, 92).
The terminus regions of the chromosomes of E. coli and S. typhimurium are inverted with respect to one another, but it is reasonable to assume that the general organization of their Ter sites is similar. The inversion only exchanges the relative positions of the Ter sites but does not change their orientation with respect to the direction of DNA replication. That Ter site organization is conserved has been confirmed by the relationship of the TerA site and the pyrF gene in both bacteria. In E. coli, the TerA site is located 156 bp upstream of the pyrF start codon. In S. typhimurium, even though the pyrF gene and its upstream region show only 76 and 67% homology, respectively, to the E. coli sequence (83), the TerA site is perfectly conserved 155 bp upstream of the start of the pyrF gene, a testament to the importance of this site. However, not all sites are conserved, as witnessed by the divergence in the amyA gene (min 43), which has a putative Ter site in the 3' end of the coding sequences in S. typhimurium but not in E. coli (29, 81). A possible explanation for the lack of conservation of this particular region may be that Ter sites located within the traditional terminus region, which would be expected to be used most often, might be more highly conserved than Ter sites located at other points around the chromosome.
A growing body of evidence suggests that the last domain of the nucleoid to be replicated and organized is special compared to other nucleoid domains. The uniqueness of the terminal domain is exemplified by the fact that two of its most interesting features, the TRZ and the dif locus, acquire function by virtue of their position in the chromosome and not by their DNA sequences. Thus, the TRZ (65) apparently exists regardless of what DNA sequences occupy the zone, and the dif locus (12, 19, 60) functions as a dimer resolution site only when it is located in the terminal domain of the nucleoid. In this context, the primary function of the Ter sites is to confine the terminal events of replication to the terminus region, thereby facilitating the final events of chromosome separation. This also explains the apparent dispensability of the replication arrest system. Loss of the Ter sites does not prevent the terminal replication events from occurring, but in the long run, such a loss may very well reduce the overall efficiency of cell growth and division and, consequently, the evolutionary fitness of the organism.
I gratefully acknowledge all colleagues who shared in the preparation of the manuscript by providing information in advance of publication. I also extend thanks to Peter Kuempel and Rolf Bernander, who provided critical commentary on the contents of this chapter.
This work was supported by grants from the National Institutes of Health (GM43193) and the American Cancer Society (NP-828).
References
1. Asai, T., S. Sommer, A. Bailone, and T. Kogoma. 1993. Homologous recombination-dependent initiation of DNA replication from DNA damage-inducible origins in Escherichia coli. EMBO J. 12:3287–3295.
2. Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8. Microbiol. Rev. 54:130–197.
3. Baker, T. A. 1991. And then there were two. Nature (London) 353:794–795.
4. Bastia, D., J. Germino, J. H. Crosa, and J. Ram. 1981. The nucleotide sequence surrounding the replication terminus of plasmid R6K. Proc. Natl. Acad. Sci. USA 78:2095–2099.
5. Bedrosian, C. L., and D. Bastia. 1991. Escherichia coli replication terminator protein impedes simian virus 40 (SV40) DNA replication fork movement and SV40 large tumor antigen helicase activity in vitro at a prokaryotic terminus sequence. Proc. Natl. Acad. Sci. USA 88:2618–2622.
6. Béjar, S., F. Bouché, and J.-P. Bouché. 1988. Cell division inhibition gene dicB is regulated by a locus similar to lamboid bacteriophage immunity loci. Mol. Gen. Genet. 212:11–19.
7. Bierne, H., S. D. Ehrlich, and B. Michel. 1991. The replication termination signal terB of the Escherichia coli chromosome is a deletion hot spot. EMBO J. 10:2699–2705.
8. Bierne, H., S. D. Ehrlich, and B. Michel. 1994. Flanking sequences affect replication arrest at the Escherichia coli terminator terB in vivo. J. Bacteriol. 176:4165–4167.
