HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 279, Issue 40, 41546-41556, October 1, 2004

This Article Abstract Full Text (PDF) All Versions of this Article: 279/40/41546    most recent M407149200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Arbeloa, A. Articles by Arthur, M. Search for Related Content PubMed PubMed Citation Articles by Arbeloa, A. Articles by Arthur, M. Social Bookmarking                      What's this?

Synthesis of Mosaic Peptidoglycan Cross-bridges by Hybrid Peptidoglycan Assembly Pathways in Gram-positive Bacteria^*

Ana Arbeloa, Jean-Emmanuel Hugonnet, Anne-Charlotte Sentilhes, Nathalie Josseaume, Lionnel Dubost¶, Christelle Monsempes, Didier Blanot||, Jean-Paul Brouard¶, and Michel Arthur**

From the INSERM E0004, Laboratoire de Recherche Moléculaire sur les Antibiotiques, 15 rue de l'Ecole de Médecine, 75270 Paris, cedex 06, France, ^¶Développement et Diversité Moléculaire, Muséum National d'Histoire Naturelle, USM0502-CNRS UMR8041, 75005 Paris, France, and ^||Enveloppes Bactériennes et Antibiotiques, UMR 8619 CNRS, Btiment 430, Université de Paris-Sud, 91405 Orsay, France

Received for publication, June 25, 2004 , and in revised form, July 26, 2004.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The peptidoglycan cross-bridges of Staphylococcus aureus, Enterococcus faecalis, and Enterococcus faecium consist of the sequences Gly₅, L-Ala₂, and D-Asx, respectively. Expression of the fmhB, femA, and femB genes of S. aureus in E. faecalis led to the production of peptidoglycan precursors substituted by mosaic side chains that were efficiently used by the penicillin-binding proteins for cross-bridge formation. The Fem transferases were specific for incorporation of glycyl residues at defined positions of the side chains in the absence of any additional S. aureus factors such as tRNAs used for amino acid activation. The PBPs of E. faecalis displayed a broad substrate specificity because mosaic side chains containing from 1 to 5 residues and Gly instead of L-Ala at the N-terminal position were used for peptidoglycan cross-linking. Low affinity PBP2a of S. aureus conferred -lactam resistance in E. faecalis and E. faecium, thereby indicating that there was no barrier to heterospecific expression of resistance caused by variations in the structure of peptidoglycan precursors. Thus, conservation of the structure of the peptidoglycan cross-bridges in members of the same species reflects the high specificity of the enzymes for side chain synthesis, although this is not essential for the activity of the PBPs.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The bacterial cell wall peptidoglycan is a net-like macromolecule that completely surrounds the cytoplasmic membrane and supplies the cell with mechanical protection against its own osmotic pressure (1). The stress-bearing peptidoglycan is polymerized from a subunit containing -1,4-linked GlcNAc and N-acetylmuramic acid (MurNAc)¹ substituted by a peptide stem (2). In the pathogenic Gram-positive bacteria belonging to the genera Enterococcus, Streptococcus, and Staphylococcus, the stem peptide consists of a conserved pentapeptide (L-Ala₁-D-iGln₂-L-Lys₃-D-Ala₄-D-Ala₅) and a variable side chain (see Fig. 1) (3).

View larger version (27K): [in this window] [in a new window]   FIG. 1. Schematic representation of peptidoglycan cross-linking in E. faecalis (A), E. faecium (B), and S. aureus (C). Peptidoglycan is polymerized from a subunit comprising three elements: (i) a disaccharide composed of -1,4-linked N-acetylglucosamine (GlcNAc) and MurNAc, (ii) a conserved pentapeptide stem (L-Ala1-D-iGln2-L-Lys3-D-Ala4-D-Ala5) linked to the D-lactoyl residue of MurNAc, and (iii) a variable side chain linked to the -amino group of the third residue (L-Lys3) of the pentapeptide stem. The structure of the side chain is generally highly conserved in members of the same species but variable in different species (3). Incorporation of glycine and of L-amino acids into the side chain is performed by a unique family of nonribosomal peptide bond-forming enzymes that use aminoacyl-tRNA as the substrate (Fem and Bpp transferases in S. aureus and E. faecalis, respectively) (25). In E. faecium, an unidentified ATP-dependent ligase catalyzes peptide bond formation between the -carboxyl group of D-Asp and the -amino group of L-Lys3 (33). The -carboxyl group of D-Asp is secondarily partially amidated (D-Asx, D-Asn, or D-Asp). The two disaccharide peptide substrates participating in the D,D-transpeptidation reaction are referred to as carbonyl acceptor and donor. The leaving groups are circled. The amino acids of the cross-bridge are boxed. The orientation of the peptide bonds is indicated by arrows.

Diversification of the side chain structure during speciation (see Fig. 1) is potentially associated with diversification of the substrate specificity of the D,D-transpeptidases. In S. aureus, the FmhB transferase for incorporation of the first residue of the pentaglycine side chain (see Fig. 1) is an essential enzyme. This indicates that unsubstituted pentapeptide stems cannot be cross-linked (11). In addition, femA and femB mutants of methicillin-resistant S. aureus are susceptible to methicillin. This suggests that a complete pentaglycine side chain is essential for the D,D-transpeptidase activity of PBP2a (12). Similarly, the side chain is essential for penicillin resistance in Streptococcus pneumoniae (13). Inactivation of bppA1 encoding the transferase for incorporation of the first residue of the L-Ala-L-Ala side chain in E. faecalis has not been obtained (14). Deletion of bppA2 led to production of precursors substituted by a single L-Ala and to impaired expression of intrinsic -lactam resistance (14). For these reasons, transferases of the Fem family are considered to be potential targets for the development of novel antibiotics active against -lactam-resistant Gram-positive cocci (15).

