INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Escherichia coli cell cycle analyses often rely on a temporary inhibition of the initiation stage of chromosome replication in an exponentially growing cell population, such that replication initiation occurs in synchrony when the inhibition is lifted. It is important to identify the conditions under which the synchronization process is optimally efficient, as well as to find out how well the synchrony is maintained during a single cell cycle. With flow cytometry, it is possible to analyze the DNA content of a large number of individual cells during the synchronization period and after reinitiation.

Thermosensitive strains containing mutations in the dnaC gene are often used for synchronization (6, 14, 21-23, 28, 29, 31). The full role of the DnaC protein in the replication process is not clear, but one of its functions is believed to involve the loading of the DnaB helicase onto the prepriming complex during initiation of chromosome replication (17).

Mutations in the dnaC gene fall into two distinct classes: those which block initiation of replication and those which interfere with the elongation stage (30). Members of the former group, which includes dnaC2 (7) and dnaC28 (4, 26), are ideal for cell cycle studies. In a standard experiment, strains containing either dnaC2 or dnaC28 are grown exponentially at 30 or 28.5°C, respectively, and initiation of replication is blocked by shifting the temperature to 40°C or 42°C. Replication is then synchronously restarted by shifting the cultures back to the lower temperature.

Here, we have used flow cytometry to analyze runout of chromosome replication when dnaC mutants were shifted to a nonpermissive temperature, as well as reinitiation of chromosome replication and subsequent cell division after return to the permissive temperature. Antibiotics were used to study the transcription and translation requirements in the mutants under different conditions, and the effects of different cell concentrations at the time of reinitiation were also analyzed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains, media, and growth conditions. All strains used in this study were derivatives of E. coli K-12 (2) and are listed in Table 1. Strains PC2 and NK5184 were kindly provided by William Donachie, and LN681 was provided by Thomas Hill. The dnaC2 and dnaC28 alleles were transferred by P1 transduction (20) from PC2 and LN681, respectively, using a Tn10 insertion in the nearby thr operon as selectable marker. The Tn10 insertion was obtained from NK5184.

1

1

Temperature shift experiments. Bacterial cultures were grown overnight, diluted into fresh medium, and allowed to grow at the permissive temperature for at least five generations. The cell concentration was monitored by measuring the optical density at 600 nm (OD₆₀₀). At an OD₆₀₀ of 0.05, the cultures were shifted to the nonpermissive temperature for 120 min. The cultures were then shifted to the permissive temperature, either (i) for the duration of the experiment or (ii) for 10 min followed by a shift back to the nonpermissive temperature. Variations of this procedure are indicated in Results.

Flow cytometry. For flow cytometry (reviewed in reference 27), 1-ml samples were collected from the cultures and centrifuged at 14,000 × g for 10 min at 4°C. The cell pellet was resuspended in 100 µl of 10 mM Tris buffer (pH 7.5) containing 10 mM EDTA, and the cells were fixed in 1 ml of cold 77% ethanol and stored at 4°C. Prior to measurements, the cells were centrifuged, washed in 1 ml of 10 mM Tris buffer (pH 7.5) containing 10 mM MgCl₂, centrifuged again, and resuspended in 75 µl of the same solution. For DNA staining, an equal volume of stain containing 40 µg of ethidium bromide ml¹ and 200 µg of mithramycin A ml¹ in Tris-MgCl₂ buffer was added to the sample. Sample analysis was carried out on a Bryte-HS flow cytometer (Bio-Rad) equipped with a mercury-xenon arc lamp as a light source. The instrument was calibrated with plastic beads (Bio-Rad) with a diameter of 1.5 µm and labelled with the fluorescent coumarin 6 dye. The beads were uniform, with coefficients of variation of <1.2 and <1.5% for size and fluorescence, respectively.

