Purification and Characterization of a Carboxypeptidase-Transpeptidase of Bacillus megaterium Acting on the Tetrapeptide Moiety of the Peptidoglycan*

The enzyme carboxypeptidase-IIW of Bacillus megaterium incorporates free diaminopimelate into puri- fied bacterial walls. This enzyme can be solubilized from toluene-treated cells by LiCl extraction and has now been purified 106-fold to one major band on polyacrylamide gel electrophoresis. The enzyme has an ap- parent molecular weight of approximately 60,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and by Sephadex G-100 gel filtration. Carboxy- peptidase-IIW requires divalent cations and thiol group(s) for optimal activity. Product analysis indicates that the enzyme can hy-drolyze the terminal D-alanine from the tetrapeptide of the peptidoglycan or replace it with a variety of amino acids with D-asymmetric centers for transpeptidation.

The enzyme carboxypeptidase-IIW of Bacillus megaterium incorporates free diaminopimelate into purified bacterial walls. This enzyme can be solubilized from toluene-treated cells by LiCl extraction and has now been purified 106-fold to one major band on polyacrylamide gel electrophoresis. The enzyme has an apparent molecular weight of approximately 60,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and by Sephadex G-100 gel filtration. Carboxypeptidase-IIW requires divalent cations and thiol group(s) for optimal activity. Product analysis indicates that the enzyme can hydrolyze the terminal D-alanine from the tetrapeptide of the peptidoglycan or replace it with a variety of amino acids with D-asymmetric centers for transpeptidation. Substrate specificity studies reveal that the enzymatic activity depends on the presence of N-acetyl-D-glucosamine of the GlcNAc-MurNAc-tetrapeptide.
This specificity of carboxypeptidase-IIW for the N-acetyk-glucosamine explains in part the affinity of the enzyme for the cell wall of B. megaterium.
The enzyme is compared to the carboxypeptidases-transpeptidases of other organisms with the similarities and differences discussed.
In 1972, Wickus and Strominger (1,2) proposed a reaction sequence for the incorporation of free DD-and meso-diaminopimelate into the peptidoglycan synthesized by membrane particles of Bacillus megaterium KM. This reaction sequence in the presence of concurrent peptidoglycan synthesis involved the replacement of the terminal n-alanine of the disaccharidepentapeptide' of the peptidoglycan by diaminopimelate. Consequently, D-alanine was released. Some of the incorporated DD-diaminopimelate resulted in a peptide cross-linkage unique to B. megaterium (l), whereas the remaining portion of the incorporated DD-and meso-diaminopimelate resulted in no cross-linkage. These authors suggested that this activity was due to enzyme(s) similar to the carboxypeptidase(s)-transpeptidase(s) of other organisms ( should be addressed. ' The abbreviations used are: pentapeptide, L-Ala-D-y-Glu-meso-Azpm-D-Ala-D-Ala; tetrapeptide, L-Ala-D-Glu-meso-Azpm-D-Ala; tripeptide, L-Ala-n-Glu-meso-Azpm; disaccharide-peptide, GlcNAc-MurNAc-pentapeptide, tetrapeptide, and tripeptide in an unknown ratio; Alpm, a,+diaminopimelate. Taku et al. (4,5) have also studied the incorporation of diaminopimelate into the peptidoglycan synthesized by toluene-treated cells and membrane particles of B. megaterium 899. They found that toluene-treated cells of B. megaterium could utilize the externally added nucleotide precursors UDP-GlcNAc and UDP-MurNAc-Ala-Glu-Azpm-Ala-Ala to synthesize peptidoglycan and that these cells could also incorporate free diaminopimelate into the cell wall. Proteins could be extracted from these toluene-treated cells by a high concentration of LiCl. The LiCl-treated cells synthesized peptidoglycan and incorporated free diaminopimelate less efficiently than before the extraction. The loss of these synthetic activities could be restored by adding a crude extract of the LiClsoluble proteins to the LiCl-treated cells. Of the several extracted proteins, one was purified to electrophoretic homogeneity and was identified as the N-acetylglucosaminyltransferase involved in peptidoglycan synthesis and its addition stimulated this synthesis in the LiCl-treated cells (5, 6). In this communication, we examine the LiCl-extractable enzyme able to incorporate free diaminopimelate into previously formed cell walls. The diaminopimelate used was a mixture of LL, DD, and mesoisomers.