In this study, we have further investigated the synthesis of the side chains of peptidoglycan precursors by transferases of the Fem family and their use by the PBPs in the cross-linking reaction. The study was designed to test the hypothesis that a narrow specificity of the PBPs could account for the essential role of transferases of the Fem family in peptidoglycan synthesis and -lactam resistance. Heterospecific expression of genes encoding transferases and PBPs was used to manipulate the structure of the side chain of the peptidoglycan precursors and the PBPs responsible for their polymerization. We first determined whether heterospecific expression of the S. aureus fmhB, femA, and femB genes in E. faecalis leads to synthesis of mosaic side chains. Mass spectrometry and tandem mass spectrometry (MS/MS) analyses of uncross-linked muropeptide monomers revealed that the Fem transferases of S. aureus were functional in E. faecalis and retained their substrate specificity as displayed in the original host. The participation of the mosaic precursors in the cross-linking reaction was then deduced from the structure of cross-linked muropeptide dimers. Mosaic side chains were detected both at the donor and acceptor positions of the dimers, indicating that the transpeptidases of E. faecalis had a broad substrate specificity. In a second set of experiments, the mecA, pbp5_fs, and pbp5_fm were heterologously expressed in E. faecalis and E. faecium to evaluate the capacity of the corresponding low affinity PBPs to confer -lactam resistance. PBP2a and PBP5_fm were found to be functional in the heterologous hosts producing precursors with L-Ala-L-Ala and D-Asx side chains, reflecting again a broad substrate specificity. Thus, analysis of the different hybrid peptidoglycan assembly pathways revealed that the high specificity of the enzymes for side chain synthesis accounts for conservation of the structure of the peptidoglycan cross-bridges in staphylococci and enterococci, although the conservation of substrate structure is not essential for the transpeptidase activity of the PBPs.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Bacterial Strains and Growth Conditions—Bacterial strains were grown in brain heart infusion (BHI) broth or agar (Becton Dickinson, le Pont de Claix, France) at 37 °C. Population analysis profiles were performed as previously described (16). Briefly, the bacteria were grown overnight at 37 °C in BHI broth, and serial 10-fold dilutions were plated on BHI agar containing increasing concentrations of ceftriaxone (Roche Applied Science). Colony forming units were determined after 48 h of incubation at 37 °C. To study the stability of ceftriaxone resistance phenotypes, the bacteria were sequentially subcultured for 5 days in 10 ml of BHI broth by using 0.1 ml of the preceding overnight culture as the innoculum.

Derivatives of expression vector pNJ2 harboring fem and pbp genes (see below) were introduced into E. faecalis JH2Sm::Tn916 (17) by electroporation and transferred by conjugation to E. faecalis JH2–2 (18), E. faecalis JH2–2bppA2 (14), E. faecalis JH2–2pbp5 (17), E. faecium D344S (19), and E. faecium BM4107 (20), as previously described (17). Spectinomycin (60 µg/ml) was used in all experiments to counter select loss of the plasmids.

Plasmid Construction—The open reading frame and ribosome-binding site of the fem and mecA genes of S. aureus and of the pbp5_fm gene of E. faecium were amplified and cloned under the control of the aph-A-3p promoter of the shuttle expression vector pNJ2 (17). Briefly, the mecA and fem genes of S. aureus Mu50 (21) were amplified with the following primers that contained SacI or XbaI restriction sites (underlined): mecA, 5'-TTGAGCTCATATAAGGAGGATATTGATG-3' and 5'-TTTCTAGACGGATTGCTTCACTGTTTTG-3'; fmhB, 5'-TTGAGCT-CAGGTATTGTTAAATAGAAGG-3' and 5'-TTTCTAGAGAGCGTTCA-GATTTCAGTCG-3'; femA,5'-TTGAGCTCATTAACGAGAGACAAATA-GG-3' and 5'-TTTCTAGACCTTCCTAAAAAATTCTGTC-3'; and femB, 5'-TTGAGCTCACAGAATTTTTTAGGAAGGG-3' and 5'-TTTCTAGAG-CCCTAACATCATTTACATC-3'. The pbp5_fm gene of E. faecium D344 (22) was amplified with 5'-TTGAGCTCTTCCTCAAAGACATATGT-3' and 5'-TTTCTAGATTATTGATAATTTTGGTTG-3'. The amplicons were cloned into pCR-Blunt (Invitrogen) and subcloned into pNJ2 by using SacI and XbaI. The nucleotide sequence of the inserts in the resulting recombinant plasmids was determined. The construction of the pNJ2 derivative for expression of the pbp5_fs gene of E. faecalis JH2–2 has been previously described (17).

The fmhB, femA, and femB genes were cloned together into pNJ2 to obtain a policistrionic operon under control of the aphA-3p promoter in two steps. A SacI restriction fragment containing fmhB was recovered from pCR-BluntfmhB and cloned into the SacI site located upstream from femB in pNJ2femB. The resulting plasmid, pNJ2fmhB femB, contained a unique SpeI restriction site located between fmhB and femB that was used to introduce the XbaI fragment of pCR-BluntfemA containing femA. The orientation of the three genes in the resulting plasmid, pNJ2fmhB femA femB, was determined by restriction analysis and partial nucleotide sequencing.

Preparation of Disaccharide Peptides—Bacteria were grown in 500 ml of BHI broth at 37 °C to an optical density at 650 nm of 0.7. Peptidoglycan was extracted by treating the bacterial pellet with 14 ml of 4% SDS at 100 °C for 30 min. Peptidoglycan was washed five times by centrifugation (12,000 x g for 10 min at 20 °C) with 20 ml of water. Peptidoglycan was serially treated overnight at 37 °C with Pronase (200 µg/ml) in 1 ml of Tris-HCl (10 mM, pH 7.4) and with trypsin (200 µg/ml) in 1 ml of phosphate buffer (20 mM, pH 7.8). Peptidoglycan was washed twice with 20 ml of water and digested overnight with mutanolysin (200 µg/ml; Sigma-Aldrich) and lysozyme (200 µg/ml; Sigma-Aldrich) at 37 °C in 1 ml of phosphate buffer (25 mM, pH 6.0) containing MgCl₂ (0.1 mM). Mutanolysin and lysozyme were inactivated for 3 min at 100 °C, and soluble disaccharide peptides were recovered by ultracentrifugation (100,000 x g for 30 min at 20 °C).

