Penicillin-binding Protein 2x of Streptococcus pneumoniae: Three New Mutational Pathways for Remodelling an Essential Enzyme into a Resistance Determinant

J. Mol. Biol. (2008) 376, 1403–1416

doi:10.1016/j.jmb.2007.12.058

Available online at www.sciencedirect.com

Penicillin-binding Protein 2x of Streptococcus pneumoniae: Three New Mutational Pathways for Remodelling an Essential Enzyme into a Resistance Determinant Patrick Maurer 1 , Barbara Koch 1 , Ilka Zerfaß 1 , Jan Krauß 1 , Mark van der Linden 1 , Jean-Marie Frère 2 , Carlos Contreras-Martel 3 and Regine Hakenbeck 1 ⁎ 1

Department of Microbiology, University of Kaiserslautern, Paul-Ehrlich Straße 23, D-67663 Kaiserslautern, Germany 2

Institut de Chimie B6, Université de Liège, B-4000 Liège, Belgium 3 Institut de Biologie Structurale Jean-Pierre Ebel (CEA/CNRS/ UJF/PSB), Laboratoire des Protéines Membranaires, 41 rue Jules Horowitz, 38027 Grenoble, France

Received 31 August 2007; received in revised form 14 December 2007; accepted 21 December 2007 Available online 4 January 2008

Mutations in the transpeptidase domain of penicillin-binding protein 2x (PBP2x) of Streptococcus pneumoniae that reduce the affinity to beta-lactams are important determinants of resistance to these antibiotics. We have now analyzed in vitro and in vivo properties of PBP2x variants from cefotaximeresistant laboratory mutants and a clinical isolate. The patterns of two to four resistance-specific mutations present in each of the proteins, all of which are placed between 6.6 and 24 Å around the active site, fall into three categories according to their positions in the three-dimensional structure. The first PBP2x group is characterized by mutations at the end of helix α11 and carries the well-known T550A change and/or one mutation on the surface of the penicillin-binding domain in close contact with the C-terminal domain. All group I proteins display very low acylation efficiencies, ≤1700 M− 1 s− 1, for cefotaxime. The second class represented by PBP2x of the mutant C505 shows acylation efficiencies below 100 M− 1 s− 1 for both cefotaxime and benzylpenicillin and contains the mutation L403F at a critical site close to the active serine. PBP2x of the clinical isolate 669 reveals a third mutational pathway where at least the two mutations Q552E and S389L are important for resistance, and acylation efficiency is reduced for both beta-lactams to around 10,000 M− 1 s− 1. In each group, at least one mutation is located in close vicinity to the active site and mediates a resistance phenotype in vivo alone, whereas other mutations might exhibit secondary effects only in context with other alterations. © 2008 Elsevier Ltd. All rights reserved.

Edited by F. Schmid

Keywords: penicillin-binding protein; penicillin resistance; Streptococcus pneumoniae; PBP2x; cefotaxime

Introduction *Corresponding author. E-mail address: [email protected]. Present addresses: B. Koch, Department of Biophysics, University of the Saarland, Homburg, Germany; M. van der Linden, German National Reference Center for Streptococci, Department of Medical Microbiology, University Hospital RWTH, Aachen, Germany. Abbreviations used: PBP, penicillin-binding protein; GST, glutathione S-transferase; MIC, minimal inhibitory concentration.

Penicillin resistance in several Gram-positive pathogenic bacteria is related to alterations in the target proteins of beta-lactam antibiotics, essential penicillin-binding proteins (PBPs). 1,2 PBPs are enzymes involved in metabolism of murein, a macromolecule exclusively found in prokaryotes. The transpeptidation reaction between two muropeptides resulting in the cross-linked structure of the mature peptidoglycan network is the crucial penicillin-sensitive step. PBPs are inhibited by beta-

0022-2836/$ - see front matter © 2008 Elsevier Ltd. All rights reserved.

Streptococcus pneumoniae PBP2x

1404 lactams by forming a relatively stable, covalent penicilloyl–enzyme complex via the active-site serine residue located in the penicillin-binding/transpeptidase domain of all PBPs.3 Alterations that confer resistance reduce the affinity to beta-lactams; thus, higher concentrations of the drug are required for in vivo inhibition of the cells. Whereas such mutations interfere with the interaction of the PBP enzyme with the inhibitor, they should not seriously affect their in vivo functions. Streptococcus pneumoniae represents a paradigm of the evolutionary power of a human pathogen where the dramatic emergence and spread of penicillinand multiple-antibiotic-resistant strains over the past decades has been observed worldwide.4 S. pneumoniae contains six PBPs: the three high molecular weight class A PBPs 1a, 1b and 2a, the class B high molecular weight PBPs 2x and 2b, and the low molecular weight D,D-carboxypeptidase PBP3. Although all six PBPs in S. pneumoniae have been implicated in the resistance process,5,6 alterations that result in low-affinity variants in the class B high molecular weight PBP2x and PBP2b are a prerequisite for high-level resistance mediated by PBP1a, and PBP2a, 1b and 3 have been recognized as low-affinity PBPs only occasionally (for a review, see Ref. 1 and references within). PBP2x of S. pneumoniae that has one of the highest affinities for penicillin among the penicillin-interactive proteins is an ideal model for studying the structure–function relationship of mutant derivatives that affect the interaction with beta-lactam antibiotics. It is one of the primary determinants of beta-lactam resistance in this organism. It is the first essential class B PBP whose three-dimensional structure has been resolved at high resolution,7,8 and the structure of a clinical isolate containing multiple alterations is also available.9 Since PBPs are perfect antibiotic targets, modeling of mutant proteins with a low affinity to the classical betalactam antibiotics will help in the design of new compounds that are able to interact with such lowaffinity protein variants and hence represent potent antimicrobial agents. Therefore the identification of sites crucial for the interaction with beta-lactams is of major interest. A variety of point mutations related to resistance have been identified in PBP2x of laboratory mutants as well as in clinical isolates.10–17 PBP2x mutations selected in a set of independently obtained spontaneous cefotaxime-resistant laboratory mutants revealed a surprising multitude of sites involved in resistance development for this essential protein, all of which were scattered throughout the penicillinbinding domain.10,11 Mutations relevant for resistance in the mosaic genes of clinical isolates appear to be distinct, and only the mutation T550A, which can easily be selected in the laboratory,18,19 has been found occasionally in high-level cefotaxime-resistant isolates12,20 and as a single mutation in lowlevel resistant strains.21 Although residues important for the interaction with beta-lactams have been identified in clinical isolates, the multitude of amino

acid alterations and the variability of the mosaic blocks observed have not allowed the identification of all the mutations relevant for the resistance phenotype. In fact, comparison of a large number of diverse mosaic PBP2x revealed only two sites common to many distinct mosaic designs whose impact on resistance could be confirmed by functional tests in vitro: T338 adjacent to the active-site S337 is altered in one group of mosaic PBP2x (mostly to A338), whereas another group contains the mutation Q552E often without a mutated T338.22,23 All other mutations revealed so far occur only in subfamilies of related mosaic PBP2x or are restricted to rare variants. We have therefore analyzed PBP2x of cefotaximeresistant laboratory mutants where the entire set of mutations selected during resistance development is known. The order in which the PBP2x mutations were introduced during the selection procedure has now been determined for four mutant families. A mosaic PBP2x from a clinical isolate was also included, since it contained only five changes that have not been found in homologues from sensitive Streptococcus mitis, all of which differed from those found in the laboratory mutants. Soluble PBP2x mutant derivatives were isolated from Escherichia coli strains overexpressing the protein for in vitro determination of their acylation efficiencies. With the high-resolution structure of PBP2x the position of the amino acid changes was determined, revealing new structural elements important for the interaction with beta-lactam antibiotics. A series of transformants was isolated from the sensitive laboratory strain S. pneumoniae R6 using the cloned pbp2x variants in order to determine the resistance potential mediated in vivo by the PBP2x variants.

