![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Journal of Bacteriology, June 2007, p. 4332-4337, Vol. 189, No. 11
0021-9193/07/$08.00+0 doi:10.1128/JB.01829-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
,
Laetitia Fontaine,1
Jean-Luc Mainardi,2
Richard Daniel,3
Jeff Errington,3
Bernard Hallet,1 and
Pascal Hols1*
Unité de Génétique, Institut des Sciences de la Vie, Université catholique de Louvain, 1348 Louvain-la-Neuve, Belgium,1 INSERM, U655-LRMA, Université Paris 6, Université Paris-Descartes, Centre de Recherche Biomedicales des Cordeliers, Paris, F-75006, France,2 Institute for Cell and Molecular Biosciences, The Medical School, University of New Castle upon Tyne, New Castle upon Tyne NE2 4HH, United Kingdom3
Received 6 December 2006/ Accepted 19 March 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT TEXTREFERENCES |
|---|
|
TEXT |
|---|
|
TopABSTRACTTEXTREFERENCES |
|---|
1,000-fold-lower affinity (for reviews, see references 10 and 15). This resistance mechanism results from the acquisition of at least three genes (vanH, vanA [or vanB], and vanX) generally clustered in an operon (2, 3). The vanH gene encodes a D-lactate dehydrogenase responsible for the conversion of pyruvate into D-Lac (5); vanA (or vanB) encodes a specific D-Ala-D-Lac ligase required for the formation of D-Ala-D-Lac depsipeptides, which are incorporated at the end of the modified precursors (2, 7, 12); and the product of vanX is a highly specific dipeptidase needed to selectively eliminate D-Ala-D-Ala dipeptides produced by the host endogenous D-Ala-D-Ala ligase Ddl (1, 25) (Fig. 1). Additional accessory genes have been reported to contribute to vancomycin resistance in different enterococcal strains (10, 15). For example, vanY encodes a D,D-carboxypeptidase capable of hydrolyzing the terminal D-Ala of the UDP-N-acetylmuramyl (MurNAc) pentapeptide precursors, further increasing the resistance level (1, 2) (Fig. 1).
|
Genes homologous to the resistance genes vanH, vanA (or vanB), and vanX are also found together or separately in the genomes of other organisms, where they appear to have distinct physiological roles. In particular, homologues of the VanX dipeptidase were identified in a wide variety of bacterial species, including gram-negative bacteria, which are normally not challenged by vancomycin because of the failure of the antibiotic to penetrate the outer membrane barrier. However, only a few of these VanX variants have been studied at the functional level (20).
In Escherichia coli, ddpX, encoding a VanX-related dipeptidase, is cotranscribed with a dipeptide transport system (ddpABCDF) during the stationary phase of growth. It was proposed that D-Ala-D-Ala dipeptides released from cell wall peptidoglycan can be reimported from the periplasm through the Ddp transporter and subsequently hydrolyzed by DdpX to provide D-Ala as a nitrogen and carbon source (18, 29). Likewise, the D-Ala-D-Ala dipeptidase PcgL, found in the pathogen species Salmonella enterica, confers the ability to use D-Ala-D-Ala as a sole carbon source and is involved in survival under starvation conditions (16, 23).
Lactobacillus plantarum NCIMB8826 is an intrinsically vancomycin-resistant lactic acid bacterium that produces D-Lac-ended peptidoglycan precursors (13). This bacterium possesses the requisite D-Ala-D-Lac ligase (DdlLp) for depsipeptide synthesis (13) and a D-Lac dehydrogenase for D-Lac production (13) (Fig. 1). We recently reported that D-Lac is also produced by L-lactate racemization in L. plantarum (14) (Fig. 1). Racemization seems to represent an important rescue pathway to supply the cell with D-Lac for peptidoglycan synthesis. An L. plantarum mutant totally impaired in D-Lac production strictly requires exogenous D-Lac in the medium for growth, reflecting a strong selectivity for D-Lac incorporation into the peptidoglycan of L. plantarum (14).
