Structure and inhibition studies of a type II beta-carbonic anhydrase psCA3 from Pseudomonas aeruginosa

Bioorganic & Medicinal Chemistry xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc

Structure and inhibition studies of a type II beta-carbonic anhydrase psCA3 from Pseudomonas aeruginosa Melissa A. Pinard a,⇑, Shalaka R. Lotlikar b, Christopher D. Boone a, Daniela Vullo c,d, Claudiu T. Supuran c,d, Marianna A. Patrauchan b, Robert McKenna a,⇑ a

Department of Biochemistry and Molecular Biology, College of Medicine, University of Florida, Gainesville, FL 32610, USA Department of Microbiology and Molecular Genetics, Oklahoma State University, Stillwater, OK 74078, USA Università degli Studi di Firenze, Laboratorio di Chimica Bioinorganica, Rm. 188, Via della Lastruccia 3, I-50019 Sesto Fiorentino (Firenze), Italy d Università degli Studi di Firenze, Polo Scientifico, Dipartimento Neurofaba, Sezione di Scienze Farmaceutiche e Nutraceutiche, Via Ugo Schiff 6, 50019 Sesto Fiorentino (Firenze), Italy b c

a r t i c l e

i n f o

Article history: Received 31 March 2015 Revised 10 May 2015 Accepted 19 May 2015 Available online xxxx Keywords: Carbonic anhydrase Beta-carbonic anhydrase Pseudomonas aeruginosa psCA3 Type I Type II Sulfonamides Anionic inhibitors

a b s t r a c t Carbonic anhydrases (CAs) are metallo-enzymes that catalyze the reversible hydration of carbon dioxide into bicarbonate and a proton. The b-class CAs (b-CAs) are expressed in prokaryotes, fungi, plants, and more recently have been isolated in some animals. The b-CA class is divided into two subclasses, termed type I and II, defined by pH catalytic activity profile and active site structural configuration. Type I b-CAs display catalytic activity over a broad pH range (6.5–9.0) with the active site zinc tetrahedrally coordinated by three amino acids and a hydroxide/water. In contrast, type II b-CAs are catalytically active only at a pH 8 and higher where they adopt a functional active site configuration like that of type I. However, below pH 8 they are conformationally self-inactivated by the addition of a fourth amino acid coordinating the zinc and thereby displacing the zinc bound solvent. We have determined the structure of psCA3, a type II b-CA, isolated from Pseudomonas aeruginosa (P. aeruginosa) PAO1 at pH 8.3, in its open active state to a resolution of 1.9 Å. The active site zinc is coordinated by Cys42, His98, Cys101 and a water/hydroxide molecule. P. aeruginosa is a multi-drug resistant bacterium and displays intrinsic resistance to most of the currently used antibiotics; therefore, there is a need for new antibacterial targets. Kinetic data confirm that psCA3 belongs to the type II subclass and that sulfamide, sulfamic acid, phenylboronic acid and phenylarsonic acid are micromolar inhibitors. In vivo studies identified that among six tested inhibitors representing sulfonamides, inorganic anions, and small molecules, acetazolamide has the most significant dose-dependent inhibitory effect on P. aeruginosa growth. Ó 2015 Published by Elsevier Ltd.

1. Introduction Carbonic anhydrases (CAs, EC 4.2.1.1) are a family of mostly zinc metallo-enzymes that catalyze the reversible hydration of carbon dioxide into a bicarbonate ion and a proton.1 The b-class CAs (b-CAs) are found in prokaryotes as well as in plants, algae and some animals.2,3 In prokaryotes b-CAs provide endogenous bicarbonate and support the activity of bicarbonate dependent carboxylases required for fatty acid biosynthesis.4 In addition, the enzymatic activity of b-CAs may contribute to a broad range of physiological functions including cyanate degradation,5 colonization6 and survival7–9 of pathogenic bacteria in their hosts. In photosynthetic prokaryotes such as cyanobacteria, these enzymes are Abbreviations: CA, carbonic anhydrase; b-CA, beta-carbonic anhydrase.

⇑ Corresponding authors. Tel.: +1 352 392 5696; fax: +1 352 392 3422. E-mail address: rmckenna@ufl.edu (R. McKenna).

expressed in the carboxysome where they play an essential role in carbon fixation.2 In plants like Pisum sativum, b-CAs are transcriptionally linked to Rubisco and may act as an accessory enzyme.10 Less is known about their role in animals; DmBCA, the first ever animal b-CA was identified in the mitochondria of Drosophila melanogaster.11 The b-CAs are classified into two distinct subclasses by the organization of their active site configuration and catalytic pH range, as either type I (broad pH range) or type II (selective alkaline pH range). In type I b-CAs, the zinc is coordinated by two cysteines and a histidine, permitting an ‘open’ active site with the presence of an exchangeable zinc-bound water/hydroxide as the fourth ligand.4 In the open active site state, an arginine stabilizes an aspartate through salt-bridge interactions (forming an Asp-Arg dyad), thereby opening the active site.4 Type I b-CAs show catalytic activity at pH from as low as 6.5 to greater than 9.0 in some cases and always maintain an ‘open’ active configuration4 (Fig. 1A).

http://dx.doi.org/10.1016/j.bmc.2015.05.029 0968-0896/Ó 2015 Published by Elsevier Ltd.

