Multipartite Regulation of rctB, the Replication Initiator Gene of Vibrio cholerae Chromosome II

Bacterial strains, phages, plasmids, and growth conditions.Relevant details of strains used in this study are described in Table 1.

CVC250, a recA derivative of the sequenced V. cholerae El Tor strain N16961 (CVC16) (Table 1) (17), was constructed as follows. First, about 3 kb of DNA (573022 to 576657) spanning the recA gene of N16961 was amplified by PCRusing primers PS39 (5′-ACATGCATGCGTGACACAATGAAACAGAAG CGAGAGGC) and PS40 (5′-TGCAGAGCTCCATCCCTGCGTTGGTAAACA ACAG) containing sites for restriction enzymes SphI and SacI, respectively, at their 5′ end. The PCR product after digestion with these enzymes was cloned into a suicide vector, pDS132 (25), yielding pPS25. Next, the recA gene was deleted and substituted with a Kan cassette as follows. pPS25 was digested with MfeI and NotI, and between the two sites a Kan cassette (about 1.2 kb, obtained from pRFB110 after digestion with HindIII) was ligated after blunting of all four ends, yielding pPS26 (pPSΔrecA::Kan). The plasmid was transferred to V. cholerae strain CVC205 by conjugation using SM10λpir/pPS26 as donor. The transconjugants were selected and purified on plates containing 25 μg/ml chloramphenicol (Cm), 40 μg/ml kanamycin (Kan), and 20 μg/ml tetracycline (Tet). Transconjugants were screened by colony PCR to confirm integration of pPS26 at the recA locus using screening primers PS41 (5′-GCAAAGTAGCCCGAGCGTCCTACGCTGGC) and PS42 (5′-CGTTGAGCGGTCGTGGCGCGAGTGTAGGCGCGC), which flanked the 3-kb recA locus, and primers PS37 (5′-CAGTCATAGCCGAATAGCCT) and PS38 (5′-CGGTGCCCTGAATGAACTGC)within the Kan gene. Out of 24 colonies tested, only 3 gave the expected PCR products. A Cm-sensitive but Kan- and Tet-resistant derivative was obtained from one of the colonies by sucrose selection (25). The recA locus was again examined by PCR, and the loss of recA function of the strain was confirmed by its sensitivity to UV light compared to the parent strain. The resultant strain was named CVC250.

Fragments containing the rctB control region were obtained from N16961 by PCR amplification. The fragments were fused to the lacZ gene present in promoter and protein fusion vectors pMLB1109 (21) and pMLB1034 (31), respectively, and the resultant plasmids are described in Table 1. The identity of all cloned fragments was verified by DNA sequencing. They matched with published sequences in all cases.

Both E. coli and V. cholerae strains were grown in L broth at 37°C, and log-phase cultures (optical density at 600 nm [OD₆₀₀] ≈ 0.4) were used to measure promoter activities.

Primer extension analysis.Primer extension analysis was done using α-³²P-labeled primer (Univ RF) and an E3030 kit (Promega, Madison, WI). Total RNA was isolated from DH5Δlac (33) carrying pDP309 grown in L broth to an OD₆₀₀ of ∼0.4 using an RNeasy Mini kit (QIAGEN). Typically, approximately 10 μg of total RNA was mixed with 1 pmol primer in a buffer containing 50 mM Tris-HCl, pH 7.5, 10 mM MgCl₂, 0.1 mM spermidine, and 5 mM dithiothreitol in an 11-μl reaction volume. The mixture was incubated at 58°C for 20 min to anneal the primer to RNA before the reaction was cooled at room temperature for 10 min in a Perkin-Elmer PCR system 2400. A mixture of 1.4 μl of 40 mM sodium pyrophosphate, 5 μl of 2× avian myeloblastosis virus primer extension buffer, 0.1 U of avian myeloblastosis virus reverse transcriptase (0.05 U μl⁻¹), and 0.6 μl water was added to the annealed reaction products and incubated at 42°C for 30 min. An equal volume (20 μl) of loading dye (98% formamide, 10 mM EDTA, 0.1% xylene cyanol, and 0.1% bromophenol blue) was added to terminate the reactions. A 2-μl aliquot of the reaction was loaded onto a 6% sequencing gel together with a sequencing ladder generated using the primer and plasmid identical to those in the primer extension reaction. The ladder was generated using an fmol DNA sequencing kit (Promega, Madison, WI). The bands were visualized using a Phosphoimager (Fujix BAS 2000).

