Catalytic mechanism and cofactor preference of dihydrodipicolinate reductase from methicillin-resistant Staphylococcus aureus

Archives of Biochemistry and Biophysics 512 (2011) 167–174

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi

Catalytic mechanism and cofactor preference of dihydrodipicolinate reductase from methicillin-resistant Staphylococcus aureus Sudhir R. Dommaraju a,1, Con Dogovski a,1, Peter E. Czabotar b,c, Lilian Hor a,d, Brian J. Smith b,c, Matthew A. Perugini a,⇑ a

Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, 3010 Victoria, Australia The Walter and Eliza Hall Institute of Medical Research, Parkville, 3052 Victoria, Australia Department of Medical Biology, The University of Melbourne, Parkville, 3010 Victoria, Australia d School of Chemistry, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, 3010 Victoria, Australia b c

a r t i c l e

i n f o

Article history: Received 5 May 2011 and in revised form 8 June 2011 Available online 16 June 2011 Keywords: Antibiotic Antibiotic resistance Dihydrodipicolinate reductase Lysine biosynthesis MRSA Staphylococcus aureus

a b s t r a c t Given the rapid rise in antibiotic resistance, including methicillin resistance in Staphylococcus aureus (MRSA), there is an urgent need to characterize novel drug targets. Enzymes of the lysine biosynthesis pathway in bacteria are examples of such targets, including dihydrodipicolinate reductase (DHDPR, E.C. 1.3.1.26), which is the product of an essential bacterial gene. DHDPR catalyzes the NAD(P)H-dependent reduction of dihydrodipicolinate (DHDP) to tetrahydrodipicolinate (THDP) in the lysine biosynthesis pathway. We show that MRSA–DHDPR exhibits a unique nucleotide specificity utilizing NADPH (Km = 12 lM) as a cofactor more effectively than NADH (Km = 26 lM). However, the enzyme is inhibited by high concentrations of DHDP when using NADPH as a cofactor, but not with NADH. Isothermal titration calorimetry (ITC) studies reveal that MRSA–DHDPR has 20-fold greater binding affinity for NADPH (Kd = 1.5 lM) relative to NADH (Kd = 29 lM). Kinetic investigations in tandem with ITC studies show that the enzyme follows a compulsory-order ternary complex mechanism; with inhibition by DHDP through the formation of a nonproductive ternary complex with NADP+. This work describes, for the first time, the catalytic mechanism and cofactor preference of MRSA–DHDPR, and provides insight into rational approaches to inhibiting this valid antimicrobial target. Ó 2011 Elsevier Inc. All rights reserved.

Introduction Staphylococcus aureus is a Gram-positive facultative anaerobe pathogen that commonly colonizes the anterior nares, respiratory system, and urinary tract of the host [1]. The organism can also enter open wounds and is thus capable of causing multi-systemic life-threatening infections in humans [2]. Methicillin-resistant S. aureus (MRSA)2 is a strain that has developed antibiotic resis-

⇑ Corresponding author. Address: Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, 30 Flemington Road, Parkville, Victoria 3010, Australia. Fax: +61 3 9348 1421. E-mail address: [email protected] (M.A. Perugini). 1 These authors contributed equally to this work. 2 Abbreviations used: (S)-ASA, aspartate semi-aldehyde; DHDP, dihydrodipicolinate; DHDPR, dihydrodipicolinate reductase; DHDPS, dihydrodipicolinate synthase; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; ITC, isothermal titration calorimetry; MRSA, methicillin-resistant Staphylococcus aureus; NAD+, nicotinamide adenine dinucleotide (oxidized); NADH, nicotinamide adenine dinucleotide (reduced); NADP+, nicotinamide adenine dinucleotide phosphate (oxidized); NADPH, nicotinamide adenine dinucleotide phosphate (reduced); 2,6-PDC, 2,6-pyridine dicarboxylate; THDP, tetrahydrodipicolinate; v, initial velocity. 0003-9861/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2011.06.006

tance to all penicillin-based antibiotics, including methicillin [3]. Two major variants of MRSA have been described, namely, hospital-acquired MRSA (HA-MRSA) and community-acquired MRSA (CA-MRSA) [4]. Currently, all available b-lactam antibiotics are ineffective against HA-MRSA and CA-MRSA strains [5,6]. Therefore, vancomycin is employed as the antibiotic of last resort for treating MRSA infections, but reports of resistance to vancomycin are nonetheless emerging at a rapid rate [7,8]. As a result, the morbidity and mortality rates of MRSA infections are increasing worldwide [9], and there is thus an urgent need to discover new antibiotics for the treatment of MRSA infections and an equally urgent need to characterize novel antibiotic targets. One such target is the lysine biosynthesis pathway (also known as the diaminopimelate pathway) in bacteria [7,10]. The products of the pathway include meso-diaminopimelate and lysine, that are essential building blocks for the synthesis of the bacterial cell wall, housekeeping proteins, and virulence factors [7,10]. Accordingly, the enzymes that catalyze critical steps in the pathway serve as excellent antimicrobial targets [7,10–12]. This is further validated given that humans do not contain the enzymatic machinery used to synthesize lysine, instead acquiring this essential amino acid from dietary sources [7,10].

168

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

Fig. 1. Schematic representation of the lysine biosynthesis pathway showing the DHDPS and DHDPR catalyzed reactions.