Results PBP2x in cefotaxime-resistant laboratory mutants The family trees of the cefotaxime-resistant laboratory mutants used in the present study are shown in Fig. 1a. C501, C503, C505 and C606 represent independently selected derivatives of the penicillinsensitive laboratory strain S. pneumoniae R6 obtained after five or six selection steps in the mutant families C001, C003, C005 and C006, and differ in terms of cefotaxime resistance levels and cross-resistance to other beta-lactams.24 Each of these mutants contains a distinct PBP2x variant with three to four mutations.10,11 The selection procedure has been described in detail.24 In brief, R6 cells were plated on cefotaxime concentrations at and above the minimal inhibitory concentration (MIC) value (0.015 – 0.02 µg/ml). Spontaneous mutants resisting these concentrations were used for the second selection step using higher antibiotic concentrations and so forth, until several independent mutant lineages

Streptococcus pneumoniae PBP2x

1405

Fig. 1. Mutations in PBP2x implicated in resistance development for beta-lactams. (a) The cefotaximeresistant mutant lineages and the occurrence of PBP2x mutations during the selection procedure is shown. The first digit indicates the selection step and the last digit specifies the mutant lineage. (b) The central penicillin-binding/transpeptidase domain of PBP2x with the three conserved boxes S*337TMK, S395SN and K547SG is shown above. The amino acid residues of the beta-lactam susceptible R6 strain are indicated on top, and the position of the PBP2x mutations in individual cefotaxime-resistant laboratory mutants are shown below. In PBP2x of the clinical isolate 669, the three alterations are shown in bold lines that are present in the resistant R6-transformant (R6-T669) and are distinct from alterations found in PBP2x homologues of penicillin-susceptible S. mitis; other amino acid changes are indicated by thin lines. (c) Amino acid changes in PBP2x of S. pneumoniae 669 compared to the sensitive R6 strain. The vertical numbers on top indicate the amino acid residues. The region present in R6-T669 is boxed. Amino acids not found in homologues of sensitive streptococci are bold and shaded gray.

consisting of up to six mutants resisting increasing concentrations of cefotaxime were obtained. They are numbered according to the selection step (first digit) and the mutant lineage (last digit). The PBP2x of the mutants, C501, C503 and C606, contain mutations at similar sites located between residues 597 to 601 at the C-terminal end of the penicillin-binding domain (Fig. 1b) and are summarized in group I. C501 and C503 have the mutation T550A adjacent to the conserved KSG

box that specifically mediates cefotaxime resistance,11,12,18 and another region affected in two of the mutants is marked by the changes G422D (C606) and R426C (C503). C606 contains a fourth mutation M289T at the N-terminus of the penicillin-binding domain. The second group of mutants is represented by C505 where none of its three PBP2x mutations occurs at positions close to those of the other mutant proteins. PBP2xC505 is also unusual in that no binding to penicillin derivatives could be demon-

Streptococcus pneumoniae PBP2x

1406

Fig. 2 (legend on next page)

Streptococcus pneumoniae PBP2x

strated by use of either radioactive compounds or nonradioactive benzylpenicillin up to a concentration of 100 μg/ml.24 We have now identified the selection step at which each of the PBP2x mutations occurred in the mutant families C001, C003 and C005 (Fig. 1a); those in the C006 family have been described.10 Mutations located at the C-terminal region of the penicillinbinding domain between Gly597 and Gly601 were selected as the first mutations in lineage C001, C003 and C006, respectively, indicating that this region plays an important role for primary levels of cefotaxime resistance. T550A (in C301 and C303) is selected as the second mutation. In the lineage C005, the most C-terminal mutation T526S was the first one selected (C205), and introduction of the L403F change in C405 correlates with the apparent loss of penicillin binding in PBP2x of this mutant. PBP2x of a clinical isolate S. pneumoniae 669 was isolated in Great Britain in 1987, and it is unlikely that resistance in this case is exclusively the result of cefotaxime selection. The PBP2x669 contains 22 alterations in its penicillinbinding domain compared to the PBP2xR6 in its unique mosaic gene structure (Fig. 1c). However, most amino acid changes occur also in penicillinsensitive S. mitis strains,25–27 leaving only five amino acid changes within the penicillin binding with potential relevance for resistance (Fig. 1c). These include Q552E, which has been shown to contribute to cephalosporin resistance,9,13,28 and S389L, suggested to contribute to a destabilizing effect of the active site.17 In addition, there are three unusual alterations not found in other clinical isolates including major resistant clones: G286P, F388L and A604K. Location of the PBP2x mutations in the three-dimensional structure The bases for this analysis were the structures of the native and acyl–enzyme forms of PBP2x,7,28 which are refined to resolutions of 2.4 and 2.8 Å, respectively. The two groups of laboratory mutant proteins show two completely different mutational patterns in the three-dimensional context, and PBP2x669 reveals a third type of mutational network (Fig. 2a–c). In the first group (PBP2x from the mutants C501, C503 and C606), at least one mutation on α11 is present in each PBP2x: G597D, L600W and G601V (Fig. 2a). They introduce bulkier side chains at the C-terminal region of helix α11, and G597D adds a

1407 negative charge close to the active site. All these residues face similar directions on the α-helix, which is positioned close to strand β3 and β4 and might affect the active site indirectly by changing the overall topology. T550A (in C501 and C503) is the only mutation close to the active-site nucleophile, S337 (6.6 Å distance between the two Cα atoms), that removes two H-bonds between protein and antibiotic.8 G422D and R426C (in C606 and C503, respectively) are positioned on the surface of the transpeptidase domain where they are in close contact with the C-terminal domain but distant from the active site: G422D within 17.5 Å and R426C within 24.5 Å (Fig. 2a). These two mutations, together with M289T on β1 (in C606), appear to have only minor effects with respect to the active site and might play a role only in the context of the other mutations. In the case of PBP2xC505, the T526S mutation is located in the loop α9–β3 close to the active site (Fig. 2b). The T526 side chain establishes an H-bond with β3 and a hydrophobic interaction with H394 (α4), and the mutation T526S introduces a higher degree of freedom to H394. At the end of α4, H394 is the last residue before the short loop α4–α5, where the residues S395 and N397 are in close contact with the antibiotic (H bond and ionic bond). A perturbation in the position of S395, after the T526S mutation, could disturb the S395–antibiotic H-bond. The L403F mutation lies on α5 and points into a hydrophobic niche that also stabilizes α4, α2 and the loop α2–β2a; these three helices lie in the vicinity of the catalytic cleft. This mutation modifies the arrangements of helices α4 and α5, a perturbation that must be transferred to the loop α4–α5 mentioned before. The introduction of this mutation in C405 coincides with the failure to detect PBP2x on fluorograms10 and hence in C505 (Fig. 3), indicating that it has a strong impact on antibiotic interaction. The effect of the third mutation Q458K, which is introduced as the third mutation in C505, cannot be deduced from the structure. In the mosaic PBP2x669, 12 of the 22 alterations occur on the surface of the penicillin-binding domain and are therefore unlikely to result in affinity changes (positions 286, 355, 358, 378, 382, 384, 401, 462, 483, 516, 565, 567 and 578) with the exception of 462, which points into a hydrophobic region in the vicinity of the central β-sheet. Most of these changes occur in sensitive PBP2x variants, and only G286P is unique to PBP2x669 (see Figs. 1c and 2c). The two alterations Q447M and S449A also found in sensitive PBP2x are located on the surface defining the active-site groove and might contribute

