Biochemical and Biophysical Characterization of an Unexpected Bacteriolytic Activity of VanX, a Member of the Vancomycin-resistance vanA Gene Cluster

Background: VanX belongs to the vanA gene cluster, which confers vancomycin resistance by protecting the bacterial cell wall.

Results: Bacteriolysis occurred when isolated VanX was expressed in E. coli, which originated from the dipeptidase activity of VanX.

Conclusion: This is the first direct characterization of VanX-mediated bacteriolysis.

Significance: This finding suggests a new class of therapeutic agents taking advantage of VanX-mediated bacteriolysis.

Abstract

VanX is a d-alanyl-d-alanine (d-Ala–d-Ala) dipeptidase encoded in the vancomycin-resistance vanA gene cluster. Here we report that strong bacteriolysis occurred when isolated VanX was expressed in Escherichia coli at temperatures lower than 30 °C, which was unexpected because the vanA operon confers vancomycin resistance by protecting the cell wall. Therefore, we monitored cell lysis by measuring sample turbidity with absorbance at 590 nm and VanX expression using SDS-PAGE. No cell lysis was observed when VanX was expressed, even in large quantities, in the cell inclusion bodies at 37 °C, suggesting that a natively folded VanX is required for lysis. In addition, VanX mutants with suppressed dipeptidase activity did not lyse E. coli cells, confirming that bacteriolysis originated from the dipeptidase activity of VanX. We also observed shape changes in E. coli cells undergoing VanX-mediated lysis with optical microscopy and classified these changes into three classes: bursting, deformation, and leaking fluid. Optical microscopic image analysis fully corroborated our interpretation of the turbidity changes in the samples. From a practical perspective, the finding that VanX expressed in isolation induces cell lysis suggests that inhibitors of VanA and VanH that act downstream from VanX could provide a new class of therapeutic chemicals against bacteria expressing the vancomycin-resistance gene cluster.

Introduction

Cell lysis can occur through various distinct molecular processes and mechanisms, each involving different factors (1). A well known enzyme that induces cell lysis is lysozyme (2), which hydrolyzes the 1,4-β linkage between N-acetylmuramic acid and N-acetylglucosamine in the peptidoglycan layer (3). In Gram-negative bacteria, such as Escherichia coli, the external membrane layer largely prevents the access of lysozyme to the peptidoglycan layer (4), reducing its lytic activity. An artificial cell lysis process requiring the coexpression of several enzymes has also been reported (5) in which holin, a membrane pore-forming protein, and several endolysins (amidase, muramidase, transglycosylase, and endopeptidase), which are muralytic enzymes originating from bacteriophages (6, 7), attack already-synthesized peptidoglycan layers.

Vancomycin is an antibiotic that inhibits the synthesis of the bacterial cell wall by binding to the d-alanyl–d-alanine (d-Ala–d-Ala)³ termini of the peptidoglycan monomers, preventing their cross-linkage and polymerization into the cell wall (Fig. 1A) (8, 9). Resistance to vancomycin is conferred by the combined action of several enzymes encoded in the vanA operon (Fig. 1B) (10). vanA operon identifies a gene cluster encoding five enzymes and is synonymous to the vancomycin-resistance gene cluster. First, VanX, a d-Ala–d-Ala dipeptidase, hydrolyzes d-Ala–d-Ala to two d-Ala monomers, and VanH produces d-lactate by the hydrolysis of pyruvate. The VanA enzyme, one of the five enzymes encoded in the vanA gene cluster, then synthesizes a peptide bond between d-alanine and d-lactate to yield d-Ala–d-Lac. d-Ala–d-Lac binds to the C-terminus of the peptide chain of the peptidoglycan monomer and allows it to cross-link, inducing the biosynthesis of altered cell walls made of peptide chains that terminate in d-Ala–d-Lac instead of the normal d-Ala–d-Ala. Cell-wall synthesis with peptidoglycan terminating with d-Ala–d-Lac is not inhibited by vancomycin, which can bind d-Ala–d-Ala but not d-Ala–d-Lac.