In recent years there has been heightened interest in studying the enzymatic machinery of bacterial lysine biosynthesis, particularly in characterizing the kinetic properties, regulation, and threedimensional structures of both wild-type and mutant enzymes [13–25]. Studies have also focused on determining the essentiality of the enzymes to bacteria [10]. For example, in a landmark study by Kobayashi et al. [26], the authors systematically knocked out all 4118 genes comprising the Bacillus subtilis genome and showed that only 271 genes were essential for viability, including the dapB gene that encodes the enzyme dihydrodipicolinate reductase (DHDPR). The first committed step of the lysine biosynthesis pathway is catalyzed by dihydrodipicolinate synthase (DHDPS), which catalyzes the condensation of pyruvate and aspartate semi-aldehyde [(S)-ASA] forming the unstable heterocycle, HTPA (Fig. 1). HTPA is then non-enzymatically dehydrated to form dihydrodipicolinate (DHDP), which is subsequently reduced by DHDPR to form tetrahydrodipicolinate (THDP) using NAD(P)H as the reductant (Fig. 1) [27]. DHDPR has been characterized from a number of bacterial species, including Bacillus cereus, Bacillus megaterium, B. subtilis, Corynebacterium glutamicum, Escherichia coli, Methylophilus methylotrophus, Mycobacterium tuberculosis, and Thermotoga maritima [28–35]. The three-dimensional structure of this enzyme has been determined from E. coli, M. tuberculosis and T. maritima in the absence and/or presence of cofactor [29–31,36]. Like most nucleotide-dependent reductases, dual specificity and/or preference for one of the nucleotides NADH and/or NADPH has been reported for DHDPR from different organisms [28–30]. Moreover, DHDPR from E. coli and M. tuberculosis have been shown to have dual cofactor specificity, with E. coli DHDPR having a two-fold greater affinity for NADH over NADPH [28,31,32]. In contrast, DHDPR from T. maritima possesses significantly greater affinity for NADPH compared to NADH, but is inhibited by the latter at higher concentrations [29]. Whereas, DHDPR from M. tuberculosis is able to utilize both cofactors with equal efficiency [30]. Given that DHDPR from diverse bacterial species demonstrate different cofactor preferences, the aim of this study was to kinetically and thermodynamically characterize the enzymatic mechanism and cofactor preference of MRSA–DHDPR. The results presented in this study demonstrate that MRSA–DHDPR catalyzes the reduction of DHDP to THDP using a compulsory-order ternary complex mechanism with cofactor preference for NADPH. However, the enzyme displays substrate inhibition at high DHDP concentrations when using NADPH as the cofactor. This study offers important insight into rational drug design strategies for inhibiting a novel antibiotic target from methicillin-resistant S. aureus.

Cloning, expression and purification of MRSA–DHDPR and E. coli DHDPS The dapB gene from MRSA strain 252 was amplified by PCR and cloned into the pET11a expression vector. Recombinant protein was produced in E. coli BL21 (DE3) cells, and purified to yield a >95% homogeneous enzyme preparation as described in [14]. Dihydrodipicolinate synthase (DHDPS) from E. coli (required for the coupled enzyme kinetic assay) was expressed and purified to yield >95% pure enzyme as described in [23]. Coupled enzyme kinetics assay The coupled assay employing DHDPS and DHDPR [28,29] was measured at a wavelength of 340 nm at 30 °C using a 1 cm acrylic cuvette and a Cary 4000 spectrophotometer. The assay was performed in HEPES buffer, pH 8.0, at a final concentration of 100 mM with a final assay volume of 800 ll. The standard constituents of the assay include 70 nM MRSA–DHDPR (final assay concentration), an excess amount of E. coli DHDPS as the coupling enzyme, NADPH or NADH as the cofactor, (S)-aspartate semi-aldehyde [(S)-ASA], and pyruvate. E. coli DHDPS was considered to be in excess when addition of further amounts of the enzyme did not result in an increase to the initial reaction rate (i.e. final assay concentration of 1.6 lM). The assay mixture was incubated at 30 °C for approximately 3 min with E. coli DHDPS and initiated by the addition of MRSA–DHDPR. All assays were performed in duplicate to ensure that Km and Vmax values were reproducible. To determine the kinetic parameters for the substrate (i.e. DHDP), (S)-ASA concentration was varied from 0.025 mM to 1.0 mM, with pyruvate maintained at a constant concentration of 2 mM, and with the concentration of cofactor (NADH or NADPH) fixed at either a concentration of 20 lM, 40 lM, 60 lM, 80 lM or 100 lM. The coupled assay is accurate when using cofactor (i.e. NADH or NADPH) concentrations P20 lM and (S)-ASA concentrations P0.025 mM. Initial rates for MRSA–DHDPR using both NADH and NADPH were determined from the change in absorbance at 340 nm (DA340nm) as a function of time using e340 [NAD(P)H] = 6220 M1 cm1 over the linear portion of the A340nm versus time profiles. Kinetic data analysis Data were analyzed by nonlinear regression using the program ENZFITTER (Biosoft, Cambridge, UK). Eq. (1) was employed to fit a compulsory-order ternary-complex mechanism, and Eq. (2) to fit a compulsory-order ternary-complex mechanism with substrate inhibition by DHDP.