Fig. 2. Positions of PBP2x mutations in the three-dimensional structure (stereo views). Ribbon representation for the transpeptidase domain from Protein Data Bank 1QMF and ball and sticks for amino acid side chains and cefuroxime (magenta). Secondary structure elements, α-helices and β-strands are labeled.7 Amino acid positions for all mutations referred to in Fig. 1b are shown. (a) Amino acid modifications found in mutants C501, C503 and C606. (b) Amino acid modifications found in mutant C505 (20° rotation around the y-axis from the orientation of Fig. 1a). (c) Only specific modifications to PBP2x669 mutant are depicted (30° rotation around the y-axis from the orientation of Fig. 1a). Figures were generated with PyMOL (DeLano, W. L. (2002). The PyMOL Molecular Graphics System. http://www.%20pymol.org) and possible interactions analyzed with Ligplot.29

Streptococcus pneumoniae PBP2x

1408

In vitro characterization of soluble PBP2x derivatives The acylation efficiency of the PBP2x mutants was examined using purified PBP2x derivatives (PBP2x*) where the putative hydrophobic membrane-anchoring domain encoding amino acids 19– 48 was deleted. In addition to the PBP2x derivatives from the four mutants C501, C503, C505 and C606, the second-level mutant C206 containing only the two mutations G597D and G601V was included as well as the PBP2x from the clinical isolate for determination of in vitro activities. PBPs interact with beta-lactams according to the following scheme:30 K

k2

k3

E þ I → Ed I → EI* → E þ P

Fig. 3. PBP2x in beta-lactam-resistant strains and transformants of the R6 strain obtained with the respective pbp2x genes. Cell lysates were labeled with [3H]propionylampicillin (a) or Bocillin™FL (b), and PBPs were detected after SDS-PAGE and fluorography, respectively, with the FluorImager. The PBP profile of the recipient strain S. pneumoniae R6 is included for comparison. The mutants and transformants are indicated on top. The positions of the PBPs are marked on the left side.

to the overall structure of the active site. The other three changes, which can be identified in sensitive PBP2x (positions 572, 574 and 576), are located on β4, facing helix α11. A604K on α11 introduces a bulkier side chain and a positive charge; this mutation is specific to PBP2x669 (Fig. 2c). F388L, S389L and Q552E, also specific to PBP2x669, are located close to the active site. F388L is the equivalent mutation in α4 of L403F in α5. F388 and L403 are in the core of this hydrophobic niche and are key components in the arrangement between helices α4 and α5. S389L has been proposed to be involved in an ‘open’ conformation of the active site in a resistant PBP2x of distinct mosaic structure,17 and the negative charge introduced by Q552E located on strand β3 interferes with the binding of negatively charged beta-lactams such as cefotaxime.9

where E = active enzyme, I = beta-lactam compound, E·I = noncovalent complex, EI* = covalent acyl–enzyme complex and P = biologically inactive product. The k2/K parameter accounts for the acylation efficiency; k2 and k3 are first-order rate constants for the acylation and deacylation steps, respectively. The acylation efficiency indicated by the k2/K values of the purified proteins (PBP2x*) was determined by following beta-lactam quenching of the intrinsic fluorescence of the PBPs31 and was obtained for two beta-lactams: cefotaxime, the selective compound in the case of the laboratory mutants, and benzylpenicillin (Table 1). For comparison, published data on the single mutations T338A, T550A and Q552E that have been identified in mosaic PBP2x of resistant clinical isolates are included. In the three group I PBP2x*C501, PBP2x*C503 and PBP2x*C606, the acylation efficiency for cefotaxime decreased considerably to ≤ 1% of that of the wildtype PBP2x*, whereas that for benzylpenicillin was reduced only to 17–50% (Table 1). This documents effects of mutations other than T550A in the PBP2x*C501 and PBP2x*C503, since the T550A mutation alone had been reported to confer a 20-fold reduction of k2/K for cefotaxime and had no effect on benzylpenicillin.28 The two mutations G601V and G597D in the PBP2x*C206 hardly affected the interaction with benzylpenicillin and only the acylation efficiency for cefotaxime was reduced 20-fold. This shows that at least one of the two mutations in PBP2x*C606, M289T and G422D, that occurred during the last two steps of the six-step selection procedure further decreased considerably the affinity to benzylpenicillin as well as to cefotaxime. The PBP2x*C505 was distinct in that it hardly reacted at all with either beta-lactam in agreement with the difficulties in visualizing this protein via radioactive penicillins on fluorograms,24 and the k2/K values below 100 M− 1 s− 1 are of the lowest reported so far. Thus, all PBP2x mutants obtained via cefotaxime selection were highly affected with respect to their reactivity with this compound, and each set of mutations conferred a distinct reactivity profile.

Streptococcus pneumoniae PBP2x

1409

Table 1. Comparison of kinetic parameters for the interactions between PBP2x derivatives and β-lactams and thiolester substrate S2d PBP2x R6 C206G601V-G597D C503T526S-T550A-R426C C501L600W-T550A C606G601V-G597D-M289T-G422D C505T526S-L403F-Q458K 669 R6b T338Ab T550Ab Q552Eb a b

Cefotaxime, k2/K (M− 1 s− 1)

Benzylpenicillin, k2/K (M− 1 s− 1)

S2d, kcat/Km (M− 1 s− 1)

162,000 ± 4000a ≥8600 1700 ± 900 370 ± 16 47 ± 9 61 ± 25 11,500 ± 3200 209,000 ± 18,000 114,000 ± 2400 11,900 ± 900 59,400 ± 3100

58,000 ± 5000a 60,000 ± 7000 22,700 ± 300 10,000 ± 1800 27,000 ± 2000 52 ± 27 8050 ± 1200 99,000 ± 12,000 36,700 ± 3900 112,000 ± 10,000 30,600 ± 200

2600 ± 1000 1400 ± 58 143 ± 27 130 ± 56 b50 b50 92 ± 15 2500 ± 200 510 ± 20 1360 ± 140 1690 ± 270

Data obtained from independent experiments. From Mouz et al.28

In the PBP2x*669 the k2/K value was reduced 6-fold for benzylpenicillin and 20-fold for cefotaxime. The single point mutation Q552E as described previously28 showed only approximately 30% activity with both compounds (Table 1), demonstrating that additional amino acid changes contribute to the reduced affinity in the PBP2x*669. The hydrolytic activity was determined with the thiol ester substrate analogue S2d (Table 1). There are no natural substrates for estimating enzymatic activity of PBP2x. In many cases, thiol ester compounds can be used in transpeptidation reactions, demonstrating that they are mimicking true substrates.32 Although PBPs may react differently with this compound than the natural peptide substrates as indicated in previous studies, the kinetic parameters help to define the effect of mutations on peptide hydrolysis of PBP variants. All PBP2x derivatives showed a reduced reactivity with S2d compared to the R6 enzyme. The hydrolytic activity of PBP2x*C206 was reduced only by a factor of approximately 2, similar to the single mutation T550A.28 A decrease in kcat/Km values was observed for PBP2x*C501, PBP2x*C503 and PBP2x*669, retaining approximately 5% of the activity of the wild-type protein. The reactivities of PBP2x*C606 and PBP2x*C505 were below detection limits. The relative effect of the mutations on the second-order rate constants characterizing the acylation reaction with S2d was similar compared to the effect on acylation efficiency for cefotaxime in case of the laboratory mutants but not in the PBP2x669. Transformation of PBP2x mutations into R6 PBP2x is not the only compound that contributes to the resistance in the mutants and the clinical isolate. In the laboratory mutants, resistance is also mediated by mutations in the histidine protein kinase CiaH33,34 in addition to those in pbp2a and unknown genes (R.H., unpublished results). PBP2b is never affected in the cefotaxime-resistant laboratory mutants since it does not interact with cefotaxime and therefore is not a target for this compound.35 Therefore, in order to evaluate in vivo effects of the mutant PBP2x alone, the cloned PBP2x

