Evidence that the Rhizobium regulatory protein RirA binds to cis-acting iron-responsive operators (IROs) at promoters of some Fe-regulated genes

K. H. Yeoman1,{dagger}, A. R. J. Curson1,{dagger}, J. D. Todd1, G. Sawers2 and A. W. B. Johnston1

1 School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, UK
2 John Innes Centre, Norwich Research Park, Colney Lane, Norwich NR4 7UH, UK

Correspondence
Andrew W. B. Johnston
a.johnston{at}uea.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Mutations in rirA of Rhizobium have been shown to deregulate expression of several genes that are normally repressed by iron. A conserved sequence, the iron-responsive operator (IRO), was identified near promoters of vbsC (involved in the synthesis of the siderophore vicibactin), rpoI (specifies an ECF {sigma} factor needed for vicibactin synthesis) and the two fhuA genes (encoding vicibactin receptor). Removal of these IRO sequences abolished Fe-responsive repression. Most of these genes were constitutively expressed in the heterologous host, Paracoccus denitrificans, but introduction of the cloned rirA gene repressed expression of these Rhizobium genes in this heterologous host if the corresponding IRO sequences were also intact. These observations are the first to examine the mechanisms of RirA, which has no sequence similarity to well-known iron-responsive regulators such as Fur or DtxR. They provide strong circumstantial evidence that RirA is a transcriptional regulator that binds to cis-acting regulatory sequences near the promoters of at least some of the genes whose expression it controls in response to Fe availability.


Abbreviations: IRO, iron-responsive operator; SDM, site-directed mutagenesis; VB, vicibactin

{dagger}Both authors contributed equally to this work.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Iron is needed as a component of many proteins. Although abundant on Earth, Fe acquisition presents difficulty because of its extremely low solubility in most biotic conditions, and microbes use many strategies to sequester it from their environments. Although essential, intracellular Fe is also potentially toxic, since it catalyses the formation of harmful radicals. Therefore the genes involved in Fe uptake, storage and utilization are tightly regulated in bacteria (see Andrews et al., 2003).

In Gram-negative bacteria, the most widely used (and most studied) Fe-responsive gene regulator is Fur (ferric uptake regulator). In model {gamma}-Proteobacteria (e.g. Escherichia coli or Pseudomonas), in the presence of Fe2+, this dimeric protein (Kolade et al., 2002; Pohl et al., 2003) binds to sequences, known as fur boxes, and prevents transcription of the target promoters. Fur mutants in E. coli also have reduced expression of several transcripts (McHugh et al., 2003). In some cases, this is via Fur-mediated repression of a small RNA molecule, RyhB, which then represses transcription of genes whose expression appeared to be induced by Fur (Masse & Gottesman, 2002).

Other Fe-responsive regulators include those of the ‘Dtx’ class, which are important in Gram-positive bacteria and cyanobacteria (see Andrews et al., 2003). Although differing in their primary sequences, the structures of the Dtx and the Fur families share broad structural similarities (see Pohl et al., 2003).

A third class of wide-ranging, Fe-responsive transcriptional regulators, termed RirA, was recently described for Rhizobium leguminosarum, the symbiotic N2-fixing {alpha}-proteobacterium that induces root nodules on peas, beans and clover (Todd et al., 2002), a process that is very demanding for iron (see Johnston et al., 2001). Mutations in rirA cause deregulated expression of several genes whose transcription is normally repressed in cells grown in Fe-replete medium. These include the vbs and fhu genes, which are involved, respectively, in the synthesis and uptake of the siderophore vicibactin (VB) (Carter et al., 2002; Stevens et al., 1999; Yeoman et al., 2000); hmuPSTUV, which specifies a haem ABC transporter (Wexler et al., 2001); tonB (Wexler et al., 2001); and rpoI, which encodes an ECF {sigma} factor that specifically recognizes promoters of the vbsADL and vbsGSO VB biosynthetic genes (Yeoman et al., 1999, 2003). Transcription of rirA itself is enhanced in Fe-replete conditions and is auto-regulated (Todd et al., 2002).

There is no detectable sequence similarity between RirA and members of the Fur or Dtx families, suggesting that RirA represents a novel class of Fe-responsive regulator. There are close homologues (>70 % amino acid identity) of RirA, but these are confined (so far) to Sinorhizobium, Mesorhizobium, Agrobacterium and the animal pathogen Brucella, all of which are near relatives of Rhizobium. However, RirA has ~30 % amino acid identity to the large (181 members) ‘Rrf2’ family, which occurs in all eubacteria, with many genera (including Rhizobium) having more than one paralogue. Rrf2 itself regulates cytochrome biosynthesis in Desulfovibrio (Keon et al., 1997). The only other member to have been studied in detail, IscR of E. coli, regulates the adjacent iscRSUA, operon, which is involved in the synthesis of FeS clusters. IscR is itself an FeS-cluster protein, repressing iscRSUA transcription, but is only fully effective if charged with this co-factor (Schwartz et al., 2001). IscR does not repress the iscRSUA operon in Fe-depleted or oxygen-stressed E. coli cells, indicating that this regulator responds to both of these environmental signals (Outten et al., 2004).

It is possible that a subset of the Rrf2 superfamily evolved into a specialized Fe-sensing regulator in Rhizobium and its very close relatives, with RirA mediating many functions that in E. coli are undertaken by Fur. Consistent with this, Rhizobium has a gene whose product resembles a ‘classical’ Fur, but mutations in the fur-like gene do not affect Fe-responsive regulation of the fhu, vbs, rpoI, hmu and tonB genes alluded to above (Wexler et al., 2003). Indeed, this Fur-like protein is actually a manganese-sensing regulator, termed ‘Mur’, which represses the operon that specifies a Mn2+ ABC transporter, SitABCD, in Mn2+-replete cells (Diaz-Mireles et al., 2004; Chao et al., 2004).

