| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Centre for Infectious Diseases, Royal (Dick) School of Veterinary Medicine, Chancellor's Building, University of Edinburgh, Edinburgh EH16 4SB, UK
2 Biomedical Research Group, Department of Biosciences, University of Kent, Kent CT2 7NJ, UK
3 Department of Molecular Biology, Umeå University, SE-901 87 Umeå, Sweden
4 Department of Genome Sciences, University of Washington, 1705 NE Pacific, Seattle, WA 98195, USA
Correspondence
David L. Gally
dgally{at}ed.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
integrase family, and invert a 314 bp DNA switch that controls the expression of type 1 fimbriae in Escherichia coli. FimB and FimE differ in their activity towards the fim switch, with FimB catalysing inversion in both directions in comparison to the higher-frequency but unidirectional on-to-off recombination catalysed by FimE. Previous work has demonstrated that FimB, but not FimE, recombination is completely inhibited in vitro and in vivo by a regulator, PapB, expressed from a distinct fimbrial locus. The aim of this work was to investigate differences between FimB and FimE activity by exploiting the differential inhibition demonstrated by PapB. The research focused on genetic changes to the fim switch that alter recombinase binding and its structural context. FimB and FimE still recombined a switch in which the majority of fimS DNA was replaced with a larger region of non-fim DNA. This demonstrated a minimal requirement for FimB and FimE recombination of the Fim binding sites and associated inverted repeats. With the original leucine-responsive regulatory protein (Lrp) and integration host factor (IHF)-dependent structure removed, PapB was now able to inhibit both recombinases. The relative affinities of FimB and FimE were determined for the four ‘half sites’. This analysis, along with the effect of extensive swaps and duplications of the half sites on recombination frequency, demonstrated that FimB recruitment and therefore subsequent activity was dependent on a single half site and its context, whereas FimE recombination was less stringent, being able to interact initially with two half sites with equally high affinity. While increasing FimB recombination frequencies failed to overcome PapB repression, mutations made in recombinase binding sites resulted in inhibition of FimE recombination by PapB. Overall, the data support a model in which the recombinases differ in loading order and co-operative interactions. PapB exploits this difference and FimE becomes susceptible when its normal loading is restricted or changed.
A supplementary table listing the strains, constructs, plasmids and primers used is available with the online version of this paper.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
-D-mannosides via the FimH adhesin (Krogfelt et al., 1990
). Type 1 fimbriae are considered to be important for the colonization of the human gastrointestinal tract (Adlerberth et al., 1995) and oropharynx (Donaldson et al., 1991; Guerina et al., 1983). They are a virulence factor for urinary tract infections (Connell et al., 1996; Iwahi et al., 1983; Schaeffer et al., 1987) and degranulation triggered by these fimbriae contributes to renal scarring (Mizunoe et al., 1997).
Phase-variable expression of type 1 fimbriae exhibited by single bacterial cells results from the activity of two site-specific recombinases that invert a 314 bp DNA element termed the fim switch (or fimS) (Abraham et al., 1985; Klemm, 1986) (see Fig. 1a). A promoter that drives expression of the fim genes required for export and structural assembly of type 1 fimbriae is located within fimS (Klemm et al., 1985). The fim operon is only expressed when the fim switch is inverted to the on orientation, allowing transcription of the fim genes. FimB and FimE, required to invert fimS, are members of the integrase family of site-specific recombinases (Eisenstein et al., 1987) (for a review see Hallet & Sherratt, 1997). Recombination of fimS is distinct from the related Xer-mediated recombination in that the recombinases act independently to invert fimS (Blakely et al., 1993; Blomfield et al., 1991b; Gally et al., 1996; Klemm, 1986; McClain et al., 1991). Each inverted repeat (IR) is flanked by overlapping FimB and FimE binding sites, and following occupancy of these sites they recombine the switch within the IR sequence. As for phage chromosomal integration and excision, fim recombination also requires accessory proteins, specifically integration host factor (IHF) and the leucine-responsive regulatory protein (Lrp) (Blomfield et al., 1993, 1997). These proteins are believed to contribute to the overall architecture of the fim switch that facilitates synapse of the 9 bp IRs. The genetic organization and associated protein binding sites of fimS are represented in Fig. 1(a).
|
). However, the level of these amino acids could also act as a signal of inflammation, as aliphatic amino acids are particularly sensitive to oxidation by reactive oxygen species (Stadtman & Levine, 2003). Recently, it has been proposed that Lrp can act as a recombination directionality factor (RDF) that favours recombination in the off-to-on direction under certain conditions (Kelly et al., 2006).
However, it is the activity of the FimB and FimE recombinases that predominantly determines the orientation of the fim switch. While FimB and FimE share 52 % amino acid identity and bind to overlapping sites, both their activity and their directionality are different (Gally et al., 1996). FimB catalyses inversion in both directions, although with a slight bias for the off-to-on orientation, while FimE predominantly catalyses on-to-off inversion. Control of FimE expression is important in bringing about its orientation bias (Kulasekara & Blomfield, 1999); as the fim switch is located at the end of fimE, the orientation of fimS determines the length and 3' sequence of the fimE transcript (Hinde et al., 2005; Sohanpal et al., 2001). As a consequence, fimE mRNA is likely to be subject to more rapid 3' to 5' degradation when the switch is in the off orientation than when it is in the on orientation. In addition, FimE preferentially binds to fimS in the on orientation, as has been demonstrated in vitro and in vivo (Kulasekara & Blomfield, 1999), which adds to the directional bias. A further difference between FimB and FimE is that FimB inversion frequencies are markedly lower than those exhibited by FimE, both in vitro and in vivo (Blomfield et al., 1991b; Gally et al., 1993).
