Kinetic Evidence for a Second Ligand Binding Site on Streptococcus

Feb 27, 2018 - In vitro, these enzymes do not catalyze peptide hydrolysis or aminolysis, their in vivo reaction, but some, such as PBP2x, do catalyze ...
0 downloads 0 Views 1MB Size
Download PDF
Article Cite This: Biochemistry XXXX, XXX, XXX−XXX

pubs.acs.org/biochemistry

Kinetic Evidence for a Second Ligand Binding Site on Streptococcus pneumoniae Penicillin-Binding Protein 2x S. A. Adediran, Kumar Subarno Sarkar, and R. F. Pratt* Department of Chemistry, Wesleyan University, Lawn Avenue, Middletown, Connecticut 06459, United States S Supporting Information *

ABSTRACT: High molecular mass penicillin-binding proteins (PBPs, DDpeptidases) of class B, such as Streptococcus pneumoniae PBP2x, catalyze the cross-linking of peptidoglycan in bacterial cell wall biosynthesis and are thus important antibiotic targets. Despite their importance in this regard, structure− function studies of ligands of these enzymes have been impeded by the absence of useful substrates. In vitro, these enzymes do not catalyze peptide hydrolysis or aminolysis, their in vivo reaction, but some, such as PBP2x, do catalyze these reactions of certain thioesters such as PhCH2CONHCH2COSCH(D-Me)CO2− (2). We have now prepared several peptidoglycan-mimetic thioesters that we expected to more closely resemble the natural substrates of these enzymes. To our surprise, however, these compounds, although indeed substrates of PBP2x, did not, unlike 2, appear to form an acyl-enzyme intermediate during hydrolysis, and their turnover was inhibited by certain peptides and N-acylamino acids much more weakly than that of 2. An inhibitor of this type, N-benzyloxycarbonyl-D-glutamic acid, also quenched the fluorescence of PBP2x that had been labeled at the DD-peptidase active site by 6-dansylamidopenicillanic acid. These results were interpreted in terms of a model where the peptidoglycan-mimetic thioesters preferentially bound to and hydrolyzed at a site other than the classical DD-peptidase active site. This second site is likely to represent part of an extended binding site that accommodates a peptidoglycan substrate or regulator in vivo. Such a site may be a target for future inhibitor/ antibiotic design.

B

in the search for new antibiotics to combat pathogenic bacteria that are becoming more and more resistant to the current therapies.11 Streptococcus pneumoniae PBP2x is a quite typical HMMB DD-peptidase that has been studied for some time. A solubilized version has been crystallized both in the native form and in covalent complexes with β-lactams.12−14 Neither the holoenzyme15 nor the solubilized version7,16 has yet been convincingly shown to catalyze the hydrolysis or aminolysis of peptides, including peptidoglycan-mimetics. PBP2x does, however, catalyze the hydrolysis and aminolysis by D-amino acids of thioesters of general structure 1, where Ar/R represents the side chains found in typical β-lactam antibiotics.16 In an attempt to advance knowledge of the reactivity and substrate specificity of PBP2x we prepared the peptidoglycan-mimetic thioesters 3−6. The stem peptide of S. pneumoniae contains L-lysine at position 3. Compounds 4 and 5 contain the α-N-Ac-L-lysyl side chain characteristic of the S. pneumoniae cell wall. For comparison, we used compound 2, a representative of the class of substrates of structure 1. Unexpectedly, we found that hydrolysis of 3−6 was catalyzed by PBP2x but at a site on the enzyme other than the classical active site. Reaction of 3−6 at this site could be inhibited by peptides and certain N-acylamino acids and by β-lactams. These results form the basis of the present paper.