Number of replication origins. To determine the number of replication origins in each cell at a given time (reviewed in reference 27), either 300 µg of rifampin ml¹ plus 10 µg of cephalexin ml¹ or 200 µg of chloramphenicol ml¹ plus 10 µg of cephalexin ml¹ was added, and the samples were incubated at the permissive temperature for 2 h to allow all replication forks to run to completion. Rifampin and chloramphenicol block initiation of chromosome replication while ongoing rounds of replication continue to termination, and cephalexin prevents cell division during the runout period. After the runout period, the DNA content of the cells was determined by flow cytometry as described above, and the number of chromosome equivalents per cell showed the number of replication origins that were present at the time of drug addition.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

DNA content and coordinated replication initiations in exponentially growing dnaC2 and dnaC28 mutants. The dnaC2 and dnaC28 alleles were transduced from strains PC2 and LN681 (Table 1) into the wild-type MG1655 strain to obtain an isogenic background. The DNA content in the exponential growth phase was examined by flow cytometry both in the MG1655 derivatives and in PC2 and LN681. The dnaC2-carrying derivatives were grown in supplemented minimal medium (see Materials and Methods) at 30°C. The dnaC28-containing strains were grown in LB, since LN681 grew poorly in minimal medium, and at 25°C, since DNA replication in this strain was perturbed at higher temperatures (see below).

View larger version (23K): [in this window] [in a new window]   FIG. 1.   DNA content distributions of wild-type E. coli and dnaC mutants. The strains were grown exponentially and sampled for flow cytometry either directly (left) or after addition of rifampin and cephalexin followed by incubation for 120 min (right). (A) Strains MG1655, MG1655dnaC2, MG1655dnaC28, and PC2 were grown in supplemented M9 medium (see text for details) at 30°C except for MG1655dnaC28, which was grown at 25°C. (B) Strains MG1655dnaC28 and LN681 were grown at 25°C in LB supplemented with glucose and thymine.

Inhibition of initiation of chromosome replication and runout of replication in dnaC2 and dnaC28 mutants shifted to different temperatures. Strain MG1655dnaC2 was grown in supplemented minimal medium at 30°C until an OD₆₀₀ of approximately 0.05 was reached. The culture was then split into several smaller cultures which were incubated at different temperatures at 2°C intervals, and samples for flow cytometry were collected after 60, 90, and 120 min. An increase in temperature of 4°C (to 34°C) was sufficient to cause a significant alteration in the DNA content of the cells (Fig. 2A), presumably due to a less efficient initiation of replication. At higher temperatures, replication initiation was blocked while ongoing replication and subsequent cell division continued, with the result that most cells contained a single chromosome equivalent (Fig. 2A). The lowest temperature at which replication initiation was efficiently blocked was 36°C. A significant proportion, 20 to 25%, of the populations contained two chromosomes after the runout, representing cells that were unable to divide despite completion of replication. Also, 7 to 12% of the cells had a DNA content of between one and two chromosome equivalents, representing a subpopulation that was unable to complete replication at the nonpermissive temperature.

View larger version (27K): [in this window] [in a new window]   FIG. 2.   Effect of temperature on DNA replication in strains MG1655dnaC2 and PC2. The strains were grown exponentially at 30°C in supplemented M9 medium and then shifted to the indicated temperatures. (A) DNA content distributions of strain MG1655dnaC2 120 min after a shift to the indicated temperature. (B) DNA content distributions of strains MG1655dnaC2 and PC2 shifted to 38°C and sampled after 60, 90, and 120 min. (C) DNA content distributions of strains MG1655dnaC2 and PC2 shifted to 40°C and sampled after 60 and 120 min.

Runout kinetics in different growth media. The DNA content of cells grown in rich media (fast-growing cells) is higher than that of cells grown in minimal media (reviewed in reference 5). The effect of medium composition on the length of time required for replication runout and subsequent division was investigated for strain MG1655dnaC2 at 38°C. After 60 min at 38°C in M9 medium containing 0.2% glucose or 0.2% Casamino Acids or in LB containing 0.2% glucose, the DNA content distributions were essentially similar (Fig. 3, column 2). Thus, the relative amounts of cells containing one or two chromosomes were similar at this time point, despite the large differences in the DNA content of the exponentially growing cells (Fig. 3, column 1). After 90 min, the cells grown in LB contained mainly one chromosome equivalent, and little change was observed after further incubation. However, in both M9 media an incubation time of 120 min was required to obtain the maximum number of cells containing one chromosome equivalent. Again, populations in which all cells contained one chromosome after the runout were not observed. Finally, during exponential growth in LB-glucose medium, a small number of cells containing either one or two chromosome equivalents were observed (Fig. 3, column 1, lower panel). These peaks might represent a small subpopulation of nonreplicating or dead cells.