We report here the purification and initial characterization of this enzyme which we demonstrate to be a diaminopimelyl-D-alanyl-carboxypeptidase-transpeptidase.
We propose to call this enzyme carboxypeptidase-IIW since it is wall-bound and has a mode of action similar to that of the diaminopimelyl-n-alanyl-carboxypeptidase-II of other organisms, which also removes the terminal alanine from the MurNAc-tetrapeptide of peptidoglycan. In addition, most bacteria possess a D-alanyl-n-alanyl-carboxypeptidase-I which cleaves the terminal alanine from the MurNAc-pentapeptide of peptidoglycan (3). The study of the mode of action of this enzyme should give additional information about the variety of ways free diaminopimelate can be added to the peptidoglycan of the cell wall. As described below, carboxypeptidase-IIW has both significant differences and similarities with other carboxypeptidasestranspeptidases.
Unlabeled amino acid-containing compounds were located by spraying with ninhydrin. Radioactive samples in aqueous solutions were counted by liquid scintillation using Beckman BioSolv solubilizer. Affinity Column-Six milliliters of the above monomer solution (0.8 mM), 1.0 ml of 1.0 M sodium phosphate buffer, pH 6.5, and 3.0 ml of water were shaken with 0. The enzyme has been purified 106-fold to a single major band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The details of the purification including Figs. 1 to 3 are presented in the accompanying miniprint.
In brief, the enzyme was extracted from toluene-treated B. megaterium cells using LiCl. Then the enzyme was purified using hydroxylapatite, Sephadex G-100 gel filtration, and affinity chromatography steps.
Before choosing the LiCl extraction of toluene-treated B. megaterium cells as the source of the enzyme, initial studies were performed to determine the best method for extracting the enzyme. When fresh cells were extracted by 1.5 M LiCl with and without prior toluene treatment, it was observed that the enzyme could be extracted without toluene pretreatment. Carboxypeptidase-IIW had a specific activity a-fold higher in these extracts than in LiCl extracts of cells that were also toluene-treated.
However, the yield of the enzyme in these extracts were 5-fold less compared to extracts of toluenetreated cells. Moreover, the proteins in these extracts could not be concentrated by a simple ammonium sulfate precipitation step.
Cells were sonicated to obtain membrane particles and wall preparations as described elsewhere (13). The endogenous content of carboxypeptidase-IIW in cell walls was measured by the incorporation of diaminopimelate into freshly isolated native walls. This incorporation was approximately 3 times higher than that of membranes mixed with purified walls. However, only 4% of the wall-bound enzyme could be extracted by 1.5 M LiCl. This was observed both for native walls with associated endogenous enzyme or purified walls bound with externally added enzyme. The inefficient extraction of carboxypeptidase-IIW from enzyme bound to purified walls suggested that this enzyme was difficult to remove by salt extraction rather than the alternative of the bacteria possessing more than one diaminopimelate-incorporating enzyme. The facts that no more than 4% of the enzyme located in the walls could be extracted by LiCl and that toluene treatment resulted in 5-fold increase in the yield of enzyme extracted, suggested that carboxypeptidase-IIW extracted by the above method was mainly a membrane-bound enzyme. The affinity step took advantage of the observation that the enzyme bound to walls and hence was likely to bind to wall fragments.