FIGURE 1.

Schematics of bacteriolysis by vancomycin and VanX. A, cell lysis by vancomycin. Peptidoglycan (PG) monomers are made of N-acetylmuramic acid (black bar) and N-acetylglucosamine (gray bar) polymerized with a peptidoglycan peptide chain (stripe and white circles), which cross-links with other peptide chains to synthesize the cell wall. Cell wall synthesis is blocked by vancomycin binding to the d-Ala-d-Ala (two white circles) termini of the peptide chain inhibiting the bridging of the peptidoglycan monomers. B, resistance to vancomycin is conferred by the VanA gene cluster. Several enzymes act in concert as follows; VanX hydrolyzes d-Ala-d-Ala dipeptide into d-Ala monomers, and VanH dehydrogenates pyruvate (black circle) to d-Lactate (gray circle). d-Ala and d-Lac are synthesized into d-Ala-d-Lac by VanA. d-Ala-d-Lac is then attached to the peptide chain of peptidoglycan. Vancomycin cannot recognize d-Ala-d-Lac, and the peptidoglycan can bridge with each other synthesizing an altered but complete cell wall. C, in the absence of VanH and VanA, no d-Ala dipeptide is available for attachment to the peptide chain as all d-Ala-d-Ala are hydrolyzed by VanX into d-Ala monomers. Peptidoglycan layer with incomplete peptide cannot cross-link causing cell lysis.

We recently noticed that VanX (23.3 kDa; 202 residues) exhibits strong lytic activity when expressed in isolation in E. coli, and we previously used this phenomenon to extract recombinant proteins expressed in E. coli (11). The bacteriolytic activity of VanX was unexpected because it is usually involved in protecting bacterial cell walls and thus has a positive effect on bacterial survival. We were thus interested in clarifying the lysis mechanism which negatively influences bacterial survival. Here, we report a biochemical and biophysical analysis of VanX-mediated lysis. We show that the bacteriolysis caused by the expression of VanX in isolation from the other genes of the vanA operon is attributable to the hydrolysis of the peptidoglycan component d-Ala–d-Ala before its attachment to the peptide termini (8, 12). This molecular mechanism differs from all previously reported enzymatic lysis mechanisms. From a practical perspective, these findings suggest that inhibitors of VanA could provide a new class of therapeutic chemicals against bacteria acquiring antibiotic resistance via the vanA gene cluster (13).

EXPERIMENTAL PROCEDURES

Materials

Nickel-nitrilotriacetic acid Superflow column was purchased from Qiagen, benzamidine-agarose 6B resin was from GE Healthcare, and ninhydrin, d-alanine, d-alanyl–d-alanine, and all other chemicals were purchased from Wako Pure Chemical Industries Ltd or Nacalai Tesque Inc.

Construction of VanX Mutants

VanX gene was inserted at a NdeI and BamHI of pET11 or pET15b plasmid (pET-11-VanX and pET-15-VanX hereafter). In addition, pET15b expression vectors of inactivated VanX mutants were constructed by replacing residues at the catalytic site (Arg-71 and Glu-181) and the zinc ion binding site (His-116, Asp-123, and His-184) (Lessard and Walsh (10)) by an alanine using the site-directed mutagenesis strategy similarly to our previous studies (14, 15).

Assessment of VanX Expression and Turbidimetric Measurements

E. coli cells (BL21(DE3), JM109(DE3), and AD494(DE3)) were transformed with the pET15-VanX expression vector and spread on LB agar plates. Colonies grown overnight onto the plates were suspended with LB medium containing ampicillin at a concentration of 50 μg/ml. 5 ml of LB/Amp cultures were inoculated with the suspension to a concentration of A_{590 nm} (optical absorbance at 590 nm) = 0.1, and they were grown at 37 °C with shaking at 250 rpm and cooled on ice for 30 min when the A_{590 nm} reached 0.5–0.8. Protein expression was induced by adding IPTG to a final concentration of 1 mm and incubating the culture at the specified temperature (usually 20 °C, 25 °C, 30 °C, and 37 °C). The sample turbidity was assessed at regularly timed intervals by measuring the A_{590 nm}.

