|
|
|
Journal of Bacteriology, September 2004, p . 6110-6117, Vol . 186, No . 18 Structural Determinants Required To Target Penicillin-Binding Protein 3 to the Septum of Escherichia coliAndré Piette,1 Claudine Fraipont,1 Tanneke den Blaauwen,2 Mirjam E . G . Aarsman,2 Soumya Pastoret,1 and Martine Nguyen-Distèche1* Centre d'Ingénierie des Protéines, Institut de Chimie, Université de Liège, Liège, Belgium,1 Swammerdam Institute for Life Sciences, University of Amsterdam, Amsterdam, The Netherlands2 Received 12 April 2004/ Accepted 22 June 2004
The functions of most of these proteins are not known . How the divisome works in term of protein-protein interaction is poorly understood . Until now, the only known direct interactions arethose between FtsZ and FtsA or ZipA [17, 26, 29] . The bitopicproteins FtsB and FtsL seem to interact in a coiled-coil structurethrough their periplasmic domain [6] . Other protein-proteininteractions should be involved in the functioning of the divisome. The multimodular class B PBP3 specifically catalyzes peptide cross-bridges of the septal cell wall peptidoglycan during cell division . It consists of a short intracellular M1-R23 peptidefused to an F24-L39 membrane anchor that is linked via a G40-S70peptide to an R71-I236 noncatalytic module, itself linked toa D237-V577 catalytic penicillin-binding module . It has beenproposed that the activity of the transpeptidase module of PBP3is regulated by the interaction of its N-terminal noncatalyticmodule with other cell division proteins [11, 24] . Previous experiments allowed identification of three peptide segmentsin the noncatalytic module of PBP3 that have protein-proteininteraction potentials and specific functions . The G40-S70 sequenceand the membrane anchor-containing module appear to play animportant role in the proper insertion of the protein withinthe divisome at the division site [24, 30] . It might interact with FtsW, which is essential for the recruitment of PBP3 at the cell septation site [25] . The H160-G172 segment, located at the intermodule junction, seems to be involved in intramolecular interactions and plays an important role in the conformationof the protein . The E206-V217 segment, which is exposed at thesurface of the noncatalytic module of PBP3, plays an importantrole in cell septation, presumably by interacting with othercomponents of the divisome . FtsQ, FtsL, and FtsW are plausiblepartners [24]. In order to assess the role of the G40-S70 and E206-V217 segments and their interaction with other cell division proteins, the localization of PBP3 modified in these segments was analyzed.A search for suppressors of the dominant-negative effect dueto mutations in the E206-V217 segment was carried out by analyzingthe effect of overproduction of FtsQ, FtsL, or FtsW on E . coli cells producing a PBP3 mutant altered in this segment . The results reported hereafter show that the first 56 amino acid residuesof PBP3 contain the structural determinants required for targetingof PBP3 at the cell division site and that none of the putativeprotein interaction sites in the noncatalytic module is essentialfor the positioning of the protein to the division site . FtsQcould play a role in regulating the assembly of the divisionproteins FtsL, FtsB, FtsW, and PBP3.
E . coli EC548 cells expressing GFP fusion proteins were grown
for two mass doublings in LB supplemented with 0.2% L-arabinose
at 37°C from an overnight culture grown in the same conditions.