Reduction of Disaccharide Peptides—For reduction of MurNAc to N-acetylmuramitol, equal volumes (200 µl) of the solution of disaccharide peptides and of borate buffer (250 mM, pH 9.0) were mixed. Two mg of sodium borohydride were added, and the solution was incubated for 20 min at room temperature. The pH of the solution was adjusted to 4.0 with 20% orthophosphoric acid.

Preparation of Lactoyl Peptides—The ether link internal to MurNAc was cleaved under alkaline conditions (23) to produce lactoyl peptide peptidoglycan fragments. To the solution of unreduced disaccharide peptides (200 µl), 32% ammonium hydroxide (64 µl) was added, and the mixture was incubated for 5 h at 37 °C. It was neutralized with acetic acid (61 µl), lyophilized, and dissolved in 200 µl of water containing 0.05% trifluoroacetic acid.

Purification of Peptidoglycan Fragments—Reduced disaccharide peptides (200 µl) or lactoyl peptides (200 µl) were separated by reversed-phase high pressure liquid chromatography (HPLC) on a C₁₈ column (3 µm, 4.6 x 250 mm; Interchrom, Montluçon, France) at a flow rate of 0.5 ml/min. A 0 –20% gradient was applied between 10 and 90 min (solvent A: 0.05% trifluoroacetic acid in water; solvent B: 0.035% trifluoroacetic acid in acetonitrile). UV detection was performed at 210 nm. The relative abundance of peptidoglycan fragments was estimated as the percentage of the total integrated area. The peaks were individually collected, lyophilized, and dissolved in 100 µl of water.

Determination of the Mass of Peptidoglycan Fragments—The mass spectral data were collected with an electrospray time-of-flight mass spectrometer operating in the positive mode (Qstar Pulsar I, Applied Biosystems, Courtabuf, France). Purified fractions (3 µl) of reduced disaccharide peptides or lactoyl peptides were directly injected into the mass spectrometer using HPLC pumps at a flow rate of 0.2 ml/min (acetonitrile 50%, water 49.5%, formic acid 0.5%, per volume). The data were acquired with a capillary voltage of 5,200 V and a declustering potential of 20 V. The mass scan range was from m/z 400 to 2,500, and the scan cycle was 1 s. Structure assignment of muropeptides based on mass was performed as previously described (14).

Determination of the Structure of Monomers—Fragmentation performed on reduced disaccharide peptides and lactoyl peptides monomers gave essentially the same information on the structure of the peptide moieties of the molecules. Typical MS/MS experiments were performed by injecting 3–20 µl of the purified fractions, depending on the abundance of the peptidoglycan fragments, at a flow rate of 0.2 ml/min, as described above. The declustering potential was set to 60 V, the ions were selected based on the m/z value ([M+H]¹⁺) in the high resolution mode, and fragmentation was performed with nitrogen as the collision gas. Collision energy was typically of 25–28 and 36–40 eV for reduced disaccharide peptide and lactoyl peptide monomers, respectively.

Raising the declustering potential to 120 V led to formation of fragments of the lactoyl peptides that were analyzed by MS/MS. Such an analysis was used to confirm the structure assignment of the fragments obtained by tandem mass spectrometry.

Determination of the Structure of Multimers—Tandem mass spectrometry performed on reduced disaccharide peptide dimers provided little information on the peptide moiety of the molecules because fragmentation occurred mainly at the -1,4-GlcNAc-MurNAc bonds. In contrast, the entire peptide sequence could be determined by analyzing lactoyl peptide peptidoglycan fragments. In a first set of experiments, fragmentation was performed on the doubly charged ions of the molecules ([M+2H]²⁺), which gave a higher current intensity than singly charged ions ([M+H]¹⁺). The declustering potential was set to 60 V, and the collision energy was typically of 26–35 eV. Fragments resulting from loss of D-Lac-L-Ala₁ moieties of the molecules (loss of 143 atomic mass units) and additional NH₃ (17 atomic mass units) or NH₃ + CONH₃ (17 + 45 atomic mass units) in various combinations gave ions of high current intensities. Fragments generated by cleavage within the cross-bridges gave ions of lower intensity. The fragmentation pattern was also complicated by the presence of singly and doubly charged forms of certain fragments.

Optimization of the fragmentation conditions for lactoyl peptide dimers was achieved by raising the declustering potential to 100 V, which increased the current intensity of the singly charged form of the molecules ([M+H]¹⁺). Fragmentation performed with a declustering potential of 100 V on singly charged ions with a collision energy of 57–65 eV exclusively produced singly charged fragments. The highest intensities were observed for cleavage within the cross-bridges. The data reported under "Results" were performed under these conditions by using 7–35 µl of the purified fractions.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Structure of the Main Monomer Resulting from Heterospecific Expression of femA in E. faecalis JH2–2bppA2—The femA gene, encoding the transferase required for incorporation of glycyl residues at the second and third positions of the pentaglycine side chain in S. aureus (Fig. 1C), was cloned into the expression vector pNJ2 to generate plasmid pNJ2femA and introduced into E. faecalis JH2–2bppA2. The latter host produces precursors substituted by a single L-alanyl residue following deletion of the bppA2 gene (Fig. 1A). The most abundant monomers of JH2–2bppA2/pNJ2femA (Fig. 2, Peak 6) had a monoisotopic mass of 744.6, which matched the calculated value for a D-lactoyl-pentapeptide stem substituted by a side chain consisting of one L-alanyl and two glycyl residues. The structure of this branched peptide was solved by MS/MS, based on the detection of specific ions generated by loss of residues from the N terminus of the side chain (L-Ala-Gly-Gly-Nter) and from the carboxyl (D-Ala₄-D-Ala₅-Cter) or hydroxyl (OH-D-Lac-L-Ala₁-D-iGln) extremities of the lactoyl-pentapeptide stem (Fig. 3). The interpretation of the fragmentation pattern (Fig. 3B) was confirmed by MS/MS performed on fragments of the molecule, as exemplified by fragmentation of the ion at m/z 474.3 (Fig. 4). The structure of the main monomer determined by these approaches (Figs. 3 and 4) indicates that FemA of S. aureus is functional in E. faecalis because it catalyzed the addition of two glycyl residues after the L-alanyl residue incorporated by BppA1.

