Biochemical and Genetic Characterization of the vanC-2 Vancomycin Resistance Gene Cluster of Enterococcus casseliflavus ATCC 25788

Abstract

The vanC-2 cluster of Enterococcus casseliflavus ATCC 25788 consisted of five genes (vanC-2, vanXY_C-2, vanT_C-2, vanR_C-2, and vanS_C-2) and shared the same organization as the vanC cluster of E. gallinarum BM4174. The proteins encoded by these genes displayed a high degree of amino acid identity to the proteins encoded within the vanC gene cluster. The putative d,d-dipeptidase-d,d-carboxypeptidase, VanXY_C-2, exhibited 81% amino acid identity to VanXY_C, and VanT_C-2 displayed 65% amino acid identity to the serine racemase, VanT. VanR_C-2 and VanS_C-2 displayed high degrees of identity to VanR_C and VanS_C, respectively, and contained the conserved residues identified as important to their function as a response regulator and histidine kinase, respectively. Resistance to vancomycin was expressed inducibly in E. casseliflavus ATCC 25788 and required an extended period of induction. Analysis of peptidoglycan precursors revealed that UDP-N-acetylmuramyl-l-Ala-δ-d-Glu-l-Lys-d-Ala-d-Ser could not be detected until several hours after the addition of vancomycin, and its appearance coincided with the resumption of growth. The introduction of additional copies of the vanT_C-2 gene, encoding a putative serine racemase, and the presence of supplementary d-serine in the growth medium both significantly reduced the period before growth resumed after addition of vancomycin. This suggested that the availability of d-serine plays an important role in the induction process.

Enterococci of the VanA, VanB, and VanD phenotypes possess high-level resistance to glycopeptide antibiotics, which is a result of the production of alternative cell wall precursors which end in d-lactate (d-Lac) and the elimination of d-alanine (d-Ala)-terminating precursors to which vancomycin binds (4, 7, 25, 29, 31). Low-level resistance to vancomycin is observed in enterococci of the VanE, VanG, and VanC phenotypes, which replace d-Ala with d-serine (d-Ser) in the C-terminal position of UDP-N-acetylmuramyl-l-Ala-γ-d-Glu-l-Lys-d-Ala-d-Ala (pentapeptide[d-Ala]) (7, 15, 19, 30). VanA-, VanB-, and VanD-type enterococci all possess genes that encode a depsipeptide ligase and a dehydrogenase that reduces pyruvate to d-Lac, both of which are essential for the production of the d-Ala-d-Lac depsipeptide used in cell wall synthesis (8, 10, 13). The removal of precursors terminating in d-Ala prevents the interaction of the glycopeptide antibiotic with its target. Two enzymes are involved in this process: VanX (VanX_B), a d,d-dipeptidase that hydrolyzes d-Ala-d-Ala, and VanY (VanY_B), a d,d-carboxypeptidase that removes the terminal d-Ala residue of late pentapeptide[d-Ala] precursors that are produced if elimination of the d-Ala-d-Ala dipeptide is incomplete (5, 29).

Three species of enterococci possess the VanC phenotype: Enterococcus gallinarum, Enterococcus flavescens, and Enterococcus casseliflavus. The vanC gene cluster of E. gallinarum BM4174 consists of five genes (1). Three genes from the cluster, vanC-1, vanXY_C-2, and vanT, are necessary and essential for resistance (1). VanC-1 is a ligase that synthesizes d-Ala-d-Ser, which is required for the production of the alternative cell wall precursor (12). VanT is a membrane-bound serine racemase that provides a source of d-Ser (2). VanXY_C-2 combines the activities of VanX and VanY in a single enzyme and contains the consensus sequences for zinc and substrate binding and for catalysis that are present in both VanX- and VanY-type enzymes (27). Genes encoding d-Ala-d-Ser ligases have also been identified within the vanE and vanG gene clusters (15, 19). Analysis of the vanG cluster of Enterococcus faecalis has revealed the presence of a putative serine racemase and d,d-peptidases (19).

