The ability to repair damaged replication forks and restart them is important for cell survival. DnaT is essential for replication restart in vitro and yet no definite genetic analysis has been done in Escherichia coli K-12. To begin, dnaT822, an in-frame six-codon (87–92) deletion was constructed. DnaT822 mutants show colony size, cell morphology, inability to properly partition nucleoids, UV sensitivity, and basal SOS expression similar to priA2::kan mutants. DnaT822 priA2::kan double mutants had phenotypes similar to those of the single mutants. DnaT822 and dnaT822 priA2::kan mutant phenotypes were fully suppressed by dnaC809. Previously, a dominant temperature-sensitive lethal mutation, dnaT1, had been isolated in E. coli 15T−. DnaT1 was found to have a base-pair change relative to the E. coli 15T− and E. coli K-12 dnaT genes that led to a single amino acid change: R152C. A plasmid-encoded E. coli K-12 mutant dnaT gene with the R152C amino acid substitution did not display a dominant temperature-sensitive lethal phenotype in a dnaT+ strain of E. coli K-12. Instead, this mutant dnaT gene was found to complement the E. coli K-12 dnaT822 mutant phenotypes. The significance of these results is discussed in terms of models for replication restart.
THE ability to reload the replication machinery at repaired replication forks, away from the site of regulated cell-cycle-dependent initiation of DNA replication, has become an important theme in the way prokaryotic and eukaryotic cells replicate their genomes (von Hippel 2000; Cox 2001; Sherratt 2003). At the core of this process is the ability to recognize a broken replication fork, repair it, and then reload or restart the replication machinery. Since replication fork collapse can occur at any place on the chromosome, replication restart must be sequence independent and occur in such a way as to reload the replication machinery in the same direction in which it was replicating before the collapse.
Much of what is understood about this process at the biochemical level has come from the study of the in vitro loading of the DnaB replicative helicase by the DnaC protein on the single-stranded ΦX174 chromosome (reviewed in Kornberg and Baker 1992; Marians 1992, 1996). For this to occur, four proteins, originally called n, n′, n″, and i, and now known to be the PriB, PriA, PriC, and DnaT proteins, respectively, need to be sequentially assembled onto the primosome assembly site (PAS) of the single-stranded ΦX174 chromosome. These proteins target the DnaBC complex to the repaired replication fork and aid DnaC to load DnaB onto the chromosome. This model system has been expanded to show that the same proteins can load DnaB at recombinational intermediates and replication fork-like structures (Liu and Marians 1999; Liu et al. 1999; Marians 2000).
An early inference of the ΦX174 primosome assembly pathway was that PriA, PriB, PriC, and DnaT would be essential for DNA replication in vivo. This was first tested genetically by the construction of mutations in priA (Lee and Kornberg 1991; Nurse et al. 1991). Such mutants were viable but extremely sick. PriA mutant phenotypes include sensitivity to UV irradiation and rich media; deficiency in recombination, stable DNA replication, and chromosome partitioning; and high basal levels of SOS expression (Lee and Kornberg 1991; Nurse et al. 1991; Masai et al. 1994; Kogoma et al. 1996; Sandler et al. 1996; McCool and Sandler 2001). In that these phenotypes seemed reminiscent of a recombination-deficient mutant, understanding these phenotypes has led to a better understanding of how DNA replication and recombination are linked in the cell (Cox et al. 2000; Kowalczykowski 2000). The ΦX174 system also incorrectly predicted that mutations in priB and priC by themselves would have no phenotypic consequence for an otherwise wild-type cell and that mutations in both of these genes would be synthetically lethal (Sandler et al. 1999). These observations and others based on the synthetic lethality between priA and priC and the ability of different dnaC mutations to indirectly suppress the absence of either just priA (dnaC809) or priA and priC (dnaC809,820) were fitted to a multiple pathway model for replication restart (Sandler 2000; Figure 1) . This model describes replication restart at repaired replication forks on the chromosome vs. loading of the replication machinery on a single-stranded DNA phage (Sandler 2000, 2001; Sandler and Marians 2000).