Previous work has demonstrated that when pyelonephritis-associated pili (P-pili) are switched on, PapB prevents expression of type 1 fimbriae in the same bacterial cell by inhibition of FimB activity (Holden et al., 2006; Xia et al., 2000). In contrast, FimE recombinational activity, and hence rapid on-to-off phase switching, is not inhibited by PapB in vivo or in vitro. In fact, FimE expression is increased at the transcriptional level, raising the frequency with which fimS in the on orientation is turned off (Xia et al., 2000). PapB regulates P-pili expression by binding to specific sites on pap DNA (Blyn et al., 1989; Forsman et al., 1989; Xia et al., 1998). PapB has also been shown to bind to fim DNA (Fig. 1a
) (Holden et al., 2001; Xia et al., 2000), which raises the possibility that inhibition may be the result of occlusion of FimB, but not FimE, from fimS. Inhibition of FimB-promoted recombination by PapB is not directional and occurs in vitro, and hence must arise from a direct effect on FimB recombination. The specificity shown by PapB suggests that the mechanism of inhibition is due to differences in how the recombinases bind to and complex with fimS, and not the recombination event per se.
In the current study we aimed to understand the basis of the differences between FimB- and FimE-mediated recombination by making use of the differential inhibition exhibited by PapB. The research made use of deletions, duplications, swaps and replacements within and adjacent to the fim switch (fimS) and also measured the relative affinities of FimB and FimE for the four ‘half sites’ that flank the 9 bp IRs.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Indicator media were minimal MOPS plates supplemented with X-Gal (40 µg ml–1; Melford Laboratories) and the appropriate antibiotic, as described in Gally et al. (1993). Protein expression was induced with arabinose or IPTG at the concentrations indicated. Antibiotics used were ampicillin (Amp; 50 µg ml–1) and tetracyline (Tet; 25 µg ml–1).
Cloning techniques and the generation of different fimS alleles
DNA was purified with Promega or Hybaid DNA purification mini columns. Restriction enzymes were purchased from either New England Biolabs or Boehringer Mannheim. All the molecular biology techniques were carried out according to standard protocols (Sambrook et al., 1989). All the primers, plasmids and bacterial strains used or generated are listed in Supplementary Table S1 available with the online version of this paper.
Individual half site cloning.
The four different half sites were amplified by PCR, with different primer sets depending on the half site to be cloned. The primers used were 2893 and 3139 [right inverted repeat (IRR) outside], 2774 and 3004 (IRR inside), 2476 and 2693 [left inverted repeat (IRL) outside] and 2632 and 2812 (IRL inside). The PCR products were digested with HaeIII, and treated with the Klenow fragment of DNA polymerase I, and each half site was cloned into SmaI-digested pUC21. Successful clones were selected using E. coli XL1. Clones were checked by sequencing and the ones used in the study all had the half sites oriented in the same direction within the vector. This was to help ensure that surrounding vector DNA had no impact on the comparison of binding affinities. Four plasmids were created, each containing different half sites inserted into the multi-cloning site of pUC21 (pJL043, IRR outside; pJL044, IRL outside; pJL046, IRL inside; pJL077, IRR inside).
Cloning half site changes.
ZAP1164: wild-type fimS in fimB– fimE– background. Plasmid pDG019 was exchanged into BGEC150 to generate ZAP1164.
ZAP1169: duplication of IRL-out to IRR-out (designed to strengthen FimB and weaken FimE affinities and to disrupt a triple A sequence). PCR was used to generate a downstream fragment with the F107 and F90 primers and an upstream fragment with the F108 and F90 primers. The fragments were digested with BglII, and annealed and amplified with the F90 and F91 primers, and the complete fragment was cloned into the SphI and BamHI sites in pDG019 to generate pNJH66, via pCR4-TOPO T/A. Strain ZAP1169 was then generated by allelic exchange of pNJH66 into BGEC144.
ZAP1170 and ZAP1171: duplication of IRR-in (on) to IRR-out (designed to strengthen FimB affinity marginally and to weaken FimE affinity as well as to disrupt triple A sequence). PCR was used to generate a downstream fragment with primers F107 and F90, and an upstream fragment with primers F109 and F90. The fragments were digested with BglII, and annealed and amplified with primers F90 and F91, and the complete fragment was cloned into the SphI and BamHI sites in pDG019 to generate pNJH67, via pCR4-TOPO T/A. Strain ZAP1170 was then generated by allelic exchange of pNJH67 into BGEC144. Strain ZAP1171 was generated by allelic exchange of pNJH67 into BGEC132.
ZAP1172 and ZAP1173: deletion of 185 bp downstream of IRR-out. pNJH67 was digested with BglII and BamHI, the intervening 185 bp were removed, and the plasmid religated to generate pNJH92. Strain ZAP1172 was generated by allelic exchange of pNJH92 into BGEC144. Strain ZAP1173 was generated by allelic exchange of pNJH92 into BGEC132.
ZAP1176: designed to alter IRR-out to more closely resemble IRL-in (on), also to delete flanking DNA upstream of IRL-out and IRR-out. PCR was used to generate the fragment containing the alteration at IRR-out, using primers F135 and F142, and this was cloned into the SphI and BamHI sites of pDG019 via pCR4-TOPO T/A. Strain ZAP1176 was generated by allelic exchange of pNJH137 into BGEC144.
ZAP1177: designed to alter IRR-out to more closely resemble IRL-in (on), with swapped IRs and deletion of flanking DNA upstream of IRL-out and IRR-out. PCR was used to generate the fragment containing the alteration at IRR-out, using primers F143 and F144, and this was cloned into the SphI and BamHI sites of pDG019 via pCR4-TOPO T/A. Strain ZAP1177 was generated by allelic exchange of pNJH138 into BGEC144.
ZAP1178: designed to combine duplication of IRR-in (on) to IRL-in (on) and IRR-out with deletion of flanking DNA upstream of IRL-out and IRR-out. PCR was used to generate the fragment containing the alteration at IRL-in and IRR-out, using primers F145 and F141, and this was cloned into the SphI and BamHI sites of pDG019 via pCR4-TOPO T/A. Strain ZAP1178 was generated by allelic exchange of pNJH139 into BGEC144.
ZAP1179: designed as ZAP1178, but with flanking region upstream of IRL-out restored by PCR to maintain functional fimE. PCR was used to generate the fragment to repair fimE, using primers F150 and F90, and this was cloned into the SphI site of pNJH139 to produce pNJH140. Strain ZAP1179 was generated by allelic exchange of pNJH140 into BGEC144.
