Independent Control of Replication Initiation of the Two Vibrio cholerae Chromosomes by DnaA and RctB

Abstract

Although the two Vibrio cholerae chromosomes initiate replication in a coordinated fashion, we show here that each chromosome appears to have a specific replication initiator. DnaA overproduction promoted overinitiation of chromosome I and not chromosome II. In contrast, overproduction of RctB, a protein that binds to the origin of replication of chromosome II, promoted overinitiation of chromosome II and not chromosome I.

The genome of Vibrio cholerae, the curved gram-negative bacterium that causes the severe diarrheal disease cholera, is composed of two circular chromosomes. The V. cholerae genome is unequally divided between the two chromosomes. Most of the genes essential for V. cholerae growth are found on the 2.96-Mb chromosome I (chrI); however, the presence of essential genes on the 1.07-Mb chromosome II (chrII) indicates that this replicon is not a dispensable megaplasmid but a bona fide chromosome (9). The genomes of the many other species that are included in the family Vibrionaceae are also thought to be divided between two chromosomes (19, 25, 26). Furthermore, a diverse set of bacteria outside of the Vibrionaceae have divided genomes that consist of more than one chromosome (1, 4, 13, 17, 21). Currently there is very little knowledge of the mechanisms controlling chromosome replication in bacteria with multipartite genomes, as nearly all studies of bacterial chromosome replication to date have used organisms with single chromosomes.

We previously defined and characterized the origins of replication of the two V. cholerae chromosomes, oriCI_Vc and oriCII_Vc (6). oriCI_Vc is similar in sequence to the origin of replication of the Escherichia coli chromosome, oriC. oriCI_Vc, like oriC, includes five binding sites for DnaA, a protein that promotes unwinding of oriC to initiate replication (18), as well as a binding site for integration host factor (IHF), a protein that bends DNA and is thought to stimulate the unwinding activity of DnaA (23), and several sites for methylation by DNA adenine methyltransferase (Dam), an enzyme that is involved in the timing of replication in E. coli (16). oriCII_Vc has very little sequence similarity to oriCI_Vc but contains a single binding site for DnaA, a binding site for IHF, and several sites for methylation by Dam (6). Both dam and dnaA proved to be required for replication of oriCI_Vc- and oriCII_Vc-based minichromosomes, suggesting that at least two common factors contribute to the regulation of initiation of chrI and chrII replication.

oriCII_Vc-based replication in E. coli requires two V. cholerae genes that flank the origin of chrII. One of these genes, rctA, codes for an untranslated RNA whose function in oriCII_Vc-based replication is unknown (6). The other gene, rctB, codes for a multifunctional protein, RctB, that is conserved in many diverse vibrio species. RctB appears to function as a chrII-specific initiator protein as it is required for oriCII_Vc-based and not oriCI_Vc-based replication and binds to several sites in oriCII_Vc but not to oriCI_Vc (6). Also, it was recently reported that the copy number of an oriCII_Vc-based minichromosome in E. coli correlates with RctB levels (20). RctB also functions as a trancriptional repressor, inhibiting its own transcription as well as that of rctA (3, 20). Together these observations suggest that rctB autoregulation determines cellular RctB levels and thereby controls chromosome II replication.

The requirement of rctA and rctB for oriCII_Vc- and not oriCI_Vc-based replication indicates that at least partially distinct control mechanisms govern initiation of replication of the two chromosomes. However, when we investigated the kinetics of V. cholerae chromosome replication using Meselson-Stahl density shift and flow cytometry experiments, we found that both chromosomes initiate replication with a constant inter-replication time and within a narrow time interval of each cell cycle (5). Since DnaA levels and/or activity play a critical role governing replication in E. coli (8), we speculated that this well-conserved initiator protein that is required for both oriCI_Vc- and oriCII_Vc-based replication may coordinate the replication of the two chromosomes.

Here we analyzed the effect of overproduction of DnaA and RctB on initiation of chrI and chrII replication using three techniques: fluorescence microscopy, quantitative PCR (Q-PCR), and flow cytometry. The conclusions from each of these approaches were concordant. DnaA overproduction promoted overinitiation of chrI and not chrII replication; whereas RctB overproduction promoted overinitiation of chrII and not chrI replication. Thus, DnaA and RctB appear to independently control initiation of chrI and chrII replication, respectively.

DnaA overproduction promotes chrI overinitiation.