Luciferase assay.The luciferase assay was performed using the luciferase assay system (Promega, Madison, WI). To 50 μl of culture, 40 μl of a BR8706 culture (OD₆₀₀ = 0.4) (carrier cells) and 10 μl of 1 M K₂HPO₄ (pH 7.8)-20 mM EDTA solution were added. The mixture was placed on dry ice for 5 min and allowed to equilibrate to room temperature for 15 min. The cells were then lysed with 300 μl of a lysis mix, and 10 μl of the lysed extract was added to 100 μl of the substrate in a flat-bottom 96-well microtiter plate (Costar 3596, Corning, NY). Light emission was measured for 1 s at intervals of 10 s for a total of 10 readings using a Wallace 1420-Victor Multilabel counter (Perkin-Elmer, Shelton, CT). The mean value of 10 readings was determined and used in the figures.

Western blot analysis.Two OD units of log-phase cells were resuspended in 100 μl of 1× sodium dodecyl sulfate sample buffer (NEB, Beverly, MA), boiled for 5 min, and centrifuged. Twenty-five microliters of the supernatant was loaded on a 7.5% Tris-HCl Criterion gel (Bio-Rad, Hercules, CA). After electrophoretic separation, the bands were blotted onto a nitrocellulose filter (Schleicher & Schuell, Keene, NH) using 1× CAPS [3-(cyclohexylamino)-1-propanesulfonic acid] transfer buffer (pH 11, USB) containing 10% methanol. The membrane was then blocked by incubation with 5% membrane-blocking agent (Amersham Biosciences, Piscataway, NJ) and treated with 1:2,000 diluted anti-RctB antiserum (Medicore, Gaithersburg, MD) and finally with 1:5,000 diluted anti-rabbit horseradish peroxidase-conjugated secondary antibody from donkey (Amersham Biosciences, Piscataway, NJ). The anti-RctB antiserum was raised against untagged RctB purified using an IMPACT system (NEB, Beverly, MA). The blot was developed in ECL Western blotting detection reagents (Amersham Biosciences, Piscataway, NJ), and the light intensities were captured by a LAS1000 system (Fujifilm, Stamford, CT).

Plasmid copy number.The plasmid copy number was measured as has been described previously (4).

Mapping of a promoter for rctB.A minichromosome carrying the two initiator genes rctA and rctB and the intergenic region (ig2) of chrII of the sequenced Vibrio strain N16961 replicates in E. coli, suggesting that the promoters for the two genes must be functional in the surrogate host (11). To identify the sequences responsible for rctB transcription, the first eight codons of the rctB open reading frame (ORF) and various sequences upstream were fused to lacZ in a pBR322-derived promoter-cloning vector, pMLB1109 (21) (Fig. 1). The sequences included the minimal origin and some of the elements identified earlier to be involved in the negative regulation of replication initiated from oriCII_VC. In E. coli, a strong promoter activity was detected in all of the fragments tested. The shortest one containing 209 of bp upstream sequences from the putative start codon of the rctB ORF (pDP336) (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 1.

