Crystal Structure of a Peptidoglycan Synthesis Regulatory Factor(PBP3) from Streptococcus pneumoniae*

Received for publication, July 26, 2004, and in revised form, November 29, 2004Published, JBC Papers in Press, December 13, 2004, DOI 10.1074/jbc.M408446200

Cecile Morlot‡§, Lucile Pernot¶, Audrey Le Gouellec‡, Anne Marie Di Guilmi‡, Thierry Vernet‡,Otto Dideberg¶, and Andrea Dessen¶�

From the ¶Laboratoire de Cristallographie Macromoleculaire and ‡Laboratoire d’Ingenierie des Macromolecules,Institut de Biologie Structurale Jean-Pierre Ebel (CNRS/CEA/UJF), 41 rue Jules Horowitz, Grenoble 38027, France

Penicillin-binding proteins (PBPs) are membrane-as-sociated enzymes which perform critical functions inthe bacterial cell division process. The single D-Ala,D-Ala(D,D)-carboxypeptidase in Streptococcus pneumoniae,PBP3, has been shown to play a key role in control ofavailability of the peptidoglycal substrate during cellgrowth. Here, we have biochemically characterized andsolved the crystal structure of a soluble form of PBP3 to2.8 Å resolution. PBP3 folds into an NH2-terminal, D,D-carboxypeptidase-like domain, and a COOH-terminal,elongated �-rich region. The carboxypeptidase domainharbors the classic signature of the penicilloyl serinetransferase superfamily, in that it contains a central,five-stranded antiparallel �-sheet surrounded by �-hel-ices. As in other carboxypeptidases, which are presentin species whose peptidoglycan stem peptide has a ly-sine residue at the third position, PBP3 has a 14-residueinsertion at the level of its omega loop, a feature thatdistinguishes it from carboxypeptidases from bacteriawhose peptidoglycan harbors a diaminopimelate moietyat this position. PBP3 performs substrate acylation in ahighly efficient manner (kcat/Km � 50,500 M�1�s�1), anevent that may be linked to its central role in control ofpneumococcal peptidoglycan reticulation. A model thatplaces PBP3 poised vertically on the bacterial mem-brane suggests that its COOH-terminal region could actas a pedestal, placing the active site in proximity to thepeptidoglycan and allowing the protein to “skid” on thesurface of the membrane, trimming pentapeptides dur-ing the cell growth and division processes.

Bacterial division is a complex phenomenon that requiresthe coordination of diverse processes including chromosomalsegregation, FtsZ ring-dependent membrane constriction, andcell wall synthesis at the site of septation. The latter processinvolves the polymerization of glycan chains and transpeptida-tion of pentapeptidic moieties within the structure of the pep-tidoglycan, a highly cross-linked mesh that is crucial for main-

taining bacterial shape and providing protection from osmoticshock and lysis (1). Both reactions are catalyzed by penicillin-binding proteins (PBPs),1 membrane-associated molecules,which can be classified as high molecular mass (hmm; oftenbifunctional) and low molecular mass (lmm; monofunctional)and play key roles in the bacterial life cycle. The pathogenicbacterium Streptococcus pneumoniae offers a unique opportu-nity for the study of the relationship between cell division andcell wall synthesis, since it carries a relatively simple set of sixPBPs, compared with other well studied organisms which pres-ent much higher complexity (2). In this organism, PBP1a, -1b,and -2a catalyze both glycosyltransfer and transpeptidation;PBP2b and -2x only catalyze the latter reaction, and PBP3, thesingle lmm PBP in S. pneumoniae, has been shown to act as aD-Ala,D-Ala (D,D) carboxypeptidase (3).

The central role of hmm PBPs in the cell growth and divisionprocesses has been recently confirmed through the study oftheir localization within the cell cycle through the employmentof immunofluorescence techniques (4). In S. pneumoniae, theconstriction of the FtsZ-ring is spatially coupled to PBP2x- andPBP1a-mediated septal peptidoglycan synthesis, with the for-mer process preceding the latter by approximately 5 min (4). Atthe beginning of the cell cycle, PBP3 localizes throughout thewhole bacterial surface but seems to be absent from the futuredivision site (5). Since the D,D-carboxypeptidase activity ofPBP3 removes the COOH-terminal D-alanine of the peptidogly-can pentapeptide side chains, its hemispheric localization im-plies that the cellular region neighboring the future divisionsite will be the only one where full-length pentapeptides will beavailable as substrates for other PBPs. Interestingly, a mutantpneumococcal strain which lacks PBP3 displays abnormal mor-phology and exhibits multiple septa initiated at aberrant loca-tions (6). Thus, it is likely that the availability of intact pen-tapeptidic substrates dictates the localization of the hmmPBPs. Therefore, by guaranteeing that pentapeptides are avail-able uniquely at the future division site, PBP3 may ensure thespatial coordination of the FtsZ-ring with the septum synthesismachinery.

PBP3 is associated to the bacterial membrane through aCOOH-terminal amphiphilic helix. In the D,D-carboxypeptidasereaction catalyzed by PBP3, an active serine residue reactswith the D-Ala-D-Ala COOH terminus of a peptide chain of thepeptidoglycan to form a transient acyl-enzyme complex that issubsequently hydrolyzed. The reaction results in the formationof a tetrapeptide that can only serve as an acceptor for asubsequent transpeptidation reaction by other PBPs (3). As in

* This work was supported in part by European Commission GrantLSHM-CT-2003-503335 (COBRA). The costs of publication of this arti-cle were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and structure factors (code 1XP4) have beendeposited in the Protein Data Bank, Research Collaboratory for Struc-tural Bioinformatics, Rutgers University, New Brunswick, NJ(http://www.rcsb.org/).