View larger version (23K): [in this window] [in a new window]   FIG. 3.   Effect of different growth media on the kinetics of replication runout and cell division in strain MG1655dnaC2. The strain was grown exponentially at 30°C in M9 medium with glucose (M9glu), M9 with Casamino Acids and glucose (M9cas), or LB with glucose (LBglu) and sampled for flow cytometry. The cultures were then shifted to 38°C and sampled after 60, 90, or 120 min.

Restart of chromosome replication and cell division after return to permissive temperature. After runout at 38°C (synchronization of replication), strain MG1655dnaC2 was returned to 30°C. Replication was reinitiated, and elongation proceeded such that the peak at one chromosome started to gradually move towards the two-chromosome position (Fig. 4A). At 30 to 45 min the peak started to broaden, indicating that a second round of replication had initiated in a significant part of the cell population. Also, a one-chromosome peak could again be distinguished. Since replication lasts more than 40 min and division lasts at least 20 min (reviewed in reference 12), these cells could not have progressed through the entire cell cycle in this short time period, and this peak may instead have represented a subpopulation of cells in which reinitiation of replication had been delayed. This proportion was found to be in the range of 5 to 10% of the cell population. In support of the suggestion that initiation was delayed, the peak disappeared at subsequent time points. At 90 min, a proportion of the cells contained more DNA than an undisturbed exponentially growing culture (Fig. 3, middle row, first panel), suggesting that overreplication occurred to some extent. Furthermore, the general broadening of the DNA content distribution with time suggested that the cell population was returning to the asynchronous replication characteristics of an exponentially growing population. However, the characteristic DNA content distribution of an exponentially growing culture was not fully reestablished, since at late time points in the experiment the culture was already approaching stationary phase, as evidenced by an OD value of near 0.8. In conclusion, the cells initiated replication in synchrony, but after approximately 30 min the synchrony was rapidly lost.

View larger version (14K): [in this window] [in a new window]   FIG. 4.   Reinitiation of chromosome replication in strain MG1655dnaC2 after synchronization at 38°C. The strain was grown exponentially in supplemented M9 medium at 30°C, shifted to 38°C for 120 min, and then shifted back to 30°C. Samples for flow cytometry were taken at the times indicated. DNA content distributions of exponential- and stationary-phase cultures are shown to the left and right, respectively, for comparison. (A) DNA content distributions at different time points after a shift to 30°C at time zero. (B) DNA content distributions at different time points after a shift to 30°C at time zero followed by a shift back to 38°C after 10 min at 30°C.

Restart of replication after a transient shift to permissive temperature. After synchronization at 38°C, the MG1655dnaC2 mutant was shifted to 30°C for a period of 10 min and then returned to 38°C. Initiation of replication occurred, and replication proceeded at the nonpermissive temperature until part of the one-chromosome peak had moved to the two-chromosome position (Fig. 4B). As seen previously in the nontransient shift to the permissive temperature (Fig. 4A) a significant proportion of the peak, in this case 12 to 16% of the population, remained at the one-chromosome position, representing cells that did not initiate replication during the short time period at 30°C. At 90 min, the one-chromosome peak increased significantly in relative size, showing that cell division was occurring. The peak continued to increase in relative size up to the end of the experiment at 120 min. However, the total proportion of the population that went through cell division was small (the proportion is overestimated in the one-chromosome peak, since a single division event gives rise to two one-chromosome cells). Thus, after a transient shift, synchronous initiation of chromosome replication was observed in the majority of the population, but only a small fraction of those cells were able to carry out cell division, and the level of division synchrony was low.