Identification by deletion mapping of a region containing a promoter for the initiator gene rctB. The top line shows the region involved in chromosome II replication. It consists of rctB, the origin (oriCII_VC), and elements of negative control of replication. The region has two kinds of repeats, 11-mers (open triangles) and 12-mers (filled triangles), and these were used as guides for choosing deletion end points (next five lines). The coordinates are as published previously (17). All fragments ended at coordinate 1157 and included the first eight codons of the rctB ORF. They were cloned in front of the lacZ reporter in a promoter-cloning vector (pMLB1109), and their promoter activities were measured after the resultant plasmids were transferred to E. coli (DH5Δlac) and V. cholerae (CVC250). The arrows indicate the orientation in front of the lacZ gene. The reporter (β-galactosidase) activities were measured and presented in Miller units as described previously (20). Values presented are averages from at least two independent transformants, each assayed in duplicate.

In order to locate the start of the rctB message, a primer extension analysis was performed with total RNA isolated from an E. coli strain, DH5Δlac, carrying pDP309 (Fig. 2A). A clear signal was detected corresponding to a start site at an A at coordinate 1092. Suitable matches to consensus −10 and −35 boxes were detected at the expected positions. To confirm the promoter, some of the bases were mutated in the putative −10 and −35 boxes (Fig. 2B). The activity of the promoter was reduced in both cases, confirming that the sequences are essential elements of the rctB promoter, hereafter called rctBp (Fig. 2C). The results were qualitatively similar in a recA mutant derivative of N16961, CVC250, indicating that rctBp is also functional in the native host (Fig. 1 and 2C). The promoter activities in general were considerably reduced in V. cholerae compared to E. coli, suggesting that host-specific genes could negatively regulate rctBp in trans.

• Open in new tab
• Download powerpoint

FIG. 2.

Precise location of rctBp. (A) Primer extension results showing the sequence around the start site of the rctB message. The starting base is shown in bold. (B) The predicted −10 and −35 boxes, assuming that +1 is at coordinate 1092, are bracketed. Also shown are the predicted start codon (in bold at 1134) and a putative SD sequence (underlined) of the rctB ORF. The promoter region has four heptameric sequences (boxes) containing GATC that are partially homologous to 11- and 12-mer repeats. Their sequence relationships are shown in the box at the bottom. Also shown are two potential ToxT binding sites (gray arrows) overlapping the −35 and −10 boxes. (C) Promoter activities when the putative −10 and −35 boxes of rctBp were mutated. The mutated sequences are shown above the boxes in panel B.

Regulation of rctBp.Since initiator genes are often autoregulated, we determined the effect of RctB in trans on the activity of rctBp. The protein was supplied from an arabinose-inducible promoter, P_BAD, from plasmid pTVC11 (Fig. 3; Table 1) (16). The initiator expression was varied by varying the inducer (arabinose) concentration, and the host used (BR8706) carried a constitutively expressing arabinose transport system to avoid all-or-none induction at suboptimal inducer concentration (12, 19). pTVC11 also contained the firefly luciferase gene downstream of rctB, so that both genes were transcribed from P_BAD and luciferase activity could serve as a reporter for the relative levels of rctB expression. When rctB expression was induced with arabinose, the activity of rctBp present in pDP309 initially decreased but then reached a plateau, with significant activity still remaining (Fig. 3A, pDP309/pTVC11). Measurement of luciferase activity confirmed that rctB transcription increased in the plateau region in proportion to the inducer concentration (Fig. 3C) and suggests that the incomplete repression of rctBp was unlikely to be due to insufficient RctB supply. This was further confirmed by Western analysis (Fig. 3D, lanes 7 to 13). Increasing the arabinose concentration from 0.002 to 0.02% did increase RctB concentration about 2.1-fold.

• Open in new tab
• Download powerpoint

FIG. 3.