Here we try to elucidate how RirA regulates Fe-responsive genes in R. leguminosarum and identify potential RirA-binding sites near promoters of some of its target genes.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strains, plasmids, media and genetic manipulations in vivo and in vitro.
Strains and plasmids are described in Tables 1 and 2. R. leguminosarum and E. coli were grown as described by Beringer (1974), and Paracoccus denitrificans as described by Carter et al. (2002). Techniques for conjugal gene transfer and assaying {beta}-galactosidase in Fe-depleted media were as described by Wexler et al. (2001) and references therein.


View this table:
[in this window]
[in a new window]
 
Table 1. Plasmids constructed and used in this study

 
Nucleic acid manipulations.
Routine methods were used for cloning and plasmid isolations. Transcriptional fusions were made by cloning promoter fragments in the wide host-range reporter plasmid pMP220 (Spaink et al. 1987), in some cases having cloned the fragment initially in pUC18 (Messing et al., 1983).

To map transcript starts (Sawers & Böck, 1989), cells were grown in Fe-replete or Fe-depleted media and total RNA was isolated. Using primers located within the structural, promoter-proximal gene, primer extensions were done, together with manual sequencing of the corresponding cloned segment of DNA to identify transcription initiation sites. The sequences of the primers used for these experiments were as follows: vbsC (5'-GATCGGCAGAGCTCCGCCTTG-3'); fhuA1 (5'-CTTGTGAACACACGCAACTTTCCC-3'); fhuA2 (5'-GGCTATCTCTATAAATTCGCG-3'); fhuC (5'-CGGCGACAAAAGTTTCAGCCGTTC-3').

Site-directed mutagenesis (SDM).
The 10 bp deletion in IRORi was generated using the ExSite kit (Stratagene) and primers that hybridized on either side of the designated deletion. The template DNA was plasmid pBIO1327, which contains a 400 bp DNA fragment spanning the rpoI–vbsC intergenic space, cloned in pUC18. Mutant plasmids were verified by DNA sequence analysis and the mutated insert was recloned into the wide host-range transcriptional fusion vector pMP220 in the appropriate orientation. Site-directed substitution mutations were generated using the QuikChange XL kit (Stratagene) to mutate the wild-type plasmid pBIO1328, which contains the same insert as in pBIO1327, but cloned in pMP220, forming a rpoI–lacZ fusion in pMP220. Mutagenic primers, 25 bp in length (synthesized by MWG-Biotech) and containing a single base pair substitution compared to wild-type (Table 2), were used in a PCR-based mutagenic reaction, and the products were transformed into E. coli DH10B-T1R (Invitrogen), selecting tetracycline resistance. Mutations were verified by DNA sequencing. To mutate the IROVc, the vbsC–lacZ fusion plasmid pBIO1306, which contains 780 bp, spanning the rpoI–vbsC intergenic space, cloned in pMP220 was used as a template. Using a QuikChange XL kit (Stratagene), 8 bp of IROVc were removed and replaced with a NotI site (5'-GCGGCCGC-3') 16 bp downstream from the vbsC transcriptional start site. Mutations were verified by sequencing. The resultant mutant plasmid was termed pBIO1466. The mutant rpoI–lacZ fusion plasmid pBIO1467, which lacks IROVc, was derived from pBIO1466, from which the insert was cloned into pMP220, but in the opposite orientation.


View this table:
[in this window]
[in a new window]
 
Table 2. Effect of mutating individual nucleotides in IRORi on the expression of rpoI–lacZ fusions

Plasmid pBIO1328 is an rpoI–lacZ fusion derived from pMP220 (see Methods). Derivatives of pBIO1328 with individual nucleotide substitutions (A->T, T->A, G->C or C->G) were made in the IRORi sequence (see Methods); the altered bases are underlined and in bold type. Wild-type and mutant plasmids were mobilized to wild-type R. leguminosarum J251, and assayed in triplicate for {beta}-galactosidase after growth in high- or low-level Fe media.

 

   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Location of IRO sequences relative to the transcript starts of vbsC and rpoI
In R. leguminosarum, RpoI is the {sigma} factor responsible for transcribing the vbsADL and vbsGSO operons that specify enzymes involved in the synthesis of the siderophore VB (Yeoman et al., 1999, 2003). Divergently transcribed from rpoI is vbsC, which specifies an enzyme that likely acetylates VB (Carter et al., 2002). Expression of both rpoI and vbsC is reduced in wild-type, Fe-replete cells, but in rirA mutants this regulation is lost (Todd et al., 2002). Transcription of vbsC and rpoI is normal in rpoI mutants; thus RpoI is not the {sigma} factor for either of these genes (Carter et al., 2002).

In the 163 bp intergenic region between rpoI and vbsC of R. leguminosarum strain J251, we noted two similar (11 of 17 bp identical) sequences, termed iron-responsive operators (IROs), with opposing orientation (see Fig. 1). Similar sequences occur near other Fe-regulated promoters (see below). The conserved sequence 5' of rpoI, which we term IRORi, is centred 40 bp 5' of the previously identified (Yeoman et al., 2003) transcription initiation site of rpoI. The other, termed IROVc, is immediately 5' of the rpoI transcript start site (Fig. 1).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 1. (a) Architecture of the promoter regions of the RirA-regulated genes. Representation of the rpoI–vbsC intergenic region. Thick arrows show locations and orientation of genes in the fhuA2F–vbsCrpoI region. The sequence of the vbsC–rpoI intergenic space is shown below. Locations of IROVc and IRORi are shown under the brackets, with the nucleotides that were removed (for IRORi) or replaced (for IROVc) by SDM (see text) shown by white type on a black background. The rpoI and vbsC transcription initiation sites are indicated, as are the distances to the translational starts of these two genes. (b) Comparison of 17 bp IRO sequences in the fhu and vbsC promoter regions. The 17 bp IRORi sequence is shown at the top. Nucleotides in the IRO motifs 5' of the fhuA1, fhuA2 and vbsC promoters that are identical to those in IRORi are underlined. Distances (bp) to the corresponding transcript start sites are shown.

