DNA methylation affects the cell cycle transcription of the CtrA global regulator in Caulobacter
2002; Springer Nature; Volume: 21; Issue: 18 Linguagem: Inglês
10.1093/emboj/cdf490
ISSN1460-2075
Autores Tópico(s)Protist diversity and phylogeny
ResumoArticle16 September 2002free access DNA methylation affects the cell cycle transcription of the CtrA global regulator in Caulobacter Ann Reisenauer Corresponding Author Ann Reisenauer Developmental Biology, Stanford University, Stanford, CA, 94305-5329 USA Search for more papers by this author Lucy Shapiro Lucy Shapiro Developmental Biology, Stanford University, Stanford, CA, 94305-5329 USA Search for more papers by this author Ann Reisenauer Corresponding Author Ann Reisenauer Developmental Biology, Stanford University, Stanford, CA, 94305-5329 USA Search for more papers by this author Lucy Shapiro Lucy Shapiro Developmental Biology, Stanford University, Stanford, CA, 94305-5329 USA Search for more papers by this author Author Information Ann Reisenauer 1 and Lucy Shapiro1 1Developmental Biology, Stanford University, Stanford, CA, 94305-5329 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (2002)21:4969-4977https://doi.org/10.1093/emboj/cdf490 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The Caulobacter chromosome changes progressively from the fully methylated to the hemimethylated state during DNA replication. These changes in DNA methylation could signal differential binding of regulatory proteins to activate or repress transcription. The gene encoding CtrA, a key cell cycle regulatory protein, is transcribed from two promoters. The P1 promoter fires early in S phase and contains a GAnTC sequence that is recognized by the CcrM DNA methyltransferase. Using analysis of CcrM mutant strains, transcriptional reporters integrated at different sites on the chromosome, and a ctrA P1 mutant, we demonstrate that transcription of the P1 promoter is repressed by DNA methylation. Moreover moving the native ctrA gene to a position near the chromosomal terminus, which delays the conversion of the ctrA promoter from the fully to the hemimethylated state until late in the cell cycle, inhibited ctrA P1 transcription, and altered the time of accumulation of the CtrA protein and the size distribution of swarmer cells. Together, these results show that CcrM-catalyzed methylation adds another layer of control to the regulation of ctrA expression. Introduction The interaction of regulatory proteins and methylated DNA is important to cell physiology in both eukaryotes and prokaryotes. In eukaryotes, a major consequence of chromosome methylation is transcriptional silencing (Bird and Wolffe, 1999). In prokaryotes, DNA methyltransferases (MTases) are best known for their role in restriction–modification systems (Bickle and Kruger, 1993). However, these enzymes also have regulatory roles in the bacterial cell. Two examples of regulatory MTases are the Escherichia coli Dam and the Caulobacter crescentus CcrM proteins. Neither Dam nor CcrM have known cognate restriction enzymes, but rather these proteins act to coordinate cell cycle events. Dam methylation governs several cellular functions, including the initiation of DNA replication (Barras and Marinus, 1989; Boye and Lobner-Olesen, 1990) and the transcription of certain genes, such as the pap pili operon in uropathogenic E.coli (Nou et al., 1993; Braaten et al., 1994) and plasmid-encoded fimbriae (Pef) in Salmonella typhimurium (Nicholson and Low, 2000). In addition, Dam methylation regulates Tn10 transposition by altering the activity of the transposase promoter (Roberts et al., 1985). Dam is also required for virulence in S.typhimurium, where it either directly or indirectly controls the expression of a number of genes that are induced during infection (Garcia-Del Portillo et al., 1999; Heithoff et al., 1999). The CcrM MTase, which methylates the adenine in GAnTC target sequences, is widespread among α-proteobacteria and has been shown to be essential for viability in C.crescentus, Sinorhizobium meliloti, Brucella abortus and Agrobacterium tumefaciens (Stephens et al., 1996; Wright et al., 1997; Robertson et al., 2000; Kahng and Shapiro, 2001). CcrM activity is cell cycle regulated in both C.crescentus