acterial DD-peptidases, otherwise known as penicillinbinding protein (PBPs), catalyze transpeptidation and carboxypeptidation reactions in cell wall biosynthesis and are the targets of β-lactam antibiotics.1,2 These proteins were originally subdivided into classes based on molecular mass,3,4 but these divisions are now known to be related to structure and function.2,5 A high molecular mass (HMM) class (50−100 kDa) has been divided into two subgroups A and B. The former subgroup comprises bifunctional enzymes that catalyze both the transglycosylase and transpeptidase reactions necessary to incorporate a peptidoglycan monomer into the cell wall polymer, while enzymes of the latter subgroup catalyze only the transpeptidation reaction. The low molecular mass (LMM) enzymes (40−50 kDa) catalyze carboxypeptidase and endopeptidase reactions of the peptidoglycan polymer. Although, as noted above, the HMM enzymes catalyze transpeptidation reactions in vivo, it has been difficult to study this activity in vitro when using the complete membrane-bound enzymes and the natural substrate, lipid II. Only very recently has progress been made in this area.6 Substrate structure−activity studies, however, have not yet been performed. Attempts to detect DD-peptidase activity in HMM enzymes that have been solubilized from the bacterial membrane by removal of anchor peptides have not been successful.7,8 It is these solubilized enzymes, however, that have yielded most of the available crystal structures.2,5 It now seems likely that specific in vivo activity of HMMB DD-peptidases requires the close proximity of an HMMA enzyme, peptidoglycan and, probably, other proteins.6,9,10 Studies of the details of these enzymes activities are important © XXXX American Chemical Society

Received: December 1, 2017 Revised: February 14, 2018

A

DOI: 10.1021/acs.biochem.7b01209 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

■

MATERIALS AND METHODS Materials. The purified DD-peptidase of S. pneumoniae PBP2x was a gift from Dr. Otto Dideberg of CNRS-CEM, Grenoble, France. The following chemicals were purchased commercially and used as received: D-alanine (Sigma-Aldrich), D-alaninamide hydrochloride (Chem-Impex), glycine (SigmaAldrich), D-norleucine (Sigma-Aldrich), D-lysine (Sigma-Aldrich), N-benzyloxycarbonyl-D-aspartic acid (15, Chem-Impex), N-benzyloxycarbonyl-D-glutamic acid (12, Chem-Impex), N-benzyloxycarbonyl-D-glutamine (14, Chem-Impex), N-benzyloxycarbonyl-L-glutamic acid (13, Chem-Impex), Nα,Nεdiacetyl-L-Lys-D-Ala-D-Ala (8, Sigma-Aldrich), Nα-Ac-L-Lys-DAla-D-Ala (10, Sigma-Aldrich), N-α-Ac-L-Lys-D-Ala-D-Lys(Ac) (7, New England Peptide), benzylboronic acid (21, Alfa Aesar), 5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB, Aldrich), benzylpenicillin sodium salt (23, Sigma-Aldrich), cefuroxime sodium salt (24, Sigma-Aldrich). N-(Phenylacetyl)glycyl-Dthiolactate 2,17 3-(D-cysteinyl)-propanoyl-D-alanyl-D-thiolactate 6,18 N-(phenylacetyl)glycyl-D-alanine 9,19 N-(phenylacetyl)-D-alanyl-D-alanine 11,19 6-(dansylamido)penicillanic acid 22,20 and the boronic acids 18 and 2021 were synthesized in this laboratory. N-(Benzyloxycarbonyl)-D-α-aminoadipic acid 16 was prepared by the method of Hachisako et al.,22 disodium benzylpenicilloate 17 by the alkaline hydrolysis of benzylpenicillin as described by Mozingo and Folkers,23 the boronic acid 19 by the method of Prati and co-workers,24 and the thiolactates 3−5 by standard methods of peptide synthesis, as described in Supporting Information. Analytical and Kinetic Methods. All kinetics measurements were carried out in 10 mM phosphate buffer at pH 7.0 and 37 °C and containing 0.2 M NaCl. Absorption spectra and spectrophotometric reaction rates were measured with HewlettPackard 8452A and 8453 spectrophotometers. Substrate Kinetics. Enzyme-catalyzed reactions were monitored spectrophotometrically, at 240 nm for 2−6 and at 230 nm for 7 and 8. Steady-state kinetics parameters for the reactions of 2 were obtained from measurements of initial rates as a function of substrate concentration and analyzed by a nonlinear least-squares procedure to fit the Michaelis−Menten equation. Total progress curves for reactions of 3−6 in the absence and presence of the enzyme, the former to determine the rate constants, k0, of spontaneous hydrolysis, were analyzed by means of the Dynafit program25 in terms of Scheme 1. Enzyme concentrations were 0.26 μM in these experiments. Spectrophotometric measurements were also used to obtain initial rates of reactions of 2 and 3 in the presence of various amino acids and amino amides. These data were analyzed