 
To locate IROVc precisely, relative to the vbsC promoter, the vbsC transcription initiation site was identified by primer extension, using total RNA isolated from R. leguminosarum grown in Fe-replete and Fe-depleted media. A single vbsC transcript was found, whose intensity was, as expected, greater with mRNA obtained from cells grown in low-Fe medium (Fig. 2). The IROVc sequence was centred around 17 bp 5' of this vbsC transcriptional start site, closer than is IRORi relative to the rpoI transcript start. The 5' ends of the rpoI and vbsC transcripts are separated by 34 bp; thus there is overlap in the promoter regions of each of these two divergently transcribed genes (Fig. 1).



View larger version (81K):
[in this window]
[in a new window]
 
Fig. 2. Transcription of vbsC is regulated in response to iron levels. The transcription initiation site of vbsC was determined by primer extension with total RNA (25 µg) isolated from R. leguminosarum after growth in media with high and low concentrations of Fe. The DNA sequence was determined with the same primer and with a cloned DNA fragment spanning the corresponding region of the vbsC gene. The 5' end of the transcript is indicated by the arrow and the precise location of the initiation site relative to the DNA sequence is shown.

 
Effect of mutating the IRO sequences on the expression of rpoI and vbsC
Given their locations, relative to the transcription initiation sites of vbsC and rpoI, IROVc and IRORi might be cis-acting elements involved in Fe-responsive gene expression. To examine this, a series of mutations was made in which parts of these IRO sequences were individually mutated and the effects on expression of vbsC and rpoI determined. To do this, similar fragments spanning the intergenic region were cloned, in both orientations, into the wide host-range promoter-probe lac fusion plasmid pMP220 to form pBIO1328 (rpoI–lacZ fusion) and pBIO1306 (vbsC–lacZ fusion). From each of these, mutant derivatives were constructed in which 10 bp were removed from the 5' end of IRORi to form pBIO1335 (rpoI–lacZ) and pBIO1354 (vbsC–lacZ), respectively. Similarly, 8 bp from the 5' end of IROVc in pBIO1328 and pBIO1306 were replaced by a NotI site to form pBIO1467 and pBIO1466, respectively (Fig. 3). Each plasmid was mobilized into wild-type R. leguminosarum strain J251 and transconjugants, grown in Fe-replete or Fe-depleted media, were assayed for {beta}-galactosidase.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 3. Effect of mutating IRORi and IROVc on expression of rpoI–lacZ and vbsC–lacZ fusions. The two diagrams indicate the organization of the oppositely oriented inserts in the two parental transcriptional lacZ fusion plasmids pBIO1328 (rpoI–lacZ) and pBIO1306 (vbsC–lacZ), in both of which IRORi (vertical stripes) and IROVc (horizontal stripes) are intact. Mutant plasmids, which were deleted for 10 bp in IRORi (pBIO1335 and pBIO1354) or which had an 8 bp substitution in IROVc (pBIO1467 and pBIO1466) were derived from pBIO1306 and pBIO1328. The wild-type parental fusions and each mutant plasmid were mobilized into wild-type R. leguminosarum and the transconjugants grown in high-Fe and low-Fe media before assaying {beta}-galactosidase enzyme activities in triplicate. Results, in Miller Units, are shown for each plasmid, for both media.

 
As expected, with the two wild-type fusions, expression of both vbsC–lacZ and rpoI–lacZ was greater in cells grown in Fe-depleted medium (Fig. 3). As noted previously (Carter et al., 2002; Yeoman et al., 1999), with rpoI–lacZ there is still significant {beta}-galactosidase enzyme activity even in high-Fe medium, but vbsC–lacZ expression is strongly repressed by Fe. Strikingly, the mutations in IRORi (in pBIO1335) and IROVc (in pBIO1466) caused a large increase in rpoI–lacZ and vbsC–lacZ expression, respectively, and totally abolished Fe-dependent repression (Fig. 3). Indeed, with these mutant derivatives, expression of rpoI–lacZ and vbsC–lacZ was enhanced, rather than repressed, by growth in Fe-replete conditions. Although the most striking effects of removing IRORi and IROVc were on the expression of rpoI–lacZ and vbsC–lacZ, respectively, there was a significant enhancement of vbsC expression (in plasmid pBIO1354) when IRORi was removed, although Fe-responsive expression was retained. This 10-mer deletion would potentially relocate the vbsC transcript start site (see Fig. 1a), perhaps enhancing the initiation of vbsC transcription. In contrast, the 8-mer IROVc mutation in the rpoI–lacZ fusion plasmid pBIO1467 reduces rpoI expression, leading to constitutive low-level transcription in both high-Fe and low-Fe media (Fig. 2). Since IROVc is located just 5' of the rpoI transcription start site, this effect may be due to the loss of binding of a trans-acting protein near the rpoI promoter, or it may arise from the alteration of the –10 sequence at the rpoI promoter.

Having established its importance in the control of rpoI expression, we examined IRORi in more detail. Mutant derivatives of pBIO1328, which contains the intact rpoI promoter and regulatory region, were constructed by site-directed mutagenesis (SDM) in which individual bases of IRORi were substituted. The resulting mutant plasmids were each mobilized into wild-type R. leguminosarum and the transconjugants assayed for {beta}-galactosidase activity after growth in high-Fe and low-Fe media. As shown in Table 2, most of these substitutions had little or no effect on rpoI–lacZ expression in either medium, but those at the 3' end of the motif abolished Fe-responsive repression of the fusion. However, these mutations at the 3' end of the IRORi also reduced rpoI expression, even in –Fe medium, indicating they affect promoter activity and the response to Fe availability. This is consistent with their location near the –35 region relative to the rpoI transcriptional start. Two other adjacent mutations, in pBIO1473 and pBIO1427 (Table 2), elevate rpoI–lacZ expression, especially in low-Fe medium. Although upstream from the –35 region, these two mutations may enhance rpoI promoter activity, either directly through interactions with RNA polymerase or, perhaps, by indirect interaction with another, unknown, factor.