and A.tumefaciens (Stephens et al., 1996; Kahng and Shapiro, 2001). In Caulobacter, this enzyme is present and is active only at the end of S phase when it brings the newly replicated DNA from the hemimethylated to the fully methylated state (Stephens et al., 1996). CcrM is restricted to this period of the cell cycle by three regulatory mechanisms: activation of ccrM transcription by the CtrA response regulator (Quon et al., 1996; Reisenauer et al., 1999), inhibition of ccrM transcription by methylation of the GAnTC sites immediately downstream of the transcription start site (Stephens et al., 1995), and rapid proteolysis of the CcrM protein (Wright et al., 1996). In mutants that express CcrM throughout the cell cycle, the control of DNA replication initiation is relaxed and the cells have abnormal morphology (Zweiger et al., 1994), suggesting that differential CcrM methylation helps to regulate these processes. Although CcrM is required for viability, the essential functions of this MTase are unknown. In Caulobacter, cell differentiation is coordinated with progression through the cell cycle (see Figure 3B). The motile swarmer cell present in G1 phase ejects its flagella and differentiates into a non-motile stalked cell. During the swarmer to stalked cell (G1–S) transition, chromosome replication is initiated on a fully methylated chromosome. As the stalked cell progresses through S phase, a new flagellum is assembled at the pole opposite the stalk. Consequently, two distinct cell types are produced at each cell division, a replication-repressed swarmer cell and a stalked cell, which immediately begins another round of DNA synthesis (Hung et al., 2000). The chromosome is fully methylated at the start of replication, but progressively becomes hemimethylated as replication proceeds bidirectionally from the origin to the terminus (Dingwall and Shapiro, 1989). Re-methylation of the newly synthesized DNA is restricted to the end of S phase when the CcrM MTase is synthesized (Stephens et al., 1996; Marczynski, 1999). This successive change in the methylation state of the chromosome during S phase reflects the progression of DNA replication. Figure 1.Activity of the ctrA P1–lacZ and P1UM–lacZ transcriptional reporters integrated at different sites on the chromosome. (A) Diagram of the C.crescentus chromosome showing the locations of the origin of replication (Cori), the terminus region (Ter), the ctrA gene and the hrcA (site 1), recA (site 2), and trpE (site 3) integration sites. (B) Schematic of the methylation state of GAnTC motifs at the three integration sites during the cell cycle. All sites are fully methylated (FM) in the swarmer cell. After the initiation of DNA replication, the time when each GAnTC site becomes hemimethylated (HM) depends on its distance from Cori. Near the end of S phase, CcrM (shown as a gray bar) methylates the newly synthesized DNA strand, restoring the chromosome to the fully methylated state. The Caulobacter cell cycle is shown schematically. The θ and ring structures inside the cells represent replicating and non-replicating DNA, respectively. (C) Activity of the wild-type ctrA P1 (P1–lacZ) and unmethylated P1 (P1UM–lacZ) transcription probes integrated into sites 1, 2 and 3. Promoter activity is the mean ± SD of three experiments. Download figure Download PowerPoint There are nearly 4500 GAnTC sites in the Caulobacter genome, whereas ∼12 000 sites are expected statistically (Nierman et al., 2001). In addition, 22% of these sites are found in the 10% of the genome located between open reading frames. The concentration of the limited number of GAnTC sites in intergenic DNA suggests that changes in the methylation state of these sites may alter the interactions of regulatory proteins with their target DNA. To explore the possibility that DNA methylation plays a role in controlling transcription in Caulobacter, we examined temporally regulated genes that have GAnTC sites in their promoter regions. These genes include ctrA, encoding a global transcriptional regulator (Quon et al., 1996), ftsZ, encoding a tubulin-like protein required for cell division (Quardokus et al., 1996), and groESL, encoding a molecular chaperone (Avedissian and Gomes, 1996). Of these candidate