In the present study, we show that L. plantarum also contains a functional D-Ala-D-Ala dipeptidase belonging to the VanX family. The biological function of this endogenous dipeptidase and its implications for intrinsic vancomycin resistance and D-Lac selectivity for peptidoglycan synthesis were investigated.
Identification of a vanX-like gene in L. plantarum. Analysis of the genome of L. plantarum WCFS1 (17) revealed the presence of a gene homologous to the resistance gene vanX (lp_0769; designated aad for D-alanyl-D-alanine dipeptidase). This gene is not located in the same locus as the gene coding for the D-Ala-D-Lac ligase (DdlLp) of L. plantarum, nor is it in a locus involved in D-Lac production. Inspection of the surrounding sequences indicates that aad is likely to be transcribed as a monocistronic unit, with the adjacent open reading frames being on the complementary strand (data not shown).
Phylogenetic analysis of VanX-like sequences from gram-positive and gram-negative bacteria reveals that they form a rather heterogeneous family of enzymes (see Fig. S1A in the supplemental material). The Aad protein clearly belongs to a distinct subgroup, displaying only 30% identity with VanXA from Enterococcus faecium BM4147 (3, 4) and 32% and 28% identity with the dipeptidases DdpX and PcgL from E. coli (18) and S. enterica (16), respectively. This Aad subgroup contains VanX-like proteins identified in the genomes of Lactobacillus brevis ATCC 367 (accession number YP_796214) and Lactobacillus salivarius UCC118 (accession number YP_536261) (75% and 72% identical to Aad, respectively; see Fig. S1A in the supplemental material). These two VanX-like-protein-encoding genes are present in a different genetic context than aad, but, as found in L. plantarum, they are not located in loci involved in peptidoglycan biosynthesis or D-Lac production.
Enzymes of the VanX family are metallopeptidases that use Zn2+ as a cofactor. Despite a relatively low level of overall similarity, members of the family are characterized by several highly conserved residues involved in catalysis (E181 and R71 in E. faecium VanXA) and Zn2+ cofactor binding (H116, D123, and H184) (8, 18, 19, 22). In the L. plantarum protein Aad, all these amino acids are conserved except H184, which is replaced by an aspartate (see Fig. S1B in the supplemental material).
The L. plantarum aad gene encodes a functional D-Ala-D-Ala dipeptidase. To determine whether the L. plantarum aad gene encodes a functional D-Ala-D-Ala dipeptidase, the levels of D-Ala-D-Ala hydrolysis were examined in the wild-type strain, an aad knockout strain, and an Aad-overproducing strain.
Overexpression was achieved by cloning aad under the control of the nisin-inducible promoter PnisA from Lactococcus lactis (24). The resulting plasmid (pGIM023) was introduced into L. plantarum strain NZ7100, a WCFS1 derivative carrying the nisR and nisK regulatory genes required for nisin induction (M. Kleerebezem, unpublished data). The aad knockout mutant (MD119) was constructed in NZ7100 using a two-step homologous recombination procedure as previously described (14). Aad inactivation resulted in the stable replacement of the aad open reading frame (encoding 180 amino acid residues out of 185) by an erythromycin resistance cassette. For details on the construction of these recombinant strains, see the supplemental text and Table S1 in the supplemental material.
The D-Ala-D-Ala dipeptidase activities of the different strains were measured in crude cell extracts prepared from exponentially growing cultures (MRS broth [Becton Dickinson, Cockeysville, MD]; optical density at 600 nm [OD600] = 2). Nisin (50 ng/ml) was added when required to allow expression from the PnisA promoter. The enzymatic assay consisted of measurement of the amount of D-Ala released from the D-Ala-D-Ala substrate (100 mM) using a D-amino acid oxidase coupled to a peroxidase (2).