Cloned rirA confers Fe-responsive repression of rpoI and vbsC in P. denitrificans
Carter et al. (2002) showed that the cloned vbsGSO and vbsADL genes were not expressed in the {alpha}-proteobacterium P. denitrificans unless the cloned R. leguminosarum {sigma} factor gene rpoI was also present in this heterologous host. Thus, the promoters of vbsGSO and vbsADL are not recognized by a native P. denitrificans {sigma} factor, but rpoI itself is expressed in this bacterium. We therefore directly examined rpoI expression in P. denitrificans and, for comparison, the divergently transcribed vbsC gene. Plasmids pBIO1328 and pBIO1306 (rpoI–lacZ and vbsC–lacZ fusions, respectively), were each mobilized into P. denitrificans, and the transconjugants were grown in medium with high and low concentrations of Fe. Both fusions were expressed in this host, but their expression was unaffected by the Fe status of the medium (Table 3), showing that P. denitrificans cannot mediate Fe-responsive regulation of rpoI or vbsC.


View this table:
[in this window]
[in a new window]
 
Table 3. Expression of rpoI and vbsC genes in P. denitrificans in the presence and absence of the cloned rirA gene

Transconjugants of P. denitrificans containing various rpoI–lacZ and vbsC–lacZ fusion plasmids (see Fig. 3) were assayed for {beta}-galactosidase enzyme activity after growth in high-Fe and low-Fe media. In some strains, the cloned rirA gene, in pBIO1451, was also present.

 
This suggested a genetic approach to test if RirA bound to IRO sequences near the vbsC and rpoI promoters. First, by using a rirA–lacZ fusion plasmid, we showed that rirA itself was expressed from its own promoter in P. denitrificans (not shown). We then cloned rirA into the wide host-range vector pOT2 (Allaway et al., 2001), to form pBIO1451. [Although pOT2 was designed primarily to be used as a promoter-probe plasmid (it has a promoterless gfp gene 3' of its polylinker), in pBIO1451 rirA gene is expressed from its own native promoter.] This plasmid was then transferred (selecting the gentamicin resistance, specified by pOT2) into P. denitrificans harbouring the rpoI–lacZ or vbsC–lacZ fusion plasmids (see above). Transconjugants were grown in high-Fe and low-Fe media before assaying {beta}-galactosidase enzyme activity. As shown in Table 3, the introduction of the cloned rirA gene caused a large reduction in the expression of both fusions. With the rpoI–lacZ fusion plasmid pBIO1328, this RirA-dependent repression was more marked in the high-Fe than in the low-Fe medium. However, with the vbsC–lacZ fusion pBIO1306, the repressive effects of introducing rirA were very similar in both sets of media.

To show that an IRO box is needed for this repression, the mutant rpoI–lacZ fusion plasmid pBIO1335, which has a deletion in IRORi, was transferred to P. denitrificans. As with its intact parent, pBIO1328, there was constitutive, high-level expression of rpoI–lacZ in P. denitrificans harbouring pBIO1335. Significantly, introduction of the cloned rirA gene no longer conferred Fe-responsive repression to this mutated rpoI–lacZ fusion (Table 3). For reasons that are not clear, the presence of rirA further enhanced the level of expression from pBIO1335.

With pBIO1466, the mutated vbsC–lacZ fusion plasmid that lacks an intact IROVc, a similar pattern was found. Whereas the expression of {beta}-galactosidase from the parental plasmid pBIO1306 was repressed by introducing cloned rirA into P. denitrificans, when IROVc was removed, the inhibitory effect of RirA in this heterologous host was abolished (Table 3).

Taken together, these experiments, in which P. denitrificans was used as a ‘null’ background, provide strong circumstantial evidence that RirA protein acts directly on the IRO motifs of vbsC and rpoI, resulting in a very significant drop in the levels of transcription of both genes. In the case of the rpoI–lacZ fusion, this repression is much greater in the high-Fe than in the low-Fe medium just as it is in R. leguminosarum itself. However, with the vbsC fusion, RirA is almost equally repressive in both media. The reasons for this difference are unknown, but may suggest the presence of other regulatory factors that are needed for the precise regulation of some, but not all, RirA-responsive promoters. It might also be, though, that the presence of extra copies of the cloned rirA can confer repression of vbsC, even in the absence of Fe.

Different arrangements of fhuA genes in different field isolates of R. leguminosarum
In R. leguminosarum field-isolated J251, we had shown that the uptake of VB is mediated by the Fhu transport proteins, the outer-membrane receptor being FhuA, the periplasmic binding protein FhuD, the inner-membrane transporter FhuB and the ATPase FhuC (Stevens et al., 1999; Yeoman et al., 2000). Analysis of the genome of the sequenced R. leguminosarum field isolate 3841 showed that it has two fhuA genes, with ‘fhuA1 being on the symbiotic plasmid, pRL10JI, ~145 kb from the nod genes (http://www.sanger.ac.uk/Projects/R_leguminosarum/) and the other copy, ‘fhuA2’ being in the fhuA2F–rpoI–vbs gene cluster on a different native plasmid, pRL12JI. The finding of two apparently intact copies of fhuA in strain 3841 was surprising, since field isolate J251 had been shown to have one intact fhuA, in the same location relative to rpoI as that of fhuA2 in field isolate 3841. In the locus occupied by fhuA1 in strain 3841, field isolate J251 has a non-transcribed pseudogene, {psi}fhuA, with many stop codons, which, like fhuA1 of 3841, is orientated divergently from fhuCDB. Despite this, {psi}fhuA is 81 % identical over 460 bp at its 3' end to fhuA1 of strain 3841. Thus, J251 probably had an intact fhuA1, but its function was lost by accumulated mutations (Yeoman et al., 2000). In strain 3841, both fhuA1 and fhuA2 are functional: when individually cloned, each can confer VB uptake ability to a fhuA mutant of strain J251 (results not shown).