genes, the transcription of ctrA and ftsZ changed in response to changes in the methylation state of the chromosome. The CtrA response regulator directly controls the transcription of at least 55 operons (Laub et al., 2002), including those required for DNA methylation (ccrM), cell division (ftsZ), and flagella and pili biogenesis (Quon et al., 1996; Kelly et al., 1998; Laub et al., 2000; Skerker and Shapiro, 2000). CtrA also prevents the initiation of DNA replication in swarmer cells by binding to the Caulobacter origin of replication (Cori; Quon et al., 1998). CtrA activity during the cell cycle is highly regulated. The transcription of ctrA is controlled by feedback regulation (Figure 1A). At the beginning of S phase, ctrA is transcribed from the ctrA P1 promoter, which contains a GAnTC site near the −35 region. As CtrA protein accumulates during S phase, it activates transcription from the ctrA P2 promoter and represses the P1 promoter (Domian et al., 1999). The activity of this global transcriptional regulator in turn is governed by temporally controlled phosphorylation and targeted proteolysis (Domian et al., 1997). Here we show that the methylation state of the P1 promoter adds another layer of control to the regulation of CtrA expression. Figure 2.Feedback control of ctrA transcription. (A) Diagram of the ctrA promoter region. The P1 and P2 transcription start sites are indicated by bent arrows, GAnTC sites are marked by asterisks, and CtrA binding sites are shown as gray boxes. As CtrA protein (gray oval) accumulates during S phase, it activates transcription from P2 and inhibits P1 transcription. (B) Nucleotide sequence of the ctrA promoter. The P1 and P2 transcription start sites (bent black arrows) and the P2FM alternate start site described in this study (bent gray arrow) are marked. CcrM methylation sites are underlined and the C(–25)T mutation in P1 is indicated. CtrA recognition motifs and the translation start site are in bold. Download figure Download PowerPoint Results ctrA promoter activity is altered in CcrM mutant strains The gene encoding the CtrA response regulator is transcribed from two promoters, P1 and P2, which are expressed at different times during the Caulobacter cell cycle (Domian et al., 1999). CtrA binds to consensus motifs in each promoter (Figure 1A, light gray boxes), repressing the transcription of the early P1 promoter and activating P2 transcription. In addition, there are two GAnTC sites (shown by the asterisks in Figure 1A) that are found at −29 relative to the P1 promoter and at +16 relative to the P2 promoter. Their location in the ctrA promoter suggests that changes in the methylation of these sites could influence ctrA transcription. To test the effect of maintaining the chromosome in the fully methylated state throughout the cell cycle on ctrA transcription, we expressed ccrM constitutively using strain LS1. In the LS1 strain, two copies of ccrM are present on the chromosome: one is expressed from its native promoter, and the other is expressed from a constitutive Plac promoter. As a result, ccrM is transcribed continually and chromosomal DNA is maintained in the fully methylated state throughout the cell cycle (Zweiger et al., 1994). RNase protection assays showed that the ctrA P1 transcript was significantly reduced but the P2 transcript was unaffected under these conditions (Figure 2A). In this experiment, RNA isolated from the wild-type and LS1 strains was probed with 32P-labeled antisense RNA probes for ctrA and rrnA. The 16S ribosomal RNA (rrnA) probe is an internal control used to normalize for differences in the amount of RNA applied to each lane. Figure 3.ctrA transcript levels in CcrM mutant strains. (A) Representative phosphorimage of ctrA and rrnA transcripts assayed by RNase protection in wild-type (wt), LS1 and LS2144 cultures grown in PYE + 0.1% xylose (X) or PYE + 0.1% glucose for 3 h (G). The excess, undigested ctrA probe and transcripts from the P1 and P2 promoters are marked. The 16S ribosomal RNA (rrnA) was probed as an internal control. 