9. Bierne, H., and B. Michel. 1994. When replication forks stop. Mol. Microbiol. 13:17–23.
10. Bird, R. E., J. Louarn, J. Martuscelli, and L. Caro. 1972. Origin and sequence of chromosome replication in Escherichia coli. J. Mol. Biol. 70:549–566.
11. Bitner, R. M., and P. L. Kuempel. 1981. P1 transduction map spanning the replication terminus of Escherichia coli K12. Mol. Gen. Genet. 184:208–212.
12. Blakely, G., S. Colloms, G. May, M. Burke, and D. Sherratt. 1991. Escherichia coli XerC recombinase is required for chromosomal segregation at cell division. New Biol. 3:789–798.
13. Blakely, G., G. May, R. McCulloch, L. K. Arciszewska, M. Burke, S. T. Lovett, and D. Sherratt. 1993. Two related recombinases are required for site-specific recombination at dif and cer in E. coli K12. Cell 75:351–361.
14. Bouché, J. P. 1982. Physical map of a 470 kilobase pair region flanking the terminus of DNA replication in the Escherichia coli K12 genome. J. Mol. Biol. 154:1–20.
15. Bouché, J. P., J. P. Gélugne, J. Louarn, J. M. Louarn, and K. Kaiser. 1982. Relationships between the physical and genetic maps of a 470 × 103 base-pair region around the terminus of Escherichia coli K12 DNA replication. J. Mol. Biol. 154:21–32.
16. Brewer, B. J., and W. L. Fangman. 1988. A replication fork barrier at the 3' end of yeast ribosomal RNA gene. Cell 55:637–643.
17. Brewer, B. J., D. Lochshon, and W. L. Fangman. 1992. Arrest of replication fork barrier in the rDNA of yeast occurs independently of transcription. Cell 71:267–276.
18. Burton, P., and I. B. Holland. 1983. Two pathways of division inhibition in UV-irradiated E. coli. Mol. Gen. Genet. 190:309–314.
19. Clerget, M. 1991. Site-specific recombination promoted by a short DNA segment of plasmid R1 and by a homologous segment in the terminus region of the Escherichia coli chromosome. New Biol. 3:780–788.
20. Colloms, S. D., P. Sykora, G. Szatmari, and D. J. Sherratt. 1990. Recombination at ColE1 cer requires the Escherichia coli xerC gene product, a member of the lambda integrase family of site-specific recombinases. J. Bacteriol. 172:6973–6980.
21. Cornet, F., I. Mortier, J. Patte, and J.-M. Louarn. 1994. Plasmid pSC101 harbors a recombination site, psi, which is able to resolve plasmid multimers and to substitute for the analogous chromosomal Escherichia coli dif site. J. Bacteriol. 176:3188–3195.
22. Coskun-Ari, F. F., A. Skokotas, G. R. Moe, and T. M. Hill. 1994. Biophysical characteristics of Tus, the replication arrest protein of Escherichia coli. J. Biol. Chem. 269:4027–4034.
23. Dasgupta, S., R. Bernander, and K. Nordström. 1991. In vivo effect of the tus gene on cell division in an Escherichia coli strain where chromosome replication is under control of plasmid R1. Res. Microbiol. 142:177–180.
24. deMassy, B., S. Bejar, J. Louarn, J. M. Louarn, and J. P. Bouché. 1987. Inhibition of replication forks exiting the terminus region of the Escherichia coli chromosome occurs at two loci separated by 5 min. Proc. Natl. Acad. Sci. USA 84:1759–1763.
25. Fouts, K. E., and S. D. Barbour. 1982. Insertions of transposons through the major cotransduction gap of Escherichia coli K-12. J. Bacteriol. 149:106–113.
26. François, V., J. Louarn, and J.-M. Louarn. 1989. The terminus region of the Escherichia coli chromosome is flanked by several polar replication pause sites. Mol. Microbiol. 3:995–1002.
27. François, V., J. Louarn, J. Patte, J.-E. Rebollo, and J.-M. Louarn. 1990. Constraints in chromosomal inversions in Escherichia coli are not explained by replication pausing at inverted terminator-like sequences. Mol. Microbiol. 4:537–542.