The VanX expression assessment by SDS-PAGE was carried out by collecting the culture, which was centrifuged at 20,100 × g for 15 min at 4 °C to separate the cell culture supernatant fraction from the cell pellet. The cell pellets were homogenized with glass beads beating and separated into intracellular soluble fraction and intracellular insoluble fraction. The proteins were concentrated by acetone precipitation, and their concentrations were adjusted so that the amount of proteins loaded onto the SDS-PAGE gels corresponded to an initial fraction volume of 50 μl.

Large Scale Expression and Purification of VanX and Its Mutants

Large scale expression of VanX was performed using the corresponding pET15-VanX vectors with BL21 (DE3) E. coli as a host cell. The cells were grown at 37 °C in 1 liter of LB medium containing ampicillin (50 μg/ml), and VanX expression was induced when A_{590 nm} reached 0.5–0.8 by the addition of IPTG at a final concentration of 1 mm. After induction, the cells were grown for 3 h at 25 °C or 37 °C. A small amount of DNase was added to the culture of wild type VanX expressed at 25 °C 1 h before harvest due to the culture viscosity arising from the leaked E. coli genome DNA. The culture media were centrifuged at 8500 × g for 10 min at 4 °C, and the culture supernatant of the wild type VanX expressed at 25 °C was harvested. At 37 °C, the wild type VanX as well as the inactive mutants expressed in the inclusion body. The cell pellets were resuspended in 50 ml of Lysis buffer (50 mm NaH₂PO₄, pH 8.0, 300 mm NaCl, 10 mm imidazole, 1% Triton X-100, and 0.5 mm DTT) and homogenized by glass beads. The suspension was separated into the supernatant and the pellet by centrifugation at 8500 × g for 30 min at 4 °C, and the supernatant was filtered through a 0.45-μm pore filter. Then the supernatant was loaded onto a nickel-nitrilotriacetic acid Superflow column, and the column was washed with a 10× column volume of wash buffer (50 mm NaH₂PO₄, pH 8.0, 300 mm NaCl, 20 mm imidazole, and 0.5 mm DTT). VanX and its mutants were eluted with 10 ml of elute buffer (50 mm NaH₂PO₄, pH 8.0, 300 mm NaCl, 250 mm imidazole, and 0.5 mm DTT). The His tag was cleaved overnight with thrombin with dialysis at 25 °C against 50 mm NaH₂PO₄, pH 8.0, 100 mm NaCl, and 0.5 mm DTT. The removal of the His tag was assessed by SDS-PAGE chromatography.

Secondary Structure Content by Circular Dichroism (CD)

CD measurements were performed using a JASCO-J820 spectrophotometer essentially in line with our previous reports (16). In short, the proteins were dialyzed against 5 mm NaH₂PO₄, pH 7.0, 100 mm NaCl. CD spectra were measured at a protein concentration of 24 μm in a 0.1-cm optical path length quartz cuvette at 15 °C, 25 °C, 37 °C, and 45 °C with a wavelength range of 190–250 nm and a scanning rate of 50 nm/min. Spectra were averaged over 10 scans.

Measurement of d-Ala-d-Ala Dipeptidase Activity

The dipeptidase activity was determined quantitatively by measuring the release of free amino acid with the modified cadmium-ninhydrin method (17). A cadmium-ninhydrin stock solution was prepared by dissolving 1 g of ninhydrin in 100 ml of ethanol followed by the addition of 12.5 ml of acetic acid and 1.25 g of CdCl2 dissolved in 12.5 ml of H₂O. The color was developed by incubating the diluted hydrolyzed product with 0.75 ml of Cd-ninhydrin stock solution at 85 °C for 5 min. The absorbance was measured at 505 nm and quantified with free amino acids as a standard.

Enzymatic Activity Measurement of VanX and Lysozyme

The enzymatic activity of VanX and lysozyme were compared by measuring the cell wall degradation of Micrococcus luteus as described in Okanojo et al. (18). VanX and lysozyme solution were prepared to 0.4 mg/ml final concentration in 50 mm phosphate buffer, pH 7.0, 100 mm NaCl, and 0.5 mm DTT and 50 mm phosphate buffer, pH 6.5, respectively. 1.5 ml of M. luteus resuspended at a 0.5 mg/ml concentration was mixed with 20 μl of enzyme solution. The lytic activity was assessed using the decrease of light scattering at A_{600 nm} for 300 s, measured every 10 s.

Optical Microscopy

BL21(DE3) E. coli cells transformed with pET15-VanX were grown at 37 °C with 250 rpm shaking until the A_{590 nm} reached 0.5–0.6 at which time the IPTG was added (no IPTG was added for negative control samples). The cells were further incubated at 25 °C for 50 min, and 50-μl aliquots of the sample were held between a slide glass and a coverslip and placed on the stage of an inverted optical microscope (IX-70; Olympus; Tokyo, Japan) equipped with a 40× objective lens (Uapo40×/340, NA = 0.9; Olympus) (19). Ten minutes after placement, images were photographed with a cooled CCD camera (Sensicam QE, PCO AG, Kelheim, Germany) every 2 s, and 300 images were collected over 10 min. The changes of cell shapes were quantified using ImageJ (NIH free software).

Electron Microscopy Measurements

E. coli cells, transformed using pET-11VanX or its inactive mutant pET-11 VanX-R71A, were harvested by centrifugation at 1200 rpm for 10 min. The precipitates were incubated in 2.5% glutaraldehyde in 0.1 m phosphate buffer (PBS) at pH 7.0 for 1–2 h, washed three times with PBS, and treated with 1.0% osmium tetroxide for 1 h. They were embedded in 4% agarose gel, which was sliced into 2 × 2-mm blocks. The blocks were dehydrated by shaking in ethanol-containing baths for 20 min. The ethanol concentration was gradually raised from 50% (v/v) to 70, 80, 95, and 100%, and the blocks were rinsed 3 times in 100% ethanol. The dehydrated blocks were washed twice by shaking in propylene oxide for 20 min. These blocks were immersed in a liquid containing epoxy resin (TAAB Co. Ltd.) and propylene oxide for 1 h, where the volume ratio of the resin was 50%, 67%, and 100%, and the blocks were immersed 3 times in 100% epoxy resin for 1 h. The samples blocks were completely embedded in 100% resin and kept for 24 h at 45 °C in a temperature bath and were further kept for 48 h at 60 °C. As a trimming operation, ultra-thin sections were cut by an ultramicrotome (ULTRACUT UC7, Leica Co. Ltd.), and were collected on 150-mesh copper grids coated with ultra-thin carbon films. We stained the samples according to a recently developed double-staining method that uses 2.5% gadolinium triacetate for 10 min and 1.0% lead citrate for 2-min intervals. This method yields pictures with contrasts higher than the traditional gadolinium triacetate staining method (20). The sections were observed on a JEOL transmission electron microscope JEM-2100 equipped with a high contrast pole piece. The accelerating voltage was 200 kV, and all images were acquired with a high resolution CCD camera (Ultra Scan, with 2000 × 2000 pixel size, Gatan Co. Ltd.).

RESULTS

Cell Turbidity and Bacteriolysis by VanX Expressed in E. coli

The first clue to the bacteriolytic activity of VanX came from an unanticipated reduction in sample turbidity when the expression temperature was reduced from 37 °C to 25 °C as we attempted to express VanX in the soluble fraction (Fig. 2A). This was confirmed by measuring the turbidity of the cell culture using the optical absorbance at 590 nm (Fig. 2, B and C). SDS-PAGE analysis indicated that all of the VanX were expressed in the insoluble fraction at 37 °C, but when expressed at 20 or 25 °C they were released into the culture medium together with the endogenous E. coli proteins (Fig. 2D).

FIGURE 2.

Turbidity E. coli cultures and expression of VanX. A, cell culture turbidity assessed using the scattering of a laser beam. The turbidity change is readily seen with the naked eye, but scattered laser beam provides a more stringent view for photographs. E. coli was transformed with an expression vector holding either a VanX gene or no gene (for control), and IPTG was added at 25 °C and 37 °C. B, growth curve of BL21(DE3) E. coli cells. The vertical axis indicates the absorbance at 590 nm, and the horizontal axis indicates the culture time measured from the time at which protein expression was initiated. The temperature was maintained at 37 °C during cell growth and decreased upon starting VanX expression to 20 °C (open diamonds), 25 °C (black diamonds), 30 °C (gray triangles), or 37 °C (gray squares). C, growth curve of E. coli cells from different strains, BL21(DE3), JM109(DE3), AD494(DE3), containing a VanX expression vector are shown with, respectively, filled circles, triangles, and squares. The respective control samples transformed with a vector containing no gene are shown with open triangles, circles, and squares. The temperature was maintained at 37 °C during cell growth and decreased upon starting VanX expression to 25 °C. D, localization of VanX protein by SDS-PAGE analysis. The extracellular culture (C), the intracellular soluble fraction (S), and the intracellular insoluble fraction (P) were sampled 2 h after starting VanX expression at the indicated temperatures. Lane C for expression temperatures between 20 °C and 30 °C shows that endogenous E. coli proteins together with VanX are leaked in the culture fraction. M, mass marker.

These observations are consistent with the hypothesis that the reduction in turbidity at low temperatures can be attributed to bacteriolysis caused by the expression of natively folded VanX in the soluble fraction, whereas VanX expressed at 37 °C in the cell inclusion bodies is inactive and cannot lyse the E. coli cells. We observed the same phenomenon with other E. coli strains (BL21[DE3], JM109[DE3], and AD494[DE3]), indicating the generality of the bacteriolytic activity of VanX (Fig. 2C).

Bacteriolysis Relates to the Dipeptidase Activity of VanX

To confirm the link between dipeptidase activity and bacteriolysis, we constructed five inactive VanX mutants and examined their lytic activities, structures, and in vitro dipeptidase activities. The inactive mutants were constructed by replacing Arg-71 and Glu-181 at the active site and His-116, Asp-123, and His-184 at the zinc ion-binding site of VanX with alanine (10). Unlike wild type (wt) VanX, none of the inactive mutants reduced the turbidity of the cultures even when expressed in the soluble fraction at 25 °C strongly suggesting that the bacteriolysis originated from the dipeptidase activity of VanX.

To assess the structure and in vitro dipeptidase activity of the VanX variants, we expressed the wild type and inactive VanX mutants at 25 °C and purified them from both the cell culture media and the intracellular soluble fractions. We also purified wtVanX expressed at 37 °C from the intracellular insoluble fractions and refolded it. Both the secondary structure content and the dipeptidase activity of wtVanX expressed at 25 °C were almost identical to those of wtVanX expressed at 37 °C and refolded from the inclusion bodies (Figs. 3 and 4, A and B) indicating that both are natively folded in a functional form. Furthermore, the inactive mutants expressed at 25 °C in the intracellular soluble fraction were fully folded according to circular dichroism measurements but had no dipeptidase activity, as expected (Fig. 4B). Taken together, these results again show that VanX-mediated bacteriolysis is associated with the enzyme dipeptidase activity, which requires that VanX be natively folded in a functional and soluble form.

FIGURE 3.

Correlation between VanX dipeptidase activity and bacteriolysis. A, growth curve of BL21(DE3) cells expressing the wild type VanX (gray diamonds) and inactive R71A (black squares), E181A (white squares), H116A (black triangles), D123A (white triangles), and H184A (white diamonds) VanX mutants at 25 °C. Cells were grown at 37 °C, and the temperature was decreased to 25 °C upon starting VanX expression. B, localization of VanX by SDS-PAGE analysis. C, S, and P indicate, respectively, the extracellular culture, intracellular soluble fraction, and intracellular insoluble fraction. The fractions were obtained after the VanX variants were expressed for 4 h at 25 °C. M, mass markers.