View larger version (37K): [in this window] [in a new window]   FIG. 2. Structural analysis of peptidoglycan from JH2–2bppA2/pNJ2femA. A, purified peptidoglycan was digested with muramidases and treated with ammonium hydroxide producing D-lactoyl peptide fragments that were separated by reversed-phase HPLC. Absorbance was monitored at 210 nm (absorbance unit x 103). B, model D-lactoyl peptide showing the main variations (boxed) in the free C terminus of the acceptor peptide stem, the cross-bridge, and the free N-terminal side chain. C, peaks 1–16 were individually collected, lyophilized, and analyzed by mass spectrometry. The relative abundance (%) of material in the 16 peaks was calculated by integration of the absorbance at 210 nm. The structure was deduced from the observed monoisotopic mass and, for most lactoyl peptides (indicated by stars), directly determined by tandem mass spectrometry. The most abundant D-lactoyl peptides, based on the relative absorbance at 210 nm and current intensity, are indicated in bold type. Tri, tripeptide L-Ala1-D-iGln2-L-Lys3; Tetra, tetrapeptide L-Ala1-D-iGln2-L-Lys3-D-Ala4; Penta, pentapeptide, L-Ala1-D-iGln2-L-Lys3-D-Ala4-D-Ala5; D-iGlu, the -carboxyl group of the second residue was not amidated. Gly C-ter, Gly instead of D-Ala at the 5th (C-terminal position) of pentapeptide stems.

View larger version (30K): [in this window] [in a new window]   FIG. 3. Analysis of the main monomer from JH2–2bppA2/pNJ2femA by tandem mass spectrometry. A, fragmentation was performed on the ion at m/z 745.6 corresponding to the [M+H]1+ from the major monomer. B, structure of the major monomer and inferred fragmentation pattern. The boxed m/z values in A originate from cleavage at single peptide bonds as represented in B. Peaks at m/z 631.5 and 560.4 matched the predicted values for loss of two N-terminal glycyl residues and of an additional L-alanyl residue, respectively. Loss of one and two D-Ala from the C terminus of the pentapeptide stem gave ions at m/z 656.5 and 585.4. Further loss of NH3 gave peaks at m/z 639.4 and 568.4. Fragmentation of the D-Lac-L-Ala1 amide bond was not observed. The peak at m/z 602.4 matched the predicted value for loss of D-Lac-L-Ala1. Further loss of NH3 and additional CONH3 led to peaks at m/z 585.4 and 540.4. Cleavage of the same peptide bond also produced peaks at m/z 144.0 and 116.0 corresponding to the D-Lac-L-Ala1 moiety of the molecule and loss of CO, respectively. Fragmentation at the D-iGln2-L-Lys3 peptide bond produced ions at 272.1 and 474.3. Additional loss of NH3 from ion at 474.3 gave an ion at 457.3. Additional ions could be accounted for by combinations of the fragmentations described above. In particular, loss of D-Ala4-D-Ala5 and D-Lac-L-Ala1 gave an ion at m/z 442.3. Further loss of NH3 and CONH3 led to peaks at m/z 425.3 and 380.2, respectively. Similarly, loss of D-Ala5 and D-Lac-L-Ala1 gave ions at m/z 513.3, 496.3, and 451.3. Peaks at m/z 471.3 matched the predicted value for loss of D-Ala5 and the L-Ala-Gly-Gly side chain. Peaks at m/z 289.2 and 314.2 matched the expected mass of L-Lys3 substituted by D-Ala4-D-Ala5 and L-Ala-Gly-Gly, respectively. Further fragmentation of the ion at m/z 314.2 led to ion at m/z 269.1 following loss of CONH3. Finally, ions at m/z 257.1 and 84.0 could correspond to D-iGln2-L-Lys3 and the immonium derivative of L-Lys. Loss of NH3 and additional CONH3 from peak at m/z 257.1 gave ions at 240.1 and 195.1, respectively. Together, these fragmentations account for the 29 ions with the highest relative current intensity. Ions of lower intensity could be accounted for by additional combinations of fragmentations (unlabeled peaks).

View larger version (18K): [in this window] [in a new window]   FIG. 4. Tandem mass spectrometry analysis of ion at m/z 474.3 corresponding to a fragment of the main monomer from JH2–2bppA2/pNJ2femA. Tandem mass spectrometry (A) and inferred fragmentation pattern (B). The boxed m/z values in A originate from cleavage at single peptide bonds as represented in B. The peak at m/z 456.2 matched the predicted value for loss of H2O. The peaks at m/z 417.2, 360.2, and 289.2 matched the predicted value for loss of N-terminal Gly, Gly-Gly, and L-Ala-Gly-Gly from the side chain. Loss of the C-terminal D-Ala5 and additional CO gave ions at m/z 385.2 and 357.2, respectively. Loss of D-Ala4-D-Ala5 and additional CO or CONH3 gave ions at m/z 314.2, 286.2, and 269.2, respectively. Peak at m/z 328.2 matched the predicted value for loss of one Gly from the side chain and D-Ala5. The additional loss of one Gly resulted in peak at m/z at 271.2 and further loss of CO resulted in peak at 243.2. Peak at m/z 200.1 could correspond to the mass of L-Lys3 substituted by D-Ala4 or L-Ala, and further loss of CONH3 would lead to peak at m/z 155.1. The peak at m/z 186.1 matched the predicted value for the side chain L-Ala-Gly-Gly. Finally, peaks at m/z 129.1 and 84.1 could correspond to the L-Lys residue and its immonium ion, respectively.

