Functional and Structural Characterization of RsbU,a Stress Signaling Protein Phosphatase 2C*

Received for publication, May 17, 2004, and in revised form, July 7, 2004Published, JBC Papers in Press, July 19, 2004, DOI 10.1074/jbc.M405464200

Olivier Delumeau‡§, Sujit Dutta¶�, Matthias Brigulla**‡‡, Grit Kuhnke**‡‡§§,Steven W. Hardwick¶¶, Uwe Volker**‡‡, Michael D. Yudkin‡, and Richard J. Lewis¶��

From the ‡Microbiology Unit and ¶Laboratory of Molecular Biophysics, Department of Biochemistry, University of Oxford,Oxford OX1 3QU, United Kingdom, the **Department of Biochemistry, Max-Planck-Institute for Terrestrial Microbiology,D-35032 Marburg, Germany, the ‡‡Laboratory for Functional Genomics, Medical School, Ernst-Moritz-Arndt-University,D-17487 Greifswald, Germany, and the ¶¶Institute of Cell and Molecular Biosciences, Faculty of Medical Sciences,University of Newcastle, Newcastle upon Tyne NE2 4HH, United Kingdom

RsbU is a positive regulator of the activity of �B, thegeneral stress-response � factor of Gram� microorgan-isms. The N-terminal domain of this protein has no sig-nificant sequence homology with proteins of knownfunction, whereas the C-terminal domain is similar tothe catalytic domains of PP2C-type phosphatases. Thephosphatase activity of RsbU is stimulated greatly dur-ing the response to stress by associating with a kinase,RsbT. This association leads to the induction of �B ac-tivity. Here we present data on the activation processand demonstrate in vivo that truncations in the N-ter-minal region of RsbU are deleterious for the activationof RsbU. This conclusion is supported by comparisons ofthe phosphatase activities of full-length and a truncatedform of RsbU in vitro. Our determination of the crystalstructure of the N-terminal domain of RsbU from Bacil-lus subtilis reveals structural similarities to the regula-tory domains from ubiquitous protein phosphatases anda conserved domain of �-factors, illuminating the acti-vation processes of phosphatases and the evolution of“partner switching.” Finally, the molecular basis of ki-nase recruitment by the RsbU phosphatase is discussedby comparing RsbU sequences from bacteria that eitherpossess or lack RsbT.

Reversible phosphorylation of proteins is the predominantregulatory mechanism in biology, modulating cellular pro-

cesses such as signaling, division, and development. The phos-phorylation of regulatory proteins by protein kinases effects achange in their function and structure (1), reversed by theaction of protein phosphatases, which restore the regulatoryproteins to their original, unphosphorylated state. Hence thecellular response is determined by controlling the enzymaticactivities of the mutually antagonistic kinases and phospha-tases. Protein phosphatases can be divided into three majorgroups, defined by their substrate specificity (2): phosphoty-rosine (further subdivided into Cdc25 and the conventional andthe low molecular weight phosphotyrosine phosphatases),phosphoserine/threonine (further subdivided into protein phos-phatase P and M families), and phosphoaspartate phospha-tases (e.g. Rap and Spo0E from Bacillus). In addition, there isthe dual-specific phosphatase group, which can dephosphoryl-ate phosphotyrosine and phosphoserine/threonine substrates,and it has been recorded that several histidine kinases alsohave phosphohistidine phosphatase activity (3).

In Bacillus subtilis, five serine/threonine protein phosphata-ses have been identified that belong to the PP2C subgroup ofthe protein phosphatase M family, namely PrpC (4), SpoIIE (5),RsbP (6), RsbX and RsbU (7). PrpC plays an important regu-latory role in stationary phase (8), and SpoIIE regulates differ-entiation in Bacillus, a process known as sporulation, by form-ing a complex with the cell division protein FtsZ (5, 9). RsbX,RsbP, and RsbU are involved in the regulation of the alterna-tive � factor, �B, which controls the general stress response ofB. subtilis and other Gram� microorganisms, such as the hu-man pathogens Listeria monocytogenes and Staphylococcus au-reus (10, 11). Depending on the nature of the stress, the signalis conveyed to �B by two separate pathways, which converge onphosphorylated RsbV (RsbV-P),1 the common substrate forRsbP and RsbU (Fig. 1). A decrease in the intracellular energylevel activates RsbP, via RsbQ (12), whereas environmentalstresses such as heat or salt shock, or ethanol treatment, acti-vate RsbU, via the kinase RsbT (7, 13, 14). DephosphorylatedRsbV subsequently liberates �B from the transcriptionally in-active �B�RsbW complex, by competing with �B for bindingsurfaces on RsbW (15, 16), freeing �B to bind to core RNApolymerase to activate transcription of the �B regulon. Thealternative binding of RsbW to �B or RsbV is a regulatorymechanism called “partner switching” (17).

RsbT is related to the anti-� factors RsbW and SpoIIAB; allare members of the GHKL family of kinases/ATPases, a groupthat includes two-component histidine kinases, topoisomer-

* This work was supported by a research grant (to M. D. Y.) and astudentship award (to R. J. L.) from the Biotechnology and BiologicalSciences Research Council, a Wellcome Trust Research Career Devel-opment Fellowship (to R. J. L.), “headroom” funds from the Universityof Newcastle (to R. J. L.), and grants from the Max-Planck-Society andthe Deutsche Forschungsgemeinschaft (to U. V.) for research in theLaboratory for Functional Genomics. The costs of publication of thisarticle were defrayed in part by the payment of page charges. Thisarticle must therefore be hereby marked “advertisement” in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and structure factors (code 1W53) have beendeposited in the Protein Data Bank, Research Collaboratory for Struc-tural Bioinformatics, Rutgers University, New Brunswick, NJ(http://www.rcsb.org/).

§ Current address: Institute of Cell and Molecular Biosciences, Fac-ulty of Medical Sciences, University of Newcastle, Newcastle uponTyne, NE2 4HH, United Kingdom.

� Current address: National Institute for Medical Research, ProteinStructure, The Ridgeway, Mill Hill, London NW7 1AA, UnitedKingdom.

§§ Current address: Institute for Cytobiology, Philipps-University,D-35037 Marburg, Germany.

�� To whom correspondence should be addressed: Institute of Cell andMolecular Biosciences, Faculty of Medical Sciences, University of New-castle, Newcastle upon Tyne, NE2 4HH, United Kingdom. Tel: 44-0-191-222-5482; Fax: 44-0-191-222-7424; E-mail: R.Lewis@ncl.ac.uk.

1 The abbreviations used are: RsbV-P, phosphorylated RsbV; IPTG,isopropyl-1-thio-�-D-galactopyranoside; TPR, tetratricopeptide repeat;PP5, protein phosphatase 5.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 39, Issue of September 24, pp. 40927–40937, 2004© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 40927

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

ases, and chaperones. However, RsbT does not bind, as in theother partner-switching mechanisms, to � or anti-anti-� fac-tors. During exponential growth RsbT is thought to be seques-tered in a large supramolecular complex composed of RsbR andRsbS (18). However, during environmental stress RsbT is lib-erated from the supramolecular complex after phosphorylatingits substrates, RsbR and RsbS. RsbT is then free to associatewith, and activate, RsbU (7). RsbU is not a substrate for thekinase activity of RsbT (19), and the precise mechanism of theactivation of RsbU by RsbT remains unknown. Analysis ofthe sequence of RsbU shows that it is composed of two domains,a C-terminal domain of �200 amino acids with sequence ho-mology to other PP2C-type phosphatases and a N-terminaldomain of �110 amino acids with no significant homology toany non-RsbU sequences. The simplest hypothesis regardingthe role of the N-terminal domain of RsbU is that it exerts aninhibitory influence on the C-terminal, catalytic domain, whichis relieved by the binding of RsbT to RsbU.

We report here an investigation into the activation process ofRsbU that integrates genetics and molecular and structuralbiology. The phenotype of B. subtilis strains containing dele-tions in the N-terminal domain of RsbU support the view thatthis domain plays a critical role in the activation of the phos-

phatase by RsbT. These results are discussed in the light of ourdetermination of the crystal structure at 1.6-Å resolution of theN-terminal domain of RsbU, and conclusions are drawn as tothe nature of the activation process of RsbU.