Materials and methods

v ¼ ðV max  a  bÞ=ððK iA  K mBÞ þ ðK mB  aÞ þ ðK mA  bÞ þ ða  bÞÞ

Materials

v ¼ ðV max  a  bÞ=ððK iA  K mB Þ þ ðK mB  aÞ þ ðK mA  bÞ

NAD+, NADH, NADP+, NADPH, 2,6-pyridine dicarboxylate and sodium pyruvate, were purchased from Sigma–Aldrich (Australia), HEPES was obtained from Ajax Finechem Pty., Ltd. (Australia), and (S)-aspartate semi-aldehyde was synthesized as described in [37].

þ ða  bÞ  ð1 þ ðb=K siB ÞÞÞ

ð1Þ

ð2Þ

where v is the initial velocity, a is the co-factor [NAD(P)H] concentration, b is the substrate (DHDP) concentration, Vmax is the limiting maximal velocity/rate, KmB and KmA are the Michaelis-Menten constants for substrates a and b, respectively, KiA is the inhibition

169

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

constant of substrate a, and KsiB is a constant that defines the strength of substrate inhibition. Isothermal titration calorimetry (ITC) ITC was performed on a MicroCal VP-ITC unit based on optimization of established methods [38–43]. The protein samples were buffer exchanged prior to each experiment into 200 mM HEPES, pH 8.0 using a 5 ml Hi-Trap desalting column (GE Healthcare). The same buffer was used for blanking experiments, preparing ligand solutions and washing the VP-ITC instrument. All experiments were carried out in duplicate at 30 °C. Protein and ligand solutions were degassed prior to taking measurements using a ThermoVac vacuum device (MicroCal). For studies where NADH and NADPH were the titrant, the initial concentrations of protein and ligand used in the experiments were 80 lM and 760 lM, respectively. For experiments where 2,6-PDC was the titrant, the initial concentration of protein was 80 lM pre-incubated in the absence or presence of 160 lM NADPH or 160 lM NADP+. Each titration typically included a 26 injection protocol with 12 ll of ligand per injection spaced 180 sec apart from an injection syringe rotating at 310 rpm. The data were analyzed using MicroCal ORIGIN (version. 7.0) software in order to calculate the thermodynamic parameters: dissociation constant (Kd), reaction stoichiometry (n), enthalpy (DH) and entropy (DS). The heats of dilution obtained from titration of the ligands into the buffer were subtracted from the data for the protein sample before the data was fitted to the appropriate model. Results Determination of the enzyme kinetic mechanism of MRSA–DHDPR Recombinant MRSA–DHDPR was expressed and purified to homogeneity as previously described [14]. We, therefore, sought to study the kinetic properties of the recombinant enzyme and deduce the catalytic mechanism using the quantitative DHDPS– DHDPR coupled assay, initially using NADH as the cofactor [28,29]. Initial velocity studies were conducted by varying the concentration of (S)-ASA (and thus DHDP) at several fixed concentrations of NADH. The change in absorbance at 340 nm was carefully measured to ensure the initial reaction rate was obtained, which remained linear when less than 20% of the cofactor and substrate were turned over, corresponding to the first 10–20 s of the reaction. The kinetics of dual substrate enzyme-catalyzed reactions usually follow a ‘ternary complex’ or ‘ping-pong’ mechanism [44,45]. Indeed, previous studies demonstrate that DHDPR from E. coli and M. tuberculosis follow a ternary complex mechanism with the binding of the cofactor preceding the binding of substrate, which is referred to as a ‘compulsory order’ ternary complex mechanism [28,30]. Primary plots of MRSA–DHDPR kinetic data employing NADH as the cofactor were constructed (data not shown). These plots show a characteristic pattern of lines that intersect at a common point to the left of the abscissa axis indicating a sequential ternary complex mechanism [44,45]. The data obtained for NADH was subsequently used to construct an initial velocity versus substrate (i.e. Michaelis–Menten) plot (Fig. 2A). The v versus [DHDP] plots generated at various fixed NADH concentrations show a typical rectangular hyperbola relationship (Fig. 2A). These data were globally fitted (Fig. 2A, solid lines) to a ternary complex mechanism, which yielded an excellent fit (R2 = 0.98) resulting in a Vmax of 1.4 ± 0.1 lM s1 and Km values of 39 ± 2.0 lM and 26 ± 0.5 lM for DHDP and NADH, respectively (Table 1). Similarly, MRSA–DHDPR was kinetically characterized utilizing NADPH as the cofactor. In contrast to the NADH results, primary plot analysis

Fig. 2. Kinetic analyses of MRSA–DHDPR using (A) NADH and (B) NADPH as the cofactor. The initial rate (v) is plotted as a function of [DHDP] at varying NADH (Panel A) and NADPH (Panel B) concentrations of 20 lM (}), 40 lM (d), 60 lM (s), 80 lM (j), and 100 lM (h). The experiments were performed in duplicate with the error bars shown representing the standard error of the replicate measurements. The solid lines represent the nonlinear least squares best fits to a ternary complex mechanism without substrate inhibition (Panel A) and a ternary complex mechanism with substrate inhibition (Panel B).