FIGURE 4.

Activity and secondary structure content of the wt VanX and its mutants. A, d-Ala-d-Ala dipeptidase activity of the wild type VanX and inactivated mutants. The wt VanX was expressed at 25 °C and purified from the extracellular culture fraction (black circles) or expressed at 37 °C and purified from the intracellular soluble fraction (gray circles). The inactivated mutants were expressed at 25 °C and purified from the intracellular soluble fraction. Symbols for the inactivated mutants are the same as in Fig. 3. The dipeptidase activity was measured at a protein concentration of 40 μm in 50 mm phosphate buffer, pH 7.0, containing 100 mm NaCl and 0.5 mm DTT. B, circular dichroism spectra of the wild type VanX and inactivated mutants VanX. The final protein concentration was 24 μm, and the buffer condition was the same as for the activity measurement. Symbols are the same as in A. C, enzymatic bacteriolytic activity of VanX. Lysozyme, WT25, and WT37 are shown, respectively, by gray squares, black circles, and gray triangles.

Finally, we measured the enzymatic activity of VanX using M. luteus and showed that VanX does not hydrolyze completed cell walls (Fig. 4C). Therefore, the bacteriolysis mechanism of VanX differs from that of lysozyme and the muralytic enzymes (21, 22), which act on the synthesized peptidoglycan cell wall. However, VanX acts before cell wall synthesis during peptidoglycan elongation.

Single-cell Observations with Optical Microscopy

In this section we report the morphological changes induced in E. coli cells by VanX expression, observed with optical microscopy. An aliquot of E. coli cells incubated at 25 °C for 50 min after the addition of IPTG, which is the time period immediately preceding the observed reduction in turbidity (Fig. 2, B and C), was placed on a slide glass. After the slides were allowed to stand for 10 min we collected images every 2 s for 10 min. We identified 848 E. coli cells containing pET11 VanX and analyzed the shapes of 601 cells in detail. The controls were E. coli cells without the VanX plasmid, and we observed 305 of 393 cells in detail. No morphological changes were observed in 90.8% of the control E. coli cells lacking the VanX-expressing plasmid, and all of the remaining 28 cells (9.2%) showed morphological changes associated with cell division. In E. coli transformed with the VanX-expressing plasmid, morphological changes associated with cell division or bacteriolysis were observed in 4 (0.7%) or 106 cells (17.6%), respectively (Fig. 5A and Table 1). For the purpose of discussion, we computed the relative areas of individual E. coli cells by calculating the area at the time of normalization by that at the beginning of the observation period (Fig. 5, B and C). In the control sample, the cell areas of the dividing cells were 25–75% that of the initial cell area, and the areas of the non-dividing cells ranged from 75% to 135%, both with distributions close to ideal normal distributions. In samples of cells containing pET11 VanX, the areas of the dividing E. coli cells ranged from 25 to 65% and the areas of the non-dividing cells were 65–125%, which were similar to the distribution pattern of the control samples not expressing VanX. In contrast, the areas of cells undergoing bacteriolysis were distributed over a wider range of 25–125%, reflecting the large variation in shape changes.

FIGURE 5.

Optical microscopy observation of VanX-expressing E. coli. A, relative frequency of E. coli in different status. Fraction of cells without morphological change are white. Gray and black indicate, respectively, the fraction of dividing cells and those undergoing bacteriolysis (classified as deformation, bursting, and fluid leakage in Table 1). B, histogram of the relative area change of E. coli cells that do not hold a VanX expression vector. The cells were incubated at 25 °C during observation, and the area changes were measured for 10 min. The same color coding as in A was used. C, histogram of the relative area change of E. coli cells holding the VanX expression vector. D, optical microscopy images of cells with and without expression of VanX. The image size is 8.75 μm × 8.75 μm, and the time is measured starting from the first image.

TABLE 1.

Numbers of cell for each shape change class

Cell Shape	Without VanX	With VanX
No change	277	491
Cell division	28	4
Bursting	0	50
Deformation	0	35
Fluid leakage	0	21
Total	305	601

The shape changes of cells undergoing lysis displayed distinct patterns, which we classified into three types based on visual inspection: bursting, deformed, and leaking fluid (Fig. 5D, Table 1). Bursting cells appeared as knots which burst, allowing the leakage of the cytoplasm, after which the entire cell gradually faded. The early changes during deformation were similar to those that occurred during a knot bursting, but the knot bent the cell body without any reduction in the intensity of the cell image, suggesting that there was no leakage from these E. coli cells. In the cells undergoing fluid leakage, the cytoplasmic contents leaked vigorously from the centers of the cells, and the intensity of the whole cell gradually faded.

Cell Morphology Analyzed with Electron Microscopy

We further characterized VanX-mediated cell lysis by observing the morphological changes that occurred in E. coli cells with transmission electron microscopy (Fig. 6). To this end we collected the cells 1 h after VanX expression was induced with IPTG, when the cell density started to decline with increasing cells lysis. E. coli cells expressing the inactive VanX R71A showed clear borders, which were deemed to be the cell walls. In contrast, the borders of many E. coli cells expressing active VanX were smeared, especially in their central regions. The slight constriction (Fig. 6D) at the cell center suggests that the wall “melts” as the cell undergoes cell division. This is consistent with the assumption that overexpressed VanX hydrolyzes d-Ala–d-Ala in the cytosol, which produces defective peptidoglycan lacking the d-Ala–d-Ala extension and thus impedes cell wall growth. The staining of the cells expressing VanX was also brighter than that of cells expressing inactive VanX, which might reflect a thinner cell wall. Overall, the transmission electron microscopy images show the lack of a cell wall in VanX-expressing E. coli, which was not observed in samples where VanX was not expressed.

FIGURE 6.

Transmission electron microscopy images of E. coli cells with and without expression of active VanX. E. coli cells were collected 1 h after IPTG induction. A and C show E. coli that expresses inactive VanX with a magnification of ×5000 and ×20000, respectively. C is an expanded image of the region indicated by a black square in A. The cell wall shape is very regular, and a dark gray band at the cell border, probably corresponding to the cell wall, is clearly visible. The E. coli center shows a constriction presumably attributable to a starting cell division. B and D show E. coli that express active VanX with a magnification of ×6000 and ×20000, respectively. D, expansion of the region delineated by a black square on B. Similar to panel C, a constriction is visible near the center of the E. coli, but the cell borders appear rough, which presumably reflects a melted and dislocated cell wall.

The lysis mechanism emerging from the transmission electron microscopy images and the biochemical and biophysical analyses discussed above is that intracellular VanX hydrolyzes the peptidoglycan elongation component d-Ala–d-Ala in the cytoplasm. Incomplete peptidoglycan inhibits the growth of the cell wall at the elongation site during cell division, leading to an incomplete and thinner cell wall that is conceivably susceptible to breakage by osmotic shock, thus resulting in the release of intracellular substances into the culture medium.

DISCUSSION

The present study demonstrates that VanX, a member of the vanA operon, well known to protect bacteria from vancomycin, lyses bacterial cells when expressed in isolation and in large quantities. The melting of E. coli cells expressing VanX was observed with optical microscopy, and we could classify the lysis patterns into three characteristic classes. The conclusions drawn from the optical microscopy analysis are consistent with the bulk VanX-mediated bacteriolysis monitored with turbidimetry.

Overall, this study exemplifies the multifaceted aspects of an enzyme biological functions or roles and how adaptive nature can be in using them. The ability of VanX to induce bacteriolysis by inhibiting cell wall synthesis is particularly surprising given its natural role in protecting the cell wall. From a more practical perspective, the present findings strongly suggest that inhibitors of VanA, such as those recently reported by Sova et al. (23), could provide a new class of therapeutic chemicals against vancomycin-resistant bacteria.