IRO-like sequences and the promoters of fhu genes involved in VB uptake
The reason for describing the differences in the organization of fhuA genes in these two strains is that when we examined the R. leguminosarum strain 3841 genome for IRO-like sequences (based on IRORi), the closest match (14 of 17 identical, Figs 1a and 4) was located 5' of fhuA2. A near-identical sequence was also seen in the corresponding region 5' of fhuA2 of field isolate J251. We also examined the DNA 5' of the other two fhu operons, fhuA1 and fhuCDB, in strain 3841 for IRO-like sequences. As seen in Figs 1(b) and 4, there was a region 88 bp 5' of fhuA1 with limited similarity to IRORi, this being most pronounced at the 3' end of the motif, around the –35 bp region. In contrast, we found no IRO-like motifs near the transcript start of the fhuCDB operon, which had been located by primer extension experiments, even though fhuCDB expression is regulated in response to Fe availability, this being mediated by RirA (Todd et al., 2002; unpublished observations).



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 4. Expression of various fhulacZ fusions in R. leguminosarum and P. denitrificans. The DNA sequences in the regions upstream of the transcriptional start sites of fhuA1 and fhuA2F are shown. Sequences of the corresponding IRO motifs are shown as underlined bases. Two wide host-range fhuA1–lacZ fusion plasmids, pBIO1508 and pBIO1510, were constructed by amplifying fragments of 338 and 318 bp, respectively, from strain J391 genomic DNA. The 3' ends of these plasmids are within fhuA1, and their respective 5' ends are shown as the open, vertical arrows. Two fhuA2–lacZ fusion plasmids, pBIO1512 and pBIO1514, containing 369 and 198 bp, respectively, were constructed in a similar way. All PCR products had EcoRI and SphI sites at their termini, to facilitate cloning into pUC18, and thence into the wide host-range vector pMP220. Transconjugants of wild-type R. leguminosarum (Rl) or P. denitrificans (Pd) containing each of these four fusions were grown in high-Fe (+Fe) or low-Fe medium (–Fe) and assayed for {beta}-galactosidase. In addition, strains containing these lac fusion plasmids plus the cloned rirA gene (in pBIO1451) were also assayed as indicated. Activities are shown in Miller units, the mean of three independent measurements. NT, Not tested.

 
To establish the roles, if any, of the IRO-like sequences (‘IROFa1 and ‘IROFa2’) 5' of fhuA1 and fhuA2, respectively, we first located their transcription-initiation sites, by primer extensions with RNA obtained from R. leguminosarum grown in high-Fe and low-Fe medium and with primers located in fhuA1 or fhuA2. Both genes had a single transcription-initiation site, whose intensity was greater in the Fe-depleted cells (not shown). Thus, the IROFa1 motif overlaps the fhuA1 promoter region (see Fig. 4). However, IROFa2 was ~50 bp 5' of the fhuA2F transcript start, more distant than in the other IRO motifs for the other genes described above (see Figs 1b and 4). The promoter regions of the fhuA1 and fhuA2F operons were studied in more detail.

fhuA1.
A transcriptional fhuA1–lacZ fusion plasmid, pBIO1508, which contains DNA extending 60 bp 5' of the fhuA1 transcriptional start, exhibited normal Fe-responsive expression in wild-type R. leguminosarum (Fig. 4). In the RirA mutant J397 (Todd et al., 2002), expression of this fhuA1–lacZ fusion was at a high level, even in high-Fe medium (data not shown), formally demonstrating that RirA mediates Fe-responsive regulation of fhuA1. When the slightly smaller fusion plasmid pBIO1510, which lacks IROFa1, was mobilized to wild-type R. leguminosarum, the expression of {beta}-galactosidase was essentially the same in +Fe and –Fe media (Fig. 4). Thus, removal of IROFa1 abolishes Fe-responsive regulation of fhuA1.

These two fhuA1–lacZ fusion plasmids were also mobilized into P. denitrificans. In this background, both of them expressed {beta}-galactosidase constitutively. Thus, as with rpoI and vbsC, this heterologous species can transcribe the introduced R. leguminosarum gene but cannot effect Fe-responsive regulation of its expression. When pBIO1451 (rirA cloned in pOT2), was introduced into P. denitrificans that harboured pBIO1508, there was significant reduction in the level of {beta}-galactosidase activity. This was more pronounced in the high-Fe than in the low-Fe medium, but in both regimes there was still a significant level of expression, showing that the repression was not as effective as that of the analogous experiments using the rpoI–lacZ fusion plasmid in P. denitrificans (see above). Importantly, though, the presence in P. denitrificans of the cloned rirA gene had no effect on the expression from pBIO1510, which lacks an intact IROFa1 (Fig. 4).

The IROFa1 motif is therefore likely to be a cis-acting regulatory element that interacts with RirA, despite its considerable sequence divergence compared to IRORi and IROVc.

fhuA2.
As mentioned above, the best match to IRORi in the genome of R. leguminosarum is to a motif (IROFa2) 5' of the fhuA2F promoter, but further upstream of the fhuA2F promoter than were the other IRO sequences relative to their promoters. Nevertheless, this IROFa2 motif appears to be important in the Fe-responsive regulation of fhuA2F, as seen by the behaviour of two fhuA2–lacZ fusion plasmids in wild-type R. leguminosarum grown in high-Fe or in low-Fe medium. One of these, pBIO1512, which contains the intact IROFa2 and extends 87 bp 5' of the fhuA2 transcriptional start site, exhibited Fe-responsive repression of {beta}-galactosidase in R. leguminosarum grown in Fe-replete medium (Fig. 4). However, the smaller plasmid pBIO1514, whose 5' end is within IROFa2, shows no Fe-dependent regulation, its expression being essentially the same in the +Fe and –Fe media.

