Flexibility in Repression and Cooperativity by KorB of Broad Host Range IncP-1 Plasmid RK2

doi:10.1016/j.jmb.2005.03.062

J. Mol. Biol. (2005) 349, 302–316

Flexibility in Repression and Cooperativity by KorB of Broad Host Range IncP-1 Plasmid RK2 Lewis E. H. Bingle, Donia P. Macartney, Anaı¨s Fantozzi Susan E. Manzoor and Christopher M. Thomas* School of Biosciences University of Birmingham Edgbaston, Birmingham B15 2TT, UK

KorB, encoded by plasmid RK2, belongs to the ParB family of active partitioning proteins. It binds to 12 operators on the RK2 genome and was previously known to repress promoters immediately adjacent to operators OB1, OB10 and OB12 (proximal) or up to 154 bp away (distal) from OB2, OB9 and OB11. To achieve strong repression, KorB requires a cooperative interaction with one of two other plasmid-encoded repressors, KorA or TrbA. Reporter gene assays were used in this study to test whether the additional KorB operators may influence transcription and to test how KorB acts at a distance. The distance between OB9 and trbBp could be increased to 1.6 kb with little reduction in repression or cooperativity with TrbA. KorB was also able to repress the promoter and cooperate with TrbA when the OB site was placed downstream of trbBp. This suggested a potential regulatory role for OB sites located a long way from any known promoter on RK2. OB4, 1.9 kb upstream of traGp, was shown to mediate TrbA-potentiated KorB repression of this promoter, but no effect on traJp upstream of OB4 was observed, which may be due to the roadblocking or topological influence of the nucleoprotein complex formed at the adjacent transfer origin, oriT. Repression and cooperativity were alleviated significantly when a lac operator was inserted between OB9 and trbBp in the context of a LacIC host, a standard test for spreading of a DNA-binding protein. On the other hand, a standard test for DNA looping, movement of the operator to the opposite face of the DNA helix from the natural binding site, did not significantly affect KorB repression or cooperativity with TrbA and KorA over relatively short distances. While these results are more consistent with spreading as the mechanism by which KorB reaches its target, previous estimates of KorB molecules per cell are not consistent with there being enough to spread up to 1 kb from each OB. A plausible model is therefore that KorB can do both, spreading over relatively short distances and looping over longer distances. q 2005 Elsevier Ltd. All rights reserved.

*Corresponding author

Keywords: transcriptional repression; plasmid maintenance; plasmid transfer; DNA looping; DNA silencing

Introduction Control of promoter activity depends on the direct interaction of RNA polymerase (RNAP) with DNA as well as the presence of one or more activators or repressors at the promoter. Integration of multiple signals at a promoter depends on how Abbreviations used: RNAP, RNA polymerase; tsp, transcription start-point. E-mail address of the corresponding author: [email protected]

different regulatory proteins interact with the promoter DNA or with RNAP and the net outcome of activation or repression. Since the promoter region can become crowded, the ability of some regulatory proteins to act at a distance from their binding sites can be important and may be essential for the outcome in terms of gene expression. Regulatory circuits controlling the expression of genes on IncP-1 plasmids illustrate two important principles of general interest: the ability of a repressor to act from a promoter-distal operator and the ability of repressors to act cooperatively.1

0022-2836/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

Repression and Cooperativity by IncP-1 KorB

Both of these effects depend on communication between different repressors or between a repressor and RNAP. The repressor protein that is the focus of this study, KorB, is encoded by, and acts on, the 60 kb broad host range IncP-1 plasmid RK2.2 Several plasmid-encoded protein partners may interact with KorB to perform roles in global regulation of plasmid genes and stable plasmid inheritance.1,3 Indeed, KorB is not a strong repressor when acting alone, the maximum inhibitory effect observed in our reporter gene assays being approximately tenfold, even when expressed in trans from a highly induced tac promoter. Many bacterial and plasmid genomes encode proteins of the ParB family, which are involved in the active partitioning of chromosome or plasmid copies to daughter cells at cell division, and are homologous to KorB.3,4 These homologues are found throughout the bacterial world, on plasmids and on the chromosomes of most bacterial species, a notable exception being the chromosome of Escherichia coli. The binding of KorB to its operator sequence is modulated by its interaction with IncC,5–7 a homologue of the ParA proteins that generally interact with ParB partners to effect partitioning. In regulatory mode, KorB represses the promoters of genes involved in plasmid transfer, vegetative replication and stable maintenance in the host cell in a way that is potentiated by its interaction with IncC,7 and by cooperative interactions with the global regulators KorA and TrbA.8,9 Of the plasmid ParAB homologues, the IncP-1 IncC and KorB proteins are more closely related to their chromosomal equivalents than to other well-studied plasmid systems such as SopA-SopB of F plasmid and ParA-ParB of the “plasmid” P1 prophage.3,10–13 KorB is known to bind at 12 sites (OB1–OB12) of sequence 5 0 -TTTAGCSGCTAAA-3 0 on the RK2 genome (Figure 1).14 According to their position relative to the known promoters, the OB sites may be classified into three groups: class I OBs are centred 45/46 bp upstream of the transcription start-point (tsp) and 10 bp upstream of the K35 hexamer, class II sites are within 45–154 bp of the tsp and class III sites are more than 500 bp away from any known promoter (Figure 1). KorB has been shown to repress transcription through both class I15–18 and class II sites.19 However, the regulatory role, if any, of the six class III OB sites on the RK2 genome is unknown. KorB repression may either utilise a single mechanism that acts efficiently from both promoter-proximal and promoter-distal sites, or may require distinct mechanisms. In vitro footprinting experiments on korAp and trbBp promoter fragments indicate that KorB allows RNAP simultaneous access to the promoter where it is bound,19,20 and some KorB deletion mutants show a phenotype of improved DNA binding combined with reduced or no repressor activity.10 Both of these lines of evidence suggest that repression does not result from a simple exclusion

303 of RNAP from the promoter. Thus KorB repression may require the formation of a multiprotein–DNA structure or repressosome capable of preventing functioning of RNAP.21 The repressor action of KorB seems to be due to prevention of isomerisation of the promoter from closed to open complex, via what could conceivably be inhibitory contacts with RNAP or changes in promoter DNA conformation.19,20 Cooperativity in repression by KorB together with KorA correlates with the observed synergism of DNA-binding when both proteins are present, and we propose that cooperativity with TrbA occurs via the same mechanism.9 It seems most likely that these cooperativities result from direct protein–protein interactions or from longrange interactions mediated by DNA. A common mechanism that can be involved in such cases is DNA looping between proteins bound to two widely separated DNA sites.22,23 An alternative mechanism for action at a distance is illustrated by the DNA-silencing activity that has been demonstrated for some homologues of KorB, in which promoters are made inaccessible to RNAP, possibly by protein polymerisation along the DNA strand.24–27 However, in these cases the levels of ParB proteins are considerably higher than that observed for KorB in vivo (e.g. 7000 dimers of ParB in an E. coli cell carrying P1 prophage compared with 500 dimers of KorB in a cell carrying RK2).28,29 In other regulatory systems where looping is known to be involved, there is a helical periodicity in the effect of operator movements on regulation, i.e. moving the DNA binding-site halfway around the helix reduces regulation.30–32 Comparison of trbBp sequences of RK2 and the related IncP-1 plasmid R751 reveals that while OB and promoter sequences are highly conserved, the intervening distance differs by 4 bp, i.e. approximately half a helical turn. This would seem to favour an explanation for KorB’s action as a repressor binding at a distance from RK2 promoters that is not dependent on DNA looping; for example, a silencing type spreading along the DNA strand, which allows KorB to reach the promoter and interact with RNAP there. Here, we show for the first time KorB repression from class III operators, indicating that all the KorB binding sites on RK2 may play a regulatory role. To help distinguish between various models for how KorB interacts with both RNAP and the small repressors KorA/TrbA, we describe the results of varying the distance between a KorB binding site and its target in such a way that the orientation of bound KorB on the face of the DNA should be rotated by up to 1808. We describe the effects of introducing an operator for a second DNA-binding protein between KorB operator and promoter. The results indicate that there are few limits on the action of KorB, a plausible explanation for which is that KorB can both spread and loop so that it can do either, depending on the context.

