A defect in cell wall recycling triggers autolysis during the stationary growth phase of Escherichia coli
1999; Springer Nature; Volume: 18; Issue: 15 Linguagem: Inglês
10.1093/emboj/18.15.4108
ISSN1460-2075
Autores Tópico(s)Peptidase Inhibition and Analysis
ResumoArticle2 August 1999free access A defect in cell wall recycling triggers autolysis during the stationary growth phase of Escherichia coli Markus F. Templin Markus F. Templin Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Astrid Ursinus Astrid Ursinus Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Joachim-Volker Höltje Corresponding Author Joachim-Volker Höltje Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Markus F. Templin Markus F. Templin Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Astrid Ursinus Astrid Ursinus Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Joachim-Volker Höltje Corresponding Author Joachim-Volker Höltje Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany Search for more papers by this author Author Information Markus F. Templin1, Astrid Ursinus1 and Joachim-Volker Höltje 1 1Abteilung Biochemie, Max-Planck-Institut für Entwicklungsbiologie, Spemannstrasse 35, 72076 Tübingen, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:4108-4117https://doi.org/10.1093/emboj/18.15.4108 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The first gene of a family of prokaryotic proteases with a specificity for L,D-configured peptide bonds has been identified in Escherichia coli. The gene named ldcA encodes a cytoplasmic L,D-carboxypeptidase, which releases the terminal D-alanine from L-alanyl-D-glutamyl-meso-diaminopimelyl-D-alanine containing turnover products of the cell wall polymer murein. This reaction turned out to be essential for survival, since disruption of the gene results in bacteriolysis during the stationary growth phase. Owing to a defect in muropeptide recycling the unusual murein precursor uridine 5′-pyrophosphoryl N-acetylmuramyl-tetrapeptide accumulates in the mutant. The dramatic decrease observed in overall cross-linkage of the murein is explained by the increased incorporation of tetrapeptide precursors. They can only function as acceptors and not as donors in the crucial cross-linking reaction. It is concluded that murein recycling is a promising target for novel antibacterial agents. Introduction The cell envelope of most bacteria is stabilized by an exoskeleton made of the cross-linked heteropolymer murein (peptidoglycan). Its basic building block, a peptidyl-disaccharide, is polymerized by the formation of glycosidic and peptide bonds. Thereby, a bag-shaped structure, the murein sacculus, is synthesized which completely encloses the cell (reviewed by Höltje, 1998). In the case of Escherichia coli the subunits of murein consist of the peptide D-alanyl-D-glutamyl-meso-diaminopimelyl-D-alanyl-D-alanine (L-Ala-D-Glu-m-A2pm-D-Ala-D-Ala) or truncated forms of this peptide, which are linked via an amide bond to the lactyl group of the muramic acid of the disaccharide N-acetylglucosamine-β-1.4-N-acetylmuramic acid (GlcNAc-MurNAc; van Heijenoort, 1994). Importantly, the peptide side chains consist of D- and L- (or meso-) amino acids giving rise to the presence of unusual L,D-, as well as D,D-, peptide bonds in murein. Cross-linkage of the murein strands is achieved by a head-to-tail cross-linking of the peptides via a peptide bond between the carboxyl group of the terminal amino acid of one peptide moiety and the non-alpha amino group present in a peptide side chain protruding from a neighboring glycan strand. In E.coli two types of cross-links can be formed (Glauner et al., 1988). The majority is found between the carboxyl group of a terminal D-Ala and the ω-amino group at the D-center of the meso-diaminopimelic acid resulting in a cross-linking D,D-peptide bond. To a lesser extent also the carboxyl group at the L-center of a terminal meso-diaminopimelic acid present in a tripeptide stem peptide is linked to the ω-amino group at the D-center of the meso-diaminopimelic acid of another peptide side chain giving rise to the formation of an L,D-cross-linking peptide bond. Not surprisingly, L,D- and D,D-peptide bonds are import ant targets for highly specific antibacterial agents since they are not found in eukaryotes. The most important group of antibiotics, the β-lactams, inhibit the enzymes specifically involved in the formation and cleavage of the D,D-peptide bonds (Waxman and Strominger, 1983). Since these enzymes bind penicillin covalently, they are referred