EXPERIMENTAL PROCEDURES

Bacterial Strains, Media, and Growth Conditions—The experimentsconducted in this study (Table I) were performed with derivatives of B.subtilis wild type strain 168 (20). For the construction of the varioustruncated versions of RsbU, there were two approaches. First, the rsbUgene, and truncated versions thereof, were amplified by PCR withchromosomal DNA of B. subtilis strain 168 as the template usingappropriate primer pairs and a proofreading Taq polymerase. Subse-quently, all these PCR products were digested with HindIII and SalIand ligated into the non-integrative plasmid pDG148 (21), which hadalso been digested with the same enzymes. The different rsbU alleleswere thus placed under the control of the IPTG-regulated promoterPspac. After transformation into Escherichia coli strains TOP10 or TG2,and confirmation of the correct DNA sequence of the rsbU inserts bysequencing, plasmids carrying the wild type rsbU sequence (pGK01)and the truncations �rsbU1–19 (pMB21), �rsbU1–38 (pGK02),�rsbU1–77 (pGK03), �rsbU1–93 (pGK04), and �rsbU1–134 (pGK05)were selected for transformation into B. subtilis BSA140 (Table I). TheB. subtilis strain BSA140 carries a ctc::lacZ transcriptional reportergene fusion, but lacks a functional copy of rsbU because of a deletion ofan NdeI fragment internal to the rsbU structural gene. Transformants

FIG. 1. Model for the regulation of �B. Pre-stress, the anti-� factor RsbW sequesters �B in a transcriptionally inactive complex. RsbW alsoinactivates RsbV by phosphorylation through its kinase activity. During the initial response to an imposed stress (solid arrows) either RsbP(energy) or RsbU (environmental) phosphatases are activated, which dephosphorylate RsbV-P. The product, RsbV, attacks the RsbW��B complexand forms an alternative complex with RsbW by liberating �B (dashed arrows). Free �B then forms a complex with RNA polymerase, initiatingtranscription of the �130 general stress response genes. RsbT activates RsbU in the environmental stress response pathway. In the absence ofstress, RsbT is held in an inactive form in a large molecular weight complex, the “stressosome” with at least RsbR and RsbS. During stress, RsbRpresumably becomes phosphorylated and stimulates the rate of phosphorylation of RsbS; both reactions are catalyzed by RsbT. RsbT is thenreleased from the stressosome to activate RsbU (dashed arrows). The phosphatase for RsbS-P, RsbX, resets the level of �B activity to the pre-stresslevels. Ringed plus signs indicate which events regulate �B activity in a positive fashion, ringed minus signs indicate those that are negative. Thismodel has been adapted from, among others, Refs. 18, 54, and 55.

Activation of a Stress-responsive Phosphatase40928

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

were selected for their resistance to kanamycin (20 �g ml�1) creatingstrains BSG14, BSG15, BSG16, BSG17, BSG18, and BSG19, respec-tively (Table I).

Bacteria were routinely grown under vigorous agitation in a minimalmedium described previously (22) supplemented with 0.2% (w/v) glu-cose as a carbon source and L-tryptophan (0.78 mM). The cultures wereinoculated from overnight cultures propagated in minimal media con-taining kanamycin to an optical density at 540 nm of 0.05. The expres-sion of the plasmid-encoded rsbU variants was induced by the additionof IPTG to a final concentration of 1 mM. Ethanol stress was imposed onthe cells during exponential growth phase (A540 � 0.3) by the additionof ethanol to a final concentration of 4% (v/v).

In the second approach, B. subtilis strain PB291, previously deleted forrsbU, was transformed with derivatives of plasmid pMLK (23), whichdirects integration at the amyE locus. A fragment of DNA correspondingto PA-rsbR-rsbS-rsbT-rsbU (where PA represents the �A-dependent pro-moter region of the operon) was excised from plasmid pAW70 (24) byBamHI restriction digest and cloned in pMLK using the BamHI site inamyE. For the cloning of rsbR, rsbS, rsbT, and a fragment of rsbU codingfor the C-terminal catalytic domain only (�N-rsbU-(1–112), the regionscorresponding to PA-rsbR-rsbS-rsbT and �N-rsbU-(1–112) were amplifiedseparately by PCR using primers that allow overlaps between the 3� and5� ends, respectively. The annealed PCR products were subsequently usedas the template for a further round of PCR to yield a DNA fragment ofPA-rsbR-rsbS-rsbT-�N-rsbU-(1–112), which was also cloned in pMLK atthe BamHI site in amyE. Transformants of these strains, which arediploid for rsbR, rsbS, and rsbT, were selected by loss of amylase activityon starch plates. Bacteria were grown in buffered LB, and subjected to 4%ethanol stress at time 0.

In both cases, the �-galactosidase activity of the ctc::lacZ reportergene fusion was determined by harvesting 1-ml aliquots of cells atappropriate time points by centrifugation at 4 °C. �-Galactosidase en-zyme assays were conducted as previously described (14, 25).

Purification of Full-length RsbU and Truncated Proteins—RsbU wasoverexpressed in E. coli (BL21) and was purified using a procedureslightly modified from that previously published (18). Cells were dis-rupted by sonication in 30 ml of lysis buffer (50 mM Tris-HCl, pH 8, 1mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 10 mM MgCl2)and centrifuged for 30 min at 15,000 rpm. The supernatant was then

applied to a DEAE-Sepharose (Amersham Biosciences) column pre-equilibrated with 50 mM Tris-HCl, pH 8, 1 mM dithiothreitol, 10 mM

MgCl2 and the chromatogram was developed with a NaCl gradient from0 to 400 mM. The RsbU-containing fractions were concentrated bycentrifugal filtration (Amicon) and then loaded onto a Superdex-200 gelfiltration column (Amersham Biosciences). To remove the few remain-ing contaminants, RsbU was further purified using high resolutionMono-Q ion exchange chromatography using the same buffer conditionsas for the earlier DEAE-Sepharose column. For C-RsbU, which wascloned into pET15b, the overexpression and purification protocols weresimilar to those for full-length RsbU, except that a Superdex-75 gelfiltration column was used instead of Superdex-200. The two N-RsbUconstructs (residues 1–84 and 1–112) were purified as described previ-ously (26); RsbT was purified as described previously (18) except that allbuffers were supplemented with 100 �M ATP to maintain RsbT in asoluble form. Estimates of the molecular sizes of RsbU, N-RsbU, andC-RsbU were obtained by the use of Superdex-200 (Amersham Bio-sciences) gel filtration chromatography, calibrated against proteins ofknown molecular sizes.

Measurement of the Phosphatase Activity—RsbV-P, the phosphatasesubstrate, was prepared as previously described (18). The dephospho-rylation reactions were performed in a buffer of 50 mM Tris-HCl, pH 7.5,50 mM KCl, 10 mM MgCl2, 1 mM MnCl2, and 1 mM dithiothreitol at 30 °Cwith 30 �M RsbV-P, 0.5 �M RsbU (or C-RsbU) and, unless specified, 1�M RsbT. The rates of dephosphorylation of RsbV-P were measured atvarious time intervals by removing 20-�l samples, stopping the reactionin each by the addition of 10 �l of loading buffer (40% glycerol, 200 mM

EDTA, and 0.1% bromphenol blue) and placing the sample on ice untilanalysis by native gel electrophoresis and Coomassie Blue staining.RsbV-P and RsbV bands are easily separated on a 12% acrylamide gel(16). Eight time points were normally taken per reaction. The gels werescanned and intensities of the bands corresponding to the appearance ofRsbV were measured with Scion Image software. The values were thencompared with a standard curve of known concentrations of RsbVtreated under the same electrophoretic conditions.

Crystallographic Methods—The structure of N-RsbU was deter-mined by selenomethionine MAD phasing. To prepare selenomethionyl-labeled N-RsbU, the E. coli methionine auxotroph, B834 (DE3), wastransformed with pETNRsbUL, a pET15b derivative that directs the

TABLE IPlasmids and strains used in this study

Strain or plasmid Relevant genotype or features of plasmid Construction or Ref.