ZAP1174 and ZAP1175: IRL-out swapped with IRR-out. pHK156 was digested with SacI to remove the pUC19-derived origin of replication, creating pNJH91. Strain ZAP1174 was generated by allelic exchange of pNJH91 into BGEC144. Strain ZAP1175 was generated by allelic exchange of pNJH91 into BGEC132.
ZAP1180 and ZAP1181: complete replacement of internal fimS with part of cdsH also containing the Ptac promoter to drive lacZYA expression. PCR was used to generate the replacement switch of cdsH' flanked by fimS IRL and IRR, using primers F293 and F240; the resulting fragment was cloned into the SphI and BamHI sites of pDG019, via pCR4-TOPO T/A. The Ptac promoter was then inserted by PCR amplification using primers K12 and K14, digested with BamHI and XmaI, end-filled with Klenow, and cloned into the XmaI site in pNJH127 (located within cdsH'), generating pNJH130. Strain ZAP1180 was generated by allelic exchange from pNJH130 into BGEC144, and strain ZAP1181 was generated by allelic exchange from pNJH130 into ZAP1164 (see Supplementary Table S1).
ZAP1182: contains wild-type fimS in a fimE– lrp– background. The lrp mutation was introduced by P1 transduction from strain BGEC429A into strain AAEC370A to generate ZAP1182.
ZAP1183: complete replacement of internal fimS with part of cdsH and containing the Ptac promoter to drive lacZYA expression, in a fimE– lrp– background. The lrp mutation was introduced by P1 transduction from strain BGEC429A into strain ZAP1180 to generate ZAP1183.
Cloning fimB, fimE and papB.
pNJH55 was generated by cloning fimB, amplified by PCR using primers F96 and F97, into the EcoRI and KpnI sites in pBAD18. pNJH56 was generated by cloning fimE, amplified by PCR using primers F100 and F101, into the EcoRI and KpnI sites in pBAD18. pNJH20, containing papB in pBAD18, was published previously (Holden et al., 2001).
Measurement of recombination frequencies on agar media.
Inversion frequencies of the fim switch were determined using fimA–lacZYA strains cultured on MOPS minimal agar plates, supplemented with ampicillin (50 µg ml–1) and/or tetracycline (25 µg ml–1) and X-Gal (40 µg ml–1). PapB was expressed from pNJH20, induced with arabinose (0.2 %, w/v), or expressed from pHMG88, induced with IPTG (1 mM). FimB, expressed from pNJH55, and FimE, expressed from pNJH56, were induced with arabinose (0.0004 %, w/v). To assess the effect of PapB on FimB-promoted recombination, pNJH20 and pBAD18 (control) were transformed into the following strains: AAEC370A, BGEC576, BGEC605, ZAP973, ZAP1169, ZAP1170, ZAP1172, ZAP1174, ZAP1176, ZAP1177, ZAP1178, ZAP1180 and ZAP1183. To assess the effect of PapB on FimE-promoted recombination, pNJH20 and pBAD18 (control) were transformed into the strains AAEC198A, ZAP1171 and ZAP1173. To assess the inhibitory effect of multi-copy FimB or FimE on the inhibitory role of PapB, pNJH55 or pNJH56 were co-transformed with pHMG88 into ZAP1180 and ZAP1183 (the relevant controls were pBAD18 and pACYC184). The assays were carried out as described in Gally et al. (1993) and Xia et al. (2000). Briefly, recombination frequencies were determined in strains that contained fimA–lacZYA, where blue colonies represented fimS in the on phase and white colonies, fimS in the off phase. Blue (for on-to-off recombination) or white (for off-to-on recombination) colonies were picked from an agar plate containing X-Gal. The colony was suspended in PBS, diluted and replated onto a similar indicator plate, and the resulting proportion of blue and white colonies that had arisen from the original colony, as well as the age of the original colony (in number of generations), were determined. This allowed measurement of the frequency of recombination from one phase to another on a per cell, per generation basis, also described in Gally et al. (1993). Each frequency is derived from at least two sets of independent experiments and each set from approximately five different colonies. The mean and SEM for each frequency were determined. Comparison of different recombination frequencies was carried out using a two-sample t test, and differences were considered significant at the 95 % confidence level.
Measurement of fimS on/off proportions.
The orientation of non-fim switches was determined by PCR amplification with primer combinations of F246 and F128 for on-specific products and F90 and F246 for off-specific products. The orientation of wild-type fimS was determined by PCR amplification with primer combinations of F85 and F106 for on-specific products and F90 and F85 for off-specific products. The products were resolved on a 1.5 % agarose gel. For proportional determination of non-fim switch populations, DNA was amplified by PCR using primers F90 and F128, and then subjected to digestion with AatII. For wild-type fimS, DNA was amplified by PCR using primers F1 and F2, and then subjected to digestion with HinfI. The products were then resolved on a 4 % polyacrylamide gel.
Preparation of proteins for protein–DNA interactions.
Crude cell extracts containing FimB or FimE were generated from E. coli NEC11 transformed with pIB378 and pIB382, respectively, after induction with 0.5 mM IPTG for 2.5 h during mid-exponential growth in rich defined MOPS media, and subsequent sonication. A control extract was obtained in the same way from E. coli NEC11 transformed with pET11. Sonicates were cleared by centrifugation at 10 000 g for 30 min at 4 °C. The cleared lysates were stored in 50 % (v/v) glycerol at –20 °C.
Electrophoretic mobility shift assay (EMSA).