In contrast to other lac fusions described here, when pBIO1512 was transferred to P. denitrificans, its expression was affected by the Fe availability of the medium, the levels of {beta}-galactosidase being approximately twofold lower in the high-Fe than in the low-Fe medium. With the smaller fusion plasmid pBIO1514, which lacks an intact IROFa2, no such Fe-responsive regulation was seen. This suggests that the region 5' of fhuA2 contains a sequence that is recognized by an (unknown) Fe-responsive regulator that is native to P. denitrificans. Introducing the cloned rirA into P. denitrificans harbouring the fhuA2–lacZ fusion pBIO1512 had little further effect, the ratio of expression in high-Fe and low-Fe media being similar (~0·3) to that when the P. denitrificans did not contain rirA.

This difference in behaviour of the fhuA2 regulatory region in the heterologous host was not due to the use in P. denitrificans of a promoter different to that used in R. leguminosarum. From primer extension experiments, using RNA from P. denitrificans harbouring the cloned fhuA2 gene, it was found that the fhuA2 transcriptional start was identical in the two bacteria, and that the levels of expression were significantly greater in cells that had been grown in Fe-depleted medium (not shown).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Todd et al. (2002) showed that mutations in R. leguminosarum rirA cause constitutive expression of several genes involved in iron uptake, which are normally repressed under Fe-replete conditions. However, it was not determined whether the RirA protein was a transcriptional regulator that recognizes motifs near the promoters of its target genes or not. The only other member of the Rrf2 family to be studied in detail is IscR, which binds at the operator of the iscRSUA operon (Schwartz et al., 2001). It would be surprising, a priori, if RirA is not a DNA-binding regulator, which may also contain an [FeS] cofactor.

Here, a combination of deletion analyses and expression studies of various Fe-responsive Rhizobium genes in Paracoccus has provided strong circumstantial evidence that RirA likely does recognize IRO motifs. Although this work revealed some similarities in the ways in which the IROs identified here affect Fe-responsive gene expression, there are also some differences, pointing to differences in the exact way in which RirA mediates its regulatory effects on individual target genes. In what follows, it should be kept in mind that although we know that expression of vbsC, rpoI and the fhu genes does not involve the RpoI {sigma} factor, it has not been established whether transcription of one or more of these genes is mediated by the housekeeping {sigma}70 or by one (or more) of the other {sigma} factors that Rhizobium possesses. It cannot be discounted, therefore, that the different responses of the genes examined here may be due to their being transcribed by different {sigma} factors.

Here, we have used an operational definition of a consensus of a 17-mer as the IRO sequence. However, since in some cases (rpoI and fhuA1) the 3' end of the IRO motif overlaps the –35 region of the promoter, we have not formally distinguished bases that may be conserved because they constitute part of the RNA polymerase recognition sequence from those that may be specifically involved in interacting with RirA. Indeed, some nucleotides may be involved in both these functions.

The most detailed studies were done on rpoI and the divergently transcribed vbsC. Removal (in the case of rpoI) or replacement (for vbsC) led to their enhanced expression, especially in medium that was Fe-supplemented. Thus, the importance of these sequences in Fe-responsive regulation is unequivocal. The locations of the IRORi and IROVc motifs relative to the transcription starts of the corresponding genes were different, but in both cases would be in accord with transcriptional repression of rpoI and vbsC. Further, SDM of individual bases in the 5' regions of the IRORi motif had little or no effect on Fe-responsive regulation of rpoI. These sets of observations suggest that there may be considerable flexibility in the sequences required for the interaction with RirA. This was further borne out by the examination of the IRO-like sequences in the promoter regions of fhuA1 and fhuA2. In the former, the IROFa1 motif overlaps the –35 region of the fhuA1 transcript, just as does IRORi relative to the rpoI promoter. However, the DNA sequences of IROFa1 and IRORi are not strikingly similar, especially towards their 5' ends. The reverse is true for the IROFa2 motif, which strongly resembles the sequence of IRORi, but is significantly further upstream of the fhuA2F promoter than are the other IROs from their corresponding transcription start sites. Such limited sequence similarity between factor-binding sites has been observed previously for other transcriptional regulators. For example, with OxyR from E. coli, the nucleotide contacts required for protein recognition are separated by 10 bp intervals (Toledano et al., 1994) and different genes have different spacings between the OxyR-binding sites and the promoter.

All the fusions studied here were expressed in the heterologous {alpha}-proteobacterium P. denitrificans, but with one exception (fhuA2F) they were not subject to Fe-responsive repression in this background. The behaviour of most of the lac fusions in this study indicates that P. denitrificans does not have a functional equivalent of RirA. This is not unexpected, since close homologues of RirA exist only in the very near relatives of the rhizobia and are not found even in other {alpha}-proteobacteria.

However, the Fe-responsive repression of the fhuA2FlacZ fusion indicates that, at least in some cases, the regions 5' of Fe-regulated R. leguminosarum genes can be recognized by regulators in this {alpha}-proteobacterium. It is becoming increasingly clear that bacterial genomes are remarkably mosaic in form, with large regions having likely been acquired by lateral gene transfer (Lawrence & Hendrickson, 2003). Possibly, therefore, fhuA2F was acquired from another bacterium relatively recently, and had retained some of the cis-acting regulatory sequences that were used in its previous host, in addition to having acquired the cognate IRO sequences appropriate for RirA-dependent regulation in R. leguminosarum. Thus, it might be potentially subject to one of two different mechanisms of Fe-responsive regulation, one appropriate for Paracoccus and the other for Rhizobium. We do not know what is/are the Fe-responsive regulator(s) in Paracoccus. Indeed, little or nothing is known of Fe-responsive gene regulation in any {alpha}-proteobacterium, even in such genetically well-characterized genera as Caulobacter and Rhodobacter.