32P-labeled ssDNA markers are shown on the left. (B) Quantitation of ctrA P1 and P2 transcripts in LS2144 cultures grown in PYE + xylose (X) or glucose (G) using a PhosphorImager. Values were normalized using the rrnA internal control and expressed relative to the PYEX value. Data are the mean ± SD of three experiments. (C) Southern blots showing the methylation state of the dnaA and fliG chromosomal loci in wild-type cultures (wt), and in LS2144 cultures grown in PYE + xylose (X) or glucose (G). FM and HM mark fully methylated and hemimethylated DNA, respectively. Download figure Download PowerPoint To analyze ctrA P1 and P2 promoter activity when CcrM is depleted, we used strain LS2144 in which the chromosomal ccrM locus is inactivated and ccrM under the control of the conditional xylX promoter is present on a low copy number plasmid (Stephens et al., 1996). Transcription of PxylX is induced by xylose (Meisenzahl et al., 1997). When LS2144 cultures are shifted from growth in peptone-yeast extract (PYE) + 0.1% xylose (PYEX) to PYE + 0.1% glucose (PYEG), cell viability drops after 4 h and cell growth ceases after 6–8 h (Stephens et al., 1996). In addition, CcrM protein levels fall. To confirm that growth of this strain in PYEG reduces CcrM activity and methylation of the chromosome, we examined the methylation state of two sites on the chromosome, the dnaA and fliG loci, using an overlapping restriction site assay (Campbell and Kleckner, 1990; Zweiger and Shapiro, 1994). As shown in Figure 2C, hemimethylated DNA (HM) at these sites increased 4- to 6-fold when cultures were shifted to PYEG for 3 h, demonstrating that methylation of the chromosome was impaired when CcrM was depleted. We used RNase protection assays to compare ctrA P1 and P2 mRNA levels in the wild-type strain and strain LS2144 with xylose-dependent expression of CcrM. As shown in Figure 2A, transcription from the ctrA P1 promoter increased when CcrM was depleted (LS2144 cultures grown in PYEG). The ratio of P1 to total (P1 + P2) ctrA transcripts in this experiment was 0.38 (wild-type cultures), 0.14 (LS1 cultures), 0.34 (LS2144 cultures grown in PYEX) and 0.61 (LS2144 cultures grown in PYEG). The bar graph in Figure 2B summarizes the results of three separate RNase protection experiments using the CcrM depletion strain LS2144. As CcrM was depleted, ctrA P1 transcript levels doubled, but there was no substantial change in ctrA P2 mRNA levels. Thus both depleting CcrM and expressing the enzyme throughout the cell cycle altered ctrA P1 transcription, suggesting that CcrM methylation either directly or indirectly affects the transcription of this promoter. Because P2 transcription did not change in these experiments, we focused on the effect of the DNA methylation state on P1 transcription in subsequent experiments. Transcription of a ctrA P1–lacZ fusion integrated at different chromosomal locations The previous experiments using ccrM mutant strains showed that changing the timing or level of CcrM expression altered ctrA P1 transcription. To test the possibility that ctrA P1 transcription is directly regulated by the methylation state of the GAnTC site in the ctrA P1 promoter, we constructed a P1 transcription probe containing an Ω-ctrA P1–lacZ transcriptional fusion (pAR263). This reporter was integrated at three different sites on the chromosome: near the origin (site 1, generating NA1000 hrcAΩ::pAR263), approximately halfway between the origin and the terminus (site 2, generating NA1000 recA::Tn5Ω::pAR263), and near the terminus (site 3, generating NA1000 trpE::Tn5Ω::pAR263). The position of these sites relative to Cori is shown in Figure 3A. In each of these strains, chromosomal ccrM is transcribed from its native promoter so the timing and amount of CcrM expression is normal. Previous studies have demonstrated that DNA methylation at the sites used in this study varies during the cell cycle. GAnTC sites near Cori become hemimethylated soon after the initiation of DNA replication and remain hemimethylated until the end of S phase when the CcrM MTase is present and active (Stephens et al., 1996; Marczynski, 1999). In contrast, the GAnTC sites engineered into a transposon-based methylation probe integrated near the terminus (site 3) are hemimethylated only for a short period at the end of S phase. When this methylation probe was integrated midway between the origin and terminus (site 2), the GAnTC sites are hemimethylated