to as penicillin-binding proteins (PBPs; Spratt, 1975). In addition to transpeptidases and endopeptidases, carboxypeptidases have also been described (Izaki et al., 1966). Being PBPs many of the D,D-enzymes from different sources have been characterized in some detail (Ghuysen, 1991). By way of contrast, knowledge about enzymes with a specificity for L,D-peptide bonds is limited. In particular, no gene encoding an L,D-transpeptidase, an L,D-endopeptidase or an L,D-carboxypeptidase has been cloned to date and the physiological functions of these proteins are still under discussion (Templin and Höltje, 1998). The best characterized example of an enzyme of this group is an L,D-carboxypeptidase isolated from E.coli (Metz et al., 1986a,b; Ursinus et al., 1992). The enzyme cleaves the D-alanine in position 4 that is linked via an L,D-peptide bond to the L-center of the meso-diaminopimelic acid in position 3 of the peptide side chains. Such an activity was shown earlier to oscillate during the cell cycle of E.coli with highest activity during cell division (Beck and Park, 1976, 1977). The function(s) of L,D-carboxypeptidases remain(s) obscure, although an involvement in cell division has been proposed (Begg et al., 1990; Höltje, 1998). Here we report on the cloning, overexpression and deletion of a cytoplasmic L,D-carboxypeptidase which turned out to be essential at the onset of the stationary phase of growth by specifically interfering with murein recycling. Results Identification of the gene encoding the L,D-carboxypeptidase A Classical protein purification methods yielded a highly enriched fraction of an L,D-carboxypeptidase, which could be used for biochemical characterization of the enzyme (Ursinus et al., 1992). Nevertheless, the amount of protein was not high enough to obtain a partial peptide sequence that could have been used for identification of the gene encoding this enzyme. Since a simple and fast assay for L,D-carboxypeptidase using uridine 5′-pyrophosphoryl N-acetylmuramyl-L-alanyl-D-glutamyl-meso-diaminopimelyl-D-alanine (UDP-MurNAc-tetrapeptide) as a substrate, which can be used even with crude cell extracts, was available (Ursinus et al., 1992), we decided to screen an expression library for the corresponding gene. Therefore, the Kohara miniset library (Noda et al., 1991) was scored for phage-based overexpression of L,D-carboxypeptidase activity. Screening the 479 phages from the miniset library led to the identification of one clone with a reproducible increase in L,D-carboxypeptidase activity. This phage (λ 245) carries a 17.4 kb insert containing 13 potential reading frames. Only one of the predicted proteins, f304 (Blattner et al., 1997) had a calculated molecular mass close to the experimentally determined value of the L,D-carboxypeptidase (Ursinus et al., 1992). Therefore, this coding region was amplified from phage DNA by PCR and directly cloned into an expression vector. A transformant of MC1061 carrying this construct (pLD2) was induced to overproduce the cloned gene product by adding isopropyl-thio-β-D-galactoside (IPTG). A 4500-fold increase in L,D-carboxypeptidase activity was detected and it could be concluded that the gene for an L,D-carboxypeptidase active on UDP-MurNAc-tetrapeptide had been identified. This high degree of overproduction had no effect on growth rate or cell viability. Since this is the first example of the cloning of a gene coding for an enzyme possessing L,D-carboxypeptidase activity the gene was named ldcA (L,D-carboxypeptidase A). Sequence analysis suggests a prokaryotic protease family The reading frame for the LdcA consists of 912 base pairs, coding for a protein with a predicted molecular mass of 33 572 Da. No apparent N-terminal signal sequence could be detected by using the TopPred II algorithm (Claros and von Heijne, 1994); therefore, a cytosolic location was expected. Interestingly, the ldcA gene and emtA, a recently identified gene coding for a lytic endo-transglycosylase (Kraft et al., 1998), are found side-by-side (Figure 1). Whereas the genes are transcribed in opposite directions, the promotors for both genes seem to overlap (M.F.Templin, unpublished). A coordinated expression might be significant, since some of the muropeptides released by EmtA are substrates for the L,D-carboxypeptidase. Figure 1.Map of the region from 1241 to 1259 kb of the E.coli W3110 chromosome. The inserts encoded by the Kohara phages λ244, λ245 and λ246 are indicated. The unassigned open reading frame f304 (black arrow) codes for the L,D-carboxypeptidase , emtA encodes the membrane-bound lytic endotransglycosylase A, an enzyme which releases muropeptides that are substrate for LdcA (Kraft et al., 1998). Download figure Download PowerPoint Similarity searches with the predicted coding sequence were performed using the PSI-BLAST algorithm (Altschul et al., 1997) and revealed the presence of homologs in different bacterial species including Synechocystis sp., Bacillus subtilis, Streptomyces coelicolor, Vibrio cholerae, Rickettsia prowazekii and Treponema pallidum (e values <10−63 in a PSI-BLAST search after the second iteration). No significantly similar proteins were detected in any eukaryote. When extending the searches to the available 'unfinished bacterial genomes' (last update 15 March 1999; preliminary sequence data were obtained from the Institute for Genomic Research website at http://www.tigr.o..) further homologs of LdcA were detected in Salmonella typhi, Neisseria gonorrhoeae, Neisseria meningitidis, Bordetella pertussis, Chlorobium tepidum, Pseudomonas aeruginosa, Streptococcus pyogenes, Streptococcus mutans and Deinococcus radiodurans (e values in the tblastn search 1%) were found in the membrane or in the periplasmic fraction (data not shown). The same result was obtained after Tris–EDTA treatment of the cells performed as described by Beck and Park, (1977). This shows that the LdcA is a cytoplasmic enzyme. Table 1. L,D-carboxypeptidase activity in different cell fractionsa Acid phosphataseb G6P-DHb L,D-carboxypeptidaseb A: Tris–EDTA–lysozyme treatment Cytoplasm 19.5% 91.5% 91.2% Periplasm 80.5% 8.5% 8.8% B: Tris–EDTA treatment Cytoplasm 55.3% 89.3% 93.1% Periplasm 44.7% 10.7% 6.9% a Cell fractionation was carried out following two different methods: A, according to Withold et al. (1979) and B, as described by Beck and Park, 1977. b Enzyme activities were determined using published methods for acid phosphatase (Atlung et al., 1989) as a marker for periplasmic proteins, for glucose-6-phosphate dehydrogenase (G6P-DH; Bergmeyer, 1974) as a marker for cytoplasmic enzymes and for LdcA as described in Materials and methods. Substrate specificity and inhibition of the LdcA With the construction of an N-terminally His6-tagged version of the enzyme it was possible to obtain sufficient amounts of highly enriched enzyme preparations (2.53×106 units/mg protein). This allowed us to determine the substrate specificity of the LdcA in some detail and to examine the sensitivity against different antibiotics. Cytoplasmic fractions containing 1.34 units of enzyme were incubated with a variety of murein derivatives carrying L-Ala-D-Glu-meso-A2pm-D-Ala tetrapeptide side chains. As listed in Table II, LdcA does not interact with isolated high molecular mass murein sacculi or with cross-linked muropeptides. Thus, only monomeric muropeptides, free tetrapeptide and UDP-activated murein precursors are substrates for LdcA. This finding is in accordance with results obtained by others (Leguina et al., 1994), who showed that an L,D-carboxypeptidase is present in E.coli that acts on low molecular mass muropeptides, but not on the native polymer murein. The specificity determined is consistent with the fact that LdcA is a cytoplasmic enzyme that cannot come into contact with the murein sacculus in the periplasm. Table 2. Substrate specificity of LdcA and inhibition of the enzyme by different antibiotics A: Substrate specificity Substrate Relative activity (%) UDP-MurNAc-tetrapeptidea 100 Disaccharide-tetrapeptideb 23.0 Anhydrodisaccharide-tetrapeptideb 25.7 MurNAc-tetrapeptideb 100 Bis-disaccharide-tetrapeptideb n.d.d Bis-anhydrodisaccharide-tetrapeptideb n.d.d Tetrapeptideb 132.8 Murein sacculic n.d.d B: Inhibition by antibiotics Antibiotic Concentration (μg/ml) Inhibition (%) Imipenem 100 10.1 250 13.9 Penicillin G 100 26.6 250 23.7 CGP31608 100 35.3 250 34.0 Cefminox (MT-141) 100 33.7 250 50.5 Cephalosporin C 100 53.1 250 63.1 Nocardicin A 100 81.7 250 86.9 a L,D-carboxypeptidase activity with UDP-MurNAc-tetrapeptide was determined at a concentration of 0.7 mM with 1.3 U of the enzyme under standard conditions as described in Materials and methods. The value for UDP-MurNAc-tetrapeptide was set to 100%. b L,D-carboxpeptidase activity with the muropeptides indicated was determined by separating substrate and product by RP-HPLC using the system to separate muropeptides described by Glauner (1988) except for terminating the elution when 50% of buffer B was reached, i.e. after 67.5 min. c The reaction with isolated murein sacculi was determined by