EMSA analysis of FimB and FimE on the fimS half sites was carried out as described in Gally et al. (1996) and Holden et al. (2001). DNA incorporating each recombinase half site from pJL043, pJL044, pJL046 and pJL077 was amplified by PCR using primers Puc1 and Puc2. FimB and FimE extracts generated as above were serially diluted (fivefold) in sonication buffer and used in gel retardation assays, as described previously (Gally et al., 1996). The proportion of free DNA was measured using a phosphoimager (Molecular Dynamics) and the relative concentrations of FimB or FimE extract required to shift 50 % of the DNA were calculated. To demonstrate the relative affinities using these figures, an arbitrary value of 1 was assigned to the lowest-affinity half site and the other sites were scaled to this. In the case of FimB, even at the highest concentration used it failed to shift 50 % of the DNA for its lowest-affinity site, and therefore the neat extract concentration was taken as 1 for calculating the relative affinities of the remaining sites and scored as <1 for this half site. The recombinase affinities shown are therefore only comparable for each recombinase and cannot be used to compare FimB and FimE.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Blomfield et al., 1997). With such structural constraints removed, would this alter recombinase susceptibility to PapB? A minimal-switch-region unit has previously been constructed in the chromosome at fim by removal of sequences either side of the recombinase external binding sites that flank the IR units; 227 bp upstream of IRL and 200 bp downstream of IRR (strain ZAP973; Fig. 1b). This removes IHF site I (Blomfield et al., 1997). Fusion of fimA to lacZYA allowed FimB-promoted recombination to be determined from the proportion of blue and white colonies, as described in Methods. These changes alone only had a minor effect in the off-to-on direction: 2.63x10–3 (±0.15x10–3) per cell per generation compared to 2.34x10–3 (±0.18x10–3) for wild-type fimS, and a reduction of 2.7-fold in the on-to-off direction from 1.90x10–3 (±0.46x10–3) for wild-type fimS to 0.73x10–3 (±0.19x10–3) for the minimal switch. Recombination of this construct in both orientations was still completely inhibited in the presence of PapB. The FimE rate was not measured for this construct as the left-hand flanking region deletion leads to a fimE truncation.
The fim switch region (270 bp) between the internal recombinase half sites was then replaced within the minimal switch construct, described above, with a larger fragment of non-fim DNA (1167 bp phage DNA) (Fig. 1c). The novel sequence was screened and did not contain consensus sites for IHF or Lrp binding. While we cannot rule out the possibility that either protein has an interaction with this element, the context is now entirely different and any interaction is unlikely to promote alignment of the IRs. The on and off orientation of this novel switch was defined accordingly to the position of the recombinase half sites in wild-type fimS. FimB-promoted recombination of this replacement switch in the off orientation occurred and could be detected by orientation-specific PCR amplification from individual colonies (Fig. 2a). The lack of a requirement for Lrp for recombination was confirmed by transducing an lrp mutation into a strain background containing the replacement switch (strain ZAP1183; Supplementary Table S1). Recombination in the off-to-on direction was still detected in the lrp– strain by orientation-specific PCR. In the corresponding wild-type fimS strain containing the lrp mutation (ZAP1182), the fim switch starts in the on orientation and no on-to-off recombination was detected (Fig. 2a). This indicated that lrp was required for FimB-mediated recombination of wild-type fimS but not of the replacement switch. Despite the lack of putative PapB binding sites in the replacement switch, PapB still inhibited FimB-promoted recombination. FimE was supplied in trans (as the chromosomal copy is truncated) and shown to recombine the replacement switch, demonstrated by PCR amplification of the whole switch and subsequent restriction enzyme digestion to generate proportional amounts of on- and off-specific products (Fig. 2b). In addition, the recombination frequency for the replacement switch from on phase to off phase was determined to be 3.31x10–4 (±0.98x10–4) per cell per generation. However, no FimE-promoted recombination of the replacement switch was detected in the presence of PapB, in either the recombination assay or the PCR assay (Fig. 2b). Taken together, these data demonstrate a minimal requirement of just the IRs and the associated recombinase binding sites for recombination by both FimB and FimE. Of note was that FimB recombination of this novel construct was blocked by PapB, indicating that the inhibitory activity is unlikely to be dependent on accessory protein interactions with the fim switch. Also, FimE-promoted recombination was shown to be inhibited by PapB, in contrast to its previously reported effect on the wild-type fimS (Xia et al., 2000), and this is shown in Fig. 5 for strain AAEC198A.
|
|
). To investigate the importance of sequence differences in these sites, relative DNA binding affinities for FimB and FimE were determined. For this, the four half sites were amplified by PCR and cloned into pUC21, as described in Methods. Bacterial extracts containing either FimB or FimE were prepared as described previously, using fimE and fimB clones in the expression vector pET11 (Gally et al., 1996). EMSA experiments were carried out with serial dilutions of each recombinase preparation. It must be noted that binding affinities would ideally be determined using purified recombinase proteins but, despite considerable efforts from a number of groups, functional proteins have not been purified, so the studies described here are restricted to relative affinities. Loss of free DNA was quantified and the relative binding affinities are shown in Fig. 3(a, b); the nomenclature used for the ‘half sites’ is that from the on orientation. No discernible protein–DNA complexes were seen using control extracts without FimB or FimE (data not shown). The experiments were repeated with different FimB and FimE preparations and gave similar relative affinity values. The binding sites and relative binding affinities are given in Table 1. It is clear from the binding data that FimE binds with higher affinity to two of the half sites, the IRL internal (on orientation) and IRR external sites. It bound with a lower affinity to the IRL external site, and binding to the internal site at IRR (on-orientation) was only just detectable. In comparison, FimB bound with highest affinity to only one half site, the IRL internal site (on-orientation): eightfold more FimB protein was needed to give 50 % occupancy at the external IRL site, and binding was very weak at both IRR sites (on-orientation). From these results, the key bases required in FimE and FimB recombinase binding sequences have been predicted, but remain to be proven experimentally (Table 1).
|
|
, the recombinase half sites were varied to examine their role in recombinase activity as well as PapB inhibition, by construction of half site duplications, swaps and alterations in an otherwise wild-type fim switch, followed by chromosomal integration into the fim locus in E. coli K-12 (shown diagrammatically in Figs 4 and 5 and listed in Supplementary Table S1). PapB was added to each strain by transformation with an inducible plasmid. We understand that the levels of PapB expression are physiologically relevant, since we have previously demonstrated that single-copy papB driven from its native promoter is also sufficient to prevent fimS recombination (Holden et al., 2006).
|
). The frequency of recombination for the fim switch containing the duplicated internal half site is therefore at least 20-fold higher than that of wild-type fimS (2x10–3 per cell, per generation). Therefore, the lower-affinity IRR internal (on) site must limit FimB recombination. Despite the high level of FimB-promoted recombination in this strain, addition of PapB, induced from a plasmid expression vector (pNJH20), resulted in colonies locked in either the on or the off phase, as determined by a colony-switching assay and orientation-specific PCR amplification reactions (data not shown).
