Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
Microbiology. Author manuscript; available in PMC Sep 3, 2008.
Published in final edited form as:
PMCID: PMC2528290

Swarming-coupled expression of the Proteus mirabilis hpmBA haemolysin operon


The HpmA haemolysin toxin of Proteus mirabilis is encoded by the hpmBA locus and its production is upregulated co-ordinately with the synthesis and assembly of flagella during differentiation into hyperflagellated swarm cells. Primer extension identified a σ70 promoter upstream of hpmB that was upregulated during swarming. Northern blotting indicated that this promoter region was also required for concomitant transcription of the immediately distal hpmA gene, and that the unstable hpmBA transcript generated a stable hpmA mRNA and an unstable hpmB mRNA. Transcriptional luxAB fusions to the DNA regions 5′ of the hpmB and hpmA genes confirmed that hpmB σ70 promoter activity increased in swarm cells, and that there was no independent hpmA promoter. Increased transcription of the hpmBA operon in swarm cells was dependent upon a 125 bp sequence 5′ of the σ70 promoter −35 hexamer. This sequence spans multiple putative binding sites for the leucine-responsive regulatory protein (Lrp), and band-shift assays with purified Lrp confirmed the presence of at least two such sites. The influence on hpmBA expression of the key swarming positive regulators FlhD2C2 (encoded by the flagellar master operon), Lrp, and the membrane-located upregulator of the master operon, UmoB, was examined. Overexpression of each of these regulators moderately increased hpmBA transcription in wild-type P. mirabilis, and the hpmBA operon was not expressed in any of the flhDC, lrp or umoB mutants. Expression in the mutants was not recovered by cross-complementation, i.e. by overexpression of FlhD2C2, Lrp or UmoB. Expression of the zapA protease virulence gene, which like hpmBA is also upregulated in swarm cells, did not require Lrp, but like flhDC it was upregulated by UmoB. The results indicate intersecting pathways of control linking virulence gene expression and swarm cell differentiation.

Keywords: flagella, LRP, toxin, virulence gene expression


Proteus mirabilis undergoes swarming, multicellular behaviour in which differentiated long aseptate hyperflagellated swarm cells migrate over surfaces (Fraser et al., 2000). During swarming, P. mirabilis co-ordinately upregulates expression of the flagellar gene hierarchy and production of several virulence factors (Allison et al., 1992; Rozalski et al., 1997), including the haemolysin HpmA (Welch, 1987; Uphoff & Welch, 1990; Senior & Hughes, 1987), urease (Mobley et al., 1995), and Zap protease (Walker et al., 1999). In particular, HpmA production is severely reduced in non-swarming flagellar gene mutants (Gygi et al., 1995, 1997), suggesting that biogenesis of flagella and expression of the hpmBA haemolysin operon are coupled.

Several transcriptional regulators of swarming differentiation have been identified, including the FlhD2C2 activator complex encoded by the flagellar hierarchy master operon flhDC (Furness et al., 1997; Claret & Hughes, 2000a, b), the leucine-responsive regulatory protein (Lrp) (Hay et al., 1997), and the cell envelope Umo proteins, especially UmoB (Dufour et al., 1998). It seems that these proteins might act as part of a regulatory network that integrates a number of stimuli and co-regulates virulence and flagellar gene expression (Fraser & Hughes, 1999). While signals that govern expression of the flagellar gene hierarchy are assumed to funnel through flhDC, this need not be the case for the co-regulated virulence genes. In particular, Lrp appears to regulate both flhDC and hpmBA, as a non-swarming lrp::Tn5 mutant is non-haemolytic even when its hyperflagellation and swarming is recovered by overexpression of flhDC (Hay et al., 1997). In this report, we characterize hpmBA transcription, and investigate the influence of the swarm regulators Lrp, FlhD2C2 and UmoB upon expression of hpmBA and also zapA.


Bacterial strains, recombinant DNA techniques and sequencing

The Proteus mirabilis lrp::Tn5 (Hay et al., 1997) and umoBΩ (Dufour et al., 1998) strains are derivatives of wild-type P. mirabilis U6450 (Allison & Hughes, 1991). Bacteria were grown at 37 °C in Luria-Bertani (LB) broth or on LB agar supplemented with ampicillin (100 μg ml−1), chloramphenicol (40 μg ml−1) or spectinomycin (50 μgml−1), when necessary. Plasmids were introduced by electroporation. Swarming differentiation on seeding plates (Gygi et al., 1995) was initiated by spreading 200 μl stationary-phase LB cultures (approx. 5 × 10) cells ml− 1) onto 8 cm diameter LB agar plates. Plasmid DNA was manipulated by standard techniques (Ausubel et al., 1988; Sambrook et al., 1989) and maintained in Escherichia coli XL-Blue (recA1[FlacIqZΔM15Tn10]; Stratagene). Plasmids are listed in Table 1. DNA sequence was obtained from double-stranded pBluescript II templates by the chain-terminating method (Sanger et al., 1977), using T7 DNA polymerase (Pharmacia), and was analysed using GCG software (Devereux et al., 1984). Database searches were performed at the NCBI using the blast network service (Altschul et al., 1990).

Table 1

Northern blot analysis

Total cellular RNA was isolated using the hot phenol method (Koronakis & Hughes, 1988). Formamide/formaldehyde-denatured RNA (10 μg) was resolved by electrophoresis through 1·2% agarose formaldehyde gels (Sambrook et al., 1989) and transferred onto nitrocellulose membranes (Hybond-C; Amersham) by vacuum blotting (Pharmacia LKB). Duplicate blots were stained with methylene blue to ensure that rRNA levels in the samples were equivalent (not shown). Probes were a 1·5 kb XhoI–EcoRV fragment internal to hpmB, a 1 kb ClaI fragment internal to hpmA, a 0·8 kb AvaI fragment internal to flhDC, a 0·9 kb BamHI-HindIII fragment internal to lrp and a 1·1 kb XbaI–XhoI fragment internal to zapA. Probes were labelled with [α-32P]dATP (Amersham) by random priming (Feinberg & Vogelstein, 1984), and hybridization and detection were carried out as described by Gygi et al. (1995).