Scheme 1. Kinetic Mechanism for Turnover of Thiodepsipeptides by PBP2x

in terms of Scheme 2 by means of the Dynafit program.25 In Scheme 2, ES and E-S represent noncovalent and covalent Scheme 2. Kinetic Mechanism for Competing Hydrolysis and Aminolysis of Thiodepsipeptides by PBP2x

(acyl-enzyme) complexes of enzyme and substrate, respectively, P1 and P2 represent the hydrolysis products, and Q represents the aminolysis product (see Scheme 4, below). 1 H NMR spectra of a reaction mixture containing 3 (8.1 mM), 2 D-lysine (20 mM), and PBP2x (0.04 μM) in H2O containing 16.6 mM phosphate buffer, pD = 7.9, were taken at zero time after enzyme addition and after 24 h. Initial rates of reaction of 2 in the presence of methanol (0−2.5 M) were also determined spectrophotometrically at 240 nm, as described above. The increase of rate with methanol concentration was analyzed by means of eq 1, as previously described.26 The rate constants of eq 1 have the same meaning as in Scheme 2, but where A = MeOH. The value of S0/Km was 5.38, and it was assumed that k2 ≫ k3. In eq 1, v and v0 represent the initial rates in the presence and absence of methanol, respectively, α = (k2/k3) + [H2O]0, β = [H2O] + (k4/k3)[MeOH], [H2O]0 is the water concentration in the absence of methanol (55.56 M), and [H2O] = 55.56−2.25 [MeOH], where 2.25 is the ratio of molar volumes of methanol and water. A nonlinear least-squares fit of the data to this equation (not shown) gave a value for k4/k3 and thus k4. v / v0 =αβ(1 + S0 /K m)/{αβ + [H 2O]0 (k 2 /k 3 + β)(S0 /K m)} (1)

Inhibition Kinetics. Enzyme-catalyzed reactions were monitored spectrophotometrically with 2 and 3 as the substrates. B

DOI: 10.1021/acs.biochem.7b01209 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry The substrate (50 μM, 2; 200 μM, 3) was incubated with the different inhibitors at appropriate concentrations (0−50 mM) in buffer. Enzyme (final concentration 0.26 μM) was then added and initial rates of hydrolysis of the substrate were monitored spectrophotometrically at 240 nm. In several cases (12−21), DTNB was used as a thiol leaving group reporter and the reaction was monitored at 412 nm. The measured initial rates were plotted versus inhibitor concentration. The data were analyzed by their fit to eq 2, where competitive inhibition was assumed, by a nonlinear least-squares procedure; values of Km of 9.3 μM and ≫200 μM for 2 and 3, respectively, were employed. v = v0(K m +[S])/(K m(1 + [I]/ K i) + [S]

N-benzyloxycarbonyl-D-glutamate (12, 0−15 μL, 0−13 mM) and the change in fluorescence emission intensity monitored.