The dnaT gene codes for a 20-kD protein. By itself, this protein has no known intrinsic activity. It is, however, essential for the sequential loading of PriA, PriB, and PriC onto several DNA substrates so that DnaC can mediate the loading of the DnaB helicase (reviewed in Kornberg and Baker 1992; Marians 1992, 1996). In this reaction, the PriA protein first binds the DNA substrate. The PriB protein then facilitates the loading of the DnaT protein (Liu et al. 1996). Increasing the amount of DnaT in a reaction can circumvent the need for PriB in that reaction and rescue the defects of certain mutant PriA proteins (Liu et al. 1996). Once associated with the PriA-DNA complex, the DnaT protein is stably maintained in the complex through the remainder of the reaction (Ng and Marians 1996b). The protein is thought to act in this reaction as a trimer (Ng and Marians 1996b). While PriA, PriB, and DnaT are essential for this reaction, PriC plays a poorly understood, nonessential role. Elimination of PriC from this reaction decreases the loading of DnaB and thus DNA replication by a third (Ng and Marians 1996a; Xu and Marians 2003).
Given the differences between the predictions based on the ΦX174 in vitro DNA replication system and the observed phenotypes associated with priA, priB, and priC mutations, it was of importance to make a mutant of the dnaT gene and test its role in replication restart in vivo. The dnaT gene is located at 99 min on the Escherichia coli K-12 chromosome and resides in an operon with dnaC and two uncharacterized genes, yjjA and yjjB (Figure 2) . While there are no published reports of E. coli K-12 dnaT null mutants, dnaT1, a temperature-sensitive dominant mutation has been isolated in E. coli 15T− by Lark et al. (1978). The early interpretation of the function of the dnaT gene based on the study of this mutant was that DnaT was involved in termination of DNA replication (Lark et al. 1978).
To begin an analysis of the dnaT gene in E. coli K-12, an 18-bp, six-codon (87–92) in-frame deletion of dnaT was created. This article describes the construction and initial phenotypic characterization of this mutant, dnaT822. It was found for each phenotype tested that the dnaT822 mutant behavior was similar to that of a priA2::kan mutant. These similarities included poor growth and viability, sensitivity to UV irradiation, high basal levels of SOS expression, and two populations of cells when viewed microscopically: one wild type and one filamentous with poorly partitioning nucleoids. In some cases, the dnaT822 mutant phenotype was slightly more severe than the priA2::kan mutant phenotype. A priA2::kan dnaT822 double mutant had the same phenotype of either single mutant. The dnaT822 single- and priA2::kan dnaT822 double-mutant phenotypes were fully suppressed by the priA2::kan extragenic suppressor, dnaC809. Lastly, the dnaT sequence of the E. coli 15T− strain carrying the dnaT1 mutation was determined. This dnaT gene showed one difference from the predicted amino acid sequence of the wild-type dnaT from E. coli 15T− or K-12: a change of an arginine residue for a cysteine residue at codon 152 (R152C). The construction of this mutant dnaT gene onto a plasmid and its introduction into E. coli K-12 failed to create a temperature-sensitive phenotype.
MATERIALS AND METHODS
Bacterial strains and plasmids:
All bacterial strains used in this work are derivatives of E. coli K-12 or E. coli 15T− (Table 1) . Plasmids used in this work are shown in Table 2 . Strains for all experiments were grown in either 56/2 minimal medium (Willetts et al. 1969) or Luria-Bertani medium. Minimal medium for UT205 and UT3062 was supplemented with 0.4 μg/ml thymine. Antibiotics were added when appropriate to the following final concentrations: kanamycin (25 μg/ml), tetracycline (10 μg/ml), ampicillin (25 μg/ml), and chloramphenicol (25 μg/ml).