genes were used in transformation assays with the parental beta-lactam-sensitive R6 strain as recipient. Several cefotaxime-resistant transformants were isolated in each experiment. A range of cefotaxime concentrations was used for selection starting with 0.02 μg/ml cefotaxime. Up to 10 transformants were picked at 0.08 μg/ml or higher, and the presence of PBP2x mutations was confirmed by DNA sequencing of the entire pbp2x. The CiaH gene was also sequenced to ensure the presence of the wild-type allele, since we have experienced that ciaH mutations can be selected very easily with cefotaxime contributing to resistance as well. Transformants containing distinct sets of up to four PBP2x mutations were obtained (Table 2). All the mutations present in the donor pbp2x could be transformed in the case of the group I PBP2x, i.e., from the mutants C503 (three mutations), C501 (two mutations) and C606 (four mutations) (see also Fig. 1). In the transformants containing one or two mutations, only the C-terminal mutations on α11 (G597D, G601V) and/or the T550A mutation were present [R6-T501(i), R6-T503(i), R6-T606(i)]. With the group II PBP2xC505, only two mutations were obtained: T526S and Q458K. Therefore, PBP2x* of the mutant C405 was also used in a second set of transformation experiments, and in this case both mutations of the mutant T526S and L403F were obtained in transformants selected at 0.04 μg/ml cefotaxime. The transformant R6-T669 contained only part of the mosaic block with the region encoding amino acids 301–559, leaving only three positions that are distinct from related PBP2x of penicillin-susceptible streptococci: F388L, S389L and Q552E (see Fig. 1c). PBP pattern in cell lysates Cell lysates of the transformants were used to visualize a low-affinity PBP2x using either radioactive propionyl-ampicillin or the fluorescent Bocillin™FL, and examples are shown in Fig. 3a and b. Presence of the T550A mutation in C503 and C501 did not eliminate binding to the penicillin derivative, since this mutation specifically interferes with the interaction of third-generation cephalosporins such

Streptococcus pneumoniae PBP2x

1410 Table 2. Resistance profile of mutants and transformants Strains (mutation)a Group I Mutants C606G422D-M289T-G597D-G601V/ciaH/pbp2a C506M289T-G597D-G601V/ciaH/pbp2a C406G597D-G601V/ciaH/pbp2a C306G597D-G601V/ciaH C206G597D-G601V C106G601V Transformants R6-T606(ii)M289T-G422D-G597D-G601V R6-T606(i)G597D-G601V Mutants C503R426C-T550A-G597D/ciaH/pbp2a C403T550A-G597D/ciaH/pbp2a C303T550A-G597D/ciaH C203G597D/ciaH C103ciaH Transformants R6-T503(ii)R426C-T550A-G597D R6-T503(i)T550A-G597D Mutants C501T550A-L600W/unknown/ciaH/unknown C401T550A-L600W/unknown/ciaH C301T550A-L600W/unknown C201L600W/unknown C101L600W Transformants R6-T501(ii)T550A-L600W R6-T501(i)T550A Group II Mutants C505Q458K-L403F-T526S/ciaH/unknown C405L403F-T526S/ciaH/unknown C305T526S/ciaH/unknown C205T526S/unknown C105unknown Transformants R6-T405L403F-T526S R6-T505Q458K-T526S Group III Mutant 669pbp2x/pbp2b/pbp1a Transformants R6-T669 M31Q552E Parental strain R6 a b c

MIC (μg/ml)b cefotaxime

Increase in resistance mediated by PBP2xc

0.8–1.3 0.8–1.2 0.5–0.7 0.35–0.4 0.14–0.16 0.08–0.09

0–0.1 0.3–0.5 0.05–0.07 0.08–0.09

0.25 0.13–0.16

0.09–0.12 0.13–0.16

0.85–1.5 0.85–1.5 0.8–0.9 0.15–0.2 0.04

Not detectable 0.65–0.7 0.11–0.16

0.25 0.2–0.25 0.6–0.64 0.6–0.64 0.18–0.2 0.04–0.08 0.04–0.08

0.04–0.08

0.2 0.15–0.2

0.13–0.18

0.4–0.5 0.15–0.2 0.05–0.1 0.06–0.08 0.02

0.12–0.14

0.25–0.3 0.1 0.04–0.06

0.1–0.15 0.1

1.2 0.1–0.14 0.05–0.06 0.015–0.02

Point mutations of PBP2x are indicated. Mutations in the pbp2a (R.H., unpublished results) and ciaH 32,36 are not specified. MIC values were obtained after incubation at 30 °C for 48 h. The numbers represent the increase in MIC obtained upon introduction of one PBP2x mutation.

as cefotaxime (Fig. 3a). In all other cases, a lowaffinity PBP2x in the transformants could easily be detected. PBP2x in R6-TC606(ii) with all four mutations was labeled very poorly compared to R6-TC606 (i) with two mutations, confirming the low in vitro activity of the PBP2x*C606 derivative (see Table 1). PBP2x was also traced with affinity purified antiPBP2x antibodies. In most mutants, PBP2x could be labeled perfectly well similar to the parental R6 strain. Surprisingly, the amount of PBP2x was clearly reduced in mutants C505 and C606 (Fig. 4a) and in C405 but not in the other mutants or in the R6 transformants R6-TC606(ii) and R6-T405 (not shown). This effect was not due to a reduced reactivity of the antibodies, since no difference was

visible using the purified soluble PBP2x derivatives (Fig. 4b). The phenomenon thus coincides with the introduction of the mutation L403F in C405 and G422D in C606, demonstrating that mutations in PBP2x not only affect the affinity for penicillin but also the apparent stability of the protein itself. Cefotaxime susceptibility mediated by PBP2x mutations in vivo Cefotaxime susceptibilities were determined using a narrow spectrum of concentrations in order to detect minor differences mediated by the different PBP2x mutations (Table 2). Transformants containing the entire set of group I PBP2x (from

Streptococcus pneumoniae PBP2x

1411 compared to the mutant containing only the Q552E mutation, confirming that at least one of the mutations F388L/S389L present in the transformant is involved in resistance. R6pbp2x669 exhibited clear cross-resistance to benzylpenicillin (0.2 μg/ml) thereby differing from the transformants obtained with pbp2x from the laboratory mutants (≤ 0.05 μg/ ml), implicating a distinct structural modification of the active site. The cefotaxime MICs of the transformants did not exceed 0.25 μg/ml, which is more than 10-fold the MIC of the sensitive R6 strain, strongly suggesting that this value represents the critical limit that can be achieved by the combinations of PBP2x mutations tested here.

Discussion

Fig. 4. PBP2x in beta-lactam-resistant mutants. Proteins of cells lysates (a) or 50 ng purified glutathione S-transferase (GST)-PBP2x derivatives (b) were separated by SDS-PAGE and transferred onto polyvinylidene fluoride membrane. The Western blots were incubated with affinity purified anti-PBP2x antibodies. The mutants are indicated on top; M, marker. The black arrow indicates the position of PBP2x (a) and GST-PBP2x (b); the white arrow shows the position of GST (1 μg), which was included as control (b).