analyzing the muropeptide composition of 50 μg sacculi treated with 27 U L,D-carboxypeptidase under standard conditions by RP-HPLC as described by Glauner (1988). d No activity detected. L,D-carboxypeptidase activity had been shown to be inhibited competitively by β-lactam antibiotics containing a D-configured amino acid as a side chain (Hammes, 1978; Hammes and Seidel, 1978a; Ursinus et al., 1992). Metz et al. (1986b) reported an inhibition of the activity by penem antibiotics. The enriched enzyme fraction was used to test the senitivity of LdcA: little or no inhibition could be shown for Penicillin G and the penems Imipemem and CGP31608 (kindly supplied by Ciba-Geigy, Basel, Switzerland); β-lactams containing a D-configured amino acid as a side chain including Cefminox, Cephalosporin C and Norcardicin A (kindly supplied by Eli-Lilly, Indianapolis, IN) inhibited enzyme activity (Table II). Construction and growth phenotype of a deletion mutant To get an idea of the function of the ldcA, a plasmid-coded deletion was constructed and transferred to the chromosome of E.coli MC1061 by λ-phage-mediated transduction (Kulakauskas et al., 1991). The obtained mutant named LD3 did not contain the coding sequence for the protein. No L,D-carboxypeptidase activity could be detected in soluble extracts when UDP-MurNAc-tetrapeptide was used as a substrate (data not shown), suggesting that the identified enzyme is the only activity of this type present in E.coli. Despite the fact that viable clones could be obtained, growth during the stationary phase was severely affected. After growth on Luria-Bertani (LB) agar for 72 h at 37°C, only a small percentage of the cells remained viable. Liquid cultures of wild-type bacteria and of the LdcA deletion mutant growing aerobically at 37°C had the same growth rate during the exponential phase (Figure 3). However, upon onset of the stationary phase, growth of the mutant culture seemed to slow down and the culture eventually started to lyse. Phase-contrast microscopy (Figure 4) showed cells that had lost their shape and, in addition, many lysed cells were visible. The number of viable cells determined after 12 h of growth was reduced to 3.5×107 in the LdcA mutant culture compared with 1.8×109 in the control culture. Therefore, spontaneous autolysis occurred when the culture entered the stationary growth phase. This phenotype could be suppressed by a plasmid carrying the ldcA gene. Figure 3.Growth of the ldcA deletion mutant LD3 and its isogenic parent strain MC1061. Growth of E.coli LD3 (○) and MC1061 (●) in 50 ml LB medium with aeration at 37°C was followed by optical density readings. Inoculation was carried out with one single colony from a fresh plate. Download figure Download PowerPoint Figure 4.Morphology of the ldcA deletion mutant LD3 and its isogenic parent strain MC1061. Phase-contrast micrographs were taken from cells grown aerobically at 37°C in LB medium at an OD578 of 1.5 (A, MC1061; B, LD3) and after growth overnight (C, LD3). The bar represents 5 μm. Download figure Download PowerPoint Changes in the murein composition of the mutant Lysis of the LdcA deletion mutant in the stationary growth phase could be the result of some defects in the structure of the murein sacculus. Therefore, the composition of the murein of an exponentially growing culture (OD578 0.5) and a culture just entering the stationary phase (OD578 1.5) was analyzed. Isolated sacculi were digested completely using a muramidase and the degradation products (muropeptides) were separated and quantified by reversed-phase high-pressure liquid chromatography (RP-HPLC) (Glauner, 1988). The results shown in Table III reveal significant differences in the relative amounts of the various muropeptides between the ldcA mutant and the parental strain. Whereas these differences were obvious in the murein from cells harvested in the exponential phase of growth, they were more pronounced for a culture grown to an OD578 of 1.5. In particular the overall cross-linkage was reduced by 23% leading to an increase in the monomeric muropeptide disaccharide–tetrapeptide by 16% and disaccharide–tripeptide by 46%. Whereas the number of dimeric D,D-cross-bridges was reduced by 12%, an even more dramatic decrease was observed for the trimeric and tetrameric structures containing D,D-cross-links: the relative amounts decreased by 67 and 58%, respectively. Interestingly, the unusual L,D-cross-linkage between two m-A2pm residues was increased. This finding is consistent with a number of reports showing a