Regulation of the expression of the vancomycin resistance gene clusters is controlled by a two-component regulatory system (24). These systems consist of VanR-type proteins, which are response regulators, and VanS-type proteins, which are histidine kinases (3, 17, 35). In the vanA, vanB, and vanD clusters the genes encoding the two-component regulatory system are present upstream of the structural genes encoding resistance proteins, whereas in the vanC cluster they are present downstream of the genes encoding resistance proteins (1, 6, 10, 13). However, the vanC cluster of E. gallinarum BM4174 is expressed constitutively, and two regions upstream of vanC-1 and vanR_C have been identified as potential promoters (1). Other strains of E. gallinarum in which resistance is inducible have been investigated (32).

Prior to this investigation a single gene from the vanC-2 cluster of E. casseliflavus ATCC 25788 had been cloned and characterized. VanC-2 is a d-Ala-d-Ser ligase that displays 71% amino acid identity to VanC-1 (21, 23). This work describes the cloning and sequencing of the remaining genes of the vanC-2 cluster and examines the expression of vancomycin resistance in E. casseliflavus.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Enterococci were grown in brain-heart-yeast extract (BHY) broth or on BHY agar. Gentamicin (60 μg/ml) was added to the medium for E. casseliflavus ATCC 25788 containing derivatives of pAT392. Induction of resistance was initiated by the addition of vancomycin (2 μg/ml). Escherichia coli XL1-Blue (9) was used for cloning the vancomycin resistance genes and was grown in Luria-Bertani broth or agar containing either 50 μg of ampicillin per ml when derivatives of pUC18 were present (22) or gentamicin (8 μg/ml) to maintain derivatives of pAT392 (5).

DNA manipulation.

Total DNA from E. casseliflavus ATCC 25788 was extracted as described previously (26). Cloning, digestion with restriction endonucleases (Roche Molecular Biochemicals, Mannheim, Germany), isolation of plasmid DNA (Wizard Plus SV Minipreps; Promega), ligation, and transformation were carried out by standard methods (33). Plasmid constructs based on pAT392 were purified from E. coli and electroporated into E. casseliflavus as described previously (11).

Cloning and sequencing of the vanC-2 gene cluster.

The sequences of the vanXY_C-2, vanT_C-2, and vanR_C-2 genes and the 5′ end of the vanS_C-2 gene were obtained from the inserts present in plasmids pUCX1, pUCT1, pUCR1, and pUCS1 (Fig. 1). The remaining portion of the vanS_C-2 gene was obtained by inverse PCR after the digestion of chromosomal DNA with ClaI. A digoxigenin-labeled probe consisting of the vanR_C-2 gene hybridized to a 3.1-kb ClaI fragment of chromosomal DNA. Total DNA was then digested with ClaI and self-ligated at 16°C for 16 h at a concentration of 10 μg/ml. An inverse PCR with Pwo polymerase (Roche Molecular Biochemicals) was performed with primers R4 and S3 (Table 1). The PCR product, of the expected size of 2.5 kb, was digested with XbaI and SacI and cloned into pUC18 digested with the same enzymes to create plasmid pUCS2. DNA sequencing was performed by the dideoxy chain terminator method with fluorescent cycle sequencing with dye-labeled terminators (ABI Prism TM Dye Terminator Cycle Sequencing Ready Reaction Kit; Perkin-Elmer) on a model 373A automated DNA sequencer.

FIG. 1.

Physical map of the vanC-2 gene cluster of E. casseliflavus ATCC 25788. The fragments cloned in plasmids pUCX1, pUCT1, pUCR1, pUCS1, pUCS2, pIC1, pIC2, and pIC3 are indicated by solid lines. Arrows represent each open reading frame.

TABLE 1.

