Figure 1.
ParB2 spreads into the replication origin of V. cholerae chromosome II (chrII).
The origin region has three functional units: incII, oriII and rctB. incII is required for controlling replication, oriII for initiating replication, and rctB for supplying the chrII-specific replication initiator protein, RctB. RctB binds to two kinds of site: the 11- and 12-mers (arrow heads) and 39-mers (black rectangles). The origin also has binding sites for DnaA and IHF, and the two promoters PrctA and PrctB. The rctA locus has a 39-mer and a binding site for ParB2, parS2-B. The bottom panel shows binding of ParB2 (grey profile) and RctB (black profile) in the origin and flanking regions, determined by ChIP-chip using specific antibodies (denoted as α) against the two proteins. The dashed line represents the average signal for ParB2 (1.1±1.1) over the entire genome. The corresponding value for RctB is 1.1±0.9. parS2-A and parS2-C are the nearest neighbors of parS2-B. The error bars here and elsewhere represent one standard deviation.
Figure 2.
ParB2 can silence PrctA and PrctB in the presence of parS2-B in E. coli.
The top map shows the origin region as in Figure 1. The lines below the map show the regions of the origin that were cloned upstream of a promoter-less lacZ gene and tested for promoter activity in E. coli. Fragments marked 1A–3A are present in pTVC122-124, respectively, and are orientated to record PrctA activity. The fragments marked 1B–3B are present in pTVC210-11 and pAS1, respectively, and are oppositely oriented to record PrctB activity. The white bars represent β-galactosidase activities obtained in the presence of an empty vector (pACYC184), while the grey bars indicate the activities in the presence of pTVC236 that supplied ParB2 at about 14-fold the physiological level. The copy number of lacZ-carrying plasmids was about 60 per cell, assuming there are four oriC copies in newborn E. coli cells in LB. The activities shown are mean values from three cultures inoculated with independent single colonies.
Figure 3.
Silencing of a reporter promoter PrepA by ParB2 without requiring parS2-B.
A three-plasmid system was used in E. coli (cartoon in the top right corner) to monitor the span of ParB2-mediated silencing within incII. One plasmid carried various incII DNA fused to PrepA, itself fused to a promoter-less lacZ gene, and the other two plasmids carried Plac-parB2 (pTVC501) and Pbad-rctB (pTVC499), to supply ParB2 and RctB, respectively. The PrepA-lacZ plasmids from the top were pTVC234, −239, −505, −504 and −509. The white and the grey bars represent β-galactosidase activities in the presence of uninduced (−) and induced (+) levels of ParB2. The induced level was about 10-fold the physiological level and the uninduced level was about an order of magnitude lower (Figure S2). The induced level (+) of RctB was about two-fold higher than the physiological level (Figure S1 of [27]). The uninduced level (−) of RctB was undetectable. The copy number of lacZ-carrying plasmids was about 60 per cell. The dotted lines represent β-galactosidase activities with induced levels of ParB2 alone, and are provided for comparison with activities under other conditions.
Figure 4.
parS2-B and the central 39-mer are the only specific ParB2 binding sites in chrII origin.
Binding was tested by EMSA using purified ParB2. The DNA fragments tested are indicated below the origin map. Fragments 1 to 10 were obtained by PCR from plasmids pTVC291, −526, −221, −248, −515, −270, −514, −228, −139 and −400, respectively. In addition to the origin sequences, all of the fragments had 100 bp of adjoining vector (pTVC243) sequences at each end. The fragment used as the negative control had only the identical vector sequences. Each fragment was run in three consecutive lanes with 0, 0.7 and 1.4 µM ParB2, marked by −, +, and ++, respectively, and the order was maintained for all the fragments. The stars mark the positions of specific ParB2 binding. The graph below shows the fraction of total DNA retarded by ParB2 at the indicated concentrations. The mean retarded fractions from three independent gels were plotted.
Figure 5.
A ParB2 half-site is necessary but not sufficient for 39-mer binding.