In fact, the enzyme bound to one of the wall fragments obtained by lysozyme digestion, namely the uncross-linked disaccharide-peptide (monomer) of the peptidoglycan. Hence, the enzyme was attached to an affinity column containing this monomer. A high salt concentration was used to elute the enzyme because the enzyme could be extracted from the cells and walls using the same conditions (5). The main reason for attaching the enzyme to disaccharide-peptide (monomer) linked to immobile beads rather than undegraded walls was that recovery of the enzyme bound to intact walls was no more than 4%. In addition, intact walls have several other components besides the disaccharide-peptide (monomer) (3). Thus, some proteins besides carboxypeptidase-IIW could have bound to these components, leading to a less specific purification.

Purification Summary
The results of the above four purification steps are summarized in Table I. Minimal loss of activity was observed when the enzyme was maintained at pH 6.5, even though the optimal pH for the assay of the enzyme activity was 8.0.

5674
Purification of Carboxypeptidase-IIW Hence, sodium phosphate buffer, pH 6.5, was used during the purification whenever feasible. When the final preparation from the affinity chromatography step was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( Fig.  4), one major and one minor band were observed. The major band had a molecular weight of 60,000 (Fig. 5B). The enzyme from the affinity chromatography step was also applied to a Sephadex G-100 column for molecular weight determination and a value of 56,000 was obtained. The fact that a single component of similar molecular weight was found for the enzyme using both an enzymatic (Fig. 5A) and a protein ( Fig.  5B) measurement, indicated that the enzyme was purified close to electrophoretic homogeneity during the three chromatographic steps employed. The similarity in molecular weight suggested that carboxypeptidase-IIW is a monomeric enzyme which does not dissociate into subunits in the presence of sodium dodecyl sulfate and a reducing agent.  Fig. 3 was electrophoresed in a sodium dodecyl sulfate-polyacrylamide gel as described under "Materials and Methods." The gel was stained for protein using Coomassie blue and then a densitometer tracing was made at 540 nm.  Fig. 3 were pooled and concentrated. These fractions (0) were compared to reference compounds with the indicated molecular weights (0): rabbit muscle D-k&ate dehydrogenase tetramer (136,000), bovine serum albumin (68,GOO), ovalbumin (43,OOO), rabbit muscle n-lactate dehydrogenase monomer (36,000), bovine pancreas carboxypeptidase A (34,000), beef pancreas deoxyribonuclease (31,000), hen egg white lysozyme (14,300), and horse cytochrome c (12,400). Not all reference proteins were used for both frames. Frame A, carboxypeptidase-IIW from the affinity column was analysed on a Sephadex G-100 column (2.3 X 90 cm) which had been equilibrated with 0.05 M sodium phosphate buffer, pH 6.5, 1 IIIM 2-mercaptoethanol, and 0.4 M KCl. Elution was carried out using the same buffer and fractions of 4.5 ml were collected at a flow rate of 14 ml/h. The enzyme was assayed by the binding assay while the reference proteins were detected by their ASMU). Frame B, enzyme preparation from the affinity column was analyzed by sodium dodecyl sulfate-lo% polyacrylamide gel electrophoresis. The gels were electrophoresed for 4 to 6 h at 7 to 9 mA/tube as described under "Materials and Methods." All samples were stained for protein using Coomassie blue.

Properties of Carboxypeptidase-IIW
Assay Conditions-Unless otherwise stated, experiments described below were performed using the partially purified enzyme from Step 2. Its specific activity ranged from 600 to 900 units/mg of protein. The direct assay described under "Material and Methods" was employed. The incorporation of diaminopimelate into cell walls by the enzyme was linear with respect to time up to 90 min (Fig. 6A). The assay also was linear with respect to enzyme concentration until a plateau was reached (Fig. 6B). The maximum amount of diaminopimelate that could be incorporated into walls purified from 1.5 x lo8 cells under the given assay conditions was in the range of 23 to 32 pmol. The optimal temperature for the incorporation of the diaminopimelate into walls by the enzyme was in the range of 30-60°C. From O-3O"C, the incorporation of diaminopimelate increased linearly. The relative incorporation at O"C, lO"C, and 60°C were O%, 27%, and 99%, respectively, of the incorporation at 30°C. Carboxypeptidase-IIW was totally inactivated after 5 min in a boiling water bath in 0.05 M sodium phosphate buffer, pH 6.5, 1 mM 2-mercaptoethanol. This is contrary to the data published earlier which states that the activity is insensitive to heat to some extent (4). Probably the heat sensitivity of the enzyme depends upon the exact inactivation conditions, the buffers used and the degree of purity of the enzyme.