In the presence of Zn2+, low but significant D-Ala-D-Ala dipeptidase activity was measured in extracts of the wild-type strain, whereas D-Ala-D-Ala hydrolysis detected in cell extracts of the aad mutant was in the range of the background level measured in the absence of substrate or cosubstrate (Table 1). Overexpression of aad increased the observed activity about 200-fold compared to the activity in the wild-type strain or the control strain containing the empty expression vector pGIM020 (Table 1).
|
In contrast to what has been observed for the enterococcal dipeptidase VanXA and other enzymes of the same family, which remain sufficiently charged in cofactors when extracted from the cells (18; M. Deghorain, unpublished data), the optimal activity of L. plantarum Aad required the addition of Zn2+ to the in vitro reaction mixture (Table 1). This may reflect a lower affinity for the cofactor as a consequence of the substitution of the conserved histidine (H184) in the Zn2+-binding motif of the protein.
The D-Ala-D-Ala dipeptidase specific activity of the wild-type strain remained constant throughout growth (data not shown), suggesting that the expression of aad is not regulated by a growth-dependent mechanism as was reported for the dipeptidase DdpX of E. coli (18).
Aad contributes to the selectivity for the incorporation of D-Ala-D-Lac at the terminal position of the peptidoglycan precursor side chain. As mentioned above, L. plantarum is highly selective for the production and incorporation of D-Ala-D-Lac-ended peptidoglycan precursors in the cell wall (13, 14). A key enzyme for this selectivity is DdlLp, which is responsible for the production of D-Ala-D-Lac depsipeptides that are incorporated at the end of the peptidoglycan precursor peptide chain by MurF (Fig. 1).
To determine whether Aad contributes to this selectivity, the intracellular pools of UDP-N-acetylmuramyl (UDP-MurNAc) pentadepsipeptides and UDP-MurNAc pentapeptides were compared in exponentially growing cells of the wild type and an aad knockout strain. Peptidoglycan precursors were extracted as reported before (14).
As previously observed (13), precursors ending in D-Ala-D-Ala were undetectable in the wild-type strain, whereas they accounted for 7% of the mature precursors found in the aad mutant (Table 2). This result shows that the DdlLp ligase produces both D-Ala-D-Lac depsipeptides and D-Ala-D-Ala dipeptides for their incorporation into the peptidoglycan precursors by MurF and that Aad prevents the formation of D-Ala-D-Ala-ended precursors in the wild-type strain by eliminating the D-Ala-D-Ala dipeptides. Production of UDP-MurNAc-pentapeptide precursors ending in D-Ala-D-Ala had already been observed in an L. plantarum mutant deficient for the L- and D-lactate dehydrogenases (TF103; ldhL ldhD double mutant) (13). This mutant is impaired in D-Lac production, which was not the case with the aad mutant.
|
D-Ala-D-Ala-ended peptidoglycan precursors can be transported to the external face of the plasma membrane. In order to further investigate the ability of Aad to control the intracellular composition in peptidoglycan precursors and its impact on cell wall biosynthesis, the monospecific D-Ala-D-Ala ligase from L. lactis (encoded by ddlLc), a vancomycin-sensitive gram-positive bacterium, was expressed in the L. plantarum wild-type strain (NZ7100) and in the aad mutant. It was shown previously that DdlLc is capable of promoting the synthesis of peptidoglycan precursors ending in D-Ala-D-Ala in L. plantarum (14).
Different levels of ddlLc expression were obtained by cloning the ddlLc gene as a transcriptional and a translational fusion with the PnisA promoter, on a multicopy plasmid, yielding pGIM021 and pGIM121, respectively (see the supplemental text and Table S1 in the supplemental material). Dipeptidase activity in cell extracts from strains expressing the ligase gene ddlLc was unchanged compared to that of the wild-type strain, indicating that Aad activity is not regulated by the presence of D-Ala-D-Ala dipeptides (data not shown).
For both constructs, induction of the PnisA promoter by adding nisin in the culture medium had toxic effects, making the observations difficult to reproduce. In the absence of nisin, basal expression of the ddlLc gene from the transcriptional and translational fusion in the wild-type strain led to the accumulation of 5% and 58% of D-Ala-D-Ala-ended peptidoglycan precursors, respectively [NZ7100 (pGIM021) and NZ7100 (pGIM121)] (Table 2). This proportion was increased to about 97% in the aad deletion mutant expressing the PnisA::ddlLc translational fusion [MD119 (pGIM121)] (Table 2).
These results further demonstrate the contribution of the Aad dipeptidase in preventing the accumulation of peptidoglycan precursors ending in D-Ala-D-Ala for cell wall synthesis.
Strains producing increasing amounts of D-Ala-D-Ala-ended precursors were found to be gradually less resistant to vancomycin, with the strains expressing the PnisA::ddlLc translational fusion being completely sensitive to the antibiotic (Table 2). This indicates that noncanonical D-Ala-D-Ala-ended UDP-MurNAc-pentapeptides are correctly transported to the external face of the plasma membrane to serve as a substrate for the PBPs at the outside of the cell, where they become accessible to the antibiotic in the form of the lipid II complex.
To confirm this, and to visualize the sites of precursor incorporation into the growing peptidoglycan network, exponentially growing cells from the different strains were incubated with a fluorescent derivative of vancomycin (Van-FL) and examined by epifluorescence microscopy as previously reported (11). Consistent with the high level of resistance to vancomycin and the absence of detectable peptidoglycan precursors ending in D-Ala-D-Ala, cells of the wild-type strain NZ7100 were not stained by the Van-FL probe (Table 2; Fig. 2A). In contrast, strains expressing the ddlLc gene were labeled to different extents, depending on the ratio of D-Ala-D-Ala- and D-Ala-D-Lac-ended precursors that they produced (Table 2). Fluorescence was specifically localized at the division sites and along the longitudinal axis of the cells, forming peripheral spots and transversal bands suggestive of the formation of a short spiral around the cell (Fig. 2B), as previously reported for Bacillus subtilis (11).
|
Since the strains described above still produce peptidoglycan precursors ending in D-Ala-D-Lac (Table 2), it cannot be ruled out that their growth exclusively relies on the incorporation of D-Ala-D-Lac-ended precursors into the growing peptidoglycan. This is not the case for the ldhD lar double mutant of L. plantarum (strain PG1104, a derivative of PG6212 ldhD lar containing a chromosomal integration of the empty expression plasmid pMec10), which is totally impaired in D-Lac production. This mutant is unable to grow in a chemically defined medium (MPL) (9) deprived of D-Lac (Fig. 3) (14). However, expression of ddlLc from L. lactis can partially rescue this growth defect through the production of D-Ala-D-Ala-ended peptidoglycan precursors (Fig. 3) (strain PG1174, a derivative of PG6212 containing the chromosomally integrated PnisA::ddlLc fusion) (14).
|
In the presence of D-Lac in the culture medium, growth of both strains was identical to the growth of the similarly D-Lac-supplemented parental ldhD lar mutant (Fig. 3 and data not shown). In a D-Lac-deprived medium, inactivation of the aad gene neither suppressed the growth defect of the ldhD lar mutant nor improved the growth of the mutant expressing ddlLc (Fig. 3). Thus, it appears that the pool of D-Ala-D-Ala is not the limiting growth factor in this background but rather that D-Ala-ended peptidoglycan precursors are not optimal substrates for cell wall synthesis in L. plantarum.
Concluding remarks. This study reports the characterization of a VanX-like dipeptidase, Aad, in L. plantarum and its role in resistance to vancomycin. This enzyme was found to direct incorporation of D-lac into peptidoglycan precursors, rather than D-Ala, through the specific hydrolysis of D-Ala-D-Ala peptides. However, a second key enzyme in promoting the synthesis of this specific form of peptidoglycan precursor is the D-Ala-D-Lac ligase DdlLp. We show here that DdlLp produces D-Ala-D-Ala dipeptide in L. plantarum, even under conditions where D-Lac is not limiting. Therefore, Aad contributes to the exclusive formation of D-Ala-D-Lac-ended precursors by hydrolyzing D-Ala-D-Ala arising from this dual specificity (Fig. 1). This conclusion does not rule out the possibility that Aad may have additional functions, such as providing D-Ala as a carbon or nitrogen source, as was found in E. coli and S. enterica (16, 18, 23, 29).