A second polymorphism was generated by the presence of stem peptides lacking one (tetrapeptide) or both (tripeptide) C-terminal D-alanyl residues (Fig. 2). As previously described (14), these stem peptides could be generated by D,D-carboxypeptidases and L,D-carboxypeptidases that cleave the D-Ala₄-D-Ala₅ and L-Lys₃-D-Ala₄ peptide bonds, respectively.

A third polymorphism was generated by the presence of Gly instead of D-Ala at the fifth position of lactoyl-pentapeptide stems. Pentapeptide stems with a C-terminal glycyl residue were also detected in E. faecalis JH2–2 and JH2–2bppA2 and are therefore unrelated to the activity of FemA. The PBPs are likely to be responsible for exchange of the C-terminal D-Ala by Gly because cytoplasmic UDP-MurNAc-pentapeptide ending in a glycyl residue were not detected in the cytoplasm (data not shown).

A fourth polymorphism was generated by the lack of amidation of the -carboxyl group of the D-isoglutaminyl residue at the second position of the stem peptide. This polymorphism is unrelated to the activity of FemA because it was also detected in the control strains JH2–2 and JH2–2bppA2.

Structures of Multimers of E. faecalis JH2–2bppA2/pNJ2 femA—Expression of femA in JH2–2bppA2 led to production of novel dimers differing from those present in preparations from the JH2–2bppA2 host by an increment of 228 atomic mass units, as expected for the presence of a total of four additional glycyl residues (Fig. 2). Tandem mass spectrometry provided the sequence of the cross-bridge and of the free N-terminal side chain, which were both found to contain the sequence L-Ala-Gly-Gly (see Fig. 5 for the analysis of the main dimer in peak 11). As described for the monomers, MS/MS was also performed on fragments of the molecules to confirm the interpretation of the fragmentation patterns. The structure of the main dimer (Fig. 5B) indicates that the D,D-transpeptidases of E. faecalis cross-linked glycyl-containing precursors. Peptide stems substituted by L-Ala-Gly-Gly had participated in the transpeptidation reaction both as an acceptor and a donor.

View larger version (31K): [in this window] [in a new window]   FIG. 5. Analysis of the main dimer from JH2–2bppA2/pNJ2femA by tandem mass spectrometry. A, fragmentation was performed on the ion at m/z 1400.7 corresponding to the [M+H]1+ form of the major dimer. B, the structure of the major dimer was inferred from the fragmentation pattern. The boxed m/z values in A originate from cleavage at single peptide bonds as represented in B. Other peaks could originate from multiple cleavage. For example, the loss of additional D-Lac-L-Ala1 from ions at m/z 1240.6, 841.4, 656.3, and 585.3 gave ions at m/z 1097.5, 698.3, 513.2, and 442.2, respectively. The loss of additional NH3 and CONH3 from ion at m/z 513.2 gave an ion at m/z 451.2. The loss of additional NH3 from ion at m/z 1097.5 gave an ion at m/z 1080.5. Loss of additional D-Lac-L-Ala1-D-iGln2 from ions at m/z 1240.6 gave an ion at 969.5.

Expression of femA in E. faecalis JH2–2—Analysis of the monomers of E. faecalis JH2–2/pNJ2femA indicated the participation of BppA1, BppA2, and FemA in side chain synthesis in various combinations (Table I). The wild-type sequence (L-Ala-L-Ala) was generated by BppA1 and BppA2. The sequence L-Ala-Gly-Gly, also observed in JH2–2bppA2/pNJ2femA (see above), was generated by BppA1 and FemA. The sequence L-Ala-L-Ala-Gly involved the participation of BppA1, BppA2, and FemA. Thus, FemA and BppA2 competitively added Gly and L-Ala, respectively, at the second position of the side chain. The sequence L-Ala-L-Ala-Gly also indicated that FemA can add a single residue. The three types of sequences (L-Ala-Gly-Gly, L-Ala-L-Ala, and L-Ala-L-Ala-Gly) were present both in the cross-bridges and in the free side chains of multimers.

Expression of femB—The FemB transferase is required for incorporation of glycyl residues at the fourth and fifth positions of the pentaglycine side chain in S. aureus (Fig. 1). Expression of femB did not lead to incorporation of glycyl residues in the peptidoglycan of JH2–2 or JH2–2bppA2 (Table I). Thus, FemB was unable to elongate side chains comprising one or two L-alanyl residues.

Synthesis of Pentaglycine Side Chains by FmhB, FemA, and FemB—Co-expression of the fmhB, femA, and femB genes in JH2–2bppA2 led to production of muropeptides monomers substituted by five glycyl residues (Table I). Thus, synthesis of the complete pentaglycine side chain of S. aureus was obtained in E. faecalis. Side chains consisting of one L-alanyl and four glycyl residues were also present because of competitive incorporation of L-Ala and Gly by BppA1 and FmhB at the first position. Amino acid residues other than Gly were not observed at positions 2–5. Incomplete side chains were also present. Analysis of multimers indicated that precursors containing pentaglycine side chains were used as donors and acceptors in the cross-linking reaction. Expression of fmhB, femA, and femB in JH2–2 gave essentially the same results except that Gly and L-Ala were competitively incorporated also at the second position of the side chain (Table I).

Heterospecific Expression of Genes Encoding Low Affinity PBPs—Resistance to -lactams mediated by heterospecific expression of mecA, pbp5_fs, and pbp5_fm was used as a screen to evaluate the capacity of the corresponding PBPs to catalyze cross-linking of pentapeptide stems substituted by L-Ala-L-Ala in E. faecalis and by D-Asx in E. faecium. mecA, pbp5_fs, and pbp5_fm were cloned into the expression vector pNJ2 and introduced into E. faecalis JH2–2pbp5 and E. faecium D344S. The latter hosts are previously characterized mutants susceptible to -lactams because of deletion of their respective species specific pbp5 genes (17, 19). Ceftriaxone, a third generation cephalosporin, was used to monitor expression of -lactam resistance, because deletion of the pbp5 genes produced large decreases (>1000-fold) in the minimal inhibitory concentration of this antibiotic both in E. faecium and E. faecalis.

