JOURNAL OF BACTERIOLOGY, Dec. 2006, p. 8044–8053 Vol. 188, No. 230021-9193/06/$08.00�0 doi:10.1128/JB.00824-06Copyright © 2006, American Society for Microbiology. All Rights Reserved.

GroES/GroEL and DnaK/DnaJ Have Distinct Roles in Stress Responsesand during Cell Cycle Progression in Caulobacter crescentus�

Michelle F. Susin, Regina L. Baldini, Frederico Gueiros-Filho, and Suely L. Gomes*Departamento de Bioquımica, Instituto de Quımica, Universidade de Sao Paulo, Sao Paulo, Brasil

Received 9 June 2006/Accepted 6 September 2006

Misfolding and aggregation of protein molecules are major threats to all living organisms. Therefore, cellshave evolved quality control systems for proteins consisting of molecular chaperones and proteases, whichprevent protein aggregation by either refolding or degrading misfolded proteins. DnaK/DnaJ and GroES/GroEL are the best-characterized molecular chaperone systems in bacteria. In Caulobacter crescentus thesechaperone machines are the products of essential genes, which are both induced by heat shock and cell cycleregulated. In this work, we characterized the viabilities of conditional dnaKJ and groESL mutants underdifferent types of environmental stress, as well as under normal physiological conditions. We observed that C.crescentus cells with GroES/EL depleted are quite resistant to heat shock, ethanol, and freezing but are sensitiveto oxidative, saline, and osmotic stresses. In contrast, cells with DnaK/J depleted are not affected by thepresence of high concentrations of hydrogen peroxide, NaCl, and sucrose but have a lower survival rate afterheat shock, exposure to ethanol, and freezing and are unable to acquire thermotolerance. Cells lacking thesechaperones also have morphological defects under normal growth conditions. The absence of GroE proteinsresults in long, pinched filamentous cells with several Z-rings, whereas cells lacking DnaK/J are only somewhatmore elongated than normal predivisional cells, and most of them do not have Z-rings. These findings indicatethat there is cell division arrest, which occurs at different stages depending on the chaperone machine affected.Thus, the two chaperone systems have distinct roles in stress responses and during cell cycle progression in C.crescentus.

Caulobacter crescentus, an aquatic bacterium and a memberof the � subdivision of the Proteobacteria, produces two celltypes: motile, DNA replication-quiescent “swarmer cells” andsessile, DNA replication-competent “stalked cells.” Theformer are important for dispersion, and the latter are impor-tant for reproduction (65).

Each motile swarmer cell has a single polar flagellum andseveral pili at one pole. CtrA, a DNA-binding response regu-lator that directly controls transcription of at least 95 genes in55 operons (48), is present in swarmer cells, where it binds tothe C. crescentus origin of replication and blocks replicationinitiation (61). Simultaneously, CtrA directly represses tran-scription of gcrA (35), ftsZ (44), and podJ (12), blocking theearly steps in cell division and polar development. Swarmercells undergo differentiation to stalked cells, during which thepolar pili, flagellum, and chemotaxis apparatus are lost and arereplaced by a stalk that grows at the pole previously occupiedby the flagellum. Concurrent with the swarmer cell-stalked celltransition, CtrA is degraded (60), while DnaA levels increase(31). The presence of DnaA, not just the absence of CtrA, isrequired to trigger an increase in GcrA levels and start the nextwave of cell cycle transcription, which includes expression ofgenes encoding nucleotide biosynthesis and DNA replicationenzymes. Thus, DnaA not only initiates DNA replication butalso promotes the transcription of the components necessary

for successful chromosome duplication and the transcription offtsZ and podJ, starting the cell division and polar organelledevelopment processes that prepare the cell for asymmetricdivision (35).

More than 19% of the C. crescentus genes have discretetimes of transcriptional activation and repression during a nor-mal cell cycle (49). For each cell cycle-regulated event, a set ofassociated genes is induced immediately before or coincidentwith the event (49). The DnaK chaperone is synthesized atdefined times in the C. crescentus cell cycle (29, 30). The dnaK/Jgenes are transcribed just before the S phase during the tran-sition from swarmer cells to stalked cells and again in latepredivisional cells just before the initiation of DNA replicationin the progeny stalked cells (29). Expression of the groESLoperon is also under cell cycle control in C. crescentus, andGroEL levels are higher in predivisional cells (4, 30). Theseobservations indicated possible roles of DnaK and GroELchaperones in specific events of the C. crescentus cell cycle.

In bacteria, the major molecular chaperones include theDnaK machine (67) and the GroE machine (GroES andGroEL). Molecular chaperones protect newly synthesized orstress-denatured polypeptides from misfolding and aggrega-tion in the highly crowded cellular environment, often in anATP-driven process (24, 33). The C. crescentus dnaKJ (3, 29)and groESL operons (4, 5), in addition to lon (81), hrcA/grpE(63), ftsH (23), and clpB (69), are heat shock inducible and areall positively controlled by the specific heat shock sigma factor�32. The rpoH gene encoding �32 is also heat shock inducible,as one of its promoters was shown to be �32 dependent,indicating that there is autogenous control of rpoH tran-scription in C. crescentus (62, 82). Similar to its role in

* Corresponding author. Mailing address: Departamento de Bio-quımica, Instituto de Quımica, Universidade de Sao Paulo, Av. Prof.Lineu Prestes 748, 05508-000, Sao Paulo, SP, Brasil. Phone: 55-11-3091-3826. Fax: 55-11-3091-2186. E-mail: sulgomes@iq.usp.br.

� Published ahead of print on 15 September 2006.

8044

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

Escherichia coli, DnaK is a negative modulator of the heatshock response in C. crescentus, acting by inhibiting �32

activity and stimulating degradation of this molecule (13).However, despite the strong effect of DnaK levels on theinduction phase of the response, the shutoff of heat shockprotein (HSP) synthesis is not affected by changes in theamount of this chaperone. Competition between �32 and�73, the major sigma factor in C. crescentus, which wasshown also to be heat shock inducible, has been proposed asthe important factor controlling the downregulation of HSPsynthesis during the recovery phase (13). Moreover, theabsence of the chaperone ClpB delays the shutoff of HSPsynthesis in C. crescentus. Reactivation of heat-inactivated�73 in this bacterium was shown to be dependent on ClpBchaperone activity, indicating that ClpB levels control down-regulation of the heat shock response in C. crescentus (69).

In E. coli, the ribosome-associated trigger factor (TF) coop-erates with DnaK and its DnaJ and GrpE cochaperones toassist in the de novo folding of at least 340 cytosolic proteins.The TF/DnaK-dependent proteins have a broad size range,between 16 and 167 kDa, and the number of multidomainproteins is particularly high (15, 73). Neither TF nor DnaK isabsolutely essential for E. coli viability, but deletion of both ofthem results in synthetic lethality at temperatures of �30°C(15, 73). Other chaperones, including GroES/GroEL, can par-tially compensate for the combined loss of TF and DnaK (25,74, 76). Indeed, the chaperonin GroEL and its cofactor GroESare the only E. coli chaperones that are essential for viabilityunder all growth conditions tested (21, 36). GroES/EL interactwith about 5 to 15% of newly synthesized proteins, and thepredominant size range of these proteins is 20 to 60 kDa (8, 19,38, 39), suggesting that the GroE system plays a role in the denovo folding of these proteins. Some proteins that are too largeto be encapsulated can nevertheless utilize GroEL for foldingby cycling on and off the GroEL ring in trans to bind GroES(10). Recently, Kerner et al. (45) described characterization ofthe GroEL substrate proteome by a combination of biochem-ical analyses and quantitative proteomics. Approximately 250of the �2,400 cytosolic E. coli proteins interact with GroEL inwild-type cells, and the number increases substantially in cellslacking the upstream chaperones TF and DnaK. However, only�85 substrates exhibit an obligate dependence on GroEL forfolding under normal growth conditions, occupying 75 to 80%of the GroEL capacity (45).