rise of L,D-cross-links when D,D-cross-links decrease for various reasons (Driehuis and Wouters, 1987; Kohlrausch and Höltje, 1991a). Therefore, formation of L,D-cross-bridges is considered to represent a salvage mechanism to keep the overall degree of cross-linkage at a high level even under conditions where normal D,D-cross-links, which depend on the presence of sufficient pentapeptides, can no longer be formed. It is important to note that the number of chain ends, which reflects the average length of the glycan strands, did not increase in the mutant indicating that the impending lysis event was not caused by autolytic processes described previously (Kitano and Tomasz, 1979; Kohlrausch and Höltje, 1991a). We conclude that autolysis of the ldcA mutant is the result of a dramatically reduced overall murein cross-linkage. Table 3. Muropeptide composition of the ldcA mutant and its isogenic parent straina Relative amounts (%) Exponential growth (OD578 0.5) Stationary growth (OD578 1.5) MC1061 LD3 Δ% MC1061 LD3 Δ% Muropeptidesb Monomeric 55.35 60.43 +9 51.47 60.89 +18 Tri 7.35 9.86 +34 7.43 10.83 +46 Tetra 44.39 46.54 +5 38.75 44.81 +16 Penta 0.08 0.03 −63 0.05 0.03 −40 Dimeric 39.06 36.22 −7 38.54 36.01 −7 TetraTetra 27.71 23.24 −16 22.52 19.05 −15 TetraTri 2.80 2.77 −1 3.03 3.03 0 Trimeric 5.29 3.24 −39 9.68 2.98 −69 Tetrameric 0.30 0.12 −60 0.31 0.13 −58 Anhydro 4.19 3.83 −9 6.01 4.93 −18 LysArg- containing 1.81 2.33 +29 2.92 3.28 +12 Cross-linked 23.28 20.35 −13 25.96 20.08 −23 Ala-A2pm 21.81 18.30 −16 23.03 16.81 −27 A2pm-A2pm 1.49 2.02 +36 3.07 3.34 +9 a Murein composition of isolated murein sacculi was determined as described previously (Glauner, 1988). b Tri, disaccharide tripeptide; Tetra, disaccharide–tetrapeptide; Penta, disaccharide–pentapeptide; TetraTri/TetraTetra, bis-disaccharide-tetra-tri-/tetra-tetra-peptide; disaccharide, GlcNAc-β-1.4-MurNAc; tri-/tetra-/penta-peptide, -L-Ala-D-Glu-m-A2pm-(D-Ala)-(D-Ala); A2pm-A2pm indicates a D,L-A2pm-2pm cross-bridge; Ala-A2pm indicates a D,D-Ala-A2pm cross-bridge. Determination of murein precursors and phenotypic suppression of ldcA mutants Since LdcA is present in the cytoplasm it can interact only with cytoplasmic murein precursors and murein turnover products. The turnover products, however, are known to be recycled by their incorporation into the precursor molecules (Goodell, 1985). Therefore, the intracellular pools of the final UDP-activated murein precursors were determined in an LdcA mutant and compared with the pools present in wild-type cells. The precursors were radioactively labeled by growth of the cells in the presence of [3H]diaminopimelic acid, extracted with hot water and separated by HPLC. As shown in Figure 5, in wild-type cells only UDP-MurNAc-pentapeptide could be found in the cytoplasm. By way of contrast in the deletion mutant HfrH-ldcA only a small amount of pentapeptide is detected but there is a prominent peak corresponding to UDP-MurNAc-tetrapeptide which has never been detected in E.coli before. In addition, free tetrapeptide is also found. Figure 5.Determination of the intracellular pools of soluble murein precursors. Murein precursors, radiolabeled with meso-2,6-diamino-[3,4,5-3H]pimelic acid were extracted from E.coli HfrH (wt), HfrH ldcA (ldcA) and JRG582 ldcA (ampDE ldcA). Separation the precursors was performed by HPLC on Nucleosil ODS as described in Materials and methods. The identity of the indicated peaks was verified by authentic standards. A, tetrapeptide; B, UDP-NAcMur-tripeptide; C, UDP-NAcMur-tetrapeptide; D, UDP-NAcMur-pentapeptide. Download figure Download PowerPoint With a possible role in recycling, it was interesting to investigate the effects of mutations in other genes coding for proteins involved in muropeptide recycling. It was particularly interesting to study the effect of an ampD mutation, which is known to cause a severe block in muropeptide recycling, since cytoplasmic hydrolysis of muropeptides into the sugar and peptide part is abolished (Höltje et al., 1994; Jacobs et al., 1995). It was therefore expected that a mutation in ampD would counteract the toxic effect of a lack in LdcA. Indeed, a mutant carrying deletions in both ampD and ldcA showed no lysis during stationary growth. Analysis of precursor pools in this mutant showed a decrease in the amounts of free tetrapeptide and UDP-MurNAc-tetrapeptide, whereas MurNAc-pentapeptide was increased compared with the ldcA single mutant (Figure 5). The observed phenotypic suppression by a mutation in ampD further supports the