Table 1 Kinetic parameters of MRSA–DHDPRa. Cofactor

Km (lM)

Km (DHDP) (lM)

Vmax (lM s1)

KsiB (DHDP) (lM)

NADH NADPH

26 ± 0.5 12 ± 1.0

39 ± 2.0 28 ± 1.5

1.4 ± 0.1 1.4 ± 0.1

– 280 ± 36

a Enzyme assays were performed in duplicate and thus all values reported are expressed as mean ± standard error.

with respect to NADPH shows a series of lines that do not have a common point of intersection, which is indicative of a compulsory-order ternary complex mechanism with substrate inhibition (data not shown) [44,45]. Accordingly, the v versus [DHDP] profiles generated at various fixed NADPH concentrations were fitted to a compulsory-order ternary complex mechanism incorporating substrate inhibition by DHDP (Eq. (2)) (Fig. 2B). The global nonlinear least squares best fit (Fig. 2B, solid lines) yielded a R2 = 0.96, Vmax = 1.4 ± 0.1 lM s1, Km (DHDP) = 28 ± 1.5 lM, Km (NADPH) = 12 ± 1.0 lM, and a KsiB (DHDP) = 280 ± 36 lM (Table 1). Despite the substrate inhibition evident at high DHDP concentrations, the lower Km obtained for NADPH compared to NADH suggests MRSA–DHDPR shows cofactor preference for the phosphorylated reductant. We, therefore, set out to measure the affinity of

170

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

MRSA–DHDPR for NADPH compared to NADH using isothermal titration calorimetry (ITC). Thermodynamic characterization of MRSA–DHDPR with respect to phosphorylated and non-phosphorylated cofactors Based on the kinetic parameters reported in Table 1, it was anticipated that MRSA–DHDPR would show tighter binding to NADPH compared to NADH given the 2-fold lower Km determined for the phosphorylated cofactor. ITC studies were thus performed titrating recombinant MRSA–DHDPR with NADH (Fig. 3A) or NADPH (Fig. 3B). The data was subsequently fitted to various models using nonlinear regression analyses, which resulted in best fits for both cofactors to a single site model yielding Kd values of 29 ± 1.8 lM for NADH and 1.5 ± 0.1 lM for NADPH (Table 2). The resulting thermodynamic parameters from the nonlinear least squares best fit indicate that NADPH binds with substantially greater enthalpy compared to NADH, but also exhibits greater entropic loss in binding. ITC studies with the substrate analog 2,6-PDC ITC studies reported in Fig. 3 and Table 2 demonstrate that MRSA–DHDPR can bind cofactor independent of the substrate, DHDP. This observation, along with kinetic studies described above (Fig. 2 and Table 1), suggests that the cofactor binds first in a compulsory-order ternary complex mechanism. We, therefore, set out to confirm the order of substrate binding to MRSA–DHDPR.

However, given that DHDP is an unstable transient substrate [27,28], a substrate analog was sought to employ in ITC studies. 2,6-pyridine dicarboxylate (2,6-PDC) has been shown to be a competitive inhibitor versus DHDP in previous studies with E. coli DHDPR [28]. This DHDP analog was thus used as the titrant in ITC studies against MRSA–DHDPR in the absence and presence of NADPH (Fig. 4). Results show that in the absence of cofactor, 2,6PDC does not bind to MRSA–DHDPR (Fig. 4A, Table 2). However when 2,6-PDC was titrated against NADPH bound MRSA–DHDPR (Fig. 4B), the resulting isotherm demonstrated a detectable change in free energy (DGs) yielding a Kd of 100 ± 17 lM (Table 2). These results demonstrate that an enzyme-cofactor complex forms prior to binding of the substrate analog 2,6-PDC, and thus confirms a compulsory-order ternary complex kinetic mechanism.

Calorimetric studies reveal the mode of substrate inhibition of MRSA– DHDPR Kinetic analysis shows that MRSA–DHDPR is inhibited by high concentrations of DHDP when utilizing NADPH as the cofactor (Fig. 2B). We, therefore, sought to investigate if the observed substrate inhibition is the result of non-productive ternary complex formation with the oxidized cofactor (i.e. NADP+). Accordingly, ITC studies were performed using 2,6-PDC as the titrant with MRSA–DHDPR pre-incubated in a molar excess of NADP+ (Fig. 5). The resulting nonlinear regression analysis best fit indicates that the substrate analog binds the enzyme pre-incubated with oxidized cofactor with a Kd of 13 ± 0.6 lM, (Table 2). This data con-

Fig. 3. Microcalorimetric titration of MRSA–DHDPR with (A) NADH and (B) NADPH. Raw data is shown in the top panel, with each peak corresponding to a single injection of the titrant. The integrated titration curve is plotted in the bottom panel showing heats of interaction (kcal/mole of injectant) as a function of molar ratio, fitted to a one site model using the ORIGIN software. Titrations were performed in duplicate. 80 lM MRSA–DHDPR was employed in the titrations shown with an initial NAD(P)H concentration of 760 lM.