28. François, V., J. Louarn, J.-E. Rebollo, and J.-M. Louarn. 1990. Replication termination, nondivisible zones, and structure of the Escherichia coli chromosome, p. 351–359. In K. Drlica and M. Riley (ed.), The Bacterial Chromosome. American Society for Microbiology, Washington, D.C.
29. Friedrich, M. J., N. E. Kinsey, J. Vila, and R. J. Kadner. 1993. Nucleotide sequence of a 13.9 kb segment of the 90kb virulence plasmid of Salmonella typhimurium: the presence of fimbrial biosynthesis genes. Mol. Microbiol. 8:543–558.
30. Germino, J., and D. Bastia. 1981. Termination of DNA replication in vitro at a sequence-specific replication terminus. Cell 23:681–687.
31. Gottlieb, P. A., S. Wu, X. Zhang, M. Tecklenburg, P. Kuempel, and T. M. Hill. 1992. Equilibrium, kinetic, and footprinting studies of the Tus-ter protein-DNA interaction. J. Biol. Chem. 267:7434–7443.
32. Greenfeder, S. A., and C. S. Newlon. 1992. Replication forks pause at yeast centromeres. Mol. Cell. Biol. 12:4056–4066.
33. Guest, J. R., and G. C. Russell. 1992. Complexes and complexities of the citric acid cycle in Escherichia coli. Curr. Top. Cell. Regul. 33:231–247.
34. Hakkart, M. J. J., P. J. M. van der Elzen, E. Veltkamp, and N. J. J. Nijkamp. 1984. Maintenance of multicopy plasmid CloDF13 in E. coli cells: evidence for site-specific recombination at parB. Cell 36:203–209.
35. Henson, J. M., and P. L. Kuempel. 1985. Deletion of the terminus region (340 kilobase pairs) from the chromosome of Escherichia coli. Proc. Natl. Acad. Sci. USA 82:3766–3770.
36. Hernández, P., S. S. Lamm, C. A. Bjerknes, and J. Van’t Hof. 1988. Replication termini in the rDNA of synchronized pea root cells (Pisum sativum). EMBO J. 7:303–308.
37. Hiasa, H., and K. Marians. 1992. Differential inhibition of the DNA translocation and DNA unwinding activities of DNA helicases by the Escherichia coli Tus protein. J. Biol. Chem. 267:11379–11385.
38. Hiasa, H., and K. Marians. 1994. Tus prevents overreplication of oriC plasmid DNA. J. Biol. Chem. 269:26959–26968.
39. Hidaka, M., M. Akiyama, and T. Horiuchi. 1988. A consensus sequence of three DNA replication terminus sites on the E. coli chromosome is highly homologous to the terR sites of the R6K plasmid. Cell 55:467–475.
40. Hidaka, M., T. Kobayashi, and T. Horiuchi. 1991. A newly identified DNA replication terminus site, terE, on the Escherichia coli chromosome. J. Bacteriol. 173:391–393.
41. Hidaka, M., T. Kobayashi, Y. Ishimi, M. Seki, T. Enomoto, M. Abdel-Monem, and T. Horiuchi. 1992. Termination complex in E. coli inhibits SV40 DNA replication in vitro by impeding the action of T antigen helicase. J. Biol. Chem. 267:5361–5365.
42. Hidaka, M., T. Kobayashi, S. Takenaka, H. Takeya, and T. Horiuchi. 1989. Purification of a DNA replication terminus (ter) site-binding protein in Escherichia coli and identification of the structural gene. J. Biol. Chem. 264:21031–21037.
43. Hill, T. M. 1992. Arrest of bacterial DNA replication. Annu. Rev. Microbiol. 46:603–633.
44. Hill, T. M., J. M. Henson, and P. L. Kuempel. 1987. The terminus region of the Escherichia coli chromosome contains two separate loci that exhibit polar inhibition of replication. Proc. Natl. Acad. Sci. USA 84:1754–1758.