However, the lack of Fe-responsive regulation of most of the RirA-regulated operons studied here in P. denitrificans presented an effective means of studying the effects of RirA on its target genes in a heterologous null background. There was a striking repression of fhuA1–lacZ, vbsC–lacZ and rpoI–lacZ fusion expression when the cloned rirA was also introduced into P. denitrificans, and this required the corresponding cognate IRO motif to be present. Although the introduced rirA gene may affect the expression of another regulator, native to P. denitrificans, which in turn regulates the fusion plasmid, we feel that this is inherently unlikely. It was noticeable that in P. denitrificans the cloned rirA gene markedly reduced expression of rpoI, fhuC and fhuA1, even in cells that had been grown in Fe-depleted medium. With rpoI–lacZ and, to a lesser extent, with fhuA2–lacZ, growth in high-Fe medium enhanced the repression, but this was not the case with the vbsC–lacZ fusion. Therefore, R. leguminosarum (but not P. denitrificans) may have some other factor that contributes to its ability to sense the availability of Fe, ensuring the appropriate level of Fe-dependent control in the native host.

This is the first report to examine the factors involved in RirA-mediated gene regulation in response to Fe availability. Although IRO motifs are clearly important to the ability of RirA to repress gene expression, the differences in the exact behaviour of the promoters that were studied here point to other unknown factors that influence the ability of RirA to affect the expression of specific transcriptional units. Indeed, one of the operons, fhuCDB, whose transcription quite clearly responds to Fe availability in an RirA-dependent way, does not even have a recognizable IRO motif near its promoter.

In addition to the genes involved in the synthesis and uptake of VB, Todd et al. (2002) showed that the Fe-dependent repression of other R. leguminosarum genes, namely haem-uptake (hmu) genes, ABC transporter (fbp) genes and tonB, also requires RirA. The transcriptional start sites of the hmuPSTUV and the orf1–tonB operons have been identified (Wexler et al., 2001); in both cases, a sequence similar to the 3' end of an IRO motif was located at the ~–35 position. More recently, proteomic studies have shown that >100 polypeptide gene products are differentially expressed in rirA mutant strains of R. leguminosarum, compared to the near-isogenic wild-type strains (J. D. Todd, unpublished observations). The transcript initiation sites for the corresponding genes have not been determined, but it does appear that at least some of these newly discovered RirA-responsive genes do not have a motif with similarity to the IRO motifs in their putative promoter regions.

Lynch et al. (2001) identified a more ‘local’ form of Fe-responsive gene control in Sinorhizobium meliloti. In this species, rhrA specifies a positive regulator that activates expression of the rhb genes, which are involved in the synthesis of the siderophore rhizobactin 1021. It will be of interest to see if there is any interplay between the wide-ranging regulator RirA and the more specialised RhrA of S. meliloti.

We already know that RirA participates in one regulatory cascade, repressing the transcription of rpoI, whose {sigma} factor product is, in turn, involved in the transcription of the vbsGSO and vbsADL genes that are required for siderophore biosynthesis (Carter et al., 2002; Yeoman et al., 2003). It will be interesting to discover how many of the large number of genes whose expression is affected by RirA involve a direct interaction between this regulator and their cognate promoter regions.


   ACKNOWLEDGEMENTS
 
This work was supported by the BBSRC. We are grateful to Marg Wexler for useful discussions. Access to the R. leguminosarum genomic sequence, produced by Julian Parkhill's group at the Sanger Centre, Hinxton, UK is gratefully acknowledged.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Allaway, D., Schofield, N. A., Leonard, M. E., Gilardoni, L., Finan, T. M. & Poole, P. S. (2001). Use of differential fluorescence induction and optical trapping to isolate environmentally induced genes. Environ Microbiol 3, 397–406.[CrossRef][Medline]

Andrews, S. C., Robinson, A. K. & Rodriguez-Quinones, F. (2003). Bacterial iron homeostasis. FEMS Microbiol Rev 27, 215–237.[CrossRef][Medline]

Beringer, J. E. (1974). R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84, 188–198.[Medline]

Carter, R. A., Worsley, P. S., Sawers, G. & 7 other authors (2002). The vbs genes that direct synthesis of the siderophore vicibactin in Rhizobium leguminosarum: their expression in other genera requires ECF {sigma} factor RpoI. Mol Microbiol 44, 1153–1166.[CrossRef][Medline]

Chao, T.-C., Becker, A., Buhrmester, J., Pühler, A. & Weidner, S. (2004). The Sinorhizobium meliloti fur gene regulates with dependence of Mn(II) the transcription of the sitABCD operon encoding a metal type transporter. J Bacteriol 186, 3609–3620.[Abstract/Free Full Text]

Diaz-Mireles, E., Wexler, M., Sawers, G., Bellini, D., Todd, J. D. & Johnston, A. W. B. (2004). The Fur-like protein Mur of Rhizobium leguminosarum is a Mn(2+)-responsive transcriptional regulator. Microbiology 150, 1447–1456.[CrossRef][Medline]

Johnston, A. W. B., Yeoman, K. H. & Wexler, M. (2001). Metals and the rhizobial-legume symbiosis – uptake, utilization and signalling. Adv Microbial Physiol 45, 113–156.[Medline]

Keon, R. G., Fu, R. & Voordouw, G. (1997). Deletion of two downstream genes alters expression of the hmc operon of Desulfovibrio vulgaris subsp. vulgaris Hildenborough. Arch Microbiol 167, 376–383.[CrossRef][Medline]