Transcriptional autorepression of rctBp. (A) The promoter activities from two different fragments were determined: when they were present in transcriptional fusion vector pMLB1109, the resultant plasmids were called pDP309 and pDP310 (Fig. 1), and when in translational fusion vector pMLB1034, the resultant plasmids were called pDP311 and pDP312. To study autorepression, various levels of RctB protein were supplied from pTVC11 (pTVC12+rctB) where the protein was under arabinose-inducible promoter P_BAD present in vector pTVC12. The host was BR8706 in these experiments, with a constitutive source of arabinose transporter. (B) To show specificity of promoter repression by RctB, activity of another fusion, repAp-lacZ (present in pALA326), was tested. RepA was supplied from pALA169, where repA is constitutively made. pST52 is the vector used to clone repA that generated pALA169. (C) The arabinose induction of P_BAD was confirmed by measuring the activity of the luciferase gene present downstream of rctB in pTVC11. pTVC12 is identical to pRFG110 except that rctB and the downstream luciferase gene replaced the lacI-yfp fusion gene. (D) RctB production after P_BAD induction was measured by immunoblotting with antibody against RctB. Extracts from equal ODs of cells were added to all lanes, except that the ODs of V. cholerae cells were two times higher (lanes 14 and 15). Purified RctB (5, 10, and 20 ng) (arrow) were added to extracts from vector (pTVC12)-carrying cells for reference purposes (lanes 1 to 3). The band inferred as RctB (arrow, lanes 14 and 15) comigrated with purified RctB when it was mixed with the V. cholerae extract (data not shown), confirming the identity of the protein. Samples were loaded in duplicates (lanes 4 to 15).

To confirm the specificity of RctB in promoter repression, another well-characterized autoregulated promoter, repAp of plasmid P1, present in plasmid pALA326, was used (3). Supplying RctB from pTVC11 repressed the activity of rctBp as before but not that of repAp (Fig. 3B). Under identical conditions, repAp was efficiently repressed when the cognate repressor RepA was supplied in trans from pALA169. The repression activity of RctB, therefore, appears specific to its own promoter.

A second level of regulation of rctBp was evident when the repressed levels of transcription in plasmids pDP309 and pDP310 were compared (Fig. 3A, pDP309/pTVC11 and pDP310/pTVC11). The repressed level was lower in pDP309. An effect of upstream sequences was also evident in V. cholerae strain CVC250 (Fig. 1). The level of repression depended in a complex way on the extent of upstream sequences present, pDP310 being the least repressed of all. (pDP309, which contains oriCII_VC as well as the negative control locus, could not be established in the V. cholerae strain, most likely because of incompatibility with the chromosomal origin as suggested previously [11].) Thus, upstream sequences seem to regulate the promoter both in E. coli and in V. cholerae. When a few fragments containing the additional upstream sequences present in pDP309 compared to pDP310 were provided in trans from the pACYC184 vector (yielding pDP348, pDP349, and pDP350 [Table 1]), they were ineffective in exerting any repressive effect (data not shown). The cis action of upstream sequences suggests that they aid in repression, possibly by DNA looping or altering the structure of the promoter DNA (14).

Additional promoters of rctB.To understand the basis of why the repression in pDP309 was higher than that in pDP310, we searched for sequences upstream of rctBp for additional promoters not regulated by RctB. Weak promoter activities were detected when fragments without the rctBp sequences were analyzed by fusion to lacZ, but the levels of their activities (≤800 Miller units [MU]) were not high enough to fully account for the plateau level seen from plasmid pDP310 (Fig. 4). For pDP310, the plateau level was around 1,500 MU, but the maximal promoter activity obtained from sequences upstream of rctBp was around 300 MU in the presence of RctB. It appears that additional promoters are not adequate to explain the difference in rctBp activity between pDP309 and pDP310 in the presence of RctB.

• Open in new tab
• Download powerpoint

FIG. 4.

Search for additional promoters of rctB. The fragments shown were fused to lacZ of the promoter-cloning vector pMLB1109, as in Fig. 1. The region containing rctBp is deleted in all but pDP309. In the +RctB column, the protein was supplied from pTVC11 after adding 0.2% arabinose to the culture; otherwise, the experiments were done as described in the legend to Fig. 3. In the −RctB column, the vector pTVC12 replaced pTVC11.