They were then diluted 400-fold to an OD600 of 0.0025 in LB
containing 0.2% glucose at 37°C and harvested after sevenmass
doubling periods [OD600 of
E . coli TOP10 F' cells coexpressing PBP3 mutants and FtsL, FtsQ, or FtsW were grown at 37°C in LB medium inoculated withan overnight culture [inoculum 2%] . When the OD600 reached a value of 0.3, 0.1 mM IPTG was added, and the culture was maintained for 30 min at 37°C before the addition of 0.005 or 0.01% L-arabinose [arabinose induction was omitted when GFP-PBP3 constructscontrolled by a weakened trc promoter were used] . Cells wereharvested 3 h after induction. Molecular biological procedures. Plasmids and oligonucleotides are described in Table 1 . Oligonucleotides were purchased from Amersham Biosciences and Eurogentec . All constructs were sequenced to verify their integrity, using an ALFexpress DNA sequencer [Amersham Biosciences]. Recombinant plasmids. Construction of GFP fusions . Plasmid pDSW234 encoded the wild-typePBP3 fused to the C-terminal end of GFP . In this construct,the initiating methionine of PBP3 is absent from the fusion[30] . The SacII-NruI fragment of pDSW234 was exchanged withthe corresponding fragment of pUCBM20/ftsI[R166Q, R167Q], pUCBM20/ftsI[R210Q,R213Q], and pHR275 carrying the truncated ftsI gene encodingthe M1-R239 PBP3 [24] . The resulting plasmids, pDML2481, pDML2482,and pDML2484, encoded GFP-PBP3[R166Q, R167Q], GFP-PBP3[R210Q,R213Q], and GFP-PBP3[K2-R239], respectively, and the gene fusionwas under the control of a weakened trc promoter [30]. The ftsI gene encoding the PBP3[D58V] mutant was amplified by PCR using pHR458 as a template, the gfp-ftsI primer bearingan EcoRI site at its 5' end, and the ftsI-NruI primer . The PCRproduct was digested by EcoRI and NruI and inserted in the corresponding sites of pDSW234, digested by the same enzymes . The resulting plasmid, pDML2480, encoded GFP-PBP3[D58V] . Plasmid pDML2483was obtained by replacing the SacII-NruI fragment of pDML2480with the corresponding fragment of pUCBM20/ftsI[R210Q, R213Q]and encoded the GFP-PBP3[D58V, R210Q, R213Q] triple mutant.To construct pDML2485, which encoded GFP-PBP3[K2-S70], the SacII-HindIIIfragment of pDSW234 was excised . The plasmid was treated withT4 DNA polymerase to make blunt ends and then ligated on itself.Because of the construction, the encoded PBP3[K2-S70] had anadditional heptapeptide [QLGCFGG] at the carboxy end . To generate GFP-PBP3[K2-E56] and GFP-PBP3[K2-V42], the MluI-HindIII fragmentof pDSW234 encoding the R41-S588 sequence of PBP3 was replacedby double-stranded oligonucleotides dupE56 and dupV42 to giverise to pDML2486 and pDML2487, respectively. To fuse GFP-PBP3[K2-V42] to the penicillin-binding domain ofBlaR, MluI and HindIII restriction sites were added, respectively,upstream and downstream of the sequence encoding the carboxy-terminaldomain of the Bacillus licheniformis BlaR protein by PCR, using pDML995 as the template and the oligonucleotides MluI-blaR and blaR-HindIII as primers . The resulting PCR product was thendigested by MluI and HindIII and cloned in the correspondingsites of pDSW234 to give rise to pDML2488 . pDSW234 was digestedby EcoRI and HindIII and treated with the mung bean nucleasebefore being circularized to give pDML2489, which encoded GFPunder the control of a weakened trc promoter. Construction of plasmids coding for PBP3, FtsL, and FtsQ. An NheI site was first inserted upstream of the ftsI gene by site-directed mutagenesis with the QuikChange kit [Stratagene]using pUCBM20/ftsI as template and the primers NheI-ftsIfwdand NheI-ftsIrev . The PCR product was then restricted with NheIand EcoRI and cloned in the NheI-EcoRI sites of pET-28a to givepDML2494, which encoded the His-tagged PBP3 . In order to placethis fused gene under the control of the arabinose promoter,the NcoI-EcoRI fragment was excised from pDML2494 and clonedinto the NcoI-EcoRI sites of pBAD/His A to create pDML2495.The SacII-PstI fragments containing the double mutations [R166QR167Q or R210Q R213Q] were excised from pUCBM20/ftsI recombinantsand exchanged with the unmodified SacII-PstI fragment of pDML2495to give rise to pDML2496 and pDML2497, respectively . The PBP3[D58V,R210Q, R213Q] triple mutant was obtained by site-directed mutagenesis[QuikChange kit from Stratagene], using pDML2497 as a templateand the primers ftsID58V and ftsIcplD58V . The PCR product wasdigested with SacII and PstI and exchanged with the unmodifiedfragment of pDML2495 to create pDML2498. The ftsL and ftsQ genes were amplified by PCR, using as templatesplasmids pUCBM20/ftsI and pNB2, respectively, and the primersBamHI-ftsL and ftsL-HindIII [for ftsL] and XbaI-ftsQ and ftsQ-HindIII[for ftsQ] . The PCR products were digested with BamHI