JOURNAL OF BACTERIOLOGY,0021-9193/97/$04.0010

Aug. 1997, p. 4741–4746 Vol. 179, No. 15

Copyright © 1997, American Society for Microbiology

A Nonswarming Mutant of Proteus mirabilis Lacks the LrpGlobal Transcriptional Regulator

NICOLE A. HAY, DONALD J. TIPPER,† DANIEL GYGI, AND COLIN HUGHES*

Department of Pathology, University of Cambridge, Cambridge CB2 1QP, United Kingdom

Received 18 February 1997/Accepted 3 June 1997

Proteus swarming is the rapid cyclical population migration across surfaces by elongated cells that hyper-express flagellar and virulence genes. The mini-Tn5 transposon mutant mns2 was isolated as a tight non-swarming mutant that did not elongate or upregulate flagellar and hemolysin genes. Individual cell motilitywas retained but was reduced. The transposon had inserted in the gene encoding the global transcriptionalregulator Lrp (leucine-responsive regulatory protein), expression of which was upregulated in differentiatingswarm cells. Swarming was restored to the lrp mutant by artificial overexpression of the flhDC flagellarregulatory master operon. Lrp may be a key component in generating or relaying signals that are required forflagellation and swarming, possibly acting through the flhDC operon.

Swarming migration in Proteus mirabilis involves the coordi-nate differentiation of short motile vegetative cells with a fewperitrichous flagella into multinucleate, aseptate swarm cells ofup to 40-fold vegetative cell length and with a more than50-fold greater density of surface flagella. Swarm cells migraterapidly away from the colony as multicellular rafts until theypause (consolidate) and undergo some dedifferentiation. Cy-cles of differentiation and consolidation follow, generating amacroscopic pattern of concentric rings or terraces (1). Swarm-ing is suggested to play a role in urinary tract colonization byProteus mirabilis (4, 6), and hyperexpression of fliC and otherflagellar genes is paralleled by the coordinate upregulation ofvirulence genes, e.g., hpmBA encoding hemolysin toxin (3).

In contrast to differentiation among other bacteria, swarm-ing appears not to be a starvation response but is rather stim-ulated by high growth rates (29) and is probably influenced bymultiple environmental signals, such as cell density and thepresence of amino acids and possibly peptides (5, 23, 38).Characterization of swarming-defective Proteus transposonmutants has indicated the involvement of a substantial numberof genes involved in differentiation and subsequent populationmigration (2, 9, 10, 23, 25–27). Both hyperflagellation and asurface polysaccharide are needed for translocation (26), andthe nonswarming mutants lacking the flagellar proteins FlhAand FlgN have confirmed that cell elongation and upregulationof flagellar and virulence genes are mechanistically coupled(25–27). Artificial overproduction of the flagellar regulatorymaster operon flhDC in Serratia and Proteus cells induces elon-gation and hyperflagellation, suggesting that it could be a pri-mary site for swarm cell induction (19, 22), but no regulatorycomponent has been identified that specifically couples physi-ological or environmental signals to swarm cell changes in geneexpression. We report a mini-Tn5Cm mutant of P. mirabilisthat is unable to swarm and show that the mutation interruptsthe lrp gene encoding a global regulator, the leucine-responsiveregulatory protein.

MATERIALS AND METHODS

Mutagenesis and assays of swarming, motility, and differentiation. Wild-typeP. mirabilis U6450 was mutagenized by the pUT mini-Tn5Cm system (15), andmutants resistant to chloramphenicol (80 mg ml21) were selected. Swarming wasassessed on 1.5% Luria-Bertani (LB) agar plates. Swarming inhibitor a-p-nitro-phenyl-glycerol (100 mg ml21, Sigma) was added to isolate single colonies.Motility was assessed in 0.3% LB agar. Differentiation was initiated by spreading200-ml stationary-phase LB cultures (ca. 5.108 ml21) onto 8-cm-diameter LBagar plates and incubating them at 37°C (25). Growth on seeding plates wasassayed by harvesting cells in LB and normalizing A600 to constant volume. Celllength was assessed by phase-contrast microscopy. Levels of surface FliC weredetermined by vortexing of washed cells for 5 min, centrifugation, and trichlo-roacetic acid precipitation of supernatants, followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Coomassie brilliant blue staining. Cell-associated hemolysin activity was calculated by normalizing hemoglobin releasedfrom erythrocytes (A543) to A600 (30). Migrating cells were collected from theswarm front (within 3 mm) of parallel swarming colonies following centralinoculation with a stationary-phase LB culture.

Recombinant DNA and sequencing. Routine DNA manipulation and electro-poration were carried out with Escherichia coli XL1Blue {recA1 endA1 gyrA96thi-1 hsdR17 supE44 relA1 lac [F9 proAB lacIqlacZDM15 Tn10(Tetr)]}. Plaquehybridizations and Southern analysis were carried out with digoxigenin (DigSystem User’s Guide [11a]). DNA fragments were subjected to exonuclease IIIdigestion (Erase-a-Base; Promega), and DNA sequencing was performed withthe T7 sequencing kit (Pharmacia). Sequences were analyzed by GCG, version 8(16).

mRNA hybridization. Total RNA was isolated with hot phenol (35). Dena-tured RNA samples (10 mg) were separated by agarose gel electrophoresis andtransferred onto nitrocellulose filters (Hybond C; Amersham). In each experi-ment, parallel samples were separated, transferred, and stained with 0.04%methylene blue to confirm equivalent quantities of total RNA in each track.Probe labelling, hybridization, and detection were carried out as previouslydescribed (25).

Nucleotide sequence accession number. The DNA sequence of lrp has beensubmitted to the EMBL database (accession no. Y10417).

RESULTS

Isolation and characterization of the motile nonswarmingmutant mns2. Following mutagenesis of P. mirabilis U6450 bythe pUT mini-Tn5Cm system, mutant mns2 was isolated as anonswarming colony unable to migrate even after prolongedincubation (Fig. 1A). Microscopic examination of cells andinoculation of the mutant into semisolid agar showed that itwas motile, although less so than the wild type (Fig. 1B).Growth rates in LB medium appeared unchanged (not shown).

Differentiation in the absence of migration was assessed byseeding of stationary-phase cells at high density onto the entiresurface of 1.5% LB agar plates. Cells differentiate synchro-nously with kinetics similar to that of a swarm cycle, althoughdifferentiation is not as extreme as that seen in cells at a

* Corresponding author. Mailing address: Department of Pathology,University of Cambridge, Tennis Court Rd., Cambridge CB2 1QP,United Kingdom. Phone/Fax: 01223 333732. E-mail: ch@mole.bio.cam.ac.uk.

† Present address: Molecular Genetics and Microbiology, Universityof Massachusetts Medical School, Worcester, MA 01655.

4741

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

migrating front (25). The wild type showed the typical ca.50-fold increase in surface flagellin FliC at peak differentiation(4 to 6 h), followed by a return to near vegetative levels at 8 h(Fig. 2A). The mutant made much less flagellin throughout,but subdued increase and decrease were still evident (Fig. 2A).

While most wild-type cells were highly elongated by 4 h, allmns2 cells remained short (Fig. 2B). Wild-type hemolytic ac-tivity peaked sharply at 4 h, and although mns2 again showedsubdued induction, the level achieved was only 0.5% of that ofthe wild type (Fig. 2C). In contrast, urease activity remained at70% of that of the wild type at 3 to 4 h, and protease activitywas unaffected (not shown).

A mutation in the lrp global regulator gene. A restrictionmap of the mutant locus was constructed by Southern hybrid-ization with a probe internal to the transposon, and the locuswas cloned into the vector pBluescript SK (Stratagene) as aPstI-SalI fragment, selecting Tn5Cm-encoded chlorampheni-col resistance. This generated the plasmid p200PS (Fig. 3A). Afragment of the cloned chromosomal region was then used toprobe a l DashII phage library of partial Sau3A fragmentsfrom the wild-type P. mirabilis chromosome (25). Severalphage hybridized and were used to assemble a restriction mapof the locus (Fig. 3A), and the plasmid p200XP was con-structed by subcloning a 2.03-kb fragment of the phage insert(Fig. 3A) into pBluescript SK. Transformation by p200XP re-stored swarming to the mutant, albeit determining a mildlydendritic pattern with swarm terraces closer together thanthose of the wild type.

Sequencing of p200XP and of p200PS showed that the trans-poson had inserted between bases 101 and 102 of a 492-bp,164-codon open reading frame (ORF) (Fig. 3B). The predictedamino acid sequence has 97% identity with the E. coli leucine-responsive regulatory protein (Lrp), which is highly conservedin members of the Enterobacteriaceae (20). Lrp is a globaltranscriptional regulator that influences many pathways, in-cluding several of amino acid metabolism (13, 36). E. coli lrpmutants grow poorly in the absence of exogenous serine be-cause the biosynthetic gene serA is positively regulated by Lrp(41). This was true of the P. mirabilis mns2 mutant (notshown). Downstream (128 bp) of the Proteus lrp gene, withinp200XP, are the first 804 bp of an ORF with 55% identity toftsK, an E. coli cell division gene that lies similarly 39 of lrp (8).Proteus ftsK is preceded by potential 235 and 210 promotercomponents and a potential lex box, an element allowing in-duction by the SOS response also identified 59 of E. coli ftsK(34). The loss of swarming cannot be explained by a polareffect upon ftsK, since swarming is restored by the comple-menting plasmid p200XP, which contains only the predictedfirst 20% of the ftsK gene (the E. coli protein has 1,330 aminoacids). This was confirmed by subsequent complementationwith a plasmid (constructed by exonuclease digestion ofp200XP) encoding Lrp but only the first 23 amino acids of FtsK(p200xp4 [Fig. 3]) and by the absence of complementation bya plasmid lacking ftsK and in which lrp was missing the last 14codons (p200xp5 [Fig. 3]).

Upregulation of the lrp transcript during swarm cell differ-entiation. Northern blots of RNA from wild-type cells with aprobe internal to the lrp gene identified major transcripts withsizes of about 650, 750, and 1,100 nucleotides, considerablylarger than the 492-bp ORF1, but none large enough to en-compass the downstream ftsK gene (Fig. 4A). In E. coli, thestart of the lrp transcript has been mapped by primer extensionat 267 bp upstream of the ORF (43). The 600-bp sequenceupstream of lrp lacks detectable ORFs but also contains noclear 59 promoter consensus sequences for s70 or s28, althoughthere are several poor matches. There is a potential stem-loopstructure 65 bp downstream of lrp that might act as a transcrip-tion terminator.

Analysis of cultures from seeding plates showed that expres-sion of all of the lrp transcripts is upregulated during swarmcell differentiation (Fig. 4A), following kinetics similar to those

FIG. 1. Swarming and motility of the P. mirabilis wild type (wt) and mns2mutant. Overnight cultures (1 ml) were inoculated onto 1.5% LB agar (A) andinto 0.3% LB agar (B) and incubated for 6 h at 37°C.

FIG. 2. Cell differentiation of the P. mirabilis wild type (wt) and mns2 mutantinitiated by seeding onto 1.5% LB agar. (A) Surface flagellin. (B) Elongation (at4 h; phase-contrast microscopy at 31,000 magnification). (C) Cell-associatedhemolysin activity, with hemoglobin released from erythrocytes (A543) normal-ized to cell mass. (Inset) mns2 hemolytic activity, expanded scale.

4742 HAY ET AL. J. BACTERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

of fliC, but reaching a maximum about 30 min earlier. Levels ofan rRNA transcript remained constant. A growth curve of theseeding plate culture indicated that this transient lrp inductionoccurs in the late exponential phase of growth, decreasingagain as cells enter the stationary phase (Fig. 4B). Whether thisbatch growth phase is a determinant of swarm cell differenti-ation is not clear, but to ensure that the upregulation was notspecific to the culture growing after high-density seeding, theresult was confirmed with mRNA isolated from cells at the

edge of actively swarming colonies. Both the fliC and lrp tran-script levels peaked in maximally differentiated cells shortlybefore migration (Fig. 4C [6 h]), with lower levels of transcriptin cells of migrating populations (Fig. 4C [4 h]). So, whether inresponse to growth phase or to other signals, lrp is upregulatedin differentiating swarm cells.

Downregulated flagellar and hemolysin gene expression inthe lrp mutant. Hemolysin and flagellar gene mRNAs wereassayed in the lrp mutant at 4 h in the seeding assay at maxi-

FIG. 3. (A) The mns2 locus and chromosomal regions carried in the vector pBluescript SK. The hatched box indicates DNA derived from the transposon. Darkboxes indicate ORFs. S, SalI; N, NsiI; H, HindIII; E, EcoRI; X, XbaI; P, PstI; B, BglII. (B) Nucleotide sequence of the p200XP insert, with deduced amino acidsequences of Lrp and FtsK shown below. SD, Shine-Dalgarno site; Tn, site of transposon insertion; xp4 and xp5, 39 extent of inserts carried in plasmids p200xp4 andp200xp5, respectively; 235 and 210, putative promoter elements. The lex box is indicated, and the putative terminator is underlined.

VOL. 179, 1997 Lrp AND SWARMING 4743

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

mum differentiation. The fliC message concentration was re-duced (Fig. 5A) to a low level corresponding to the level ofFliC filament protein detected on the surface of these cells(Fig. 2A). The levels of expression of flagellar genes flgM(another class III gene), flgB (class II), and flhDC (class I) wereall substantially reduced (Fig. 5A). Levels of hpm transcriptencoding hemolysin were strongly reduced, such that radiola-belled message from the mutant was detected by autoradiog-raphy only after overexposure of the wild-type mRNA signal.Six randomly cloned U6450 transcribed sequences did notshow any reduction in transcript in the lrp mutant (Fig. 5B).

In cells grown in liquid culture, levels of flhDC transcriptwere also lower in the lrp mutant than in the wild type (Fig.

5C), in agreement with the lower flagellin protein levels seen inovernight cultures (Fig. 2A) and impaired motility in semisolidagar (Fig. 1). This suggests that the influence of Lrp on flagel-lar gene expression may not be limited to swarm cells, but thereremains some doubt, because these conditions do support alow level of differentiation of the wild type (17), and introduc-tion of an E. coli lrp mutation (12) into a motile strain of thatspecies (14) did not impair motility (not shown).

Overexpression of flhDC restores swarming to the lrp mu-tant. Swarming is a demanding process with a substantial pro-portion of metabolism given over to the assembly and opera-tion of flagella. It has been argued that the physiologicalchanges in an E. coli lrp mutant result in overall metaboliccompromise (7), so it seemed possible that the Proteus lrpmutant was metabolically incapable of supporting hyperflagel-lation. Flagellar gene expression was artificially induced in thelrp mutant by placing the P. mirabilis flhDC master operon andthe preceding 60 bp under the control of the arabinose pro-moter in the vector pBAD18 (24), from which expression isrepressed by glucose and induced by arabinose. Transformantsgrown on LB agar with 0.2% glucose remained nonswarming,although transformants of the wild type swarmed normally onthat medium. However, induction of gene expression by theinclusion of 0.2% arabinose restored swarming in the lrp mu-tant, both mutant and wild-type transformants swarmed atalmost twice the normal rate (Fig. 6A). High levels of surfaceFliC were also restored in the pBADDC-complemented mu-tant (Fig. 6B), and the cells became fully elongated (data notshown), but the hemolytic activity of the lrp mutant was notrestored (data not shown).

DISCUSSION

The mns2 mutant was one of five motile nonswarming mu-tants among more than 3,500 Cmr transconjugants (the others

FIG. 4. (A) Hybridization of lrp, fliC, and an rRNA probe, r2, to RNA fromwild-type P. mirabilis sampled during a differentiation cycle initiated by seedingonto 1.5% LB agar. (B) Growth of wild-type cell culture during a differentiationcycle initiated by high-density seeding onto 1.5% LB agar. (C) Hybridization oflrp fliC and an rRNA probe, r2, to RNA from wild-type P. mirabilis sampled fromthe front (3 mm) of expanding swarm colonies. The times shown are afterinoculation of the colony. Sw, periods of population migration (swarming).

FIG. 5. (A) Hybridization of fliC, flgM, flgB, flhDC, and hpmA probes toRNA from wild-type (wt) and mns2 (lrp) cells, sampled at peak differentiation,4 h after being seeded onto 1.5% LB agar. (B) Hybridization of six randomProteus EcoRI-HindIII probes to mRNA of the wild type (wt) and the mns2mutant (lrp) sampled at 4 h. The wt9 and lrp9 samples contain 10-fold moreRNA. (C) Hybridization of an flhDC probe to mRNA of the wild type (wt) andthe mns2 mutant (lrp) prepared from cells grown in liquid LB medium to an A600of 1.2.

4744 HAY ET AL. J. BACTERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

were mutated in different loci). The gene disrupted in mns2encodes the Lrp global transcriptional regulator. Although aunifying role for Lrp has not been defined, it is pivotal to theexpression of at least 40 genes in E. coli, including many in-volved in amino acid biosynthesis and degradation, nitrogenassimilation, biosynthesis of one-carbon units, peptide trans-port, and pilin synthesis (13, 36), repressing expression of somegenes and activating others. It is a small basic protein thatbends DNA and is relatively abundant in the cell, promptingthe proposal that it acts as a chromosome organizer (36, 42).Expression of lrp itself is, in E. coli, higher in poor growthmedia and in the stationary growth phase (31).

The Proteus lrp mutant was profoundly defective in bothswarming and differentiation, being unable to hyperflagellate,elongate, or hyperinduce hemolysin toxin. Protease and ureaseproduction were not significantly decreased in the mutant,indicating that their swarm cell induction follows a differentbranch. The effect of Lrp upon hyperflagellation seems likelyto be mediated through the flhDC master regulon, since tran-scription of this operon and of the genes below it in the flagel-lar hierarchy was strongly reduced in the lrp mutant, andswarming was fully restored by artificial induction of flhDCexpression. In both Serratia and Proteus, hyperflagellation, cellelongation, and swarming can be induced by upregulation offlhDC, the operon at the top of the flagellar hierarchy (19, 22),indicating that flhDC is a focal point of the mechanistic rela-tionship between hyperflagellation and inhibition of cell divi-sion (25) and may be a key control point in swarming. Therestoration of swarming following artificial induction of flhDCsuggests that the lrp mutant fails to induce flhDC upregulationrather than being physiologically incapable of supporting hy-perflagellation. In keeping with a role for Lrp in the inductionof swarming, lrp transcription was upregulated in differentiat-ing swarm cells, peaking prior to swarming migration slightlyearlier than fliC. The reduction in surface flagellin and flhDCtranscription seen in lrp cells grown in liquid culture suggeststhat Lrp also has some influence over flagellation in Proteusvegetative cells. Lrp-mediated regulation of flagellation mightbecome critical during the hyperinduction involved in swarm-ing. Hemolytic activity was not restored by flhDC overproduc-tion, implying that despite being coupled to swarm cell differ-

entiation, it is not under flhDC control. Alternatively, Lrp maybe directly required for hemolysin induction.

Lrp may itself positively regulate the flagellar and otherswarming-associated genes, or its effect may be indirect, me-diated by an Lrp-dependent factor or factors. In E. coli, flhDCis subject to regulation by the cyclic AMP-receptor proteinCRP and the osmoregulator OmpR (39, 40). H-NS mutantsare also defective in flagellation, and there are possible con-sensus sequences for regulation by the integration host factor(11, 39). Like Lrp, the latter proteins alter DNA topology uponbinding, and these proteins are global regulators (21, 28).Complex transcriptional regulation involving Lrp along withother master regulators has been reported for osmY and ilvIH(32, 33), and the possibility that DNA topology changes allowmultiple layers of gene regulation in response to environmen-tal or physiological factors is attractive (18, 37). Lrp may acttogether with other regulators to integrate a range of signalsinto the control of flagellation and swarming.

ACKNOWLEDGMENTS

We thank L. Guzman (Harvard Medical School) for the pBADvector, R. Belas (University of Maryland) for the fliC gene, and D. Low(University of Utah Medical Center) for the E. coli lrp mutant.

This work was supported by a project grant from the Medical Re-search Council and a Programme grant from the Wellcome Trust.

REFERENCES

1. Allison, C., and C. Hughes. 1991. Bacterial swarming: an example of pro-karyotic differentiation and multicellular behaviour. Sci. Prog. Edinburgh75:403–422.

2. Allison, C., and C. Hughes. 1991. Closely linked genetic loci required forswarm cell differentiation and multicellular migration by Proteus mirabilis.Mol. Microbiol. 5:1975–1982.

3. Allison, C., H.-C. Lai, and C. Hughes. 1992. Co-ordinate expression ofvirulence genes during swarm-cell differentiation and population migrationof Proteus mirabilis. Mol. Microbiol. 6:1583–1591.

4. Allison, C., N. Coleman, P. L. Jones, and C. Hughes. 1992. Ability of Proteusmirabilis to invade human urothelial cells is coupled to motility and swarmingdifferentiation. Infect. Imm. 60:4740–4746.

5. Allison, C., H.-C. Lai, D. Gygi, and C. Hughes. 1993. Cell differentiation ofProteus mirabilis is initiated by glutamine, a specific chemoattractant forswarming cells. Mol. Microbiol. 8:53–60.

6. Allison, C., L. Emody, N. Coleman, and C. Hughes. 1994. The role of swarmcell differentiation and multicellular migration in the uropathogenicity ofProteus mirabilis. J. Infect. Dis. 169:1155–1158.

7. Ambartsoumain, G., R. D’Ari, R. T. Lin, and E. B. Newman. 1994. Alteredamino acid metabolism in lrp mutants of Escherichia coli K12 and theirderivatives. Microbiology 140:1737–1744.

8. Begg, K. J., S. J. Dewar, and W. D. Donachie. 1995. A new Escherichia colicell division gene, ftsK. J. Bacteriol. 177:6211–6222.

9. Belas, R., D. Erskine, and D. Flaherty. 1991. Proteus mirabilis mutantsdefective in swarmer cell differentiation and multicellular behavior. J. Bac-teriol. 173:6279–6288.

10. Belas, R., M. Goldman, and K. Ashliman. 1995. Genetic analysis of Proteusmirabilis mutants defective in swarmer cell elongation. J. Bacteriol. 177:823–828.

11. Bertin, P., E. Terao, E. H. Lee, P. Lejeune, C. Colson, A. Danchin, and E.Collatz. 1994. The H-NS protein is involved in the biogenesis of flagella inEscherichia coli. J. Bacteriol. 176:5537–5540.

11a.Boehringer. 1993. Dig system user’s guide. Boehringer, Mannheim, Ger-many.

12. Braaten, B. A., L. B. Blyn, B. S. Skinner, and D. A. Low. 1991. Evidence fora methylation-blocking factor (mbf) locus involved in pap pilus expressionand phase variation in Escherichia coli. J. Bacteriol. 173:1789–1800.

13. Calvo, J. M., and R. G. Matthews. 1994. Leucine-responsive regulatoryprotein—a global regulator of metabolism in Escherichia coli. Microbiol.Rev. 58:466–490.

14. Casadaban, M., and S. N. Cohen. 1980. Analysis of gene control signals byDNA fusion and cloning in Escherichia coli. J. Mol. Biol. 138:179–207.

15. de Lorenzo, V., M. Herrero, U. Jakubzik, and K. N. Timmis. 1990. Mini-Tn5transposon derivatives for insertion mutagenesis, promoter probing, andchromosomal insertion of cloned DNA in gram-negative eubacteria. J. Bac-teriol. 172:6568–6572.

16. Devereux, J., P. Haeberli, and O. Smithies. 1984. A comprehensive set ofsequence analysis programs for the VAX. Nucleic Acids Res. 12:387–395.

FIG. 6. (A) Swarming on 1.5% LB agar (0.2% arabinose) of P. mirabiliscomplemented with P. mirabilis flhDC. wt, wild type containing the vectorpBAD18; wt pflhDC, wild type containing pBAD flhDC; lrp pflhDC, lrp mutantcontaining pBADflhDC; lrp, lrp mutant containing pBAD18 (lrp). (B) Surfaceflagellin from the cells listed above seeded onto LB agar and harvested at 4 h.

VOL. 179, 1997 Lrp AND SWARMING 4745

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

17. Dick, H., R. G. E. Murray, and S. Walmsley. 1985. Swarmer cell differenti-ation of Proteus mirabilis in fluid media. Can. J. Microbiol. 31:1041–1050.

18. Dorman, C. J., and N. Ni Bhriain. 1993. DNA topology and bacterial viru-lence gene regulation. Trends Microbiol. 1:92–99.

19. Eberl, L., G. Christiansen, S. Molin, and M. Givskov. 1996. Differentiationof Serratia liquefaciens into swarm cells is controlled by the expression of theflhD master operon. J. Bacteriol. 178:554–559.

20. Friedberg, D., J. V. Platko, B. Tyler, and J. M. Calvo. 1988. The amino acidsequence of Lrp is highly conserved in four enteric microorganisms. J. Bac-teriol. 177:1624–1626.

21. Friedman, D. I. 1988. Integration host factor: a protein for all reasons. Cell54:713–721.

22. Furness, R. B., G. M. Fraser, N. A. Hay, and C. Hughes. Negative feedbackfrom a Proteus class II flagella export defect onto the flhDC master operoncontrolling cell division and flagella assembly. Submitted for publication.

23. Gaisser, S., and C. Hughes. 1997. A locus coding for putative non-ribosomalpeptide/polyketide synthase functions is mutated in a swarming-defectiveProteus mirabilis strain. Mol. Gen. Genet. 253:415–427.

24. Guzman, L.-M., D. Belin, M. J. Carson, and J. Beckwith. 1996. Tight regu-lation, modulation, and high-level expression by vectors containing the arab-inose PBAD promoter. J. Bacteriol. 177:4121–4130.

25. Gygi, D., M. J. Bailey, C. Allison, and C. Hughes. 1995. Requirement forFlhA in flagella assembly and swarm-cell differentiation by Proteus mirabilis.Mol. Microbiol. 15:761–769.

26. Gygi, D., M. M. Rahman, H.-C. Lai, R. Carlson, J. Guard-Petter, and C.Hughes. 1995. A capsular polysaccharide required for rapid population mi-gration by differentiated swarm cells of Proteus mirabilis. Mol. Microbiol.17:1167–1175.

27. Gygi, D., G. Fraser, A. Dufour, and C. Hughes. A motile but non-swarmingmutant of Proteus mirabilis lacks FlgN, a facilitator of flagella filament as-sembly. Mol. Microbiol., in press.

28. Hulton, C. S. J., A. Seirafi, J. C. D. Hinton, J. M. Sidebotham, L. Waddell,G. D. Pavitt, T. Owen-Hughes, A. Spassky, H. Buc, and C. F. Higgins. 1990.Histone-like protein H-1 (H-NS), DNA supercoiling and gene expression inbacteria. Cell 63:631–642.

29. Jones, H. E., and R. W. A. Park. 1967. The influence of medium compositionon the growth and swarming of Proteus. J. Gen. Microbiol. 47:369–378.

30. Koronakis, V., M. Cross, B. Senior, E. Koronakis, and C. Hughes. 1987. Thesecreted hemolysins of Proteus mirabilis, Proteus vulgaris, and Morganellamorganii are genetically related to each other and to the alpha-hemolysin ofEscherichia coli. J. Bacteriol. 169:1509–1515.

31. Landgraf, J. R., J. Wu, and J. M. Calvo. 1996. Effects of nutrition and growthrate on Lrp levels in Escherichia coli. J. Bacteriol. 178:6930–6936.

32. Lange, R., M. Barth, and R. Hengge-Aronis. 1993. Complex transcriptionalcontrol of the ss-dependent stationary-phase-induced and osmotically regu-lated osmY (csi-5) gene suggests novel roles for Lrp, cyclic AMP (cAMP)receptor protein-cAMP complex, and integration host factor in the station-ary-phase response of Escherichia coli. J. Bacteriol. 175:7910–7917.

33. Levinthal, M., P. Lejeune, and A. Danchin. 1994. The H-NS protein modu-lates the activation of the ilvIH operon of Escherichia coli K12 by Lrp, theleucine regulatory protein. Mol. Gen. Genet. 242:736–743.

34. Lewis, L. K., M. E. Jenkins, and D. W. Mount. 1992. Isolation of DNAdamage-inducible promoters in Escherichia coli: regulation of polB (dinA),dinG, and dinH by LexA repressor. J. Bacteriol. 174:3377–3385.

35. Melton, D. A., P. A. Krieg, M. R. Rebagliati, K. Zinn, and M. R. Green. 1984.Efficient in vitro synthesis of biologically active RNA and RNA hybridizationprobes from plasmids containing a bacteriophage SP6 promoter. NucleicAcids Res. 12:7035–7056.

36. Newman, E. B., and R. Lin. 1995. Leucine-responsive regulatory protein: aglobal regulator of gene expression in E. coli. Annu. Rev. Microbiol. 49:747–775.

37. Perez-Martin, J., F. Rojo, and V. de Lorenzo. 1994. Promoters responsive toDNA bending: a common theme in prokaryotic gene expression. Microbiol.Rev. 58:268–290.

38. Rauprich, O., M. Matsushita, C. J. Weijer, F. Seigert, S. E. Esipov, and J. A.Shipiro. 1996. Periodic phenomena in Proteus mirabilis swarm colony devel-opment. J. Bacteriol. 178: 6525–6538.

39. Shin, S., and C. Park. 1995. Modulation of flagellar expression in Escherichiacoli by acetyl phosphate and the osmoregulator OmpR. J. Bacteriol. 177:4696–4702.

40. Silverman, M., and M. Simon. 1974. Characterization of Escherichia coliflagellar mutants that are insensitive to catabolite repression. J. Bacteriol.120:1196–1203.

41. Tuan, L. R., R. D’Ari, and E. B. Newman. 1990. The leucine regulon ofEscherichia coli K-12: a mutation in rblA alters expression of L-leucine-dependent metabolic operons. J. Bacteriol. 172:4529–4535.

42. Wang, Q., and J. M. Calvo. 1993. Lrp, a major regulatory protein in Esch-erichia coli, bends DNA and can organize the assembly of a higher-ordernucleoprotein structure. EMBO J. 12:2495–2501.

43. Wang, Q., J. Wu, D. Friedberg, J. Platko, and J. M. Calvo. 1994. Regulationof the Escherichia coli lrp gene. J. Bacteriol. 176:1831–1839.

4746 HAY ET AL. J. BACTERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from