Promoter mapping by primer extension

Oligonucleotides A1 (5′-CGCCCGGAGGGTGAAAGTTTAAG-3′, complementary to nucleotides 15 to 38 of hpmA), B1 (5′-CAGCTTAATAGTGTTAATAAAAC-3′, complementary to nucleotides 16 to 38 of hpmB), and B2 (5′-CCAGTTACTCTTTAATTG-3′, complementary to nucleotides −304 to −321 5′ of hpmB) were 5′ end-labelled with [γ-32P]ATP (Amersham) by T4 polynucleotide kinase. Primer extension reactions were performed as described by Dufour et al. (1998).

Luciferase reporter assays

Transcriptional fusion assays were performed as previously described (Gygi et al., 1995). P. mirabilis wild-type and the flhDCΩ strains carrying derivatives of the reporter plasmid pQF120 (Ronald et al., 1990) containing transcriptional fusions of putative hpmBA promoter regions to the luxAB genes of Vibrio fischeri were monitored in a Bio-Orbit 1253 single-tube luminometer in the presence of 0·02% (v/v) dodecanal (Aldrich). The baseline luciferase activities of wild-type cells carrying the pQF120 parental vector were subtracted from the activities produced by cells containing hpmBA–luxAB fusions. Each data collection point was replicated five times and the mean and standard error were calculated.

Purification of N-terminally His-tagged Lrp protein

The lrp gene was PCR-amplified with Pfu Polymerase (Promega) from P. mirabilis U6450 chromosomal DNA using the mutagenizing oligonucleotide primers Lrp1 (5′-GAAAGAGAGTATACATATGATTGATAAT-3′; NdeI site) and Lrp2 (5′-GCACTAACCTACCAGCAGGCTTAGCGTG-3′; BamHI site). The PCR fragment was inserted into pET15b (Novagen). N-His-Lrp was produced in E. coli BL21(DE3) (Studier & Moffart, 1986), by induction at mid-exponential phase with 1 mM IPTG for 3 h. Protein purification was carried out under non-denaturing conditions using His-Bind Quick 900 columns (Novagen) as described by the manufacturer.

Gel retardation assays

DNA fragments were PCR-amplified with Pfu Polymerase (Promega) using pGF74 as a template. The downstream oligonucleotide primer B1 (5′-CAGCTTTAATAGTGTTAATAAAA-3′) is internal to the hpmB gene. The upstream oligonucleotide primers hpm AccI (5′-GATTTTATTTGTCGACAACGCTTATTCTC-3′) and hpm ClaI (5′-CGTTTATTGATCGATAATACTAAAAAAAG-3′) were used to amplify the 239 bp fragment spanning nucleotides 473 to 711 and the 183 bp fragment from nucleotide 529 to 711 (see Fig. 4), respectively. PCR products were digested with XhoI, dephosphorylated with calf intestinal phosphatase and labelled using T4 polynucleotide kinase and [γ-32P]dATP. The control 157 bp labelled DNA fragment was obtained as described by Claret & Hughes (2000b). The 183 bp, 239 bp and 157 bp fragments were purified on a polyacrylamide gel as described by Sambrook et al. (1989). Labelled fragments (0·02 pmol) in 10 μl buffer A (20 mM Tris/HCl pH 7·4, 50 mM NaCl, 0·1 mM EDTA, 0·1 mM DTT, 20%, v/v, glycerol, 100 μg BSA ml−1) were incubated with 0-30 ng purified N-His-Lrp. After 20 min incubation at room temperature, samples were run on a 6% polyacrylamide gel (2 h, 130 V, room temperature) containing 0·5× TBE; the gel was dried and autoradiographed.

Fig. 4
DNA sequence of the mutT-hpmB intergenic region showing the hpmB transcription start site, the putative σ70 and σ28 promoter sequences, and four putative Lrp-binding sites (Lrp 1-4).

Generation of the P. mirabilis flhDC null mutant

A 250 bp fragment between nucleotides 62 and 312 (HinP1I and HincII sites, respectively) of the 351 bp P. mirabilis flhD gene was removed and replaced by the 2 kbp spectinomycin resistance omega (Ω) cassette (Fellay et al., 1987), creating the plasmid pRF10Ω. The approximately 4·6 kbp flhDCΩ locus of pRF10Ω was inserted into the suicide vector pGP704 (Miller & Mekalanos, 1988). Conjugal transfer of the suicide plasmid from E. coli SM10::λpir to P. mirabilis U6450, and isolation of flhDCΩ mutants, were carried out as described by Swihart & Welch (1990). Colony blots were probed with an approximately 1·4 kbp BamHI fragment of pGP704 and an approximately 2·0 kbp HindIII fragment of the Ω cassette.


Identification of a swarming-regulated σ70 promoter 5′ of hpmBA

DNA sequence analysis of the hpmBA locus has previously identified a putative σ70 promoter 78 bp 5′ of the proximal hpmB gene and a second weak match to the σ70 promoter consensus 104 bp 5′ of the distal hpmA gene, in the hpmB coding region (Uphoff & Welch, 1990). We identified a third potential promoter sequence (TAAA-N15-GTAGATAA) 233 bp 5′ of hpmB that is similar to the consensus sequence (TAAA-N15-GCCGATAA) recognized by the flagellar-gene-specific sigma factor, σ28 (Helmann & Chamberlin, 1987; Hellmann, 1991). To determine which of these putative promoters is active during swarming, the 5′ termini of the hpmB and hpmA mRNAs were mapped by primer extension using template RNA isolated from wild-type P. mirabilis vegetative cells and from differentiated swarm cells. Vegetative and differentiated cells were recovered as previously from seeded agar plates after 1·5 and 4 h, respectively (Gygi et al., 1995).