Several reports have shown that GroES/GroEL and DnaK/DnaJ are induced during environmental conditions other thanheat shock, such as osmotic and saline stresses, oxidative stress,pH extremes, UV radiation, and the presence of toxic com-pounds (ethanol, antibiotics, heavy metals, and aromatic com-pounds) (20, 32, 58, 72). However, the importance of thesechaperones for bacterial survival under these harmful condi-tions has not been comprehensively investigated. In this work,we analyzed C. crescentus groESL and dnaKJ conditional mu-tant strains SG300 and SG400 to determine their abilities tosurvive in the presence of various environmental stresses, in-cluding heat shock at 42°C and 48°C, oxidative stress (2.5 mMH2O2), a high concentration of ethanol (15%), freezing(�80°C), saline and osmotic stresses (85 mM NaCl and 150mM sucrose, respectively), and acquisition of thermotolerance.In addition, we also analyzed the aberrant cell division pheno-

types related to the lack of GroES/GroEL and DnaK/DnaJ inC. crescentus.

MATERIALS AND METHODS

Bacterial strains and growth conditions. C. crescentus strains were grown at30°C in PYE complex medium (57) supplemented with either 0.2% glucose(PYEG) or 0.2% xylose (PYEX). The C. crescentus strains used were NA1000, aholdfast mutant derivative of wild-type strain CB15 (18); SG300, containing thegroESL operon under control of the PxylX promoter; and SG400, containing thednaKJ operon under control of the PxylX promoter (13).

Determination of cell viability. Strains SG300 and SG400 were tested forsensitivity to different stress conditions, and parental strain NA1000 was used asa control in all experiments. Overnight NA1000 cultures were diluted to anoptical density at 600 nm (OD600) of 0.1 in PYE and incubated in a rotary shakerat 30°C for 6 h. Overnight cultures in PYEX of conditional mutants SG300 andSG400 were washed several times in PYE to remove all the remaining xylose,diluted in PYE, PYEX, or PYEG, and incubated at 30°C for 6 h. In PYE andPYEG, this time was long enough to deplete GroES/GroEL or DnaK/DnaJ fromthe cells. Aliquots of cells were exposed to heat shock at 42°C, 15% ethanol, or2.5 mM hydrogen peroxide. For each data point, serial dilutions of the cultureswere made and plated on PYEX agar plates. Typically, 10 �l or 100 �l of a 10�4

or 10�6 dilution was plated in order to obtain a convenient number of coloniesfor counting (30 to 300 colonies). The plates were then incubated for 2 days at30°C, and colonies were counted to determine the number of CFU/ml for eachtime. In the freezing experiments, exponentially growing bacterial cultures werefrozen and incubated for up to 144 h at �80°C, and the numbers of viable cellswere determined as described above. The responses to saline and osmoticstresses were also tested with exponential-phase cultures in PYE by adding 85mM (final concentration) NaCl and 150 mM (final concentration) sucrose, re-spectively, and incubating the cells for up to 8 h at 30°C. The numbers of viablecells were determined as described above. To test for the involvement of GroES/GroEL and DnaK/J in induced thermotolerance, mutant strains SG300 andSG400 and parental strain NA1000 were grown in liquid PYE at 30°C for 6 h.Each culture was then divided into two aliquots; one of the aliquots was main-tained at 30°C, and the other was incubated at 40°C for 30 min. After this, bothaliquots were subjected to heat treatment at 48°C for 60 min. Cultures were thenserially diluted in PYE and plated on PYEX agar to determine the numbers ofviable cells.

Immunoblot assays. Samples of C. crescentus cells were taken at regularintervals during heat shock or during growth in the absence of xylose, centri-fuged, and resuspended in Laemmli sample buffer, and equal amounts of totalprotein were separated by denaturing sodium dodecyl sulfate-polyacrylamide gelelectrophoresis (47) and transferred to nitrocellulose membranes. The mem-branes were treated as previously described (4) using anti-DnaK and anti-GroELantisera of C. crescentus (5). The blots were developed using an enhancedchemiluminescence kit (Amersham). Quantification of the immunoblots wascarried out by densitometry scanning of the X-ray films using the ImageMaster-VDS software (Pharmacia Biotech).

Preparation of cells for light microscopy. Overnight cultures of conditionalmutants SG300 and SG400 were washed several times to remove the xylose andthen diluted to obtain an OD600 of 0.1 in PYEX or PYEG. Samples werecollected after 1, 6, 10, and 24 h of growth, and cell morphology was assessed bylight microscopy with a Nikon TE300 microscope, using a Planfluor 100� ob-jective lens. Cells grown for 10 h were stained with a LIVE/DEAD Baclightviability kit to determine the in situ viability or with FM1-43 (Molecular Probes)for fluorescent staining of the cytoplasmic membrane. Cells were prepared bydiluting cultures 1:1 with PYE and adding 1 �l of mixed LIVE/DEAD stain andthen were observed immediately. FM1-43 was added from a 1 mM stock solutionin dimethyl sulfoxide directly to the cells in growth medium to obtain a finalconcentration of 1 �M. Samples used for imaging the cells and membranes weremounted on poly-L-lysine-treated slides. SYTO 9 from the LIVE/DEAD kit andFM1-43 were detected using a fluorescein isothiocyanate (FITC) filter (NikonEF-4 B-2E/C), and propidium iodide was detected with a red band-pass filter(Nikon EF-4 G-1B).

For FtsZ immunolocalization, samples were prepared by methanol fixation aspreviously described (34) and incubated overnight at room temperature withanti-FtsZ antibody (1:1,000; kind gift from Y. V. Brun). Goat anti-rabbit FITC-conjugated secondary antibody was used at a dilution of 1:50 in phosphate-buffered saline containing 2% bovine serum albumin and incubated for 1 h atroom temperature. Cells were imaged using an FITC filter, and the images werecaptured and processed using the Metamorph software, version 4.5 (UniversalImaging, Media, PA).

VOL. 188, 2006 DISTINCT ROLES OF GroES/EL AND DnaK/J IN CAULOBACTER 8045

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

RESULTS

GroEL and DnaK levels in conditional mutant strainsSG300 and SG400. To analyze the effects of different stresseson the survival of C. crescentus cells lacking the chaperonesDnaK/J or GroES/EL, we used conditional mutants SG300 andSG400, which contain single functional chromosomal copies ofgroESL and dnaKJ, respectively, under control of the xylose-inducible promoter from the xylX gene (13). Since the regula-tion of these operons in parental strain NA1000 is complex, weinitially compared the levels of GroEL and DnaK in SG300and SG400 cells under both permissive and restrictive condi-tions with the corresponding wild-type levels.

As shown in Fig. 1, the GroEL levels in SG300 cells grownin the presence of xylose were about 50% lower than the levelsin NA1000, whereas the opposite situation was observed forDnaK levels in SG400 cells, which were 80% higher than thelevels in the parental strain, under the same growth conditions(Fig. 1). This probably reflected the strength of the wild-typepromoters of each operon compared with the strength of thePxylX promoter. In addition, whereas the GroEL levels weresimilar in NA1000 and SG400 cells growing in the presence ofxylose (Fig. 1), the amount of DnaK was 50% larger in SG300cells grown under the same conditions than the amount in theparental strain (Fig. 1).