Construction of dnaT+ plasmids:
pJDM14 was used during the initial cloning and mutagenesis of dnaT. PCR primers prSJS197 (5′-AATGGTCGCGCTGGTATCGG) and prSJS198 (5′-GCATTCATCTGCCCTTCACAC) were used to amplify the dnaT coding region and native dnaT promoter (Figure 2). The 1.1-kb PCR product was inserted into the SmaI site of pUC18, resulting in pJDM4. The DNA sequence of this 1.1-kb PCR fragment was verified by DNA sequencing. A SphI fragment containing a gene encoding spectinomycin resistance (SpecR) from plasmid pSJS985 was inserted at the HpaI site of pJDM4, resulting in pJDM14. pJDM59 is a pACYC184 derivative and was used for complementation studies. PCR primers prSJS524 (5′-GGTGCGTTATCTAGAGGTCTTTCCATTCC) and prSJS523 (5′-GGAAGACATACTGTTTCTCATATATGGAATCATGAACGAGAAGGG) were used to amplify the dnaT gene with its native promoter (dnaTp). The PCR product was inserted into pCR2.1 (Invitrogen, Carlsbad, CA) by TOPO TA cloning methods (Invitrogen; pJDM58). A BamHI-XhoI fragment from this plasmid was mixed with a BamHI-SalI fragment of pACYC184 and treated with T4 DNA ligase. The resulting plasmid was called pJDM59.
Construction of plasmids containing dnaT822 and placement of dnaT822 on the chromosome:
pJDM14 was restricted with NcoI and BspEI and treated with Klenow and T4 DNA ligase to produce pJDM16. This protocol deleted 18 nucleotides in the middle of the dnaT coding region corresponding to amino acids 87–92 (GKFAMY). pJDM16 was restricted with EcoRI-BamHI and treated with DNA polymerase I Klenow fragment in the presence of dNTPs to blunt the ends. The 2.4-kb EcoRI-BamHI fragment from pJDM16 was then combined with pBR322 restricted with SphI and treated with DNA polymerase I Klenow fragment as above. The two fragments were then treated with T4 DNA ligase to produce pJDM20. pJDM20 was restricted with BamHI and SalI and the fragment that contained the dnaT region was combined with similarly restricted pKO3. The fragments were then treated with T4 DNA ligase to produce pJDM21. This plasmid was used to transfer dnaT822 to the chromosome according to the method of Link et al. (1997). Briefly, this involves the formation of a co-integrate created by a recombination event between the dnaT locus on the chromosome and the dnaT region that has been inserted on the plasmid by the selection of chloramphenicol resistance at 42° (pKO3 encodes the cat gene and a temperature-sensitive replication gene). Resolution of the co-integrate occurs by recombination. This time, however, the recombination event is on the other side of the dnaT822 mutation. This is done by growing the cells containing the co-integrate in the absence of chloramphenicol at 30° and selecting individual colonies on solid Luria media containing 5% sucrose (pKO3 also encodes the counterselectable gene, sacB). The presence of dnaT822 on the chromosome of SS306 was confirmed by PCR and restriction analysis of the PCR product. The 18-bp deletion creates a BglI site at the site of the deletion that is diagnostic of dnaT822. The sequence of dnaT822 in this strain was confirmed by DNA sequence analysis.
Construction of plasmids containing the E. coli K-12 dnaT gene encoding an amino acid change R152C:
The gene sequence directing the expression of a mutant dnaT protein with the R152C amino acid change was cloned into pACYC184 using PCR. The crossover PCR strategy (Coppi et al. 2001) required first the amplification of two fragments of the dnaTC region such that each fragment had an overlapping region of 20 bp that included the introduction of the coding sequence for the R152C amino acid change. PCR primers prSJS524 (5′-GGTGCGTTATCTAGAGGTCTTTCCATTCC) and prSJS526 (5′-CCGCCGTTGCTGGCACAACCGATTTGC) were used to amplify the first fragment. This was a portion of the dnaT region spanning just upstream of the dnaT promoter to ∼460 bp within the dnaT gene. The second fragment was amplified using primers prSJS525 (5′-CCGCCGTTGCTGGCACAACCGATTTGC) and prSJS523 (5′-GGAAGACATACTGTTTCTCATATATGGAATCATGAACGAGAAGGG). This fragment spanned from approximately the 440 bp position to the middle of the dnaC gene. The two fragments were then mixed and 10 rounds of thermocycling were performed in the absence of primers, followed by 30 rounds in the presence of prSJS524 and prSJS523. The full-length PCR product was inserted into pCR2.1 using the TOPO TA cloning protocol (Invitrogen). This sequence was verified by DNA sequencing. A BamHI-XhoI fragment from this vector was combined with BamHI-SalI-digested pACYC184 and treated with T4 DNA ligase to form pJDM57.