for an intermediate period of time (Marczynski, 1999). The changes in the methylation state of these sites during the cell cycle are shown in Figure 3B. Activity of the P1–lacZ transcription probe varied dramatically when integrated at these three sites on the chromosome (Figure 3C, left panel). β-galactosidase activity was maximal when the probe was integrated near Cori (site 1), reduced by 60% when the reporter was integrated 1.1 Mb from Cori (site 2), and reduced by 85% when it was integrated near the terminus (site 3). Thus, ctrA P1 activity correlates well with the position of the P1 transcription probe on the chromosome, and reflects the period of time that GAnTC sites at these locations remain in the fully methylated state during the cell cycle (Marczynski, 1999). These results support the previous experiments indicating that hemimethylation of the ctrA P1 promoter is required for its full expression and further suggest that the effect is direct. To assess the role of CcrM in the direct regulation of ctrA P1 transcription, we generated a C(–25)T mutation in the P1 promoter that eliminated the GATTC methylation site (see Figure 1B). We then constructed a lacZ transcriptional fusion plasmid containing only the mutant P1 promoter (pP1UM-lacZ), introduced the plasmid into wild-type cells and measured promoter activity. Mutating the methylation site had little effect on P1 activity; promoter activity was 3250 ± 100 and 3980 ± 160 Miller units in wild-type cultures bearing the pctrA-P1 and pP1UM-lacZ transcriptional fusion plasmids, respectively. This is not surprising since the timing of P1 transcription was similar for both the wild-type and unmethylated promoters (Figure 4A). We also constructed a transcription probe containing the mutant ctrA P1UM promoter fused to lacZ (pAR579), integrated this reporter at the same three sites on the chromosome, and measured β-galactosidase activity. As shown in Figure 3C, the activity of the P1UM–lacZ transcription probe was maximal at site 1 and reduced by 20 and 32% at sites 2 and 3, respectively. This modest drop in promoter activity reflects the changes in the copy number of the reporter during DNA replication. During most of S phase, there are two copies of the reporter integrated near the origin (site 1), but only one copy of the reporter integrated near the terminus (site 3). The activity of both the wild-type P1–lacZ and the unmethylated P1UM–lacZ reporters was similar at site 1, reflecting the similar timing of the transcription of these promoters (Figure 4A). However, activity at sites 2 and 3 increased 2- and 5-fold, respectively, when P1 contained a point mutation that blocks CcrM methylation. These data indicate that changes in the methylation of P1 are responsible for the inhibition of promoter activity in the P1–lacZ strains at sites 2 and 3 and imply that full methylation of the P1 promoter represses ctrA transcription. When promoter activity is corrected for approximate gene dosage, we observed that P1–lacZ activity at site 3 was still significantly reduced. Figure 4.The timing of ctrA transcription when P1 cannot be methylated and when the ctrA gene is relocated near the chromosomal terminus. (A) The Caulobacter cell cycle is shown schematically. In synchronized wild-type cultures, ctrA P1 (open cirles) and P2 (closed circles) transcription was monitored using plasmids pctrA-P1 and pctrA-P2 (Domian et al., 1999). Samples were pulse-labeled with [35S]methionine at the indicated times, and β-galactosidase synthesis was assessed by immunoprecipitation. To determine the timing of P1 and P2 transcription in synchronous cultures with ctrA transcribed from an unmethylated P1 promoter (LS3551) or with ctrA located near the terminus (LS3355), total cellular RNA was isolated at 20 min intervals and ctrA transcript abundance was analyzed by primer extension. The primer extension products were resolved on a sequencing gel and quantitated with a PhosphorImager. (B) Immunoblot of CtrA in cells from synchronized wild-type and LS3355 cultures. An equivalent cell mass (based on OD600) was applied to each lane. Cell proteins were separated on SDS–12% polyacrylamide gels, and probed with an antibody to the C.crescentus CtrA protein. Download figure