and HindIII[for ftsL] and XbaI and HindIII [for ftsQ] and then ligatedinto the corresponding sites of pMCL210 to give pDML2490 andpDML2491, respectively . In these constructs the ftsL and ftsQgenes were under the control of the lac promoter. Microscopy and image analysis. Cells were fixed in culture medium with 2.8% formaldehyde and0.04% glutaraldehyde for 15 min at room temperature, collectedby centrifugation at 4,500 x g for 5 min, washed twice in phosphate-bufferedsaline [140 mM NaCl, 27 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4 [pH7.2]], suspended in water, and immobilized on agarose slidesas described by Koppelman et al . [20] . They were photographedwith a Photometrics CoolSNAPfx CCD camera [Roper Scientific]mounted on an Olympus BX60 microscope through a UPlanFI x100/1.3 oil-immersion objective, both in phase contrast and fluorescence [with U-MNB filter cube, 470- to 490-nm band-pass excitationfilter, >515-nm long-pass emission filter] . Images were analyzedby using the public-domain software Object-Image 2.08 by N.Vischer [University of Amsterdam http://simon.bio.uva.nl/object-image.html],based on NIHImage by W . Rasband. Membrane preparation and fluorescent labeling of penicillin-binding protein. Cells were harvested by centrifugation and suspended in a solutioncontaining 20 mM Tris-HCl [pH 8], 5 mM EDTA, and 100 µg of lysozyme ml–1 on ice for 30 min . MgCl2 [final concentration,15 mM], Benzonase [4 U ml–1; Merck], and 10–4 Mphenylmethylsulfonyl fluoride were then added, and the cell suspension was frozen and thawed three times . Membranes were spun down at 30,000 x g for 30 min and resuspended in a solutioncontaining 10 mM Tris-HCl [pH 8], 500 mM NaCl, 10% glycerol,and 10% ethylene glycol. Membranes were diluted in 20 mM phosphate [pH 7.6]-150 mM NaCland incubated for 10 min at 30°C with 5 x 10–5 M ß-iodopenicillanicacid to inactivate any residual ß-lactamase . 5'-Fluorescein-ampicillinwas then added to a final concentration of 5 x 10–5 M,and membranes were further incubated for 25 min at 30°C[22] . Proteins were then boiled for 5 min in sodium dodecylsulfate [SDS] sample buffer to denature GFP. Separation and detection of proteins. Proteins were solubilized and separated by standard SDS-polyacrylamidegel electrophoresis [PAGE] methods [21] . For immunological detectionof GFP, proteins were electroblotted to a polyvinylidene difluoride[PVDF] membrane and revealed with mouse anti-GFP monoclonalantibodies and horseradish peroxidase-coupled goat anti-mouseimmunoglobulin G antibodies . Detection was performed using theECL chemiluminescence reagent [Amersham Biosciences]. Fluorescence detection after SDS-PAGE was done with a Molecular Imager FX [Bio-Rad Laboratories] instrument using the parameterset designed for fluorescein isothiocyanate detection [excitationby 488-nm Ar-ion laser, 515- to 545-nm band-pass emission filter].For GFP fluorescence, samples were not boiled but solubilizedfor 30 min at 37°C in SDS sample buffer . Quantificationwas performed with the Quantity One 4.1 software [Bio-Rad Laboratories].
As derived from Western blotting of the plasma membrane isolated from LMC500 transformants, the amount of the GFP-PBP3 variantswas estimated at about four times that of the natural chromosomally encoded PBP3 [results not shown]. As shown in Fig . 2, each of the PBP3 mutants localized at thedivision site of both strains . In LMC500 transformants, PBP3 fluorescent rings were observed in 49 to 80% of the cells, and the average cell length ranged from 3.2 to 5.2 µm, dependingon the mutant [Table 2; Fig . 1] . The increase of the cell lengthreflects a delay in the division process . This was also observedwith transformants producing the native GFP-PBP3 fusion . A numberof cells expressing GFP-PBP3[D58V] showed some fluorescent aggregatesin their membranes [Fig . 2C] . To assess if this could reflectan instability of the protein, its thermostability was measuredby incubating the membranes isolated from the transformantsfor 10 min at different temperatures and then by measuring theamount of PBP3 left in an active form by the subsequent bindingwith fluorescent ampicillin for 25 min at 30°C . GFP-PBP3[D58V]showed the same thermostability as wild-type PBP3, PBP3[R210Q,R213Q], and PBP3[D58, R210Q, R213Q] under the conditions used[Fig . 3] . Interestingly, the PBP3[R166Q, R167Q] mutant, whichdoes not bind penicillin and does not show a dominant-negativeeffect, was targeted to the septal site [Fig. 2E] . These resultsindicate that the segments containing D58 and R210 and R213residues that have protein-protein interaction potentials arenot involved in septal targeting of PBP3 and that this targetingdoes not rely on the proper folding of the catalytic module.