■

RESULTS AND DISCUSSION PBP2x catalyzed the hydrolysis of several small molecule thioesters 2−6, releasing the thiol leaving group, as observed previously for 1,16 and shown in Scheme 4. Figure 1A, for example, shows the plot of initial rates of hydrolysis of 2 as a function of the concentration of 2. From these data, kcat and Km values for 2 could be determined. The value of the rate constant for spontaneous hydrolysis of 2 was found to be 1.1 × 10−4 s−1. The enzyme kinetic parameters that we obtained for 2 were quite similar to those for a related compound S2c (1: Ar = Ph, R2 = H, R3 = Me) described in the literature (kcat = 0.4 s−1, Km = 130 μM, kcat/Km = 3200 M−1 s−1).16 Thioesters 3−6 were hydrolyzed less efficiently than 2, with high Km values; e.g., the Km value for 3 was in excess of 5 mM and that of 4 greater than 2 mM. Consequently, kcat/Km values for 3−6 were obtained from exponential total progress curves at the concentrations employed, e.g., as shown for 3 in Figure 1B. As demonstrated previously,7,16 the hydrolysis of peptidoglycan-mimetic peptides 7 and 8 was not catalyzed at measurable rates by PBP2x. Steady state data thus obtained for compounds 2−8 are presented in Table 1. The reaction of 2 with PBP2x is accelerated by D-amino acids such as D-alanine (Figure 2). This phenomenon is commonly observed with DD-peptidases, both with analogues of 216,27,28 and also with peptidoglycan-mimetic thioesters such as 3−6.18,29 These data are customarily interpreted in terms of Schemes 2 and 4, where an acyl-enzyme E-S is formed that can be either hydrolyzed or aminolyzed; P1 and P2 represent the hydrolysis products and P1 and Q the aminolysis products. In cases where hydrolysis of E-S is rate-determining under substrate saturation conditions, the overall reaction is accelerated by addition of a D-amino acid (RNH2), such as observed for 2 in Figure 2. Thus, in the hydrolysis of 2, catalyzed by PBP2x, breakdown of the acyl-enzyme intermediate is rate-determining. From fitting of the data of Figure 2 to Scheme 2, a value of k4 = 1.3 × 102 M−1 s−1 was obtained. Values of k4 for various nucleophiles are presented in Table 2. Little specificity is seen between the reported amino acids. Replacement of the carboxylate, as in D-alaninamide, also had little effect, probably because the leaving group site (R′) of PBP2x must be bispecific, to accommodate a D-alanine residue in the acylation step of cross-linking and the N-terminus of 29,30 L-lysine in deacylation. Qualitatively similar data for D-alanine have been described in the literature for an analogous thiolester substrate.16 Reaction of 2 in the presence of PBP2x was also accelerated by methanol. These data were treated as described in Materials and Methods to yield a value of k4 of 0.11 M−1 s−1, smaller than those of the amino acids, as expected. The PBP2x-catalyzed reaction of 3, however, was not accelerated by D-amino acids, e.g., D-alanine (see Figure 2). Furthermore, a 1H NMR spectrum revealed only the hydrolysis product (Supporting Information, Figure S1). These results suggest that PBP2x catalyzes hydrolysis of 3 without formation of an acyl-enzyme intermediate, by direct water attack presumably. The same result was obtained with 4 (data not shown).

(2)

In eq 2, v and v0 represent the initial velocity in the presence and the absence of the inhibitor, respectively. Inhibition by the boronic acids (18−21) was studied similarly. The nature of the inhibition of PBP2x by peptide 7 was studied in more detail at three concentrations of 2 (19, 31, and 93 μM), with the concentration of 7 varied from 0 to 4.25 mM. The data were fitted by the use of the Dynafit program25 to Scheme 3 where Ki is the competitive inhibition constant of the observed inhibition. Scheme 3. Kinetic Mechanism for Competitive Inhibition of PBP2x