B. subtilisBSA46 PY22 SP� ctc::lacZ erm cat 86a 58BSA140 PY22 rsbU�NdeI rsbX::pWH25 SP� ctc::lacZ erm cat 86a spcb 14BSG14 PY22 rsbU�NdeI rsbX::pWH25 pGK01 SP� ctc::lacZ erm cat 86a spcb This studyBSG15 PY22 rsbU�NdeI rsbX::pWH25 pGK02 SP� ctc::lacZ erm cat 86a spcb This studyBSG16 PY22 rsbU�NdeI rsbX::pWH25 pGK03 SP� ctc::lacZ erm cat 86a spcb This studyBSG17 PY22 rsbU�NdeI rsbX::pWH25 pGK04 SP� ctc::lacZ erm cat 86a spcb This studyBSG18 PY22 rsbU�NdeI rsbX::pWH25 pGK05 SP� ctc::lacZ erm cat 86a spcb This studyBSG19 PY22 rsbU�NdeI rsbX::pWH25 pMB21 SP� ctc::lacZ erm cat 86a spcb This studyPB291 PB2 �rsbU::ermC SP�ctc-lacZ trpC2 24

E. coliBL21(DE3) F� ompT hsdSB(rB�mB�) gal dcm (DE3) NovagenB834(DE3) F� ompT hsdSB(rB�mB�) gal dcm met (DE3) NovagenTOP10 F� mcrA � (mrr-hsdRMS-mcrBC) �80lacZ�M15�lacX74recA1 deoR

ara�139 � (ara-leu)7697 galU galK rpsL (StrR) endA1 nupGInvitrogen

TG2 F� traD36 lacIq � (lacZ)M15 proA�B�/ supE hsd�5 thi � (lac-pro) �(srl-recA)306::Tn10 (TetR)

Bethesda Research Laboratories, USA

PlasmidsPAW70 bla cat amyE::(PA-rsbR-rsbS-rsbT-rsbU) 24PMLK bla cat amyE:: 23pET15b amp lacI PT7lac NovagenPETNRsbUB amp lacI PT7lac::rsbU1–112 26PETNRsbUL amp lacI PT7lac::rsbU1-84 This studyPETCRsbU amp lacI PT7lac::rsbU118–335 26pDG148 bla kan pleo lacI Pspac 21pGK01 bla kan pleo lacI Pspac::rsbU This studypGK02 bla kan pleo lacI Pspac::�rsbU1–38 This studypGK03 bla kan pleo lacI Pspac::�rsbU1–77 This studypGK04 bla kan pleo lacI Pspac::�rsbU1–93 This studypGK05 bla kan pleo lacI Pspac::�rsbU1–134 This studypMB21 bla kan pleo lacI Pspac::�rsbU1–19 This study

a The cat86 gene conferring chloramphenicol resistance is linked to the ctc::lacZ reporter gene fusion.b The spc gene conferring resistance to spectinomycin was introduced by a Campbell-like recombination into the B. subtilis chromosome and is

linked to the rsbU�NdeI allele.

Activation of a Stress-responsive Phosphatase 40929

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

IPTG-inducible expression of N-RsbU-(1–112) (26). Transformantswere initially grown in LB media, before harvesting, washing, andfinally inoculating selenomethionyl media. This culture was grown fora further 4 h before the addition of IPTG to a final concentration of 1 mM

and the cells were harvested by centrifugation 4 h later, and the cellpellets were stored at �80 °C overnight. Purification and crystallizationwere carried out as reported previously (26). The near complete incor-poration of selenomethionine was confirmed by mass spectrometry.

MAD diffraction data were collected at beamline BM14 of the ESRF,Grenoble, France, from a single crystal of SeMet N-RsbU at threedifferent wavelengths to �2.1-Å resolution. Each data set was individ-ually integrated and reduced using the HKL suite (27) scaling theBijvoet pairs separately. Native diffraction data were collected sepa-rately on ID14-EH2 to a resolution of 1.6 Å for refinement purposes.Data collection statistics are summarized in Table II. The structure wasdetermined with the program SOLVE in its automatic mode (28). Atotal of two heavy atom sites were found corresponding to the twointernal methionines, with a SOLVE Z-score of 9.1. MAD phasing fromthese sites produced an electron density map with a clear boundarybetween protein and solvent, and a mean FOM to 2.3 Å of 0.34. Thesephases were improved by density modification in RESOLVE (meanFOM of 0.47), producing an electron density map of sufficient quality forRESOLVE to automatically build amino acids 3–34 (helices 1–2) and44–58 (helix 3) with the correct sequence. After initial refinement inREFMAC (29), the higher resolution data became available and refine-ment proceeded against these data, where the Rfree flags were main-tained from the MAD data. Building of the rest of the structure, mostlyloops and helix 4, was performed by hand in QUANTA (Acclerys). Allother calculations were performed using programs from the CCP4 suite(30). The loop between helices 3 and 4 was built in two stages, com-mencing only when its path could be traced unequivocally. Successiverounds of rebuilding and refinement were interspersed until refinementconverged, at an Rwork of 19.6% and Rfree of 22.9%. Greater than 97% ofthe residues of the final model fall within the core region of the Ram-achandran plot, and the overall G factor is 1.6. Statistics for the finalmodel are presented in Table II. Atomic coordinates and structurefactor amplitudes have been deposited with the RCSB, with accessioncode 1w53.

RESULTS

Effect of Deletions of the N-terminal Domain of RsbU inVivo—If the N-terminal domain of RsbU indeed acts as aninhibitory element preventing activation of RsbU in the ab-sence of stress, partial or complete deletion of the N-terminalpart of RsbU should render the phosphatase constitutivelyactive. To test this hypothesis experimentally, a series of N-

terminal truncated RsbU variants were designed and theireffect on �B activity was tested both in exponentially growingcells and in cells exposed to ethanol, a well known environmen-tal stress stimulus. Specifically, the truncations were of thefirst 19, 38, 77, 93, and 134 amino acids. A wild-type copy andeach of the truncated rsbU genes were cloned by PCR into theself-replicating plasmid pDG148 and thus were placed underthe control of the IPTG-inducible promoter, Pspac. Plasmidswere transformed into a derivative of B. subtilis wild-typestrain BSA46, in which the chromosomal copy of rsbU had beeninactivated by a deletion of an internal NdeI fragment in rsbU(BSA140) (14). The expression of all truncated versions of RsbUon IPTG induction was verified by Western blot analysis withmonoclonal antibodies directed against RsbU (data not shown).�B activity was monitored under the same conditions with actc-lacZ reporter gene fusion. This fusion is known to be strictly�B-dependent (31) and thus the �-galactosidase activity ofthose strains is directly correlated to �B activity.

Surprisingly, although all the RsbU variants were ex-pressed, none of the strains harboring the plasmid-encodedtruncated rsbU alleles displayed any significant ctc-lacZ re-porter gene activity during exponential growth in the presenceof the inducer IPTG (data not shown). Only the strain carryinga plasmid-encoded full-length copy of rsbU displayed a modest�-galactosidase, and hence �B activity, in the presence of IPTG.Therefore, partial or complete deletion of the N-terminal do-main of RsbU does not render RsbU constitutively active invivo. Next, the ability of the truncated RsbU variants to medi-ate environmental stress-triggered activation of �B was tested.Cells grown in the presence or absence of the inducer IPTGwere exposed to 4% ethanol, a well known strong inducer of the�B regulon (32, 33). As expected, wild-type strain BSA46 dis-played strong transient induction of the ctc-lacZ fusion, whichpeaked 20 min after ethanol addition and was not altered bythe inclusion of the inducer IPTG (Fig. 2). The internal deletionin rsbU in strain BSA140 completely abrogated any �-galacto-sidase activity. Complementation of the rsbU disruption, byproviding a full-length copy of rsbU from plasmid pGK01 toform strain BSG14, resulted in an IPTG-dependent increase inctc-lacZ-encoded �-galactosidase activity that also peaked �20

FIG. 2. Effect of N-terminal truncations in RsbU on �B activity. A set of B. subtilis ctc-lacZ fusion strains was cultivated in a syntheticmedium (22) at 37 °C in the absence (A) or presence (B) of 1 mM IPTG. During exponential growth (time zero), ethanol was added to a finalconcentration of 4% (v/v) to induce the �B regulon via the environmental stress sensing pathway. Samples were removed for �-galactosidase assaysat the indicated time points. �-Galactosidase activities are expressed in Miller units (MU). The gray line indicates the growth of the wild-type strain(BSA46); growth of the rsbU mutant strains did not differ from that of the wild-type strain under these growth conditions. The activities of thectc-lacZ reporter gene fusion of the different strains are indicated by the following symbols: open squares, BSA46 (wild-type); filled squares, BSA140(rsbU�NdeI); open diamonds, BSG14 (rsbU�NdeI; PSPAC::rsbU); filled diamonds, BSG19 (rsbU�NdeI; PSPAC::�rsbU1–19); open circles, BSG15(rsbU�NdeI; PSPAC::�rsbU1–38); filled circles, BSG16 (rsbU�NdeI; PSPAC::�rsbU1–77); open triangles, BSG17 (rsbU�NdeI; PSPAC::�rsbU1–93);filled triangles, BSG18 (rsbU�NdeI; PSPAC::�rsbU1–134).

Activation of a Stress-responsive Phosphatase40930

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

min after exposure to ethanol, but with a peak height that wasroughly half that of BSA46. We do not consider that the differ-ence in �-galactosidase activity between strains BSG14 andBSA46 is significant; rsbU is not under the control of Pspac instrain BSA46, unlike in strain BSG14, and the relative levels ofRsbU in these strains, and hence the activity of �B, may differ.