Kolade, O. O., Bellini, P., Wexler, M., Johnston, A. W. B., Grossmann, J. G. & Hemmings, A. M. (2002). Structural studies of the Fur protein from Rhizobium leguminosarum. Biochem Soc Trans 30, 771–774.[Medline]

Lawrence, J. G. & Hendrickson, H. (2003). Lateral gene transfer: when will adolescence end? Mol Microbiol 50, 739–749.[CrossRef][Medline]

Lynch, D., O'Brien, J., Welch, T., Clarke, P., Cuiv, P. O., Crosa, J. H. & O'Connell, M. (2001). Genetic organization of the region encoding regulation, biosynthesis, and transport of rhizobactin 1021, a siderophore produced by Sinorhizobium meliloti. J Bacteriol 183, 2576–2585.[Abstract/Free Full Text]

Masse, E. & Gottesman, S. (2002). A small RNA regulates the expression of genes involved in iron metabolism in Escherichia coli. Proc Natl Acad Sci U S A 99, 4620–4625.[Abstract/Free Full Text]

McHugh, J. P., Rodriguez-Quinones, F., Abdul-Tehrani, H., Svistunenko, D. A., Poole, R. K., Cooper, C. E. & Andrews, S. C. (2003). Global iron-dependent gene regulation in Escherichia coli: a new mechanism for iron homeostasis. J Biol Chem 278, 29478–29486.[Abstract/Free Full Text]

Messing, J., Crea, R. & Seeburg, P. H. (1983). A system for shotgun DNA sequencing. Nucleic Acids Res 9, 309–314.

Outten, F. W., Djaman, O. & Storz, G. (2004). A suf operon requirement for Fe-S cluster assembly during iron starvation in Escherichia coli. Mol Microbiol 52, 861–872.[CrossRef][Medline]

Pohl, E., Haller, J. C., Mijovilovich, A., Meyer-Klaucke, W., Garman, E. & Vasil, M. L. (2003). Architecture of a protein central to iron homeostasis: crystal structure and spectroscopic analysis of the ferric uptake regulator. Mol Microbiol 47, 903–915.[CrossRef][Medline]

Sawers, G. & Böck, A. (1989). Novel transcriptional control of the pyruvate formate-lyase gene: upstream regulatory sequences and multiple promoters regulate anaerobic expression. J Bacteriol 171, 2485–2498.[Medline]

Schwartz, C. J., Giel, J. L., Patschkowski, T., Luther, C., Ruzicka, F. J., Beinert, H. & Kiley, P. J. (2001). IscR, an Fe-S cluster-containing transcription factor, represses expression of Escherichia coli genes encoding Fe-S cluster assembly proteins. Proc Natl Acad Sci U S A 18, 14895–14900.[CrossRef]

Spaink, H. P., Okker, R. J. H., Wijffelman, C. A., Pees, E. & Lugtenberg, B. J. J. (1987). Promoters in the nodulation region of the Rhizobium leguminosarum symbiotic plasmid pRL1JI. Plant Mol Biol 9, 27–39.

Stevens, J. B., Carter, R. A., Hussain, H., Carson, K. C., Dilworth, M. J. & Johnston, A. W. B. (1999). The fhu genes of Rhizobium leguminosarum, specifying siderophore uptake proteins: fhuDCB are adjacent to a pseudogene version of fhuA. Microbiology 145, 593–601.[Medline]

Todd, J. D., Wexler, M., Sawers, G., Yeoman, K. H., Poole, P. S. & Johnston, A. W. B. (2002). RirA, an iron-responsive regulator in the symbiotic bacterium Rhizobium leguminosarum. Microbiology 148, 4059–4071.[Medline]

Toledano, M. B., Kullik, I., Tringh, F., Baird, P. T., Schneider, T. D. & Storz, G. (1994). Redox-dependent shift of OxyR-DNA contacts along an extended DNA-binding site: a mechanism for differential promoter selection. Cell 78, 897–909.[Medline]

Wexler, M., Yeoman, K. H., Stevens, J. B., de Luca, N. G., Sawers, G. & Johnston, A. W. B. (2001). The Rhizobium leguminosarum tonB gene is required for the uptake of siderophore and haem as sources of iron. Mol Microbiol 41, 801–816.[CrossRef][Medline]

Wexler, M., Todd, J. D., Kolade, O., Bellini, D., Hemmings, A. M., Sawers, G. & Johnston, A. W. B. (2003). Fur is not the global regulator of iron uptake genes in Rhizobium leguminosarum. Microbiology 149, 1357–1365.[CrossRef][Medline]

Yeoman, K. H., May, A. G., de Luca, N. G., Stuckey, D. B. & Johnston, A. W. B. (1999). A putative ECF sigma factor gene, rpoI, regulates siderophore production in Rhizobium leguminosarum. Mol Plant–Microbe Interact 12, 994–999.[Medline]

Yeoman, K. H., Wisniewski-Dye, F., Timony, C., Stevens, J. B., deLuca, N. G., Downie, J. A. & Johnston, A. W. B. (2000). Analysis of the Rhizobium leguminosarum siderophore-uptake gene fhuA: differential expression in free-living bacteria and nitrogen-fixing bacteroids and distribution of an fhuA pseudogene in different strains. Microbiology 146, 829–837.[Medline]

Yeoman, K. H., Mitelheiser, S., Sawers, G. & Johnston, A. W. B. (2003). The ECF sigma factor RpoI of R. leguminosarum initiates transcription of the vbsGSO and vbsADL siderophore biosynthetic genes in vitro. FEMS Microbiol Lett 223, 239–244.[CrossRef][Medline]

Received 18 June 2004; revised 30 August 2004; accepted 31 August 2004.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Yeoman, K. H.
Articles by Johnston, A. W. B.
Articles citing this Article
PubMed
PubMed Citation
Articles by Yeoman, K. H.
Articles by Johnston, A. W. B.
Agricola
Articles by Yeoman, K. H.
Articles by Johnston, A. W. B.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.