A single extension product terminating 70 bp 5′ of hpmB and 8 bp 3′ of the −10 hexamer of the putative hpmB σ70 promoter was detected in the swarm-cell RNA but not in that from the vegetative cells (Fig. 1a). No extension product was evident immediately 3′ of the putative σ28 promoter, and this was verified by carrying out primer extension with an oligonucleotide primer complementary to sequence 5′ of the σ70 consensus (not shown). Several swarm-cell-specific primer extension products terminated in the region 5′ of hpmA (Fig. 1b), but none was compatible with transcription initiation from the putative hpmA σ70 promoter. These data indicate that the σ70 promoter 5′ of the hpmB gene is the only promoter active and upregulated during swarming. They suggest that the hpmA mRNA might arise from processing of a bicistronic hpmBA transcript.

Fig. 1
Primer extension mapping of mRNA termini 5′ of (a) hpmB and (b) hpmA. Wild-type vegetative (v) cells and differentiated swarm (s) cells were isolated from LB agar plates at 1·5 and 4 h after seeding, respectively. T, G, C and A indicate ...

Swarming-coupled transcription of hpmA requires the hpmB promoter region

The primer extension data indicate that hpmBA might be transcribed as an operon. To establish whether the hpmB promoter region is indeed essential for the high-level transcription of hpmA during swarming, hpmA RNA was assayed following expression of the hpmBA operon with and without the hpmB 5′ region. The plasmids pGF74, carrying hpmBA preceded by 725 bp encompassing the hpmB promoter region, and pGF72, carrying hpmA preceded by a 5′ truncated hpmB allele lacking the hpmB promoter region and nucleotides 1 to 612 of hpmB, were introduced into wild-type P. mirabilis cells in parallel with the parental pBluescript vector. Total cellular RNA was prepared from differentiated swarm cells carrying the plasmids and Northern blots were probed with a ClaI fragment internal to hpmA (Fig. 2). In cells carrying pGF74, the amount of hpmA mRNA was about 20-fold higher than that in cells carrying the pBluescript vector control, i.e. in which hpmA mRNA is produced by transcription of the chromosomal hpmBA locus. The amount of hpmA mRNA in cells carrying pGF72, encoding the 5′-truncated hpmBA determinant, was comparable to that in the control cells. There was therefore no swarm-regulated transcription of the plasmid-encoded hpmA in the absence of a cis hpmB promoter region. These data further support the hypothesis that hpmBA is transcribed as an operon from a swarm-regulated promoter 5′ of hpmB.

Fig. 2
Northern hybridization of a hpmA probe to total cellular RNA isolated from wild-type differentiated swarm cells carrying pBluescript (−), pGF74 (hpmBA) or pGF72 (ΔphpmBA). A 281- to 6583-base RNA ladder (Promega) was used to determine ...

The hpmBA transcript yields a stable hpmA mRNA

Previous Northern blot analysis of hpmA mRNA detected a major 5·2 kb transcript and a minor 6·9 kb transcript, both upregulated during swarming differentiation (Allison et al., 1992). The primer extension data presented here (Fig. 1) suggest that hpmB mRNA is also abundant, as expected from expression of an operon. To assess the size and abundance of the hpmBA transcript, total cellular RNA was prepared from wild-type P. mirabilis cells isolated from LB agar plates after 2, 3 and 4 h seeding differentiation, and Northern blot analysis was carried out using probes internal to hpmB and hpmA (Fig. 3). After a 4 h exposure of the fluorogram, an approximately 5 kb hpmA transcript was detected that was strongly induced during differentiation (Fig. 3, left-hand panel). However, no hpmBA or hpmB transcripts were visible. Only after a 240 h exposure of the fluorogram was an approximately 6·9 kb hpmBA transcript detected by the hpmB probe (Fig. 3, right-hand panel). This transcript was, as expected, also recognized by the hpmA probe (not shown). The hpmBA transcript was visible in the 2 h and 3 h samples, but was barely visible in the 4 h sample isolated from differentiated swarm cells. Degradation products of the hpmBA and hpmB transcripts, seen only after long exposure, were detected by the hpmB probe as a heterogeneous population migrating between about 0·2 and 6·9 kb, although most of the products appeared to be about 1·7 kb and smaller. This suggests that the hpmBA transcript might be cleaved to produce a stable approximately 5 kb hpmA transcript and a highly unstable approximately 1·7 kb hpmB transcript. Extensive attempts were made to generate stable hpmB’A and hpmBA’ transcripts by deleting the hpmB–hpmA intergenic region and various segments of the hpmBA coding region such that predicted RNase cleavage sites were removed. We were unable to isolate any stable mRNA containing hpmB (not shown), suggesting that the degradation is not simple and might involve endonucleolytic cleavage of the hpmB mRNA at multiple sites.

Fig. 3
Northern hybridization of hpmA and hpmB probes to total cellular RNA isolated from differentiating wild-type cells collected 2, 3, and 4 h after seeding onto LB agar. The positions of the hpmA and hpmBA transcripts are indicated and hpmB degradation products ...