The experiments were done similarly to Figure 4, except for the insert sequences were as indicated. DNA fragments were used at 4 nM each and ParB2 at 1.4 µM (+). As in Figure 4, the insert sequences were flanked by BamHI (5′-GGATCC) and XhoI (5′-CTCGAG) sites of the vector, pTVC243. The fragments used in EMSA had additional 100 bp flanking sequences of the vector beyond those restriction sites. Fragments 1–7 were from plasmids pBJH200, pTVC222, pTVC132, pBJH216, pTVC120, pBJH221 and pBJH201, respectively.
Figure 6.
ParB2 and RctB bind simultaneously to rctA, but competitively to the central 39-mer.
The fragment in the top panel contained the entire rctA, which has parS2-B and a 39-mer for binding ParB2 and RctB, respectively. The fragment in the bottom panel contained only the 39-mer central to incII. The fragments were PCR amplified from pTVC291 and pTVC222, respectively. (The rctA fragment is identical to the fragment 1 of Figure 4 but the 39-mer fragment does not have the natural flanks of the fragment 5 of Figure 4). The fragments (2 nM each) were subjected to EMSA with purified RctB and ParB2, each at two concentrations: 3 nM (+) and 30 nM (++) for RctB, and 0.7 µM (+) and 1.4 µM (++) for ParB2. Arrows indicate the bands representing single or simultaneous binding.
Figure 7.
Increase in copy number of oriII plasmids by ParB2 in the absence of parS2-B in E. coli.
The copy number was measured in E. coli by supplying RctB from Pbad-rctB present in pTVC499, and ParB2 from Plac-parB2 present in pTVC501 (cartoon on the top). Lines below the origin map indicate the extent of origin DNA present in different plasmids. The copy-numbers were from cultures containing 0.002% arabinose that supplied a near-physiological level of RctB, and either no IPTG (− ParB2) or 100 µM IPTG (+ ParB2) that supplied ParB2 at about 10-fold the physiological level (Figure S2). The copy numbers are the mean values from three cultures inoculated with independent single colonies. The copy number 1 corresponds to four copies of oriII is per cell. % increase (last column) = 100×[Copy # (+ParB2)−Copy # (−ParB2)]/Copy # (−ParB2).
Figure 8.
Effect of a roadblock on ParB2-mediated stimulation of oriII activity in E. coli.
Two oriII plasmids, pTVC20 and pBJH218, isogenic except for about 100 bp extra sequences carrying an array of five P1 RepA binding sites (iterons) in the latter, were used to transform BR8706 (recA minus) cells without (#1, #2) and with (#3, #4) an integrated λ prophage (λDKC311). The prophage supplied P1 RepA constitutively. The recipient cells also had pTVC499 and pTC501 to supply RctB and ParB2, respectively, as in Figure 7. Cultures from single colonies were grown to OD600≅0.1 with (+) or without (−) 100 µM IPTG. One µl of the cultures was streaked onto respective plates with and without IPTG for comparison of colony sizes. For determining generation times, the cultures were diluted 50× and grown to early log phase. The generation times are from five different cultures, and they decreased in the presence of IPTG in all cases, but less so under the condition RepA was expected to bind to P1 iterons and create roadblock (Table row #4). Note that in the IPTG carrying plate (+I), the growth was least robust for #4 compared to the other three. In that sector, a larger colony forming revertant is also conspicuous (+I plate; arrow). In these experiments, ParB2 was supplied at about 10× the physiological level, and oriII number was approximately one per cell.
Figure 9.
Model of activation of V. cholerae chromosome replication by Par proteins using multiple mechanisms.
The upper diagram shows ParB2 interaction with the chrII origin by two mechanisms: spreading from the centromeric site parS2-B into a neighboring 39-mer and direct binding to another 39-mer. We suggest that these interactions promote chrII replication by out-competing initiator binding to the 39-mers. The lower diagram shows activation of chrI replication by ParA1 interaction with DnaA, which is negatively controlled by ParB1. For both chromosomes, the Par proteins target the most powerful regulators, by DNA-protein interactions for chrII and by protein-protein interactions for chrI.