Phenotypic analyses of mutant strains:
The phenotypic analyses of cell viability, UV resistance, and SOS expression have been described elsewhere (Sandler et al. 1999). Cells were prepared for microscopy as described previously (McCool and Sandler 2001). Microscopy was carried out using a Nikon E600 microscope. An ORCA-ER cooled CCD camera (Hamamatsu) and Openlab software (Improvision) were used for all image acquisition and processing. The lengths of 500 individual cells were measured. Filaments are defined as a cell >6 μm in length (McCool and Sandler 2001).
Construction of the dnaT822 mutant:
The dnaT gene is the second of four genes (yjiB, dnaT, dnaC, and yjjA) that share an overlapping pattern of transcription regulation (Masai and Arai 1988). Promoters and termination sites have been mapped. The first and last genes in this group, yjjB and yjjA, have not yet been genetically characterized and are of unknown function. The third gene, dnaC, is an essential gene in E. coli. DnaC loads the DnaB helicase at two different types of protein-DNA complexes: DnaA at oriC (Bramhill and Kornberg 1988) and assemblages of combinations of PriA, PriB, PriC, Rep, and DnaT at repaired replication forks (see Bramhill and Kornberg 1988; Masai and Arai 1988). To create a mutation of dnaT that was not polar on dnaC expression, an in-frame deletion of 18 bp was constructed. This procedure removed codons 87–92 (see materials and methods). This deletion mutation, called dnaT822, was transferred to the chromosome of a dnaC809,820 strain using the method of Link et al. (1997). As mentioned above, dnaC809,820 indirectly suppresses the absence of proteins in both the PriA-dependent and the PriA-independent pathways (Sandler 2000). A dnaC priA2::kan suppressor was used because it was predicted that a dnaT null mutant would have a phenotype like that of a priA2::kan mutant and would be difficult to recover in a dnaC+ strain. DnaC809,820 was used instead of dnaC809 in the event that dnaT may have an unpredicted role in the PriA-independent pathway. The sequence of dnaT822 on the chromosome was verified by DNA sequencing (data not shown). The dnaT822 mutation was then separated from dnaC809,820 by recombination after P1 transduction (see Table 1).
Phenotypes of dnaT822:
The multiple pathways model (Figure 1) predicted that mutations in dnaT would have a phenotype similar to that of priA2::kan mutants. As mentioned above, priA2::kan mutants have pleiotropic phenotypes. Some of these include: small colony size, high basal levels of SOS expression, sensitivity to UV irradiation, and two types of cells when viewed microscopically (wild-type-like cells and filaments with poorly segregating nucleoids). To assess whether dnaT822 phenotypes are like those of a priA2::kan mutant, a strain containing only dnaT822 was subjected to a series of tests. The data (Table 3 ; Figure 3) show that a dnaT822 mutant, like a priA2::kan mutant, has a small colony phenotype, about eight-fold higher basal levels of SOS expression, and is sensitive to UV irradiation. Finally, microscopic analysis revealed that cells from a culture of a dnaT822 strain, again like a priA2::kan strain, could be separated into two classes, wild-type cells and filaments with poorly partitioning nucleoids, in almost the same proportions as had been reported for priA2::kan cultures. In the cases of colony size and UV sensitivity (Table 3), the phenotypes of dnaT822 appeared slightly more severe (Figure 3). It is concluded that except for small quantitative differences, the phenotypes produced by dnaT822 are very similar to those produced by priA2::kan.