The presence of Mg2+ ions stimulated the activity of the enzyme approximately 2-to 2?+fold with the optimal concentration being in the range of 10 mM (Fig. 7). Higher concentrations were inhibitory. Table II shows that Mg2+ ions could be replaced by Mnx' or Ca2+ ions but not by Cu2+ or Zn2+ ions.
The pH curve for the enzyme assay was determined using various buffers ranging from pH 3 to pH 10. The optimum pH range was 7.5 to 8.2 (Fig. 8).
The incorporation of diaminopimelate into walls could be Step 2 (32 pg) was assayed by the direct assay as a function of time. Frame B, partially purified enzyme was assayed by the direct assay as a function of enzyme concentration.
The incorporation of diaminopimelate into walls by carboxypeptidase-IIW approached a maximum at a concentration of 12 mM diaminopimelate in the typical assay mixture (Fig. 9), using enzyme from Step 4 of the purification with specific activity of 6,000 units/mg of protein. In both the direct and the binding assay, a much lower concentration of diaminopimelate (0.25 mM) was used in order to increase the specific radioactivity of [3H]diaminopimelate and hence the sensitivity of the assay. Since the amount of diaminopimelate incorporated by the enzyme at these suboptimal concentrations of substrate was very low (23 pmol in Fig. 9) in the direct and the binding assay, the possibility existed that the compound incorporated into walls was not diaminopimelate but rather some contaminant.
This possibility was ruled out by analysis of the walls which had incorporated [3H]diaminopimelate. These walls were hydrolyzed in 4 N HCl for 18 h after complete removal of all unused substrates by centrifugation and washing. The hydrolyzed material was subjected to paper chromatography in isobutyric acid:1 N NH40H (5:3, v/v) for 18 h and the region of the paper corresponding to diaminopimelate was cut out and counted. All of the radioactivity incorporated FIG. 7. Effect of magnesium ion concentration on the incorporation of diaminopimelate into purified walls by carboxypeptidase-IIW. Partially purified enzyme from Step 2 (25 pg) was assayed by the direct assay at various magnesium ion concentrations. Purification of Carboxypeptidase-IIW into walls by the enzyme was found in the diaminopimelate position. From the Lineweaver-Burk plot of the result of Fig.  9 (inset), the K, value for diaminopimelate was about 2 mM. The turnover number for carboxypeptidase-IIW for diaminopimelate incorporation was 0.075 mol/mol/min from this figure. However, the K, is a maximum estimate and the turnover number is a minimum estimate because the diaminopimelate concentration curve was constructed using only the concentration of walls that were purified from 1.5 X 10' cells. Besides, as discussed below, the enzyme has a hydrolytic activity perhaps more efficient than the diaminopimelate-incorporating activity.

Effect of Sulfhydryl
Inhibitors on Carboxypeptidase-IIW-Treatment of the enzyme with either 3 IIIM N-ethylmaleimide or 3 mM iodoacetamide resulted in the loss of approximately 60% of the activity while one-tenth this concentration of inhibitors had no substantial effect. Dithiothreitol and 2-mercaptoethanol at concentrations of 0.6 mu and 2.0 mM, respectively, stimulated the activity of the enzyme to 141% and 139% of the value without the reducing agent (Table III). Hence, during all purification procedures 1 IDM 2-mercaptoethanol was included. The inactivation of the enzyme by sulfhydryl inhibitors appeared to be partially reversible. Pretreatment with N-ethylmaleimide did not prevent subsequent stimulation by dithiothreitol, although full stimulation was not achieved. These observations suggested the requirement of free sulfhydryl group(s) for optimal activity of the enzyme.

Alanine
Release-Since carboxypeptidase-IIW can bind to walls and incorporate diaminopimelate in the absence of an exogenous energy source, it was surmised that a peptide bond was being broken to obtain energy necessary for the incorporation of diaminopimelate.
Other enzymes of this nature have already been reported (see review by Blumberg and Strominger (3)). To test this possibility, cell walls radioactively labeled in the n-alanine positions of the peptidoglycan were incubated with the enzyme. Upon chromatography of the reaction mixture, free radioactive alanine was observed to be released (described more fully below and in Table IV). When the same experiment was carried out using cell walls labeled with radioactive diaminopimelate at the third position of the pentapeptide of the peptidoglycan, no free radioactive amino acid or small peptide was released (data not shown). This Partially purified enzyme from Step 2 (32 pg) was incubated with the indicated amounts of sulfhydryl inhibitors in 0.05 M Tris-HCl, pH 7.4, in a total volume of 30 ~1 for 25 min at 0-4°C. Where indicated, 5 ~1 of dithiothreitol in 0.05 M Tris-HCl was then added and the incubation was continued for another 5 min. In the samples without dithiothreitol, buffer was used instead. The treated enzyme was then assayed for incorporation of diaminopimelate into purified walls by the direct assay.  ' Prepared as a mixture from peptidoglycan made in vitro without penicillin.
'Walls from 1.4 x 10' cells were used in the incubation.
suggested that the tripeptide of the peptide moiety of peptidoglycan was not used as a substrate by the enzyme. The fact that only free alanine was released suggested that the enzyme has a hydrolytic activity on the tetrapeptide or the pentapeptide containing 1 or 2 n-alanine residues, respectively. Since carboxypeptidase-IIW also incorporated diaminopimelate, this amino acid must be replacing the terminal or the subterminal D-ahnh2 of the peptide moiety if the enzyme acted as a transpeptidase.
Transpeptidation in Organic Solvents-The alanine release experiments described above indicate that the enzyme has a hydrolytic as well as transpeptidase activity. That both activities are due to the same enzyme was demonstrated by an experiment of a type described by Zeiger et al. (15). In a nonaqueous environment where the hydrolysis is limited by the absence of water molecules, the release of n-alanine should be dependent on the presence of diaminopimelate.
Therefore, cell walls labeled with 14C at the n-alanine positions were incubated with carboxypeptidase-IIW in varying concentrations of a mixture of aqueous and nonaqueous solvents. The amount of [14C]alanine released and ["Hldiaminopimelate incorporated was measured. The diaminopimelate-incorporating activity of the enzyme decreased with increasing proportions of the nonaqueous solvent. At 55% HzO, the activity of carboxypeptidase-IIW was 14% of that of 100% Hz0 (data not shown). Fig. 10 shows that within the limits of discernible activity of the enzyme, the hydrolytic activity is partially dependent on the presence of diaminopimelate.
Even in the presence of Hz0 alone, slightly more alanine was released in the presence of diaminopimelate than in its absence. Complete dependence of hydrolytic activity on the presence of diaminopimelate was not obtained at any of the conditions used and may require complete removal of the water from the Purification of Carboxypepttdase-IIW 5677 assay. Moreover, the enzyme was progressively inactivated as the HZ0 is replaced. Therefore, it was difficult to study alanine release in the absence of HzO.
These observations support the conclusion that carboxypeptidase-IIW is capable of carrying out both the alanine release and diaminopimelate incorporation.
In this respect it resembles the carboxypeptidases of other organisms studied (3). Substrate Specificity of Carboxypeptidase-IIW-Various compounds generally used as substrates for carboxypeptidases were tested for their suitability as substrates for carboxypeptidase-IIW (Table IV). The enzyme released very little alanine from UDP-MurNAc-pentapeptide, MurNAc-pentapeptide, or disaccharide-pentapeptide.
The enzyme also did not hydolyze peptidoglycan made in the presence of 100 pg/ml of penicillin G which contains uncross-linked polymer with predominantly disaccharide-pentapeptide units (9). All the above compounds are generally substrates used by carboxypeptidases-I (3,9,(16)(17)(18). The enzyme also failed to remove alanine from UDP-MurNAc-tetrapeptide and MurNAc-tetrapeptide, which are generally substrates used by carboxypeptidases-II (19,20). Therefore, the enzyme is dissimilar to the previously reported carboxypeptidases-I and -11. Moreover, carboxypeptidase-IIW failed to demonstrate any endopeptidase activity as evidenced by its failure to convert bis(disaccharide-peptide) (crosslinked dimer, see "Materials and Methods") to disaccharidepeptide (monomer) (data not shown). However, the enzyme was able to use as substrate, cell walls, peptidoglycan made in uitro, and disaccharide-peptide (monomer). Fig. 2 Table IV and Fig. 11 were made from peptidoglycan synthesized in vitro from UDP-MurNAc-tripeptide-n-['4C]Ala-n-['4C]Ala.
The radioactive monomeric material was a mixture of disaccharide-tetrapeptide and disaccharide-pentapeptide in an unknown ratio. It is technically difficult to separate these two kinds of monomers cleanly. But if the disaccharide-tetrapeptide is the actual substrate for carboxypeptidase-IIW and if there is a replacement of its terminal n-alanine by diaminopimelate, then the disaccharide- is accompanied by the release of n-alanine. To distinguish between these possibilities, disaccharide-peptide labeled with [ 14C]alanine at the terminal positions were incubated with carboxypeptidase-IIW and ["Hldiaminopimelate. At the end of the incubation period, the reaction mixture was boiled and the products of the reaction were analyzed. The first step involved chromatography on a Dowex 1-formate column to remove the salts in the reaction mixture. These salts would have interfered with subsequent electrophoresis. The Dowex column was then eluted with increasing concentrations of ammonium formate. The bulk of the unused diaminopimelate eluted first. These fractions which contained 3H but had no trace of 14C radioactivity were discarded. The unused disaccharide-peptide eluted later and the reaction product containing ["Hldiaminopimelate eluted after that by virtue of the extra carboxyl group on the incorporated diaminopimelate.
However, the two kinds of monomers were poorly separated. The radioactive eluate consisting of unreacted disaccharide-peptide monomers, the product of carboxypeptidase-IIW activity and the remaining unused diaminopimelate not separated by the Dowex fractionation, were combined and lyophilized to remove the formate completely. This material was subjected to paper electrophoresis at pH 3.5 whereby the reaction product was well separated from the two substrates, namely disaccharide-peptide and any residual diaminopimelate.
That disaccharide-tetrapeptide was the substrate for carboxypeptidase-IIW was demonstrated by the data in Fig. 12A. (2 @i) in a volume of 50 ~1 for 2 h at 30°C in the standard assay buffer. The reaction mixture was boiled and suspended in 1 ml of H20. The pH was raised to 10 with 1 N NaOH.
The material was then charged onto a Dowex I-formate column (1 x 4 cm) equilibrated with HzO. After washing with H20 to remove all unadsorbed material, the column was eluted with a gradient of 0 to 0.5 M ammonium formate, pH 3.0. The total volume of the gradient was 200 ml. Fractions of 3 ml were collected and assayed for 3H and %. All fractions containing only 3H radioactivity and eluting before the fist trace of "'C counts were discarded.