Primers used in this study

Primer	Origin nta (source or reference)	Nucleotide sequence (5′-3′)	Relevant property
DEGX	512-486	CTAAGAGCTCCCCKACATAMCKRAAATGCCAWGGTTC	SacI site (underlined)
DEGT	1727-1704	CTAAGAGCTCRCCRAGCCRTAGGCRTCGGCYTT	SacI site (underlined)
DEGR	3465-3446	GTAATCTAGACCNACHCCSCRBAYVGTT	XbaI site (underlined)
DEGS	4447-4464	CTAAGAGCTCAATCGSMAGICCMAGWCC	SacI site (underlined)
S3	4259-4276	ACTTGATCCGCAATGCCA
C3	487-503 (22)	GTAATCTAGACTCCTACGATTCTCTTG	XbaI site (underlined)
X3	431-449	AACAGGAGATTACAGGGAT
R3	3377-3395	GTAATCTAGAACAGTGATGGCACATATCG	XbaI site (underlined)
R4	2909-2890	GTAATCTAGAAGGCCAAAGCGCTCGTTCCA	XbaI site (underlined)
C1	8-26 (22)	CTAAGAGCTCTCGGAAAAGCGGAAGGAAG	SacI site (underlined)
C2	1085-1068 (22)	GTAATCTAGATCATTTGACTTCCTCCTT	XbaI site (underlined)
X1	1053-1070 (22)	CTAAGAGCTCCTGCTTGTCTTAGCAAAG	SacI site (underlined)
X2	573-556	GTAATCTAGATCATGCGAACTGCCTCGC	XbaI site (underlined)
T1	544-561	CTAAGAGCTCCTTGAACAAACTGCGAGG	SacI site (underlined)
T2	2690-2673	GTAATCTAGACTACTTTGAACTAGAGGT	XbaI site (underlined)

^aWhere appropriate, nucleotide (nt) positions refer to the numbering for the sequence submitted to GenBank under accession number AY033089.

Plasmid construction.

Plasmid pUCX1 contained the 5′ end of vanXY_C-2 and was constructed by cloning the 1.0-kb PCR product, obtained through the use of a combination of a specific primer (primer C3) targeted against the vanC-2 gene and a degenerate primer (primer DEGX) targeted against a vanXY-type gene, into SmaI-digested pUC18. Plasmid pUCT1 was obtained by cloning the 2.8-kb PCR product synthesized by using primers X3 and DEGT into the SmaI site of pUC18. Primer DEGT was a degenerate primer based on the sequences of the pyridoxal phosphate binding site of VanT and alanine racemases. Plasmid pUCR1 was constructed by cloning the 3.0-kb PCR product, obtained by using the specific primer X3 and degenerate primer DEGR, into the SmaI site of pUC18. The degenerate primer DEGR was designed by using a series of conserved residues present in the C-terminal domain of enterococcal response regulators. Plasmid pUCS1 contained the 1.5-kb PCR product obtained by using primers R3 and DEGS cloned into SmaI-digested pUC18. The sequence of the degenerate primer was based on a series of conserved residues found in the C-terminal domain of histidine protein kinases. The construction of plasmid pUCS2 is described above. Plasmids pIC1, pIC2, and pIC3 were constructed by cloning vancomycin resistance genes into the SacI and XbaI sites of the vector pAT392. Plasmid pIC1 contained the vanC-2 gene and its ribosomal binding site (RBS) placed under the control of the P₂ promoter. The vanC-2 gene and its RBS were amplified by PCR with primers C1 and C2, digested with SacI and XbaI, and cloned into pAT392. Plasmid pIC2 contained the vanXY_C-2 gene together with its RBS, which were amplified by PCR with primers X1 and X2 and cloned into pAT392. Plasmid pIC3 was constructed by cloning the vanT_C-2 gene and its RBS, amplified by PCR with primers T1 and T2, into the XbaI-SacI site of pAT392. The primers used in this study are shown in Table 1.

Sequence analysis.

DNA sequence analysis was performed with the programs of the Wisconsin Package (version 10.2; Genetics Computer Group, Madison, Wis.). Protein alignments were produced with the CLUSTALW program (34). Transmembrane helix predictions were produced with the TM PRED program (16).

Analysis of peptidoglycan precursors and biochemical activities.

Cytoplasmic peptidoglycan precursors were extracted from E. casseliflavus and analyzed by high-pressure liquid chromatography (HPLC) as described previously (20). The activities of the d,d-dipeptidase and serine racemase present in the cytoplasm and cell membrane, respectively, were determined as described earlier by using an assay for d-amino acids (2, 27).

Nucleotide sequence accession number.

The nucleotide sequence of the vanC-2 vancomycin resistance gene cluster of E. casseliflavus ATCC 25788 has been deposited in GenBank under accession number AY033089.