The dnaC gene is immediately downstream of the dnaT gene. As mentioned above, dnaC is essential for initiation of DNA replication at oriC and replication restart. Even though the dnaT822 mutation was designed not to be polar on dnaC, it is still possible that the phenotypic effects elicited by this mutation could be produced by some unknown effect of dnaT822 on the expression of dnaC. Additionally, the dnaT822 deletion mutation removes only six codons from the middle of the dnaT gene. It is therefore likely that this mutant gene produces a polypeptide that is similar to the dnaT+ protein. DnaT performs its role in the cell as a member of a complex of proteins. It is possible that dnaT822 produces its phenotypic effect either by inactivating the dnaT gene product so that it cannot interact with the other primosomal proteins during replication restart or by allowing a partially active DnaT protein to interact with the other replication restart proteins and thereby somehow poison the complex. If the former is true, then dnaT822 would be recessive to dnaT+. If the latter is true, then dnaT822 would dominant to dnaT+. To test both of these ideas, dnaT+ was cloned into pACYC184 (see materials and methods for construction of pJDM59) and used to transform a strain with dnaT822 on the chromosome. Figure 3 shows that this strain has wild-type colony size and cell morphology. Therefore, dnaT+ on a plasmid is able to complement dnaT822 for the mutant phenotypes tested.
We conclude that it is likely that the phenotypes seen in a dnaT822 mutant are not due to any polar effect on dnaC and that dnaT822 behaves as a mutation that is recessive to dnaT+. Other interpretations of this result are discussed below.
PriA and dnaT are in the same pathway:
The phenotypic analyses of dnaT822 and priA2::kan mutants show that these two mutants have similar phenotypes. Liu and Marians have shown that PriA-PriB-DnaT and PriA-DnaT complexes can assemble on PAS in vitro and that either complex is competent for ΦX174 replication (Liu et al. 1996). Taken together, these findings suggest that priA and dnaT may be in the same pathway for replication restart in vivo. This hypothesis makes two predictions. First, a priA2::kan dnaT822 double-mutant strain should have phenotypes like either of the single-mutant strains. Second, a dnaC mutation that suppresses priA2::kan phenotypes should also suppress dnaT822 phenotypes.
To test the first prediction, a priA2::kan dnaT822 double mutant (SS467) was constructed (Figure 4) . The donor in the cross was zji-202::Tn10 dnaT822 and the recipient was priA2::kan dnaC809. Tetracycline-resistant transductants were selected. It was predicted that, in most transductants, dnaC809 in the recipient would be replaced by dnaC+ from the donor since dnaT and dnaC are adjacent genes and tightly linked with zji-202::Tn10. Transductants of two different sizes (large and small) were seen. PCR analysis revealed that, of the small colony isolates tested, all were dnaT822 priA2::kan (dnaC+) double mutants (crossovers at 1 and 4 in Figure 4). Analysis of some of the large colony isolates indicated that these were dnaT822 priA2::kan dnaC809 (crossovers at 1, 2, 3, and 4 in Figure 4). One of the resulting dnaT822 priA2::kan double mutants (SS467) was inspected for colony size (Figure 4) and cell morphology (data not shown). These phenotypes were exactly like those of either single mutant (Figure 3), supporting the notion that dnaT and priA function in the same pathway of replication restart.
To test the second prediction of whether dnaC809 could suppress the phenotypes of a dnaT822 by itself, dnaT822 and dnaC809 (the donor was SS466) were introduced simultaneously into DM4000 and JC13509 (see Table 1). From the data summarized in Table 3 and Figure 5 , it is clear for the phenotypes tested (colony size, UV survival, and basal levels of SOS expression) that dnaC809 rescued the dnaT822-associated phenotype in each case. It was additionally tested whether or not priA was required for this suppression by analysis of the dnaT822 priA2::kan dnaC809 triple mutant (SS466) constructed above. SS466 formed wild-type-sized colonies and had wild-type nucleoid organization and wild-type basal levels of SOS expression (Table 3 and data not shown). We conclude that dnaC809 suppresses the effects of dnaT822 and that this does not depend on the presence or absence of the priA gene product.