Two derivatives of El Tor V. cholerae strain N16961, MForiI (formerly referred to as MFfc7) and MForiII (formerly referred to as MFfc11) (7), enabled us to use fluorescence microscopy to investigate the effect of elevated cellular DnaA or RctB concentrations on initiation of chrI and chrII replication. MForiI and MForiII contain arrays of tet operator sequences inserted near oriCI_Vc and oriCII_Vc, respectively. The origin regions in these strains can be visualized as fluorescent foci after expression of TetR-YFP, the Tet repressor protein fused to the yellow fluorescent protein (YFP). Enumeration of the fluorescent foci in MForiI provides a way to count the number of oriCI_Vc regions; similarly, enumeration of the fluorescent foci in MForiII provides a way to count the number of oriCII_Vc regions. When MForiI was grown in M9 minimal medium supplemented with 0.2% glucose to an optical density at 600 nm (OD₆₀₀) of ∼0.5, more than 95% of the cells contained one or two fluorescent foci (Fig. 1A), suggesting that the great majority of cells grown under these conditions had not initiated replication or had initiated one round of replication. This was also the case for MForiII; like MForiI, more than 95% of MForiII cells contained one or two fluorescent foci (Fig. 1D). There appear to be more MForiI than MForiII cells with two foci (Fig. 1A versus D). This may reflect the fact that duplicated oriCI_Vc foci initiate segregation earlier and thus are detectable as discrete foci, prior to duplicated oriCII_Vc (7).

FIG. 1.

Distribution of the number of origins of the two V. cholerae chromosomes, established in four or more independent experiments. Only cells with one or more foci were analyzed. The number of cells used to determine each distribution is shown in the upper right in each panel. Fluorescence microscopy was carried out as described previously (7). Expression of TetR-YFP was induced with 0.04% arabinose, and expression of RctB and DnaA was induced with 100 μM IPTG.

Overproduction of DnaA had distinct effects on initiation of replication of chrI and chrII. When DnaA was overproduced from the isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible promoter in pGZ119 (14), the number of cells with multiple oriCI_Vc foci significantly increased (compare Fig. 1C and A). The percentage of MForiI cells with more than two oriCI_Vc foci increased from 4.3% to 28% following overexpression of DnaA. Furthermore, there was a statistically significant difference (P < 0.01) in the distribution of the numbers of oriCI_Vc foci in MForiI in the absence versus the presence of DnaA overproduction by the Student's t test. In contrast, the number of oriCII_Vc foci in MForiII was not significantly altered by overproduction of DnaA; most cells still had one or two oriCII_Vc foci, and there was no difference between the distributions of the number of oriCII_Vc foci in the absence versus the presence of DnaA overexpression (Fig. 1D versus F).

We performed Q-PCR analyses as another way to assay the influence of DnaA on initiation of chrI and chrII replication. In the Q-PCR experiments, cells were grown to an OD₆₀₀ of ∼0.5 and then chromosomal DNA was extracted and used as a template for Q-PCRs. In these reactions, we measured the ratio of oriCI_Vc to terI_Vc (terminus of chrI) DNA and the ratio of oriCII_Vc to terII_Vc (terminus of chrII) DNA. In N16961 harboring the empty expression vector pGZ119, these ratios were 1.86 and 1.59, respectively (Table 1). Following overexpression of dnaA for approximately 3 h (at which time DnaA levels were ∼3× greater than normal), the oriCI_Vc/terI_Vc ratio increased to 3.17, a statistically significant change (P < 0.01) from cells grown without excess DnaA. In contrast, dnaA overexpression did not increase the oriCII_Vc/terII_Vc ratio (Table 1). This level of DnaA overproduction did not appreciably alter cell growth. In time course experiments, following induction of DnaA, the oriCI_Vc/terI_Vc ratio increased for ∼30 min and then reached a plateau (Fig. 2A), suggesting that there are limits to the extent to which DnaA can promote initiation of chrI replication. Neither the oriCII_Vc/terI_Vc nor terII_Vc/terI_Vc ratio changed after DnaA induction (Fig. 2A). Overall, like the fluorescence microscopy analyses, these quantitative PCR experiments indicate that overproduction of DnaA promotes initiation of chrI and not chrII replication.

TABLE 1.

Influence of overproduction of RctB or DnaA on the oriCI_Vc/terI_Vc and oriCII_Vc/terII_Vc DNA ratios^a

Expression vector	Ratiob
	oriCIVc/terIVc	oriCIIVc/terIIVc
Vector alone	1.86 ± 0.2	1.59 ± 0.5
With RctB	1.75 ± 0.5	5.33 ± 2*
With DnaA	3.17 ± 0.5*	0.93 ± 0.4

^aFor the Q-PCR assays, 1 ng of template DNA was used for PCR amplification. The primers used were 5′-GTGCAGTGGCTTAGCAATGA-3′ and 5′-ATTGCCAACGGTGATAAAGC-3′ for oriCI, 5′-CATCAGCGGGCTTTGTTATT-3′ and 5′-GCCAGCTATGCTCGTAAAGG-3′ for oriCII, 5′-AGCCGAAACCTTCCAACTTT-3′ and 5′-CATAGGCGGTTGCTTGGTAT-3′ for terI, and 5′-CAAGGCGAAGAAATTCAAGC-3′ and 5′-TAAGGCTGAAACGGGAGAAA-3′ for terII.