45. Hill, T. M., B. J. Kopp, and P. L. Kuempel. 1988. Termination of DNA replication in Escherichia coli requires a trans-acting factor. J. Bacteriol. 170:662–668.
46. Hill, T. M., and K. J. Marians. 1990. Escherichia coli Tus protein acts to arrest the progression of DNA replication forks in vitro. Proc. Natl. Acad. Sci. USA 87:2481–2485.
47. Hill, T. M., A. J. Pelletier, M. Tecklenburg, and P. L. Kuempel. 1988. Identification of the DNA sequence from the E. coli terminus region that halts replication forks. Cell 55:459–466.
48. Hill, T. M., A. J. Pelletier, M. Tecklenburg, and P. L. Kuempel. 1989. tus, the trans-acting gene required for termination of DNA replication in Escherichia coli, encodes a DNA-binding protein. Proc. Natl. Acad. Sci. USA 86:1593–1597.
49. Horiuchi, T., Y. Fujimura, H. Nishitani, T. Kobayashi, and M. Hidaka. 1994. The DNA replication fork blocked at the ter site may be an entrance for the RecBCD enzyme into duplex DNA. J. Bacteriol. 176:4656–4663.
50. Horiuchi, T., and M. Hidaka. 1988. Core sequence of two separable terminus sites of the R6K plasmid that exhibit polar inhibition of replication is a 20 bp inverted repeat. Cell 54:515–523.
51. Khatri, G. S., T. MacAllister, P. R. Sista, and D. Bastia. 1989. The replication terminator protein of E. coli is a DNA sequence-specific contra-helicase. Cell 59:667–674.
52. Khoury, A. M., H. J. Lee, M. Lillis, and P. Lu. 1990. lac repressor-operator interaction: DNA length dependence. Biochim. Biophys. Acta 1087:55–60.
53. Kobayashi, T., M. Hidaka, and T. Horiuchi. 1989. Evidence of a ter specific binding protein essential for the termination reaction of DNA replication in Escherichia coli. EMBO J. 8:2435–2441.
54. Kobayashi, T., M. Hidaka, M. Nishizawa, and T. Horiuchi. 1992. Identification of a site required for DNA replication fork-blocking activity in the rRNA gene cluster in Saccharomyces cerevisiae. Mol. Gen. Genet. 233:355–362.
55. Kogoma, T., K. G. Barnard, and X. Hong. 1994. RecA, Tus protein and constitutive stable DNA replication in Escherichia coli rnhA mutants. Mol. Gen. Genet. 244:557–562.
56. Kolter, R., and D. Helinski. 1978. Activity of the replication terminus of plasmid R6K in hybrid replicons in Escherichia coli. J. Mol. Biol. 124:425–441.
57. Kuempel, P. L., and S. A. Duerr. 1979. Chromosome replication in Escherichia coli is inhibited in the terminus region near the rac locus. Cold Spring Harbor Symp. Quant. Biol. 43:563–567.
58. Kuempel, P. L., S. A. Duerr, and P. D. Maglothin. 1978. Chromosome replication in an Escherichia coli dnaA mutant integratively suppressed by prophage P2. J. Bacteriol. 134:902–912.
59. Kuempel, P. L., S. A. Duerr, and N. R. Seeley. 1977. Terminus region of the chromosome in Escherichia coli inhibits replication forks. Proc. Natl. Acad. Sci. USA 74:3927–3931.
60. Kuempel, P. L., J. M. Henson, L. Dircks, M. Tecklenburg, and D. F. Lim. 1991. dif, a recA-independent recombination site in the terminus region of the chromosome of Escherichia coli. New Biol. 3:799–811.
61. Lee, E. H., and A. Kornberg. 1992. Features of replication fork blockage by the Escherichia coli replication termination-binding protein. J. Biol. Chem. 267:8778–8784.