Inactivation of PBP2x by β-lactams. The activity of PBP2x (0.13 μM) was assessed against 2 and, separately, 3, after incubation (300 s) of the enzyme with benzylpenicillin (23, 20 μM) and, separately, cefuroxime (24, 20 μM). Initial rates of reaction of 2 and 3 were measured spectrophotometrically at 240 nm. These experiments were repeated where DTNB (1.0 mM) was included in the activity measurements, and the reactions were monitored at 412 nm. In all cases, control experiments were performed in the absence of β-lactams. 6-Dansylamidopenicillanic acid labeling and titration. An aliquot of a solution of 6-dansylamidopenicillanic acid was added to a solution of PBP2x, both in 10 mM phosphate buffer at pH 7.0 containing 0.2 M NaCl. Final concentrations of 6-dansylamidopenicillanic acid and PBP2x were 50 μM and 7.2 μM, respectively, in a volume of 0.9 mL. The mixture was allowed to react at room temperature for 2 h, after which it was passed down a Biogel 6DG column (1 × 20 cm) and eluted with the phosphate buffer in 0.5 mL fractions. Fluorescent-labeled protein was detected in fractions 11−13 and identified by absorption maxima at 277 nm (ε277 nm = 61 800 M−1 cm−1)16 and 340 nm. The fractions also exhibited dansyl fluorescence: λex = 340 nm, λem = 522 nm. The original 6-dansylamidopenicillanic acid displayed λex = 340 nm and λem = 556 nm. A sample of the labeled protein (100 μL, 0.1 μM) was titrated with

Scheme 4. Chemistry of Competing Hydrolysis and Aminolysis of Thiodepsipeptides by PBP2x

C

DOI: 10.1021/acs.biochem.7b01209 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 2. Initial rates of total reaction (hydrolysis plus aminolysis) of 2 (500 μM, ○) and 3 (200 μM, □) in the presence of D-alanine at varying concentration and of PBP2x (0.26 μM). The points are experimental and the lines represent fits of the data to Scheme 2 (see text).

Table 2. Rate Constants for Acylation of Nucleophiles by 2, Catalyzed by PBP2x acyl acceptor D-alanine D-alaninamide D-lysine D-norleucine

glycine methanol water (k3/55.56)

k4 (M−1 s−1) 130 ± 15 90 ± 30 190 ± 50 140 ± 20 80 ± 30 0.11 ± 0.01 0.0043 ± 0.0002

When the data were fitted to a mixed inhibition model, values of Ki of 1.6 ± 0.3 mM and Kis of 50 ± 70 mM were obtained. The peptide 7 was, however, a much poorer inhibitor of the hydrolysis of 3 with a Ki value of 43 ± 14 mM. If it is assumed, most reasonably, that 2 reacts at the conventional DD-peptidase active site, forming an acyl-enzyme by acylation of Ser337, peptide 7 for competitive inhibition would most likely bind there also. Since the presence of 7 at 1.5 mM does not significantly inhibit the hydrolysis of 3, one must most simply conclude that the hydrolysis of 3 occurs at another site, where, as shown above, an acyl-enzyme intermediate is not involved. Thus, 7 must bind most strongly to the enzyme at the classical active site where it inhibits reaction of 2 and less strongly at a second site where it inhibits reaction of 3. Hydrolysis of 2 and 3 are

Figure 1. (A) Spectrophotometric (240 nm) initial rates of hydrolysis of 2 catalyzed by PBP2x (0.26 μM) as a function of the concentration of 2. The points are experimental and the line represents a nonlinear least-squares fit of the data to the Michaelis−Menten equation. (B) Spectrophotometric (240 nm) progress curves showing hydrolysis of 3 (200 μM) in the absence (□) and presence (○) of PBP2x (0.26 μM). The points are experimental, and the lines represent fits of the data to Scheme 1 (see text).

The peptide 7 was studied as an inhibitor of thioester hydrolysis. Peptide 7 appeared to act as a competitive inhibitor of the hydrolysis of 2 (Figure 3) with a Ki value of 1.5 ± 0.2 mM.