Swarming-coupled hpmBA transcription requires regulatory sequences upstream of the hpmB σ70 promoter

To identify elements that might be involved in the transcriptional regulation of the P. mirabilis hpmBA operon during swarming, the DNA sequence of the 1195 bp region 5′ of hpmB was determined (GenBank accession number AJ250100). A 384 nt ORF was identified 697 bp 5′ of the hpmB initiation codon. The ORF encodes a putative 128 aa polypeptide with 85% and 48% identity, respectively, to the MutT nucleoside triphosphatases of E. coli and Proteus vulgaris (Akiyama et al., 1987; Kamath & Yanofsky, 1993). Several putative Lrp-binding sequences (Cui et al., 1995) were evident in the mutT–hpmB intergenic region (Fig. 4), which seems significant as a non-swarming P. mirabilis mutant unable to hyper-produce haemolysin or flagella has been reported to have a lesion in the lrp gene (Hay et al., 1997). Each of the putative Lrp-binding sequences 5′ of hpmB has four mismatches to the 15 bp consensus derived by Cui et al. (1995), i.e. approximately 73% identity.

To assess the elements required for swarming-coupled expression of hpmBA, transcriptional fusions of regions 5′ of hpmB and of hpmA were made to the luxAB luciferase genes of Vibrio fischeri. Two derivatives of the reporter plasmid pQF120 (Ronald et al., 1990) were initially constructed (Fig. 5a), one with the entire 712 bp non-coding region 5′ of hpmB (pLUX712), the second (pLUX125) with a truncated 5′ 125 bp region containing the hpmB σ70 promoter but lacking the σ28 consensus and three of the four putative Lrp-binding sites. These plasmids were introduced into wild-type P. mirabilis and its derivative non-swarming flhDC null mutant, which was constructed by deleting a 250 bp internal fragment of flhD and replacing it with the spectinomycin resistance omega (Ω) cassette (see Methods).

Fig. 5
(a) Transcriptional fusions of the non-coding region 5′ of hpmBA to luxAB. A 712 bp region 5′ of hpmB encompassing putative promoter sequences (black boxes, σ70 and σ28) and four putative Lrp-binding sites (arrows, 1-4) ...

Luciferase activity was monitored over the swarm cell differentiation cycle. In wild-type cells carrying pLUX712, activity was strongly upregulated, with peak activity in differentiated cells being about ninefold higher than in swarm cells carrying pLUX125 (Fig. 5a, b). There seemed to be no substantial plasmid copy number titration of positive or negative regulatory factors. Wild-type P. mirabilis carrying pLUX712 retained wild-type haemolytic activity over the differentiation cycle, suggesting that the presence of multiple copies of the hpmB upstream region did not disrupt regulation of the chromosomal hpmBA. In the flhDCΩ mutant, which grew like the wild-type but does not differentiate into swarm cells, peak luciferase activities in cells carrying either pLUX712 or pLUX125 were approximately 10-fold lower than that observed in wild-type cells carrying pLUX712 (Fig. 5a). The peak activity produced by pLUX125 in the wild-type was not significantly different to those of the flhDC mutant carrying either plasmid, and may reflect transcription from the hpmB promoter in the absence of a response to differentiation signals. The data suggest that regulatory sequences 5′ of the hpmB promoter are required for full transcriptional upregulation in response to swarming signals.

Three derivative reporter fusions, pLUX415, pLUX227 and pLUX183, contained truncations of the 712 bp hpmB upstream region (Fig. 5a). The largest, pLUX415, contained a 5′ truncation to 415 bp, retaining the putative σ28 consensus sequence, the four putative-Lrp-binding sites and the hpmB σ70 promoter. In wild-type cells it produced a peak luciferase activity comparable to pLUX712 (Fig. 5a), suggesting that the sequences required for full transcriptional upregulation during swarming are within this 415 bp region. Further 5′ truncation to 227 bp removed the potential σ28 −35 consensus and the putative Lrp1 site to generate pLUX227, which in the wild-type generated a peak luciferase activity about 50% of that produced by pLUX425 and pLUX712 (Fig. 5a).

Nevertheless, pLUX227 peak activity in wild-type was about sixfold higher than that in the flhDC mutant, indicating that the hpmB promoter on pLUX227 was still substantially upregulated in response to differentiation. Further truncation of the hpmB promoter region to 183 bp in pLUX183 removed the putative Lrp1 and Lrp2 sites and caused severe reduction in the promoter’s activity, such that wild-type and flhDC cells carrying pLUX183 produced comparably low peak luciferase activities (Fig. 5a). In addition to these hpmB σ70 promoter fusions, a luxAB transcriptional fusion was constructed containing a 413 bp region encompassing the hpmB σ28 consensus but not the σ70 promoter (pLUXD413). Peak luciferase activities in the wild-type and the flhDC mutant carrying pLUXD413 were comparably low (Fig. 5a), in agreement with the primer extension data (Fig. 1a, b), which indicated that the putative σ28 consensus is not an active promoter. A second transcriptional reporter plasmid containing the 230 bp region immediately 5′ of hpmA (pLUXA230, not shown) was also constructed; it generated very low activities in wild-type and flhDC mutant cells, again indicating that there is probably not an independently regulated promoter immediately 5′ of hpmA