171

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174 Table 2 Thermodynamic parameters for cofactor and substrate analog binding to MRSA–DHDPR obtained from fitting the data to a one-site binding modela.

a b

Sample

Titrant

Kd (lM)

DG° (kcal mol1)

DH° (kcal mol1)

DS° (cal mol1)

TDS° (kcal mol1)

MRSA–DHDPR MRSA–DHDPR MRSA–DHDPR NADPH bound MRSA–DHDPR NADP + bound MRSA–DHDPR MRSA–DHDPR MRSA–DHDPR

NADH NADPH 2,6-PDC 2,6-PDC 2,6-PDC NAD+ NADP+

29 ± 1.8 1.5 ± 0.1 NBb 100 ± 17 13 ± 0.6 NBb NBb

5.5 ± 0.7 6.4 ± 1.7

11 ± 0.5 16 ± 1.3

20 ± 1.1 33 ± 1.4

6.0 ± 0.3 10 ± 0.4

7.2 ± 1.4 6.2 ± 0.5

13 ± 1.1 33 ± 2.6

19 ± 1.1 90 ± 10

5.8 ± 0.3 27 ± 3.1

Titrations were performed in duplicate and thus all values reported are expressed as mean ± standard error. NB: no binding of titrant detected.

Fig. 4. Microcalorimetric titration of MRSA–DHDPR with (A) 2,6-PDC alone, and (B) 2,6-PDC + NADPH. Raw data is shown in the top panel, with each peak corresponding to a single injection of the titrant. The integrated titration curve is plotted in the bottom panel showing heats of interaction (kcal/mole of injectant) as a function of molar ratio, fitted to a one site model using the ORIGIN software. Titrations were performed in duplicate. The MRSA–DHDPR concentration employed in all titrations was 80 lM either in the absence of cofactor (panel A) or pre-incubated in the presence of 160 lM NADPH (panel B). The initial 2,6-PDC (titrant) concentration was 760 lM.

firms that substrate inhibition by DHDP is the result of non-productive ternary complex formation with NADP+ (Fig. 6). Titration of NAD(P)+ into the apo enzyme resulted in no detectable binding by ITC (Table 2), consistent with a capture mechanism illustrated in the Cleland plot (Fig. 6). This indicates that inhibition by DHDP results after release of the product (THDP), but before the oxidized cofactor (NADP+) dissociates from the enzyme. Discussion The overall aim of this study was to determine the catalytic mechanism and potential cofactor preference of dihydrodipicolinate reductase from methicillin-resistant S. aureus. We show that MRSA–DHDPR catalyzes the NAD(P)H-dependent reduction of

DHDP to form THDP using a compulsory-order ternary complex mechanism. However, kinetic (Fig. 2 and Table 1) and thermodynamic (Fig. 3 and Table 2) studies demonstrate that MRSA–DHDPR binds the phosphorylated cofactor (NADPH) with significantly greater affinity (20-fold) than the non-phosphorylated form (NADH). This suggests MRSA–DHDPR has evolved to utilize NADPH in preference to NADH, which is in contrast to E. coli DHDPR that has been shown to bind NADH with 4- to 5-fold greater affinity than NADPH [28,32]. Kinetic studies of DHDPR from M. tuberculosis show that the enzyme can utilize NADH and NADPH with equal efficiency [30]; whilst studies of DHDPR from T. maritima [29] suggest that this thermophilic enzyme, like MRSA–DHDPR, has higher affinity for NADPH. It has also been reported that T. maritima DHDPR is inhibited by a high concentration of NADH [29]; however

172

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

Fig. 5. Thermodynamic analysis of substrate analog binding. Calorimetric titration of 2,6-PDC into MRSA–DHDPR pre-incubated with NADP+. The MRSA–DHDPR concentration employed in the titration was 80 lM pre-incubated in the presence of 160 lM NADP+ using an initial 2,6-PDC (titrant) concentration of 760 lM.

an underlying mechanism for inhibition is not described. We show that MRSA–DHDPR is inhibited by high concentrations of DHDP when utilizing NADPH as the cofactor, forming a nonproductive ternary complex with oxidized NADP+ (Fig. 6). The structural mechanisms underpinning the apparent differences in cofactor binding and substrate inhibition are therefore of interest to this study. From the structures of DHDPR bound to NAD(P)H from E. coli [31,32] and M. tuberculosis [30] it has been demonstrated that the conserved nucleotide binding motif (GXXGXXG) and the acidic residue (D or E) located 19-20 residues downstream of the glycinerich region is important in binding the cofactor (Fig. 7A). Although the conserved acidic residue is present in the MRSA–DHDPR sequence (i.e. Glu31), it is interesting to note that the glycine-rich motif is incomplete, defined in the case of the MRSA–DHDPR as the sequence GXGXXN (Fig. 7). Despite this deviation from the consensus sequence, it was clear from kinetic (Fig. 2) and thermodynamic studies (Fig. 3) that MRSA–DHDPR binds and turns over