Table 1. Steady State Parameters for Hydrolysis of Substrates by PBP2x substrate 2 3 4 5 6 7 8 a

k0 (s−1) −4

1.1 × 10 1.54 × 10−5 3.1 × 10−5 2.8 × 10−5 2.6 × 10−5 NRa NR

Km (μM)

kcat (s−1)

9.3 ± 1.6 >200 >200 >200 >200

0.24 ± 0.01 >0.07 >0.11 >0.03 >0.02 NR NR

kcat/Km (M−1 s−1) 2.53 × 104 (3.4 ± 0.1) × (5.55 ± 0.04) (1.45 ± 0.02) (0.83 ± 0.06)

102 × 102 × 102 × 102

NR = No reaction observed in the time frame of the experiments. D

DOI: 10.1021/acs.biochem.7b01209 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

peptides 7, 8, and 10 seem to bind most tightly while to site 2 (inhibitors of the hydrolysis of 3), 9, the peptide analogue of 2, appears to bind more tightly. Of the N-benzyloxycarbonyl-Ddicarboxylic acids 12, 15, and 16, the glutamate 12 binds more strongly to both sites than the shorter chain aspartate 15 and the longer chain adipate 16. N-Benzyloxycarbonyl-L-glutamate 13 binds more weakly to both sites than the D-enantiomer. Finally, the effects of β-lactams on turnover of 2 and 3 by PBP2x were investigated. β-Lactams are well-known, of course, to covalently inhibit the DD-peptidase site.12,16 Benzylpenicillin inactivates the DD-peptidase site of PBP2x rapidly (t1/2 ≈ 0.6 s at 20 μM benzylpenicillin concentration)12 and cefuroxime would be expected to do so also.12,16 Strikingly, however, we found that both benzylpenicillin (23, 20 μM) and cefuroxime (24, 20 μM) inactivated PBP2x toward turnover of 3 as well as that of 2 (Table 4). If it is assumed that β-lactams, and 3, do not acylate the second site (supported by the crystal structure of the PBP2x/cefuroxime crystal structure; see discussion below), this inactivation of the enzyme toward 3 must derive from a negative cooperative interaction between the sites when the DD-peptidase site is acylated by a β-lactam. With respect to this point, however, it is interesting that acylation of the DD-peptidase site by the dansylpenicillin 22 has little effect on the noncovalent binding of 12 (see above). Boronic acids are known to reversibly but covalently inhibit DD-peptidases by formation of anionic tetrahedral adducts with the active site serine hydroxyl group.31,32 We therefore examined the effects of the boronic acids 18−21 on the rates of hydrolysis of 2 and 3 by PBP2x. As seen in Table 2, the boronic acids inhibited the hydrolysis of both substrates, but as with compounds 9−17, to different extents and more weakly in the case of 3. It should be noted that competitive inhibition was assumed in all cases although this was not proven. If the boronic acids inhibited the hydrolysis of 3 in a cooperative allosteric/noncompetitive fashion after binding to the DD-peptidase active site, the apparent Ki values should be equal to or less than those for inhibition of 2. Apparently then the boronic acids also bind at the second site, perhaps noncovalently as anions or by formation of a covalent adduct, possibly with Thr665 (see below). The stronger inhibition demonstrated by the more electrophilic 18−20 than by 21 would support either of these possibilities. Overall, therefore, there seems to be good reason to believe that 2 and 3 are largely hydrolyzed at separate sites on PBP2x and can be separately inhibited at those sites. Small amounts of hydrolysis of 2 at the second site and of 3 at the DD-peptidase site, however, cannot at this stage be discounted. The absence of significant hydrolysis of 3 at the DD-peptidase site is quite surprising but is reminiscent the low reactivity of β-lactams bearing peptidoglycan-mimetic side chains with LMMA and HMMB DD-peptidases.33 The reason for this is not known but probably relates to the need for extended induced fit phenomena at these active sites on interaction with specific substrates, as seen, for example, at a HMMB1 active site.34 The thioesters 4−6 are probably also hydrolyzed preferentially at the second site (the PBP2x catalyzed rate of reaction of 4 was not increased in the presence of 200 mM D-alanine). Structural Interpretation. The kinetics results described above suggest that certain substrates such as 2−6 and peptides of similar structure interact with PBP2x not only at the DD-peptidase site but also at a second site. In this regard, it is interesting to recall that the crystal structure of PBP2x in the presence of the β-lactam cefuroxime shows the presence of two ligands.12 At the DD-peptidase site, cefuroxime is found covalently bound to