C501, C503 and C606) revealed MIC values of 0.2– 0.25 μg/ml for cefotaxime, well below that of the mutants (0.5–1.5 μg/ml; see Table 2), confirming the impact of alterations in other genes present in the parental mutants such as ciaH and pbp2a. All other transformants had lower MICs. Concerning group I PBP2x, only the T550A mutation was transformed as a single mutation [R6-T501(i)] and conferred a somewhat higher MIC value (0.15–0.2 μg/ml cefotaxime) compared to the two C-terminal mutations G597D and G601V present in R6-T606(i) (0.13–0.16 μg/ml). Single mutations of PBP2x as present in the original first-step mutants C101 (L600W) and G601V (C106) mediated MIC values below 0.1 μg/ml (Table 2). The combination of T550A with G597D or L600W resulted in a slightly higher cefotaxime resistance compared to the level achieved with T550A alone, and was always higher than MICs achieved with other combinations of two mutations in PBP2x (≥ 0.2 versus ≤ 0.15 μg/ml). Transformants with the group II PBP2xC505 (R6T505 and R6-T405) with two mutations in each case showed MICs of 0.1–0.15 μg/ml, with the R6-T405 containing the L403F mutation being somewhat more resistant. The MIC of the R6pbp2x669 transformant with 0.1– 0.14 μg/ml for cefotaxime was considerably higher

The present study continues investigations of the structure/function relationship of one of the major resistance determinants in S. pneumoniae, PBP2x. The collection of cefotaxime-resistant laboratory mutants used here provides a unique tool for analyzing the mutational pathways during a defined selection procedure with one beta-lactam, cefotaxime. This compound has not been used clinically during the period when resistant clinical isolates evolved initially, and it is therefore not surprising that different sets of mutations were detected in the laboratory mutants as compared to those identified in clinical isolates. Only T550A, which confers cefotaxime resistance specifically and occurs in two independent laboratory mutant lineages, has been observed occasionally in clinical isolates with unusually high cefotaxime MICs, probably because third-generation cephalosporins have been introduced for therapy in areas where the prevalence of penicillin-resistant strains is high.12 PBP2x mutations in the three-dimensional structure The mutational sites of the PBP2x variants are scattered throughout the penicillin-binding domain, illustrating the importance of alterations in the protein within 20-Å radius from the active-site S337 and for the mutation R426C, even further. Three patterns of mutational sites became apparent, with group I consisting of the three mutant PBP2xC501, PBP2xC503 and PBP2xC606, group II being represented by PBP2x C505 , and group III by PBP2x669 of the clinical isolate. In all these groups, at least one mutation is located close to the activesite S337 (T550A and T526S) or introduces a negative charge into the active-site cavity (Q552E and G597D). T550A and Q552E have been analyzed in detail previously.9,28 T550 is located on strand β3 close to the K547SG motif and is in direct contact with cefuroxime according to the crystal structure of the PBP–cefuroxime complex. The negative charge introduced by Q552E located at the beginning of strand β3 may interfere with the binding of

1412 negatively charged beta-lactams such as cefotaxime.9 Mutations in the group I PBP2x (L600W and G601V), located at the beginning of α11, lead to movement of α11 and might have a steric effect on the active site indirectly, and both mutations were the first mutations selected in C101 and C106, respectively. Mutations in the same region that contribute to resistance have also been described in clinical isolates (T605N15 and Y595F14). Introduction of the mutation L403F in the group II PBP2xC505 practically abolishes the interaction with the beta-lactam. This shows the impact that mutations located close to the loop between α4 and α5 could have, such as T526A and F388L. It would be helpful to study the impact of the mutations in the three-dimensional context of crystallized mutant proteins. In vivo effects of PBP2x mutations Transformation experiments using PBP2x genes with up to four mutations confirmed that some of the mutations decrease cefotaxime susceptibility in the wild-type background, without the context of mutations in other genes that occur during the selection procedure. This was not surprising for T550A and Q552E, mutations that have been shown to be selected easily with cefotaxime in the laboratory.13,18,19 Considering the group I PBP2x, mutations located between G597 and G601 also decreased the cefotaxime susceptibility in the R6 background. The variability of mutations that are located at the N-terminal end of α11 (G597D, L600W and G601V) strongly suggests that this region is critical for the active-site conformation. Another mutant, C604, investigated previously also contains mutations on α11, S596L and G601E.10,37 Curiously, the unusual alteration in PBP2x669 from the clinical isolate, A604K, is located close to this region, and although it was not found in the transformant investigated here it is possible that it contributes to resistance in the original strain. None of the PBP2x mutations introduced late during selection in C503 (R426C) and C606 (G422D and M289T), or any of the C505 mutations, were found as single mutations in the transformants isolated here. It is possible that the concentrations used for isolation of the transformants were above the resistance potential mediated by some of these mutations. However, this could also be due to the fact that the genetic background is important for the mutations to be effective. In fact, introduction of the M289T mutation in the original mutant C506 results in an increase of 0.3–0.5 μg/ml cefotaxime, and each of the C505 mutations also leads to an increase in MIC (Table 2). The increase in MIC mediated by the G597D mutation, for example, is 0.05–0.07 μg/ml in C206 compared to C106 in combination with G601V, but 0.11–0.16 μg/ml in the mutant C203 in combination with the ciaH mutation (Table 2). Similarly, T550A alone results in an increase in MIC of 0.13– 0.18 μg/ml when introduced into R6, but of 0.65– 0.7 μg/ml in the mutant C303 (Table 2). On the other

Streptococcus pneumoniae PBP2x

hand, neither R426C nor G422D appears to have any detectable effect, and therefore they could be compensatory mutations. All the transformants with PBP2x had a generation time similar to that of the R6 strain (not shown). However, this does not necessarily mean that the actual function of the mutated PBP2x is not affected. When PBP2x mutants are studied in combination with deletion mutants in the regulatory system CiaRH, a slower growth rate and rapid lysis during stationary phase are observed, and the severity of the effect apparently depended on the type of the PBP2x mutation.38 Since no tests with true substrates are available for the transpeptidation reaction of PBP2x so far, the depsipeptide S2d has been introduced to determine the rate of hydrolysis for estimating the activity of PBPs.39 This thiol ester gives rise to linear acyl–enzymes, which usually are easily hydrolyzed or undergo aminolysis, in contrast to the complex formed with most beta-lactams, which retains a cyclic moiety. In all cases studied so far, PBP2x mutants that showed a reduced acylation efficiency also reacted with the depsipeptide S2d considerably poorly and the same result was obtained with all the PBP2x variants studied here, in agreement with an effect of the mutations on the in vivo activity of the proteins. One surprising finding was the apparent effect of the mutations L403F in C405 and G422D in C606, namely, a reduced amount of the protein in the cell. Purified PBP2x from all mutants reacted perfectly well with the antibodies, excluding the possibility that the mutations affect the reactivity of the antibodies. This is the case with mosaic PBP2x, and therefore these experiments cannot be performed with clinical isolates. In these two mutants, but also in C501 and C503 where PBP2x was present in wild-type quantities, additional mutations in ciaH are present, and it is unlikely that the cia regulon is affected differently in C405 and C606. It could be that individual mutations affect the stability of the protein, whereas it seems unlikely that mutations within the structural gene affect the expression level. Further experiments are required to clarify the molecular basis for this phenomenon. PBP2x in the clinical isolate Previous publications focused on mosaic PBP2x from clinical isolates and mutations therein.9,15,22,28,40 Due to the extensive changes in the mosaic blocks because these regions have evolved from homologues in sensitive streptococci that differ by about 10% in their amino acid sequence from PBP2x in sensitive S. pneumoniae, identification of amino acid changes, i.e., the actual mutations relevant for resistance, is a challenging task. In fact, previously identified individual mutations such as those at positions T338, T550 and Q552 hardly accounted for the low acylation efficiencies in the resistant mosaic proteins where they occur,22,28 and even extensive semiautomated searches for other relevant sites in one particular mosaic PBP2x could not reveal all