NAD(P)H. When the sequence comparison is extended to include a larger subset of bacterial DHDPR enzymes whose structures have not yet been determined (Fig. 7B), the MRSA–DHDPR sequence remains the only one containing an incomplete glycine-rich motif (Fig. 7). Furthermore, it has been shown by structural analyses of E. coli DHDPR bound to either NADH or NADPH that the conserved acidic residue (i.e. Glu38) and the adjacent basic residue (i.e. Arg39) play a role in accommodating the different cofactors [31,32]. Moreover, when NADH is bound to E. coli DHDPR, Glu38 is shown to hydrogen-bond with the O2’ and O3’ hydroxyls of the ribose moiety. However, when NADPH is bound, the O2’ phosphate is shown to interact electrostatically with Arg39 [31]. By contrast, structural studies of wild-type and mutant M. tuberculosis DHDPR enzymes [30] show that the equivalent of Glu38 and Arg39 are an acidic Asp33 and a neutral Ala34, respectively. The conserved acidic residue Asp33 of M. tuberculosis DHDPR plays the equivalent role of Glu38 in E. coli DHDPR when NADH is bound; however, the lysine residues (Lys9 and Lys11) located within the GXXGXXG motif of the M .tuberculosis enzyme (Fig. 7) are shown to electrostatically anchor the phosphate moiety when NADPH is bound [30]. These lysine residues thus serve the same function as Arg39 in the E. coli structure [30]. Interestingly, the MRSA– DHDPR sequence lacks a basic residue equivalent to Arg39 in E. coli DHDPR and also the lysine pair within the GXXGXXG motif as observed in M. tuberculosis DHDPR (Fig. 7). It is therefore unclear at the sequence level why the MRSA–DHDPR enzyme preferentially binds NADPH. However, it is possible that the cluster of non-conserved basic residues (i.e. R14, R17 and K22) downstream of the GXGXXN motif in MRSA–DHDPR sequence may play a role in anchoring the phosphate moiety. The three-dimensional structure of MRSA–DHDPR in complex with NAD(P)H, which to date has remained elusive, will confirm this hypothesis. Nevertheless, it is evident from the sequence alignment provided in Fig. 7 that the unique GXGXXN motif is likely to play a role in defining the cofactor preference of MRSA–DHDPR for NADPH (Figs. 2 and 3 and Tables 1 and 2). The cofactor preference observed for MRSA–DHDPR may have biological significance when one considers the metabolic properties of S. aureus. Given that this organism is a facultative anaerobe, it is likely that NADH is conserved primarily for ATP production via anaerobic glycolysis. Whereas, anabolic pathways that include reduction steps, such as the DHDPR-catalyzed reaction of the lysine biosynthesis pathway, may have evolved to utilize NADPH as a means to conserving the intracellular NADH pool. Indeed, it has been shown that azoreductase from S. aureus uses only NADPH as the reductant and cannot utilize NADH at all [46]. This supports the notion that NADH levels may be tightly regulated and thus conserved in S. aureus cells. Therefore, it is anticipated that a more thorough understanding of the utilization of NADH in S. aureus will better inform novel drug discovery programs. In addition, the thorough characterization of thermodynamic parameters governing nicotinamide-based cofactor binding to antibiotic targets from S. aureus, such as MRSA–DHDPR, can also be exploited for rational drug design. We show using ITC studies that the tighter binding of NADPH by MRSA–DHDPR (relative to NADH)

Fig. 6. A Cleland plot of the catalytic mechanism of MRSA–DHDPR employing NADPH as the cofactor depicting a compulsory order sequential ternary-complex mechanism incorporating substrate inhibition by non-productive ternary complex formation. Note: MR is the abbreviation for MRSA–DHDPR.

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

173

Fig. 7. Multiple sequence alignment of bacterial DHDPR. Alignment of N-terminal residues of DHDPR enzymes from various bacterial species encompassing the conserved nucleotide binding motif (GXXGXXG) and the ‘acidic residue’ (E/D). The sequences included in the alignment are – SA: Methicillin-resistant Staphylococcus aureus (query sequence), against (A) three DHDPR enzymes whose three-dimensional structures have been determined – EC: Escherichia coli (PDB ID 1ARZ), MTB: Mycobacterium tuberculosis (PDB ID 1C3V), and TM: Thermotoga maritima (PDB ID 1VM6); and (B) DHDPR sequences from other bacterial species whose structures have not yet been determined – NM: Neisseria meningitidis, NG: Neisseria gonorrhoea, PA: Pseudomonas aeuriginosa, CJ: Clostridium jejuni, CB: Clostridium botulinum, SP: Streptococcus pneumoniae, and BA: Bacillus anthracis.

is due to a 1.5-fold favored enthalpic contribution (Fig. 3 and Table 2). This suggests a greater network of hydrogen bonds and/or van der Waal interactions are involved in binding NADPH compared to NADH. Further, this is likely to be due to the phosphate moiety of NADPH, given that this polar functional group offers enhanced van der Waals and/or H-bonding potential. Although, the TDS contribution of binding either NADH or NADPH is negative (Table 2), thus suggesting hydrophobic interactions play only a minor role in the cofactor interaction, the approximate 2-fold lower TDS contribution for binding NADPH can once again be explained by the polar nature of the phosphate moiety given that it reduces the overall hydrophobicity of the cofactor. Nevertheless, the large enthalpic and negative entropic contribution for both cofactors clearly indicates that nucleotide binding to MRSA–DHDPR is driven by polar interactions. This is consistent with the observations of Reddy et al. [32], who demonstrate similar thermodynamic trends with respect to DH and TDS contributions of NADH and NADPH binding to E. coli DHDPR, albeit with the opposite cofactor preference demonstrated in this study of the MRSA enzyme. The observation that MRSA–DHDPR exhibits substrate inhibition when utilizing NADPH as a cofactor, due to the non-productive ternary complex formed by NADP+ (Fig. 6), offers new insights into rational inhibitor design. Moreover, non-reactive analogs of NADP+ that interact tightly with the nucleotide binding site of DHDPR offer enormous potential for the development of new antimicrobials for the treatment or prevention of MRSA infection. This putative new class of inhibitor provides a dual mode of action, namely the capacity to (i) inhibit binding of reduced NAD(P)H cofactors thus preventing reduction of DHDP, and (ii) promote substrate (i.e. DHDP) inhibition through the nonproductive ternary complex MR–NADP+ DHDP (Fig. 6) determined in this study. In conclusion, this paper describes for the first time the catalytic mechanism and cofactor preference of an essential enzyme and antibiotic target from an important human pathogen that offers new insight into structure-based drug design for antimicrobial intervention.