In an independent genetic approach, the promoter and poly-cistronic coding sequences for rsbR, rsbS, rsbT, and rsbU werecloned into pMLK (23). This plasmid, which permits integra-tion at the amyE locus, was transformed into strain PB291(24), in which rsbU had previously been deleted in a ctc-lacZfusion background. This strain responds to ethanol shock inmuch the same way as BSA46 (data not shown), that is, it isindependent of IPTG. A second strain was constructed thatcontained rsbR-rsbS-rsbT, but here a truncated copy of rsbUwas cloned, corresponding just to the catalytic domain (resi-dues 118–335), instead of the full-length gene. This particulargene product could not induce �B activity with or without theimposition of stress (data not shown).

RsbU provided in trans can thus convey environmentalstress signals. In contrast, none of the strains carrying N-terminal truncated rsbU genes conferred constitutive activityon RsbU variants, nor could these strains activate �B in re-sponse to environmental stress. Thus any disruption to theN-terminal domain of RsbU, even a deletion of just the first 19amino acids (Fig. 2B), is detrimental to its function and hasprofound effects on �B activation.

Comparison of the Phosphatase Activities of RsbU and itsC-terminal Domain—The above results show that genetic de-letions in the regions of rsbU that encode its N-terminal do-main lead to an inability of the cells to activate �B in vivo andso to respond to stress. These results suggest that the activa-tion of the phosphatase function of RsbU has been affected bytruncations in its N-terminal domain, and that the C-terminaldomains created by deletion have no phosphatase activity. Weaddressed these points in vitro by measuring the phosphataseactivity of the recombinant C-terminal domain of RsbU (C-RsbU, residues 118–335), and comparing it to the activity ofRsbU alone. We also compared the phosphatase activity ofC-RsbU and of RsbU each in the presence of a 2-fold molarexcess of RsbT. The enzymatic assay revealed that while C-RsbU alone was active as a phosphatase toward RsbV-P, therate of dephosphorylation of RsbV-P was only 4-fold greaterthan that of full-length RsbU in the absence of RsbT (Fig. 3). Incontrast, the rate of dephosphorylation of RsbV-P by RsbU was�40-fold higher when stimulated by the addition of a 2-foldmolar excess of RsbT over RsbU (Fig. 3). C-RsbU displayedexactly the same, low phosphatase activity in the presence andabsence of RsbT (Fig. 3).

The observation of diminished phosphatase activity of C-RsbU in vitro is consistent with our genetic data, which revealthat N-terminal truncated RsbU has insufficient phosphataseactivity to trigger a �B-directed response to stress. Therefore, itappears unlikely that the function of N-RsbU is to inhibitC-RsbU, with RsbT relieving this inhibition via an interactionpredominantly, but not necessarily wholly, with the N-terminaldomain of RsbU. We conclude that the N-terminal domain ofRsbU does not exert an inhibitory influence on the C-terminalcatalytic domain but rather is absolutely required for the acti-vation of RsbU.

Interactions between RsbU and RsbT—In previous studies,the partner switching behavior of RsbW, RsbV, and �B wasmonitored by the use of non-denaturing polyacrylamide gelelectrophoresis (16). We applied the same technique to studythe interactions between different combinations of RsbT withRsbU, C-RsbU, and the N-terminal 112 residues of RsbU (N-

RsbU-(1–112)), which was previously identified as a proteolyti-cally stable fragment of RsbU (26). RsbT alone did not form asharp, single band in native gels, and a significant quantity ofmaterial was found in the well of the gel (Fig. 4, lane 2). Incontrast, RsbU and the isolated N- and C-terminal domainsmigrated well through the gel, with Rf (Rf � relative mobilityto the dye front) values of 0.41, 0.71, and 0.82 (Fig. 4, lanes 1,3, and 4).

When RsbT was mixed with an excess of RsbU (Fig. 4, lane5), an additional species was observed (Rf � 0.27) that mi-grated slower through the gel than RsbU, and which we pre-sume is a complex of RsbT and RsbU. Similarly, an additionalband Rf value of 0.41, presumably corresponding to the com-plex of RsbT and N-RsbU-(1–112), was observed when thesetwo proteins were mixed (Fig. 4, lane 6). However, when RsbTand C-RsbU were mixed (Fig. 4, lane 7), no additional specieswere observed, only those of the individual proteins. RsbT andRsbU are therefore capable of forming a complex stable enoughto be observed in the conditions of gel electrophoresis. The datapresented in Fig. 4 are consistent with the interaction betweenthese two proteins being mediated predominantly, or solely,through the N-terminal domain of RsbU. The isolated domainsof RsbU did not interact with each other under the electro-phoresis conditions to reconstitute a “full-length” RsbU (Fig. 4,lane 8). It is thus unlikely that the two domains of RsbU forma stable intramolecular complex. Other, multidomain proteinswhose structures undergo a conformational change as part of aregulatory mechanism cannot always be reconstituted from thepurified individual domains. For instance, the two isolated

FIG. 3. A comparison of the phosphatase activity of RsbU andC-RsbU. The phosphatase activities of 0.5 �M RsbU and 0.5 �M C-RsbUin the presence and absence of 1 �M RsbT were compared, where the yaxis corresponds to the rate of dephosphorylation of RsbV-P (mol ofRsbV�mol of phosphatase�1 s�1). The rates of dephosphorylation ofRsbV-P were calculated by measuring the intensity of bands corre-sponding to RsbV on non-denaturing polyacrylamide gels over a seriesof time points of the reaction. The activity of C-RsbU was unaltered bythe presence of N-RsbU-(1–112), or both RsbT and N-RsbU-(1–112)(data not shown). In this study, when RsbT was present in a 2-foldmolar excess over RsbU, the phosphatase activity of RsbU was stimu-lated by �40-fold. This stimulation is greater than that previouslyreported (7), where the phosphatase activity of RsbU was stimulated�7-fold in the presence of a 5-fold molar excess of RsbT. The differencebetween these amplification factors might come from the fact that wehave discovered that RsbT, which is quite unstable and has a tendencyto precipitate, behaves much better when ATP or ADP is addedthroughout the purification process. We therefore might have used amore “active” RsbT in our experiments. The results shown are theaverage of four independent experiments and the error bars representthe S.D. � mean.

Activation of a Stress-responsive Phosphatase 40931

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

domains of Spo0A do not interact under native gel electro-phoresis conditions.2

Stoichiometry and Stability of the RsbU�RsbT Complex—Wehave demonstrated that the presence of a 2-fold molar excess ofRsbT over RsbU increased the latter’s phosphatase activitysome 40-fold (Fig. 3). To determine the ratio of RsbU:RsbT thatis required for maximum activation of RsbU, the rate of de-phosphorylation of RsbV-P by RsbU was measured in the pres-ence of increasing concentrations of RsbT. The phosphataseactivity of RsbU increased linearly until the ratio betweenRsbT and RsbU reached 6:1, and reached its maximum at aratio of 10:1 (Fig. 5). This maximum activity was �200 timeshigher than that of RsbU in the absence of RsbT. Either RsbUbinds 6 or more molecules of RsbT to gain maximum activation,or the ratio at which maximal activity is observed is a reflectionof a low-affinity interaction between RsbT and RsbU.

Because it was possible by native gel electrophoresis to de-tect a complex between RsbU (or N-RsbU-(1–112)) and RsbT,the band corresponding to the RsbU�RsbT complex was excisedfrom the gel, and the proteins within the band were electro-eluted and then analyzed by SDS-PAGE. Bands correspondingto RsbU and RsbT were observed, and their relative intensitieswere measured after digitization of the gel. By comparison toknown standards of RsbU and RsbT, we conclude that thestoichiometry of the RsbU�RsbT complex under these condi-tions is 1:1.

To characterize the interactions between RsbU and RsbTfurther, we used gel filtration chromatography to estimate thesize of the purified individual proteins, isolated domains, andcomplexes. The same technique showed previously that RsbT isa monomer (18). The molecular mass of RsbU was found to bearound 80 kDa, a value that corresponds to a dimer of RsbUcomposed of 38-kDa subunits (data not shown). This result is inclose agreement with previous gel filtration experiments per-formed with a B. subtilis cellular extract from which RsbUeluted as a protein of �90 kDa (34). The apparent molecularmass of the N-terminal 112 residues of RsbU was 34 kDa,suggesting that this molecule is dimeric, or perhaps trimeric.The size of the C-terminal construct of RsbU (residues 118–

335, C-RsbU) predicted from its gene sequence is 25 kDa, whichwas consistent with the gel filtration elution profile that indi-cated that C-RsbU is monomeric. Hence it would appear thatthe dimerization determining motifs of RsbU are restricted tothe N-terminal domain. Despite testing a variety of differentconditions (pH, NaCl, ATP, and Mg2� concentration), we couldnot isolate any complex by size exclusion chromatography, be itbetween RsbT and RsbU, or between RsbT and N-RsbU-(1–112). Given that the measured stoichiometry of the RsbU�RsbTcomplex is 1:1, yet RsbT must saturate RsbU for full activation,and that the RsbU�RsbT complex could not be isolated by gelfiltration, these observations argue that there is only a weakaffinity interaction between RsbU and RsbT.