Streptococcus pneumoniae PBP2x

mutations leading to the low affinity of the protein.15 Another approach based on comparison of sequences from sensitive streptococci representing potential ancestor genes of the mosaic blocks 27 highlighted the same sites identified in other studies as relevant for resistance, i.e., positions L339, F364, T371, A369, S389, Y595F and N605,14–17 none of which are present in the PBP2x669. Using this strategy in the present study, only three amino acid changes remained as potential resistance determinants in the transformant obtained with PBP2x669: Q552E, and F388L/S389L. The latter two changes were probably too close to be separated easily during the recombination process during transformation. The S389L change is observed frequently in some classes of mosaic PBP2x from resistant isolates and has been suggested to be involved in an ‘open’ conformation of the active site in a resistant PBP2x of distinct mosaic structure.9 This change and possibly the adjacent F388L are likely to be responsible for the MIC of the transformant R6-T669 compared to that mediated by the Q552E mutation alone (Table 2). Nevertheless, it cannot be ruled out that other changes observed in sensitive homologues are also relevant for the phenotype of the 669 strain in the context of the mosaic framework even though they might have no effect per se. There are clear differences in the mutated sites that affect acylation efficiencies in the resistant PBP2x of laboratory mutants versus those found in clinical isolates. This is most likely due to differences in the selective beta-lactam antibiotics. Clearly, there are several ways of restructuring this essential protein into a resistance determinant given the variability observed in the laboratory mutants and the enormous variation in mosaic sequences identified in the clinical isolates. In addition to mutations that contribute to a reduced acylation efficiency of PBP2x, compensatory mutations that affect the actual function of the protein must also be taken into account. Moreover, the genetic background— mutations in PBP genes and non-PBP genes—is highly variable and contributes to the overall resistance of a strain, and it is probably impossible to predict the number of mutational pathways that result in a beta-lactam-resistant pneumococcus. In vitro activities of PBP2x variants The determination of the k2/K values of the purified PBP2x* mutant derivatives confirmed the cefotaxime specificity of the laboratory mutations, and values below 100 M− 1 s− 1 for cefotaxime have not been reported for any of the PBP2x derivatives from clinical isolates.16,22,28,40 For benzylpenicillin, only with the C505-PBP2x were similar low values obtained (Table 1), confirming the special remodeling of this mutant protein compared to all others. In summary, only three or four mutations in PBP2x suffice to practically abolish its interaction with beta-lactam antibiotics, making these protein variants interesting objects to target other inhibitory compounds.

1413

Materials and Methods Bacterial strains, plasmids and growth conditions S. pneumoniae R6 and cefotaxime-resistant mutant derivatives have been described.10 S. pneumoniae M31 containing the Q552E mutation in PBP2x has been described and was kindly provided by Laurent Gutmann.13 S. pneumoniae strains were grown in liquid culture in a casein-based semisynthetic medium supplemented with 0.2% yeast extract.41 MICs of beta-lactam antibiotics were tested by agar dilution on blood agar plates (3% sheep blood). A narrow range of antibiotic concentrations was used in order to also detect subtle differences in susceptibility between different transformants. E. coli DH5α was used as host strain for plasmid pCG31 and its derivatives. Plasmid pCG31 encodes PBP2x* from the parental strain S. pneumoniae R6; a deletion of amino acids 19–48, the membrane-anchoring region, results in a high-level overproduction of a soluble PBP2x derivative.31 Derivatives expressing different PBP2x mutants were constructed by replacing the parental 1793-bp PstI–EcoRI fragment with that of the respective mutant strain. Plasmids were purified following published procedures.42 The PBP2x genes from S. pneumoniae R6, C405, C503 and C606 were also cloned into the expression vector pGEX-6P1 tet43 after PCR amplification using the oligonucleotides 5′-CCGGAATTCGGGACAGGCACTCGCTTTGGAACAGATTTAGCGAAG-3′ (PM11) and 5′-CCGCTCGAGTTAGTCTCCTAAAGTTAATGTAATTTTTTTAATGTC-3′ (PM12) in conjunction with the high-fidelity enzyme iProof Polymerase (BioRad). The resulting 2127-bp product was digested with the restriction enzymes EcoRI and XhoI and ligated into the similarly digested vector pGEX-6P-1 tet and were transformed into the E. coli strain JM101.43 The vector pGEX-6P-1 tet is a derivative of the pGEX-6P-1 (Amersham Biosciences) in which ampR was replaced with tetR. The resulting plasmids were designated pPM20 (R6 pbp2x), pPM21 (C405), pPM22 (C503) and pPM23 (C606), respectively. Overexpression and purification of PBP2x derivatives All plasmids highly overexpressed a soluble PBP2x derivative in E. coli DH5α, which was purified from the cytoplasmic fraction by affinity chromatography on Procion blueherd mix 3967 coupled to Merck Fractogel TSK HW-65 (Merck, Darmstadt, Germany) as described.31 PBP2x-containing fractions were identified by SDS-PAGE and concentrated and desalted by centrifugation through Centricon-30 microconcentrators (Amicon). Between 3% and 7% of the total cellular protein was recovered in the PBP2x-containing fractions. Purified PBP2x was stored at − 20 °C in the presence of 10% glycerol in 10 mM Tris/HCl, pH 8; no loss of activity in terms of beta-lactam binding was noted during a several month period. Transformation of S. pneumoniae S. pneumoniae R6 and derivatives were transformed essentially according to published procedures, using 30 min incubation in the presence of DNA at 30 °C followed by a 2-h phenotypic expression period at 37 °C.44,45 Transformants were selected in blood agar

Streptococcus pneumoniae PBP2x

1414 plates at the antibiotic concentrations specified in the results. Detection of penicillin-binding proteins Cells of an exponentially growing culture were collected by centrifugation and resuspended in 10 mM sodium phosphate buffer, pH 7.2, 0.1% (w/v) Triton X-100. PBPs were labeled using 10 μl of a cell suspension corresponding to approximately 1 × 107 cells with [3H]propionylampicillin or Bocillin™FL for 30 min at 37 °C.24,36 Routinely, approximately 1 μCi radioactive beta-lactam was used per sample. The specific radioactivity of 3[H]propionylampicillin was not determined directly and was estimated to be close to that of the 3[H]succinimidylpropionate (90– 100 Ci/mmol, Amersham, England) used for its synthesis. Proteins were separated by PAGE with an acrylamide/ bisacrylamide ratio of 30:0.8, and PBP–beta-lactam complexes were visualized after fluorography for detection of the radioactive PBP–beta-lactam complexes. Alternatively, PBPs were labeled with BocillinTMFL (5 μM final concentration), and after separation of the proteins by PAGE, Bocillin–PBP complexes were visualized with a FluorImager at 488 nm (GE Healthcare). Western blotting Proteins of 5 μl cell lysate were separated on SDS-PAGE and transferred onto a polyvinylidene fluoride membrane. After incubation with affinity purified anti-PBP2x antibodies, PBP2x was detected with alkaline phosphataseconjugated goat anti-rabbit IgG (Sigma) after incubation with nitroblue tetrazolium chloride and 5-bromo-4-chloro3-indolyl phosphate (Roche). Rabbit antiserum obtained with soluble PBP2x*31 was affinity purified using PBP2x coupled to NHS-activated Sepharose (1 ml, Amersham Bioscience), washed with 20 ml sodium phosphate buffer (10 mM, pH 7.2) and PBP2x-specific antibodies eluted with glycine (100 mM, pH 2.2). Fractions (3 ml) were collected directly in 250 μl Tris buffer (1.5 M, pH 8.8) and used at a dilution of 1:10,000. Determination of kinetic parameters Kinetic parameters were determined at 37 °C in 1 ml 10 mM sodium phosphate, pH 7.0. Various concentrations between 2 and 50 μg protein (corresponding to 10 to 30 μl protein solution) were tested. PBP2x shows intrinsic fluorescence that is quenched upon acylation with betalactams.31 For determination of k2/K, decrease of fluorescence in the presence of antibiotic (at least fivefold molar excess) was determined using a Perkin-Elmer LS 50 spectrofluorimeter (excitation and emission wavelengths were 280 and 342 nm and the bandwidths 5 and 10 nm, respectively). For determination of the reactivity of the wild-type enzyme PBP2x*-R6, a stopped-flow apparatus was used as described.32 Hydrolytic activity was assayed with the depsipeptide S2d, a synthetic thiol ester analogue of a muropeptide.39 The reaction scheme shown above is valid for both betalactams and thiol esters such as S2d, the difference being that the k3 value is much higher for S2d. In this case, I = S, and k2/K = kcat/Km. Spectrophotometric measurements were performed with a Perkin-Elmer 554 UV–VIS spectrophotometer. Hydrolysis of the substrate corresponding to a decrease of the absorption upon addition of the PBP2x derivatives was monitored at 250 nm in 10 mM sodium