acknowledge the Defense Threat Reduction Agency (DTRA) (DTRA Project I.D. AB07CBT004) for project support, and the Australian Research Council for providing a Future Fellowship for M.A.P. (FT0991245) and P.E.C. (FT0992105). B.J.S. and P.E.C. acknowledge infrastructure support from the NHMRC (Grant No. 361646) and a Victorian State Government OIS Grant.

Acknowledgments

[21]

We would firstly like to thank all members of the Perugini laboratory for helpful discussions during the preparation of this manuscript, in particular Renwick C.J. Dobson, Michael W.D. Griffin, Natalia E. Ketaren, and Voula Mitsakos. We would also like to

[22]

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14]

[15] [16] [17] [18]

[19] [20]

[23]

G.L. Archer, Clin. Infect. Dis. 26 (1998) 1179–1181. P.C. Appelbaum, Clin. Microbiol. Infect. 12 (Suppl. 2) (2006) 3–10. H.F. Chambers, Clin. Microbiol. Rev. 1 (1988) 173–186. H.F. Chambers, Emerg. Infect. Dis. 7 (2001) 178–182. W.L. Hand, Adolesc. Med. 11 (2000) 427–438. C.J. Weber, Urol. Nurs. 28 (2008) 143–145. C.A. Hutton, M.A. Perugini, J.A. Gerrard, Mol. Biosyst. 3 (2007) 458–465. M.B. Edmond, R.P. Wenzel, A.W. Pasculle, Ann. Intern. Med. 124 (1996) 329– 334. E. Klein, D.L. Smith, R. Laxminarayan, Emerg. Infect. Dis. 13 (2007) 1840–1846. C. Dogovski, S.C. Atkinson, S.R. Dommaraju, L. Hor, R.C.J. Dobson, C.A. Hutton, J.A. Gerrard, M.A. Perugini, Encyclopedia of life support systems, in: H.W. Doelle, S. Rokem (Ed.), Biotechnology Part I, vol. 11, 2009, 116– 136. J.F. Caplan, R. Zheng, J.S. Blanchard, J.C. Vederas, Org. Lett. 2 (2000) 3857–3860. A.M. Paiva, D.E. Vanderwall, J.S. Blanchard, J.W. Kozarich, J.M. Williamson, T.M. Kelly, Biochim. Biophys. Acta 1545 (2001) 67–77. R.J. Cox, Nat. Prod. Rep. 13 (1996) 29–43. S.R. Dommaraju, M.A. Gorman, C. Dogovski, F.G. Pearce, J.A. Gerrard, R.C.J. Dobson, M.W. Parker, M.A. Perugini, Acta Crystallogr., Sect. F 66 (2010) 57– 60. L. Hor, C. Dogovski, R.C.J. Dobson, C.A. Hutton, M.A. Perugini, Acta Crystallogr., Sect. F 66 (2010) 37–40. N.E. Sibarani, M.A. Gorman, C. Dogovski, M.W. Parker, M.A. Perugini, Acta Crystallogr., Sect. F 66 (2010) 32–36. J.M. Wubben, C. Dogovski, R.C.J. Dobson, R. Codd, J.A. Gerrard, M.W. Parker, M.A. Perugini, Acta Crystallogr., Sect. F 66 (2010) 1511–1516. J.E. Voss, S.W. Scally, N.L. Taylor, S.C. Atkinson, M.D.W. Griffin, C.A. Hutton, M.W. Parker, M.R. Alderton, J.A. Gerrard, R.C.J. Dobson, C. Dogovski, M.A. Perugini, J. Biol. Chem. 285 (2010) 5188–5195. M.D.W. Griffin, R.C.J. Dobson, J.A. Gerrard, M.A. Perugini, Arch. Biochem. Biophys. 494 (2010) 58–63. J.E. Voss, S.W. Scally, N.L. Taylor, C. Dogovski, M.R. Alderton, C.A. Hutton, J.A. Gerrard, M.W. Parker, R.C.J. Dobson, M.A. Perugini, Acta Crystallogr., Sect. F 65 (2009) 188–191. B.R. Burgess, R.C.J. Dobson, M.F. Bailey, S.C. Atkinson, M.D.W. Griffin, G.B. Jameson, M.W. Parker, J.A. Gerrard, M.A. Perugini, J. Biol. Chem. 283 (2008) 27598–27603. B.R. Burgess, R.C.J. Dobson, C. Dogovski, G.B. Jameson, M.W. Parker, M.A. Perugini, Acta Crystallogr., Sect. F 64 (2008) 659–661. M.D.W. Griffin, R.C.J. Dobson, F.G. Pearce, L. Antonio, A.E. Whitten, C.K. Liew, J.P. Mackay, J. Trewhella, G.B. Jameson, M.A. Perugini, J.A. Gerrard, J. Mol. Biol. 380 (2008) 691–703.