Resistance Phenotype Mediated by mecA in E. faecalis and E. faecium—Population analysis profiles indicated that the mecA gene of S. aureus can confer resistance to ceftriaxone both in E. faecalis JH2–2pbp5 (Fig. 6A) and E. faecium D344S (Fig. 6B). The resistance phenotypes were characterized in both hosts by the presence of subpopulations of bacteria highly resistant to the drug. Representatives of the resistant subpopulations obtained on agar containing ceftriaxone were homogeneously resistant to high levels of ceftriaxone. The resistance trait remained stable after five serial subcultures in the absence of the drug. Sequencing of the mecA open reading frame and upstream sequences comprising the relevant promoter and ribosome-binding site did not reveal any mutation. Moreover, introduction of pNJ2mecA isolated from representatives of the resistant subpopulations into E. faecalis JH2–2plp5 or E. faecium D344S resulted in expression of high level ceftriaxone resistance only in subpopulations of the bacteria, as initially observed for the wild-type plasmid. Thus, plasmid pNJ2mecA of homogeneously resistant variants did not harbor any mutation affecting mecA or cis-acting sequences required for its expression. Together, these observations indicate that expression of wild-type mecA of S. aureus can confer high level resistance to ceftriaxone in E. faecalis and E. faecium. Modification of an unknown host factor was required for full expression of resistance, as also observed for the introduction of mecA in methicillin susceptible S. aureus (24).

View larger version (18K): [in this window] [in a new window]   FIG. 6. Ceftriaxone resistance mediated by heterospecific expression of pbp genes in E. faecalis JH2–2pbp5 and E. faecium D344S. A so-called population analysis profiles (16) was performed by plating serial 10-fold dilutions on increasing concentrations of ceftriaxone. The number of colony forming unit (cfu) was determined after 48 h of incubation at 37 °C. A, derivatives of E. faecalis JH2–2pbp5 harboring pNJ2 (open circles), pNJ2pbp5fs (open triangles), pNJ2pbp5fm (X), and pNJ2mecA (closed circles). Closed squares, representative clone of the subpopulation of E. faecalis JH2–2pbp5/pNJ2mecA highly resistant to ceftriaxone. B, derivatives of E. faecium D344S harboring pNJ2 (open circles), pNJ2pbp5fm (open triangles), and pNJ2mecA (closed circles). Closed squares, representative clone of the subpopulation of E. faecium D344S/pNJ2mecA highly resistant to ceftriaxone. C, closed circles, E. faecium D344S/pNJ2pbp5fs; closed squares, representative clone of the subpopulation of E. faecium D344S/pNJ2pbp5fs highly resistant to ceftriaxone. The population analysis profiles were reproducible as shown by testing a minimum of five independent cultures for each strain.

View larger version (25K): [in this window] [in a new window]   FIG. 7. Structural analysis of peptidoglycan from derivatives of E. faecalis JH2–2pbp5 grown in the presence of ceftriaxone at 100 µg/ml. Purified peptidoglycan was digested with muramidases and treated with ammonium hydroxide producing D-lactoyl peptide fragments that were separated by reversed-phase HPLC. A, E. faecalis JH2–2pbp5/pNJ2pbp5fs.B, E. faecalis JH2–2pbp5/pNJ2mecA (representative of the subpopulation highly resistant to ceftriaxone). C, mass spectrometry was performed on individually collected peaks. The calculated mass (not shown) of the proposed structure for the D-lactoyl peptides differed from the observed monoisotopic mass by less than 0.2 atomic mass unit. The structure of the monomers and dimers was determined by tandem mass spectrometry. The relative abundance (%) of material in peaks 1–10 was calculated by integration of the absorbance at 210 nm. NA, not applicable.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The Fem proteins form a family of highly diverse tRNA-dependent aminoacyl-transferases with respect to the type, number, and position of the amino acids added to the side chain of peptidoglycan precursors (Fig. 1) (25). The catalytic activity of the transferases has been assessed only of purified FemX_Wv from Weissella viridescens (26, 27) and the Bpp transferases from E. faecalis (14, 28). In S. aureus, transferases essential for incorporation of residues at specific positions of the pentaglycine side chain (Fig. 1) have been identified based on gene inactivation (femA and femB) and conditional gene expression (fmhB) (11, 12). Individual inactivation of two additional fem-related genes had little impact on viability and amino acid composition of the peptidoglycan (29). We show here that FmhB competes with BppA1 for the incorporation of Gly at the first position of the side chain in E. faecalis JH2–2 and JH2–2bppA2. FemA was sufficient for addition of two residues because expression of femA in JH2–2bppA2 led to the incorporation of two Gly in the absence of the other four fem genes present in the chromosome of S. aureus (Figs. 2 and 3). The two reactions can occur independently because FemA mediated incorporation of one or two Gly to generate the sequence L-Ala-L-Ala-Gly and L-Ala-Gly-Gly in E. faecalis JH2–2 because of competition between FemA and BppA2 at the second position (Table I). FemB added two Gly and required a side chain composed of three residues, indicating that this transferase was highly specific for the fourth and fifth positions. Co-expression of fmhB, femA, and femB led to the production of stem peptides substituted by a pentaglycyl moiety, showing for the first time that synthesis of the complete side chain of S. aureus can be reproduced in a heterologous host.

The Fem transferases of S. aureus incorporated glycyl residues in E. faecalis, indicating that the enzymes retained the specificity displayed in the original host (Fig. 2 and Table I). This implies that E. faecalis produces glycyl-tRNAs, which were efficiently used by the Fem transferases. Fractionation of the tRNAs isoacceptor of glycine and sequence analysis have raised the possibility that staphylococci produce unusual tRNAs dedicated to peptidoglycan synthesis (30, 31). Our analysis indicates that these putative tRNAs are unlikely to play a key role in the specificity of the Fem transferases for the amino acids incorporated into the side chains of peptidoglycan precursors.