Crystal Structure of N-RsbU-(1–112)—To understandfurther the function of the N-terminal domain of RsbU, thestructure of N-RsbU-(1–112) was determined by x-ray crystal-lography, using the MAD technique from crystals of selenome-thionyl-labeled N-RsbU-(1–112), and refined against diffrac-tion data to 1.6 Å (Fig. 6A). Statistics of the diffraction data andfinal refined model are presented in Table II. The electrondensity map does not reveal the conformation of residues 85–112, and these amino acids are missing from the final struc-ture; we assume them to be disordered. One molecule ofN-RsbU-(1–112) comprises four anti-parallel �-helices, ar-ranged into a rough “L” shape (Fig. 6B), forming a four-helicalbundle. Such bundles are found commonly in biology in avariety of different circumstances; they are thermodynamicallyfavored in protein folding, as seen in the globin superfamily,and also mediate interactions with themselves in dimer inter-faces and with other molecules, such as DNA or proteins, forinstance in Spo0B, homeodomains, and focal adhesion kinase.To identify structural neighbors, the structure of N-RsbU wassubmitted to the DALI (35) server. More than 460 structureswere identified as having structural similarity to N-RsbU, withZ-scores of 2.0 or greater. Not surprisingly, the sequence iden-tities between N-RsbU and its structural neighbors were low,mostly less than 10%. No single identifiable biological themelinked the structures, and examples were found from the con-texts where four helical bundles are prevalent (e.g. residuesfrom hemoglobin (Z-score 3.1, Protein Data Bank code 1CG5),2 R. J. Lewis, unpublished observations.

FIG. 4. An analysis of the protein-protein interactions betweenRsbT and RsbU by native gel electrophoresis. Each lane contains�10 �g of total protein. Lane 1, RsbU alone; lane 2, RsbT alone; lane 3,N-RsbU-(1–112) alone; lane 4, C-RsbU alone; lane 5, RsbU and RsbT;lane 6, N-RsbU-(1–112) and RsbT; lane 7, C-RsbU and RsbT; lane 8,N-RsbU-(1–112) and C-RsbU. The positions of protein�protein com-plexes are marked with an asterisk, the individual proteins are markedwith an arrow. Note that in lanes 2 and 7, RsbT does not enter the gel,and can be seen stuck in the well. When RsbT is mixed with either RsbU(lane 5), or N-RsbU-(1–112) (lane 6), RsbT enters the gel to form a newprotein�protein complex as marked by the asterisks.

FIG. 5. Activation of RsbU as a function of the concentration ofRsbT. The phosphatase activity of 0.25 �M RsbU against 30 �M of thesubstrate, RsbV-P was monitored as a function of increasing concentra-tions of RsbT. The rates of dephosphorylation of RsbV-P were calculatedby measuring the intensity of bands corresponding to RsbV on non-denaturing polyacrylamide gels over a series of time points of thereaction. The rate of dephosphorylation of RsbV-P (mol of RsbV�molphosphatase�1 s�1) is plotted on the y axis as a function of the increas-ing RsbT:RsbU ratio along the x axis.

Activation of a Stress-responsive Phosphatase40932

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Spo0B (2.5, Protein Data Bank code 1IXM), engrailed home-odomain (3.2, Protein Data Bank code 2HDD), and FAK tar-geting domain (2.0, Protein Data Bank code 1K04)).

Potentially relevant structural matches found by DALI in-clude domain three of �A (�3) from Thermus aquaticus (2.1,Protein Data Bank code 1KU2), and representatives from thearmadillo (�-importin, 4.5, Protein Data Bank code 1QGR;�-catenin, 4.2, Protein Data Bank code 1DOV; �-catenin, 2.1,Protein Data Bank code 3BCT), HEAT (PP2A, 3.8, ProteinData Bank code 1B3U), and tetratricopeptide repeat (TPR)(PP5, 3.3, Protein Data Bank code 1A17) families of helicalrepeat proteins; these are scaffolding molecules that mediateprotein-protein interactions (36). �3, which harbors residuesthat are important for the recognition of the �10 region of thepromoter, is also found in contact with SpoIIAB residues in thecrystal structure of the �F�SpoIIAB complex (37). A recentstudy of the SpoIIAB homologue and RsbU activator, RsbT,suggests that RsbT residues Arg19-Gln20, Arg23-Asn24, andAsp35-Gln36 are involved in RsbT-RsbU interactions (38).These residues are equivalent to those in SpoIIAB that bind �F.However, the SpoIIAB-contacting residues in �F have spatialequivalents in N-RsbU that are not solvent-accessible and thuscan only bind RsbT after significant conformational changes

occur in N-RsbU. Therefore, the functional significance of thesimilarity between N-RsbU and �3 remains to be elucidated.

On examination of the crystal packing, it was evident thatN-RsbU-(1–112) had crystallized as a dimer around a crystal-lographic 2-fold axis. 28% of the available surface area, some1600 Å2, is buried in the dimer interface; 80% of the atoms inthe interface are non-polar and no water molecules are ob-served in the interface at the resolution of 1.6 Å. This extensivedimer interface is composed of residues from all four helices ofeach protomer, but is dominated by helix 4, which contributesabout half of the buried surface in the dimer interface, whereashelices 1 to 3 each contribute �1/6th. Five hydrogen bondsbetween residues in protomer one and two are duplicated bythe 2-fold symmetry of the interface, to make a total of 10 thatstabilize the dimer. The residues involved, Gln31, Thr37, His51,Tyr59, Asp65, Ser69, Glu75, and Tyr80, are conserved across theRsbU family, except for Thr37 and the interacting pair of Tyr59

and Asp65 (Fig. 7). In this latter case, a change at one positionin RsbU orthologues is compensated for by an alteration in theother position such that either hydrogen bonding or hydropho-bic packing is conserved (e.g. pairings are found of Asp and Asn,or Gly and Phe, at positions 59 and 65 in the orthologues ofRsbU). Variation to Leu, Phe, Ala, or Val is permitted at Thr37

FIG. 6. The structure of N-RsbU. Aribbon diagram color-ramped from blue tored, from the N to C terminus in eachprotomer is drawn in three, mutually per-pendicular views with the middle panelas the reference point (A). Helices 1(amino acids 2–20) and 2 (24–40), andhelices 3 (44 and 58) and 4 (64 and 84)form two pairs, which are separated byabout 45° (C). In panel B, the view isalong the 2-fold symmetry axis of thedimer, revealing the L-shaped form ofeach protomer; in panel C, the relativedisposition of helices 1 and 2, relative tohelices 3 and 4� (where � indicates thathelix 4 is from the alternative protomer inthe dimer) is illustrated. This figure wascreated with BOBSCRIPT (56) andRaster3D (57).

TABLE IIData collection and refinement statistics

Values in parentheses refer to the highest resolution shell. r.m.s.d., root mean square deviation.

Data set

Se-Peak Se-Edge Se-Remote Native

Data collectionWavelength (Å) 0.97922 0.97936 0.93028 0.933Resolution limits (Å) 50–2.08 50–2.09 50–2.25 31–1.60Number of measurements 52293 34571 28893 48906Number of reflections 8366 8310 6804 10180Completeness (%) 95.9 (84.3) 96.7 (96.5) 98.6 (86.9) 100 (100)Rsym

a (�100) 10.3 (36.9) 9.7 (36.2) 8.9 (20.6) 7.2 (16.4)I/(�)I 15.2 (3.4) 13.0 (2.8) 14.0 (7.7) 4.4 (4.1)

RefinementNumber of non-H atoms 784Average B factor all atoms (Å2) 20.6Rwork

b (%) 19.8Rfree

c (%) 22.9r.m.s.d. bond distances (Å) 0.019r.m.s.d. bond angles (°) 1.61r.m.s.d. B factor main chain bonds (Å2) 1.53r.m.s.d. B factor side chain bonds (Å2) 2.92

a Rsym � �I � I��/ I, where I� is the average intensity from multiple observations of symmetry related reflections.b Rwork � h�Fo� � �Fc�/ h�Fo�, where Fo and Fc are the observed and calculated structure factor amplitudes.c Rfree was calculated with a 5% randomly-selected subset of the reflections not used in refinement.

Activation of a Stress-responsive Phosphatase 40933

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

because residues in this position can interact with Tyr80 eitherby forming a hydrogen bond, or by hydrophobic packing. Otherresidues found at the dimer interface are well conserved acrossthe RsbU orthologues sequenced to date.