304

Repression and Cooperativity by IncP-1 KorB

Figure 1. Locations of KorB operators (OB ) in relation to known promoters of plasmid RK2. (a) The RK2 plasmid genome with distances between class III (orphan) OB sites and nearest promoter(s) indicated (centre OB to upstream end of K35 motif). (b) Locations of class I, II and III OB sites in relation to nearest operators. Filled circle, KorB operator (OB); shaded circle, KorA operator (OA); open circle, TrbA operator (OT).

Results KorB can still repress trbBp when OB9 is more than 1.5 kb away To assess the distance constraints on KorB action, a series of spacers of various lengths were inserted between OB9 and the trbB promoter, resulting in a range of O B –transcriptional start-point (tsp)

distances of between 189 bp (pMZT39, wild-type) and 1563 bp (pLB105) (Figure 2). These constructions were cloned in front of the xylE reporter gene, prior to determination of regulation by KorB and cooperativity with TrbA (binding at the promoter) in vivo. Expression from these promoters was assessed first in the absence of any RK2-encoded regulators (data not shown). Compared with a construction

Repression and Cooperativity by IncP-1 KorB

305

Figure 2. Insertion of spacer DNA between KorB operator (OB) and recombinant trbBp does not prevent cooperative regulation by KorB and TrbA. (a) Construction of promoter fragments with spacer DNA between OB and recombinant trbBp. B, BssHII site. (b) Effect of increasing OB–promoter distance on promoter activity regulated by pRK24 in trans. The RK2 derivative pRK24 should supply KorB and TrbA at natural levels. Promoter– (KorB) operator distance (x-axis) is measured from transcriptional start point (tsp) to the centre of OB. Circles, mean XylE activities with pRK24 in trans; errors bars indicate standard deviations. No OB indicates activity for plasmid pLB49, which carries the trbBp promoter without upstream sequences including KorB operator. All activity values are normalised to the activity value for promoter without pRK24 in trans. (c) Effect of increasing the OB–OT distance on KorB-TrbA cooperativity. Values shown are mean repression indices (from three replicates). KorB was expressed from pDM1.21, TrbA was expressed from pMZT24; protein expression was induced with 0.05 mM IPTG. The centre of the TrbA operator OT is 12 bp upstream of the tsp. O B –O T distances (measured from the centre of OB to the centre of OT ): pLB49, no upstream OB site; pMZT39, 176 bp; pLB103, 520 bp; pLB105, 1550 bp.

that carried only the trbBp promoter and TrbA operator, introduction of the upstream OB site at its wild-type distance caused a slight decrease in reporter gene expression (less than 10%), suggesting that sequences upstream of the promoter are not having a great effect on promoter activity. Reporter gene activity was assayed also in bacteria that carried the RK2 derivative pRK24, which should provide KorB and TrbA proteins at approximately natural levels (Figure 2).33 To compensate for any small changes in promoter activity caused by introduction of the spacer DNA, each regulated promoter activity value was normalised by dividing by the activity of the same promoter in the absence of pRK24. In the cells carrying pRK24, introduction of the upstream OB site at its wild-type

distance decreased expression from the promoter by approximately 17-fold. In the absence of OB the presence of pRK24 had essentially no effect. As the OB was moved further away, this repression decreased only slowly. At the greatest operator– promoter distance (1563 bp) normalised XylE activity in trans to pRK24 was approximately eightfold lower than the unregulated promoter. To check whether the overall level of regulation observed with pRK24 in trans could be the effect of just KorB and TrbA, rather than involvement of additional plasmid-encoded regulators, experiments were carried out with both TrbA and KorB provided in trans from tac promoters (Figure 2). We previously reported that Western blotting shows KorB levels from uninduced tac promoter are

306 similar to the natural level from RK2.34 However, induction with 0.05 mM IPTG was necessary to measure repression by KorB and TrbA alone. Under these conditions the major repression seen depends on the presence of both repressors. The factor by which the repression observed was greater than the product of repressor indices for each repressor alone (cooperativity index) was generally in the region of tenfold. Repression by TrbA and KorB acting alone was largely unaffected by changes in OB9-trbBp distance (there was a 22% decrease in KorB repression at an OB–tsp distance of 1563 bp). The combined repression by KorB and TrbA was still significant, at 40% below the wild-type level, at the greatest O B–O T distance tested (1550 bp, pLB105). The repression by pRK24 is greater than that observed by KorB alone supplied from an expression plasmid, even after over-expression through addition of IPTG, suggesting that this level of repression must be due to the combined effects of KorB and TrbA. Since essentially no repression was seen with TrbA alone, there is very likely to be an interaction between these repressors that strengthens the action of each repressor to produce a much more potent inhibitory effect. Class III KorB operators on plasmid RK2 can have a regulatory role The finding that KorB was able to cooperate with TrbA to give significant repression of trbBp at a distance of up to 1563 bp (Figure 2) suggested that there may be a regulatory role for some, or all, of the class III operators on RK2, which in some cases are closer than this to the nearest promoter (Figure 1). A large BspEI-ApoI restriction fragment from the Tra1 region of RK2, which included traJp-traKp-OB4traGp, was cloned into the promoter probe vector pPT01 so that the traG promoter was firing into xylE with O B 4 located 1990 bp upstream (pLB13, Figure 3). When this construction was tested for KorB regulation of the traG promoter, either alone or in combination with TrbA (binding at traGp), no strong effect was seen (Figure 3). This may be due to the traJ promoter upstream of OB4 masking the

Repression and Cooperativity by IncP-1 KorB

regulation of traGp, as traJp is regulated by the relaxosome proteins binding oriT and by TrbA but is not known to be regulated by KorB.35 Therefore, PCR was used to make a second DNA fragment that carried only traGp and the upstream OB4 (pLB21, Figure 3). On examination of the plasmid profile after restriction digestion on ethidium bromidestained agarose gels, it became apparent that the copy number of pLB13 was lower than that of pLB21. This difference, quantified using image analysis software (Quantity One; BioRad Laboratories) was found to be 1.5–2-fold. In reporter assays with pLB21, a small repression by KorB alone was observed but this became more significant when TrbA was present also (Figure 3). The values obtained for plasmid copy number were used to adjust results from the reporter gene assays, resulting in a final repression index of 2.2-fold for KorB acting alone (with induction by 0.05 mM IPTG) and a cooperativity index of 11-fold for KorB with TrbA, leading to a total repression index of 26-fold. This is the first demonstration of a regulatory effect for a class III OB. KorB is able to repress at a distance from an operator downstream of the promoter Although we have reported that KorB at OB10, centred 68 bp downstream of the tsp of trbAp causes repression of this promoter, our subsequent observation that traJp is apparently unaffected by KorB at OB4 (L.E.H.B. & C.M.T., unpublished results) raised the question of how universal is the effect of downstream repression by KorB. Two synthetic promoter fragments were constructed with either OB9 (TTTAGCCGCTAAA; pLB133) or a mutant version of this operator (TccAGCCGC TAAA; pLB134), positioned 130 bp downstream of the K35 motif of trbBp (the promoter fragment was cloned without its normal upstream OB9). KorB repression alone and in cooperation with TrbA binding at trbBp were assessed for these constructs in reporter gene assays and compared with the regulation of trbBp via its natural upstream OB9 (Figure 6). All three promoters showed a similar

Figure 3. Long-distance cooperativity between KorB binding at OB4 and TrbA binding at traGp. Diagrams indicate RK2 fragments cloned in OB4-traGp reporter constructs pLB13 and pLB21. Grey box, TrbA operator; black box, KorB operator. Promoter strength is the mean unregulated value (plasmid copy number corrected) relative to pLB13G standard deviation. Ri is the mean repression index ratio (fold repression) for (K) KorB, (T) TrbA. KorB was expressed from pDM1.21, TrbA was expressed from pMZT24, induced with 0.05 mM IPTG.

307

Repression and Cooperativity by IncP-1 KorB

Helical position of the KorB operator has no effect on KorB repression or cooperativity with TrbA or KorA

Figure 4. Constructions used to determine the effect of rotation of OB position around the DNA helix relative to the promoter and to TrbA/KorA operators at trbB and kfrA promoters. (a) Recombinant trbBp with proximal OB. (b) Recombinant kfrAp with distal OB.

level of activity. The construct with a downstream OB9 (pLB133) was repressed sevenfold by KorB, compared to tenfold for the upstream O B9 (pMZT39). This may reflect the asymmetry of the sequences flanking OB, which are known to have a role in binding KorB.14 When the downstream OB site was mutated (pLB134) KorB repression was almost completely removed, indicating that the observed repression of pLB133 was indeed due to KorB acting via the downstream operator. Repression by TrbA acting alone was weak in each case, but when TrbA and KorB were provided together in trans to pLB133, there was a clear cooperative regulation of trbBp, indicating that KorB is able to interact with TrbA at trbBp from a downstream position. There was also a significant increase in repression of pLB134 when both proteins were present, suggesting that KorB is still binding to the operator to some small extent.

Since the distance from which KorB can act seems very flexible, it was important to determine whether the orientation of the KorB binding site on the DNA affected the repression observed. A series of synthetic promoters, based on trbBp, were constructed where the OB site was placed just upstream of the K35 region, as in a class I promoter, and then the distance from the centre of the OB to the tsp was increased in steps of 5 bp from 46 to 56 bp (Figure 4(a)). These promoters were assayed for transcriptional activity in reporter gene assays, with TrbA and KorB provided in trans (Table 1). Moving the promoter-distal KorB operator to a proximal position had no significant effect on the ability of KorB to repress trbBp (Table 1; pLB40 compared with pMZT39,8 which has wild-type spacing). Moving the OB site back along the DNA strand away from the promoter by approximately a half (pLB117) or a whole (pLB118) helical turn depressed promoter activity slightly in each case; this may explain the small increases in repression we observed (approximately 50% higher than with pLB40 or pMZT39). There is certainly no reduction in KorB repression for either pLB117 or pLB118. The trbB promoter is repressed also by TrbA protein, which binds to an operator overlapping the K10 hexamer.36,37 When both KorB and TrbA were present, all four synthetic promoters were repressed very strongly; TrbA-KorB cooperativity did not seem to be affected by any of the rearrangements of OB site position relative to promoter. A second series of reporter plasmids was constructed in which the kfrA promoter fires into xylE (Figure 4(b)). This promoter is regulated by KorB binding at a distal OB site (centre OB–tsp distance of 81 bp).38,39 Mutant promoters with an XhoI site between the OB and OA sites were obtained via sitedirected mutagenesis at position K60. This was followed by small insertions or deletions at the XhoI site, between the OB site and tsp, as shown in Figure 4(b).38 xylE reporter plasmids were constructed from these promoters and assayed for

Table 1. Repression and cooperativity by KorB binding at positions proximal to the trbB promoter

Reporter plasmid pMZT39 (wt trbBp) pLB40 (proximal OB) pLB117 (proximal OB C5 bp) pLB118 (proximal OB C10 bp) pLB49 (no OB) a

OB–tsp distancea (bp)

Relative promoter strengthb

Ric (KorB)

Ric (TrbA)

Ric (KorBCTrbA)

Cid (KorBCTrbA)

189 45 50 55 –

1.00G0.12 1.17G0.05 0.67G0.04 0.83G0.06 1.10G0.23

7.0G1.0 7.1G1.5 10.7G0.9 11.4G1.0 1.3G0.2

1.7G0.2 1.6G0.2 2.1G0.1 2.3G0.2 1.4G0.1

635G99 417G85 1233G158 1363G399 1.5G0.1

54 37 56 51 1

Measured from tsp to the centre of OB. Mean promoter strength relative to pMZT39Gstandard deviation. Mean repression index ratiosGstandard deviation. KorB was expressed from pDM1.21, TrbA was expressed from pMZT24, induced with 0.05 mM IPTG. d Cooperativity index. b c

308

Repression and Cooperativity by IncP-1 KorB

Table 2. Effect of altering OB–promoter distance at kfrAp on KorB repression

Reporter plasmid pDM300 (wt kfrAp) pDM301 (XhoI site) pDM302 (C13 bp) pDM304 (C4 bp) pDM305 (K7 bp) pDM306 (K8 bp) pDM307 (K17 bp)

OB–tsp distancea (bp)

Relative promoter strengthb

Ric

81 81 94 85 74 73 64

1.00G0.13 1.13G0.28 1.27G0.43 1.44G0.28 1.14G0.23 1.16G0.17 1.28G0.20

8.3G2.7 5.8G2.3 5.3G1.1 6.3G2.4 5.0G0.8 5.6G2.2 4.5G1.6

a

Measured from tsp to the centre of OB. Mean unregulated promoter strength relative to pDM300G standard deviation. c Mean repression index ratiosGstandard deviation. KorB was expressed from pMMV811, induced with 0.5 mM IPTG. b

transcriptional activity in the presence of KorB repressor (Table 2). None of the changes in operatorpromoter distance had any significant effect on the regulation of kfrAp (pDM301) by KorB alone. The cooperative interaction between TrbA and KorB seems to be remarkably flexible in its toleration of different operator spacings (see Table 1 and above). A second plasmid-encoded regulatory protein, KorA, interacts cooperatively with KorB, and is better characterised than TrbA in vitro.9,40 However, KorA has previously been shown to cooperate with KorB only in situations where the KorA operator (OA) is overlapping the promoter and the KorB operator (OB) is promoterproximal. To investigate the nature of the interaction between KorA and KorB, we used the reporter plasmids described above, in which the kfrA promoter fires into xylE (Figure 4(b)). When both KorB and KorA were present, the level of repression was much higher than when either was present alone (Table 3). This positive cooperativity was quantified by dividing the index of repression obtained when both proteins are present by the product of the repression indices obtained for single proteins, giving a measure of the direct or indirect interaction of KorB and KorA. We found that this cooperativity was not affected significantly by any of our changes to the inter-operator distance (Table 3). To see if this interaction was promoterspecific, we made a synthetic promoter fragment

that was a hybrid of the KorA regulated korAp with the upstream region and distal OB site from trbBp, where cooperativity is known to occur between TrbA and KorB. Regulation by KorB and KorA was assessed in reporter gene assays (Table 4). Replacing the normal upstream sequences of korAp with the region upstream of trbBp did not seem to have any significant effect on the unregulated promoter strength (compare pDM3.1 with pLB125; Table 4). A cooperative interaction was clearly taking place between KorB binding at the distal operator and KorA binding at korAp (pLB125), although this interaction does not seem to be as strong as the proximal KorB-KorA (pDM3.1) or the distal KorBTrbA (pGBT63) interactions. The action of KorB at a distance is partially blocked by insertion of a protein binding site between OB and the regulated promoter The tolerance to helical permutation described above is not consistent with repression by classical looping. An alternative model, spreading, can be investigated by determining whether a second DNA binding protein can act as a roadblock to the process.24,41 If KorB acts by polymerising along the DNA strand, then such a roadblock may impede this process, leading to a reduced efficiency of repression. We introduced a BssHII restriction site between OB and trbBp and ligated-in synthetic oligonucleotide linkers carrying binding sites for lac repressor proteins (Figure 5). The roadblock linkers were asymmetric and were inserted in both orientations, resulting in two pairs of constructs in which the roadblock protein binding sites were placed on opposite faces of the DNA helix (the centre of the Lac operator was moved by 5 bp). A third linker was the same length but had no lac operator sequence. We placed these reporter constructs in trans to pRK24, which would supply the RK2 repressors KorB and TrbA at approximately wild-type levels, and measured XylE activity with and without the addition of IPTG to the growth medium, which should prevent LacI binding and thereby allow KorB to repress the promoter. The reporter plasmid pLB49, which has no upstream OB site and therefore should not be regulated by KorB, was included as a negative control.

Table 3. Effect of altering OB–promoter distance at kfrAp on KorA-KorB cooperativity

Reporter plasmid pDM300 (wt kfrAp) pDM302 (C13 bp) pDM304 (C4 bp) pDM305 (K7 bp bp) a

OB–OA distancea (bp)

Relative promoter strengthb

Ric (KorB)

Ric (KorA)

Ric (KorBCKorA)

Cid (KorBCKorA)

36 49 40 29

1.00G0.02 0.87G0.10 1.03G0.03 0.76G0.05

2.6G0.1 2.6G0.3 2.8G0.5 2.7G0.7

4.9G0.8 4.8G0.7 5.5G1.0 4.9G0.8

225G43 215G25 205G7 171G21

18 17 13 13

Measured from centre of OB to the centre of OA. Mean unregulated promoter strength relative to pDM300Gstandard deviation. Mean repression index ratiosGstandard deviation. KorB was expressed from pDM1.21, KorA was expressed from pGBT37, induced with 0.05 mM IPTG. d Index of cooperativity. b c

309

Repression and Cooperativity by IncP-1 KorB

Table 4. Cooperativity at a distance between KorA and KorB

Reporter plasmid

Relative promoter strengtha

Rib (TrbA)

Rib (KorA)

Rib (KorB)

Rib (KorBCTrbA/KorA)

Cic

pGBT63 (trbBp distal OB9) pDM3.1 (korAp proximal OB) pLB125 (korAp distal OB)

1.00G0.06 1.90G0.27 1.78G0.19

1.6G0.1 – –

– 90.1G12.0 80.6G26.3

8.6G0.7 3.6G0.5 4.8G0.7

238G92 4644G1032 1816G974

17 14 5

a

Mean XylE activity normalised to pGBT63Gstandard deviation. Repression index ratios (calculated as XylE activity in presence of repressor(s)/XylE activity in absence of repressor)Gstandard deviation. KorB was expressed from pDM1.21, KorA was expressed from pGBT37, and TrbA was expressed from pMZT24. All cultures were induced with 0.05 mM IPTG. c Index of cooperativity. b

The results from these experiments indicate that, in cells growing in the absence of IPTG, repression of trbBp is reduced when there is a lacO site between the OB and the promoter (Figure 5(b), compare constructions pLB112 and pLB113 to pLB114). This suggests that binding of lac repressor between OB site and promoter partially impedes communication between KorB binding at its promoter and RNAP or TrbA binding at the promoter. The roadblock constructs pLB112 and pLB113 still show a reduced promoter activity (by about 50%) in the absence of IPTG, when compared to trbBp that is not regulated by KorB (pLB49 has no upstream OB site). There is only a small difference in the alleviation of repression caused by lac repressor proteins binding to opposite faces of the DNA helix (compare pLB112 and pLB113) (Figure 5(b)).

Discussion Understanding the ability of regulatory proteins to influence the activity of a promoter a long distance from their binding sites is of considerable importance in modelling genome-wide expression.42 The IncP-1 plasmids represent an interesting model system because of the highly conserved binding sites for KorB protein, half of which are more than 500 bp from the nearest promoter. From the two known roles for KorB, these sites should be involved in either regulation or partitioning. Here, we have explored further their role in regulation. The first question was whether the orphan sites, distant from any promoter, could play a regulatory role. By increasing the distance between trbBp and OB9, we showed that KorB was still capable of a significant level of

Figure 5. Insertion of a DNAbinding protein roadblock between distal OB and recombinant trbBp partially alleviates KorB repression. (a) Construction of promoter fragments with a roadblock between distal O B and recombinant trbBp. X, XhoI restriction site; B, BssHII restriction site. OT, TrbA binding site; OB, KorB binding site. lacO, roadblock protein (lac repressor) binding site linker inserted into BssHII site (destroying site). (b) Promoter activity shown in the presence of RK2 repressor proteins (supplied in trans from pRK24). All values are normalised to activity without pRK24 in trans. Filled bars, no IPTG in the growth medium; open bars, 2 mM IPTG in the growth medium.

310

Repression and Cooperativity by IncP-1 KorB

Figure 6. Repression of trbBp by KorB acting from a downstream operator. Black rectangle, KorB operator (OB ); grey rectangle, TrbA operator (OT ). Distances marked are between the centre of OB and the beginning (upstream end) of the K35 promoter motif. The cross over the OB site in pLB134 indicates that it is mutated.

repression over the longest distance tested (greater than 1.5 kb) even when the regulatory proteins were supplied at natural levels. That the ability to mediate KorB repression at such a distance is not a unique property of this promoter was demonstrated by the repression of traGp, which has OB4 located 1990 bp upstream, by KorB and TrbA in a reporter gene analysis. The apparent lack of repression of the traJp promoter located 575 bp upstream of OB4 may be due to roadblocking of KorB spreading by relaxosome proteins binding around oriT. Alternatively, if DNA looping is required for long-distance regulation by KorB, the lack of effect on traJp might be because there is an intrinsic bend in the oriT DNA to which TraK binds, wrapping the DNA around itself, and this nucleoprotein structure may reduce the chances of the apex of a supercoil moving to the region between OB4 and traGp.43 It is clear, therefore, that KorB can act at a considerable distance and an important question is how this occurs. Described in the Introduction are two wellestablished ways in which a repressor can act at a distance, either by spreading along the DNA from its initial binding site or by looping between distant sites in a supercoiled molecule. The former mechanism has been proposed as the way that other members of the ParB protein family silence adjacent genes.24,27 One previous observation that argues against spreading as the only mode of action for KorB is the estimation of KorB concentration in vivo at approximately 500 dimers per E. coli cell.28 Assuming five to ten copies of RK2 per cell,28 this allows an average of only four to eight dimers of KorB per operator. The footprint of KorB covers about 21–23 bp.20 Given that KorB is a largely acidic protein, despite its DNA-binding region having a local basic character,44 it seems likely to need a number of KorB dimers together to form a protein core around which DNA could wrap like a histone. Thus, while the estimated level of KorB per operator might be sufficient to spread up to 200 bp, the amount of DNA in a nucleosome, it seems insufficient to spread 2 kb, as would be needed to explain its ability to act on traGp from OB4 described here. To test experimentally whether KorB must spread along the DNA to repress at a distance, a roadblock experiment was conducted similar to those

performed in the E. coli bgl and prophage P1 systems.24,41 Our results showed that the LacI/ lacO complex reduced KorB repression by a factor of about 5, but did not prevent the action of KorB completely. If LacI acts only as a block to spreading and does not affect the ability of the DNA on either side to loop, this suggests that a major mode by which KorB represses trbBp is by spreading. An alternative explanation for these results is that lac repressor tetramers are “handcuffing” pairs of reporter plasmids, and that this causes steric hindrance of KorB binding or looping. It seems unlikely that such handcuffing would affect the binding of KorB binding more strongly than the relatively massive RNAP. Since the LacI/lacO complex has little effect (!20%) on promoter activity for pLB112 and pLB113 in the absence of KorB and TrbA, we therefore think an effect on protein binding is disfavoured. We also disfavour the idea that the LacI/lacO complex is affecting loop formation between KorB binding at the OB site and RNAP binding at the promoter. This is because insertions of n C0.5 turns (where n is an integer) described here (particularly data shown in Figure 4 and Tables 1–3) and previously,38 that should affect looping, do not affect KorB repression. The most plausible explanation for these results appears therefore to be that the LacI/lacO complex is preventing the non-covalent polymerisation of KorB from its initial binding site (OB) by forming a roadblock. This is thus consistent with KorB action being mediated, at least partly, by spreading. Spreading might even be able to explain all the effects of KorB if either the estimates of KorB level referred to in the paragraph above were inaccurate, if KorB does not spread evenly but is concentrated around certain sites, or if the way that KorB spreads out from an operator site is not a simple occupation of successive 20 bp segments so that our calculation of the amount of KorB needed is wrong. While spreading appears to fit the road block data, it is remarkable that changing the helical face of DNA to which KorB binds, at positions either distal or proximal to a promoter, had little or no effect on regulation or cooperation at that promoter. Previous in vitro data19,20 have suggested that KorB may not simply occlude RNAP from promoter binding, but rather interacts with promoter-bound RNAP to prevent isomerisation. When KorB is

Repression and Cooperativity by IncP-1 KorB

repressing a promoter cooperatively with KorA or TrbA it may well act via a different mechanism, for example exclusion of RNAP from the promoter. On a simple spreading model, KorB should still be able to interact with RNAP or a second regulator because, as it is moved away from the promoter, additional KorB molecules can bind to fill the gap. However, since the KorB footprint on DNA, in terms of protection against DNase I or hydroxyl radical cleavage is about 21–23 bp,20,28 it would be surprising if movement by 5 bp would create enough space for an extra KorB molecule to bind. This suggests that KorB is flexible in its regulatory interactions with promoter DNA/RNAP and with cooperating regulators, KorA or TrbA. Recent structural data on KorB44,45 and Spo0J of Thermus thermophilus46 suggest that ParB proteins may wrap around the DNA helix when they bind to their operators. If there were multiple patches on KorB that can interact with RNAP or KorA/TrbA then the rotation or orientation of KorB may simply bring different regions into play. Alternatively, KorB may distort or melt the DNA locally so that there is increased flexibility that allows KorB to rotate more freely. If there is a very high potential energy for DNA loop formation, via KorB binding at OB and interacting with some factor at the promoter, and this exceeds the energy required to untwist the DNA and align promoter and operator sites for a productive KorB interaction, then it may reduce the effect of moving the operator position around the DNA helix. Highly energetically favoured loop formation would be consistent with the ability of KorB to act over large distances. However, in the comparable DeoR system of E. coli, where looping can occur over distances of several kilobases, there is still an observable helical phase to repression.47 The apparent flexibility of KorB in its interactions is in contrast to other well-studied regulators; for example, in the lac and ara operons of E. coli.30–32 At proximal positions, any helical periodicity in repression should be highly pronounced due to the stiffness of a relatively short looping DNA, in contrast to our observation that KorB can repress equally well from opposing faces of the DNA.32,48,49 An example is the CRP/FNR family of transcriptional activators found in Gram-negative bacteria, which also interact with RNAP, but exhibit a strict requirement for the spatial relationship between binding site and promoter in class I activation, where only the insertion of integral helical turns is tolerated.50–54 In the case of lac repressor, the promoter sequence context and relative operator position also have a strong effect on regulation.55 In these systems, optimal regulation is generally observed at a naturally occurring operator– promoter distance, and increasing this separation resulted in a reduced activation. In RK2 three naturally occurring class I (promoter proximal) OB sites on the RK2 genome (at korAp, trfAp and klaAp) have an identical K35 hexamer–OB spacing to pLB40, despite there apparently being no

311 consequence from increased spacing. It may be that this common spacing is due to duplication of promoters and operators, as is clearly the case for the klcA, kleA and kleC promoters of RK2, followed by divergence of intra-motif regions.56,57 While repression via KorB binding at class I sites could be due to direct interaction with RNAP and at class II sites could be due to spreading, as argued above, there does not seem to be enough KorB per cell for its repression of promoters at distances as great as 1.5 kb or more to be explained in this way. It seems likely, therefore, that interactions occur between distant sites (OB and promoter) as they contact each other in the supercoiled plasmid molecule via reptation (slithering).58 The idea that KorB may both loop and spread would be consistent with experiments on ParB protein of P1, which has been observed to both silence24 (spreading) and pair plasmid molecules (the same mechanism, operating in cis, could result in looping),59 although ParB is not thought to be a global transcriptional regulator. The results presented here for KorB show that its ability to repress is very flexible and are not consistent with action solely via looping or via spreading. The data suggest an interaction between KorB and RNAP (or other proteins) with sufficient flexibility to operate between molecules on either the same or opposing faces of the DNA strand. This could possibly be mediated via a direct but highly flexible protein–protein interaction between the proteins or via an effect of KorB on local DNA topology, which would then affect promoter melting and/or binding of KorA or TrbA. Recent structural studies have indicated that the wellestablished C-terminal dimerisation domain, which holds the monomers together in solution, is joined to the DNA-binding domain by a putative flexible linker.44,45 There is a second, more N-terminal, dimerisation region that can function in the complex with DNA, so that an operator may accommodate two dimers, each with a free monomer that could either promote spreading by attracting other KorB molecules or contact RNAP or a second repressor (Figure 7).10,46 In conclusion, an interaction of this flexibility leading to gene regulation is highly unusual in the bacteria. The only examples of which we are aware come from eukaryotic systems, in particular the Gal4 activator, where DNA looping in combination with a DNA flexibility-enhancing action of high mobility group (HMG) proteins allows activation from a wide range of positions.60,61 Future studies will need to assign specific functions to different parts of KorB in addition to the DNA-binding domain, dimerisation domains and IncC interaction domain.10 C-terminal deletions are known to abolish repression at a distance.19 It may be that a mutational analysis can have differential effects on the ability to loop and to spread and thus provide a way for determining how important each activity is for repression in different situations.

312

Repression and Cooperativity by IncP-1 KorB

Figure 7. Model for the interactions of KorB with other transcription factors and with DNA based on results presented here and in the reports of recent structural studies.44,46 (a) KorB binds and encloses its operator as a dimer, the predominant form found in solution. Alternatively, it may form an operator-tetramer complex ((b)), which may occur in cis and lead to polymerisation/spreading along the DNA strand ((d)) or may occur in trans, resulting in either looping (dotted lines) or pairing of DNA molecules ((c)). If KorB encounters the second repressors KorA or TrbA (e) or RNA polymerase (f) at a promoter, then it can interact with them to repress transcription. Domains shown interacting are purely for illustration: we do not know which domain of KorB interacts with the C terminus of KorA/TrbA or with RNA polymerase.

Materials and Methods

recommended by the manufacturers, using Sprint or OmnE thermal cyclers (Hybaid).

Bacterial strains and plasmids Cloning of promoter fragments The E. coli K12 strains used for DNA manipulations were DH5a [supE44 DlacU169 (f80 lacZ DM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1]62 and JM109 [recA1 supE44 endA1 hsdR17 gyrA96 relA1 thi D(lac-proAB) F 0 (traD36 proABC lacIq lacZ DM15)].63 XylE assays were carried out in E. coli K12 strain C600KK [thr-1 leu-6 thi-1 lacY1 supE44 tonA21 galK3].64 Bacteria were grown in L-broth65 at 37 8C with shaking at 200 rpm; solid media were obtained by the addition of agar at 1.5% (w/v). Antibiotics were added, when required, at the following concentrations: ampicillin (Ap), 100 mg/ml; kanamycin (Km) 50 mg/ml, streptomycin (Sm), 30 mg/ml. Plasmids constructed and used in this study are described in Table 5. Purification, manipulation and sequencing of plasmid DNA Plasmid DNA was prepared routinely by an alkalinelysis miniprep method.66 When a higher level of purity was required, the Wizard Plus SV miniprep kit (Promega) or Plasmid Midi kit (Qiagen) was used. Sequencing was performed on an automated 373A DNA Sequencer or 3700 DNA Analyser (Applied Biosystems) using a dyeterminator kit (Big-Dye; PE Applied Biosystems). DNA manipulations were carried out using standard techniques, 67 or according to the manufacturer ’s instructions. PCR Polymerase chain reaction (PCR) was performed using a Taq-Pwo DNA polymerase mixture (Expand, Roche), or the proofreading enzymes Pwo (Roche) or KOD (Novagen). Thermal cycling was performed as

Promoter fragments for reporter gene assay, derived as described below, were digested with BamHI and ligated into the unique BamHI site of the xylE reporter vector pPT01.68 All constructs were checked by DNA sequencing, across the restriction sites used in cloning and along the length of any PCR derived region. trbBp with proximal OB site To construct the reporter plasmid pLB40, which has a synthetic promoter based on trbBp, with a proximal OB site cloned in front of the xylE gene (Figure 4), the following procedure was employed. Two synthetic oligonucleotides, 5 0 -AATTCGGATCCTTTAGCCGCTA AAGTTCTTGCAGGC-3 0 and 5 0 -TCGAGCCTGCAAGAA CTTTAGCGGCTAAAGGATCCG-3 0 , were annealed together (in 10 mM Tris–HCl (pH 8); 50 mM NaCl; 1 mM EDTA) by heating to 95 8C for two minutes, followed by gradual cooling to 25 8C over a period of 45 minutes. These annealed oligonucleotides make a synthetic K35 region with proximal OB site and EcoRIXhoI sticky ends. A BamHI fragment containing trbBp with an XhoI site inserted between the K10 and K35 regions was excised from pMZT398 and cloned into pUC18. After digestion with EcoRI (site in pUC18 polylinker) and XhoI to remove the K35 and distal upstream OB site, the annealed linker was ligated in its place. The resulting promoter was excised as a BamHI restriction fragment. The proximal OB site-promoter fragments cloned in plasmids pLB117 and pLB118 were constructed using the method described above, with synthetic linker DNA molecules made by annealing together the following pairs of oligonucleotides: pLB117, 5 0 -TCGAGCCTGCAAGAA

313

Repression and Cooperativity by IncP-1 KorB

Table 5. Plasmids used in this study Plasmid

Replicon

Selective marker(s)

pDM1.2 pDM1.21 pDM300 pDM301 pDM302 pDM304 pDM305 pDM306 pDM307 pGBT30 pGBT37 pGBT63 pLB1 pLB25 pLB49 pLB100 pLB101 pLB102 pLB103 pLB104 pLB105 pLB112 pLB113 pLB114 pLB40 pLB117 pLB118 pLB125 pLB133 pLB134 pMZT24 pMZT39 pPT01 pRK24

IncQ IncQ pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pMB1 pMB1 pSC101 pMB1 IncQ pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pSC101 pMB1 pSC101 pSC101 IncP-1a

SmR SmR KmR KmR KmR KmR KmR KmR KmR ApR ApR KmR ApR SmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR KmR ApR KmR KmR TrpEC, TcR, ApR

Description

Reference

tacp expression vector tacp-korB expression vector OB2-kfrAp-xylE OB2-kfrAp-xylE with XhoI site between OB site and promoter OB2-kfrAp-xylE with 13 bp insertion between OB site and promoter OB2-kfrAp-xylE with 4 bp insertion between OB site and promoter OB2-kfrAp-xylE with 7 bp deletion between OB site and promoter OB2-kfrAp-xylE with 8 bp deletion between OB site and promoter OB2-kfrAp-xylE with 17 bp deletion between OB site and promoter tacp expression vector tacp-korA expression vector trbBp-xylE with distal upstream OB9 pUC18 derivative with modified MCS tacp-trbA expression vector trbBp-xylE with wild-type OT trbBp with BssHII site between O-B9 and –35 motif tsp–OB centre distance 278 bp tsp–OB centre distance 366 bp tsp–OB centre distance 533 bp tsp–OB centre distance 636 bp tsp–OB centre distance 1563 bp lac binding site inserted between OB & trbBp lac binding site inserted between OB & trbBp (opposite face of helix) No binding site spacer inserted between OB & trbBp trbBp-xylE with proximal upstream OB; OT–OB centre-centre distance 45 bp trbBp-xylE with proximal upstream OB; OB centre–tsp distance 50 bp trbBp-xylE with proximal upstream OB; OT–OB centre–centre distance 55 bp korAp C distal OB9 in pPT01 trbBp with downstream OB9 trbBp with downstream OB–M3 mutation tacp-TrbA expression vector trbBp-xylE with distal upstream OB; tsp–OB centre distance 189 bp xylE promoter probe vector KmS derivative of RK2

17 17 This work This work This work This work This work This work This work 39 39 72 This work This work 36 This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work 37 8 68 33

CCTGACTTTAGCGGCTAAAGGATCCG-3 0 and 5 0 -AAT TCGGATCCTTTAGCCGCTAAAGTCAGGTTCTTGCAG GC-3 0 pLB118, 5 0 -AATTCGGATCCTTTAGCCGCTAAA GTCAGG TAGTCTTCTTGCAGG C-3 0 and 5 0 -TCGAG CCTGCAAGAAGACTACCTGACTTTAGCGGCTAAA GGATCC G-3 0 . Insertion of spacer DNA between OB and trbB promoter The BamHI fragment carrying OB9 and trbBp from reporter plasmid was cloned into the BamHI site of pUC18. A BssHII restriction site was inserted between OB9 and trbBp by inverse PCR using primers with the BssHII recognition sequence at their 5 0 ends (CGGCGCGCGCATCATATCGACATCCTC and ATAT GATGCGCGCGCCGACGGCCCGCAGAG). 69 Spacer fragments, ligated into this BssHII site, were obtained by BssHII digestion of DNA from the mupirocin gene cluster and should all be from within a single long transcriptional unit (i.e. transcriptionally inert). Promoters with no spacer inserted (pLB100) and with spacers of varying lengths (pLB101–105) were cut from the resulting pUC18-derived plasmids by digestion with BamHI. traG promoter fragments A cloning vector (pLB1) with multiple cloning site (MCS) suitable for cloning large restriction fragments of the Tra1 region of RK2 was constructed by cloning a

linker between BamHI and EcoRI sites of pUC18, destroying the EcoRI site and inserting recognition sequences for: BamHI-BsrGI-PpuMI-BspEI-AclI-ApaIApoI-BamHI. To make pLB13, a large BspEI-ApoI restriction fragment from the Tra1 region of RK2 was cloned it into pLB1, allowing excision of the fragment by BamHI restriction digestion for cloning into pPT01. pLB21 was constructed by PCR amplification of a fragment, with BamHI recognition sites at each end, stretching from just upstream of the KorB operator OB4 (primer: TGGATCCGAGGGCAGAGCCATGAC) to just downstream of traGp (primer: AGGATCCATAGGACA GCGCACCAAC).

trbBp with downstream KorB operator pLB133 (WT downstream OB site) and pLB134 (mutant downstream OB site) were constructed as follows. A DNA fragment was amplified from an RK2 template using one primer binding at OB9 which introduced a BamHI recognition site (WT, GGATCCTTCCTTCTTTAGCCGC TAAAACG; mutant, GCGGATCCTTCCTTCTccAGCCG CTAAAACG) together with a primer binding between OB9 and trbBp and introducing a BglII site (AAGA TCTGCGCACCCGCCCGATGCCA). This fragment, after digestion with BamHI and BglII, was ligated to BamHIdigested pLB49.36 Wild-type and mutant fragments were reamplified from the ligation products using OB primers as above, together with the primer P6,37 which binds just upstream of trbBp.

314 Protein roadblock constructions Pairs of oligonucleotides were annealed together to make short linkers containing protein binding-site sequences with BssHII sticky ends. The sequences of these oligonucleotides were as follows: for pLB110 and pLB111 (CRP binding site), CGCGAATCGGAAATG TGATCTAGATCACATTT-3 0 and 5 0 - CGCGAAATGT GATCTAGATCACATTTCCGA TT-3 0 ; for pLB12 and pLB13 (lac repressor binding site), 5 0 - CGCGTAATTGT GAGCGGATAACAATTGGCA TT-3 0 and 5 0 - CGCG AATGCCAATTGTTATCCGCTCACAAT TA-3 0 ; for pLB14 (no binding site) 5 0 - CGCGAATCGGTTAGT TGATCTAAGTCACTA TA-3 0 and 5 0 - CGCGTATAGT GACTTAGATCAACTAACCGATT-3 0 . Cloning of kfrA promoter fragments The kfrA promoter was modified initially by a single point mutation at position K60, between the KorB operator and the K35 hexamer, which introduced an XhoI recognition site. This site-directed mutagenesis was performed in an M13mp19 phagemid background using the Sculptor system (Amersham Corp.). Insertion mutations of 4 bp or 13 bp between OB and promoter were made by digestion with XhoI followed by Klenow fill-in with or without the addition of an oligonucleotide linker. Deletion mutations of 7 bp (pDM305), 8 bp (pDM306) and 17 bp (pDM307) between OB and promoter were introduced by XhoI digestion followed by Bal31 nuclease digestion. Mutant and wild-type variants of kfrAp were cloned as EcoRI fragments into the pSC101based promoter-probe plasmid pDM3 for use in reporter gene assays.17,38 xylE reporter gene assays The level of expression of the xylE reporter gene was determined by an enzymatic assay of activity of the gene product (catechol 2,3-oxygenase) in logarithmically growing bacterial cultures.70 One unit is defined as the amount of enzyme necessary to convert 1 mmol of substrate to product in one minute under standard conditions.70 Protein concentration was determined by the Biuret method.71 TrbA repressor protein was expressed from pMZT24,37 KorA from pGBT37, 72 and KorB from pDM1.2117 or pMMV811.16 The corresponding empty vector controls were pGBT3039 or pDM1.2.17 Repression index ratios were calculated as XylE activity in the absence of repressor/XylE activity in the presence of repressor. The cooperativity index (Ci) was calculated as the repression index (Ri) obtained with two repressors present divided by the product of the repression indices for single repressors. Standard deviations were estimated for repression indices by error propagation from the standard deviations of XylE activity measurements. Means and standard deviations were calculated from two or three replicates. Data shown are typical of at least two independent experiments.

Acknowledgements We thank Elton Stephens for providing technical assistance. L.E.H.B. and S.E.M. were supported by Wellcome Trust grants 046356/Z/95 and 067526/Z.

Repression and Cooperativity by IncP-1 KorB

D.P.M. was supported by an MRC studentship for training in research methods. DNA sequencing was performed by Alta Bioscience, using an automated sequencer funded partly by a shared equipment grant from the Wellcome Trust (038654/Z/93), and by the University of Birmingham Functional Genomics Laboratory, funded by BBSRC grant 6/JIF13209. Part of this work was carried out in the context of the EU Concerted Action MECBAD (Mobile Elements Contribution to Bacterial Adaptability and Diversity).

References 1. Bingle, L. E. H. & Thomas, C. M. (2001). Regulatory circuits for plasmid survival. Curr. Opin. Microbiol. 4, 194–200. 2. Pansegrau, W., Lanka, E., Barth, P. T., Figurski, D. H., Guiney, D. G., Haas, D. et al. (1994). Complete nucleotide-sequence of Birmingham IncP-alpha plasmids–compilation and comparative analysis. J. Mol. Biol. 239, 623–663. 3. Bignell, C. & Thomas, C. M. (2001). The bacterial ParA-ParB partitioning proteins. J. Biotechnol. 91, 1–34. 4. Surtees, J. A. & Funnell, B. E. (2003). Plasmid and chromosome traffic control: how ParA and ParB drive partition. In Current Topics in Developmental Biology, vol. 56, pp. 145–180, Academic Press, San Diego. 5. Rosche, T. M., Siddique, A., Larsen, M. H. & Figurski, D. H. (2000). Incompatibility protein IncC and global regulator KorB interact in active partition of promiscuous plasmid RK2. J. Bacteriol. 182, 6014–6026. 6. Williams, D. R., Macartney, D. P. & Thomas, C. M. (1998). The partitioning activity of the RK2 central control region requires only incC, korB and KorBbinding site OB3 but other KorB-binding sites form destabilizing complexes in the absence of OB3. Microbiology, 144, 3369–3378. 7. Jagura-Burdzy, G., Kostelidou, K., Pole, J., Khare, D., Jones, A. & Williams, D. R. (1999). IncC of broad-hostrange plasmid RK2 modulates KorB transcriptional repressor activity in vivo and operator binding in vitro. J. Bacteriol. 181, 2807–2815. 8. Zatyka, M., Bingle, L., Jones, A. C. & Thomas, C. M. (2001). Cooperativity between KorB and TrbA repressors of broad-host-range plasmid RK2. J. Bacteriol. 183, 1022–1031. 9. Kostelidou, K., Jones, A. C. & Thomas, C. M. (1999). Conserved C-terminal region of global repressor KorA of broad- host-range plasmid RK2 is required for co-operativity between KorA and a second RK2 global regulator, KorB. J. Mol. Biol. 289, 211–221. 10. Lukaszewicz, M., Kostelidou, K., Bartosik, A. A., Cooke, G. D., Thomas, C. M. & Jagura-Burdzy, G. (2002). Functional dissection of the ParB homologue (KorB) from IncP-1 plasmid RK2. Nucl. Acids Res. 30, 1046–1055. 11. Hayes, F. (2000). The partition system of multidrug resistance plasmid TP228 includes a novel protein that epitomizes an evolutionarily distinct subgroup of the ParA superfamily. Mol. Microbiol. 37, 528–541. 12. Gerdes, K., Moller-Jensen, J. & Jensen, R. B. (2000). Plasmid and chromosome partitioning: surprises from phylogeny. Mol. Microbiol. 37, 455–466. 13. Yamaichi, Y. & Niki, H. (2000). Active segregation by the Bacillus subtilis partitioning system in Escherichia coli. Proc. Natl Acad. Sci. USA, 97, 14656–14661.

315

Repression and Cooperativity by IncP-1 KorB

14. Kostelidou, K. & Thomas, C. M. (2000). The hierarchy of KorB binding at its 12 binding sites on the broadhost-range plasmid RK2 and modulation of this binding by IncC1 protein. J. Mol. Biol. 295, 411–422. 15. Thomson, V. J., Jovanovic, O. S., Pohlman, R. F., Chang, C. H. & Figurski, D. H. (1993). Structure, function, and regulation of the kilB locus of promiscuous plasmid RK2. J. Bacteriol. 175, 2423–2435. 16. Motallebi-Veshareh, M., Balzer, D., Lanka, E., JaguraBurdzy, G. & Thomas, C. M. (1992). Conjugative transfer functions of broad-host-range plasmid RK2 are coregulated with vegetative replication. Mol. Microbiol. 6, 907–920. 17. Macartney, D. P., Williams, D. R., Stafford, T. & Thomas, C. M. (1997). Divergence and conservation of the partitioning and global regulation functions in the central control region of the IncP plasmids RK2 and R751. Microbiology, 143, 2167–2177. 18. Jagura-Burdzy, G. & Thomas, C. M. (1997). Dissection of the switch between genes for replication and transfer of promiscuous plasmid RK2: Basis of the dominance of trfAp over trbAp and specificity for KorA in controlling the switch. J. Mol. Biol. 265, 507–518. 19. Jagura-Burdzy, G., Macartney, D. P., Zatyka, M., Cunliffe, L., Cooke, D., Huggins, C. et al. (1999). Repression at a distance by the global regulator KorB of promiscuous IncP plasmids. Mol. Microbiol. 32, 519–532. 20. Williams, D. R., Motallebi-Veshareh, M. & Thomas, C. M. (1993). Multifunctional repressor KorB can block transcription by preventing isomerization of RNA polymerase-promoter complexes. Nucl. Acids Res. 21, 1141–1148. 21. Adhya, S., Geanacopoulos, M., Lewis, D. E. A., Roy, S. & Aki, T. (1998). Transcription regulation by repressosome and by RNA polymerase contact. Cold Spring Harbor Symp. Quant. Biol. 63, 1–9. 22. Dodd, I. B., Shearwin, K. E., Perkins, A. J., Burr, T., Hochschild, A. & Egan, J. B. (2004). Cooperativity in long-range gene regulation by the lambda cI repressor. Genes Dev. 18, 344–354. 23. Ogata, K., Sato, K. & Tahirov, T. (2003). Eukaryotic transcriptional regulatory complexes: cooperativity from near and afar. Curr. Opin. Struct. Biol. 13, 40–48. 24. Rodionov, O., Lobocka, M. & Yarmolinsky, M. (1999). Silencing of genes flanking the P1 plasmid centromere. Science, 283, 546–549. 25. Yarmolinsky, M. (2000). Transcriptional silencing in bacteria. Curr. Opin. Microbiol. 3, 138–143. 26. Hanai, R., Liu, R. P., Benedetti, P., Caron, P. R., Lynch, A. S. & Wang, J. C. (1996). Molecular dissection of a protein SopB essential for Escherichia coli F plasmid partition. J. Biol. Chem. 271, 17469–17475. 27. Lynch, A. S. & Wang, J. C. (1995). SopB proteinmeditated silencing of genes linked to the sopC locus of Escherichia coli F plasmid. Proc. Natl Acad. Sci. USA, 92, 1896–1900. 28. Balzer, D., Ziegelin, G., Pansegrau, W., Kruft, V. & Lanka, E. (1992). KorB protein of promiscuous plasmid RP4 recognizes inverted sequence repetitions in regions essential for conjugative plasmid transfer. Nucl. Acids Res. 20, 1851–1858. 29. Funnell, B. E. & Gagnier, L. (1994). P1 plasmid partition-binding of P1 ParB protein and Escherichia coli integration host factor to altered parS sites. Biochimie, 76, 924–932. 30. Hendrickson, W., Petithory, J. & Schleif, R. (1983).

31.

32.

33.

34.

35.

36.

37.

38.

39.

40. 41.

42. 43.

44.

45.

Spacing mutations between the Escherichia coli pBAD RNA polymerase binding-site and the araC (I) induction site. Nucl. Acids Res. 11, 1873–1880. Dunn, T. M., Hahn, S., Ogden, S. & Schleif, R. F. (1984). An operator at K280 base pairs that is required for repression of araBAD operon promoter: addition of DNA helical turns between the operator and promoter cyclically hinders repression. Proc. Natl Acad. Sci. USA, 81, 5017–5020. Mu¨ller, J., Oehler, S. & Mu¨ller-Hill, B. (1996). Repression of lac promoter as a function of distance, phase and quality of an auxiliary lac operator. J. Mol. Biol. 257, 21–29. Meyer, R. J., Figurski, D. & Helinski, D. R. (1977). Properties of the plasmid RK2 as a cloning vehicle. In DNA Insertion Elements, Plasmids, and Episomes (Bukhari, A. I., Shapiro, J. A. & Adhya, S. L., eds), pp. 559–566, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Bignell, C. R., Haines, A. S., Khare, D. & Thomas, C. M. (1999). Effect of growth rate and incC mutation on symmetric plasmid distribution by the IncP-1 partitioning apparatus. Mol. Microbiol. 34, 205–216. Zatyka, M., Jagura-Burdzy, G. & Thomas, C. M. (1994). Regulation of transfer genes of promiscuous IncP-alpha plasmid RK2—repression of Tra1 region transcription both by relaxosome proteins and by the Tra2 regulator TrbA. Microbiology, 140, 2981–2990. Bingle, L. E. H., Zatyka, M., Manzoor, S. E. & Thomas, C. M. (2003). Co-operative interactions control conjugative transfer of broad host-range plasmid RK2: full effect of minor changes in TrbA operator depends on KorB. Mol. Microbiol. 49, 1095–1108. Zatyka, M., Jagura-Burdzy, G. & Thomas, C. M. (1997). Transcriptional and translational control of the genes for the mating pair formation apparatus of promiscuous IncP plasmids. J. Bacteriol. 179, 7201–7209. Macartney, D. P. (1996). Molecular and genetic analyses of the central control regions of the broad host range bacterial plasmids RK2 and R751, PhD thesis, University of Birmingham. Jagura-Burdzy, G., Ibbotson, J. P. & Thomas, C. M. (1991). The korF region of broad-host-range plasmid RK2 encodes 2 polypeptides with transcriptional repressor activity. J. Bacteriol. 173, 826–833. Kostelidou, K. & Thomas, C. M. (2002). DNA recognition by the KorA proteins of IncP-1 plasmids RK2 and R751. Biochim. Biophys. Acta, 1576, 110–118. Caramel, A. & Schnetz, K. (1998). Lac and lambda repressors relieve silencing of the Escherichia coli bgl promoter. Activation by alteration of a repressing nucleoprotein complex. J. Mol. Biol. 284, 875–883. Ptashne, M. & Gann, A. (2002). Genes & Signals, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Ziegelin, G., Pansegrau, W., Lurz, R. & Lanka, E. (1992). TraK protein of conjugative plasmid RP4 forms a specialized nucleoprotein complex with the transfer origin. J. Biol. Chem. 267, 17279–17286. Khare, D., Ziegelin, G., Lanka, E. & Heinemann, U. (2004). Sequence-specific DNA binding determined by contacts outside the helix-turn-helix motif of the ParB homolog KorB. Nature Struct. Mol. Biol. 11, 656–663. Delbru¨ck, H., Ziegelin, G., Lanka, E. & Heinemann, U. (2002). An Src homology 3-like domain is responsible

316

46.

47.

48. 49. 50. 51. 52. 53. 54.

55. 56.

57.

58. 59. 60.

Repression and Cooperativity by IncP-1 KorB

for dimerization of the repressor protein KorB encoded by the promiscuous IncP plasmid RP4. J. Biol. Chem. 277, 4191–4198. Leonard, T. A., Butler, P. J. G. & Lo¨we, J. (2004). Structural analysis of the chromosome segregation protein Spo0J from Thermus thermophilus. Mol. Microbiol. 53, 419–432. Dandanell, G. (1992). DeoR repression at-a-distance only weakly responds to changes in interoperator separation and DNA topology. Nucl. Acids Res. 20, 5407–5412. Lee, D. H. & Schleif, R. F. (1989). In vivo DNA loops in araCBAD: size limits and helical repeat. Proc. Natl Acad. Sci. USA, 86, 476–480. Mu¨ller, J., Barker, A., Oehler, S. & Mu¨ller-Hill, B. (1998). Dimeric lac repressors exhibit phasedependent co-operativity. J. Mol. Biol. 284, 851–857. Gaston, K., Bell, A., Kolb, A., Buc, H. & Busby, S. (1990). Stringent spacing requirements for transcription activation by CRP. Cell, 62, 733–743. Dethiollaz, S., Eichenberger, P. & Geiselmann, J. (1996). Influence of DNA geometry on transcriptional activation in Escherichia coli. EMBO J. 15, 5449–5458. Ebright, R. H. (1993). Transcription activation at Class I CAP-dependent promoters. Mol. Microbiol. 8, 797–802. Wing, H. J., Williams, S. M. & Busby, S. J. W. (1995). Spacing requirements for transcription activation by Escherichia coli FNR protein. J. Bacteriol. 177, 6704–6710. Ushida, C. & Aiba, H. (1990). Helical phase dependent action of CRP—effect of the distance between the CRP site and the K35 region on promoter activity. Nucl. Acids Res. 18, 6325–6330. Lanzer, M. & Bujard, H. (1988). Promoters largely determine the efficiency of repressor action. Proc. Natl Acad. Sci. USA, 85, 8973–8977. Thomas, C. M., Smith, C. A., Ibbotson, J. P., Johnston, L. & Wang, N. J. (1995). Evolution of the korA-oriV segment of promiscuous IncP plasmids. Microbiology, 141, 1201–1210. Thomas, C. M., Ibbotson, J. P., Wang, N. Y., Smith, C. A., Tipping, R. & Loader, N. M. (1988). Generegulation on broad host range plasmid RK2— identification of 3 novel operons whose transcription is repressed by both KorA and KorC. Nucl. Acids Res. 16, 5345–5359. Huang, J., Schlick, T. & Vologodskii, A. (2001). Dynamics of site juxtaposition in supercoiled DNA. Proc. Natl Acad. Sci. USA, 98, 968–973. Edgar, R., Chattoraj, D. K. & Yarmolinsky, M. (2001). Pairing of P1 plasmid partition sites by ParB. Mol. Microbiol. 42, 1363–1370. Ruden, D. M., Ma, J. & Ptashne, M. (1988). No strict

61. 62.

63.

64.

65.

66. 67. 68.

69. 70.

71. 72.

alignment is required between a transcriptional activator binding-site and the “TATA box“ of a yeast gene. Proc. Natl Acad. Sci. USA, 85, 4262–4266. Ross, E. D., Hardwidge, P. R. & Maher, L. J. (2001). HMG proteins and DNA flexibility in transcription activation. Mol. Cell. Biol. 21, 6598–6605. Grant, S. G. N., Jessee, J., Bloom, F. R. & Hanahan, D. (1990). Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc. Natl Acad. Sci. USA, 87, 4645–4649. Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains—nucleotide-sequences of the M13mp18 and pUC19 Vectors. Gene, 33, 103–119. McKenney, K., Shimatake, H., Court, D., Schmeissner, U., Brady, C. & Rosenberg, M. (1981). A system to study promoter and termination signals recognised by E. coli RNA polymerase (Chirikjian, J. C. & Papas, T. S., eds). In Gene Amplification and Analysis, vol. 2, pp. 383–415, Elsevier, Amsterdam. Kahn, M., Kolter, R., Thomas, C., Figurski, D., Meyer, R., Remaut, E. & Helinski, D. R. (1979). Plasmid cloning vehicles derived from plasmids ColE1, F, R6K and RK2. In Methods in Enzymology, (Wu, R., ed.) vol. 68, pp. 268–280, Academic Press, New York. Birnboim, H. C. & Doly, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucl. Acids Res. 7, 1513–1523. Sambrook, J. & Russell, D. W. (2001). Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Thorsted, P. B., Shah, D. S., Macartney, D., Kostelidou, K. & Thomas, C. M. (1996). Conservation of the genetic switch between replication and transfer genes of IncP plasmids but divergence of the replication functions which are major host-range determinants. Plasmid, 36, 95–111. Dorrell, N., Gyselman, V. G., Foynes, S., Li, S. R. & Wren, B. W. (1996). Improved efficiency of inverse PCR mutagenesis. Biotechniques, 21, 604. Zukowski, M. M., Gaffney, D. F., Speck, D., Kauffmann, M., Findelli, A., Wisecup, A. & Lecocq, J. P. (1983). Chromogenic identification of genetic regulatory signals in Bacillus subtilis based on expression of a cloned Pseudomonas gene. Proc. Natl Acad. Sci. USA, 80, 1101–1105. Gornall, A. G., Bardawill, C. J. & David, M. M. (1949). Determination of serum proteins by means of the Biuret reaction. J. Biol. Chem. 177, 751–766. Jagura-Burdzy, G., Khanim, F., Smith, C. A. & Thomas, C. M. (1992). Crosstalk between plasmid vegetative replication and conjugative transfer— repression of the trfA operon by TrbA of broad host range plasmid RK2. Nucl. Acids Res. 20, 3939–3944.

Edited by J. Karn (Received 7 December 2004; received in revised form 21 March 2005; accepted 22 March 2005)