A narrow substrate specificity of the D,D-transpeptidases for the acceptor side chain was proposed to account for the essential role of the transferases in viability (e.g. FmhB) and -lactam resistance (e.g. FemA and FemB) (11, 12). However, this has not been directly established by kinetic analyses of purified PBPs, because the enzymes were not active in vitro, except in very special cases involving highly reactive substrates (e.g. thioester) or atypical enzymes (e.g. the soluble R61 D,D-peptidase from Streptomyces spp.) (see Ref. 32 for a recent discussion). For this reason, heterospecific expression of pbp and fem genes was used in the current study to gain insight into the specificity of the PBPs in vivo. The D,D-transpeptidases of E. faecalis tolerated variations in the number of residues in the side chains (from 1 to 5) and substitution of L-Ala by Gly at the N-terminal position both in the donor and the acceptor (Fig. 2 and Table I). The substrate specificity of the D,D-transpeptidases was also explored based on heterospecific expression of pbp genes encoding low affinity PBPs. The mecA gene of S. aureus conferred resistance to ceftriaxone in E. faecalis JH2–2pbp5 and E. faecium D344S (Fig. 6). In neither host was the expression of resistance associated with a modification of peptidoglycan structure (Fig. 7 and data not shown). Because the D,D-transpeptidases are the essential target of -lactams (5, 6), PBP2a acted as a surrogate of the host D,D-transpeptidases and therefore catalyzed peptidoglycan cross-linking. This implies a low substrate specificity of PBP2a, because the amino group of the acceptor participating in the transpeptidation reaction was located on side chains consisting of five Gly, L-Ala-L-Ala, and D-Asx in S. aureus, E. faecalis, and E. faecium, respectively. These observations establish for the first time that mecA of S. aureus can confer -lactam resistance in distantly related hosts belonging to the genus Enterococcus, despite substantial diversity in the structure of peptidoglycan precursors. Similarly, PBP5_fs from E. faecalis conferred resistance to ceftriaxone in E. faecium D344S (Fig. 6C), indicating that substitution of L-Ala-L-Ala by D-Asx was also tolerated by this D,D-transpeptidase. Horizontal gene transfer in natural conditions has been documented for mecA between related species belonging to the genus Staphylococcus. Our data indicate that there is no barrier to heterospecific expression of -lactam resistance mediated by the low affinity PBP2a and PBP5_fs in more distantly related bacterial species, and thus, intergeneric transfers of low affinity PBPs should be anticipated.

Among the three low affinity PBPs that were tested in the current study (PBP2a, PBP5_fs, and PBP5_fm), only PBP5_fm from E. faecium did not confer ceftriaxone resistance in a heterologous host (Fig. 6). To explore the basis for the lack of expression of pbp5_fm-mediated resistance in E. faecalis, the bppA1 gene of E. faecalis was expressed in wild-type E. faecium BM4107 harboring pbp5_fm. The BppA1 transferase mediated incorporation of L-Ala instead of D-Asp in the E. faecium host, and the modified precursors were used as donors and acceptors in the cross-linking reaction (data not shown). Ceftriaxone did not inhibit the formation of the L-alanyl-containing cross-bridges. Thus, the lack of heterospecific expression of resistance to ceftriaxone, which inhibits all PBPs except PBP5_fm, did not appear to be due to the incapacity of E. faecium PBP5_fm to cross-link L-alanyl-containing peptidoglycan precursors. This observation implies, unexpectedly, that the requirements for expression of PBP5_fm-mediated -lactam resistance in terms of side chain structure vary in different hosts. A similar conclusion can be drawn for PBP2a from S. aureus because this PBP conferred -lactam resistance in hosts producing precursors substituted by L-Ala-L-Ala (E. faecalis) and D-Asx (E. faecium) (Fig. 6), whereas the complete pentaglycine side chain is essential for peptidoglycan cross-linking by PBP2a in its original host (12). The different requirements for PBP2a- and PBP5_fm-mediated -lactam resistance in different hosts may indicate that an as yet unknown component of the peptidoglycan polymerization complexes participates in the recognition of the acceptor side chain. It is clear that a narrow substrate specificity of low affinity PBPs cannot, in itself, account for the essential role of specific side chains in -lactam resistance.

In conclusion, we have established that diversification of the structure of the side chains of peptidoglycan precursors associated with speciation in Gram-positive bacteria (Fig. 1) did not correlate with diversification of the substrate specificity of the PBPs. The cross-bridges of wild-type S. aureus, E. faecalis, and E. faecium were found to exclusively contain Gly, L-Ala, and D-Asx, respectively. This reflects the high in vivo efficacy and specificity of the enzymes for side chain synthesis rather than a narrow specificity of the D,D-transpeptidases.

	FOOTNOTES

 
^* This work was supported by the programs Action Concertée Incitatrice Microbiologie 2003 from the Fonds National de la Science and Combating Bacterial Resistance to Antibiotics Contract LSHM-CT-2003-503335, 6th PCRD from the European Community. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Recipient of a fellowship from the Gobierno Vasco.

^** To whom correspondence should be addressed: LRMA-E0004, Université Paris VI, 15 rue de l'Ecole de Médecine, 75270 Paris cedex 06, France. Tel.: 33-1-43-25-00-33; Fax: 33-1-43-25-68-12; E-mail: michel.arthur{at}bhdc.jussieu.fr.

¹ The abbreviations used are: MurNAc, N-acetylmuramic acid; D-Lac, D-lactate or D-lactoyl; MS/MS, tandem mass spectrometry; HPLC, high pressure liquid chromatography; PBP, penicillin-binding protein; BHI, brain heart infusion.