§ Supported by a CFR fellowship from the Commissariat a l’EnergieAtomique.

� An EMBO Young Investigator. To whom correspondence should beaddressed. Tel.: 33-4-38-78-95-90; Fax: 33-4-38-78-54-94; E-mail:dessen@ibs.fr.

1 The abbreviations used are: PBP, penicillin-binding protein; hmm,high molecular mass; lmm, low molecular mass; D,D, D-Ala,D-Ala; Ac2-KAA, N,N-diacetyl-L-Lys-D-Ala-D-Ala; SeMet, selenomethionine; r.m.s.,root mean square.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 16, Issue of April 22, pp. 15984–15991, 2005© 2005 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.org15984

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

the case of the D,D-transpeptidase activity, D,D-carboxypeptida-tion is inhibited by penicillin and other �-lactam antibioticsthat mimic the structure of the D-Ala-D-Ala carboxyl terminusof the pentapeptide chain (7). These antibiotics react with PBPsto form a stable acyl-enzyme complex, resulting in prolongedinhibition of the enzymes.

Recently, the structure of PBP5, the soluble, signal peptide-lacking form of the D,D-carboxypeptidase from Escherichia coli,was reported in wild type and mutant forms to high resolution(1.85 and 1.9 Å, respectively; Refs. 8 and 9). Wild type PBP5deacylates its acyl-enzyme complex at a very high rate, whichis reminiscent of that of a class A �-lactamase (the latter witha poor substrate). Although these reports shed light on theenzymology of D,D-carboxypeptidation and the two-domain foldof the enzyme, several points still remain unclear, includingthe nature of the D,D-carboxypeptidase substrates and the en-zymatic role of lmm PBPs in the cell division process.

In light of our previous reports on the localization of PBP3 inthe cell cycle (5) and in an effort to answer some of the ques-tions above, we performed the enzymatic characterization ofpneumococcal PBP3 and solved the structure of a soluble formof the enzyme at 2.8 Å resolution. Although the general folds ofthe pneumococcal and E. coli enzymes are similar, PBP3 har-bors a significantly longer omega-like loop, a feature subse-quently identified as a telltale motif in enzymes present inbacteria whose peptidoglycan structures contain an L-lysinegroup in the third stem peptide position. Interestingly, itscarboxypeptidase domain is highly reminiscent of that oftranspeptidase K15 from Streptomyces spp. but shares struc-tural resemblance with other peptidoglycan biosynthetic en-zymes only in the immediate vicinity of the active site. PBP3 isa highly efficient D,D-carboxypeptidase, hydrolyzing a syntheticpeptide substrate 180 times more efficiently than E. coli PBP5;however, acyl-enzyme deacylation is 20-fold slower than for itsE. coli counterpart, suggesting that PBP3 plays a particularrole in control of peptidoglycan reticulation in the Gram-posi-tive cell wall. Last, the positioning of the active site on theopposite face of the molecule from the COOH-terminal, mem-brane-interacting region may place it in optimal position tocontact the peptidoglycan layer throughout the two cellularhemispheres.

MATERIALS AND METHODS

Measurement of Kinetic and Antibiotic Recognition Parameters—Theconstruction of plasmid pGEX-sPBP3* encoding the soluble form of wildtype PBP3, which lacks both the COOH-terminal helix and the signalpeptide (sPBP3*), was described previously (5). D,D-Carboxypeptidaseactivity was assayed with N,N-diacetyl-L-Lys-D-Ala-D-Ala (Ac2-KAA).sPBP3* (10 nM) was incubated at 37 °C in 50 mM Tris-HCl (pH 8.5), 50mM NaCl, 1 mM EDTA, 0.5 mg/ml bovine serum albumin, and Ac2-KAAat concentrations ranging from 37.5 to 24,000 �M. After various timeintervals, the reaction was stopped by addition of penicillin G to 0.1 mM,and released D-Ala was measured by the method described by Johnsonet al. (10).

The functional homogeneity of the protein sample was determined bytitrating the active sites present in the preparation using [3H]ben-

zylpenicillin (20 Ci/mmol, 1 mCi/ml; Amersham Biosciences) as a re-porter. sPBP3* solutions at 2 and 5 �M were incubated for 15 min at37 °C in 50 mM Tris-HCl (pH 8.0), 200 mM NaCl containing 0.01–20 �M

[3H]benzylpenicillin. The samples were subsequently submitted toSDS-12% PAGE electrophoresis, and estimation of [3H]benzylpenicillinbound to proteins was monitored by two different procedures. The gelwas stained with Coomassie Blue, destained, incubated with Amplify(Amersham Biosciences), dried, and either exposed to film for 16 h orcut around the protein bands. In the latter case, the gel slices weremixed with 5 ml of LSC mixture (Picofluor 15, Packard), and theirradioactivity was measured using a liquid scintillation analyzer (Pack-ard model 2100TR).

To analyze the kinetics of the deacylation reaction, 2 �M purifiedsPBP3* was labeled with 1 �M [3H]benzylpenicillin at 37 °C during 15min in 50 mM Tris-HCl (pH 8.0), 200 mM NaCl. Excess of cold ben-zylpenicillin (15 mM) was then added, and the reaction was continued at37 °C. Aliquots were regularly removed, submitted to SDS-PAGE elec-trophoresis, and the amount of radioactivity was measured in theprotein bands as mentioned above.

The ability of sPBP3* to hydrolyze the pseudo substrate N-benzoyl-D-alanylmercaptoacetic acid (S2d), which is a thioester analog of thestem wall peptide, was explored to generate a comparison profile ofhydrolysis rates for other, previously characterized pneumococcalPBPs. Hydrolysis of S2d was followed by monitoring the amount of thiolgroup released using the method described by Zhao et al. (11).