As shown in Fig . 4 and Table 2, GFP-PBP3[K2-R239], GFP-PBP3[K2-S70],and GFP-PBP3[K2-E56] localized in the septal region of 75 to87% of PBP3-depleted cells, whereas the shortest form, GFP-PBP3[K2-V42],whether fused to BlaR or not, localized in only 20 to 25% ofthe cells, with a weak fluorescence signal compared to othertruncated forms . The size of EC548 expressing the nonfunctionalGFP-PBP3 derivatives varied from 18 to 53 µm, illustratingthe depletion of functional PBP3 . The variability in lengthcould reflect differences in the growth rate as well as in thedominant-negative effect of each GFP-PBP3 derivative . For ftsI+ LMC500, a faint fluorescence signal was observed at the division site of a low percentage of cells expressing GFP-PBP3[K2-R239] or GFP-PBP3[K2-S70] [Fig . 4 and Table 2] . The other shorterforms did not localize in the wild-type strain . As shown bySDS-PAGE and fluorescence measurements of membrane proteins prepared from E . coli LMC500 expressing these derivatives, productionof GFP-PBP3[K2-E56] is four times less than that of wild-typeGFP-PBP3, whereas GFP-PBP3[K2-V42] is produced in larger amountsthan the wild-type fusion [Fig . 5] . These results suggest thatthe truncated forms of PBP3 did not compete or did not competewell with the native protein for the recruiting factor but didkeep an affinity for the division site . The affinity of theshortest form, comprising the membrane spanning-segment and the first three periplasmic amino acid residues, was particularly low . These results indicate that the structural determinants required to target PBP3 at the septation site not only lie inthe transmembrane segment but extend a little further withinthe first 17 periplasmic amino acid residues . They also indicatethat the other domains of the protein are not essential in thisprocess.
E . coli TOP10 F' cells were cotransformed with plasmids pDML2497, bearing the modified ftsI gene encoding PBP3[R210Q, R213Q], and pDML2407, pDML2490, or pDML2491, encoding FtsW, FtsL, or FtsQ, respectively . The PBP3 double mutant was under the controlof the PBAD promoter, whereas the other cell division proteins were under the control of the lac promoter . Transformants were grown at 37°C in LB medium containing 100 µM IPTGand 0.005 or 0.01% arabinose as described in experimental procedures.Figure 6 shows the results . The cells producing the PBP3 double mutant were slightly longer than those producing wild-type PBP3, indicating a weak dominant-negative effect under the conditions used . The coexpression of native PBP3 and FtsW had no effecton the size of the cells, whereas that of the native PBP3 andFtsQ increased the length of the cells 1.7-fold . It was observedpreviously that overexpression of FtsQ led to the formationof filaments [10] . When the cells expressed the PBP3 doublemutant and FtsQ, they were 4 to 4.5 times longer than thoseexpressing both FtsQ and the native PBP3 or the PBP3 mutantalone . The cells coproducing FtsW and the PBP3 double mutantwere 1.5 times longer than the control cells . FtsL had no effect[data not shown] . Thus, FtsQ and FtsW to a lesser extent exacerbatedthe dominant-negative effect of the PBP3 double mutant.
Finally, the effect of FtsQ was examined with cells expressing truncated PBP3[K2-E56] or PBP3[K2-V42], both fused to GFP andunder the control of the trc promoter [from plasmids pDML2486and pDML2487, respectively] . Transformants were grown at 37°Cin LB medium in the presence of 100 µM IPTG . Under theseconditions, FtsQ had no appreciable effect on the size of thecells expressing these constructs.