	ACKNOWLEDGMENTS

 
We thank E. Collatz for critical reading of the manuscript.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 

1. Höltje, J. (1998) Microbiol. Mol. Biol. Rev. 62, 181–203[Abstract/Free Full Text]
2. van Heijenoort, J. (2001) Nat. Prod. Rep. 18, 503–519[CrossRef][Medline] [Order article via Infotrieve]
3. Schleifer, K. H., and Kandler, O. (1972) Bacteriol. Rev. 36, 407–477[Free Full Text]
4. van Heijenoort, J. (2001) Glycobiology 11, 25R-36R[Abstract/Free Full Text]
5. Jamin, M., Wilkin, J. M., and Frère, J.-M. (1995) Essays Biochem. 29, 1–24[Medline] [Order article via Infotrieve]
6. Goffin, C., and Ghuysen, J. M. (1998) Microbiol. Mol. Biol. Rev. 62, 1079–1093[Abstract/Free Full Text]
7. Wu, S. W., de Lencastre, H., and Tomasz, A. (2001) J. Bacteriol. 183, 2417–2424[Abstract/Free Full Text]
8. Signoretto, C., Boaretti, M., and Canepari, P. (1994) FEMS Microbiol. Lett. 123, 99–106[CrossRef][Medline] [Order article via Infotrieve]
9. Rice, L. B., Carias, L. L., Hutton-Thomas, R., Sifaoui, F., Gutmann, L., and Rudin, S. D. (2001) Antimicrob. Agents Chemother. 45, 1480–1486[Abstract/Free Full Text]
10. Sauvage, E., Kerff, F., Fonzé, E., Herman, R., Schoot, B., Marquette, J. P., Taburet, Y., Prevost, D., Dumas, J., Leonard, G., Stefanic, P., Coyette, J., and Charlier, P. (2002) Cell. Mol. Life Sci. 59, 1223–1232[CrossRef][Medline] [Order article via Infotrieve]
11. Rohrer, S., Ehlert, K., Tschierske, M., Labischinski, H., and Berger-Bächi, B. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 9351–9356[Abstract/Free Full Text]
12. Strandén, A. M., Ehlert, K., Labischinski, H., and Berger-Bächi, B. (1997) J. Bacteriol. 179, 9–16[Abstract/Free Full Text]
13. Filipe, S. R., Pinho, M. G., and Tomasz, A. (2000) J. Biol. Chem. 275, 27768–27774[Abstract/Free Full Text]
14. Bouhss, A., Josseaume, N., Severin, A., Tabei, K., Hugonnet, J. E., Shlaes, D., Mengin-Lecreulx, D., van Heijenoort, J., and Arthur, M. (2002) J. Biol. Chem. 277, 45935–45941[Abstract/Free Full Text]
15. Kopp, U., Roos, M., Wecke, J., and Labischinski, H. (1996) Microb. Drug Resist. 2, 29–41[Medline] [Order article via Infotrieve]
16. Tomasz, A., Nachman, S., and Leaf, H. (1991) Antimicrob. Agents Chemother. 35, 124–129[Abstract/Free Full Text]
17. Arbeloa, A., Segal, H., Hugonnet, J.-E., Josseaume, N., Dubost, L., Brouard, J.-P., Gutmann, L., Mengin-Lecreulx, D., and Arthur, M. (2004) J. Bacteriol. 186, 1221–1228[Abstract/Free Full Text]
18. Jacob, A. E., and Hobbs, S. J. (1974) J. Bacteriol. 117, 360–372[Abstract/Free Full Text]
19. Mainardi, J. L., Morel, V., Fourgeaud, M., Cremniter, J., Blanot, D., Legrand, R., Frehel, C., Arthur, M., van Heijenoort, J., and Gutmann, L. (2002) J. Biol. Chem. 277, 35801–35807[Abstract/Free Full Text]
20. Sifaoui, F., Arthur, M., Rice, L., and Gutmann, L. (2001) Antimicrob. Agents Chemother. 45, 2594–2597[Abstract/Free Full Text]
21. Hiramatsu, K., Hanaki, H., Ino, T., Yabuta, K., Oguri, T., and Tenover, F. (1997) J. Antimicrob. Chemother. 40, 135–146[Free Full Text]
22. Mainardi, J.-L., Legrand, R., Arthur, M., Schoot, B., van Heijenoort, J., and Gutmann, L. (2000) J. Biol. Chem. 275, 16490–16496[Abstract/Free Full Text]
23. Klaic, B. (1985) Carbohydr. Res. 138, 65–72[CrossRef]
24. Ryffel, C., Strässle, A., Kayser, F., and Berger-Bächi, B. (1994) Antimicrob. Agents Chemother. 38, 724–728[Abstract/Free Full Text]
25. Rohrer, S., and Berger-Bächi, B. (2003) Antimicrob. Agents Chemother. 47, 837–846[Free Full Text]
26. Hegde, S. S., and Shrader, T. E. (2001) J. Biol. Chem. 276, 6998–7003[Abstract/Free Full Text]
27. Hegde, S. S., and Blanchard, J. S. (2003) J. Biol. Chem. 278, 22861–22867[Abstract/Free Full Text]
28. Bouhss, A., Josseaume, N., Allanic, D., Crouvoisier, M., Gutmann, L., Mainardi, J.-L., Mengin-Lecreulx, D., van Heijenoort, J., and Arthur, M. (2001) J. Bacteriol. 183, 5122–5127[Abstract/Free Full Text]
29. Tschierske, M., Mori, C., Rohrer, S., Ehlert, K., Shaw, K. J., and Berger-Bächi, B. (1999) FEMS Microbiol. Lett. 171, 97–102[CrossRef][Medline] [Order article via Infotrieve]
30. Green, C. J., and Vold, B. S. (1993) J. Bacteriol. 175, 5091–5096[Abstract/Free Full Text]
31. Roberts, R. J., Lovinger, G. G., Tamura, T., and Strominger, J. L. (1974) J. Biol. Chem. 249, 4781–4786[Abstract/Free Full Text]
32. Anderson, J. W., Adediran, S. A., Charlier, P., Nguyen-Distèche, M., Frère, J. M., Nicholas, R. A., and Pratt, R. F. (2003) Biochem. J. 373, 949–955[CrossRef][Medline] [Order article via Infotrieve]
33. Staudenbauer, W., Willoughby, E., and Strominger, J. L. (1972) J. Biol. Chem. 247, 5289–5296[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg