Purification and characterization of the bacterial MraY translocase catalyzing the first membrane step of peptidoglycan biosynthesis.

The MraY translocase catalyzes the first membrane step of bacterial cell wall peptidoglycan synthesis (i.e. the transfer of the phospho-N-acetylmuramoyl-pentapeptide motif onto the undecaprenyl phosphate carrier lipid), a reversible reaction yielding undecaprenylpyrophosphoryl-N-acetylmuramoyl-pentapeptide (lipid intermediate I). This essential integral membrane protein, which is considered as a very promising target for the search of new antibacterial compounds, has thus far been clearly underexploited due to its intrinsic refractory nature to overexpression and purification. We here report conditions for the high level overproduction and for the first time the purification to homogeneity of milligram quantities of MraY protein. The kinetic parameters and effects of pH, salts, cations, and detergents on enzyme activity are described, taking the Bacillus subtilis MraY translocase as a model.

The growing emergence of multiresistance of pathogenic bacteria to currently used antibiotics requires the development of new therapeutic compounds and the identification and exploitation of novel targets (1). The enzymes of the pathway for cell wall peptidoglycan biosynthesis that are essential for bacterial growth and specific to eubacteria constitute such a set of interesting potential targets that should be explored in detail. Indeed, peptidoglycan, the heteropolymeric mesh of the bacterial cell wall, plays a critical role in protecting bacteria against osmotic lysis. It is also responsible for the maintenance of a defined cell shape and is intimately involved in the cell division process (2). The peptidoglycan monomer unit, N-acetylglucosaminyl-␤-1,4-N-acetylmuramoyl-pentapeptide (GlcNAc-MurNAc 1 -pentapeptide), is synthesized by enzymes located in the cytoplasm or at the inner side of the cytoplasmic membrane, and its polymerization, occurring at the outer side of the cytoplasmic membrane, is catalyzed by the penicillin-binding proteins, the targets of the ␤-lactam antibiotics (3). This implies the passage of the monomer unit from the cytoplasm to the periplasm through the hydrophobic environment of the membrane, a process involving the transfer of this hydrophilic unit onto a lipid carrier, undecaprenyl phosphate (C 55 -P) (3,4). The first membrane step is the transfer of the phospho-MurNAc-pentapeptide moiety onto C 55 -P, yielding C 55 -PP-MurNAc-pentapeptide (lipid I), a reaction catalyzed by the MraY enzyme (Scheme 1). Since this reaction consists of the translocation of a peptidoglycan precursor moiety from the cytoplasm to the membrane, the latter enzyme has been named MraY "translocase" (5). This reaction is reversible, but in vivo it is drawn by coupling to the subsequent reaction catalyzed by the MurG transferase (5)(6)(7)(8). The MraY translocase is an integral membrane protein whose topology has been recently determined (9); it is composed of 10 transmembrane segments, five cytoplasmic domains, and six periplasmic domains, including the N-and C-terminal ends. The latter model has been established with MraY proteins from Escherichia coli and Staphylococcus aureus and thus appears to be conserved in both Gram-negative and Gram-positive bacteria. Alignment of bacterial MraY sequences shows that the five cytoplasmic domains contain many highly conserved amino acid residues. The presence of the MraY translocase exclusively in bacteria, the fact that it is essential for viability (which has been demonstrated in E. coli and Streptococcus pneumoniae (10,11)), its accessibility from the periplasmic space, and the recent identification of some natural inhibitors explain the renewed interest for this target. MraY is inhibited by nonclinically used antibiotics such as tunicamycin, amphomycin, mureidomycin, liposidomycin, and muraymycin (12)(13)(14)(15). Recently, simplified analogues of liposidomycin, named riburamycins, have been shown to be powerful MraY inhibitors and to possess antibacterial activities against Gram-positive organisms (16,17). Moreover, this protein has been shown to be the target of the lytic protein LysE of phage X174 (18). However, all studies on MraY reported to date only involved crude membrane preparations as the source of this bacterial enzyme. We here describe the significant overexpression and for the first time the purification to homogeneity of the MraY enzyme as well as detailed investigations of its biochemical properties.

EXPERIMENTAL PROCEDURES
Chemicals N-Lauroyl sarcosine was purchased from USB Corporation; Triton X-100, n-octyl-␤-D-glucopyranoside, tunicamycin, UDP-GlcNAc, and ATP were from Sigma; CHAPS was from ICN; n-dodecyl-␤-D-maltoside (DDM) was from Fluka; Tween 20 was from VWR; and isopropyl-␤-Dthiogalactopyranoside (IPTG) was from Eurogentec. C 55 -P was provided by the Institute of Biochemistry and Biophysics of the Polish Academy of Sciences. UDP-MurNAc was prepared as described by Blanot et al.
* This work was supported by Centre National de la Recherche Scientifique UMR 8619. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Bacterial Strains and Growth Conditions
The E. coli strains DH5␣ (Invitrogen), BL21(DE3) (Promega), and C43(DE3) (Avidis) were used as hosts for plasmids and for the overproduction of the MraY enzyme. 2YT (25) was used as a rich medium, and growth was monitored at 600 nm with a Shimadzu UV-1601 spectrophotometer. For strains carrying drug resistance genes, ampicillin and kanamycin were used at 100 and 60 g⅐ml Ϫ1 , respectively.

Plasmid Constructions
Standard procedures for molecular cloning (26) and cell transformation (27) were used. The Bacillus subtilis mraY gene was amplified from strain 168 chromosome using BS1 and BS2 oligonucleotides as PCR primers (see Table I) and the Expand high fidelity polymerase system (Roche Applied Science). The PCR fragment was cut by BspLU11I and BglII and inserted into the pTrc99A (Amersham Biosciences) and pET28b plasmid vectors (Novagen) opened by compatible sites NcoI and BamHI, generating plasmids pTrcYBS62 and pETYBS62, respectively. In these constructs, the mraY gene was expressed under control of a strong IPTG-inducible promoter, and the encoded MraY protein carried an N-terminal His 6 extension. The same procedure was used to clone the mraY genes from E. coli, S. aureus, and Thermotoga maritima, using oligonucleotides EC1 and EC2, SA1 and SA2, and TM1 and TM2, respectively (Table I) and the chromosomal DNA of the concerned bacteria as a template. The resulting plasmids were named pETYEC121 and pTrcYEC121 for E. coli, pETYSA21 and pTrcYSA21 for S. aureus, and pETYTM85 and pTrcYTM85 for T. maritima. These different constructions were verified by DNA sequencing (MWG-Biotech).

Preparation of Crude Enzyme
E. coli strain C43(DE3) harboring recombinant plasmid pETYBS62 was grown at 37°C in 2YT-kanamycin medium (5-liter culture). At an A 600 of ϳ0.7, IPTG was added at a final concentration of 1 mM, and incubation was continued for 16 h at 25°C with shaking. Cells were harvested by centrifugation (8,000 ϫ g for 20 min at 4°C), washed in 100 ml of 25 mM Tris-HCl, pH 7.5, and resuspended in 10 ml of the same buffer containing 2 mM 2-mercaptoethanol, 150 mM NaCl, 30% glycerol, and 1 mM MgCl 2 (buffer A). Bacteria were broken by sonication (Bioblock Vibracell sonicator model 72412). The resulting suspension was centrifuged at 200,000 ϫ g for 30 min at 4°C in a Beckman TL100 centrifuge. The pellet consisting of membranes and associated proteins (14 g wet weight, 1.2 g of proteins) was washed three times with buffer A and then subjected to solubilization by detergents as described below.

Solubilization of MraY
Membrane vesicles containing the overexpressed MraY protein were resuspended in 20 ml of buffer A. DDM was added at a final concentration of 17.8 mM, and the mixture was incubated at 4°C for 2 h under shaking. After centrifugation (200,000 ϫ g, 30 min at 4°C), a first supernatant (DM1) was recovered. The insoluble material was then subjected to a new cycle of solubilization in buffer A containing 21.5 mM DDM. Supernatant DM2 was recovered after centrifugation. Two further rounds of solubilization/centrifugation were performed in the same conditions (21.5 mM DDM), generating supernatants DM3 and DM4, respectively.
A similar procedure was used for extraction with other detergents. In that case, two successive treatments were performed, and the final concentration of detergent was 68, 41, 32, and 27.3 mM for n-octyl-␤-Dglucopyranoside, CHAPS, Triton X-100, and N-lauroyl sarcosine, respectively.

Purification of Histidine-tagged MraY
Solubilized membrane proteins were mixed and incubated for 2 h at 4°C with Ni 2ϩ -NTA-agarose (Qiagen) (15 mg of proteins/ml of resin) pre-equilibrated in buffer B (20 mM sodium phosphate, pH 7.2, 300 mM NaCl, 30% glycerol, 3.9 mM DDM, 2 mM 2-mercaptoethanol). After incubation, the resin was transferred to an Econo-Pac chromatography column (Bio-Rad) and washed first with 5 column volumes of buffer B. Further washings and protein elution were performed with increasing concentrations of imidazole, from 5 to 300 mM, in buffer B. After assays for translocase activity and SDS-PAGE analysis, pure MraY proteincontaining fractions were freed from imidazole and concentrated using a Vivaspin concentrator (Vivascience) in 30 mM Tris-HCl buffer, pH 7.5, containing 150 mM NaCl, 10% glycerol, and 3.9 mM DDM.

Protein Monitoring
Protein concentrations were determined using the QuantiProBCA assay kit (Sigma) and bovine serum albumin as the standard and/or by quantitative amino acid analysis with a Hitachi model L8800 analyzer (ScienceTec) after hydrolysis of samples in 6 M HCl for 24 h at 105°C.

Enzymatic Synthesis of Unlabeled and Radiolabeled UDP-MurNAc-pentapeptide
The reaction mixtures contained 100 mM Tris-HCl, pH 8.6, 2 mM dithiothreitol, 30 mM MgCl 2 , 1 mM UDP-MurNAc, 20 mM ATP, 1.2 mM L-Ala (or 1 mM L-[ 14 C]Ala), 1.2 mM each D-Glu, meso-A 2 pm, and D-Ala-D-Ala, and 200 units each of enzymes MurC, MurD, MurE, and MurF. After 2 h at 37°C, the formation of UDP-MurNAc-pentapeptide was followed by analytical HPLC on a column of Nucleosil 5C 18 (250 ϫ 4.6 mm; Alltech France) using elution with 50 mM ammonium formate, pH 4.3, at a flow rate of 0.6 ml/min (28). A first purification of UDP-MurNAc-pentapeptide from the reaction mixtures was performed by gel filtration on a column of Sephadex G-25, as previously described (29). Fractions containing these products, as judged by absorbance at 262 nm or radioactivity measurement, were pooled and lyophilized. The purification was completed by HPLC on a column of Vydac 218TP1022 (250 ϫ 22 mm; Touzart & Matignon) using elution with 50 mM ammonium formate, pH 4.3, at a flow rate of 7 ml/min. The purity of UDP-MurNAcpentapeptide was checked by analytical HPLC and spectral absorbance analysis, and quantitation was obtained by amino acid analysis of a sample after hydrolysis in 6 M HCl for 16 h at 95°C.

Assays for Translocase Activity
Standard MraY Assay-The assay was performed in a final volume of 10 l containing 100 mM Tris-HCl, pH 7.5, 40 mM MgCl 2 , 1.  In all cases, the reaction was stopped by heating at 100°C for 1 min, and the radiolabeled substrates (UDP-MurNAc-pentapeptide or UDP-GlcNAc) and reaction products (C 55 -PP-MurNAc-pentapeptide (lipid I) or C 55 -PP-MurNAc(-pentapeptide)-GlcNAc (lipid II)) were separated by TLC on silica gel plates LK6D (Whatman) using 2-propanol/ammonium hydroxide/water (6:3:1; v/v/v) as a mobile phase. The radioactive spots were located and quantified with a radioactivity scanner (model Multi-Tracemaster LB285; EG&G Wallac/Berthold). One unit of MraY activity corresponds to 1 nmol of lipid I produced per min.

Extraction of Lipid I by Butanol
An MraY standard assay (scaled up to 50 l) was performed as described above, except that UDP-MurNAc-pentapeptide was unlabeled. The reaction was stopped by the addition of 50 l of 6 M pyridinium acetate. Then 100 l of 1-butanol were added, and the mixture was vortexed for 2 min and centrifuged at 10,000 ϫ g for 5 min. The aqueous and organic phases were recovered. The organic phase was evaporated and taken up in 2-propanol/methanol (1:1; v/v) while the aqueous phase was left intact. Both phases were analyzed by mass spectrometry.

Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) Mass Spectrometry Analysis
MALDI-TOF mass spectra were recorded in the linear mode with delayed extraction on a PerSeptive Voyager-DE STR instrument (Applied Biosystems) equipped with a 337-nm nitrogen laser.
MraY-The samples were prepared according to Grü ber et al. (30); 0.5 l of MraY preparation was deposited on the plate and allowed to dry. Subsequently, 0.5 l of matrix solution (10 mg/ml ␣-cyano-4-hydroxycinnamic acid in 50% (v/v) acetonitrile, 0.1% (v/v) trifluoroacetic acid) was applied to the dried sample and again allowed to dry. Spectra were recorded in the positive ion mode at an acceleration voltage of ϩ25 kV and an extraction delay time of 300 ns. The MurC protein was used as an external calibrant.
Lipid I and UDP-MurNAc-pentapeptide-One l of matrix solution (10 mg/ml 6-aza-2-thiothymine in 20 mM diammonium citrate) was deposited on the plate, followed by 0.5 l of sample (dissolved in 2-propanol/methanol (1:1; v/v) for lipid I and water for UDP-MurNAc-pentapeptide). After evaporation of the solvents, spectra were recorded in the negative ion mode at an acceleration voltage of Ϫ20 kV and an extraction delay time of 100 ns. A mixture of UDP-MurNAc, UDP-MurNAc-L-Ala-D-Glu, and UDP-MurNAc-pentapeptide was used as an external calibrant.

RESULTS AND DISCUSSION
The MraY translocase catalyzes the first membrane step of peptidoglycan synthesis, an essential step considered as a very promising target for the search of new antibacterial compounds. It has been clearly underexploited to date, most probably because of the refractory nature of this protein to overexpression and purification. Indeed, all previous attempts to overproduce to high levels the MraY translocase, to purify it, or at least to detect it by SDS-polyacrylamide gel electrophoresis were unsuccessful (31,32). In fact, only a radiolabeled form of this protein has been detected to date using in vitro translation experiments (9,10). We recently investigated the membrane topology of the E. coli MraY protein and showed that its expression as a fusion with ␤-lactamase was an advantage for its stability and/or production. A significant but moderate 30-fold increase of MraY activity was detected in cells in which the MraY-BlaM fusion was expressed (9), and the solubilized fusion protein was subsequently used for the development of a high throughput screening assay based on fluorescence detection (17). A similar (28-fold) overproduction factor was reported by Brandish et al. (32) when the E. coli mraY gene was cloned in the expression vector pTrc99A. Why the levels of overexpression were in all cases so modest was unclear given the well documented strength of the promoters used. We thus decided to revisit this question by performing a complete set of experiments using mraY genes from various bacterial species, different plasmid vectors, and different host strains and conditions for expression.
The mraY genes from E. coli, S. aureus, T. maritima, and B. subtilis were amplified by PCR and cloned into the pTrc99A and pET28b expression vectors. In all of these constructs, the mraY gene product was expressed with a His 6 tag extension to allow its easy purification on Ni 2ϩ -NTA-agarose. Strain C43(DE3), which is particularly well adapted for high level expression of membrane proteins (33), was chosen as the host strain, since the extent of MraY overexpression in that strain was systematically about 5-fold higher than that observed in the other E. coli strains tested (DH5␣ and BL21(DE3)). In the case of B. subtilis MraY, expression from pTrc99A and pET28b vectors resulted in 25-and 145-fold increased levels of translocase activity, respectively, when cells were grown and induced at 37°C and 5-and 330-fold, respectively, when cells were first grown at 37°C and then induced at 25°C. Similar results were obtained with MraY from E. coli, S. aureus, and T. maritima (Table II). We therefore essentially used pET28bderivative plasmids and induction at 25°C in all subsequent experiments.
Small scale cultures (50 ml) were then performed to identify the best MraY candidate to purify (i.e. the one, among the aforementioned species, exhibiting the best compromise between overexpression level and enzyme stability during the extraction and purification steps). Membrane proteins of IPTGinduced and noninduced cells were solubilized with DDM and tested for MraY activity. Cells carrying the different mraYoverexpressing plasmids all contained a similar 100 -500-fold increased level of MraY activity in membranes when induced with IPTG. At this step, however, no significant increase of a protein band that could correspond to MraY (calculated mass of about 36 kDa) was detected by SDS-PAGE analysis of the extracts (data not shown). Various detergents (N-lauroyl sarcosine, DDM, Triton X-100, n-octyl-␤-D-glucopyranoside, and CHAPS) were then tested for their efficiency to extract the MraY proteins from the membranes and to maintain them in an enzymatically active form (Table III). N-Lauroyl sarcosine and DDM appeared more efficient than the other detergents for MraY extraction. However, the ionic detergent N-lauroyl sarcosine did not allow a sufficient binding of MraY on Ni 2ϩ -NTAagarose, the resin used for the final step of purification. Most of the MraY protein was eluted at very low concentrations of imidazole (5-20 mM) and was thus only poorly purified in these conditions. In the case of DDM, the affinity of MraY for the resin was much better, and the purification of a protein eluted at higher concentrations of imidazole, ranging from 20 to 300 mM, was observed. Since the protein yield and purification state appeared significantly higher in the case of the B. subtilis MraY protein coded by the pETYBS62 plasmid (N-terminal His 6 -tagged form), the latter was chosen for large scale purification and enzyme characterization experiments.
Overproduction and Purification of B. subtilis MraY-The sequence of the B. subtilis mraY gene amplified from strain 168 and cloned into pETYBS62 plasmid showed some minor differences from that found in databases (Pasteur data base, available on the World Wide Web at genolist.pasteur.fr/SubtiList/); the sequence 589 CGTGAT 594 (coding for RD) repeatedly appeared as GCTCAT (coding for AH) in the products from several independent PCRs using high fidelity polymerase. The molecular mass calculated for the B. subtilis MraY protein was 36,568 Da, taking into account the N-terminal extension tag consisting in Met-Ser-His 6 .
A 5-liter culture of E. coli C43DE3(pETYBS62) was induced with IPTG, and membranes were prepared and tested for MraY activity. As previously observed with 50-ml cultures, the specific activity in this crude membrane extract was about 300-fold higher than that detected in control cells carrying the pET28b vector (66 versus 0.2 units/mg of protein, respectively). The extraction of the MraY protein was achieved by four successive treatments of membranes with DDM: one with 17.8 mM DDM (extract DM1) followed by three with 21.5 mM DDM (extracts DM2-DM4). Table IV recapitulates the levels of MraY activity detected in these different extracts. The differential extraction with DDM allowed us to remove a large amount of proteins but only little of the MraY activity in DM1. About one-third of the total MraY activity was recovered in DM2, with a 20% increase in specific activity. Protein amounts in DM3 and DM4 were lower, with a specific activity of ϳ50 units/mg of protein as compared with 80 units/mg of protein in DM2. The DM2 extract was thus chosen for the purification of the MraY protein.
The purification was carried out in a single affinity chromatography step (see "Experimental Procedures"). Solubilized membranes (DM2 extract, 280 mg of proteins) were mixed with Ni 2ϩ -NTA-agarose and incubated for 2 h at 4°C under shaking. The resin was transferred to a chromatography column, and washing and elution steps were carried out with a discontinuous gradient of imidazole (0, 5, 10, 20, 35, 45, 60, and 300 mM). Fractions were collected and analyzed for protein content and translocase activity (Table V). In the absence of imidazole and up to a 5 mM concentration of this compound, almost all of the MraY activity remained bound to the column. It started to be released from the Ni 2ϩ -NTA resin at 10 mM, but the specific activity was shown to be maximal in the 60 and 300 mM imidazole-containing fractions. About 6 mg of pure MraY protein were recovered in the latter fractions, with a specific activity of ϳ1,900 units/mg of protein, 9,500-fold higher than the basal activity detected in wild-type E. coli membranes (ϳ0.2 units/mg of protein).
Analysis of the latter purified fractions by SDS-PAGE showed a unique protein band migrating as a protein of 31 kDa (Fig. 1). It was also detected by Western blot using monoclonal antibodies directed against the histidine tag (data not shown). The molecular mass of the protein as deduced from gel migra-  tion was lower than the theoretical one (36,568 Da). However, such a difference is often observed with integral membrane proteins (9,34).
MALDI-TOF mass spectrometry analysis of the pure protein in the presence of the detergent necessary to maintain its solubility turned out to be difficult. No signal was recorded when sinapinic acid, the traditional matrix for proteins, was used. However, good quality spectra were obtained with ␣-cyano-4-hydroxycinnamic acid, a matrix recommended by Cadene (Fig. 2). This was in perfect agreement with the molecular mass of 36,568 Da calculated from the MraY sequence, indicating that the N-terminal methionine has not been eliminated during in vivo expression of the recombinant protein. It was noteworthy that no additional peak was observed up to m/z 100,000 (data not shown), confirming the purity and homogeneity of the MraY preparation already observed by SDS-PAGE analysis.
Characterization of Lipid I Produced by Purified MraY-The assay of purified MraY enzyme was based on the addition of the phospho-MurNAc-[ 14 C]pentapeptide moiety of UDP-MurNAc-[ 14 C]pentapeptide to C 55 -P, followed by TLC separation on silica gel plates (Fig. 3). Radioactive spots corresponding to the nucleotide substrate and lipid I product were located (R f ϭ 0.25 and 0.58, respectively) and quantified with a radioactivity scanner (see "Experimental Procedures"). The reaction was then scaled up and performed with unlabeled UDP-MurNAcpentapeptide. After incubation, the mixture was extracted with butanol, and both the aqueous and organic phases were subjected to MALDI-TOF mass spectrometry analysis. Peaks of m/z 1,193.0 and 1,716.2 were detected in the aqueous and organic phases, respectively (data not shown). These m/z ratios matched the expected values for the [M Ϫ H] Ϫ ions of UDP-MurNAc-pentapeptide and lipid I (1,192.9 and 1,716.1, respectively). This result confirmed that the compound in spot B (Fig.  3) consisted of lipid I.    Coupling of Purified MraY and MurG Activities-Radiolabeled C 55 -PP-MurNAc(-pentapeptide)-GlcNAc (lipid II) was formed when both MurG enzyme and UDP-[ 14 C]GlcNAc were incorporated in the MraY reaction mixture, further indicating that coupling MraY translocase/MurG transferase reactions could be successfully performed with purified preparations of the two enzymes. In our conditions, in which UDP-GlcNAc was in limited quantity (see "Experimental Procedures"), a complete incorporation of GlcNAc was observed. No lipid II was formed when MurG was omitted, demonstrating the absence of MurG in the MraY preparation. In addition, we showed that no MraY activity could be detected when either of the two substrates, UDP-MurNAc-pentapeptide or C 55 -P, was omitted in the reaction mixture, demonstrating the absence of either compound (and in particular of any traces of the C 55 -P carrier lipid) in the preparation. No radiolabeled lipid was formed when UDP-MurNAc-[ 14 C]pentapeptide was replaced by UDP-[ 14 C]GlcNAc in the standard translocase assay, indicating that the purified MraY preparation was also not contaminated by WecA enzyme (36), a membrane activity catalyzing the transfer of [ 14 C]GlcNAc from UDP-GlcNAc onto the C 55 -P carrier lipid.
Biochemical Properties of the Pure MraY Enzyme-The translocase activity of the pure enzyme preparation was characterized in more detail. First, the effects of pH, salts, metal ions, and detergent concentrations were tested. The effect of pH was determined in the range 5.9 -10.5 by using three different buffers. A plateau of maximal activity was observed between 7.2 and 8.0, as shown in Fig. 4A. It should be noted that this optimal value corresponds approximately to the internal bac-terial pH that ranges between 7.6 and 7.8 (37). The MraY enzyme exhibited almost no activity at pH values below 6.5. All subsequent experiments were performed at pH 7.6. The effect of salts was investigated with NaCl and KCl in the range 0 -1.7 M. MraY behaved similarly with respect to these two salts (Fig.  4B). The enzyme activity greatly increased (by a factor of about 5) between 0 and 150 mM and then reached a plateau value that did not significantly vary above 250 mM and up to the maximal salt concentration tested. The MraY activity showed an abso- lute requirement for a divalent cation (Fig. 4C). Optimal enzyme activity was observed with Mg 2ϩ for concentrations ranging from 40 to 75 mM. Although Mn 2ϩ could replace Mg 2ϩ , the translocase activity detected in the presence of this cation was much lower, representing only ϳ1% of that observed with Mg 2ϩ . As compared with Mg 2ϩ , the effect of Mn 2ϩ concentration on enzyme activity showed a curve shifted toward lower concentrations, and the optimal concentration range estimated in that case was 15-30 mM (Fig. 4C).
We also tested the effects of various detergents on MraY translocase activity (Fig. 4D). N-Lauroyl sarcosine clearly appeared as the most efficient one, with an optimal concentration ranging between 7 and 16 mM. At a 32 mM concentration of this detergent, the residual activity was ϳ50% of the optimum. In the presence of Tween 20 or Triton X-100 (optimal concentrations of 7.7 and 15 mM, respectively), the MraY activity represented about 30% of that observed in the presence of N-lauroyl sarcosine. The activity in the presence of DDM was weaker by an order of magnitude but surprisingly remained intact for concentrations up to 80 mM of this detergent (data not shown).
In the presence of n-octyl-␤-D-glucopyranoside, the optimal activity observed at 2.5 mM represented only 4% of that observed with N-lauroyl sarcosine, and it dramatically decreased for concentrations higher than 10 mM. Among the different detergents tested, CHAPS appeared as the worst for translocase activity, the activity being decreased by ϳ70-fold in that case as compared with N-lauroyl sarcosine (Fig. 4D). The MraY translocase activity detected was probably the result of multiple interactions between enzyme-detergent and C 55 -P-detergent complexes within detergent micelles. This could explain why N-lauroyl sarcosine, an ionic detergent with a low aggregation number, was the best detergent for translocase activity.
In the optimal conditions for MraY activity determined above, typical Michaelis-Menten kinetics were observed in the concentration ranges considered (60 M to 3.7 mM for UDP-MurNAc-pentapeptide and 70 M to 1.1 mM for C 55 -P). The K m values of the purified MraY translocase for its substrates UDP-MurNAc-pentapeptide and C 55 -P were 1.0 Ϯ 0.3 and 0.16 Ϯ 0.08 mM, respectively, and the catalytic constant k cat was 320 Ϯ 34 min Ϫ1 . Due to the great difference between the two K m constants and because radiolabeled C 55 -P was not available, the K m for C 55 -P was determined at a subsaturating concentration (0.25 mM) of UDP-MurNAc-pentapeptide and should thus be considered as an apparent K m .
Effect of Tunicamycin on Translocase Activity-Tunicamycin was earlier known to inhibit enzymes catalyzing the transfer of UDP-GlcNAc or UDP-MurNAc-pentapeptide onto polyprenyl phosphate carrier lipids (38,39). We investigated here the effect of this compound on the purified MraY activity. As shown in Fig. 5, the translocase activity was clearly inhibited by tunicamycin, and an IC 50 of ϳ12 M was determined.
Conclusions-The purification to homogeneity of the MraY translocase described for the first time in the present work allowed us to analyze the kinetic properties of this enzyme in the absence of any contaminating protein or substrate (in particular C 55 -P) originating from membranes. Its apparent stability and availability in milligram quantities now opens the way to structural analysis of this essential membrane protein.
The reproducibility of the purification procedure makes it well adapted for the purification of multiple mutants and the development of structure-activity relationship analyses.