All subsequent radioactive fractions were pooled and lyophilized to remove the ammonium formate. This material containing the disaccharide-peptide (monomer) was then electrophoresed on Whatman No. 3MM paper in a pH 3.5 buffer consisting of pyridine, acetic acid, and Hz0 ( The reaction product (Peak I) in fact had 3H and no 14C as predicted, and also migrated toward the anode 5 cm less than the unused disaccharide-peptide (Peak II). This was expected since the reaction product containing an extra amino group in the incorporated diaminopimelate should move more slowly toward the anode at pH 3.5. The larger Peak II which coincided with authentic disaccharide-peptide had very little tritium from the [3H]diaminopimelate used in the incubation (Fig. 12A). To determine clearly the amount of radioactive diaminopimelate present, part of the material of this peak was eluted from the paper (Fig. 12A, arrows).
If any product incorporated diaminopimelate and co-electrophoresed with unused disaccharide-peptide it should be nearer the cathode than the anode again by virtue of the extra amino group on the incorporated diaminopimelate.
Hence, the portion of the peak region was selected as indicated. However, in another experiment, the material in the entire Peak II region was eluted with identical results (data not shown). Thin layer chromatography of the hydrolyzed product revealed only D-("C]alanine and no [3H]diaminopimelate ( Fig. 12B) confiiing the fact that no other compound incorporated diaminopimelate and electrophoresed with unused disaccharide-peptide. This result of the reaction product containing [3H]diaminopimelate but no n-[%]alanine supports the model that carboxypeptidase-IIW incorporates diaminopimelate by the replacement of the n-alanine of the disaccharide-tetrapeptide.
Pencillin Insensitivity of Carboxypeptidase-ZZW-The studies described so far indicated that the enzyme acted on the tetrapeptide moiety of the peptidoglycan. In other organisms, carboxypeptidases-II that act on the tetrapeptide moiety are normally penicillin-insensitive.
Therefore, the penicillin sensitivity of carboxypeptidase-IIW was examined. Carboxypeptidase-IIW was incubated with cell walls and with peptidoglycan labeled at the n-alanine positions in the presence of 500 pg/ml of penicillin G. As seen in Table V, the enzyme was insensitive to penicillin even at this very high concentration.
This penicillin resistance is in close parallel to the penicillin insensitivity of the carboxypeptidase-II of other organisms that use the tetrapeptide moiety and in contrast to the penicillin sensitivity of carboxypeptidases-I that use the pentapeptide moiety. Taku et al. (4,5) also observed the penicillin insensitivity of diaminopimelate incorporation into preformed cell walls of toluene-treated and membrane particles of B. megaterium.
Incorporation of Various Amino Acids by Carboxypeptidase-ZZW-Since carboxypeptidases-transpeptidases in general can incorporate a variety of n-amino acids, the enzyme   9). This same phenomenon might explain the high residual diaminopimelate incorporation when other D-amino acids were used in the competition.
Besides the ones listed in this table, other L-amino acids tested were histidine, lysine, methionine, serine, glutamic acid, threonine, and valine.
Table VI also shows the direct incorporation of some of the amino acids besides diaminopimelate into the walls. The results are in agreement with those of the competition experiment. From these data, we presume that of the three isomers found in the racemic mixture of diaminopimelate, only the DD-or the meso-diaminopimelate, or both, are incorporated into walls while the LL isomer is not.
Effect of Carboxypeptidase-IIW on Walls Purified from Various Bacteria-It was not known whether disaccharidepeptide was the only component in the wall that was necessary for optimal enzymatic activity. Therefore, purified walls from various sources of bacteria were incubated with the enzyme and the incorporation of diaminopimelate was measured (Table VII). Of these bacteria, only M. Zysodiekticus does not have meso-diaminopimelate in the third position of the peptide moiety (22). Walls from all the bacterial species tested have the same disaccharide-tetrapeptide but they differ otherwise in their cell wall composition (22)(23)(24)(25). The fact that the enzyme failed to incorporate diaminopimelate into these walls, with the possible exception of B. cereus, suggested that other components besides the disaccharide-tetrapeptide can affect carboxypeptidase-IIW activity. However, since the amount of uncross-linked disaccharide tetrapeptide was not determined for the various bacteria tested, it is conceivable that the differences in carboxypeptidase-IIW activity might have been due to variations in the quantity of this substrate.