Translational control of rctB.The fragments containing rctBp present in plasmids pDP309 and pDP310 were fused to lacZ in the protein fusion vector pMLB1034 so that the rctB ORF continued into the reporter gene (31). The repression profiles from the resultant plasmids, pDP311 and pDP312, in the presence of various levels of RctB were similar to those seen from the transcriptional fusion plasmids pDP309 and pDP310 (Fig. 3A, dashed lines). These results indicate that RctB is autoregulated at the level of transcription only. However, we show below that some translational regulation is possible in the presence of rctA.

Since replication initiation from oriCII_VC also requires a second initiator, rctA, we examined whether it could be responsible for the higher level of repression of rctBp seen in V. cholerae. A putative promoter for rctA was identified by gene fusion studies (D. Pal, unpublished results), and a fragment excluding this promoter but containing the rctA ORF, including the putative Shine-Dalgarno (SD) sequence and considerable sequences downstream to the ORF (coordinates 1072274 to 1072845), was chosen. The fragment was placed under P_BAD control in plasmid pDP354. To test the effect of rctA in conjunction with RctB, rctB was expressed constitutively from pTVC13, a derivative of pET28a. Although the gene was under the control of a phage T7 promoter, the basal level of expression was sufficient, even though no T7 RNA polymerase was present in these experiments. The transcription initiated from unidentified vector promoters apparently produced enough RctB to support replication of an oriCII_VC plasmid in E. coli (T. Venkova- Canova, unpublished results). To the extent that it was analyzed, no significant effect of rctA expression alone or in conjunction with RctB could be seen on rctBp (Fig. 5A). In these experiments, the rctBp-lacZ transcriptional fusion was transferred to a λ vector (λRS45 [32]), and the resultant phage, λDKC372, was integrated at the att_λ site in the chromosome. In other words, no effect of rctA could be seen on rctBp even when the promoter was present in a single copy. The results were similar when a shorter fusion present in pDP336 was tested (data not shown).

• Open in new tab
• Download powerpoint

FIG. 5.

Effect of rctA on rctB expression. The rctA effect was determined either alone or in conjunction with RctB. (A) A transcriptional rctBp-lacZ fusion as in pDP336 (Fig. 1) was used, but the fusion was present in an integrated λ prophage in CVC444 (Table 1). RctB was supplied either from an inducible source, pTVC11, or constitutively from pTVC13 (a pET28a derivative, but the cells did not have a source of T7 polymerase) when rctA was also present in the same cell. When present, rctA was under P_BAD control as in pDP354. (B) Same as panel A except the fusion was translational as in pDP351. (C) Same as panel B except rctBp was replaced with repAp and the resultant strain called BR3239. pALA162 was used to supply RepA to confirm the repressibility of repAp. (D) The extent of complementarity between the rctA message and the leader of rctB message is shown by vertical lines. The first 48 nucleotides of the leader of the rctB mRNA are shown in the top line. The predicted sequence of the 5′ end of rctA message as made from the P_BAD promoter present in pDP354 is shown below. The predicted start codon of rctA ORF is shown in gray (coordinate 246), and the vector-derived 5′ end sequences are in italics. The chromosomal sequences covering rctA present in pDP354 span the coordinates 845 to 192.

We next determined the effect of rctA on an rctB′-′lacZ translational fusion after it was transferred to the chromosome as described above (CVC446 [Table 1]). In rctA-carrying cells, β-galactosidase activities were slightly but reproducibly lower, indicating that the gene could repress, albeit to a small extent, rctB translation (Fig. 5B). This effect was also seen when the cells had a source of RctB (pTVC13), but the protein did not seem to alter the level of repression by rctA. The repression also did not increase when rctA expression was induced with arabinose and the induction was confirmed by the luciferase reporter activity. A partial complementarity between the rctA RNA and leader region of rctB mRNA could be detected by sequence comparison (Fig. 5D). Thus, there could be a weak translational regulation of rctB in the presence of rctA. We note that even in the presence of rctA and RctB, considerable rctBp activity remained (about 100 MU) (Fig. 5B). rctA and RctB also did not have any significant effect on the heterologous promoter, repAp, which could be efficiently repressed by RepA from pALA162 (Fig. 5C).