RESULTS

Sequence of the vanC-2 gene cluster of E. casseliflavus ATCC 25788.

The genes of the vanC-2 cluster were amplified by PCR with a combination of specific and degenerate primers designed by using conserved residues found in VanXY_C, VanT, VanR_C, VanS_C, and related proteins. The complete nucleotide sequences of the vanXY_C-2, vanT_C-2, and vanR_C-2 genes and the 5′ end of the vanS_C-2 gene were obtained by sequencing, on both strands, the inserts present in plasmids pUCX1, pUCT1, pUCR1, and pUCS1. The 3′ end of the vanS_C-2 gene was amplified by an inverse PCR approach and cloned, and the insert of plasmid pUCS2 was sequenced on both strands. The genes of the vanC-2 cluster were organized in the same order as those of the vanC cluster of E. gallinarum BM4147, with the two regulatory genes situated downstream of the three overlapping structural genes, vanC-2, vanXY_C-2, and vanT_C-2 (Fig. 1).

The functions of the putative proteins encoded by vanXY_C-2, vanT_C-2, vanR_C-2, and vanS_C-2 were deduced on the basis of similarity to the proteins encoded within the vanC gene cluster of E. gallinarum BM4174 and to other vancomycin resistance proteins. VanXY_C-2 was identified as a putative d,d-dipeptidase-d,d-carboxypeptidase which displayed 81% amino acid identity to VanXY_C. It also displayed considerable identity to the VanXY_E-type protein (44%) and the VanY-type protein from the vanA (43% over 158 amino acids) and vanB (46% over 127 amino acids) gene clusters. VanXY_C-2 contained all the conserved residues believed to be involved in substrate binding and catalysis and Zn²⁺ binding (27). The vanT_C-2 gene was proposed to encode a serine racemase that displayed 64% amino acid identity to VanT. The N-terminal domain of VanT_C-2 contained a number of predicted transmembrane helices, as found in VanT, and the C-terminal domain contained the pyridoxal 5′-phosphate attachment site that is highly conserved in alanine racemases and VanT (2). The C-terminal domain of VanT_C-2 also demonstrated considerable amino acid identity (35 to 40%) to alanine racemases from a variety of gram-positive and gram-negative bacteria. Based on similarity to VanR_C and VanS_C, the proteins encoded by vanR_C-2 and vanS_C-2 were proposed to correspond to a two-component regulatory system controlling the expression of the vanC-2 gene cluster. VanR_C-2 displayed 91% amino acid identity to VanR_C, the highest degree of identity of the proteins encoded within the vanC-2 and vanC gene clusters, and contained the conserved motifs associated with response regulators. VanR_C-2 also exhibited extensive identity to VanR (50%) and VanR_D (65%). VanS_C-2 was identified as a putative histidine protein kinase which possessed the five conserved motifs characteristic of this class of proteins (24). It displayed 81% amino acid identity to VanS_C and contained two predicted transmembrane domains, as found in other VanS-type proteins (1, 6, 13).

Inducible expression of vancomycin resistance in E. casseliflavus ATCC 25788.

E. casseliflavus ATCC 25788 possesses intrinsic, low-level resistance to vancomycin (MIC, 8 μg/ml). Analysis of growth was used to determine if resistance was expressed inducibly or was constitutive. After a subinhibitory concentration of vancomycin (2 μg/ml) was added to an exponentially growing culture of E. casseliflavus, a delay of 4 to 5 h occurred before normal growth resumed (Fig. 2). This indicated that expression of vancomycin resistance is inducible but that an extended period of induction is required. Population analysis indicated that the delay in resumption of growth was not due to selection of a subpopulation of vancomycin-resistant cells (data not shown). Peptidoglycan precursors terminating in d-Ser did not appear until 225 min after the addition of vancomycin (Fig. 3), and this coincided with the resumption of growth of the culture. Prior to this only pentapeptide[d-Ala] and UDP-N-acetylmuramyl-l-Ala-γ-d-Glu-l-Lys-d-Ala (tetrapeptide) were present. In the fully induced state only tetrapeptide and UDP-N-acetylmuramyl-l-Ala-δ-d-Glu-l-Lys-d-Ala-d-Ser (pentapeptide[d-Ser]) were detected. This result indicated that VanXY_C-2 was active and that its activity as a d,d-carboxypeptidase had probably resulted in the production of tetrapeptide from pentapeptide[d-Ala].

FIG. 2.

Induction of vancomycin resistance in E. casseliflavus ATCC 25788. Vancomycin (2 μg/ml) was added to a growing culture of E. casseliflavus (solid line), resulting in a delay of 270 min before growth was resumed. The point at which vancomycin was added is indicated by the arrow. The dashed line represents growth of E. casseliflavus without the addition of vancomycin.

FIG. 3.

Analysis of cell wall precursors produced during induction with vancomycin (B). Precursors were extracted and analyzed by HPLC at the time points indicated by the block arrows. The solid-line arrow in panel A indicates the point at which vancomycin (2 μg/ml) was added to the culture. The dashed-line arrows in panel A indicate the time points at which cultures were harvested for analysis of cell wall precursors.

It has been shown that production of d-Ala- and d-Lac-ending precursors in similar amounts in VanA-type enterococci does not result in significant resistance to vancomycin (4). Analysis of precursors at the time of resumption of growth (225 min) demonstrated that only 34% of the precursor pool consisted of d-Ser-ending precursors, with 24% ending in d-Ala and the remainder being tetrapeptide. This ratio was apparently sufficient for the cells to overcome the inhibition by vancomycin.

Role of d-Ser in determining length of time required for induction.

In order to investigate the factors causing the long induction period, a number of plasmids containing genes from the vanC-2 cluster were constructed and introduced into E. casseliflavus. The vanC-2, vanXY_C-2, and vanT_C-2 genes were each cloned into multicopy plasmid pAT392 to allow their constitutive expression under the control of the relatively weak P₂ promoter (5). The three constructs and pAT392 were introduced into E. casseliflavus ATCC 25788 by electroporation.

E. casseliflavus/pIC1 (vanC-2) and E. casseliflavus/pIC2 (vanXY_C-2) did not resume growth faster than E. casseliflavus/pAT392 after the addition of vancomycin, and all three strains displayed similar induction periods (data not shown). E. casseliflavus/pIC2 (vanXY_C-2) possessed 38% greater cytoplasmic d,d-dipeptidase activity than E. casseliflavus/pAT392, indicating that the supplementary vanXY_C-2 gene was expressed and that the protein was active. However, the presence of additional copies of vanT_C-2, predicted to encode a serine racemase, had a significant effect on the induction period, which was reduced by approximately 50% compared to that with E. casseliflavus/pAT392 (Fig. 4A). The serine racemase activity of E. casseliflavus/pIC4 (vanT_C-2) was found to be 30% greater than that of the control strain, suggesting that the enhanced activity was responsible for the reduced induction period. In order to confirm the importance of d-Ser in shortening the induction period, E. casseliflavus/pAT392 was grown in medium supplemented with 25 mM d-Ser. Comparison of the growth of this strain with that of the same strain grown in unsupplemented medium demonstrated that d-Ser reduced the induction period from 260 to 135 min (Fig. 4B). When 25 mM d-Ser was present in the medium, the presence of additional copies of the vanC-2 or vanXY_C-2 gene did not have an observable effect on the ability of E. casseliflavus to shorten the induction period further.

FIG. 4.

Effect of the presence of additional copies of vanT_C-2 (A) or 25 mM d-Ser (B) on the ability of E. casseliflavus to resume growth after the addition of vancomycin (2 μg/ml). The points at which vancomycin was added are indicated by the arrows.

Two further derivatives of pAT392 containing the genes encoding regulatory proteins VanR_C-2 and VanS_C-2 were constructed. The constitutive expression of additional copies of vanR_C-2 from the P₂ promoter, following its introduction into E. casseliflavus, led to a shift from inducible to constitutive expression of vancomycin resistance. This was also the case when additional copies of both vanR_C-2 and vanS_C-2 were introduced (data not shown).

DISCUSSION

E. casseliflavus possesses intrinsic low-level resistance to vancomycin which is mediated by the production of cell wall precursors terminating in d-Ser. The entire vanC-2 gene cluster of E. casseliflavus ATCC 25788 consisted of 5,692 bp and contained five open reading frames. The proteins encoded within the vanC-2 cluster all displayed a high degree of amino acid identity to the vancomycin resistance proteins of the vanC cluster of E. gallinarum BM4174. The evolutionary origin of the vanC-2 vancomycin resistance gene cluster remains unclear. The high degree of amino acid identity between the proteins encoded by the vanC-2 gene cluster and the vanC cluster of E. gallinarum BM4174 (71 to 91%), similar to that of internal portions of the Ddl ligases of these organisms (14), suggests a strong evolutionary link between the two. It is evident that there have been various degrees of divergence between the vancomycin resistance proteins, but all remain closely related. Enterococci of the VanE (15) and VanG (19) phenotypes mediate resistance to vancomycin via the production of d-Ser-terminating precursors, as does Clostridium innocuum (V. David, B. Bozodogan, J. L. Mainardi, R. Legrand, L. Gutmann, and R. Leclerq, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 680, 2000). However, the resistance proteins encoded by these organisms are not as closely related to those encoded by the vanC-2 and vanC gene clusters as these two clusters are to each other. No elements, such as insertion sequences associated with transposition events, have been identified around the vanC-2 or vanC gene clusters, indicating that it is unlikely that transposition has played a recent role in the acquisition of these clusters. It has been proposed that the vancomycin resistance in enterococci mediated by the inducible production of alternative cell wall precursors terminating in d-Lac may have originated in glycopeptide-producing bacteria, which also produce d-Lac-terminating precursors inducibly (18). However, no glycopeptide producers that mediate resistance by the production of d-Ser-terminating cell wall precursors have been identified, suggesting that this type of vancomycin resistance may have evolved from elsewhere.

E. casseliflavus ATCC 25788 displayed resistance to vancomycin that was inducible. Enterococci of the VanA and VanB phenotypes also normally demonstrate inducible resistance to vancomycin. These enterococci require an induction period of 30 to 60 min after the addition of vancomycin before resistance is expressed and d-Lac-terminating precursors are detected (28). However, E. casseliflavus displayed an unusually long induction period. It was only after this extended period of induction that normal, exponential growth resumed. This phenomenon was not limited to E. casseliflavus but has also been observed in E. gallinarum isolates which express vancomycin resistance inducibly (32). The reason behind the differences between the length of induction seen in the enterococci that possess the VanA and VanB phenotypes and the VanC-type enterococci may be explained by the different cell wall precursors that they synthesize. Enterococci which possess the VanA and VanB phenotypes produce alternative cell wall precursors terminating in d-Lac in the presence of vancomycin. Although the kinetic parameters of VanH, which reduces pyruvate to d-Lac, and the availability of the other cell wall synthesis enzymes may have a significant role to play, it seems unlikely that the availability of pyruvate would limit the rate of synthesis of the alternative cell wall precursor terminating in d-Lac. In E. gallinarum and E. casseliflavus, d-Ser is made available for cell wall synthesis through the activity of VanT and VanT_C-2. It is probable that as l-Ser is essential for many cellular processes, such as protein synthesis and one-carbon metabolism, it is rapidly utilized in the cell and is not as abundant as pyruvate. Furthermore, the level of serine racemase activity is relatively low, approximately 2.5% of that of alanine racemase (data not shown). These factors may therefore limit the rate at which d-Ser is synthesized, leading to an extended period of induction before the alternative d-Ser-terminating precursor is synthesized in sufficient amounts to overcome vancomycin inhibition.