A dnaT missense mutant (R152C) does not cause temperature sensitivity in E. coli K-12:
Lark et al. (1978) isolated the first mutation in dnaT in a strain of Escherichia coli 15T−. DnaT1 produced a temperature-sensitive dominant lethal phenotype. At the time, DnaT was hypothesized to have a function important for termination of DNA replication. Given that dnaT822 and priA2::kan mutants are very similar phenotypically and that DnaT is essential for replication restart in vitro, it can be hypothesized that dnaT1 may somehow stop the process of replication restart at the nonpermissive temperature. To begin to test this idea, the dnaT genes from UT3062 (dnaT+) and UT205 (dnaT1) were determined to ascertain the mutation associated with dnaT1. Table 4 shows that there was one nucleotide difference between dnaT+ (from E. coli 15T−) and dnaT1. The C-to-T change at base number 454 caused an arginine residue at codon 152 to be replaced by a cysteine residue (R152C). The fact that this base-pair change was the only difference between the two genes suggests that the cause of the temperature-sensitive dominant phenotype of dnaT1 mutant strains was this amino acid change. Comparison with the dnaT+ gene from E. coli K-12 reveals that the dnaT1 mutation from E. coli 15T− would also cause the same change in amino acid sequence. There were four other nucleotide positions within the dnaT coding region where the dnaT+ gene from E. coli 15T− was different from the E. coli K-12 dnaT+ gene (Table 4). None of these differences, however, led to a corresponding change in amino acid sequence.
To test whether the R152C amino acid change would create a temperature-sensitive dominant lethal phenotype in E. coli K-12, this single change was combined with the E. coli K-12 dnaT gene and inserted into pACYC184 (see materials and methods). The resultant plasmid, pJDM57, was used to transform dnaT+ and dnaT822 strains. It was found that pJDM57 (dnaT R152C) complemented the defects of the dnaT822 strain at 30° and 42° (data not shown). Additionally, it did not adversely affect the viability of an E. coli K-12 dnaT+ strain at 42° compared to 30° (data not shown).
We conclude that the only mutation in the dnaT gene of the E. coli 15 T− strain carrying the temperature-sensitive dominant mutation, dnaT1, is a single base-pair change that causes an arginine residue to be replaced by a cysteine residue at codon 152. When this mutation was combined with an E. coli K-12 dnaT gene on a plasmid, it did not cause temperature-sensitive growth in a dnaT+ or dnaT822 strain of E. coli K-12. This mutation appears to cause no adverse affects to the dnaT gene since it was able to complement the dnaT822 mutant.
Several biochemical experiments have implicated the DnaT protein of E. coli K-12 in replication restart. Previously, the genetic requirement for dnaT in replication restart had not been tested. This article describes an 18-bp, six-codon in-frame deletion mutation in the dnaT gene that causes phenotypes similar to those caused by priA2::kan mutations. These include poor growth, UV sensitivity, high basal levels of SOS expression, and two populations of cells when viewed microscopically: one wild type and one with filaments that show poorly partitioning nucleoids. In some cases, the phenotypes caused by dnaT822 were found to be slightly more severe than the corresponding priA phenotypes. The parallels in phenotypes of dnaT822 mutants, vs. slight quantitative differences, suggest an overwhelming similarity to priA2::kan mutants. These similarities are further demonstrated by the observation that dnaC809, an indirect suppressor of priA2::kan, also suppresses the absence of dnaT (and priA in the cell). The mechanistic repercussions of these observations will be discussed below. These data suggest not only that dnaT has an important role in replication restart, but also that its role is similar to that of priA.