Crystallization and Structure Solution—Selenomethionine (SeMet)-substituted sPBP3* was expressed in E. coli B834. Cells were grown inLeMaster medium (12) containing 40 mg�l�1 methionine that was pro-gressively replaced by SeMet. Expression was induced at A600 �0.3 andthe purification of the protein was carried out as described previously(5), except that all of the buffers were supplemented with 10 mM

dithiothreitol. Complete replacement of methionine residues by sel-enomethionine was confirmed by electrospray mass spectrometry.

SeMet-labeled sPBP3* crystals were grown by hanging drop vapordiffusion using 1.5 �l of protein solution (4 mg�ml�1), 1.5 �l of well

TABLE IDeacylation rates of different [3H]benzylpenicillin-PBP complexes

Protein k3t1⁄2 of the acyl-enzyme

complex Ref.

� 10�5 s�1 h

sPBP3* 5.7 3.38 This workPBP2x* 3.5 5.80 26PBP1a* 1.0 19.20 23PBP2a* 3.2 6.00 24PBP1b* 5.6 3.44 42SPBP5 (E. coli) 78.0 0.25 9SPBP5� (E. coli) 3.0 6.40K15 (Streptomyces) 10 1.92 43

TABLE IIComparison of the hydrolysis efficiency of the pseudo substrate S2d by

five different PBPs from S. pneumoniae

Protein kcat/Km Ref.

M�1 s�1

SPBP3* 50,500 � 2500 This workPBP2x* 2500 44PBP2b* 80 25PBP1a* 256 23PBP2a* 220 � 20 24

TABLE IIIData collection, phasing, and refinement statistics

Data collectionCell dimensions (Å) a � 87.57, b � 120.69, c � 176.92Space group P212121

Peak Inflection RemoteWavelength (Å) 0.9792 0.9794 0.9393Resolution range (Å) 2.8 3.0 2.8No. of unique/total

reflections47,073/255,278

38,767/200,354 47,278/249,511

Completeness (%) 99.9 (100) 99.8 (99.7) 99.8 (100)Average multiplicity 5.4 (5.0) 5.2 (4.8) 5.3 (5.0)Rsym (%) 10.2 (27.9) 16.1 (50) 13.9 (46.6)

I/�I (last shell) 12.7 (4.7) 9.65 (2.93) 9.94 (2.99)Rano (%) 7.9 (16.0) 9.0 (25.5) 7.8 (23.6)

RefinementResolution (Å) 2.80Rwork/Rfree (%) 21.2 / 26.2Number of residues

Chain A/B/C/D 369/350/368/340

Number of watermolecules

477

Number of iodines 16Number of sulfates 4Average B factor (Å2)

Protein 29.51Solvent 33.30

r.m.s. bond deviation (Å) 0.01r.m.s. angle deviation (°) 2.00

Crystal Structure of PBP3 from S. pneumoniae 15985

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

solution (0.2 M K,Na-tartrate, 0.1 M trisodium citrate (pH 5.6), 1.7 M

(NH4)2SO4, and 3% (v/v) polyethylene glycol 400) and 0.5 �l of 142 mM

NaI per drop. Orthorhombic crystals grew at 20 or 15 °C within 1 week.The crystals belong to space group P212121 with cell dimensions a �87.57 Å, b � 120.69 Å, and c � 176.92 Å and have four molecules in theasymmetric unit. Prior to data collection, crystals were cryoprotected bytransfer into 20% (v/v) glycerol, 2% ethylene glycol, 0.2 M K,Na-tartrate,0.1 M trisodium citrate (pH 5.6), and 1.9 M (NH4)2SO4.

Multiwavelength anomalous diffraction data were collected at 100 Kfrom a single SeMet crystal at peak, inflection, and remote wave-lengths, on an ADSC Quantum 4R CCD detector at beamline ID14EH4at the European Synchrotron Radiation Facility, Grenoble, France. 140degrees of data collected for each wavelength were processed usingMOSFLM (13) and CCP4 (14). Using the peak anomalous data, theselenium sites were located with ShelX-d (15) and refined with SHARP(16). The resulting 44 sites were used for phasing with SHARP. Phaseimprovement with DM and SOLOMON to 2.8 Å produced a clearlyinterpretable electron density map, from which an initial model wasbuilt using ARPwARP (17). The four molecules in the orthorhombicasymmetric unit were sufficiently different so that non-crystallographicsymmetry averaging was not helpful. The model was improved byiterative rounds of manual fitting using O and QUANTA and refine-ment in CNS (18). Structural superpositions were performed usingLSQKAB (14). Figs. 1–3 were prepared with Molscript (19) andRaster3D (20).

RESULTS

sPBP3* Is a Highly Efficient D,D-Carboxypeptidase—The in-teraction between PBPs and peptidoglycan substrates or �-lac-tam antibiotics obeys a three-step reaction that is representedas follows,

E � I-|0k1

k�1

EIO¡k2

EI*O¡k3

E � P (Eq. 1)

where E is the active PBP enzyme, I is the substrate, EI is thenon-covalent Michaelis-Menten complex, EI* is the acyl-en-zyme covalent complex, and P is the product of the reaction(21).

Enzymatic parameters for carboxypeptidation were esti-mated by measuring the initial velocities (�) at various concen-trations of Ac2-KAA. Inhibition of sPBP3* by its own substratewas observed above a ligand concentration of 15 mM, a phe-nomenon that has been previously reported for a the D,D-car-boxypeptidase from Neisseria gonorrhoeae (22). The values andstandard error of kcat � 110 � 10 s�1 and Km � 19 � 3 mM

(thus, kcat/Km � 5689 M�1�s�1) were obtained by fitting thedata points, from zero to peak activity, against the equation�/[sPBP3*]T � kcat [Ac2-KAA]/(Km � [Ac2-KAA]); these esti-mates may be lower than the true kcat and Km due to theinhibition of the enzyme at high ligand concentrations. Bycomparison, the kcat/Km value measured with this substrate forE. coli PBP5 is 32 M�1�s�1 (9).

The k3 deacylation rate calculated for sPBP3* is comparablewith those reported for other S. pneumoniae PBPs (Table I) butis 20–30-fold lower than that of E. coli PBP5 (9). In addition,sPBP3* possesses an efficiency of hydrolysis of 50 500 � 2500M�1�s�1 for the pseudo substrate S2d. This value is 200–1000-fold higher than for PBP1a, PBP2a, and PBP2b and 20-foldhigher than for PBP2x (Table II) (23–27). These high hydrolyticefficiency values suggest that PBP3 may play an importanthydrolytic role during the peptidoglycan biosynthetic process.

Overview of the sPBP3* Structure—The structure of sPBP3*was determined by multiwavelength anomalous diffraction us-ing seleno-methionyl-substituted protein; data collection, phas-ing, and refinement statistics are shown in Table III. The finalstructure, which includes four 360-amino acid monomers in theasymmetric unit, has an R factor of 21.2% (Rfree � 26.2%) at 2.8Å; 83.1% of the residues lie within the most favored region ofthe Ramachandran plot and 477 water molecules are includedin the model.

The sPBP3* monomer consists of a single polypeptide chainorganized into two domains that are orientated approximatelyat right angles to each other. The spatial relationship betweenthe domains is reminiscent of that observed for its E. colihomolog, the PBP5 D,D-carboxypeptidase, with which PBP3shares 27% sequence identity (Fig. 1). Domain I (cyan in Fig. 1,A and B) comprises residues 25–292 and bears the signaturefold topology of the penicilloyl-serine transferase superfamily,thus harboring the active site of PBP3. Using the standardsecondary structure classification of the class A and class C�-lactamases as well as PBP2x (27, 28), the first domain isprincipally constituted by a central five-stranded antiparallel�-sheet (�3/�4/�5/�1/�2) and two main helices (�8/�11). Inaddition, this domain contains three two-stranded antiparallel�-sheets (�2a/�2d, �2b/�2c, and �2e/�2f) and three helices(�2a/�4/�5).

The COOH-terminal domain II bears an elongated structure

FIG. 1. Ribbon diagrams of sPBP3* and PBP5. sPBP3* from S. pneumoniae (A) and PBP5 from E. coli (B) have their NH2-terminal domains(I) in cyan and COOH-terminal domains (II) in violet. The sulfate atom (yellow and red) located in the active site of sPBP3* and the iodines (blue)that interact with the protein are represented. In this view, the cytoplasmic membrane is located at the bottom of the molecules, and the entranceof the active site is at the top.

Crystal Structure of PBP3 from S. pneumoniae15986

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

which comprises residues 293–393 and is formed by a sandwichbetween two anti-parallel �-sheets. Comparison of the fourmolecules, which exist in the asymmetric unit, reveals that thegreatest differences at the C� level map to loop regions withinthe COOH-terminal domain, where the electron density is ofpoor quality in certain regions. Notably, the large area of in-teraction between the surfaces of domains I and II (�800 Å2,including six potential hydrogen bonds) guarantees stabilityfor the full-length molecule, which is reflected by the slightvariation (�1°) of the angle between domains I and II in all fourmolecules of the asymmetric unit.

The sPBP3* Active Site—The sPBP3* active site is at thedistal end from the COOH terminus of the molecule. As ob-served with the other carboxypeptidases and penicillin-metabolyzing enzymes, the active site is mainly defined bythree conserved structural motifs: SXXK (Ser56-Ile57-Thr58-Lys59), which includes the nucleophilic Ser56 residue, posi-tioned at the NH2-terminal end of helix �2; SXN (Ser119-Ala120-Asn121), which forms the turn between helices �4 and �5 on theleft side of the cavity, and K(T/S)G (Lys239-Thr240-Gly241),which lines strand �3 (Fig. 2A). In addition, the backbone NHgroups of the essential Ser56 and Thr242 residues occupy posi-tions that are compatible with the oxyanion hole function re-quired for catalysis. The NH2 terminus of helix �11 and theloop between �6 and �2d also contribute residues to the activesite. These include Arg278, located at the right top angle of the

cavity, Thr160, and the structural Gly161, present on the ex-tended loop at the bottom of the cavity.

The hydrogen bonding network within the active site is ex-tensive (Fig. 2A) and is identical in all four molecules in theasymmetric unit. The �-NH2 group of Lys59 plays a central rolein this network, forming hydrogen bonds with the hydroxylgroup of Ser56 and Ser119, the side chain carbonyl group ofAsn121 and the backbone carbonyl group from Thr160. Twowater molecules are observed within the hydrogen bondingnetwork, one of which (O-26) is conserved in the K15 transpep-tidase (29), as well as in E. coli PBP5 (8, 9). Although thearchitecture of the active site of the three enzymes is similar(compare Fig. 2, A–C), some significant differences can be ob-served in the orientation of three important catalytic residues:the side chains of Ser110, Lys213, and Thr214 of PBP5 are ori-ented differently from the equivalent residues in sPBP3*,Ser119, Lys239, and Thr240, respectively. In particular, in PBP5,Ser110 and Lys213 point away from the active site, and conse-quently the classical hydrogen bonding network within theactive site is not formed in this molecule. It is of interest thatall three molecules, which recognize peptidic substrates, har-bor a conserved glycine residue at the bottom of the cleft(Gly161, Gly152, and Gly144 in sPBP3*, PBP5, and K15, respec-tively). The importance of the absence of a side chain at thisposition becomes evident if one considers that it is part of abinding pocket which could accommodate the penultimate

FIG. 2. Active site architecture of sPBP3* from S. pneumoniae (A) (stereo view), PBP5 from E. coli (B), and the Streptomyces K15D,D-transpeptidase (C). The main catalytic residues are represented in ball and stick and are colored by atom type. Several potential hydrogenbonds are denoted by orange dashed lines. The secondary structures are shown with ribbon diagram; the helices are colored red, and the �-sheetsare green. Proteins were crystallized at pH values of 5.6, 7.0, and 7.2, respectively.

Crystal Structure of PBP3 from S. pneumoniae 15987

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

D-alanine residue of the peptidic substrate, as is the case for thewell studied R61 D,D-peptidase (30).

The overall structure of sPBP3* is highly reminiscent of thatof PBP5 (9) (Fig. 3A). Domains I of the two structures super-impose with an r.m.s. deviation of 1.09 Å; the main differencesinclude helix �2a, which is replaced by an irregular loop inPBP5, and the omega-like loop (residues 156–181 in sPBP3*and residues 147–158 in PBP5), which is longer in sPBP3* dueto an insertion of 14 residues (yellow in Fig. 3A). A comparisonof domain II for both molecules reveals greater differencesincluding an r.m.s. deviation between C� atoms of 2.83 Å. InsPBP3*, this region is clearly less compact than in PBP5, withshorter secondary structure elements.

A particularly outstanding feature of the sPBP3* structure isthe close similarity of domain I with the general fold of theStreptomyces K15 transpeptidase (Ref. 29; Fig. 3B); this obser-vation is surprising, since PBP3 catalyzes only carboxypepti-dation, while K15 is a transpeptidase (and must first catalyzea carboxypeptidation of the terminal D-Ala residues beforetranspeptidating with an amino acceptor group). Domain I ofsPBP3* can be superimposed onto the structure of thetranspeptidase K15 (Ref. 29; Protein Data Bank code 1SKF)with an r.m.s. deviation of 1.54 Å. The same calculation per-formed with the structure of TEM-1 �-lactamase (Protein DataBank code 1BTL) reveals a C� rms deviation of 1.70 Å. As is thecase for PBP5, both K15 and TEM-1 harbor shorter omega-likeloops than sPBP3* (compare Fig. 3, A–C).

The secondary structural elements that harbor the con-served catalytic residues of the penicillin-binding domain of thetranspeptidase PBP2x of S. pneumoniae (Ref. 28; Protein DataBank code 1PMD) PBP2a of Staphylococcus aureus (31; ProteinData Bank code 1MWR) and domain I of sPBP3* are similarlypositioned (Fig. 3), with �3, �2, �4, and �5 playing importantroles. However, outside of the active site, all of the other re-

gions display large differences, as can be observed from thepoor superposition results for these molecules (Fig. 3, D and E).

The Omega-like Loop; a Key Structural Feature within D,D-Carboxypeptidases—The omega-like loop of class A �-lactama-ses harbors residues that are required both for maintenance ofactive site topology and for enzymatic activity. The large struc-tural deviations observed between sPBP3* and PBP5 at thelevel of the omega loop prompted us to explore this regionwithin the sequences of other putative “PBP5-like” D,D-car-boxypeptidase homologues in different bacterial species. Ourgenomic search focused on proteins with potential or demon-strated D,D-carboxypeptidase activity and for which a topologysimilar to sPBP3* or PBP5 was predicted, including the pres-ence of a signal peptide, a penicillin-binding domain (analogousto domain I), a COOH-terminal extension of �100 residues(analogous to domain II), and an amphiphilic helix at theCOOH terminus (Fig. 4). In addition, only carboxypeptidasesharbored by bacteria whose cell wall composition is knownwere chosen for the study.

This primary sequence analysis revealed the existence of twogroups of PBP5-like D,D-carboxypeptidases that could be differ-entiated based uniquely on the number of residues of theiromega-like loops. Interestingly, this variation in length couldbe correlated to the chemical composition of the bacterial pep-tidoglycan. Indeed, PBP5-like D,D-carboxypeptidases from bac-terial species whose peptidoglycan harbors stem peptides witha diaminopimelate moiety in the third position, such as PBP5from E. coli or the PBP from B. subtilis, are characterized by ashort omega-like loop. On the other hand, the primary se-quences of PBP5-like D,D-carboxypeptidases from specieswhose peptidoglycan possesses stem peptides with a lysineresidue in the third position, such as PBP3 from S. pneu-moniae, display a much longer omega-like loop (in this case,corresponding to an insertion of 14 residues). Interestingly, the

FIG. 3. Comparison of the structureof the NH2-terminal carboxypepti-dase domain of sPBP3* from S. pneu-moniae with: the same domain fromPBP5 from E. coli (A), StreptomycesK15 transpeptidase (B), TEM-1 �-lac-tamase from E. coli (C), the transpep-tidase domain of PBP2a from S. au-reus (D), and the transpeptidasedomain of PBP2x from S. pneu-moniae R6 (E). The overall fold ofsPBP3* is in green, with the omega-likeloop colored in yellow. The other struc-tures are in violet, with the omega-likeloop colored in blue (except for PBP2a andPBP2x, which do not harbor classicalomega loops).