It was evident from the different cell wall precursors extracted at different stages of the induction curve of E. casseliflavus that growth was not resumed until peptidoglycan precursors ending in d-Ser had been synthesized. It appeared that the d,d-carboxypeptidase activity of VanXY_C-2 played an important role in the early stages of induction, as the large amount of tetrapeptide produced suggested that there was insufficient VanX-type activity to prevent the formation of pentapeptide[d-Ala]. The proportions of tetrapeptide and pentapeptide[d-Ala] varied slightly during the first 120 min of the induction period, but no d-Ser-ending precursors were detected during this time. The first 120 min of induction corresponded to the period when growth was effectively inhibited. This was despite the fact that tetrapeptide is not bound to vancomycin and could participate in cell wall synthesis, although it could not function as a donor in transpeptidation reactions. The extent of cell wall cross-linking and cell integrity are therefore dependent on the availability of d-Ala- or d-Ser-ending precursors to act as donors in the transpeptidation reaction. It has been demonstrated that introduction of VanXY_C into a vancomycin-susceptible E. faecalis strain results in the production of large amounts of tetrapeptide but no increase in the MIC of vancomycin, indicating that this activity alone is not sufficient for vancomycin resistance (27). During the first 120 min of induction, binding of vancomycin to the d-Ala-terminating cell wall precursors attached to the lipid carrier at the external surface of the membrane was likely to be crucial for the inhibition of peptidoglycan synthesis, because formation of the drug-target complex is expected to sequester the lipid carrier and thereby prevent translocation of additional precursors (4). Therefore, further cell wall synthesis, including incorporation of d-Ser-ending precursors, would also be dependent on synthesis of new lipid carriers or the recycling of those sequestered in a complex with vancomycin.

It was evident from the introduction of additional copies of the structural genes, vanC-2, vanXY_C-2, and vanT_C-2, into E. casseliflavus that the availability of d-Ser played a crucial role in determining the length of time required for induction after the addition of vancomycin. Introduction of plasmid pIC1 (vanC-2) into E. casseliflavus did not lead to a significant decrease in the time required before normal growth resumed. This indicated that additional d-Ala-d-Ser ligase activity alone was not sufficient to reduce the period of time necessary for induction. The activity of the VanC-2 ligase would be expected to be limited by the availability of its substrates, d-Ala and d-Ser. The level of alanine racemase activity in E. casseliflavus was 40-fold greater than that of serine racemase activity; consequently, the availability of d-Ser was more likely to be limiting than that of d-Ala.

Increased amounts of VanXY_C-2 might be expected to reduce the length of time necessary for induction, but this was not observed when plasmid pIC2 (vanXY_C-2) was introduced into E. casseliflavus. It was therefore evident that additional VanXY_C-2 activity and subsequent removal of pentapeptide[d-Ala] did not result in faster alleviation of inhibition by vancomycin. This was supported by the analysis of cell wall precursors during induction, where it was apparent that there was substantial VanXY_C-2 activity in the initial stages of induction but where normal growth did not resume until the appearance of the d-Ser-ending precursor.

No significant difference in the length of induction of E. casseliflavus/pIC2 (vanXY_C-2) and E. casseliflavus/pAT392 was observed when both were grown in the presence of 25 mM d-Ser. This suggested that when d-Ser was abundant, natural competition between d-Ala-d-Ser and d-Ala-d-Ala dipeptides was sufficient to produce enough pentapeptide[d-Ser] to relieve inhibition by vancomycin and that additional VanXY_C-2 activities did not make a significant difference to the time needed for this to occur.

The importance of the availability of d-Ser was also evident in the construct E. casseliflavus/pIC4 (vanT_C-2), which resumed growth significantly faster than the control E. casseliflavus/pAT392 as the probable result of having greater serine racemase activity. Of the three enzymes essential for resistance, it therefore appeared that the activity of the serine racemase, VanT_C-2, was most important in terms of determining the length of time needed before normal growth was able to resume after the addition of vancomycin. Examination of the substrate specificity of the Ddl of E. casseliflavus demonstrated that in the presence of 25 mM d-Ser it was able to synthesize d-Ala-d-Ser, thus leading to the production of pentapeptide[d-Ser] without the expression of the vanC-2 gene cluster. Therefore, preformed pentapeptide[d-Ser] cell wall precursors may already have been present at the point when vancomycin was added, reducing the time required for induction, although it was likely that the concentration of d-Ser would have been much lower than 25 mM, the concentration used in supplementing the medium. It is therefore apparent that whether additional d-Ser is provided by VanT_C-2 activity or by supplementation of the growth medium, its availability was crucial in determining the length of time required by E. casseliflavus to resume growth after inhibition by vancomycin.