Additional control of rctBp in V. cholerae.Before invoking additional factors in the repression of rctBp, we wanted to determine whether the promoter activity is intrinsically lower in V. cholerae than in E. coli. To account for possible differences in promoter activity in the two strains, the P1 plasmid repAp, present in pPP112 (24), was again used for normalization purposes in the two strains (Table 2). Furthermore, plasmid copy numbers were measured to normalize for possible gene dosage differences between the two strains. The activity of repAp was lower in V. cholerae (946 MU versus 2,159 MU in E. coli); this could be due to the lower copy number of pPP112 in that strain (8.3 versus 25 in E. coli). However, the copy number was not sufficient to explain the difference in rctBp activity in the two strains. The copy number of pDP310 was roughly sixfold lower (4.4 versus 27), whereas the promoter activity was about 16-fold lower (279 versus 4,500 MU). Thus, the involvement of factors other than rctA and RctB cannot be ruled out in the regulation of rctBp in V. cholerae. There are two direct repeats overlapping the putative −10 and −35 boxes that are largely homologous to ToxT transcriptional factor binding sites (35), suggesting that ToxT could be an additional transcriptional repressor of rctBp.

Stimulation of rctBp activity by E. coli DNA adenine methylase (Dam).The minimal region tested for autoregulation spanned the coordinates 925 to 1157 (as cloned in pDP336). RctB apparently binds within this region for autoregulation. However, the region is devoid of the 11- and 12-mer repeat sequences, the two likely candidates for RctB binding (11) (Fig. 1, open and shaded triangles). However, we noted that a hexamer that is also present in the 11- and 12-mer repeats, ATGATC, is present between the −10 and −35 boxes (Fig. 2B). Three other similar sequences are also present in the minimal region tested for autorepression (Fig. 2B, boxes). These short sequences appear to be the best candidate for RctB binding for autorepression.

Since the short sequences contain an adenine methylation site (GATC), we tested the role of the dam gene in the regulation of rctBp in E. coli. The dam gene is also essential for oriCII_VC activity (11). rctBp activity was reduced in a dam mutant strain, particularly evident in pDP336 (Table 3). The degree of autorepression could also be somewhat stronger in the mutant strain, suggesting that RctB could be competing better with RNA polymerase for binding to the promoter region. Although small, the Dam effect was not due to plasmid copy number changes (data not shown). The effect on autorepression was also unlikely to be due to a change in RctB expression, since the unrepressed promoter activity from pDP309 and pDP336 was affected differently in the dam mutant strain.

RctB level and the copy number of an oriCII_VC-based replicon.The regulation of initiator genes is most easily justified when they are found to be rate limiting for replication, as in plasmids R1 (22) and pT181 (26). Even when not rate limiting, regulation of initiators is required to guarantee their supply without being excessive, which could be detrimental to replication or cell growth (5). To the extent that it was overproduced, RctB was not found to be detrimental to E. coli growth (Fig. 3D). To test the first possibility, we determined the copy number of an oriCII_VC-based minichromosome (pTVC31) when RctB was supplied in trans from pTVC13, where the protein was under P_BAD control. The copy number did increase with increasing concentration of RctB, suggesting that the protein can be rate limiting for replication (Fig. 6B). Thus, RctB is not only essential for replication initiation but can also be its regulator.

• Open in new tab
• Download powerpoint

FIG. 6.

Relationship of RctB concentration and the copy number of a minichromosome carrying oriCII_VC. The origin region was present in pTVC31, and various levels of RctB were supplied from pTVC11 by changing the arabinose concentration. A constant volume of a culture of pNEB193 was added to each sample before the start of plasmid DNA isolation to normalize for any loss of DNA during the isolation procedure. The DNAs were linearized using SalI before running on the gel to recover each plasmid in one band for easier comparison of DNA concentration.