Crystal Structure of PBP3 from S. pneumoniae15988

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

same carboxypeptidases that carry the longer omega-like loopsalso harbor a 6-amino acid insertion downstream from thethird catalytic motif (KTG; see Fig. 4), which, in the tertiaryfold, is in proximity to the omega loop. It is thus conceivablethat such structural characteristics provide differential sub-strate recognition within the peptidoglycan biosyntheticmachinery.

DISCUSSION

Lmm PBPs are membrane-associated enzymes that play im-portant roles in the maintenance of cell shape and in cellulargrowth and division processes (5, 6, 32–34). Here, we havebiochemically and structurally characterized a soluble form ofPBP3 from the human pathogen S. pneumoniae. This work hasconfirmed that the single lmm PBP from S. pneumoniae cata-lyzes a D,D-carboxypeptidation reaction, although its catalyticdomain is highly reminiscent of the structure of the K15transpeptidase from Streptomyces. In addition, sPBP3* is alsoable to recognize �-lactam antibiotics, as expected from thepresence of the three conserved penicilloyl serine transferasesuperfamily motifs present in domain I (SXN, SXXK, K(T/S)G).Notably, both D,D-carboxypeptidase activity and �-lactam proc-essing proceed through the formation of an acyl-enzyme inter-mediate followed by deacylation of the enzyme. Based on these

data as well as structural and mechanistic information that isavailable for other PBPs and �-lactam-recognizing enzymes (8,9, 22–31, 36), we propose that the acylation mechanism ofsPBP3* may occur in four steps. Initially, there is formation ofa non-covalent complex that may be stabilized by interactionsbetween the carboxylate group of the ligand and conserved sidechains of Thr240 and Arg278. Backbone nitrogen atoms of Thr242

and Ser56 could play the role of the oxanion hole required forstabilization of the intermediate reactive species. Ser56 is acti-vated to form the ester bond with the substrate through ab-straction of a proton from neighboring Ser119 by the COOH ofthe ligand. This is followed by protonation of the nitrogen of the�-lactam ring by the carboxylate and subsequently ring break-age (35). An alternative concerted mechanism is possible,whereby the hydrogen from Ser119 is transferred to the nitro-gen of the �-lactam ring at the same time as the hydrogen ofSer56 is transferred to Ser119, and the former forms the acylbond with the antibiotic (37).

sPBP3* is exceptionally efficient in hydrolyzing the pseudosubstrate S2d, over 20 times more active than the most efficientpneumococcal enzyme measured to date, PBP2x (Table II). In ad-dition, the kcat/Km value measured with the synthetic peptidesubstrate N,N-diacetyl-L-lys-D-Ala-D-Ala (5689 M�1�s�1) is 180-fold

FIG. 4. Alignment of sequences of PBP3 from S. pneumoniae, PBP5 from E. coli, and six PBP5-like D,D-carboxypeptidases. Thenumbering scheme follows that of the longest sequence, the E. faecalis enzyme (and thus does not relate to the one employed in the descriptionof the sPBP3* structure). In the S. pneumoniae enzyme, the carboxypeptidase domain ranges from residues 55 to 335. Note that S. pneumoniae,Streptococcus mutans, Enterococcus faecalis, and S. aureus, all of which posses a lysine residue in the third position of the stem peptide, harboran insertion at the level of the omega-loop, as well as between amino acids 314 and 319. Strictly conserved residues are in red, residues conservedin at least half of the species analyzed are in green, and residues that display side chain similarity are in blue.

Crystal Structure of PBP3 from S. pneumoniae 15989

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

greater than the value measured for PBP5 with the same substrate(32 M�1�s�1; (9)), indicating a greater catalytic efficiency for thepneumococcal enzyme. This observation suggests that PBP3 playsan important role in control of peptidoglycan reticulation inS. pneumoniae. Indeed, Gram-positive organisms have several lay-ers of peptidoglycan, while Gram-negative bacteria, such as E. coli,have much smaller amounts. In addition, E. coli possesses at leastthree D,D-carboxypeptidases (PBP5, PBP6, PBP6b), and althoughthe precise function of each one in the cell cycle is unknown, it isconceivable that they could share peptidoglycan processing dutieswithin the Gram-negative cell wall. Hence, it is not surprising thatPBP3, the only D,D-carboxypeptidase in S. pneumoniae, must be anenzyme with a very high catalytic activity, since it must limit theamount of pentapeptidic stem peptides in the peptidoglycanthroughout the entire bacterial cell surface, with the exception ofthe division site, as shown by immunofluorescence localizationstudies (5).

In the structures of class A �-lactamases, the omega-like loopis a 19-residue stretch comprised of amino acids which partic-ipate both in catalysis and maintenance of local topology. In-sertional mutagenesis studies showed that enzymes withlarger omega-like loops were still able to perform catalysis, andin addition, displayed expanded substrate specificities (38). Inthis study, we have identified that organisms which display alysine residue as the third component of the peptidoglycanpentapeptide harbor D,D-carboxypeptidases with large omega-like loops as well as an insertion of six amino acids �30 resi-dues downstream from the catalytic KTG motif. Conversely,bacteria whose peptidoglycan displays a diaminopimelate res-idue at the same position carry D,D-carboxypeptidases withshort omega loops (and lack the 6-residue insertion). The in-sertion of the amino acid at the third position in the stempeptide is catalyzed by MurE ligases, which show high speci-ficity for their respective substrates (39); notably, if the incor-rect residue is inserted at this position, one of the subsequentreactions, transpeptidation, does not take place (40). Thus, it isconceivable that, much like the transpeptidase enzymes thatwill not bind the incorrect amino acid at the third position ofthe stem peptide, D,D-carboxypeptidases also require specificityat this position for catalysis. Considering that the carboxypep-tidation reaction may require the recognition of a large portionof the stem peptide, including the third moiety and maybebeyond, it is conceivable that the binding region generated bythe omega-like loop, and the 6-residue insertion may be able toparticipate in substrate discrimination. Hence, carboxypepti-dases involved in peptidoglycan metabolism may have evolvedto perform optimal recognition of a large portion of the stempeptide.