In a previous study, the mutation P44R was identified in agenetic screen for suppressors of a deletion in rsbX (39). Strainsdeleted for rsbX exhibit constitutively high �B activation, andrsbU44PR suppresses this activity. This mutation also abol-ishes the activation of �B on the imposition of stress. In thestructure of N-RsbU, Pro44 is situated at the dimer interface,sandwiched between conserved hydrophobic residues includingMet77 and Tyr84, and in conjunction with Glu45 it stabilizes theN-terminal end of helix 3. The data of Smirnova et al. (39)indicate that although rsbU44PR accumulates in the cell, itcannot be activated, because the ability of RsbU and RsbT tointeract has been destroyed by this mutation. Mutation fromproline to arginine at this position is highly likely to disrupt thedimer interface as well as to destabilize helix 3. The inability ofrsbU44PR to be activated by RsbT could be explained by theloss of dimerization of RsbU, affecting the binding surface onRsbU for the activator RsbT.

A Potential RsbT Binding Surface on N-RsbU—The P44Rmutation in RsbU that affects activation by RsbT led us toconsider further the interaction between RsbU and RsbT. Ananalysis of the completed genomes of Gram� microorganismsthat encode a �B orthologue reveals that not all of the Rsbproteins are conserved. For instance RsbR, RsbS, RsbT, andRsbX have not been identified in the genomes of S. aureus andStaphylococcus epidermidis. Presumably, in those bacteriawhere no activator of RsbU has been discovered, RsbU is reg-ulated in a different way to that in B. subtilis. Therefore, acomparison of the amino acid sequence of RsbU orthologuesthat originates from bacterial species that contain an RsbTactivator, and those that do not, may reveal the RsbT-interact-ing surface in RsbU. The sequence alignment of the N-terminaldomains of RsbU orthologues (Fig. 7) provides insight into thestructure and function of RsbU. Those residues that are com-pletely conserved in the alignment are found at the dimer

interface, participating in either hydrogen bond (e.g. His51 andGlu75�, where � indicates that this residue is found in the otherprotomer in the dimer), or van der Waals (e.g. Tyr11, Leu27, andLeu73�) contacts. Moreover, a larger amino acid than glycine atposition 81� could not accommodate the approach of the longcarboxylic acid side chain of Glu45, which also serves to N-caphelix 3. Other residues that are particularly poorly conservedare located mainly in loops between the secondary structuralelements or on the solvent-exposed faces of the four helices.

The residues that are solvent-accessible and conserved inRsbT encoders (Bacilli and Listeriae), and not necessarily so inRsbT non-encoders (Staphylococci), are highlighted in the se-quence alignment (Fig. 7). Glu24, Tyr28, and Arg35 are found ina discontinuous group on one face of helix 2: the break atposition 31 coincides with a 1-residue lysine insertion in thesequences of RsbU from the Staphylococci. In conjunction withthe hydrophobic residues Ile74�, Ile78�, Met82�, and Ala83� fromhelix 4�, these residues form two prominent, parallel surfaceridges of edge length �17 Å (Fig. 8A). These surface ridges areonly formed as a consequence of the dimerization of N-RsbUand in effect they form a �10-Å wide groove, the floor of whichcomprises residues Ser34, Ile38, Val76�, Gly79�, and Tyr80�,which are well conserved across all RsbU orthologues. Overall,the prominent surface features of N-RsbU from B. subtilis areunlikely to be conserved in the Staphylococci, leading us toconclude that these surface patches may represent at least partof the binding surface for RsbT on RsbU. For instance, replace-ment of Tyr18 by serine in the Staphylococci would undoubtedlyalter the local structure and molecular surface. Calculation ofthe molecular surface after replacement of the tyrosine at thisposition by serine reveals that a cavity is formed (Fig. 8B).

A Comparison of RsbT Binding by N-RsbU-(1–84) andN-RsbU-(1–112)—In Fig. 4, we monitored the binding of RsbTby RsbU and N-RsbU-(1–112). However, the structure ofN-RsbU-(1–112) did not reveal the conformation of residues85–112, because this part of the protein was disordered in thecrystal. Other residues that are conserved only in those bacte-ria that code for rsbT include Glu86, Arg91, Ile97, and Ser99 (Fig.7), which cannot be modeled in the present structure. Theseamino acids may also play a role in RsbT binding. To investi-gate whether residues 85–112 of RsbU contribute significantlyto RsbT binding, we constructed a truncated form of N-RsbUthat corresponds precisely to the ordered portion of this crystalstructure, i.e. residues 1–84. We monitored whether N-RsbU-(1–84) could recruit RsbT in a binding assay, as analyzed bynon-denaturing gel electrophoresis. N-RsbU-(1–84) runs a lit-tle quicker in electrophoresis than N-RsbU-(1–112) in the ab-sence of RsbT, but in its presence an additional electrophoreticspecies can be observed for both constructs of N-RsbU, whichwe conclude corresponds to N-RsbU�RsbT complexes (Fig. 9a).Note that the band corresponding to the N-RsbU-(1–84)�RsbTcomplex in lane 1 is slightly more diffuse than the band in lane2 of the N-RsbU-(1–112)�RsbT complex. The relative diffuse-ness may reflect the fact that the interaction of RsbT withN-RsbU-(1–84) is slightly weaker than that between RsbT andN-RsbU-(1–112).

To assess whether there is a significant difference in the bind-ing affinity of the two N-RsbU constructs for RsbT, we designeda competition assay in which a molar equivalent of N-RsbUdomains in comparison to RsbU was added to RsbV-P dephos-phorylation reactions. The initial rate of RsbV-P dephosphoryl-ation was reduced by 45% for N-RsbU-(1–84) and 40% forN-RsbU-(1–112) (Fig. 9b). The results of these experiments indi-cate that both N-RsbU-(1–84) and N-RsbU-(1–112) are compe-tent in binding RsbT, and that the major RsbT-binding determi-nants are found in those amino acids that could be located in the

FIG. 7. Sequence alignment of N-RsbU orthologues. The con-struct used for structure determination, the first 112 amino acids ofRsbU, is shown with every 10th amino acid in B. subtilis RsbU num-bered and the secondary structural elements drawn above, with �-hel-ices represented by blue bars. Residues colored green are totally con-served across the RsbU family, those in red are conserved in thosebacteria that encode rsbT, but not necessarily conserved in those bac-teria that do not. The residue in black is Pro44, mutation of which to Argsuppresses a deletion in rsbX (39). Gaps in the alignment are repre-sented by -; residues that contribute to the dimer interface are high-lighted by a * for hydrogen bond donors/acceptors and a ^ for hydro-phobic packing. The sequences listed are B. subtilis (BACSU), Bacilluslicheniformis (BACLI), Bacillus halodurans (BACHA), Oceanobacillusiheyensis (OBACI), L. monocytogenes (LIMON), Listeria innocua(LIINN), which all encode rsbT and S. aureus (STAUR) and S. epider-midis (STEPI), which do not.

Activation of a Stress-responsive Phosphatase40934

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

crystallographic electron density, residues 1–84. Although wecannot rule out the possibility that residues 85–112 are involvedin the binding of RsbT during the activation of RsbU, we canconclude that any interaction must be relatively weak.

DISCUSSION

PP2C-type phosphatases act in the signaling pathways thatregulate the response to stress in eukaryotes (40, 41) as well asin B. subtilis and its close relatives (7). The B. subtilis PP2C-type phosphatase, RsbU, activates the general stress � factor,�B, following the imposition of environmental stress (see Fig.1). Induction of the large �B-dependent regulon provides thecell with a multiple and pre-emptive stress resistance (42), butalso constitutes a considerable burden. Thus, tight control ofthe activity of �B, and therefore RsbU, is crucial. The activitiesof eukaryotic phosphatases are known to be controlled by avariety of mechanisms, including post-translational modifica-

tion and the binding of accessory proteins, but little is known ofthe regulation of PP2C phosphatases (43). For instance, theeukaryotic cell cycle phosphatases Cdc25A (44) and Cdc25C(45) are activated by phosphorylation. In contrast, the kinaseactivity of RsbT is not required for activation of RsbU (19).

A simple activation mechanism for RsbU, where the N-ter-minal domain suppresses the function of the C-terminal do-main until a conformational change occurs in RsbU on thebinding of the activator, RsbT, would be an elegant solution tothe problem of control. Such a mechanism has already beenobserved for human and rabbit protein phosphatase 5 (PP5),which is stimulated by the interaction of arachidonic acid withthe TPR domain of PP5. This interaction drives a conforma-tional change in PP5 that overcomes the TPR-mediated inhi-bition of the phosphatase domain. In this instance, the enzy-matic activity of arachidonic acid-stimulated PP5 approaches

FIG. 8. The RsbT-binding surface?A, the molecular surface of the N-RsbUdimer is colored mostly blue: those partsof the surface colored lime correspond toresidues that are completely conserved inthe sequence alignment in Fig. 7; notethat few of them map to the surface, andthe ones that do (Leu27 and Glu75�; Gly81

and Glu45�) are part of the dimer inter-face, situated between helices 3 and 4�.The red patches of the molecular surfacerelate to those amino acids that are con-served in bacteria that encode RsbT intheir genomes, but are not necessarilyconserved in bacteria that do not harboran rsbT gene. B, the inset reveals a sec-tion of the surface in the vicinity of Tyr18,where the tyrosyl side chain has been re-placed with a seryl, with the result that acavity is formed. In both panels, the sameorientation as panel A in Fig. 6 is used,and this figure was prepared with Py-MOL (www.pymol.org).