62. Lee, E. H., A. Kornberg, M. Hidaka, T. Kobayashi, and T. Horiuchi. 1989. Escherichia coli replication termination protein impedes the action of helicases. Proc. Natl. Acad. Sci. USA 86:9104–9108.
63. Linskens, M. H. K., and J. A. Huberman. 1988. Organization of replication of ribosomal DNA in Saccharomyces cerevisiae. Mol. Cell. Biol. 8:4927–4935.
64. Little, R. D., T. H. K. Platt, and C. L. Schildkraut. 1993. Initiation and termination of DNA replication in human rRNA genes. Mol. Cell. Biol. 13:6600–6613.
65. Louarn, J., F. Cornet, V. François, J. Patte, and J.-M. Louarn. 1994. Hyperrecombination in the terminus of the E. coli chromosome: possible relation to nucleoid organization. J. Bacteriol. 176:7524–7531.
66. Louarn, J., J. Patte, and J.-M Louarn. 1977. Evidence for a fixed termination site of chromosome replication in Escherichia coli K12. J. Mol. Biol. 115:295–314.
67. Louarn, J., J. Patte, and J.-M. Louarn. 1979. Map position of the replication terminus on the Escherichia coli chromosome. Mol. Gen. Genet. 172:7–11.
68. Louarn, J.-M., J. Louarn, V. François, and J. Patte. 1991. Analysis and possible role of hyperrecombination in the termination region of the Escherichia coli chromosome. J. Bacteriol. 173:5097–5104.
69. Mao, Y.-M., Q. Sho, Q.-G. Li, and Z.-J., Sheng. 1991. recA dependence of replication of the Escherichia coli chromosome initiated by plasmid pUC13 integrated at predetermined site. Mol. Gen. Genet. 225:234–240.
70. Masden, C. S., S. C. Ghivizzani, and W. W. Hauswirth. 1993. Protein binding to a single termination-associated sequence in the mitochondrial DNA D-loop region. Mol. Cell. Biol. 13:2162–2171.
71. Masters, M., and P. Broda. 1971. Evidence for the bidirectional replication of the Escherichia coli chromosome. Nature (London) New Biol. 232:137–140.
72. Moir, P. D., R. Spiegelberg, I. R. Oliver, J. H. Pringle, and M. Masters. 1992. Proteins encoded by the Escherichia coli terminus region. J. Bacteriol. 174:2102–2110.
73. Natarajan, S., S. Kaul, A. Miron, and D. Bastia. 1993. A 27kd protein of E. coli promotes antitermination of replication in vitro at a sequence-specific replication terminus. Cell 72:113–120.
74. Natarajan, S., W. L. Kelly, and D. Bastia. 1991. Replication terminator protein of Escherichia coli is a transcriptional repressor of its own synthesis. Proc. Natl. Acad. Sci. USA 88:3867–3871.
75. Nishitani, H., M. Hidaka, and T. Horiuchi. 1993. Specific chromosomal sites enhancing homologous recombination in Escherichia coli mutants defective in RNase H. Mol. Gen. Genet. 240:307–14.
76. Nordström, K., R. Bernander, and S. Dasgupta. 1991. Analysis of bacterial cell cycle using strains in which chromosome replication is controlled by plasmid R1. Res. Microbiol. 142:181–188.
77. Ochman, H., and A. C. Wilson. 1987. Evolutionary history of enteric bacteria, p. 1649–1654. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. American Society for Microbiology, Washington, D.C.
78. Palaniappan, C., V. K. Sharma, M. E. S. Hudspeth, and R. Meganathan. 1992. Menaquinone (vitamin K2) biosynthesis: evidence that the Escherichia coli menD gene encodes both 2-succinyl-6-hydroxy-2,4-cyclohexadioene-1-carboxylic acid synthase and 2-ketoglutarate decarboxylase. J. Bacteriol. 174:8111–8118.
79. Pelletier, A. J., T. M. Hill, and P. L. Kuempel. 1988. Location of sites that inhibit progression of replication forks in the terminus region of Escherichia coli. J. Bacteriol. 170:4293–4998.