Kinetics of replication initiation after return to permissive temperature. Several attempts to determine the detailed kinetics of replication initiation in the MG1655dnaC2 mutant after synchronization at 38°C were made. Rifampin was added at different time points after the shift to 30°C to monitor the number of initiations over time. However, initiation was not efficiently inhibited by the drug addition, and we were therefore unable to analyze the detailed kinetics of the replication initiation by this approach (see below).

Flow cytometry analysis of a standard experiment with strain PC2. Strain PC2 has been used in numerous cell cycle studies (6, 22, 23, 28, 29, 31). We monitored a synchronization experiment using typical experimental conditions from these investigations. Strain PC2 was grown overnight at 30°C in M9 medium containing 0.1% glucose, 0.2% Casamino Acids, and 20 µg of thymine ml¹. The culture was diluted 1,000-fold into fresh medium and further grown for approximately five mass doublings, resulting in an OD₆₀₀ of about 0.1. The culture was then transferred to 40°C for 60 min. After this runout period, most of the cells contained two chromosomes, whereas only a small part of the population contained one chromosome (Fig. 5). Cells with three or four chromosomes were also present. After return to the permissive temperature (30°C), residual cell division occurred before the 15-min time point, as evidenced by an increase in the relative height of the peak at the one-chromosome position. This corresponds to the division events observed between the 60- and 90-min time points in both Fig. 2B and 3A and shows that the incubation time at 40°C was too short to allow all divisions to occur. After the 15-min time point, the peaks gradually moved to the right with time, such that the leftmost peak was positioned between the one- and two-chromosome positions after 30 min and at the two-chromosome position after 45 min. Thus, the DNA content of the cells increased with time, and replication must therefore have been initiated. Further rounds of initiation also occurred (cf. reference 13), seen as a broadening of the peaks at the 45- and 60-min time points. By 90 min the characteristic DNA distribution of an asynchronous replicating culture had been reestablished. In conclusion, initiation synchrony was observed, but the pattern was complex and the synchrony was not maintained throughout the first cell cycle.

View larger version (9K): [in this window] [in a new window]   FIG. 5.   Standard synchronization experiment with strain PC2. The strain was grown exponentially in supplemented M9 medium at 30°C and shifted to 40°C for 60 min. The culture was then shifted to 30°C, and samples for flow cytometry were taken at the times indicated.

Analysis of transcription and translation requirements for reinitiation of replication in a dnaC2 mutant. Chloramphenicol and rifampin both inhibit initiation of chromosome replication, by inhibiting the peptidyl transferase step in protein synthesis and transcription initiation, respectively. Addition of rifampin (and cephalexin to block cell division during the runout period [see above]) to an exponentially growing culture of MG1655dnaC2 yielded fully replicated chromosomes after the runout period (Fig. 6A). However, after addition of chloramphenicol (and cephalexin), replication did not continue to termination in all cells (Fig. 6A), particularly in those containing more than two chromosomes. In contrast, addition of chloramphenicol to the wild-type MG1655 strain allowed replication to proceed to termination (not shown). This shows that the elongation stage of chromosome replication was more dependent upon ongoing protein synthesis (but not RNA synthesis) in a dnaC2 mutant than in a wild-type strain.

View larger version (25K): [in this window] [in a new window]   FIG. 6.   Analysis of transcriptional and translational requirements for replication reinitiation in strain MG1655dnaC2. (A) First panel, strain MG1655dnaC2 growing exponentially at 30°C in supplemented M9 medium. Second panel, same culture after addition of rifampin and cephalexin followed by incubation at 30°C for 120 min. Third panel, same as second panel except that chloramphenicol was added instead of rifampin. (B) Upper row, strain MG1655dnaC2 was synchronized at 38°C for 120 min, after which rifampin and cephalexin were added. Aliquots were incubated at 38°C for 0, 15, or 30 min before being shifted to 30°C and incubated for a further 120 min before sampling for flow cytometry. Lower row, same as upper row except that chloramphenicol was added instead of rifampin.