A critical role for the in situ location of the highest-affinity FimB binding site (IRL internal, on orientation) was demonstrated with a construct in which the internal half sites were swapped over (strain BGEC605). The prediction was that the half site swap would have little impact on the FimB-promoted recombination frequency, although it might alter the FimB bias to the on-to-off direction, compared to a bias for the off-to-on direction demonstrated on the wild-type fim switch sequence (compare data for AAEC370A in Fig. 4a, c). However, FimB recombination was not detected in this construct. Sequence analysis (as with all constructs) confirmed that the modified fim switch region was as anticipated. Furthermore, the chromosomal copy of FimB was shown to be functional, and overexpression of FimE but not FimB inverted the altered switch in BGEC605 (data not shown). Therefore, the context of the highest-affinity FimB half site at IRL in (on orientation) is essential for FimB, but not for FimE, recombination.
While the two constructs that affect the internal sites had significant effects on FimB recombination, alterations to the outside half sites had little effect on FimB recombination frequency in the off-to-on orientation (Fig. 4a). For example, swapping of the outside half sites (strain ZAP1174) had no significant effect on the frequency of FimB recombination. Alterations that are predicted to marginally (strain ZAP1170) or substantially (strain ZAP1169) increase the affinity of FimB binding to the outside IRR half site also had little impact on FimB recombination or inhibition by PapB.
The external half site at IRR shows the weakest FimB binding, whereas FimE shows the highest affinity here (Fig. 3). To investigate whether the differential binding to IRR is fundamental to the specificity exhibited by PapB, a series of alterations were made to fimS in an attempt first to increase the binding affinity of FimB for IRR-out, and second to reduce any potential binding to IRR-external by PapB. Previous work has shown that PapB binds preferentially to sequences containing triple adenine motifs (Xia et al., 1998), so the alterations disrupted a 5' AAA sequence present at IRR-external. Although two of the changes (strain ZAP1172 and ZAP1176) increased FimB-promoted off-to-on recombination significantly, the level of increase was small (Fig. 4b). As before, PapB activity completely inhibited FimB-promoted recombination. We also investigated FimB activity in the on-to-off orientation, as well as the associated effect of PapB, on a selected number of fimS mutants. One of the alterations (strain ZAP1170) increased FimB-promoted recombination 12-fold (Fig. 4c), whereas the others had little or no effect. Furthermore, FimB activity was still inhibited by PapB in every case.
In summary, these data demonstrate that the two internal half sites are critical for FimB recombination. IRR-internal (on) limits FimB recombination frequencies, whereas the context and presence of IRL-internal (on) is essential for recombination. In contrast, alterations to the external half sites have little effect on FimB recombination, and the sequences external to fimS are not key recombination determinants.
Impact of half site changes on FimE-promoted site-specific recombination and PapB activity
Based on the relative half site affinities, it was anticipated that FimE activity would be reduced following alteration of the sites with highest relative affinity for this recombinase [IRR-external and IRL-internal (on)]. Swapping the external half sites (ZAP1175) resulted in a 4.5-fold reduction in FimE recombination frequency in the on-to-off orientation, from 3.22x10–2 (±0.49x10–2) for AAEC198A to 0.72x10–2 (±0.11x10–2) for ZAP1175. Replacing the highest-affinity site (IRR external) with a duplication of the lowest-affinity site [IRL-internal (on), strain ZAP1171] resulted in a non-significant 1.6-fold reduction in recombination frequency (Fig. 5). Deletion of DNA downstream from IRR-external in combination with duplication of IRL-in (on) at IRR-out (ZAP1173) resulted in a 12-fold reduction in FimE activity (P<0.001). FimE recombination is normally increased in the presence of PapB, as PapB upregulates fimE expression (AAEC198A) (Xia et al., 2000). However, FimE recombination does not increase in the presence of PapB in either of the related backgrounds (ZAP1173 and ZAP1171), demonstrating that FimE activity can be inhibited, at least to some extent, by PapB when half sites are altered. There remains the possibility that alterations of the half sites have an impact on fimE expression. The role of the half sites in the inhibitory activity of PapB was further investigated by using FimE recombination of a previously published fimS allele that contains a mutation in the conserved CA nucleotides of the IRR external half site (strain BGEC424). This mutation reduced the FimE switching frequency in rich defined medium by 3.7-fold (compared to previously published data; Gally et al., 1996) to 5.23x10–3 per cell, per generation, and this frequency was reduced 30-fold to 1.8x10–4 by the presence of PapB. A large degree of variation was seen for recombination of strain BGEC424 (Fig. 5). However, addition of PapB not only reduced the recombination frequency but also reduced the level of variation between single colonies. Since the CA mutation at IRR external has been shown to reduce FimE binding by approximately 40-fold (Gally et al., 1996), the data indicate that the inhibitory activity of PapB is in part reliant on recombinase binding activity.
In summary, differences in the binding affinities of the recombinases for the half sites are largely responsible for the different outcomes in fimS recombination promoted by FimB or FimE. This is reflected in the specificity of PapB inhibition of FimB on wild-type fimS. However, PapB inhibited FimE when the context of fimS was altered by either substantially reducing the affinity of FimE for the IRR outside half site or removing the switch region between the IRs and therefore the organized structure imposed by Lrp and IHF binding.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
integrase, or tyrosine-dependent site-specific recombinase family (for reviews see Chen & Rice, 2003; Hallet & Sherratt, 1997). The Fim recombinases are less-well-characterized members of this family, and much of the current body of knowledge about them stems from detailed work on analogous systems, such as XerC/D, Int, P1 phage protein Cre and yeast protein Flp. Interest in these specific recombinases comes from the fact that although they recombine the same substrate they differ in: (i) their recombinational frequencies; (ii) the preferred orientation of their substrate; (iii) their capacity to be inhibited by the PapB regulator. The main aim of this study was to investigate these differences. To do this, each step of the process required for recombination was considered in terms of the above differences.