FIG. 9. An analysis of the protein-protein interactions between RsbT and N-RsbU. A, a native gel binding assay, as described for Fig.4, where the ability of RsbT to bind to N-RsbU-(1–84) and N-RsbU-(1–112) was monitored. Lane 1, N-RsbU-(1–84) and RsbT; lane 2, N-RsbU-(1–112) and RsbT. The position of the free individual proteins are marked with an arrow, protein�protein complexes are marked with an asterisk.Each lane contains �10 �g of total protein. B, a competition assay in which the initial rate of RsbV-P dephosphorylation by 0.5 �M RsbU stimulatedby 0.5 �M RsbT was measured in the presence of 0.5 �M of the two N-RsbU constructs. Along the x axis in lane 1, RsbU alone; lane 2, RsbUstimulated by RsbT; lane 3, RsbU stimulated by RsbT in the presence of N-RsbU-(1–84); lane 4, RsbU stimulated by RsbT in the presence ofN-RsbU-(1–112). The y axis corresponds to the percentage of RsbV-P dephosphorylation activity normalized to RsbU stimulated by RsbT. Theresults shown are the average of three independent experiments and the error bars represent the S.D. � mean.

Activation of a Stress-responsive Phosphatase 40935

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

that of the isolated phosphatase domain (46). However, in vitrothe isolated catalytic domain of RsbU has a very low phospha-tase activity in comparison to RsbT-activated full-length RsbU(Fig. 3). Moreover, N-terminal truncated RsbU proteins cannotinduce �B activity in vivo, and furthermore, the basal level of�B activity in these strains resembles that of wild type (Fig. 2).The N-terminal domain of RsbU is therefore absolutely re-quired for RsbU activation, by recruiting RsbT (Fig. 3).

The results presented here support and extend the earlierconclusions by Kang et al. (19) that RsbT activates RsbU byforming a protein�protein complex. In this study, we demon-strate that the association is mediated predominantly, but notnecessarily exclusively, by the first 84 amino acids of the N-terminal domain of RsbU (Figs. 4 and 9). The crystal structureof the N-terminal domain of RsbU reveals that only a dimericform of RsbU can form what we believe to be an RsbT-bindingsurface. Deletion of the N-terminal part of rsbU (this study), ormutation of Pro44 in the dimer interface (39), has deleteriouseffects on the stress response of B. subtilis, presumably becausethe dimer form of RsbU is destroyed. The residues that wouldappear to mediate these protein-protein interactions are onlyconserved in RsbU orthologues from bacteria that include rsbTin their genomes (Fig. 7).

What molecular mechanism is used in the activation of thephosphatase domain of RsbU during the recruitment of RsbTby N-RsbU? RsbU is dependent on manganese, and one methodof RsbU activation might involve the modulation of the Km forMn2� in the presence of RsbT, leading to a more efficientdephosphorylation of RsbV-P. This model is not without prec-edent in the PP2C phosphatase family; the catalytic subunit ofbovine mitochondrial pyruvate dehydrogenase phosphatasedisplays a Km for magnesium that increases in the presence ofits regulatory subunit (47). However, the Km value for Mn2� ofRsbU, of RsbU in the presence of RsbT, and of the C-terminaldomain of RsbU alone were found to be 0.96 mM in all threecases (data not shown) and thus similar to that observed forother PP2C phosphatases (48, 49). It would therefore seemunlikely that RsbU is regulated by changes in the Km for itsrequisite co-factor.

An alternative mechanism might entail the modulation ofthe oligomeric state, a common means of control. From solutionmeasurements, we have concluded that RsbU, in the absence ofRsbT, is a dimer, and the structure of N-RsbU reveals a signif-icant dimer interface (Fig. 6). The thermodynamic barrier of

dimer dissociation in this instance is likely to be considerable.Unless the reaction is driven by the free energy of hydrolysis ofATP, for instance, the apparent robustness of the N-RsbUdimer interface should exclude an activation mechanism wherethe presence of RsbT dissociates the N-RsbU dimer, leading toenzymatic stimulation.

Alternatively, the formation of the RsbT�RsbU binary com-plex might be a prerequisite for phosphatase activation. As inN-RsbU, the non-catalytic domains from phosphatases 2A (50),2C (51), and 5 (52) are also all �-helical (Fig. 10). Despite thefact that there is little meaningful sequence homology betweenthese domains and RsbU, these non-catalytic domains were allfound with DALI, with Z-scores of 3.8 (PP2A), 3.2 (PP2C), and2.2 (PP5). For instance, helices 1, 3, and 4 of N-RsbU werealigned against all three helices in the regulatory domain ofPP2C, which has been proposed to play a role in the determi-nation of substrate specificity (51). Similarly, N-RsbU alsomatched to the TPR motifs from PP5, which are used to medi-ate macromolecular interactions and the assembly ofprotein�protein complexes. Furthermore, there is structuralsimilarity between N-RsbU and the HEAT repeat units ofPP2A, which coordinate the assembly of the phosphatase andregulatory subunits into a functional heterotrimer (50). It ap-pears that the N-terminal domain of RsbU may have evolved toperform a similar scaffolding function, recruiting RsbT, andperhaps it is this binary complex that actually recognizes theRsbV-P substrate.

In the absence of a stable RsbT�RsbU complex, determinationof the molecular basis of phosphatase activation has provenelusive. The present study reveals that for maximal activationof RsbU, at least a 6-fold molar excess of RsbT must be present(Fig. 5). We have previously demonstrated that it is likely thatthere are at least six molecules of RsbT bound per stressosomeprior to the induction of stress signals (18). It is perhaps notcoincidental that the same number of molecules of RsbT that issequestered by the stressosome is required to activate fullyRsbU. The expression of �130 �B-dependent genes in responseto stress is an energetically demanding process, and thereforehas to be kept under strict control. The partnership betweenthe stressosome and RsbU may actually sense the amplitude ofthe stress signal, and act as a tuning system to regulate thelevel of �B activity in accordance to the level of stress. Thispartnership could also act as part of a damping mechanism, ifa threshold level of RsbU activity is necessary to induce �B

FIG. 10. Evolution of phosphataseregulatory domains. Ribbon diagramsillustrating the structural similarities be-tween the regulatory domains of proteinphosphatases 2A, 2C, and 5, which havebeen aligned against the equivalent sec-tion of the structure of N-RsbU. Thealigned parts of structure are colored ingold, the non-equivalent parts are coloredsemi-transparent blue.

Activation of a Stress-responsive Phosphatase40936

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

activity, ensuring that a stress-response is triggered only undercertain conditions. RsbX, the phosphatase for RsbS-P, is re-quired to restrict the activity of �B, and to re-set the activity of�B to pre-stress levels after the imposition of stress (53). Incontrast to the stable interactions observed between RsbW and�B or RsbV, relatively weak interactions between RsbT andRsbU may have evolved to permit the rapid and transientactivation of the �B-dependent general stress regulon thatreaches its peak within 20 min in laboratory conditions.

This study reveals that the N-terminal domain of RsbUshares structural similarity to non-catalytic domains fromprotein phosphatase families 2A, 2C, and 5. We propose thatthe structural similarity extends to a functional one. Thephosphatase regulatory domains are found across the plantand animal kingdoms, and it is thus likely that these phos-phatases share a common mechanism of control with RsbU,the binding of an additional regulatory subunit. The struc-tural similarity between N-RsbU and �3 may have functionalas well as evolutionary significance. RsbT is related to agroup of anti-� factors that inhibit � factor activity by bind-ing to �3. The partner switching of RsbT and RsbU appears tohave developed in parallel to that of � and anti-� factors, touse similar structural motifs in complex formation. Unlikethe robust ��anti-� factor complexes, however, the molecularsurfaces involved in the formation of the RsbU�RsbT complexhave evolved mutually to weaken their interaction as part ofthe transient �B-activation mechanism.

Nonetheless, several questions remain unanswered. Doesthe ephemeral association of RsbT with RsbU drive a relativelylong-lived conformational change in RsbU that is necessary foractivation, or is the RsbT�RsbU binary complex active as aphosphatase? If so, is there a measurable interaction of thebinary complex with the RsbV-P substrate? Experiments areunderway to answer these questions, and to confirm if theresidues we identified by genomic sequence analysis are uti-lized in the recruitment of RsbT by RsbU.