Interaction with and Accessibility to Peptidic Substrates—Although lipid II, the natural substrate for PBP enzymes,carries only pentapeptides in its stem moiety, it is well docu-mented that the pre-existent (available) murein can also har-bor multimeric peptides (41). Although the reaction of D,D-carboxypeptidation of stem pentapeptides has been suggestedas being necessary to regulate the degree of peptidoglycanreticulation, it is still unclear whether carboxypeptidases areable to interact with more complex substrates, i.e. with reticu-lated peptidic molecules. The close similarity between the gen-eral folds as well as the active sites of K15 and sPBP3* sug-gests that it may be the case. The fact that the K15 active sitecleft must harbor two peptidic chains to catalyze the transpep-tidation reaction suggests that the highly similar sPBP3* maybe able to recognize stem peptides that are more complex thanthe regular pentapeptide. This idea is in agreement with theobservation that PBP3 is present throughout the entire bacte-rial cell surface, except at the site where hmm PBPs are posi-

tioned, where peptidoglycan synthesis, which requires sub-strates with intact stem peptides, occurs (5). PBP3 thus mustbe a highly efficient processing enzyme, which eliminates theCOOH-terminal D-Ala group from all available stem peptides,which are localized outside from the cell division site (inde-pendent of their configuration), thus ensuring that intact pep-tidoglycan will be present uniquely at its initial synthesis site.

PBP5 from E. coli has been suggested as being poised verti-cally on the bacterial membrane, i.e. with the longer axis of theprotein placed perpendicularly to the bilayer (8), an orientationthat could be shared by pneumococcal PBP3. Hence, domain IIwould serve as a pedestal that would bring the active site closeenough to the peptidoglycan layer to interact with peptidic sidechains that may extend in the direction of the inner membrane(Fig. 5). If PBP3 is placed vertically on the membrane, it wouldcover an approximate distance of 80 Å (8 nm) from the mem-brane and thus could interact with the stem or reticulatedpeptidic chains carried by the lower layers of peptidoglycan. Itis conceivable that the COOH-terminal amphiphilic helix couldadd a certain flexibility to this orientation, enhancing the ac-cessibility of the enzyme to its substrates.

In summary, PBP3 is a highly efficient D,D-carboxypeptidase,which, at the beginning of the cell cycle, trims stem and/orreticulated pentapeptides harbored by the pre-existent mureinto generate inert regions of cell envelope. Its detection, byimmunofluorescence, throughout the bacterial surface (5), cou-pled to the abovementioned model where the amphiphilic helixmay permit superficial association to the membrane, suggeststhat it may “skid” on the surface of the membrane, trimmingpentapeptides. Donor pentapeptides for the transpeptidationreaction, required for peptidoglycan synthesis, would then onlybe available at the future division site. Once murein synthesisis initiated, the D,D-carboxypeptidase could regain the regionswith young peptidoglycan where it could again interact withstem or/and reticulated pentapeptides to regulate the reticula-tion degree of the future mature murein.

Acknowledgments—We are grateful to M. Nanao and the EuropeanSynchrotron Radiation Facility ID14 EH4 beamline staff for help withdata collection as well as A. Zapun (Institut de Biologie StructuraleJean-Pierre Ebel, Grenoble, France) for critical review of themanuscript.

REFERENCES

1. Nanninga, N. (1998) Microbiol. Mol. Biol. Rev. 62, 110–1292. Williamson, R., Hakenbeck, R., and Tomasz, A. (1980) Antimicrob. Agents

FIG. 5. Possible orientation of PBP3 in the periplasm ofS. pneumoniae. The surface potential of sPBP3* is shown; the redzones are negatively charged, and the blue ones are positively charged.The D,D-carboxypeptidase active site appears as an indentation in do-main I. The entire protein is tethered to the outer face of the innermembrane by a stretch of �20 carboxyl-terminal residues that havebeen postulated to form an amphiphilic helix.

Crystal Structure of PBP3 from S. pneumoniae15990

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

Chemother. 18, 629–6373. Hakenbeck, R., and Kohiyama, M. (1982) Eur. J. Biochem. 127, 231–2364. Morlot, C., Zapun, A., Dideberg, O., and Vernet, T. (2003) Mol. Microbiol. 50,

845–8555. Morlot, C., Noirclerc-Savoye, M., Zapun, A., Dideberg, O., and Vernet, T.

(2004) Mol. Microbiol. 51, 1641–16486. Schuster, C., Dobrinski, B., and Hakenbeck, R. (1990) J. Bacteriol. 172,

6499–65057. Tipper, D. J., and Strominger, J. L. (1965) Proc. Natl. Acad. Sci. U. S. A. 54,

1133–11418. Davies, C., White, S. W., and Nicholas, R. A. (2001) J. Biol. Chem. 276,

616–6239. Nicholas, R. A., Krings, S., Tomberg, J., Nicola, G., and Davis, C. (2003) J. Biol.

Chem. 278, 52826–5283310. Johnson, K., Duez, C., Frere, J. M., and Ghuysen, J. M. (1975) Methods

Enzymol. 43, 687–69811. Zhao, G., Yeh, W. K., Carnahan, R. H., Flokowitsch, J., Meier, T. I., Alborn,

W. E., Jr., Becker, G. W., and Jaskunas, S. R. (1997) J. Bacteriol. 179,4901–4908

12. LeMaster, D. M., and Richards, F. M. (1985) Biochemistry 24, 7263–726813. Leslie, A. G. M. (2002) MOSFLM User Guide, Mosflm version 6.2.2, MRC

Laboratory of Molecular Biology, Cambridge, UK14. Collaborative Computing Project 4 (1994) Acta Crystallogr. Sect. D Biol. Crys-

tallogr. 50, 760–76615. Sheldrick, G. M. (1997) SHELXL97: Program for the Refinment of Crystal

Structure, University of Gottingen, Gottingen, Germany16. de La Fortelle, E., and Bricogne, G. (1997) Methods Enzymol. 276, 472–49417. Morris, R. J., Perrakis, A., and Lamzin, V. S. (2003) Methods Enzymol. 374,

229–24418. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-

Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read,R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr.Sect. D Biol. Crystallogr. 54, 905–921

19. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946–95020. Merritt, E. A., and Murphy, M. E. P. (1994) Acta Crystallogr. Sect. D Biol.

Crystallogr. 50, 869–87321. Frere, J. M., Ghuysen, J. M., and Iwatsubo, M. (1975) Eur. J. Biochem. 57,

343–35122. Stefanova, M. E., Tomberg, J., Olesky, M., Holtje, J. V., Gutheil, W. G., and

Nicholas, R. A. (2003) Biochemistry 42, 14614–1462523. Di Guilmi, A. M., Mouz, N., Andrieu, J. P., Hoskins, J., Jaskunas, S. R.,

Gagnon, J., Dideberg, O., and Vernet, T. (1998) J. Bacteriol. 180, 5652–565924. Di Guilmi, A. M., Mouz, N., Martin, L., Hoskins, J., Jaskunas, S. R., Dideberg,

O., and Vernet, T. (1999) J. Bacteriol. 181, 2773–278125. Pagliero, E., Chesnel, L., Hopkins, J., Croize, J., Dideberg, O., Vernet, T., and

Di Guilmi, A. M. (2004) Antimicrob. Agents Chemother. 48, 1848–185526. Mouz, N., Di Guilmi, A. M., Gordon, E., Hakenbeck, R., Dideberg, O., and

Vernet, T. (1999) J. Biol. Chem. 274, 19175–1918027. Lobkovsky, E., Moews, P. C., Liu, H., Zhao, H., Frere, J. M., and Knox, J. R.

(1993) Proc. Natl. Acad. Sci. U. S. A. 90, 11257–1126128. Pares, S., Mouz, N., Petillot, Y., Hakenbeck, R., and Dideberg, O. (1996) Nat.

Struct. Biol. 3, 284–28929. Fonze, E., Vermeire, M., Nguyen-Disteche, M., Brasseur, R., and Charlier, P.

(1999) J. Biol. Chem. 274, 21853–2186030. Pratt, R. F. (2002) J. Chem. Soc. 2, 851–86131. Lim, D., and Strynadka, N. (2002) Nat. Struct. Biol. 9, 870–87632. Nelson, D. E., and Young, K. D. (2000) J. Bacteriol. 182, 1714–172133. Nelson, D. E., and Young, K. D. (2001) J. Bacteriol. 183, 3055–306434. Popham, D. L., and Young, K. D. (2003) Curr. Opin. Microbiol. 6, 594–59935. Diaz, N., Suarez, D., Sordo, T. L., and Merz, K. M. (2001) J. Phys. Chem. B 105,

11302–1131336. McDonough, M. A., Anderson, J. W., Silvaggi, N. R., Pratt, R. F., Knox, J. R.,

and Kelly, J. (2002) J. Mol. Biol. 322, 111–12237. Oliva, M., Dideberg, O., and Field, M. J. (2003) Proteins 53, 88–10038. Hayes, F., Hallet, B., and Cao, Y. (1997) J. Biol. Chem. 272, 28833–2883639. Mengin-Lecreulx, D., Blanot, D., and van Heijenoort, J. (1994) J. Bacteriol.

176, 4321–432740. Mengin-Lecreulx, D., Falla, T., Blanot, D., van Heijenoort, J., Adams, D. J.,

and Chopra, I. (1999) J. Bacteriol. 181, 5909–591441. Severin, A., Schuster, C., Hakenbeck, R., and Tomasz, A. (1992) J. Bacteriol.

174, 5152–515542. Di Guilmi, A. M., Dessen, A., Dideberg, O., and Vernet, T. (2003) J. Bacteriol.

185, 1650–165843. Leyh-Bouille, M., Nguyen-Disteche, M., Pirlot, S., Veithen, A., Bourguignon,

C., and Ghuysen, J. M. (1986) Biochem. J. 235, 177–18244. Mouz, N., Gordon, E., Di Guilmi, A. M., Petit, I., Petillot, Y., Dupont, Y.,

Hakenbeck, R., Vernet, T., and Dideberg, O. (1998) Proc. Natl. Acad. Sci.U. S. A. 95, 13403–13406

Crystal Structure of PBP3 from S. pneumoniae 15991

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from

Vernet, Otto Dideberg and Andréa DessenCécile Morlot, Lucile Pernot, Audrey Le Gouellec, Anne Marie Di Guilmi, Thierry

Streptococcus pneumoniaeCrystal Structure of a Peptidoglycan Synthesis Regulatory Factor (PBP3) from

doi: 10.1074/jbc.M408446200 originally published online December 13, 20042005, 280:15984-15991.J. Biol. Chem. 

  10.1074/jbc.M408446200Access 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/280/16/15984.full.html#ref-list-1

This article cites 42 references, 22 of which can be accessed free at

by guest on January 13, 2020http://w

ww

.jbc.org/D

ownloaded from