174

S.R. Dommaraju et al. / Archives of Biochemistry and Biophysics 512 (2011) 167–174

[24] G. Kefala, G. Evans, M.D.W. Griffin, M.A. Perugini, J.A. Gerrard, M.S. Weiss, R.C.J. Dobson, Biochem. J. 411 (2008) 351–360. [25] M.A. Perugini, M.D.W. Griffin, B.J. Smith, L.E. Webb, A.J. Davis, E. Handman, J.A. Gerrard, Eur. Biophys. J. 34 (2005) 469–476. [26] K. Kobayashi, S.D. Ehrlich, A. Albertini, G. Amati, K.K. Andersen, M. Arnaud, K. Asai, S. Ashikaga, S. Aymerich, P. Bessieres, F. Boland, S.C. Brignell, S. Bron, K. Bunai, J. Chapuis, L.C. Christiansen, A. Danchin, M. Debarbouille, E. Dervyn, E. Deuerling, K. Devine, S.K. Devine, O. Dreesen, J. Errington, S. Fillinger, S.J. Foster, Y. Fujita, A. Galizzi, R. Gardan, C. Eschevins, T. Fukushima, K. Haga, C.R. Harwood, M. Hecker, D. Hosoya, M.F. Hullo, H. Kakeshita, D. Karamata, Y. Kasahara, F. Kawamura, K. Koga, P. Koski, R. Kuwana, D. Imamura, M. Ishimaru, S. Ishikawa, I. Ishio, D. Le Coq, A. Masson, C. Mauel, R. Meima, R.P. Mellado, A. Moir, S. Moriya, E. Nagakawa, H. Nanamiya, S. Nakai, P. Nygaard, M. Ogura, T. Ohanan, M. O’Reilly, M. O’Rourke, Z. Pragai, H.M. Pooley, G. Rapoport, J.P. Rawlins, L.A. Rivas, C. Rivolta, A. Sadaie, Y. Sadaie, M. Sarvas, T. Sato, H.H. Saxild, E. Scanlan, W. Schumann, J.F. Seegers, J. Sekiguchi, A. Sekowska, S.J. Seror, M. Simon, P. Stragier, R. Studer, H. Takamatsu, T. Tanaka, M. Takeuchi, H.B. Thomaides, V. Vagner, J.M. Van Dijl, K. Watabe, A. Wipat, H. Yamamoto, M. Yamamoto, Y. Yamamoto, K. Yamane, K. Yata, K. Yoshida, H. Yoshikawa, U. Zuber, N. Ogasawara, Proc. Natl. Acad. Sci. USA 100 (2003) 4678–4683. [27] W. Farkas, C. Gilvarg, J. Biol. Chem. 240 (1965) 4717–4722. [28] S.G. Reddy, J.C. Sacchettini, J.S. Blanchard, Biochemistry 34 (1995) 3492– 3501. [29] F.G. Pearce, C. Sprissler, J.A. Gerrard, J. Biochem. 143 (2008) 617–623.

[30] M. Cirilli, R. Zheng, G. Scapin, J.S. Blanchard, Biochemistry 42 (2003) 10644– 10650. [31] G. Scapin, S.G. Reddy, R. Zheng, J.S. Blanchard, Biochemistry 36 (1997) 15081– 15088. [32] S.G. Reddy, G. Scapin, J.S. Blanchard, Biochemistry 35 (1996) 13294–13302. [33] K. Kimura, T. Goto, J. Biochem. 81 (1977) 1367–1373. [34] K. Kimura, T. Goto, J. Biochem. (Tokyo) 77 (1975) 415–420. [35] J. Cremer, C. Treptow, L. Eggeling, H. Sahm, J. Gen. Microbiol. 134 (1988) 3221– 3229. [36] G. Scapin, J.S. Blanchard, J.C. Sacchettini, Biochemistry 34 (1995) 3502–3512. [37] S. Roberts, J. Morris, R.C.J. Dobson, B. CL, J.A. Gerrard, ARKIVOC (2004) 166–177. [38] S. Leavitt, E. Freire, Curr. Opin. Struct. Biol. 11 (2001) 560–566. [39] R. Perozzo, G. Folkers, L. Scapozza, J. Recept. Signal. Transduct. Res. 24 (2004) 1–52. [40] W.B. Turnbull, A.H. Daranas, J. Am. Chem. Soc. 125 (2003) 14859–14866. [41] T. Wiseman, S. Williston, J.F. Brandts, L.N. Lin, Anal. Biochem. 179 (1989) 131– 137. [42] C.P. Phenix, K. Nienaber, P.H. Tam, L.T. Delbaere, D.R. Palmer, ChemBioChem 9 (2008) 1591–1602. [43] C.P. Phenix, D.R. Palmer, Biochemistry 47 (2008) 7779–7781. [44] H.J. Fromm, Methods Enzymol. 63 (1979) 42–53. [45] A. Cornish Bowden, Fundamentals of Enzyme Kinetics, third ed., Portland Press Ltd., London, 2004, pp. 157–190. [46] H. Chen, S.L. Hopper, C.E. Cerniglia, Microbiology 151 (2005) 1433–1441.