DISCUSSION
In this communication we report the purification and characterization of the enzyme carboxypeptidase-IIW that incorporates diaminopimelate into purified cell walls of B. megaterium 899 in the absence of an exogenous energy source. This enzyme, purified lo&fold, has a molecular weight of 56,000 by Sephadex G-100 filtration and 60,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. It probably requires sulfhydryl group(s) for optimal activity and also has a requirement for divalent cations. The K, value for diaminopimelate has a maximum value of 2 mM (Fig. 9).
Our data showed that purified carboxypeptidase-IIW could incorporate diaminopimelate into previously formed cell wall and could release alanine. The release of alanine was partially dependent on the presence of diaminopimelate especially when the water in the assay was partially replaced. Substrate specificity studies indicated, among other things, the requirement of the tetrapeptide moiety of the peptidoglycan for enzymatic activity. Product analysis studies indicated that diaminopimelate replaces the terminal D-alanine of the disaccharide-tetrapeptide.
On the basis of these observations, we propose the mode of action of carboxypeptidase-IIW (Product A', Fig. 13).
This enzyme activity is different from the diaminopimelateincorporating activity in the membrane particles of B. megaterium KM as described by Wickus and Strominger (1, 2) (Fig. 13). The latter activity is penicillin-sensitive as would be expected from the substrate being the pentapeptide of peptidoglycan (3). Carboxypeptidase-IIW, on the other hand, is penicillin-insensitive similar to carboxypeptidase-II of other organisms that act on disaccharide-tetrapeptide (19). Moreover, the diaminopimelate-incorporating activity by the membrane particles has been reported to be dependent on concurrent peptidoglycan synthesis (l), whereas the diaminopimelate-incorporating activity of carboxypeptidase-IIW is independent of such synthesis (4).
Available evidence suggests that the diaminopimelate incorporation by membrane particles is due to carboxypeptidase-IIW in addition to the penicillin-sensitive activity described above. For example, Wickus and Strominger have noted that 13% of the diaminopimelate incorporated into peptidoglycan was independent of peptidoglycan synthesis (1). This activity could well have been due to carboxypeptidase-IIW. The finding that most of the diaminopimelate-incorporating activity was dependent on peptidoglycan synthesis suggests that carboxypeptidase-IIW activity contributed only a small portion of all the diaminopimelate incorporated by membrane particles. In agreement with this possibility, Taku et al. (4,5) have observed that toluene-treated cells and membrane particles of B. megaterium 899 have 3-fold higher diaminopimelate-incorporating activity in presence of peptidoglycan synthesis than in its absence. The relatively low activity of carboxypeptidase-IIW in the membranes might explain why Wickus and Strominger did not observe the Product A' of Fig. 13 in their chemical analyses of the reaction product of diaminopimelate incorporation into peptidoglycan. Moreover, Products A and A' might have been difficult to resolve using the solvent and the chromatographic systems they employed.
In being penicillin-insensitive and in having a substrate specificity for the tetrapeptide moiety of the peptidoglycan, carboxypeptidase-IIW is similar to other carboxypeptidases-II. However, carboxypeptidase-IIW is also significantly different from carboxypeptidases-II of other organisms. Whereas the carboxypeptidases-II of E. coli (19) have no other specificity besides the tetrapeptide moiety of the peptidoglycan, our enzyme requires the GlcNAc moiety of the peptidoglycan for activity. That this enzyme differs from carboxypeptidase-II is evident from the fact that membrane particles of B. megaterium also have a carboxypeptidase-II that is similar to the carboxypeptidase-II of E. coli (data not shown). The specificity of carboxypeptidase-IIW for the GlcNAc moiety could explain in part the high affinity of this enzyme for the cell wall. However, this enzyme fails to utilize walls derived from other organisms that contain meso-diaminopimelate in their peptidoglycan and therefore have the disaccharide-tetrapeptide of composition identical with that of B. megaterium as one of their wall components. This suggests that other wall