^bAll amplicons are 120 nucleotides long; standard curves were created with DNA extracted from overnight cultures. The mean ± standard deviation was calculated from three or more independent experiments. An asterisk indicates statistical analysis with Student's t test revealed that the result is significantly different from the others in the same column.

FIG. 2.

Overproduction of DnaA stimulates chromosome I replication. IPTG (100 μM) was used to induce DnaA synthesis from pMR10 (12) in Bah-2 (22) grown at 37°C in minimal medium containing glycerol. In panel A, samples were taken every 5 min for Q-PCR with the following primers: 5′-CGCCAACCGAGTTTGGATTC-3′ and 5′-GAAAAAGCGCGTGAGCTTGG-3′ for oriCI_Vc, 5′-CTGAGGCGGATTTGGCACTC-3′ and 5′-GCTTGCGCCGCTTTTAACTG-3′ for terI_Vc, 5′-GCTCCACCTTCGGTGTTTCG-3′ and 5′-TGGTTTCGTGTGGCAGCAAT-3′ for oriCII_Vc, and 5′-TATCCGCACAGCCTCAGCAA-3′ and 5′-CACGCAAACAGACCGACACC-3′ for terII_Vc. Values were normalized to DNA extracted from a sample incubated with rifampin. Triangles, oriCI_Vc relative to terI_Vc; squares, oriCII_Vc relative to terI_Vc; circles, terII_Vc relative to terI_Vc. In panel B, IPTG was added at an OD₄₅₀ of 0.15, and at the indicated times samples were taken for a 4-h incubation with rifampin and cephalexin. Cells were fixed in ethanol and stained with mithramycin and ethidium bromide for flow cytometry as described previously (15).

We also measured the DNA content of rifampin- and cephalexin-treated cells by flow cytometry to assess the effect of additional DnaA protein on V. cholerae chromosome replication (Fig. 2B). In such cells, the number of fully replicated chromosomes represents the number of origins at the time of drug addition (24). Cells were grown in a minimal medium containing glycerol as the sole carbon source. Under these growth conditions, we previously found that cells contain either one copy of each chromosome (peak I, 4.1 Mb, Fig. 2B, 0 min) or two copies of each chromosome (peak II, 8.3 Mb, Fig. 2B, 0 min) (5). DnaA overproduction rapidly led to a decrease in cells containing one copy of each chromosome and an increase in cells containing more DNA (Fig. 2B, 10 min and 20 min). The distinction between peak I and peak II became less clear following DnaA overproduction. By 10 min after induction of DnaA expression, cells containing more than 2 genome equivalents of DNA were observed; there was an even larger fraction of cells with more than 2 genome equivalents of DNA by 20 min after induction of DnaA (Fig. 2B). Since no increase in the ratio of oriCII_Vc to terI_Vc was observed by Q-PCR in cells grown under identical conditions (Fig. 2A), the increase in cellular DNA content detected by flow cytometry resulted only from replication of chromosome I. No change in peak I or peak II was observed in control experiments, where cells containing an empty expression vector were grown under identical conditions (data not shown). In aggregate, these observations suggest that DnaA levels (and/or activity) restrict initiation of chrI replication; in contrast, it appears that DnaA levels are not limiting for initiation of chrII replication.

RctB overproduction promotes chrII overinitiation.

RctB overproduction had distinct effects on initiation of chrI and chrII replication. Overproduction of RctB by approximately twofold did not lead to a significant change in the distribution of the numbers of oriCI_Vc foci in MForiI (Fig. 1A versus B) or alter cell growth. However, RctB overproduction increased the number of MForiII cells with more than two oriCII_Vc foci from 2% to 26.2%. Furthermore, the distributions of the numbers of oriCII_Vc foci in MForiII in the absence versus the presence of rctB overexpression were statistically significant (P < 0.01) (Fig. 1D versus E). Q-PCR assays also revealed that overproduction of RctB promoted chrII and not chrI replication. RctB overexpression did not lead to an increase in the oriCI_Vc/terI_Vc ratio, but the oriCII_Vc/terII_Vc ratio increased more than threefold when RctB was overproduced (Table 1).