Effect of cell concentration on reinitiation of replication in MG1655dnaC2. We shifted cultures of MG1655dnaC2 to the nonpermissive temperature at different ODs. Thus, the cultures contained different cell concentrations at the time of shift to the permissive temperature, and we monitored the abilities of the different populations to reinitiate chromosome replication after the downshift. In the culture with the lowest OD at the time of the shift, most cells were able to undergo initiation shortly after the shift (Fig. 7, upper row). A second round of initiation occurred between the 30- and 45-min time points, noticeable as a considerable broadening of the peaks. In contrast, the first peak remained at the one-chromosome position for a considerably longer time in cultures with higher ODs at the time of downshift, indicating that initiation was delayed (Fig. 7, rows 3 and 4). Also, significantly less peak broadening was observed at the 60-min time point.

View larger version (24K): [in this window] [in a new window]   FIG. 7.   Effect of cell concentration on reinitiation kinetics in strain MG1655dnaC2. The strain was grown exponentially at 30°C in supplemented M9 medium. At different ODs, aliquots were shifted to 38°C and incubated for 120 min. The OD600 values at which the cultures were shifted to 38°C are indicated to the right. Samples for flow cytometry were taken at the indicated time points after the cultures were shifted back to 30°C.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In this report, we show that the phenotypic consequences of mutations in the dnaC gene are affected by several parameters, including the choice of permissive and nonpermissive temperatures, the length of the time of incubation at the nonpermissive temperature, the growth medium, the type of temperature shift used for reinitiation of replication (transient or nontransient), the genetic background of the host cell, and the cell concentration. Synchronous initiation of replication could be obtained in the mutants, but the synchrony was transient and was lost before the first cell divisions occurred. Reinitiation of replication after a shift from the nonpermissive to the permissive temperature occurred in the absence of both transcription and translation, whereas the elongation stage of replication was dependent upon ongoing protein synthesis.

Our recommendations for the use of dnaC mutants in cell cycle studies are summarized in Table 2, and we will here briefly comment upon the different parameters that affected the results.

In cell cycle studies, small temperature shifts are usually desirable to minimize induction of the heat shock response, which has multiple physiological effects (reviewed in reference 10) and may complicate the interpretation of the experiments. We found that a temperature shift of 6°C was sufficient to fully inhibit replication initiation in both the dnaC2 and dnaC28 mutants, presumably to obtain full heat inactivation of the mutant DnaC protein molecules. Also, at the permissive temperature the strains should preferably show near-wild-type replication characteristics, and this required a 2°C lower incubation temperature for the MG1655dnaC28 mutant than for MG1655dnaC2. Finally, we wish to point out that even though replication runout did occur at 42°C for both dnaC alleles used, reinitiation of replication was found to be inefficient (data not shown; similar effects were noted in reference 18), perhaps due to a too-extensive denaturation of the thermosensitive DnaC protein at 42°C.

The length of the incubation period at the nonpermissive temperature affected the proportion of the population that contained a single chromosome after the runout. Although full runout was obtained even after only 60 min at the nonpermissive temperature, this resulted in cell division still taking place when the cultures were shifted to the permissive temperature. In previous studies, little increase in cell concentration was detected after similar shifts (13, 32). The difference may be explained by the higher resolution obtained by flow cytometry as compared to the OD and Coulter Counter measurements performed in those studies. The residual division activity complicates analyses of e.g., cell cycle-specific gene expression, since initiation-specific events will be scored against a background of division-related activities. Furthermore, it was interesting that under no circumstances did all cells contain a single chromosome after the runout period. The fraction of the population that contained two chromosomes after extended incubation was always similar, and we believe that this represented a subpopulation of cells which, at the time of the shift to the nonpermissive temperature, were in a particular cell cycle stage that did not support division, e.g., due to lack of a cell cycle cue. This subpopulation has previously also been observed in the intR1 system developed in our laboratory (3) (see below), and the proportion was similar to that observed in the dnaC mutants in this report.

Cells grown in rich medium required a shorter time at the nonpermissive temperature to yield the maximum number of cells containing a single chromosome, despite their considerably higher DNA content in the exponential phase. This may be explained by the shorter generation time of fast-growing cells: the time required to go through the necessary successive divisions to reduce the number of chromosomes in the cells was shorter than that for slow-growing cells, even when the total number of required division events was higher.