Download PowerPoint Temporally controlled transcription from the ctrA P1UM promoter In wild-type cells, ctrA transcription is temporally regulated during the cell cycle. P1 transcription is maximal early in S phase, while P2 transcription peaks in mid to late S phase (Figure 4A; Domian et al., 1999). To test the effect of expressing the intact ctrA gene from the unmethylated P1 promoter (P1UM) in vivo, we constructed a strain (LS3551) in which ctrA was inactivated on the chromosome, but that contained a plasmid-borne ctrA gene transcribed from the C(–25)T mutant P1 promoter and the wild-type P2 promoter. Primer extension and S1 analysis showed that both P1UM and P2 transcripts initiated from their native start sites (data not shown). To determine whether cells bearing an unmethylated P1 promoter show the same temporal pattern of transcription as those bearing the wild-type P1 promoter, we synchronized LS3551 cultures and monitored transcription by primer extension. As shown in Figure 4A, transcription from both the wild-type and the unmethylated P1 promoters was maximal early in the cell cycle. In addition, P2 was transcribed late in S phase in both wild-type and LS3551 cultures. Therefore, both the activity and the timing of ctrA P1 transcription remained unchanged when methylation of the P1 promoter was eliminated. However, as shown above, P1 transcription is repressed when the promoter is in the fully methylated state. Since P1 transcription is inhibited by feedback regulation from CtrA (Figure 1; Domian et al., 1999), full methylation must affect the initiation of P1 transcription. Changing the position of ctrA on the chromosome altered the temporal pattern of its transcription To determine whether the transcription of ctrA P1 is delayed when the chromosome is in the fully methylated state, we changed the chromosomal location of the native ctrA gene. We constructed a strain (LS3355) in which ctrA was inactivated at its wild-type position and instead placed close to the terminus of the chromosome, a site that remains fully methylated throughout most, if not all, of the cell cycle (Marczynski, 1999). The native ctrA gene and its promoters were integrated within the 400 bp region between the clpX and lon genes by homologous recombination (see Figure 5A). The clpX and lon genes are located 1.9 Mb from Cori, placing these genes near the terminus. Genomic PCR using primers located within the ctrA gene and in the sequence flanking either ctrA or clpX was used to confirm that ctrA was absent from its wild-type location and present adjacent to clpX (data not shown). To determine whether the chromosomal position of the ctrA gene influences its transcription in vivo, we isolated RNA from wild-type and LS3355 cultures, and assessed ctrA transcript levels by S1 nuclease assays. As shown in Figure 5B, the ctrA P1 and P2 transcripts were present in wild-type cells and initiated at the sites previously described (Domian et al., 1999). However, when ctrA was located near the chromosome terminus, transcription from the P1 and P2 promoters dropped and the bulk of ctrA transcription originated at P2FM, a new transcription start site located eight nucleotides upstream of the P2 start site (Figure 5B). These results were confirmed by primer extension analysis (data not shown). The location of the P2FM start site is shown in Figure 1B. Thus when PctrA is moved to a site that remains fully methylated throughout most of the cell cycle, P1 is inactivated. This is consistent with our observation that P1 transcription is repressed when the chromosome is in the fully methylated state. Furthermore, in the LS3355 strain, the native ctrA gene is transcribed from an alternate start site. Figure 5.Moving ctrA near the terminus changes the mRNA start sites and affects swarmer cell size. (A) Diagram of strain LS3355. The chromosomal copy of ctrA was inactivated and the ctrA gene and its promoters were integrated between the clpX terminator and the lon gene. (B) S1 nuclease mapping of ctrA transcription start sites in wild-type (NA1000) and LS3355 cultures. The first four lanes show a sequencing ladder generated using a primer with the same 5′ end as the S1 probe. The P1 and P2 transcription start sites were detected in