Flow cytometry analyses were also consistent with RctB promoting initiation from oriCII_Vc and not from oriCI_Vc. Following rctB overexpression, shoulders were observed on both peak I and peak II, indicating an increase in DNA content in cells containing both one and two copies of each chromosome (Fig. 3). The increase in the amount of cellular DNA content after RctB overexpression was relatively modest; given the Q-PCR findings presented above, these small increases in DNA content likely result from replication (i.e., more copies) of the 1.07-MB chrII and not from replication of the 2.96-Mb chrI. This was confirmed by determining the oriCI_Vc/terI_Vc and oriCII_Vc/terI_Vc ratios by Southern blot hybridization (not shown). These observations, coupled with the observation that RctB binds to oriCII_Vc (6), provide strong evidence that RctB functions as an initiator of chrII replication and that cellular levels (or activity) of RctB regulate initiation of chrII replication.

FIG. 3.

Overproduction of RctB stimulates chromosome II replication. Bah-2 was grown as indicated in the legend to Fig. 2, except that RctB synthesis from pKGX was induced with 0.2% l-arabinose. At the indicated times, samples were incubated with rifampin and cephalexin prior to flow cytometric analysis as described in the legend to Fig. 2. Arrows indicate the additional DNA synthesized as a result of RctB overproduction.

Our observations suggest that initiation of replication of the two V. cholerae chromosomes is governed by two distinct initiator proteins. DnaA appears to be the initiator of chrI replication, as we found that increasing the concentration of this protein promotes chrI replication. Since overexpression of DnaA did not promote chrII replication and since overproduction of RctB alone promoted chrII replication, it seems unlikely that DnaA functions as the primary initiator for chrII replication. Instead RctB appears to be the initiator of chrII replication, as increasing the concentration of this oriCII_Vc-binding protein promoted chrII replication. The presence of two distinct initiators for chrI and chrII may minimize competition between the two chromosomes for DnaA, which is required for both oriCI_Vc-based replication and oriCII_Vc-based replication (6).

If DnaA and RctB are the initiators of chrI and chrII replication, then cellular factors that determine the levels and/or activity of these proteins indirectly control initiation of chrI and chrII, respectively. In E. coli, tight regulation of chromosome replication is controlled through the level or activity of the DnaA protein. The newly replicated and hemimethylated oriC is bound (sequestered) by the SeqA protein, thereby preventing formation of the origin/DnaA initiation complex for about one-third of a generation. During sequestration, the concentration of free DnaA protein is reduced due to its titration by datA and due to sequestration of the dnaA promoter, preventing de novo DnaA synthesis. Furthermore, the overall activity of DnaA is reduced by the regulatory inactivation of DnaA) (RIDA) system, composed of DnaN and HdA, which stimulates the ATPase activity of DnaA to reduce the amount of active ATP-bound DnaA (for review, see references 2 and 10). Similar processes are likely to govern DnaA levels and/or activity in V. cholerae. Indeed, chrI appears to have a datA locus (S. Duigou, unpublished data), and seqA and dam, genes required for oriC sequestration, are present in V. cholerae. It is not known if V. cholerae has a RIDA system, although there is no apparent Hda orthologue encoded in the V. cholerae genome.

Little is known about the cellular factors that control RctB levels or activity. However, we recently found that RctB acts as an autorepressor (3). Interestingly, DnaA and Dam, two host factors that are required for replication of both chromosomes, modulate RctB autorepression (6). DnaA may also promote RctB's activity as an initiator of replication at oriCII_Vc. DnaA is known to promote replication of certain plasmids that encode their own initiators. For example, in plasmid RK2 replication, DnaA is thought to play an accessory, though essential, role to TrfA in promoting strand unwinding of the origin (11).

Our previous work revealed that chrI and chrII initiate replication synchronously, within a narrow time interval in the cell cycle. How can we reconcile synchronous initiation with the observations presented here suggesting that distinct initiator proteins control replication of the two chromosomes? Since RctB levels and not DnaA levels promote chrII initiation, it seems unlikely that DnaA directly mediates synchronous initiation of the two chromosomes. However, our experiments only explored the influence of DnaA overexpression on initiation of chrII replication. oriCII_Vc only contains a single DnaA box, and it remains possible that, regardless of the concentration of RctB, a minimum concentration of DnaA is required for initiation of chrII replication. Alternatively, Dam, a protein that is known to regulate initiation in E. coli and that is required for replication of both V. cholerae chromosomes, may be an additional replication factor that mediates synchronous initiation of the two chromosomes. Finally, additional V. cholerae factors may coordinate the activities of RctB and DnaA.