phosphate, pH 7.0, at 37 °C. Activities of PBP2x b 100 M− 1·s− 1 could not be measured accurately, since they are in the range of the spontaneous hydrolysis rate of this compound. The decrease in fluorescence reflects the pre-steady-state accumulation of EI*.32 Under the present experimental conditions, the acylation reaction was sufficiently rapid (maximum 2–3 min) to allow the deacylation (k3) step to be neglected. Since the apparent first-order rate constant remained proportional to the beta-lactam concentration with all compounds, only the k2/K ratio could be derived.30,32 Protein Data Bank accession codes The refined structures of PBP2x and the acyl–enzyme form are deposited under the accession codes 1QME and 1QMF.

Acknowledgements We thank Sonja Schröck from the Nano+Bio Center at the University of Kaiserslautern for assistance in DNA sequencing, and Andrea Dessen for help with the structural analysis. This work was supported in part by the Bundesministerium für Bildung und Forschung (grant KZ 01Kl9703/2), the EU INTAFAR LSHM-CT-2004-512138, the Schwerpunkt Biotechnologie der Universität Kaiserslautern, and the Stiftung Rheinland-Pfalz für Innovation.

References 1. Denapaite, D., Chi, F., Maurer, P., Nolte, O. & Hakenbeck, R. (2007). Mechanism of penicillin resistance in Streptococcus pneumoniae: targets, gene transfer, and mutations. In Molecular Biology of Streptococci (Hakenbeck, R. & Chhatwal, G. S., eds), pp. 290–303, Horizon Bioscience, Wymondham, Norfolk, UK. 2. Macheboeuf, P., Contreras-Martel, C., Job, V., Dideberg, O. & Dessen, A. (2006). Penicillin binding proteins: key players in bacterial cell cycle and drug resistance processes. FEMS Microbiol. Rev. 30, 673–691. 3. Goffin, C. & Ghuysen, J.-M. (1998). Multimodular penicillin-binding proteins: an enigmatic family of orthologs and paralogs. Microb. Mol. Biol. Rev. 62, 1079–1081. 4. Henriques-Normark, B. (2007). Molecular epidemiology and mechanisms for antibiotic resistance in Streptococcus pneumoniae. In Molecular Biology of Streptococci (Hakenbeck, R. & Chhatwal, G. S., eds), pp. 269–290, Horizon Press, Wymondham, Norfolk, UK. 5. Hakenbeck, R., König, A., Kern, I., van der Linden, M., Keck, W., Billot-Klein, D. et al. (1998). Acquisition of five high-Mr penicillin-binding protein variants during transfer of high-level β-lactam resistance from Streptococcus mitis to Streptococcus pneumoniae. J. Bacteriol. 180, 1831–1840. 6. Vaz Pato, M. V., Belo de Carvalho, C. & Tomasz, A.The Multicenter Study Group. (1995). Antibiotic susceptibility of Streptococcus pneumoniae isolates in Portugal. A multicenter study between 1989 and 1993. Microb. Drug Resist. 1, 59–70.

Streptococcus pneumoniae PBP2x 7. Pares, S., Mouz, N., Pétillot, Y., Hakenbeck, R. & Dideberg, O. (1996). X-ray structure of Streptococcus pneumoniae PBP2x, a primary penicillin target enzyme. Nat. Struct. Biol. 3, 284–289. 8. Gordon, E., Mouz, N., Duee, E. & Dideberg, O. (2000). The crystal structure of the penicillin-binding protein 2x from Streptococcus pneumoniae and its acyl–enzyme form: implication in drug resistance. J. Mol. Biol. 299, 477–485. 9. Pernot, L., Chesnel, L., Le Gouellec, A., Croizé, J., Vernet, T., Dideberg, O. & Dessen, A. (2004). A PBP2x from a clinical isolate of Streptococcus pneumoniae exhibits an alternative mechanism for reduction of susceptibility to beta-lactam antibiotics. J. Biol. Chem. 279, 16463–16470. 10. Laible, G. & Hakenbeck, R. (1991). Five independent combinations of mutations can result in low-affinity penicillin-binding protein 2x of Streptococcus pneumoniae. J. Bacteriol. 173, 6986–6990. 11. Krauß, J., van der Linden, M., Grebe, T. & Hakenbeck, R. (1996). Penicillin-binding proteins 2x and 2b as primary PBP-targets in Streptococcus pneumoniae. Microb. Drug Resist. 2, 183–186. 12. Coffey, T. J., Daniels, M., McDougal, L. K., Dowson, C. G., Tenover, F. C. & Spratt, B. G. (1995). Genetic analysis of clinical isolates of Streptococcus pneumoniae with high-level resistance to expanded-spectrum cephalosporins. Antimicrob. Agents Chemother. 39, 1306–1313. 13. Sifaoui, F., Kitzis, M.-D. & Gutmann, L. (1996). In vitro selection of one-step mutants of Streptococcus pneumoniae resistant to different oral β-lactam antibiotics is associated with alterations of PBP2x. Antimicrob. Agents Chemother. 40, 152–156. 14. Smith, A. M. & Klugman, K. P. (2005). Amino acid mutations essential to production of an altered PBP 2X conferring high-level beta-lactam resistance in a clinical isolate of Streptococcus pneumoniae. Antimicrob. Agents Chemother. 49, 4622–4627. 15. Carapito, R., Chesnel, L., Vernet, T. & Zapun, A. (2006). Pneumococcal beta-lactam resistance due to a conformational change in penicillin-binding protein 2x. J. Biol. Chem. 281, 1771–1777. 16. Chesnel, L., Pernot, L., Lemaire, D., Champelovier, D., Croize, J., Dideberg, O. et al. (2003). The structural modifications induced by the M339F substitution in PBP2x from Streptococcus pneumoniae further decreases the susceptibility to beta-lactams of resistant strains. J. Biol. Chem. 278, 44448–44456. 17. Dessen, A., Mouz, N., Gordon, E., Hopkins, J. & Dideberg, O. (2001). Crystal structure of PBP2x from a highly penicillin-resistant Streptococcus pneumoniae clinical isolate: a mosaic framework containing 83 mutations. J. Biol. Chem. 276, 45105–45112. 18. Grebe, T. & Hakenbeck, R. (1996). Penicillin-binding proteins 2b and 2x of Streptococcus pneumoniae are primary resistance determinants for different classes of β-lactam antibiotics. Antimicrob. Agents Chemother. 40, 829–834. 19. Negri, M. C., Morosini, M. I., Baquero, M. R., Campo, Rd. R., Blázquez, J. & Baquero, F. (2002). Very low cefotaxime concentrations select for hypermutable Streptococcus pneumoniae populations. Antimicrob. Agents Chemother. 46, 528–530. 20. Sanbongi, Y., Ida, T., Ishikawa, M., Osaki, Y., Kataoka, H., Suzuki, T. et al. (2004). Complete sequences of six penicillin-binding protein genes from 40 Streptococcus pneumoniae clinical isolates collected in Japan. Antimicrob. Agents Chemother. 48, 2244–2250. 21. Asahi, Y., Takeuchi, Y. & Ubukata, K. (1999). Diversity

1415

22.