Multiple Lrp binding to the hpmB promoter region

The fusion analysis suggested that sequences 5′ of the hpmB σ70 promoter are needed for the transcriptional upregulation of the hpmBA operon during swarming. In particular, a region spanning two putative Lrp-binding sites (Lrp1 and Lrp2, Fig. 4) appears to be essential. To establish if Lrp does indeed bind to this hpmB promoter region, interaction was assessed in band-shift experiments. Soluble N-terminally His-tagged Lrp protein was purified under non-denaturing conditions on a nickel affinity cartridge. It was incubated with a 183 bp DNA region spanning nucleotides 529 to 711 (Fig. 4) encompassing the putative Lrp sites 3 and 4, and with a 239 bp DNA fragment containing nucleotides 473 to 711 (Fig. 4) encompassing the putative Lrp sites 2, 3 and 4. Incubation was carried out in the presence of a 1000-fold excess of poly(dI-dC) non-specific competitor DNA. We estimated the concentration of His-Lrp needed for half saturation of the labelled DNA in repeated band-shift assays. For the smaller fragment (−123 to +60) this was 17 nM, for the larger fragment (−179 to +60) 5 nM. Fig. 6 shows that migration of the shorter fragment was retarded by increasing amounts of His-Lrp, indicating the formation of a nucleoprotein complex, while the larger fragment generated two retarded bands, indicating formation of two different complexes. Interactions of His-Lrp with a still larger fragment, the 416 bp region spanning nucleotides 297 to 711 (containing all four putative Lrp sites shown in Fig. 4), indicated the formation of a complex that, due to its large size, remained in the wells of the polyacrylamide gel after electrophoresis (data not shown). No binding was seen when a control shift assay was performed with a 157 bp DNA fragment carrying the upstream promoter region of the P. mirabilis flhB gene (Claret & Hughes, 2000b). These data indicate that His-Lrp binds specifically to both the 183 bp and 239 bp fragments, but there may be multiple interactions in each case; the data show that the 239 bp fragment contains at least one more site than the 183 bp fragment.

Fig. 6
Band-shift analysis of Lrp binding to the regulatory region 5′ of hpmBA. End-labelled DNA fragments of a 183 bp and a 239 bp region 5′ of hpmB encompassing Lrp sites 3 and 4, and 2, 3 and 4, respectively (see Fig. 3, nucleotides 529 to ...

Multifactorial regulation of virulence factors by FlhD2C2, Lrp and UmoB

The cytosolic DNA-binding proteins Lrp and FlhD2C2, and the UmoB cell envelope protein, have been identified as strong positive regulators of swarming differentiation (Hay et al., 1997; Furness et al., 1997; Claret & Hughes, 2000a; Dufour et al., 1998; Fraser & Hughes, 1999). Their influence on the swarming-coupled transcription of hpmBA was investigated by Northern blot analysis of mRNA isolated from the P. mirabilis wild-type, and null mutants of lrp (Hay et al., 1997), umoB (Dufour et al., 1998) and flhDC (see Methods). The hpmA transcript was not detected in any of the non-swarming lrp, flhDC and umoB mutants (Fig. 7). It has previously been shown that artificial overexpression of swarm regulators can influence aspects of differentiation; for example, transcription of the flhDC flagellar master operon is upregulated by the overexpression of UmoB (Dufour et al., 1998), and the swarming and cell elongation defects of the lrp mutant are recovered by overexpression of FlhD2C2 (Hay et al., 1997). To investigate the possibility that hpmA transcription could be increased by similar overexpression of FlhD2C2, Lrp or UmoB, the recombinant expression plasmids pBADflhDC (Hay et al., 1997), pBADlrp and pBSumoB (Dufour et al., 1998) were introduced in parallel into the wild-type strain and the flhDC, umoB and lrp mutants. Northern blot analysis was again carried out to hpmA and, as controls, to flhDC and lrp. Overproduction of FlhD2C2, Lrp and UmoB in the wild-type increased hpmA and flhDC mRNA levels by three- to fivefold (Fig. 7), and lrp transcription was also slightly increased (not shown). However, transcription of hpmA was not recovered in the flhDC, umoB and lrp mutants, suggesting that regulation of haemolysin expression during swarming is controlled by a network of regulators and not by a single pathway. Transcription of flhDC was partially recovered in the lrp mutant overexpressing UmoB (Fig. 7).

Fig. 7
Northern hybridization of hpmA (left column), zapA (middle column) and flhDC (right column) probes to total cellular RNA isolated from differentiated wild-type swarm cells and the lrp::Tn5, umoBΩ, and flhDCΩ mutants carrying pBAD (−), ...

In addition to the upregulation of hpmBA expression during swarming, several other virulence factors are hyper-produced, e.g. the Zap protease (Wassif et al., 1995). To discover if hpmBA and the zapA protease gene were similarly regulated, parallel Northern blots using a probe internal to zapA were also carried out (Fig. 7). As with hpmA, zapA mRNA could not be detected in the flhDC and umoB regulator mutants; however, zapA transcription was only slightly reduced in the lrp mutant. Transcription of zapA could be increased in the flhDC and lrp mutants by overexpression of UmoB, suggesting that zapA might be directly regulated by UmoB.