Acknowledgments—We thank Chet Price for the gift of strain PB291and plasmid pAW70, and William G. Haldenwang for providing themonoclonal antibodies directed against RsbU. We are grateful for ac-cess to the ESRF, and beamline support from Martin Walsh (BM14) andEd Mitchell (ID14-EH2), and James Murray and Lorraine Hewitt forhelp during data collection. We also thank Harry Gilbert andJan Pane-Farre for their incisive comments on the manuscript.

Note Added in Proof—We were alerted by (and are grateful to) Dr.Alexey Murzin to the structural and functional similarity betweenN-RsbU and the C-terminal domain of the clock protein KaiA. Bothdomains form similar dimers, and the KaiC-binding site in C-KaiA (seeVakonakis, I., and LiWang, A. C. (2004) Proc. Natl. Acad. Sci. U. S. A.101, 10925–10930) is in the equivalent location to the proposed RbsTbinding site on the N-RsbU dimer.

REFERENCES

1. Johnson, L. N., and Lewis, R. J. (2001) Chem. Rev. 101, 2209–22422. Kennelly, P. J. (2001) Chem. Rev. 101, 2291–23123. Zhu, Y., Qin, L., Yoshida, T., and Inouye, M. (2000) Proc. Natl. Acad. Sci.

U. S. A. 97, 7808–78134. Obuchowski, M., Madec, E., Delattre, D., Boel, G., Iwanicki, A., Foulger, D.,

and Seror, S. J. (2000) J. Bacteriol. 182, 5634–56385. Duncan, L., Alper, S., Arigoni, F., Losick, R., and Stragier, P. (1995) Science

270, 641–6446. Vijay, K., Brody, M. S., Fredlund, E., and Price, C. W. (2000) Mol. Microbiol.

35, 180–188

7. Yang, X., Kang, C. M., Brody, M. S., and Price, C. W. (1996) Genes Dev. 10,2265–2275

8. Gaidenko, T. A., Kim, T. J., and Price, C. W. (2002) J. Bacteriol. 184,6109–6114

9. Lucet, I., Feucht, A., Yudkin, M. D., and Errington, J. (2000) EMBO J. 19,1467–1475

10. Hecker, M., and Volker, U. (2001) Adv. Microb. Physiol. 44, 35–9111. Price, C. W. (2002) Bacillus subtilis and Its Closest Relatives: from Genes to

Cells (Sonenshein, A. L., Hoch, J. A., and Losick, R., eds) pp. 369–384, ASMPress, Washington, D. C.

12. Brody, M. S., Vijay, K., and Price, C. W. (2001) J. Bacteriol. 183, 6422–642813. Kang, C. M., Brody, M. S., Akbar, S., Yang, X., and Price, C. W. (1996) J.

Bacteriol. 178, 3846–385314. Volker, U., Dufour, A., and Haldenwang, W. G. (1995) J. Bacteriol. 177,

114–12215. Alper, S., Dufour, A., Garsin, D. A., Duncan, L., and Losick, R. (1996) J. Mol.

Biol. 260, 165–17716. Delumeau, O., Lewis, R. J., and Yudkin, M. D. (2002) J. Bacteriol. 184,

5583–558917. Alper, S., Duncan, L., and Losick, R. (1994) Cell 77, 195–20518. Chen, C.-C., Lewis, R. J., Harris, R., Yudkin, M. D., and Delumeau, O. (2003)

Mol. Microbiol. 49, 1657–166919. Kang, C. M., Vijay, K., and Price, C. W. (1998) Mol. Microbiol. 30, 189–19620. Kunst, F., et al. (1997) Nature 390, 249–25621. Stragier, P., Bonamy, C., and Karmazyn-Campelli, C. (1988) Cell 52, 697–70422. Stulke, J., Hanschke, R., and Hecker, M. (1993) J. Gen. Microbiol. 139,

2041–204523. Karow, M. L., and Piggot, P. J. (1995) Gene (Amst.) 163, 69–7424. Wise, A. A., and Price, C. W. (1995) J. Bacteriol. 177, 123–13325. Miller, J. H. (1972) Experiments in Molecular Genetics, Cold Spring Harbor

Laboratory, Cold Spring Harbor, NY26. Dutta, S., and Lewis, R. J. (2003) Acta Crystallogr. Sect. D Biol. Crystallogr.

59, 191–19327. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–32628. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol.

Crystallogr. 55, 849–86129. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D

Biol. Crystallogr. 53, 240–25530. CCP4 (1994) Acta Crystallogr. D Biol. Crystallogr. 50, 760–76331. Igo, M., and Losick, R. (1986) J. Mol. Biol. 191, 615–62432. Boylan, S. A., Redfield, A. R., Brody, M. S., and Price, C. W. (1993) J. Bacteriol.

175, 7931–793733. Volker, U., Volker, A., Maul, B., Hecker, M., Dufour, A., and Haldenwang,

W. G. (1995) J. Bacteriol. 177, 3771–378034. Dufour, A., Volker, U., Volker, A., and Haldenwang, W. G. (1996) J. Bacteriol.

178, 3701–370935. Holm, L., and Sander, C. (1993) J. Mol. Biol. 233, 123–13836. Groves, M. R., and Barford, D. (1999) Curr. Opin. Struct. Biol. 9, 383–38937. Campbell, E. A., Masuda, S., Sun, J. L., Muzzin, O., Olson, C. A., Wang, S., and

Darst, S. A. (2002) Cell 108, 795–80738. Woodbury, R. L., Luo, T., Grant, L., and Haldenwang, W. G. (2004) J. Bacte-

riol. 186, 2789–279739. Smirnova, N., Scott, J., Volker, U., and Haldenwang, W. G. (1998) J. Bacteriol.

180, 3671–368040. Shiozaki, K., and Russell, P. (1995) EMBO J. 30, 492–50241. Gaits, F., Shiozaki, K., and Russell, P. (1997) J. Biol. Chem. 30, 17873–1787942. Volker, U., Maul, B., and Hecker, M. (1999) J. Bacteriol. 181, 3942–394843. Hunter, T. (1995) Cell 80, 225–23644. Hoffmann, I., Draetta, G., and Karsenti, E. (1994) EMBO J. 13, 4302–431045. Hoffmann, I., Clarke, P. R., Marcote, M. J., Karsenti, E., and Draetta, G. (1993)

EMBO J. 12, 53–6346. Chen, M. X., and Cohen, P. T. (1997) FEBS Lett. 400, 136–14047. Yan, J., Lawson, J. E., and Reed, L. J. (1996) Proc. Natl. Acad. Sci. U. S. A. 93,

4953–495648. Mukhopadhyay, S., Kapatral, V., Xu, W., and Chakrabarty, A. M. (1999) J.

Bacteriol. 181, 6615–662249. Jiang, L., Whiteway, M., and Shen, S. H. (2001) FEBS Lett. 509, 142–14450. Groves, M. R., Hanlon, N., Turowski, P., Hemmings, B. A., and Barford, D.

(1999) Cell 96, 99–11051. Das, A. K., Helps, N. R., Cohen, P. T., and Barford, D. (1996) EMBO J. 15,

6798–680952. Das, A. K., Cohen, P. W., and Barford, D. (1998) EMBO J. 17, 1192–119953. Volker, U., Luo, T., Smirnova, N., and Haldenwang, W. G. (1997) J. Bacteriol.

179, 1980–198454. Scott, J. M., Mitchell, T., and Haldenwang, W. G. (2000) J. Bacteriol. 182,

1452–145655. Akbar, S., Gaidenko, T. A., Kang, C. M., O’Reilly, M., Devine, K. M., and Price,

C. W. (2001) J. Bacteriol. 183, 1329–133856. Esnouf, R. M. (1997) J. Mol. Graph. 15, 133–13857. Merritt, E. A., and Bacon, D. J. (1997) Methods Enzymol. 277, 505–52458. Benson, A. K., and Haldenwang, W. G. (1992) J. Bacteriol. 174, 749–757

Activation of a Stress-responsive Phosphatase 40937

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Uwe Völker, Michael D. Yudkin and Richard J. LewisOlivier Delumeau, Sujit Dutta, Matthias Brigulla, Grit Kuhnke, Steven W. Hardwick,

Phosphatase 2CFunctional and Structural Characterization of RsbU, a Stress Signaling Protein

doi: 10.1074/jbc.M405464200 originally published online July 19, 20042004, 279:40927-40937.J. Biol. Chem. 

  10.1074/jbc.M405464200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/279/39/40927.full.html#ref-list-1

This article cites 56 references, 24 of which can be accessed free at

by guest on August 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from