80. Pelletier, A. J., T. M. Hill, and P. L. Kuempel. 1989. Termination sites T1 and T2 from the Escherichia coli chromosome inhibit DNA replication in ColE1-derived plasmids. J. Bacteriol. 171:1739–1741.
81. Rapa, M., I. Kawagishi, V. Müller, M. Kihara, and R. M. Macnab. 1992. Escherichia coli produces a cytoplasmic α-amylase, AmyA. J. Bacteriol. 174:6644–6652.
82. Roecklein, B., and P. L. Kuempel. 1992. In vivo characterization of tus gene expression in Escherichia coli. Mol. Microbiol. 6:1655–1661.
83. Roecklein, B., A. J. Pelletier, and P. L. Kuempel. 1991. The tus gene of Escherichia coli: autoregulation, analysis of flanking sequences, and identification of a complementary system in Salmonella typhimurium. Res. Microbiol. 142:169–176.
84. Sharma, B., and T. M. Hill. 1992. terF, the sixth identified replication arrest site in Escherichia coli, is located within the rcsC gene. J. Bacteriol. 174:7854–7858.
85. Sista, P. R., C. A. Hutchison, and D. Bastia. 1991. DNA-protein interaction at the replication termini of plasmid R6K. Genes Dev. 5:74–82.
86. Sista, P. R., S. Mukherjee, P. Patel, G. S. Khatri, and D. Bastia. 1989. A host-encoded DNA-binding protein promotes termination of plasmid replication at a sequence-specific replication terminus. Proc. Natl. Acad. Sci. USA 86:3026–3030.
87. Skokotas, A., M. Wrobleski, and T. M. Hill. 1994. Isolation and characterization of mutants of Tus, the replication arrest protein of Escherichia coli. J. Biol. Chem. 269:20446–20455.
88. Stark, W. M., M. R. Boocock, and D. J. Sherratt. 1989. Site-specific recombination by Tn3 resolvase. Trends Genet. 8:304–309.
89. Summers, D. K., and D. J. Sherratt. 1984. Multimerization of high copy number plasmids causes instability: ColE1 encodes a determinant essential for plasmid monomerization and stability. Cell 36:1097–1103.
90. Sundin, O., and A. Varshavsky. 1980. Terminal stages of SV40 DNA replication proceed via multiply intertwined catenated dimers. Cell 21:103–114.
91. Tecklenburg, M., A. Maumer, O. Nagappan, and P. L. Kuempel. The dif-resolvase locus can be replaced by a 33 basepair sequence, but function depends on location. Proc. Natl. Acad. Sci. USA 92:1352–1356.
92. Tinge, S. A., and R. Curtiss. 1990. Isolation of the replication and partitioning regions of the Salmonella typhimurium virulence plasmid and stabilization of heterologous replicons. J. Bacteriol. 172:5266–5277.
93. WIllis, D. K., L. H. Satin, and A. J. Clark. 1985. Mutation-dependent suppression of recB21 and recC22 by a region cloned from the Rac prophage of Escherichia coli K-12. J. Bacteriol. 162:1166–1172.
94. Woods, S. A., J. S. Miles, R. E. Roberts, and J. R. Guest. 1986. Structural and functional relationships between fumarase and aspartase. Biochem J. 237:547–557.
95. Woods, S. A., S. D. Schwartzbach, and J. R. Guest. 1988. Two chemically distinct classes of fumarase in Escherichia coli. Biochim. Biophys. Acta 954:14–26.
96. Yoshikawa, H., and R. G. Wake. 1993. Initiation and termination of chromosome replication, p. 507–528. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and Other Gram-Positive Bacteria: Biochemistry, Physiology, and Molecular Genetics. American Society for Microbiology, Washington, D.C.
97. Zyskind, J. W., A. L. Svitil, W. B. Stine, M. C. Biery, and D. W. Smith. 1992. RecA protein of Escherichia coli and chromosome partitioning. Mol. Microbiol. 6:2525–2537.
Chapter100