idea that LdcA takes part in the recycling pathway. But it should be noted that this type of phenotypic suppression might also occur in an in vivo situation when LdcA is inactivated by specific inhibitors. This would not only lead to restored viability, but in bacteria possessing a certain type of a chromosomally encoded β-lactamase (e.g. Citrobacter freundii; Lindberg et al., 1987) this might even lead to a decreased sensitivity against penicillins or cephalosporins. A combination of the ldcA deletion with a mutation in the N-acetylglucosaminidase, encoded by nagZ, which also participates in muropeptide recycling did not change the intracellular pool of UDP-MurNAc-pentapeptide and did not suppress bacteriolysis in the stationary growth phase (data not shown). Discussion Recycling of the murein turnover products, which make up ∼50% per generation of the total murein (Goodell, 1985), is an important energy-saving mechanism. Only a very small amount of turnover material, ∼4–7% per generation, is lost to the medium (Goodell and Schwarz, 1985). Therefore, it is not surprising to find that E.coli has a number of systems which make recycling as efficient as possible. Several uptake systems have been described, which are involved in the transportation of the murein turnover products from the periplasm back into the cytoplasm. Muropeptides are imported via the AmpG transporter (Jacobs et al., 1994), whereas peptides that are released from the murein or from muropeptides by the action of periplasmic amidases are accepted by the Opp (Goodell and Higgins, 1987) and Mpp (Park et al., 1998) permeases. The fate of the major turnover products, the disaccharide-anhydro tri- and tetrapeptides, has been of considerable interest in recent research. A close connection between the regulation of expression of certain chromosomally encoded β-lactamases and murein recycling could be demonstrated (Jacobs et al., 1997). The β-lactamase work showed that the amount of cell wall material released is used to sense the state of the murein sacculus. In a situation where the incorporation of new material is inhibited, e.g. treatment with β-lactam antibiotics, the amount of released turnover products increases dramatically. Certain bacterial species exploit this increase of released muropeptides to induce β-lactamase expression. Actually, most of the known enzymes of the recycling pathway have been identified during studies of β-lactamase induction. Not only has the uptake system AmpG been identified during a screen for genes involved in β-lactamase induction, but also the cytoplasmic amidase AmpD, which releases the peptide moiety from cytosolic muropeptides (Lindberg et al., 1987). Other enzymes known to be involved in recycling are the cytoplasmic β-N-acetylglucosaminidase NagZ (Park, 1996) and the muropeptide ligase Mpl, which has been shown to catalyze the addition of the released peptide to the nucleotide-activated MurNAc (Mengin-Lecreulx et al., 1996). Whereas mutations in these genes mainly affect the efficiency of the recycling process, the enzyme identified in this study is essential for the survival of bacteria that have entered stationary phase. The reaction catalyzed by the LdcA, the conversion of cytosolic tetrapeptide-containing structures to tripeptide-containing structures, is obviously needed for recycling. The formation of the murein precursor UDP-MurNAc-tripeptide occurs by the sequential addition of amino acids to the activated sugar moiety (van Heijenoort, 1994). Alternatively, the muropeptide ligase Mpl uses peptides released during recycling directly (Park et al., 1998). In the next step the general precursor for murein synthesis UDP-MurNAc-pentapeptide is made from UDP-MurNAc-tripeptide by the addition of the dipeptide D-alanyl-D-alanine. Since a single D-alanine cannot be added to a tetrapeptide, an activated tripeptide has to be present to facilitate the synthesis of the final precursor. In the absence of the L,D-carboxypeptidase the hydrolysis of tetrapeptide-conaining structures does not take place and the consequences are striking: the formation of UDP-MurNAc-tetrapeptide occurs, a dead-end intermediate that cannot be converted into a pentapeptide precursor (Figure 6). When incorporated into the murein sacculus, it affects the number of cross-links in the polymer since tetrapeptides can function only as acceptors and not as donors for transpeptidation. A decrease in the overall degree of cross-linkage in the sacculus is the consequence of this failure to synthesize a suf
Referência(s)