23.

24. 25.

26. 27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

of substitutions within or adjacent to conserved amino acid motifs of penicillin-binding protein 2x in cephalosporin-resistant Streptococcus pneumoniae isolates. Antimicrob. Agents Chemother. 43, 1252–1255. Mouz, N., Gordon, E., Di Guilmi, D.-M., Petit, I., Petillot, Y., Dupont, Y. et al. (1998). Identification of a structural determinant for resistance to β-lactam antibiotics in Gram-positive bacteria. Proc. Natl Acad. Sci. USA, 95, 13403–13406. Hakenbeck, R., Kaminski, K., König, A., van der Linden, M., Paik, J., Reichmann, P. & Zähner, D. (1999). Penicillin-binding proteins in β-lactam-resistant Streptococcus pneumoniae. Microb. Drug Resist. 5, 91–99. Laible, G. & Hakenbeck, R. (1987). Penicillin-binding proteins in β-lactam-resistant laboratory mutants of Streptococcus pneumoniae. Mol. Microbiol. 1, 355–363. Sibold, C., Henrichsen, J., König, A., Martin, C., Chalkley, L. & Hakenbeck, R. (1994). Mosaic pbpX genes of major clones of penicillin-resistant Streptococcus pneumoniae have evolved from pbpX genes of a penicillin-sensitive Streptococcus oralis. Mol. Microbiol. 12, 1013–1023. Bergmann, C. (2003). Epidemiologie eines Mosaikgens: pbp2x in β-Lactam-sensitiven und -resistanten oralen Streptokokken aus Spanien. Universität Kaiserslautern. Chi, F., Nolte, O., Bergmann, C., Ip, M. & Hakenbeck, R. (2007). Crossing the barrier: evolution and spread of a major class of mosaic pbp2x in S. pneumoniae, S. mitis and S. oralis. Int. J. Med. Microbiol. 297, 503–512. Mouz, N., Di Guilmi, A. M., Gordon, E., Hakenbeck, R., Dideberg, O. & Vernet, T. (1999). Mutations in the active site of penicillin-binding protein PBP2x from Streptococcus pneumoniae. Role in the specificity for β-lactam antibiotics. J. Biol. Chem. 274, 19175–19180. Wallace, A. C., Laskowski, R. A. & Thornton, J. M. (1995). LIGPLOT: a program to generate schematic diagrams of protein–ligand interactions. Protein Eng. 8, 127–134. Frère, J.-M. & Joris, B. (1985). Penicillin-sensitive enzymes in peptidoglycan biosynthesis. Crit. Rev. Microbiol. 11, 299–396. Laible, G., Keck, W., Mottl, H., Frère, J.-M., Jamin, M. & Hakenbeck, R. (1992). Penicillin-binding protein 2x of Streptococcus pneumoniae. Expression in Escherichia coli and purification of a soluble enzymatically active derivative. Eur. J. Biochem. 207, 943–949. Jamin, M., Damblon, C., Millier, S., Hakenbeck, R. & Frère, J.-M. (1993). Penicillin-binding protein 2x of Streptococcus pneumoniae: enzymic activities and interactions with β-lactams. Biochem. J. 292, 735–741. Guenzi, E., Gasc, A. M., Sicard, M. A. & Hakenbeck, R. (1994). A two-component signal-transducing system is involved in competence and penicillin susceptibility in laboratory mutants of Streptococcus pneumoniae. Mol. Microbiol. 12, 505–515. Hakenbeck, R., Grebe, T., Zähner, D. & Stock, J. B. (1999). β-Lactam resistance in Streptococcus pneumoniae: penicillin-binding proteins and non penicillinbinding proteins. Mol. Microbiol. 33, 673–678. Hakenbeck, R., Tornette, S. & Adkinson, N. F. (1987). Interaction of non-lytic β-lactams with penicillinbinding proteins in Streptococcus pneumoniae. J. Gen. Microbiol. 133, 755–760. Zhao, G., Meier, T. I., Kahl, S. D., Gee, K. R. & Blaszczak, L. C. (1999). Bocillin FL, a sensitive and commercially available reagent for detection of penicillin-binding proteins. Antimicrob. Agents Chemother. 43, 1124–1128. Jamin, M., Hakenbeck, R. & Frère, J.-M. (1992).

1416 Penicillin binding protein 2x as a major contributor to intrinsic β-lactam resistance of Streptococcus pneumoniae. FEBS Lett. 331, 101–104. 38. Mascher, T., Heintz, M., Zähner, D., Merai, M. & Hakenbeck, R. (2006). The CiaRH system of Streptococcus pneumoniae prevents lysis during stress induced by treatment with cell wall inhibitors and mutations in pbp2x involved in beta-lactam resistance. J. Bacteriol. 188, 1959–1968. 39. Adam, M., Damblon, C., Jamin, M., Zorzi, W., Dusart, V., Galleni, M. et al. (1991). Acyltransferase activities of the high-molecular-mass essential penicillin-binding proteins. Biochem. J. 279, 601–604. 40. Lu, W.-P., Kincaid, E., Sun, Y. & Bauer, M. D. (2001). Kinetics of beta-lactam interactions with penicillinsusceptible and -resistant penicillin-binding protein 2x proteins from Streptococcus pneumoniae. Involvement of acylation and deacylation in beta-lactam resistance. J. Biol. Chem. 276, 31494–31501.

Streptococcus pneumoniae PBP2x 41. Lacks, S. & Hotchkiss, R. D. (1960). A study of the genetic material determining an enzyme activity in pneumococcus. Biochim. Biophys. Acta, 39, 508–517. 42. Holmes, P. S. & Quigley, M. (1981). A rapid boiling method for the preparation of bacterial plasmids. Anal. Biochem. 114, 193–197. 43. Cacciatore, G., Petz, M., Rachid, S., Hakenbeck, R. & Bergwerff, A. A. (2004). Development of an optical biosensor assay for detection of β-lactam antibiotics in milk using the penicillin-binding protein 2x. Anal. Chim. Acta, 520, 105–115. 44. Tiraby, J.-G. & Fox, M. S. (1974). Marker discrimination and mutagen-induced alterations in pneumococcal transformation. Genetics, 77, 449–458. 45. Laible, G., Hakenbeck, R., Sicard, M. A., Joris, B. & Ghuysen, J.-M. (1989). Sequences of the pbpX genes encoding the penicillin-binding protein 2x from Streptococcus pneumoniae R6 and a cefotaxime-resistant mutant, C506. Mol. Microbiol. 3, 1337–1348.