None of the previously predicted interaction sites that are present in the noncatalytic module of PBP3 appear to be essentialfor the insertion of the protein at midcell, because all PBP3mutants including the inactive PBP3[R166Q, R167Q] mutant localizedto the septum . These sites may be involved in interactions withthe other cell division proteins and the peptidoglycan-metabolizingenzymes and thus stabilize the PBP3 once it has reached itsproper location and take part in its function. Recently, Wissel and Weiss published results converging withours . They showed by random mutagenesis the importance of residuesR23, L39, and Q46 for the localization of PBP3 and of residuesG57, S61, L62, and R210, each located in our previously predictedinteraction sites, for the recruitment of FtsN [31]. The mechanism by which the K2-E56 peptide directs the proteinto the septum is not known . FtsW is required for the recruitmentof PBP3 at the division site [25] . Overexpression of FtsW enhanced slightly the dominant-negative effect of the PBP3[R210Q, R213Q] mutant, suggesting an interaction between FtsW and PBP3. FtsW depends on FtsL and FtsB for its localization to the septum. Overexpression of FtsL had no effect on E . coli cells producing the PBP3[R210Q, R213Q] mutant, probably because FtsB, which seems to interact with FtsL and stabilize it [6], is not presentin sufficient amounts . By contrast, overexpression of FtsQ exacerbatedthe dominant-negative effect of the PBP3[R210Q, R213Q] doublemutant . In addition, overexpression of FtsQ blocks division of cells producing the PBP3[R166Q, R167Q] mutant, which is completely inactive and probably misfolded . A similar effect had already been observed by Dai and Lutkenhaus when FtsQ is produced inthe thermosensitive ftsI23 strain at the permissive temperature[8] . These results suggest a functional interaction betweenthese cell division proteins that might be direct or indirect. These suggestions were in agreement with results obtained byusing an E . coli two-hybrid system that showed interactionsbetween FtsW and PBP3 and between FtsQ, FtsW, and PBP3 [9]. Mutations in the noncatalytic module of E . coli PBP3 result in perturbation of cell growth and cell division [24] . Overexpressionof FtsQ inhibits division of ftsI+ cells, producing these PBP3mutants . FtsQ might promote the interaction of the inactiveprotein with the division machinery at the expense of wild-typePBP3 . These data show the important role of the noncatalyticmodule of PBP3 in peptidoglycan assembly at the septum throughprotein-protein interaction . They also point out the importanceof the ratio between components of the divisome . The divisomeis also sensitive to overproduction of FtsQ when it contains thermosensitive FtsZ, FtsA, or PBP3 mutants [8] at the permissivetemperature . Thus, FtsQ could play a role in regulating theassembly of the division proteins and in the control of the peptidoglycan assembly machineries within the divisome. One could ask why the inactive and probably misfolded PBP3[R166Q, R167Q] mutant does not disturb cell division when it localizesas native PBP3 in the presence of endogenous PBP3, while thePBP3[R210Q, R213Q] mutant and PBP3[S307A] do . One explanationis that the divisome is a dynamic structure . Exchanges betweenthe inactive PBP3 mutant and the wild-type protein could occuronce PBP3 is located at the division site due to its first 56amino acid residues . Depending on the affinity of PBP3 mutantsfor the other cell division proteins, PBP1b, lipid II, and/orpeptidoglycan, the dominant-negative effect is more or lessstrong . Indeed, the modified PBP3s, which showed a severe dominant-negativeeffect in our previous report when they were expressed in largeamounts or in ftsI[Ts] cells, caused only a reduced lengtheningwhen slightly expressed in a wild-type background . This wasalso noticed by Wissel and Weiss [31] . This work supports thehypothesis of the existence of multiple interactions betweenPBP3 and other division proteins that are involved in distinctfunctions.
A.P . is a Fellow of the Fonds pour la Formation à laRecherche dans l'Industrie et dans l'Agriculture, Brussels. We thank D . S . Weiss [University of Iowa] for the gift of plasmid pDSW234, S . Blacher [University of Liège] for technicalassistance in phase-contrast microscopy, and J . Coyette [Universityof Liège] for fruitful discussions.
What Is Staphylococcus Aureus?,
What Is Salmonella?,
What Is Antibiotic?,
What Is Molecular Biology?,
What Is Prokaryote?,
s,
Microbes,
r,
Bacterium,
r,
Microbiology,
c,
Microorganism,
r,
Bacteria,
o,
Corynebacter,
r,
S. cerevisiae,
n,
Culture medium,
n,
Minimum inhibiting concentration,
c,
Prokaryotes,
s,
Culture medium,
c,
Growth media,
i,
Activated sludge,
n,
Petri dish,
c,
Escherichia coli,
a,
Candida albicans,
i,
S. cerevisiae,
o,
Bacteriological,
c,
Antibiotics,
c,
Yeasts,
c,
Microbial,
a,
Microbial,
r,
Bacteria,
e,
Brevibacteria,
i,
Acinetobacter,
n,
Microorganism
|
© 2005
Transgalactic Ltd (manufacturer of Bioscreen C software) |
Privacy Statement | P.O. Box
1393, 00101 Helsinki, Finland,
Last modified: May 25, 2005
| ||||||