A nontransient shift to the permissive temperature resulted in additional rounds of replication being initiated about 30 min after the first one, thereby ending the cell cycle synchrony in the population. Secondary initiations at about this time after the first initiation have also been noted by others (8, 13). The secondary rounds of replication initiations did not occur if a transient shift was used instead. However, in this case only part of the population was able to go through the cell cycle far enough to carry out division, and, importantly, the level of division synchrony in the population was low. Furthermore, Helmstetter and Krajewski found that late cell cycle events, in their case replication termination, did not occur in synchrony (13), although transient shifts were not used in that study.

At high cell concentrations, reinitiation of replication was delayed. As cultures approach stationary phase, major changes in gene expression, physiology, and replication occur in the cells (1, 16) as a preparation for nutrient starvation and a reduced growth rate, and it is possible that the reinitiation efficiency was reduced through such mechanisms. Alternatively, the reinitiation may have been affected through so-called quorum sensing (reviewed in references 9 and 24), through which cells are able to sense and respond to high concentrations of neighboring cells. We are currently investigating these possibilities.

The genetic background of the host cell also affected the results. The strains in which the dnaC alleles were originally isolated were subjected to chemical mutagenesis (7, 26). It is thus possible that other, unrelated, mutations may have contributed to the differences observed when these strains were used.

In conclusion, dnaC mutants perform best in analyses of cell cycle events occurring around initiation of replication and up to perhaps 30 min of the replication elongation stage, whereas the synchrony is not as tight around replication termination and even less so at division. The characteristics of the dnaC mutants may be compared to those of the so-called intR1 system developed in our laboratory for cell cycle analyses, in which replication synchrony also can be obtained (3). The intR1 system displays a tighter synchrony in a transient shift, and the synchrony is better maintained throughout the first cell cycle, including division, although only part of the cell population divides in the intR1 case also. In addition, a considerably smaller temperature shift (2 to 3°C) is enough to inhibit or reinitiate replication in the intR1 system.

The fact that neither transcription nor translation was required for reinitiation of replication in the dnaC mutants indicates that de novo synthesis of DnaC was not required for initiation. This is in accordance with earlier studies (8, 11, 13, 18) and suggests that the cells had become fully competent for replication initiation at the nonpermissive temperature. This would be expected if all initiation steps up to the DnaC-mediated loading of the DnaB helicase (reviewed in reference 17) could occur at the nonpermissive temperature, including possible transcriptional activation of oriC (reviewed in reference 19), which would otherwise have been inhibited by the rifampin addition. Thus, if all events prior to the action of DnaC had taken place, initiation should be limited only by the ability of DnaC to renature after the downshift. Furthermore, it has been noticed previously that the total amount of DNA synthesized is less when rifampin is added to a culture before return to the permissive temperature than when chloramphenicol is added (11, 18). Our observation that fewer initiations occurred after rifampin addition than after chloramphenicol addition (Fig. 6) is in accordance with these observations and indicates that the transcription requirement has to do with the initiation stage.

The elongation stage of replication was dependent upon ongoing protein synthesis in the mutants at the permissive temperature. This indicates that the DnaC protein was required throughout replication and that it was being turned over. However, elongation proceeded to completion when the cultures were first shifted to the nonpermissive temperature (in the absence of chloramphenicol), indicating that the mutants were affected primarily in initiation. It may be that DnaC molecules in an already active replisome established at the permissive temperature are more heat resistant than renatured DnaC molecules that are assembled into newly forming replisomes during a transient shift or that continued synthesis of new DnaC molecules at the nonpermissive temperature could compensate for the ongoing heat inactivation. The precise requirement for the DnaC protein in elongation needs to be further studied.

Finally, when rifampin or chloramphenicol was added and transient downshifts were performed (Fig. 6), cells containing three chromosomes were detected after the runout. This shows that when two origins were present in the cells, often only one of these initiated during the downshift. This may complicate cell cycle analyses, particularly in nontransient shifts in which the second origin is initiated at a later time point, when enough active DnaC protein again has become available. Still, in the exponential cultures growing at the permissive temperature, the majority of the initiations were coordinated (Fig. 1), showing that any adverse effects on coordination were largely overridden by the normal coordination control system.