Figure 3. Inhibition of the hydrolysis of 2 (19 μM, ○; 31 μM, □; 93 μM, ◇) catalyzed by PBP2x (0.26 μM) by various concentrations of 7 and presented as a Dixon plot. The points are experimental and the lines represent a fit of the data to Scheme 3 (see text).

also inhibited by other peptides, 8−16, and by benzylpenicilloate 17, with the results reported in Table 3. Table 3. Peptide Inhibition of Thioester Hydrolysis Catalyzed by PBP2x thioester substrate, Ki (mM) peptide

2

3

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1.5 ± 0.2 1.6 ± 0.4 2.3 ± 0.4 1.4 ± 0.4 2.7 ± 0.7 3.8 ± 0.3 16 ± 4 4.2 ± 0.2 23 ± 6 9±1 1.8 ± 0.2 0.19 ± 0.02 0.68 ± 0.10 0.47 ± 0.10 >1.0

43 ± 14 31 ± 14 8±3 47 ± 13 20 ± 5 17 ± 2 >50 19 ± 4 56 ± 8 >50 >5 0.68 ± 0.16 1.3 ± 0.2 0.7 ± 0.2 >1.0

Inhibition of hydrolysis of 3 by 8−16 is considerably weaker, in general, than that of 2, suggesting that 8−16 must also bind, like 7, at the putative second site, as well as at the active site. This point is confirmed by the results shown in Figure 4, where addition of N-benzyloxycarbonyl-D-Glu (12) quenches the dansyl group fluorescence of PBP2x that had been inhibited by reaction with 6-dansylamidopenicillanic acid 22 to acylate and fluorescently label the DD-peptidase active site. Determination of the dissociation constant for 12 from the data of Figure 4 yielded a value of (15 ± 3) mM, in good agreement with the value for inhibition of hydrolysis of 3 from the kinetics experiment (Table 3). Of the ligands of site 1 (inhibitors of the hydrolysis of 2, the DD-peptidase site), the aminopimelyl E

DOI: 10.1021/acs.biochem.7b01209 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 4. Titration of dansyl-labeled PBP2x (0.1 μM) with 12, monitored by fluorescence emission on excitation at 340 nm. The concentrations of 12 are given by the numbers associated with each curve and are in units of mM.

and the protein are between the cefuroxime carboxylate and the side chain terminal cation of Arg426. The side chain of the latter has moved from a more extended conformation in the apoenzyme to allow cefuroxime binding and this interaction to occur. The cation of Arg426 in the complex is also hydrogen bonded to the carboxylate groups of Glu651 and Asp428. Otherwise, the second binding site seems to be mainly hydrophobic in nature. It is possible that this site has a role in the interactions of PBP2x with peptidoglycan or with other cell surface proteins, for example, PBP1a.35 It is interesting to note that the C-terminal domain of PBP2x is constructed from two PASTA domains which are known to bind peptidoglycan and β-lactams.36,37 The second binding site on PBP2x does not appear to be a catalytic site since the cefuroxime molecule is intact, and amino acid residues that classically catalyze peptide hydrolysis are absent. Nonetheless this site is certainly a candidate for the second peptide binding site that we have detected by inhibition kinetics in our studies of turnover of 3 by PBP2x, as described above. Slow hydrolysis of 3 could occur simply by means of a protein surface water molecule, although none are observed in the vicinity of the cefuroxime binding site in the crystal structure of either the apoenzyme or the cefuroxime complex at the attainable resolution. It is possible, however, that a water molecule

Table 4. Inactivation of PBP2x Towards Both 2 and 3 by β-Lactams substrate (conc., μM) 2 2 2 3 3 3

(50) (50) (50) (200) (200) (200)

β-lactam (conc., μM) benzylpenicillin 23 (20) cefuroxime 25 (20) benzylpenicillin 23 (20) cefuroxime 25 (20)

v0 (ΔAs−1)a 1.3 × 10−4 (5.6 × 10−4)