The production of several virulence proteins is upregulated during the Proteus swarm cycle. The synthesis and secretion of HpmA haemolysin is modulated in parallel with expression of the flagellar gene hierarchy, with haemolytic activity approximately 20-fold higher in hyper-flagellated swarm cells than that in vegetative cells (Allison et al., 1992). Coordinate regulation of virulence gene expression and flagellar biogenesis has been observed in other enterobacteria. In Salmonella enterica serovar Typhimurium, expression of the hilA activator of invasion gene transcription is upregulated by FliZ, a poorly characterized flagellar protein that positively regulates class 2 flagellar genes (Lucas et al., 2000; Iyoda et al., 2001). In Serratia liquefaciens and Yersinia enterocolitica, the genes encoding phospholipase are transcribed as part of the flagellar regulon from promoters recognized by the flagellar-specific sigma factor σ28 (Givskov et al., 1995; Schmiel et al., 2000). While a putative σ28 promoter sequence has been identified 5′ of the P. mirabilis hpmB gene, our primer extension analyses indicated that hpmB is not transcribed from this promoter, but rather from a σ70 promoter sequence 78 bp 5′ of hpmB that is substantially upregulated in differentiated swarm cells. It has been reported (Ide et al., 1999) that active σ28 promoters in S. typhimurium and E. coli feature an extended −35 motif that is not present in sequences that are otherwise homologous to the consensus described by Helmann & Chamberlin (1987). The revised consensus is TAAAGTTT-N11-GCCGATAA compared to the old consensus TAAA-N15-GCCGATAA. In agreement with our experimental analyses, the suggested σ28 promoter 5′ of hpmB does not have the extended −35 sequence. Although several transcript extension products were detected immediately 5′ of hpmA, none was compatible with initiation from the putative σ70 promoter sequence identified by Uphoff & Welch (1990). The distance between the most abundant extension product and the putative hpmA σ70 −10 hexamer was 12 bp, larger than the typical 5-9 bp. No promoter consensus sequences were found 5′ of any of the hpmA primer extension products, and it might be that these mRNA termini result from endonucleolytic cleavage of a bicistronic hpmBA operon transcript originating from the swarm-upregulated hpmB promoter. This view was strengthened by Northern blot analysis of hpmA mRNA isolated from cells carrying, in trans, either the full-length hpmBA determinant or a 5′ truncated hpmBA lacking the hpmB promoter region, which indicated that there was no independently regulated promoter immediately 5′ of hpmA.

During differentiation, the steady-state levels of hpmA mRNA were considerably higher than the levels of hpmBA mRNA. This was somewhat unexpected, as the primer extension data suggest that the hpmB transcript is abundant. However, abundance of the hpmB cDNA extension product does not necessarily reflect the stability of the hpmBA messenger, as a successful primer extension would only require hpmB mRNA fragments containing the first 200 bases. When considered together, the primer extension and Northern blot data indicate that an unstable full-length hpmBA transcript is processed to produce a stable hpmA mRNA and an unstable hpmB mRNA. Similar processing of polycistronic transcripts has been widely observed: e.g. the E. coli ftsZA and malEFG transcripts, the Salmonella histidine transport operon, and the extensively studied Rhodobacter capsulatus puf operon (Cam et al., 1996; Newbury et al., 1987; Stern et al., 1988; Klug, 1993). In the case of the R. capsulatus pufQBALMX transcript, which encodes proteins involved in photosynthesis, processing of the mRNA produces transcript segments with different stabilities and this subsequently influences translation and the stoichiometry of the proteins in the photosynthetic complexes (Klug, 1993). Similarly, processing of the P. mirabilis hpmBA transcript might be one of the mechanisms that regulates the ratio of the HpmA haemolysin and its cognate exporter/activator protein, HpmB. These proteins are highly similar to ShlA and ShlB of Serratia marcescens. ShlB has been shown to form pores in artificial lipid bilayer membranes, and it is thought to form a channel in the bacterial outer membrane through which the ShlA haemolysin is exported (Konninger et al., 1999). It seems unlikely that the HpmB (ShlB) exporter and HpmA (ShlA) haemolysin would be required in equimolar amounts, and limiting the production of the pore-forming outer-membrane exporter while producing large amounts of exported haemolysin might bring maximal benefits to the bacterial cell.

As hpmBA is most likely transcribed as an operon, we investigated the role of sequences upstream of hpmB in the modulation of its transcription during the swarm cycle. Analysis of luxAB transcriptional fusions to the non-coding region upstream of hpmB showed that this 697 bp region can be truncated to 415 bp without affecting maximal promoter activity in swarm cells. This 415 bp region encompasses four putative consensus sequences for Lrp (Fig. 4, Lrp1-Lrp4). Further truncation of the hpmB upstream region, removing the Lrp1 site, reduced maximal promoter activity to 50%, and additional removal of the Lrp2 site abolished upregulation of promoter activity during swarming. This suggests that while the putative Lrp1 site enhances, but is not essential for, upregulation of the hpmB σ70 promoter, the region spanning the Lrp2 site is absolutely required. Binding of Lrp to the hpmB upstream region was confirmed by in vitro band-shift experiments, and the data indicate that there are at least two Lrp-binding sites in the 239 bp region encompassing the putative Lrp2, Lrp3 and Lrp4 consensus sequences, and at least one binding site in the 183 bp region that includes the Lrp3 and Lrp4 sequences. There may be further Lrp-binding sites in the larger 415 bp region which carries all four of the putative Lrp consensus sequences, as this fragment formed a very large nucleoprotein complex that did not migrate into the native polyacrylamide gel. Previously characterized examples of Lrp binding to promoter regions (Gazeau et al., 1994; Cui et al., 1995; Rhee et al., 1996) show that Lrp often binds at multiple sites, and that binding sometimes shows a degree of cooperativity. Therefore, we cannot exclude the possibility that the 183 bp hpmB upstream fragment contains more than one Lrp-binding site, and the 239 bp region contains more than two sites. Our results suggest that Lrp acts directly to activate the transcription of this operon. As the Lrp-binding sites 2, 3 and 4 upstream of hpmB are within 30 bp of each other, it seems possible that they bind Lrp cooperatively.

Lrp is only one component of a regulatory network that also includes the FlhD2C2 and UmoB proteins, both of which have key roles in the control of flagellar biogenesis during swarming (Furness et al., 1997; Dufour et al., 1998). Less is known about their involvement in regulating the production of virulence factors such as the HpmA haemolysin. We investigated the influence of Lrp, FlhD2C2 and UmoB on the transcription of hpmBA during the swarm cycle and found that they were all essential, as the hpmA transcript was not detected in any of the regulator mutants. In wild-type P. mirabilis, the peak level of hpmA transcript could be moderately increased by overexpression of lrp, flhDC and umoB, and while hpmA expression could be recovered in the regulator mutants by direct complementation (not shown), it was not restored by ‘cross-complementation’ by the other regulators, e.g. by overexpression of flhDC in the lrp null mutant, which overcomes the cell elongation and swarming defects of this mutant but does not restore haemolytic activity (Hay et al., 1997). This suggests that while Lrp, FlhD2C2 and UmoB can positively influence transcription of hpmBA, there is not a single pathway that controls haemolysin expression during swarming. Rather, regulation of hpmBA appears to be complex and multifactorial. The present study and previously published data suggest that Lrp directly regulates transcription of hpmBA while FlhD2C2 and UmoB could be acting indirectly. In parallel with our studies of hpmBA expression, we investigated briefly the effects of Lrp, FlhD2C2 and UmoB on transcription of the zapA protease gene. Unlike the HpmA haemolysin, which is maximally produced in actively swarming cells, peak production of the Zap protease occurs slightly later in the swarm cycle, in differentiated cells that are about to enter the consolidation phase (Walker et al., 1999). We found that zapA regulation is markedly different to that of hpmBA, as Lrp is not required for zapA upregulation during swarming; instead, zapA appears to be primarily regulated by UmoB.

The data presented outline several components of hpmBA regulation, including differential mRNA stability, and indicate direct and indirect roles for several proteins in upregulating the single σ70 operon promoter during swarming differentiation. Taken with the initial characterization of zapA expression, the data strengthen a view of swarming as a major physiological shift, involving a complex network of regulation.


The GenBank accession number for the sequence determined in this work is AJ250100.


  • Akiyama M, Horiuchi T, Sekiguchi M. Molecular cloning and nucleotide sequence of the mutT mutator of Escherichia coli that causes A:T to C:G transversion. Mol Gen Genet. 1987;206:9–16. [PubMed]
  • Allison C, Hughes C. Closely linked genetic loci required for swarm cell differentiation and multicellular migration by Proteus mirabilis. Mol Microbiol. 1991;5:1975–1982. [PubMed]
  • Allison C, Lai H-C, Hughes C. Co-ordinate expression of virulence genes during swarm-cell differentiation and population migration of Proteus mirabilis. Mol Microbiol. 1992;6:1583–1591. [PubMed]
  • Altschul SF, Gish W, Miller W, Myers EW, Lipman D. Basic logical alignment search tool. J Mol Biol. 1990;215:403–410. [PubMed]
  • Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl L. Current Protocols in Molecular Biology. New York: Wiley; 1988.
  • Cam K, Rome G, Krisch HM, Bouche JP. RNase E processing of essential cell division genes mRNA in Escherichia coli. Nucleic Acids Res. 1996;24:3065–3070. [PMC free article] [PubMed]
  • Claret L, Hughes C. Rapid turnover of FlhD and FlhC, the flagellar regulon transcriptional activator proteins, during Proteus swarming. J Bacteriol. 2000a;182:833–836. [PMC free article] [PubMed]
  • Claret L, Hughes C. Functions of the subunits in the FlhD2C2 transcriptional master regulator of bacterial flagellum biogenesis and swarming. J Mol Biol. 2000b;303:467–478. [PubMed]
  • Cui Y, Wang Q, Stormo GD, Calvo JM. A consensus sequence for binding of Lrp to DNA. J Bacteriol. 1995;177:4872–4880. [PMC free article] [PubMed]
  • Devereux J, Haeberli P, Smithies O. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 1984;12:387–395. [PMC free article] [PubMed]
  • Dufour A, Furness RB, Hughes C. Novel genes that upregulate the Proteus mirabilis flhDC master operon controlling flagellar biogenesis and swarming. Mol Microbiol. 1998;29:741–751. [PubMed]
  • Feinberg AP, Vogelstein B. A technique for radio-labelling DNA restriction endonuclease fragments to high specific activity. Anal Biochem. 1984;132:6–13. [PubMed]
  • Fellay R, Frey J, Krisch H. Interposon mutagenesis of soil and water bacteria: a family of DNA fragments designed for in vitro insertional mutagenesis of gram-negative bacteria. Gene. 1987;52:147–154. [PubMed]
  • Fraser GM, Hughes C. Swarming motility. Curr Opin Microbiol. 1999;2:630–635. [PubMed]
  • Fraser GM, Furness RB, Hughes C. Swarming migration by Proteus and related bacteria. In: Brun YV, Shimkets LJ, editors. Prokaryotic Development. Washington, DC: American Society for Microbiology; 2000. pp. 381–401.
  • Furness RB, Fraser GM, Hay NA, Hughes C. Negative feedback from a Proteus class II flagellum export defect to the flhDC master operon controlling cell division and flagellum assembly. J Bacteriol. 1997;179:5585–5588. [PMC free article] [PubMed]
  • Gazeau M, Delort F, Fromant M, Dessen P, Blanquet S, Plateau P. Structure-function relationship of the Lrp-binding region upstream of lysU in Escherichia coli. J Mol Biol. 1994;241:378–389. [PubMed]
  • Givskov M, Eberl L, Christiansen G, Bendik MJ, Molin S. Induction of phospholipase and flagellar synthesis in Serratia liquefaciens is controlled by expression of the master operon flhD. Mol Microbiol. 1995;15:445–454. [PubMed]
  • Guzman LM, Belin D, Carson MJ, Beckwith J. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol. 1995;177:4121–4130. [PMC free article] [PubMed]
  • Gygi D, Bailey MJ, Allison C, Hughes C. Requirement for FlhA in flagella assembly and swarm cell differentiation by Proteus mirabilis. Mol Microbiol. 1995;15:761–769. [PubMed]
  • Gygi D, Fraser G, Dufour A, Hughes C. A motile but non-swarming mutant of Proteus mirabilis lacks FlgN, a facilitator of flagella filament assembly. Mol Microbiol. 1997;25:597–604. [PubMed]
  • Hay NA, Tipper DJ, Gygi D, Hughes C. A non-swarming mutant of Proteus mirabilis lacks the Lrp global transcriptional regulator. J Bacteriol. 1997;179:4741–4746. [PMC free article] [PubMed]
  • Helmann JD. Alternative sigma factors and the regulation of flagellar gene expression. Mol Microbiol. 1991;5:2875–2882. [PubMed]
  • Helmann JD, Chamberlin MJ. DNA sequence analysis suggests that expression of flagellar and chemotaxis genes in Escherichia coli and Salmonella typhimurium is controlled by an alternative sigma factor. Proc Natl Acad Sci U S A. 1987;84:6422–6444. [PMC free article] [PubMed]
  • Ide N, Ikebe T, Kutsukake K. Reevaluation of the promoter structure of the class 3 flagellar operons of Escherichia coli and Salmonella. Genes Genet Syst. 1999;74:113–116. [PubMed]
  • Iyoda S, Kamidoi T, Hirose K, Kutsukake K, Watanabe H. A flagellar gene fliZ regulates the expression of invasion genes and virulence phenotype in Salmonella enterica serovar Typhimurium. Microb Pathog. 2001;30:81–90. [PubMed]
  • Kamath AV, Yanofsky C. Sequence and characterisation of mutT from Proteus vulgaris. Gene. 1993;134:99–102. [PubMed]
  • Klug G. The role of mRNA degradation in the regulated expression of bacterial photosynthesis genes. Mol Microbiol. 1993;9:1–7. [PubMed]
  • Konninger UW, Hobbie S, Benz R, Braun V. The haemolysin-secreting ShlB protein of the outer membrane of Serratia marcescens: determination of surface-exposed residues and formation of ion-permeable pores by ShlB mutants in artificial lipid bilayer membranes. Mol Microbiol. 1999;32:1212–1225. [PubMed]
  • Koronakis V, Hughes C. Identification of the promoters directing in vivo expression of haemolysin in Proteus vulgaris and Escherichia coli. Mol Gen Genet. 1988;213:99–104. [PubMed]
  • Lucas RL, Lostroh CP, DiRusso CC, Spector MP, Wanner BL, Lee CA. Multiple factors independently regulate hilA and invasion gene expression in Salmonella enterica serovar typhimurium. J Bacteriol. 2000;182:1872–1882. [PMC free article] [PubMed]
  • Miller VL, Mekalanos JJ. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J Bacteriol. 1988;170:2575–2583. [PMC free article] [PubMed]
  • Mobley HLT, Island D, Hausinger RP. Molecular biology of microbial ureases. Microbiol Rev. 1995;59:451–480. [PMC free article] [PubMed]
  • Newbury SF, Smith NH, Higgins CF. Differential mRNA stability controls relative gene expression within a polycistronic operon. Cell. 1987;51:1131–1143. [PubMed]
  • Rhee KY, Parekh BS, Hatfield GW. Leucine-responsive regulatory protein-DNA interactions in the leader region of the ilvGMEDA operon of Escherichia coli. J Biol Chem. 1996;271:26499–26507. [PubMed]
  • Ronald SL, Kropinski AM, Farinha MA. Construction of broad-host-range vectors for the selection of divergent promoters. Gene. 1990;90:145–148. [PubMed]
  • Rozalski A, Sidorczyk Z, Kotelko K. Potential virulence factors of Proteus bacilli. Microbiol Mol Biol Rev. 1997;61:65–89. [PMC free article] [PubMed]
  • Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: a Laboratory Manual. 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1989.
  • Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci. 1977;74:5463–5465. [PMC free article] [PubMed]
  • Schmiel DH, Young GM, Miller VL. The Yersinia enterocolitica phospholipase gene yplA is part of the flagellar regulon. J Bacteriol. 2000;182:2314–2320. [PMC free article] [PubMed]
  • Senior BW, Hughes C. Production and properties of haemolysins from clinical isolates of the Proteae. J Med Microbiol. 1987;24:17–25. [PubMed]
  • Stern MJ, Prossnitz E, Ames GF. Role of the intercistronic region in post-transcriptional control of gene expression in the histidine operon of Salmonella typhimurium : involvement of REP sequences. Mol Microbiol. 1988;2:141–152. [PubMed]
  • Studier FW, Moffatt BA. Use of bacteriophage T7 DNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol. 1986;189:113–130. [PubMed]
  • Swihart KG, Welch RA. Cytotoxic activity of the Proteus hemolysin HpmA. Infect Immun. 1990;58:1861–1869. [PMC free article] [PubMed]
  • Uphoff TS, Welch RA. Nucleotide sequencing of the Proteus mirabilis calcium-independent haemolysin genes (hpmA and hpmB) reveals sequence similarity with Serratia marcescens haemolysin genes (shlA and shlB ) J Bacteriol. 1990;172:1206–1216. [PMC free article] [PubMed]
  • Walker KE, Moghaddame-Jafari S, Lockatell CV, Johnson D, Belas R. ZapA, the IgA-degrading metalloprotease of Proteus mirabilis , is a virulence factor expressed specifically in swarmer cells. Mol Microbiol. 1999;32:825–836. [PubMed]
  • Wassif C, Cheek D, Belas R. Molecular analysis of metalloprotease from Proteus mirabilis. J Bacteriol. 1995;177:5790–5798. [PMC free article] [PubMed]
  • Welch RA. Identification of two different hemolysin determinants in uropathogenic Proteus isolates. Infect Immun. 1987;55:2183–2190. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...