As previously shown (13), when xylose was removed fromthe culture media, there was a marked reduction in the amountof GroEL in SG300, and this protein was undetectable after 6 hof incubation in the absence of the inducer (Fig. 1). Concom-itantly, there was an increase in the DnaK level (Fig. 1), whichwas 50% higher than the wild-type level under permissiveconditions and was threefold higher than the wild-type level inSG300 cells grown without xylose (Fig. 1). A similar behaviorwas seen when SG400 cultures were grown in the absence ofxylose, and the level of GroEL increased significantly (1.8-fold)as DnaK in the cells was depleted (Fig. 1). The increase in thelevel of DnaK when GroEL in the cells was depleted and viceversa indicated that the absence of each of the chaperones ledto accumulation of partially denatured proteins, which causedan increase in the amount of �32 in both cases, as previouslydescribed (13). The increase in �32 levels was shown to be quite

large in SG400 cells after the removal of xylose (the levels wereabout 20-fold higher than wild-type levels); the maximum wasreached by 5 h under nonpermissive conditions, and the levelremained high even 24 h later. In SG300 cells, the increase inthe level of �32 was smaller (2.5-fold lower), with a peak after5 h of growth without xylose, and the level decreased slowlyafter this peak (13). The larger amount of �32 in each depletedmutant strain in turn induced the transcription of heat shockgenes, resulting in an increased rate of synthesis of the otherchaperone, whose gene was still under control of a �32-depen-dent promoter.

DnaK/DnaJ and GroES/EL are necessary during cell growthat different temperatures. Previous work in our laboratoryshowed that both the DnaK and GroEL proteins are essentialfor C. crescentus viability, since mutants with deletions in thecorresponding genes could be obtained only when a wild-typecopy of each gene was provided in trans or when conditionalmutants were constructed (13). As these experiments werecarried out with cultures growing at 30°C, we examinedwhether this was also the case at other temperatures. In fact, inthe absence of xylose, mutant strains SG300 and SG400 werenot able to form colonies at 16°C, 30°C, or 37°C (not shown).Moreover, SG300 did not form colonies at the highest temper-ature tested (37°C) even in the presence of xylose (not shown).These results showed that both GroES/EL and DnaK/J wereessential at all temperatures tested and that the levels ofGroES/EL obtained in the presence of xylose were not suffi-cient for growth at 37°C.

DnaK/DnaJ are important during heat shock. SG300 andSG400 cells growing in the presence of xylose were just asviable as cells of parental strain NA1000 during exposure to42°C for 2 h (Fig. 2). However, when C. crescentus SG300 andSG400 cells grown in the presence of glucose were subjected tothe same stress conditions, depletion of DnaK/J led to amarked decrease in viability (the viability was 2,600-fold lowerthan the NA1000 viability), whereas the viability of cells lack-ing the GroE proteins was not affected (Fig. 2). Similar resultswere obtained when C. crescentus cells were exposed to ex-treme heat treatment at 48°C for 60 min, and cells with DnaK/J

FIG. 1. GroEL and DnaK levels in groESL and dnaKJ conditionalmutants. Cells from parental strain NA1000 and from conditionalmutants SG300 and SG400, in which the groESL and dnaKJ operons,respectively, are under control of promoter PxylX, were grown at 30°Cin the presence or absence of xylose for 6 h. At different times, sampleswere collected and proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by immunoblotting usingantisera against C. crescentus GroEL and DnaK, as described in Ma-terial and Methods. Equal amounts of protein were applied to alllanes.

FIG. 2. Induction of DnaK/J synthesis but not induction ofGroES/EL synthesis is important for survival after heat shock. Over-night SG300 and SG400 cultures growing in PYEX at 30°C werewashed, diluted in PYEX or PYEG, and incubated for 6 h at 30°C.Cells were then transferred to a 42°C water bath with agitation, andsamples were taken at different times and plated on PYEX. Thesurvival values are the averages of three independent experiments, andthe error bars indicate the standard deviations. Light gray bars, SG400diluted in PYEX; open bars, SG400 diluted in PYEG; black bars,SG300 diluted in PYEX; dark gray bars, SG300 diluted in PYEG.

8046 SUSIN ET AL. J. BACTERIOL.

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

depleted exhibited much greater sensitivity than cells withGroES/EL depleted (Fig. 3). Interestingly, SG300 cells withthe GroE proteins depleted survived in the presence of thisextreme temperature better than parental strain NA1000 cellssurvived (Fig. 3). The DnaJ/DnaK chaperone machine was alsoshown to be necessary for C. crescentus cells to acquire ther-motolerance to extreme temperatures. Cultures of SG400 pre-incubated at 40°C for 30 min survived exposure to 48°C as wellas the wild-type cells only when xylose was present in thegrowth medium (Fig. 3). In contrast, there was a significantincrease in the ability SG300 cells growing under restrictiveconditions to survive exposure to 48°C when cells were pre-treated at 40°C (Fig. 3). This was probably due to the higherlevels of DnaK/DnaJ present in SG300 cells in all conditionstested (Fig. 1). Thus, these results indicated that the presenceof DnaK/DnaJ is very important for the survival of cells ex-posed to heat stress, whereas GroES and GroEL are not nec-essary.

DnaK/DnaJ are involved in the survival of cells exposed toethanol and freezing. Several other stresses were tested in cellswith DnaK/DnaJ or GroES/GroEL depleted in order to assessthe roles of these chaperones in cell survival after exposureto different environmental conditions. Strain SG400 grownin the absence of xylose was about six times more sensitiveto 15% ethanol than wild-type or SG300 cells were underthe same conditions (Fig. 4A), indicating that DnaK/J butnot GroES/EL is important for protecting C. crescentus cellsagainst the deleterious effects of ethanol, such as dehydration,membrane disturbance, and the resulting protein denatur-ation.

It was recently shown that C. crescentus cells are quite resis-tant to freezing (E. Lang and M. V. Marques, personal com-munication). We observed that NA1000 cultures kept frozen at�80°C for several days in the absence of any antifreezing agentwere 100% viable when they were plated on nutrient medium

at 30°C. The same behavior was observed with SG300 cells inwhich GroES/EL were depleted or not depleted. However,SG400 cells lacking DnaK/J were more sensitive to freezing,exhibiting decreased viability (80%) after incubation for 1 h at�80°C; only 30% of the cells were still viable after incubationfor 18 h or more (Fig. 4B).

GroES/GroEL are important during oxidative, osmotic, andsaline stresses. Even though GroES/EL were shown to bedispensable for survival of C. crescentus cells exposed to heatshock, ethanol, and freezing, their presence during oxidative,saline, and osmotic stresses was found to be very important. Asshown in Fig. 5A, SG300 cells lacking GroES/EL were 10-foldmore sensitive to incubation with 2.5 mM H2O2 for 60 minthan wild-type cells or SG400 cells with DnaK/J depleted. Inaddition, the presence of 150 mM sucrose in the growth me-dium resulted in a fivefold decrease in the viability of SG300cells with GroES/EL depleted compared with the viability ofthe wild-type cells after 8 h of incubation under osmotic stressconditions. Exposure to 85 mM NaCl had an even more pro-nounced effect, as only 2% of SG300 cells growing in theabsence of xylose were viable after 8 h of saline stress (Fig. 5B).Under the same conditions, SG400 cells with DnaK/J depleted

FIG. 3. Absence of DnaK/J makes SG400 cells incapable of acquir-ing thermotolerance. Cultures of strains SG300 and SG400 were incu-bated at 30°C for 12 h in PYEX, and cells were subsequently washed,diluted in PYEX or PYEG, and incubated for 6 h at 30°C. A similarprocedure was used for parental strain NA1000, except that only PYEwas used. One half of the cultures were preconditioned for 30 min atthe nonlethal temperature (40°C) (gray bars), while the other half ofthe cultures were preincubated at 30°C (open bars). The cultures werethen exposed to 48°C for 30 min, and cell survival was evaluated afterdilution and plating on PYEX agar. The values are the levels of cellsurvival (averages of three independent experiments) compared withthe survival of control cultures grown in the absence of stress. Theerror bars indicate standard deviations.