DnaT822 is a deletion mutation removing six codons from the middle of the dnaT gene. The mutation was constructed so as to disturb the expression of the essential downstream gene, dnaC, as little as possible. The presence of dnaT+ on a multicopy plasmid in the dnaT822 mutant completely restored the wild-type phenotype. The simplest interpretation of these results is that dnaT822 is a recessive, complete loss-of-function mutation. More complicated interpretations of the data are also possible. Since dnaT822 deletes only six codons, it is possible that dnaT822 could be a partial activity mutant. If so, then it is possible that a mutant that removed all activity would result in cell death instead of the poor viability and the other phenotypes seen. As mentioned above, it is likely that DnaT acts in the cell as part of a complex with PriA and PriB, and it was considered that dnaT822 may exert its phenotypic effect as a dominant mutant poisoning the activity of the complex. While this was tested by the complementation analysis, it is possible that the complementation seen is due to the multiple copies of the dnaT+ gene on a plasmid and would not be observed if dnaT+ and dnaT822 were in equal copy in the cell. Further experiments can test the validity of these more complicated explanations.
Preliminary results available at the time of the writing of the article that introduced the multiple pathways model of replication restart (Sandler 2000) suggested that DnaT acted in the PriA-PriB and PriA-PriC pathways and would not be required in the PriC-Rep-dependent, PriA-independent pathway. The data within showing that dnaT822 alone and dnaT822 priA2::kan have the same phenotype and that both mutants are fully suppressed by dnaC809 support this model with respect to dnaT's contributions to the different pathways (see Figure 1). The model in Figure 1 suggests that DnaT functions after PriA and PriB (PriA-PriB pathway) or PriA and PriC (PriA-PriC pathway). This suggestion is made on the basis of several biochemical studies by Marians and colleagues of the loading of PriA, PriB, and DnaT into the primosome (see Introduction). While the genetic evidence presented here is consistent with this model, it is also consistent with a model in which DnaT and PriA form a complex and then interact with PriB or PriC. There is, however, only limited biochemical evidence to support this model (Liu et al. 1996). Future experiments will distinguish between these models.
The isolation and characterization of dnaT1 by Lark and colleagues has been an intriguing facet of the DNA replication field for some time. While the DnaT gene product was originally interpreted as having a role in termination of DNA replication (Lark et al. 1978), biochemical analysis has clearly implicated this protein in replication restart. It was therefore attractive to hypothesize that dnaT1 may be a mutation that inhibits replication restart rather than causing termination of DNA replication at the nonpermissive temperature. To begin to test this idea, the sequence of the dnaT1 gene from the original E. coli 15 T− was determined and found to have a single amino acid change (R152C) from the wild-type gene in either E. coli 15T− or E. coli K-12. A dnaT gene on a plasmid containing this mutation was constructed to test whether this dnaT mutant would create a temperature-sensitive dominant lethal phenotype in E. coli K-12. It was found that this plasmid did not confer a temperature-sensitive dominant lethal phenotype and instead complemented a dnaT822 mutant. While these data show that a plasmid-encoded dnaT gene with a mutation causing the R152C amino acid change is not sufficient to cause the temperature-sensitive dominant lethal phenotype in E. coli K-12 (as it did in E. coli 15T−), they do not definitively test whether the R152C mutation is the single cause of the temperature-sensitive dominant lethal phenotype found in the UT205 strain. A simple explanation for this series of observations is that the temperature-sensitive dominant lethal phenotype of this mutation is dependent on other aspects of the E. coli 15T− physiology compared to the E. coli K-12 physiology. Several differences have been reported between these physiologies (Alikhanian et al. 1966; Arber and Wauters-Willems 1970). Another explanation is that some other closely linked mutation is required, possibly in addition to the R152C change, to produce the temperature-sensitive phenotype. Other models are also possible.
This work was supported by grant RPG-99-194-04-GMC from the American Cancer Society.
- Received December 4, 2003.
- Accepted February 22, 2004.
- Genetics Society of America