The biological reason for the selective incorporation of D-Ala-D-Lac-ended precursors in the cell wall of L. plantarum remains unclear. The possibility that it may provide L. plantarum with a selective advantage in specific ecological niches by conferring resistance to vancomycin cannot be ruled out. However, it is unlikely that the genes involved in D-Lac production and precursor incorporation (i.e., ldhD, ddl, and aad) were transferred from glycopeptide producers to L. plantarum as was proposed for other vancomycin-resistant species, since they are not clustered in the genome and do not display obvious signs of recent lateral exchange. Instead, it is tempting to speculate that the resistance genes that were acquired by vancomycin producers, and subsequently by enterococci, initially originated from other organisms, where they contribute to a separate physiological function, as proposed here for L. plantarum.
In the absence of Aad, D-Ala-D-Ala dipeptides can be processed and incorporated into peptidoglycan precursors. Furthermore, the use of a Van-FL probe showed that D-Ala-D-Ala-ended precursors can be exported across the cell membrane, supporting the view that the enzymes catalyzing the corresponding reactions (i.e., MurF, MraY, and MurG) do not have an absolute specificity for the fifth position of the peptidoglycan peptide (26) (Fig. 1).
The observation that inactivation of aad failed to rescue a mutant strain unable to synthesize D-Lac (ldhD lar double mutant) shows that the amount of D-Ala-D-Ala-ended precursors erroneously synthesized by Ddl of L. plantarum is insufficient to support cell wall synthesis. Surprisingly, only a partial restoration of growth is obtained by the overexpression of the ddlLc ligase gene of L. lactis in the triple mutant (ldhD lar aad), which should exclusively produce D-Ala-D-Ala dipeptides. This suggests that there is selectivity for the D-Ala-D-Lac-ended precursors in the later stages of peptidoglycan biosynthesis (Fig. 1). In particular, carboxypeptidases may play a role in this specificity. Five genes encoding potential carboxypeptidases are present in the L. plantarum genome, including one homologue of the enterococcal gene vanY (lp_1010; designated dacB; 27% identity with the VanYA gene from E. faecium BM4147) (28). High-molecular-weight PBPs catalyzing the transglycosylation and transpeptidation reactions, murein hydrolases, and carboxypeptidases involved in peptidoglycan maturation may also have a high specificity for D-Lac in the last position of the peptide chain (Fig. 1).
Although D-Lac is required for growth of the ldhD lar mutant, suppressor mutations can be selected, restoring growth in the absence of D-Lac based on the production of peptidoglycan precursors ending in D-Ala-D-Ala (14). Genetic analysis of these suppressor mutants in the presence and absence of aad will help to determine the key enzymes controlling the nature of the last residue of the peptide chain and perhaps elucidate the biological significance of this specificity.
|
ACKNOWLEDGMENTS |
|---|
We warmly thank J. Delcour for helpful discussions and scientific advice. We are grateful to M. Kleerebezem for the communication of the aad locus sequence prior to publication of the L. plantarum WCFS1 sequence and for the generous gift of strain NZ7100. We acknowledge J. Guzzo for providing the vanX-like sequence from Oenococcus oeni IOEB8413 before publication and B. Dehertogh, A. Schanck, and E. Viaene for scientific advice and technical assistance.
|
FOOTNOTES |
|---|
Published ahead of print on 30 March 2007.
Supplemental material for this article may be found at http://jb.asm.org/.
Present address: Wageningen Centre for Food Sciences, NIZO Food Research, 6710 BA Ede, The Netherlands.
|
REFERENCES |
|---|
|
TopABSTRACTTEXTREFERENCES |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Appl. Environ. Microbiol. | Infect. Immun. | Eukaryot. Cell |
|---|---|---|
| Mol. Cell. Biol. | J. Virol. | Microbiol. Mol. Biol. Rev. |
| ALL ASM JOURNALS |