FIG. 4. Depletion of DnaK/J makes cells more sensitive to ethanoland freezing. (A) Overnight cultures of SG400 or SG300 in PYEXwere washed, diluted in PYEX or PYEG, and incubated for 6 h at 30°Cfor depletion of DnaK/J or GroES/EL. The cultures were then exposedto 15% (vol/vol) ethanol, and samples were collected at different timesand plated on PYEX. (B) Samples of the 6-h cultures were frozenat �80°C without any supplement. At different times, aliquots weretransferred to room temperature, allowed to defrost, diluted, andplated on PYEX. The bars indicate the averages of three independentexperiments, and the error bars indicate standard deviations. Lightgray bars, SG400 diluted in PYEX; open bars, SG400 diluted inPYEG; black bars, SG300 diluted in PYEX; dark gray bars, SG300diluted in PYEG.

VOL. 188, 2006 DISTINCT ROLES OF GroES/EL AND DnaK/J IN CAULOBACTER 8047

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

or not depleted (Fig. 5B and C) and NA1000 cells (not shown)showed no loss of viability.

Viability and morphology of cells with DnaK/J or GroES/ELdepleted. Viability studies of SG300 and SG400 cells growingunder permissive or restrictive conditions were also carried outusing the LIVE/DEAD BacLight staining mixture (MolecularProbes), which distinguishes viable bacterial cells from deadcells on the basis of membrane integrity and has been used tomonitor damage to bacterial populations (75). Our studies withLIVE/DEAD staining showed that SG300 cultures contained48% viable cells after 10 h of growth at 30°C in the absence of

xylose, whereas in the presence of xylose 95% of the cells werestill viable (not shown). In contrast, depletion of DnaK/DnaJdid not seem to significantly affect the viability of SG400 cellsduring growth for up to 10 h under restrictive conditions, since93% of the cells were found to be viable under these conditions(not shown).

Determination of the number of CFU by plating SG300 cellsgrowing at 30°C in the absence of xylose revealed a steadydecrease in viability starting after 6 h of incubation in theabsence of xylose, and the level of viability was very low after24 h (Fig. 6A). In agreement with LIVE/DEAD staining re-sults, there was a much smaller decrease in the viability ofSG400 cells even after 24 h of incubation in absence of xylose(Fig. 6B). The increase in cell mass, as determined by measur-ing the OD600 of the cultures, also agreed with the LIVE/DEAD data; i.e., SG400 cultures growing in the absence ofxylose produced a growth curve similar to that of the parentalcell culture up to 12 h, whereas the SG300 growth curve devi-ated significantly from the growth curve for the parental strainafter 8 h of incubation without xylose (not shown).

Morphologically, cells with either GroES/EL or DnaK/J de-pleted were elongated, suggesting that there was a defect incell division in both cases, although large differences between

FIG. 5. Depletion of GroES/EL results in increased sensitivity tooxidative, osmotic, and saline stresses. Overnight cultures of SG400or SG300 in PYEX were washed, diluted in PYEX or PYEG, andincubated for 6 h at 30°C. The cultures were divided and exposed to2.5 mM H2O2 (A), 85 mM NaCl (B), or 150 mM sucrose (C), andsamples were collected at different times and plated on PYEX. Thevalues are the levels of cell survival (averages of three independentexperiments) compared with the survival of control cultures grownin the absence of stress. The error bars indicate standard deviations.Light gray bars, SG400 diluted in PYEX; open bars, SG400 dilutedin PYEG; black bars, SG300 diluted in PYEX; dark gray bars,SG300 diluted in PYEG.

FIG. 6. Survival of mutant strains SG300 and SG400 under restric-tive or permissive growth conditions. Cultures of strains NA1000(A and B), SG300 (A), and SG400 (B) were grown in PYEX or PYEGliquid medium at 30°C and monitored through the logarithmic andstationary growth phases. The numbers of CFU were determined atdifferent times by plating cell samples on solid medium (PYEX). Thevalues are the averages of three independent experiments. F, SG300grown in PYEX; E, SG300 grown in PYEG; Œ, SG400 grown inPYEX; ‚, SG400 grown in PYEG; ■ , NA1000.

8048 SUSIN ET AL. J. BACTERIOL.

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

the two strains were observed. Cells with GroES/EL depletedformed very long filaments (Fig. 7B), whereas cells withDnaK/J depleted were only somewhat more elongated thanwild-type predivisional cells (Fig. 7D). In addition, in the ab-sence of DnaK/J some cells seemed to lose the vibrioid shape,especially 24 h after xylose was removed (Fig. 7F).

Most SG300 cells grown for 10 h in the absence of xylosewere found to be 8 to 13 �m long, while the cells grown withxylose were 2 to 3.4 �m long, in agreement with the sizedistribution pattern for wild-type cells (not shown). Therefore,depletion of GroES/EL led to fourfold-longer cells. In con-trast, SG400 cells with low levels of DnaK/J were only 1.4-foldlonger than the cells grown in the presence of xylose or thanwild-type cells (not shown).

DnaK/J and GroES/EL are required for C. crescentus celldivision. To better characterize the effect of DnaK/J orGroES/EL depletion on C. crescentus cell division, a mem-brane dye (FM1-43 from Molecular Probes) was used to assesswhether septa were formed in SG300 and SG400 elongatedcells.

Most SG300 cells with GroES/EL depleted had deep, irreg-ular constrictions along the filaments, and only a few cells(�4%) lacked such constrictions (Fig. 8B). The position of thedivision sites was quite variable in these cells, and some fila-ments consisted of long unpinched segments adjacent to mini-cells or cells that were the regular size. In contrast, only 11%of the SG400 cells with DnaK/DnaJ depleted had a septum orshallow constriction in the middle of the cell (Fig. 8D). Thesame strains grown under permissive conditions (Fig. 8A andC), as well as wild-type strain NA1000 (not shown), had oneseptum per cell in about 35 to 41% of the population, corre-sponding to the percentage of predivisional cells in a normalmixed-cell population.

Since an important step in septum formation is the appear-ance of a Z-ring in the equatorial region of the cell, an anti-FtsZ antibody was used to visualize this structure by immuno-fluorescence microscopy. Again, the results obtained forSG300 and SG400 cells grown in the presence of xylose weresimilar to the results obtained for NA1000 cells: a Z-ring waspresent in 70% of the cells, corresponding to the percentagefor predivisional cells and for stalked cells in transition topredivisional cells (Fig. 9A and C). Cells that lacked a Z-ringwere swarmer or stalked cells in the first stages of chromosomereplication. Under restrictive conditions, however, 92% of theSG300 cells had Z-rings (Fig. 9B), in contrast to only 8% of theSG400 cells (Fig. 9D). These results indicated that depletion of

FIG. 7. Altered morphology of SG300 and SG400 cells lackingGroES/L or DnaK/J. Strains SG300 (A and B) and SG400 (C to F)were grown in PYEX (A, C, and E) or PYEG (B, D, and F). Cells werevisualized by light microscopy using a 100� objective and differentialinterference contrast optics after 10 h (A to D) and 24 h (E and F) ofincubation at 30°C.

FIG. 8. Septum formation in SG300 and SG400 cells grown underrestrictive conditions. After incubation for 10 h in PYEX (A and C) orPYEG (B and D), SG300 (A and B) and SG400 (C and D) cells werestained with the membrane dye FM1-43, and images were capturedusing a 100� objective and an FITC filter. Arrowheads indicate invag-ination of the cytoplasmic membrane.

VOL. 188, 2006 DISTINCT ROLES OF GroES/EL AND DnaK/J IN CAULOBACTER 8049

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

DnaK/DnaJ and depletion of GroES/GroEL affect differentstages of the cell division process.