By analogy with related systems and our preliminary work, the step-wise nature of the inversion of fimS by FimB and FimE can be described. Firstly, the Fim recombinases would need to bind to their cognate sites that flank and include the IRs. Each half site is presumably occupied by a monomer. Co-operativity may then aid occupancy of each full site. In the case of the fim switch, Lrp and IHF bind within the invertible element to establish a structure competent for recombination, presumably by bringing the IRs into juxtaposition. Co-operative protein–protein interactions may be required to complete the nucleoprotein complex, for example between recombinase dimers. Single-strand cleavage will then occur, and recent work has demonstrated that this may further stabilize the complex (Mumm et al., 2006). Holliday-junction formation and subsequent resolution presumably then occurs by a well-characterized mechanism that is shared by all members of the tyrosine recombinase family.
Given the shared mechanism of recombination, the current research focused on the preceding step of recombinase binding. This was carried out by mutating sequences within and adjacent to the fim switch, predominantly at protein binding sites, and the impact was assessed by the effect on recombination frequency. In particular, attention was centred on the specificity exhibited by PapB for inhibition of FimB alone, as this had to reflect a fundamental difference between the two recombinases.
Experiments were carried out to define the minimal requirements for both recombination and PapB inhibition. The research demonstrated that both FimB and FimE were able to act on a novel substrate in which most of the switch was replaced with non-fim DNA (derived from phage), leaving only fim recombinase binding sites and the two IRs. Therefore, both recombinases can act in the absence of accessory protein, in particular Lrp and IHF, binding. Recombination by site-specific recombinases varies in terms of the complexity of the target sequence and the superhelicity of the substrate. Cre and Flp recognize relatively simple sites and do not require accessory proteins for organization of the substrate. In contrast, higher-order nucleoprotein complex formation is required for alternative systems, dependent on accessory proteins that control the architecture of the complex necessary for recombination (Hallet & Sherratt, 1997; Sadowski, 1986). For example, severe DNA bending is induced by IHF, Xis and factor for inversion stimulation (FIS) to facilitate Int-dependent recombination (Kim & Landy, 1992), and a similar role has been demonstrated for ArgR and PepA for Xer-dependent recombination (Colloms et al., 1997; Sherratt et al., 1995). Recombination of fimS requires Lrp and IHF to induce sharp bends in the DNA (Blomfield et al., 1993, 1997). To our knowledge, the current study is the first to demonstrate recombination of a DNA sequence from which the structural constraints have been removed by completely substituting the native fim DNA. An indication that this would occur comes from earlier research by Dove & Dorman (1996), who have shown that overexpression of FimB promotes fimS recombination, even in the absence of Lrp or IHF protein binding to fim DNA. Interestingly, after the removal of Lrp- and IHF-dependent intrinsic structure, PapB was able to inhibit both FimB and FimE recombination. Furthermore, this construct no longer contained either of the putative PapB binding sites that we have mapped in the fimS region (Fig. 1a), indicating that these also are not required for PapB to inhibit recombination of this novel substrate. We do not know whether the replacement switch DNA contained additional PapB binding sites that may have had an effect. Previous work has demonstrated that PapB can inhibit FimB-, but not FimE-promoted in vitro recombination (Xia et al., 2000). Furthermore, there is no evidence that PapB binds directly to any of the four half sites (including IR regions) or directly to FimB or FimE (our unpublished data). The main difference between wild-type fimS and the replacement switch would be a lack of structure to facilitate recombinase interactions that lead to synapse formation and recombination. The fact that PapB inhibits FimE recombination under these conditions supports the concept that PapB can act on these later co-operative interactions, i.e. synapse formation, rather than initial recombinase binding.
An alternative way to change the sequence or nature of the co-operative loading is to alter the affinities of the recombinase half sites at fim. It was hypothesized that weakening FimE binding at its highest-affinity site would alter the order of interaction and allow inhibition by PapB. The data support this prediction, but are complicated by the demonstration that FimE has two half sites with similarly high affinities, so there may be some flexibility in the order in which these interactions can occur. In contrast, it was demonstrated that FimB recombination is dependent on a single high-affinity half site. This research has shown that even the position of this site is essential for FimB recombination. Internal duplication of this site did increase recombination significantly, but the reaction was still completely inhibited by PapB. The fact that swapping the internal half sites prevents FimB recombination is indicative of a single pathway leading to synapse formation, and it is likely that this is effectively blocked by PapB. On the other hand, it appears that FimE uses an alternative loading pathway that circumvents PapB inhibition. Only when this pathway is restricted by either structural changes or alteration of binding affinities does FimE recombination become susceptible to PapB inhibition.
The current study shows differences in the Fim recombinase affinities for each half site (Fig. 3), and combining these produces an order of affinities comparable with those determined previously for each whole site/IR (Gally et al., 1996). One clear difference is the magnitude of the differences in binding affinities shown for the half sites in comparison to the whole IRs. This can be explained by the co-operativity of recombinase binding between the half sites. Certain half sites are almost unable to bind recombinase alone and require a stronger half site partner to give occupancy of the whole site. As the previous analysis of binding to the whole sites looked at only the loss of free DNA and did not take into account whether both half sites were occupied, it is likely that the affinity differences described here for the half sites are fundamental in determining the specificity of recombinase activity. From this study, the current working model is that the formation of a recombination-proficient complex requires initial occupancy of certain half sites by the recombinases, and that this binding leads to co-operative binding to the adjacent site. Subsequent interactions that support binding to relatively weak sites occur as a consequence of the IRs being brought into juxtaposition. This allows co-operation across the two IRs. The different affinities of FimB and FimE for the half sites mean that these secondary interactions will also be different. PapB is likely to act at this stage, as current research strongly suggests that it does not prevent initial recombinase binding. To prove this, future work will focus on measuring biophysical interactions between fimS complexes involving either FimB or FimE, as well as the impact of PapB. Understanding the mechanism of regulatory cross-talk between these two phase-variable virulence factors will provide insights into how pathogens co-ordinate the expression of cell-surface factors during infection.
|
ACKNOWLEDGEMENTS |
|---|
Edited by: L. Jannière
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adlerberth, I., Hanson, L. A., Svanborg, C., Svennerholm, A. M., Nordgren, S. & Wold, A. E. (1995). Adhesins of Escherichia coli associated with extra-intestinal pathogenicity confer binding to colonic epithelial cells. Microb Pathog 18, 373–385.[CrossRef][Medline]
Blakely, G., May, G., McCulloch, R., Arciszewska, L. K., Burke, M., Lovett, S. T. & Sherratt, D. J. (1993). Two related recombinases are required for site-specific recombination at dif and cer in E. coli K12. Cell 75, 351–361.[CrossRef][Medline]
Blomfield, I. C., McClain, M. S. & Eisenstein, B. I. (1991a). Type 1 fimbriae mutants of Escherichia coli K12 – characterization of recognized afimbriate strains and construction of new fim deletion mutants. Mol Microbiol 5, 1439–1445.[Medline]
Blomfield, I. C., McClain, M. S., Princ, J. A., Calie, P. J. & Eisenstein, B. I. (1991b). Type 1 fimbriation and fimE mutants of Escherichia coli K-12. J Bacteriol 173, 5298–5307.
Blomfield, I. C., Vaughn, V., Rest, R. F. & Eisenstein, B. I. (1991c). Allelic exchange in Escherichia coli using the Bacillus subtilis sacB gene and a temperature-sensitive pSC101 replicon. Mol Microbiol 5, 1447–1457.[Medline]
Blomfield, I. C., Calie, P. J., Eberhardt, K. J., McClain, M. S. & Eisenstein, B. I. (1993). Lrp stimulates phase variation of type 1 fimbriation in Escherichia coli K-12. J Bacteriol 175, 27–36.
Blomfield, I. C., Kulasekara, D. H. & Eisenstein, B. I. (1997). Integration host factor stimulates both FimB- and FimE-mediated site-specific DNA inversion that controls phase variation of type 1 fimbriae expression in Escherichia coli. Mol Microbiol 23, 705–717.[CrossRef][Medline]
Blyn, L. B., Braaten, B. A., White-Ziegler, C. A., Rolfson, D. H. & Low, D. A. (1989). Phase-variation of pyelonephritis-associated pili in Escherichia coli: evidence for transcriptional regulation. EMBO J 8, 613–620.[Medline]
Chen, Y. & Rice, P. A. (2003). New insight into site-specific recombination from Flp recombinase-DNA structures. Annu Rev Biophys Biomol Struct 32, 135–159.[CrossRef][Medline]
Colloms, S. D., Bath, J. & Sherratt, D. J. (1997). Topological selectivity in Xer site-specific recombination. Cell 88, 855–864.[CrossRef][Medline]
Connell, I., Agace, W., Klemm, P., Schembri, M., Marild, S. & Svanborg, C. (1996). Type 1 fimbrial expression enhances Escherichia coli virulence for the urinary tract. Proc Natl Acad Sci U S A 93, 9827–9832.
Donaldson, S. G., Azizi, S. Q. & Dal Nogare, A. R. (1991). Characteristics of aerobic Gram-negative bacteria colonizing critically ill patients. Am Rev Respir Dis 144, 202–207.[Medline]
Dove, S. L. & Dorman, C. J. (1996). Multicopy fimB gene expression in Escherichia coli: binding to inverted repeats in vivo, effect on fimA gene transcription and DNA inversion. Mol Microbiol 21, 1161–1173.[CrossRef][Medline]
Eisenstein, B. I., Sweet, D. S., Vaughn, V. & Friedman, D. I. (1987). Integration host factor is required for the DNA inversion that controls phase variation in Escherichia coli. Proc Natl Acad Sci U S A 84, 6506–6510.
Forsman, K., Goransson, M. & Uhlin, B. E. (1989). Autoregulation and multiple DNA interactions by a transcriptional regulatory protein in E. coli pili biogenesis. EMBO J 8, 1271–1277.[Medline]
Gally, D. L., Bogan, J. A., Eisenstein, B. I. & Blomfield, I. C. (1993). Environmental-regulation of the fim switch controlling type 1 fimbrial phase variation in Escherichia coli K-12 – effects of temperature and media. J Bacteriol 175, 6186–6193.
Gally, D. L., Rucker, T. J. & Blomfield, I. C. (1994). The leucine-responsive regulatory protein binds to the fim switch to control phase variation of type 1 fimbrial expression in Escherichia coli K12. J Bacteriol 176, 5665–5672.
Gally, D. L., Leathart, J. & Blomfield, I. C. (1996). Interaction of FimB and FimE with the fim switch that controls the phase variation of type 1 fimbriae in Escherichia coli K12. Mol Microbiol 21, 725–738.[CrossRef][Medline]
Guerina, N. G., Kessler, T. W., Guerina, V. J., Neutra, M. R., Clegg, H. W., Langermann, S., Scannapieco, F. A. & Goldmann, D. A. (1983). The role of pili and capsule in the pathogenesis of neonatal infection with Escherichia coli K1. J Infect Dis 148, 395–405.[Medline]
Guzman, L. M., Belin, D., Carson, M. J. & Beckwith, J. (1995). Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177, 4121–4130.
Hallet, B. & Sherratt, D. J. (1997). Transposition and site-specific recombination: adapting DNA cut-and-paste mechanisms to a variety of genetic rearrangements. FEMS Microbiol Rev 21, 157–178.[CrossRef][Medline]
Hinde, P., Deighan, P. & Dorman, C. J. (2005). Characterization of the detachable Rho-dependent transcription terminator of the fimE gene in Escherichia coli K-12. J Bacteriol 187, 8256–8266.
Holden, N. J., Uhlin, B. E. & Gally, D. L. (2001). PapB paralogues and their effect on the phase variation of type 1 fimbriae in Escherichia coli. Mol Microbiol 42, 319–330.[CrossRef][Medline]
Holden, N. J., Totsika, M., Mahler, E., Roe, A. J., Catherwood, K., Lindner, K., Dobrindt, U. & Gally, D. L. (2006). Demonstration of regulatory cross-talk between P fimbriae and type 1 fimbriae in uropathogenic Escherichia coli. Microbiology 152, 1143–1153.
Iwahi, T., Abe, Y., Nakao, M., Imada, A. & Tsuchiya, K. (1983). Role of type 1 fimbriae in the pathogenesis of ascending urinary tract infection induced by Escherichia coli in mice. Infect Immun 39, 1307–1315.
Kelly, A., Conway, C., Ó Cróinín, T., Smith, S. G. & Dorman, C. J. (2006). DNA supercoiling and the Lrp protein determine the directionality of fim switch DNA inversion in Escherichia coli K-12. J Bacteriol 188, 5356–5363.
Kim, S. & Landy, A. (1992). Lambda Int protein bridges between higher order complexes at two distant chromosomal loci attL and attR. Science 256, 198–203.
Klemm, P. (1986). Two regulatory fim genes, fimB and fimE, control the phase variation of type 1 fimbriae in Escherichia coli. EMBO J 5, 1389–1393.[Medline]
Klemm, P., Jorgensen, B. J., van Die, I., de Ree, H. & Bergmans, H. (1985). The fim genes responsible for synthesis of type 1 fimbriae in Escherichia coli, cloning and genetic organization. Mol Gen Genet 199, 410–414.[CrossRef][Medline]
Krogfelt, K. A., Bergmans, H. & Klemm, P. (1990). Direct evidence that the FimH protein is the mannose-specific adhesin of Escherichia coli type 1 fimbriae. Infect Immun 58, 1995–1998.
Kulasekara, H. D. & Blomfield, I. C. (1999). The molecular basis for the specificity of fimE in the phase variation of type 1 fimbriae of Escherichia coli K-12. Mol Microbiol 31, 1171–1181.[CrossRef][Medline]
McClain, M. S., Blomfield, I. C. & Eisenstein, B. I. (1991). Roles of fimB and fimE in site-specific DNA inversion associated with phase variation of type 1 fimbriae in Escherichia coli. J Bacteriol 173, 5308–5314.
Mizunoe, Y., Matsumoto, T., Sakumoto, M., Kubo, S., Mochida, O., Sakamoto, Y. & Kumazawa, J. (1997). Renal scarring by mannose-sensitive adhesin of Escherichia coli type 1 pili. Nephron 77, 412–416.[Medline]
Mumm, J. P., Landy, A. & Gelles, J. (2006). Viewing single site-specific recombination events from start to finish. EMBO J 25, 4586–4595.[CrossRef][Medline]
Neidhardt, F. C., Bloch, P. L. & Smith, D. F. (1974). Culture medium for enterobacteria. J Bacteriol 119, 736–747.
Roesch, P. L. & Blomfield, I. C. (1998). Leucine alters the interaction of the leucine-responsive regulatory protein (Lrp) with the fim switch to stimulate site-specific recombination in Escherichia coli. Mol Microbiol 27, 751–761.[CrossRef][Medline]
Rose, R. E. (1988). The nucleotide sequence of pACYC184. Nucleic Acids Res 16, 355
Sadowski, P. (1986). Site-specific recombinases: changing partners and doing the twist. J Bacteriol 165, 341–347.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Schaeffer, A. J., Schwan, W. R., Hultgren, S. J. & Duncan, J. L. (1987). Relationship of type 1 pilus expression in Escherichia coli to ascending urinary tract infections in mice. Infect Immun 55, 373–380.
Sherratt, D. J., Arciszewska, L. K., Blakely, G., Colloms, S., Grant, K., Leslie, N. & McCulloch, R. (1995). Site-specific recombination and circular chromosome segregation. Philos Trans R Soc Lond B Biol Sci 347, 37–42.
Sohanpal, B. K., Kulasekara, H. D., Bonnen, A. & Blomfield, I. C. (2001). Orientational control of fimE expression in Escherichia coli. Mol Microbiol 42, 483–494.[CrossRef][Medline]
Stadtman, E. R. & Levine, R. L. (2003). Free radical-mediated oxidation of free amino acids and amino acid residues in proteins. Amino Acids 25, 207–218.[CrossRef][Medline]
Studier, F. W., Rosenberg, A. H., Dunn, J. J. & Dubendorff, J. W. (1990). Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol 185, 60–89.[Medline]
Xia, Y., Forsman, K., Jass, J. & Uhlin, B. E. (1998). Oligomeric interaction of the PapB transcriptional regulator with the upstream activating region of pili adhesin gene promoters in Escherichia coli. Mol Microbiol 30, 513–523.[CrossRef][Medline]
Xia, Y., Gally, D., Forsman-Semb, K. & Uhlin, B. E. (2000). Regulatory cross-talk between adhesin operons in Escherichia coli: inhibition of type 1 fimbriae expression by the PapB protein. EMBO J 19, 1450–1457.[CrossRef][Medline]
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.[CrossRef][Medline]
Received 4 June 2007;
revised 11 September 2007;
accepted 12 September 2007.
This article has been cited by other articles:
|
|
 |
|
  Y. Wu and F. W. Outten IscR Controls Iron-Dependent Biofilm Formation in Escherichia coli by Regulating Type I Fimbria Expression J. Bacteriol., February 15, 2009; 191(4): 1248 - 1257. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
| J MED MICROBIOL | ALL SGM JOURNALS | |
Lrp; ZAP1183), and inhibited by PapB (WT+PapB; ZAP1180). The second panel shows on- and off-specific products, indicating recombination of WT fimS (on orientation) by FimB in the presence of lrp (WT; AAEC370A) but not in the absence of lrp (