wild-type cells, while P2FM was the predominant start site in LS3355 cultures. (C) The size distribution of swarmer cells in wild-type and LS3355 cultures. Swarmer cells were isolated from synchronized cultures at 20 min into the cell cycle, fixed in buffered neutral formalin, and examined by DIC microscopy. To estimate cell size, the length of at least 50 cells in each culture was measured. Download figure Download PowerPoint To determine whether moving ctrA to a site near the chromosomal terminus affects the temporal regulation of ctrA transcription, we synchronized cultures of LS3355 and monitored ctrA mRNA levels by primer extension analysis. When the native ctrA gene was located close to the terminus, P1 transcription was negligible throughout the cell cycle, while P2FM was transcribed in mid to late S phase (Figure 4A). Immunoblot analysis showed that CtrA was present in swarmer cells, rapidly degraded at the G1–S transition, and reappeared in pre-divisional cells in both wild-type and LS3355 cultures (Figure 4B). However, the reappearance of CtrA protein was delayed in the LS3355 cultures, reflecting the delay in ctrA transcription. Thus the period during which CtrA protein is absent during the cell cycle was prolonged when the only copy of the ctrA gene is moved to a position on the chromosome that remains fully methylated for the majority of the cell cycle. Although variability in the cell cycle could affect the timing of CtrA expression, it is unlikely to cause both the earlier disappearance and later reappearance of CtrA observed in the LS3355 strain. Changing the chromosomal location of ctrA affected the distribution of swarmer cell size Swarmer cells were isolated from wild-type and LS3355 cultures and examined by light microscopy. In LS3355 cultures, the swarmer cells were elongated and exhibited a broad distribution of cell lengths; 42% of the swarmer cells were longer than their wild-type counterparts (Figure 5C). This change in cell size was also observed in stalked and pre-divisional cells. In the LS3418 control strain, with the vector alone integrated between the clpX and lon genes, swarmer cell length was not affected (2.0 ± 0.2 μm). In addition, swarmer cell length was normal when a plasmid containing the native ctrA gene and promoter region (pSAL290) was introduced into strain LS3355. These data indicate that the cell elongation is due to faulty expression of ctrA P1 and not to changes in the expression of the clpX or lon genes. Thus prolonging the period of the cell cycle in which the cells remain free of CtrA results in an abnormal cell size distribution, suggesting that changes in the DNA methylation state of the ctrA promoter and the subsequent changes in the timing of CtrA expression influence cell physiology. However, it is possible that other effects of its chromosomal position may alter CtrA expression and contribute to the observed phenotype. Discussion The CtrA response regulator is a critical component of cell cycle control in Caulobacter. This DNA-binding protein directly regulates the transcription of 55 operons (Laub et al., 2002) and directly or indirectly controls ∼25% of the 550 cell cycle-regulated genes (Laub et al., 2000). CtrA also represses DNA replication initiation in swarmer cells by binding to the origin of replication (Quon et al., 1998). Not surprisingly, this essential protein is under multiple levels of control: cell cycle-regulated ctrA transcription, CtrA phosphorylation, and proteolysis of phosphorylated CtrA (CtrA∼P) (Domian et al., 1997, 1999). The ctrA gene is transcribed from two promoters, P1 and P2, which are active at different times in the cell cycle (Figure 4). Here we present evidence that DNA methylation silences the transcription of the ctrA P1 promoter and that this inhibition of P1 transcription affects cell physiology. Consequently, the methylation state of the chromosome adds another layer of regulation to the temporal expression of CtrA. Because the Caulobacter chromosome changes progressively from the fully methylated state at the start of S phase to the hemimethylated state at the end of S phase (Stephens et al., 1996; Marczynski, 1999), the differential methylation state of specific promoters could contrib
Referência(s)