DISCUSSION

Several lines of experimental evidence demonstrated thatheat shock proteins not only have essential functions at hightemperatures or under harsh environmental conditions butalso have important roles under physiological growth condi-tions. In agreement with this, we report here that both GroES/GroEL and DnaK/DnaJ are essential for C. crescentus growthat all temperatures tested (37°C, 30°C, and 16°C). A similarobservation was made for E. coli GroES/GroEL (21), whereasDnaK was found to be essential only at high temperatures (9).Furthermore, DnaK is not required for growth of Bacillussubtilis at temperatures ranging from 16 to 52°C (67), whereasGroES/EL are essential both at normal temperatures and dur-ing heat stress (51).

Our results indicate that during heat shock there is majorinvolvement of DnaK/J compared to the involvement ofGroES/EL, since a lack of DnaK/J had a great impact on cellsurvival at high temperatures (both 42°C and 48°C), whereasthe percentage of cells that survived in the absence of GroEproteins was similar to or even higher than the percentage ofparental cells that survived after exposure to 42°C or 48°C. Theimportance of the role of DnaK/J at high temperatures is

further supported by our observation that only these chaper-ones were essential for acquisition of thermotolerance. Theseresults are consistent with the suggestion that DnaK/J have agreater role as molecular chaperones than GroES/EL (11, 16).

The observation that the absence of GroES/EL did not makeSG300 cells more sensitive to heat shock could be explained bythe presence of higher levels of DnaK at 30°C in these cellsboth in the presence and in the absence of xylose compared tothe levels in parental cells (1.5- and 3-fold-higher levels, re-spectively).

Similarly, DnaK/J seemed to play a more important rolethan GroES/EL played when C. crescentus cells were exposedto high ethanol concentrations. This was probably due to thefact that ethanol mimics the effects of high-temperature stressand induces a response similar to the heat shock response (2,55, 56, 68). As expected, ethanol stress caused a transientincrease in �32 levels and a gradual increase in DnaK andGroEL levels in C. crescentus (data not shown).

Recently, our laboratory reported that the absence of ClpBmakes C. crescentus cells more sensitive to heat shock andethanol and that this chaperone is needed for acquisition ofthermotolerance (69). Based on the similar phenotypes of theclpB null mutant and the dnaK mutant under restrictive con-ditions, we believe that DnaK and ClpB may act synergistically,preventing and solubilizing protein aggregates during heatshock and ethanol stress in C. crescentus, as previously pro-posed (28).

During oxidative stress, enzymes and cell structures aredamaged, leading to a loss of viability in bacterial populations(20, 42, 71). In several bacteria, it has been shown that theoxidative stress response overlaps other stress responses, suchas the heat shock, starvation, and SOS responses (71). Expres-sion of GroEL and DnaK is induced during treatment withH2O2 in several bacteria (17, 20, 32), as well as in C. crescentus(data not shown). In addition, it has been reported that Hae-mophilus ducreyi cells with lower GroEL levels have a dimin-ished ability to survive when they are challenged by oxidativestress (56) and that E. coli dnaK mutants are able to develop anadaptive H2O2 resistance (64).

The present work showed that C. crescentus cells with lowlevels of GroES/EL are more sensitive to oxidative stress in-duced by H2O2 than the parental strain is, while cells withDnaK/J depleted have the parental phenotype. This result maybe explained by the reported ability of GroEL to functionunder oxidative stress conditions and to be highly resistant tooxidative damage compared to other proteins (53, 54).

Our data also showed that GroES/GroEL are essential forC. crescentus during salt and osmotic stresses. When exposed toosmotic upshifts, bacteria respond in three overlapping phases:dehydration, adjustment of cytoplasmic solvent composition,and rehydration and cellular remodeling (80). Similar effectsoccur when cells are subjected to high concentrations of salt(43). The reason why GroES/GroEL are needed for C. cres-centus survival in high-osmolarity conditions could be the chap-eroning of newly synthesized proteins that are rapidly upregu-lated upon osmotic shock (e.g., the proteins of the K�

transporter, KdpACDF) and/or the refolding of polypeptidesdamaged as a result of water efflux (6, 43). It has been shownthat many outer membrane proteins, such as OmpA, OmpC,OmpF, and murein lipoprotein, have a partial GroES/EL re-

FIG. 9. Immunolocalization of FtsZ in SG300 and SG400 cells.After 10 h of incubation in PYEX (A and C) or PYEG (B and D),SG300 (A and B) and SG400 (C and D) cells were collected andprepared for immunofluorescence microscopy using the anti-FtsZ an-tibody and an FITC-labeled goat anti-rabbit secondary antibody. Im-ages were captured with a Roper CoolSnap HQ cooled charge-coupleddevice camera using a 100� objective and an FITC filter. Arrowheadsindicate the Z ring.

8050 SUSIN ET AL. J. BACTERIOL.

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

quirement in E. coli (45). GroES/EL may also participate inthe translocation and insertion into the membrane of trans-porters involved in the osmotic response. In fact, several re-ports have shown that GroEL can form complexes with nativemembrane proteins in vitro (7, 14, 70). Lecker et al. (50)showed that there is a stable interaction between GroEL and apresecretory protein (proOmpA), maintaining an open confor-mation, which is essential for translocation.

The importance of GroES/EL for survival during osmoticand salt stresses is reflected by the higher groESL promoteractivity in C. crescentus under these conditions (not shown). Inagreement with a less important role for DnaK/J in cell sur-vival, dnaKJ promoter activity increases less under these stressconditions (not shown).

For cold shock and freezing, the absence of DnaK/J resultedin a decrease in the ability of C. crescentus cells to survive infreezing temperatures, whereas the absence of GroES/EL hadno effect. Protein renaturation after cold shock and freezingseems to be more crucial to bacteria than during these stresses;therefore, the chaperones are more important in the recoveryphase. In E. coli, it was also shown that it is mainly DnaK thatallows cells to survive after freezing, as a dnaK mutant per-formed worse than a groEL mutant during recovery from thisstress (11).

The fact that groESL and dnaKJ expression is regulatedduring the C. crescentus cell cycle indicates that these chaper-ones have important roles in events related to chromosomereplication and partition and/or cell division. Bacterial celldivision involves several genes, and their products might re-quire chaperones to function properly. One of the major celldivision determinants, the tubulin homologue FtsZ, is presentin almost all eubacteria and archaea, as well as in some intra-cellular organelles of eukaryotic cells. FtsZ localizes specifi-cally to the midcell division site, where it forms the cytokineticZ-ring, which constricts the cell membrane during septation. Inboth E. coli and C. crescentus, ftsZ mutant strains form un-pinched elongated cells due to their inability to initiate celldivision (1, 78). FtsA, an actin homologue, is required for celldivision progression in C. crescentus, since cells lacking FtsAform filaments that have several constrictions, indicating thatcell division has initiated but stalled at a later point than itstalls in ftsZ mutants (52). This phenotype was also observed inCaulobacter cells lacking GroES/EL, where several Z-ringswere detected in the filaments. This observation indicated thatcell division was inhibited at a stage consistent with the timewhen the level of expression of the groESL operon was maxi-mum, which was the predivisional cell stage (4).

In E. coli, FtsE, ParC, and MreB, which are required in laterstages of cell division, were found to be obligatory substratesfor the GroE machine (45). Whereas FtsE participates directlyin cell division and is important for assembly or stability of theseptal ring in E. coli (66), DNA topoisomerase IV, which is theproduct of the parC and parE genes, is required for polarlocalization of the origin of replication and for chromosomesegregation in C. crescentus (77, 79). Incubation of parC andparE temperature-sensitive mutants at the restrictive temper-ature results in the formation of filamentous chains of cellspinched at multiples sites (79), similar to the phenotype of cellswith GroE proteins depleted. Inactivation of the actin homo-logue MreB is lethal and pleiotropic in C. crescentus (22, 26)

and disrupts multiple cellular processes, including chromo-some dynamics, determination of the cell shape, polar proteinlocalization, and cell division (27). The possibility that manyproteins related to various stages of the cell cycle and cellmorphogenesis could be partially or completely inactive due toa lack of GroES/EL could explain the heterogeneity of SG300phenotypes under restrictive conditions.

After 10 h under restrictive conditions, cells lacking DnaK/Jare only slightly more elongated than wild-type predivisionalcells, indicating that there is a putative arrest of the cell cycle.Cells with DnaK/J depleted also show inhibition of septumformation, since only 8% of the cells have a septum or areconstricted in the midcell region. The absence of Z-rings frommost cells indicates that blockage of septum formation mayoccur at the initial stage of cell division. These results areconsistent with the hypothesis that cells with DnaK/J depletedare arrested at an early stage of the cell cycle, probably at theinitial steps of chromosome replication or segregation.

It is well known that initiation of replication and chromo-some segregation in C. crescentus is coupled with cell divisionand differentiation (59). This occurs because inhibition of rep-lication by CtrA blocks the complete formation of a septum,since cell division progression requires expression of ftsA andftsQ, which is dependent on DNA replication. Without FtsAand FtsQ, the Z-ring is dismantled (52) and the division pro-cess cannot proceed. During the transition from swarmer cellsto stalked cells, CtrA is degraded (60), while the DnaA levelsincrease (31). The presence of DnaA, and not just the ab-sence of CtrA, is required to trigger an increase in GcrAlevels and to start the next wave of cell cycle transcription,which includes the expression of genes encoding nucleotidebiosynthesis and DNA replication enzymes. Thus, DnaA notonly initiates DNA replication but also promotes expressionof the components necessary for successful chromosomeduplication. DnaA also activates transcription of ftsZ andpodJ, starting the cell division and polar organelle biogen-esis processes that, in addition to DNA replication, preparethe cell for asymmetric division (37).

Interestingly, the morphology of C. crescentus cells withDnaA depleted (37) is similar to that of cells with DnaK/Jdepleted. Thus, it is possible that the arrest in the SG400 cellcycle is a consequence of DnaA inactivation in the absence ofDnaK. In agreement with this hypothesis, dnaK transcriptionin C. crescentus precedes the S phase, right at the transitionbetween swarmer cells and stalked cells, when DNA replica-tion initiates (29). This is also consistent with previous ge-netic and biochemical evidence suggesting that DnaK/Jchaperones could be involved in DNA replication in E. coli.It has been demonstrated that these chaperones could pro-tect DnaA from aggregation and could dissociate DnaAaggregates in vitro, thus allowing the initiation of oriC DNAreplication (40, 41, 46).

Besides the putative cell cycle arrest observed in the absenceof DnaK, we also noted that after 24 h in the absence of xylose,SG400 cells became more elongated and lost the vibrioid shapecharacteristic of C. crescentus. This might indicate that DnaK/DnaJ also have a chaperone role in cell shape maintenance,perhaps by interacting with the intermediate filament-like pro-tein crescentin (G. Charbon and C. Jacobs-Wagner, personalcommunication).

VOL. 188, 2006 DISTINCT ROLES OF GroES/EL AND DnaK/J IN CAULOBACTER 8051

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

It is important to note, however, that the effects on celldivision observed here could have been the result of multiplefactors, many of which are not directly related to septum for-mation. Since GroES/EL and DnaK/J can act on many differ-ent substrates, their absence could result in pleiotropic alter-ations in bacterial physiology.

In conclusion, we showed that DnaK/DnaJ and GroES/GroEL have important but distinct roles in C. crescentus bothunder normal physiological growth conditions and under dif-ferent environmental stress conditions. Characterization of thespecific substrates of these chaperones is now fundamentallyimportant for unraveling their actual roles in the various bio-logical processes analyzed here.

ACKNOWLEDGMENTS

We thank Yves V. Brun for generously supplying anti-FtsZ anti-serum. We also thank Marilis V. Marques and Christine Jacobs-Wag-ner for sharing unpublished results.

This work was supported by a grant from Fundacao de Amparo aPesquisa do Estado de Sao Paulo (FAPESP). M.F.S is a fellow ofFAPESP, and S.L.G. was partially supported by Conselho Nacional deDesenvolvimento Cientıfico e Tecnologico (CNPq).

REFERENCES

1. Addinall, S. G., E. Bi, and J. Lutkenhaus. 1996. FtsZ ring formation in ftsmutants. J. Bacteriol. 178:3877–3884.

2. Arnosti, D. N., V. L. Singer, and M. J. Chamberlin. 1986. Characterizationof heat shock in Bacillus subtilis. J. Bacteriol. 168:1243–1249.

3. Avedissian, M., D. Lessing, J. W. Gober, L. Shapiro, and S. L. Gomes. 1995.Regulation of the Caulobacter crescentus dnaKJ operon. J. Bacteriol. 177:3479–3484.

4. Avedissian, M., and S. L. Gomes. 1996. Expression of the groESL operon iscell-cycle controlled in Caulobacter crescentus. Mol. Microbiol. 19:79–89.

5. Baldini, R. L., M. Avedissian, and S. L. Gomes. 1998. The CIRCE elementand its putative repressor control cell cycle expression of the Caulobactercrescentus groESL operon. J. Bacteriol. 180:1632–1641.

6. Bianchi, A. A., and F. Baneyx. 1999. Hyperosmotic shock induces the �32 and�E stress regulons of Escherichia coli. Mol. Microbiol. 34:1029–1038.

7. Bochkareva, E., A. Seluanov, E. Bibi, and A. Girshovich. 1996. Chaperonin-promoted post-translational membrane insertion of a multispanning mem-brane protein lactose permease. J. Biol. Chem. 271:22256–22261.

8. Brinker, A., G. Pfeifer, M. J. Kerner, D. J. Naylor, F. U. Hartl, and M.Hayer-Hartl. 2001. Dual function of protein confinement in chaperonin-assisted protein folding. Cell 107:223–233.

9. Bukau, B., and G. C. Walker. 1989. Delta dnaK52 mutants of Escherichia colihave defects in chromosome segregation and plasmid maintenance at normalgrowth temperatures. J. Bacteriol. 171:6030–6038.

10. Chaudhuri, T. K., G. W. Farr, W. A. Fenton, S. Rospert, and A. L. Horwich.2001. GroEL/GroES-mediated folding of a protein too large to be encapsu-lated. Cell 107:235–246.

11. Chow, K. C., and W. L. Tung. 1998. Overexpression of dnaK/dnaJ and groELconfers freeze tolerance to Escherichia coli. Biochem. Biophys. Res. Com-mun. 253:502–505.

12. Crymes, W. B., Jr., D. Zhang, and B. Ely. 1999. Regulation of podJ expres-sion during the Caulobacter crescentus cell cycle. J. Bacteriol. 181:3967–3973.

13. da Silva, A. C., R. C. Simao, M. F. Susin, R. L. Baldini, M. Avedissian, andS. L. Gomes. 2003. Downregulation of the heat shock response is indepen-dent of DnaK and sigma32 levels in Caulobacter crescentus. Mol. Microbiol.49:541–553.

14. Deaton, J., J. Sun, A. Holzenburg, D. K. Struck, J. Berry, and R. Young.2004. Functional bacteriorhodopsin is efficiently solubilized and delivered tomembranes by the chaperonin GroEL. Proc. Natl. Acad. Sci. USA 101:2281–2286.

15. Deuerling, E., A. Schulze-Specking, T. Tomoyasu, A. Mogk, and B. Bukau.1999. Trigger factor and DnaK cooperate in folding of newly synthesizedproteins. Nature 400:693–696.

16. Ellis, R. J., and F. U. Hartl. 1996. Protein folding in the cell: competingmodels of chaperonin function. FASEB J. 10:20–26.

17. Ericsson, M., A. Tarnvik, K. Kuoppa, G. Sandstrom, and A. Sjostedt. 1994.Increased synthesis of DnaK, GroEL, and GroES homologs by Francisellatularensis LVS in response to heat and hydrogen peroxide. Infect. Immun.62:178–183.

18. Evinger, M., and N. Agabian. 1977. Envelope-associated nucleoid from Cau-lobacter crescentus stalked and swarmer cells. J. Bacteriol. 132:294–301.

19. Ewalt, K. L., J. P. Hendrick, W. A. Houry, and F. U. Hartl. 1997. In vivo

observation of polypeptide flux through the bacterial chaperonin system. Cell90:491–500.

20. Farr, S. B., and T. Kogoma. 1991. Oxidative stress responses in Escherichiacoli and Salmonella typhimurium. Microbiol. Rev. 55:561–585.

21. Fayet, O., T. Ziegelhoffer, and C. Georgopoulos. 1989. The groES and groELheat shock gene products of Escherichia coli are essential for bacterialgrowth at all temperatures. J. Bacteriol. 171:1379–1385.

22. Figge, R. M., A. V. Divakaruni, and J. W. Gober. 2004. MreB, the cellshape-determining bacterial actin homologue, co-ordinates cell wall mor-phogenesis in Caulobacter crescentus. Mol. Microbiol. 51:1321–1332.

23. Fischer, B., G. Rummel, P. Aldridge, and U. Jenal. 2002. The FtsH proteaseis involved in development, stress response and heat shock control in Cau-lobacter crescentus. Mol. Microbiol. 44:461–478.

24. Frydman, J. 2001. Folding of newly translated proteins in vivo: the role ofmolecular chaperones. Annu. Rev. Biochem. 70:603–647.

25. Genevaux, P., F. Keppel, F. Schwager, P. S. Langendijk-Genevaux, F. U.Hartl, and C. Georgopoulos. 2004. In vivo analysis of the overlapping func-tions of DnaK and trigger factor. EMBO Rep. 5:195–200.

26. Gitai, Z., N. Dye, and L. Shapiro. 2004. An actin-like gene can determine cellpolarity in bacteria. Proc. Natl. Acad. Sci. USA 101:8643–8648.

27. Gitai, Z., N. A. Dye, A. Reisenauer, M. Wachi, and L. Shapiro. 2005. MreBactin-mediated segregation of a specific region of a bacterial chromosome.Cell 120:329–341.

28. Goloubinoff, P., A. Mogk, A. P. Zvi, T. Tomoyasu, and B. Bukau. 1999.Sequential mechanism of solubilization and refolding of stable protein ag-gregates by a bichaperone network. Proc. Natl. Acad. Sci. USA 96:13732–13737.

29. Gomes, S. L., J. W. Gober, and L. Shapiro. 1990. Expression of the Cau-lobacter heat shock gene dnaK is developmentally controlled during growthat normal temperatures. J. Bacteriol. 172:3051–3059.

30. Gomes, S. L., M. H. Juliani, J. C. Maia, and A. M. Silva. 1986. Heat shockprotein synthesis during development in Caulobacter crescentus. J. Bacteriol.168:923–930.

31. Gorbatyuk, B., and G. T. Marczynski. 2005. Regulated degradation of chro-mosome replication proteins DnaA and CtrA in Caulobacter crescentus. Mol.Microbiol. 55:1233–1245.

32. Hanawa, T., T. Yamamoto, and S. Kamiya. 1995. Listeria monocytogenes cangrow in macrophages without the aid of proteins induced by environmentalstresses. Infect. Immun. 63:4595–4599.

33. Hartl, F. U., and M. Hayer-Hartl. 2002. Molecular chaperones in the cytosol:from nascent chain to folded protein. Science 295:1852–1858.

34. Hiraga, S., C. Ichinose, H. Niki, and M. Yamazoe. 1998. Cell cycle-depen-dent duplication and bidirectional migration of SeqA-associated DNA-pro-tein complexes in E. coli. Mol. Cell 1:381–387.

35. Holtzendorff, J., D. Hung, P. Brende, A. Reisenauer, P. H. Viollier, H. H.McAdams, and L. Shapiro. 2004. Oscillating global regulators control thegenetic circuit driving a bacterial cell cycle. Science 304:983–987.

36. Horwich, A. L., and K. R. Willison. 1993. Protein folding in the cell: functionsof two families of molecular chaperone, hsp 60 and TF55-TCP1. Philos.Trans. R Soc. Lond. B Biol. Sci. 339:313–326.

37. Hottes, A. K., L. Shapiro, and H. H. McAdams. 2005. DnaA coordinatesreplication initiation and cell cycle transcription in Caulobacter crescentus.Mol. Microbiol. 58:1340–1353.

38. Houry, W. A. 2001. Mechanism of substrate recognition by the chaperoninGroEL. Biochem. Cell Biol. 79:569–577.

39. Houry, W. A., D. Frishman, C. Eckerskorn, F. Lottspeich, and F. U. Hartl.1999. Identification of in vivo substrates of the chaperonin GroEL. Nature402:147–154.

40. Hupp, T. R., and J. M. Kaguni. 1993. DnaA5 protein is thermolabile ininitiation of replication from the chromosomal origin of Escherichia coli.J. Biol. Chem. 268:13128–13136.

41. Hwang, D. S., E. Crooke, and A. Kornberg. 1990. Aggregated DnaA proteinis dissociated and activated for DNA replication by phospholipase or DnaKprotein. J. Biol. Chem. 265:19244–19248.

42. Imlay, J. A. 2003. Pathways of oxidative damage. Annu. Rev. Microbiol.57:395–418.

43. Kanesaki, Y., I. Suzuki, S. I. Allakhverdiev, K. Mikami, and N. Murata.2002. Salt stress and hyperosmotic stress regulate the expression of differentsets of genes in Synechocystis sp. PCC 6803. Biochem. Biophys. Res. Com-mun. 290:339–348.

44. Kelly, A. J., M. J. Sackett, N. Din, E. Quardokus, and Y. V. Brun. 1998. Cellcycle-dependent transcriptional and proteolytic regulation of FtsZ in Cau-lobacter. Genes Dev. 12:880–893.

45. Kerner, M. J., D. J. Naylor, Y. Ishihama, T. Maier, H. C. Chang, A. P. Stines,C. Georgopoulos, D. Frishman, M. Hayer-Hartl, M. Mann, and F. U. Hartl.2005. Proteome-wide analysis of chaperonin-dependent protein folding inEscherichia coli. Cell 122:209–220.

46. Konieczny, I., and K. Liberek. 2002. Cooperative action of Escherichia coliClpB protein and DnaK chaperone in the activation of a replication initia-tion protein. J. Biol. Chem. 277:18483–18488.

47. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 227:680–685.

8052 SUSIN ET AL. J. BACTERIOL.

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from

48. Laub, M. T., S. L. Chen, L. Shapiro, and H. H. McAdams. 2002. Genesdirectly controlled by CtrA, a master regulator of the Caulobacter cell cycle.Proc. Natl. Acad. Sci. USA 99:4632–4637.

49. Laub, M. T., H. H. McAdams, T. Feldblyum, C. M. Fraser, and L. Shapiro.2000. Global analysis of the genetic network controlling a bacterial cell cycle.Science 290:2144–2148.

50. Lecker, S., R. Lill, T. Ziegelhoffer, C. Georgopoulos, P. J. Bassford, Jr., C. A.Kumamoto, and W. Wickner. 1989. Three pure chaperone proteins of Esch-erichia coli—SecB, trigger factor and GroEL—form soluble complexes withprecursor proteins in vitro. EMBO J. 8:2703–2709.

51. Li, M., and S. L. Wong. 1992. Cloning and characterization of the groESLoperon from Bacillus subtilis. J. Bacteriol. 174:3981–3992.

52. Martin, M. E., M. J. Trimble, and Y. V. Brun. 2004. Cell cycle-dependentabundance, stability and localization of FtsA and FtsQ in Caulobacter cres-centus. Mol. Microbiol. 54:60–74.

53. Melkani, G. C., C. McNamara, G. Zardeneta, and J. A. Mendoza. 2004.Hydrogen peroxide induces the dissociation of GroEL into monomers thatcan facilitate the reactivation of oxidatively inactivated rhodanese. Int.J. Biochem. Cell Biol. 36:505–518.

54. Melkani, G. C., G. Zardeneta, and J. A. Mendoza. 2004. Oxidized GroEL canfunction as a chaperonin. Front. Biosci. 9:724–731.

55. Park, S. H., K. H. Oh, and C. K. Kim. 2001. Adaptive and cross-protectiveresponses of Pseudomonas sp. DJ-12 to several aromatics and other stressshocks. Curr. Microbiol. 43:176–181.

56. Periago, P. M., W. van Schaik, T. Abee, and J. A. Wouters. 2002. Identifi-cation of proteins involved in the heat stress response of Bacillus cereusATCC 14579. Appl. Environ. Microbiol. 68:3486–3495.

57. Poindexter, J. S. 1964. Biological properties and classification of the Cau-lobacter group. Bacteriol. Rev. 28:231–295.

58. Prasad, J., P. McJarrow, and P. Gopal. 2003. Heat and osmotic stressresponses of probiotic Lactobacillus rhamnosus HN001 (DR20) in relation toviability after drying. Appl. Environ. Microbiol. 69:917–925.

59. Quardokus, E. M., and Y. V. Brun. 2003. Cell cycle timing and developmen-tal checkpoints in Caulobacter crescentus. Curr. Opin. Microbiol. 6:541–549.

60. Quon, K. C., G. T. Marczynski, and L. Shapiro. 1996. Cell cycle control byan essential bacterial two-component signal transduction protein. Cell 84:83–93.

61. Quon, K. C., B. Yang, I. J. Domian, L. Shapiro, and G. T. Marczynski. 1998.Negative control of bacterial DNA replication by a cell cycle regulatoryprotein that binds at the chromosome origin. Proc. Natl. Acad. Sci. USA95:120–125.

62. Reisenauer, A., C. D. Mohr, and L. Shapiro. 1996. Regulation of a heatshock �32 homolog in Caulobacter crescentus. J. Bacteriol. 178:1919–1927.

63. Roberts, R. C., C. Toochinda, M. Avedissian, R. L. Baldini, S. L. Gomes, andL. Shapiro. 1996. Identification of a Caulobacter crescentus operon encodinghrcA, involved in negatively regulating heat-inducible transcription, and thechaperone gene grpE. J. Bacteriol. 178:1829–1841.

64. Rockabrand, D., T. Arthur, G. Korinek, K. Livers, and P. Blum. 1995. Anessential role for the Escherichia coli DnaK protein in starvation-inducedthermotolerance, H2O2 resistance, and reductive division. J. Bacteriol. 177:3695–3703.

65. Ryan, K. R., and L. Shapiro. 2003. Temporal and spatial regulation in

prokaryotic cell cycle progression and development. Annu. Rev. Biochem.72:367–394.

66. Schmidt, K. L., N. D. Peterson, R. J. Kustusch, M. C. Wissel, B. Graham,G. J. Phillips, and D. S. Weiss. 2004. A predicted ABC transporter, FtsEX,is needed for cell division in Escherichia coli. J. Bacteriol. 186:785–793.

67. Schulz, A., B. Tzschaschel, and W. Schumann. 1995. Isolation and analysis ofmutants of the dnaK operon of Bacillus subtilis. Mol. Microbiol. 15:421–429.

68. Segal, G., and E. Z. Ron. 1995. The dnaKJ operon of Agrobacterium tume-faciens: transcriptional analysis and evidence for a new heat shock promoter.J. Bacteriol. 177:5952–5958.

69. Simao, R. C., M. F. Susin, C. E. Alvarez-Martinez, and S. L. Gomes. 2005.Cells lacking ClpB display a prolonged shutoff phase of the heat shockresponse in Caulobacter crescentus. Mol. Microbiol. 57:592–603.

70. Sun, J., C. G. Savva, J. Deaton, H. R. Kaback, M. Svrakic, R. Young, and A.Holzenburg. 2005. Asymmetric binding of membrane proteins to GroEL.Arch. Biochem. Biophys. 434:352–357.

71. Tamarit, J., E. Cabiscol, and J. Ros. 1998. Identification of the major oxi-datively damaged proteins in Escherichia coli cells exposed to oxidativestress. J. Biol. Chem. 273:3027–3032.

72. Teixeira-Gomes, A. P., A. Cloeckaert, and M. S. Zygmunt. 2000. Character-ization of heat, oxidative, and acid stress responses in Brucella melitensis.Infect. Immun. 68:2954–2961.

73. Teter, S. A., W. A. Houry, D. Ang, T. Tradler, D. Rockabrand, G. Fischer, P.Blum, C. Georgopoulos, and F. U. Hartl. 1999. Polypeptide flux throughbacterial Hsp70: DnaK cooperates with trigger factor in chaperoning nascentchains. Cell 97:755–765.

74. Ullers, R. S., J. Luirink, N. Harms, F. Schwager, C. Georgopoulos, and P.Genevaux. 2004. SecB is a bona fide generalized chaperone in Escherichiacoli. Proc. Natl. Acad. Sci. USA 101:7583–7588.

75. Virta, M., S. Lineri, P. Kankaanpaa, M. Karp, K. Peltonen, J. Nuutila, andE. M. Lilius. 1998. Determination of complement-mediated killing of bac-teria by viability staining and bioluminescence. Appl. Environ. Microbiol.64:515–519.

76. Vorderwulbecke, S., G. Kramer, F. Merz, T. A. Kurz, T. Rauch, B. Zach-mann-Brand, B. Bukau, and E. Deuerling. 2004. Low temperature orGroEL/ES overproduction permits growth of Escherichia coli cells lackingtrigger factor and DnaK. FEBS Lett. 559:181–187.

77. Wang, S. C., and L. Shapiro. 2004. The topoisomerase IV ParC subunitcolocalizes with the Caulobacter replisome and is required for polar local-ization of replication origins. Proc. Natl. Acad. Sci. USA 101:9251–9256.

78. Wang, Y., B. D. Jones, and Y. V. Brun. 2001. A set of ftsZ mutants blockedat different stages of cell division in Caulobacter. Mol. Microbiol. 40:347–360.

79. Ward, D., and A. Newton. 1997. Requirement of topoisomerase IV parC andparE genes for cell cycle progression and developmental regulation in Cau-lobacter crescentus. Mol. Microbiol. 26:897–910.

80. Wood, J. M. 1999. Osmosensing by bacteria: signals and membrane-basedsensors. Microbiol. Mol. Biol. Rev. 63:230–262.

81. Wright, R., C. Stephens, G. Zweiger, L. Shapiro, and M. R. Alley. 1996.Caulobacter Lon protease has a critical role in cell-cycle control of DNAmethylation. Genes Dev. 10:1532–1542.

82. Wu, J., and A. Newton. 1996. Isolation, identification, and transcriptionalspecificity of the heat shock sigma factor �32 from Caulobacter crescentus. J.Bacteriol. 178:2094–2101.

VOL. 188, 2006 DISTINCT ROLES OF GroES/EL AND DnaK/J IN CAULOBACTER 8053

at Sistem

a Integrado de Bibliotecas-U

SP

/FO

B on A

ugust 13, 2007 jb.asm

.orgD

ownloaded from