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

2

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

In contrast, types II b-CAs are catalytically inactive below pH 8 and the zinc-bound water/hydroxide is in equilibrium with Asp for binding to the active site metal ion. At this pH the active site is ‘closed’, with the aspartate (from the Asp-Arg dyad) being released from the arginine and conformationally moved to become the fourth ligand coordinating the zinc and displacing the zinc-bound water/hydroxide, thus preventing catalysis (Fig. 1B). At pH values above 8.0, the metal-bound water is readily deprotonated, generating a much less labile hydroxide ion, significantly altering the Aspsolvent binding equilibrium toward the metal-solvent form. The displaced Asp residue can be complexed by Arg thereby reforming the conserved salt bridge.4,9,12,13 In this scenario, the primary causal event is the deprotonation of the metal-bound water molecule, whose pKa is around 8 in b-CA. Though the catalytic mechanism of the b-CAs is not fully understood, it uses the same general ping-pong metal-hydroxide mechanism utilized by the a-CAs.4 In the first step (in the hydration direction), a CO2 molecule enters the open active site and undergoes nucleophilic attack by the zinc-bound hydroxyl to form bicarbonate that readily exchanges with a water (Eq. 1). Then in the second step, the zinc bound-hydroxide is regenerated by the transfer of a proton to the surrounding solvent (Eq. 2).

E-Zn2þ  OH $ E-Zn2þ  HCO3 $ E-Zn2þ  H2 O þ HCO3

ð1Þ

E-Zn2þ  H2 O þ B $ E-Zn2þ  OH þ BHþ

ð2Þ

Pseudomonas aeruginosa is a gram-negative facultative anaerobe, which prefers aerobic respiration, but is well adapted to low oxygen conditions. It is an opportunistic human pathogen and a common cause of life-threatening chronic infections in cystic fibrosis and endocarditis patients as well as nosocomial infections.14,15 The bacterium’s intrinsic and adaptive resistance to most antibacterial agents, including aminoglycosides and quinolones has made it difficult to treat infections with currently available antimicrobial therapies.15 Thus, the identification of new antimicrobial targets is of important research interest. Sequence analysis of the P. aeruginosa PAO1 genome revealed three genes psCA1 (PA0102), psCA2 (PA2053) and psCA3 (PA4676) that encode for functional b-CAs: psCA1, psCA2 and psCA3, respectively.16 A multiple sequence alignment of psCA1, psCA2 and psCA3 shows that these enzymes share a 27–49% amino acid sequence identity amongst themselves. When aligned with the sequences of the b-CA whose crystal structures are available in the Protein Data Bank (PDB), psCA1 shares the highest sequence identity—32% with P. sativum PSCA (PDB ID: 1EKJ) a type I b-CAs, while psCA2 and psCA3 show 32% and 53% sequence identity to

Salmonella enterica, SECCA, (PDB ID: 3QY1) and E. coli, ECCA, (PDB ID: 1I6P), respectively, both type II b-CAs (Supplementary Fig. 1). Studies by the Patrauchan group have shown that all three bCAs are expressed in PAO1, contain zinc, and hydrate CO2.16 Though, the functional roles of the psCAs are not fully characterized, by using transposon mutants with individually disrupted psCAs, psCA1 was shown to play a role in PAO1 growth at ambient CO2 levels, similar to the role played by the b-CAs PCA in Streptococcus pneumoniae and Can in Escherichia coli.8,16,17 Furthermore, the most recent studies of the group suggest that psCA2 and psCA3 are involved in the formation of extracellular calcium deposits (Lotlikar et.al., in preparation). This ability of the organism may increase its virulence by contributing to soft tissue calcification or stone formation during bacterial infections which was observed, for example, in late stages of cystic fibrosis or primary ciliary dyskinesia.18,19 As the b-CAs share no sequence or structural similarity to the aCAs, the only class of CAs expressed in humans, and due to their role in P. aeruginosa growth and calcium deposition, they represent druggable targets for P. aeruginosa drug development, in as much as b-CAs from Porphyromonas gingivalis, another human pathogen.20,21 Here, we report the crystal structure of psCA3 determined at pH 8.3 to a resolution of 1.9 Å with an ‘open’ active site configuration. In addition, kinetic studies confirm psCA3 belongs to the type II subclass and its activity is inhibited by micromolar levels of sulfamide, sulfamic acid, phenylboronic acid and phenylarsonic acid. Although, the disruption of psCA3 did not affect P. aeruginosa growth, in vivo inhibition studies identified that acetazolamide (AAZ) significantly delays and reduces growth of the pathogen. 2. Materials and methods 2.1. Sequence analysis, expression and purification of psCA3 A sequence alignment of psCA3 to all the b-CAs available in the PDB was performed by using ClustalW2.22 Cloning, protein expression and purification were performed as described previously in.16 A pET15b plasmid construct containing the psCA3 gene PCR-amplified from the PAO1 genomic DNA, was transformed into E. coli Tuner BL21 (DE3) cells for the production of a 6 His fusion protein. The cells were grown in Luria–Bertani (LB) broth containing 100 mg ml1 ampicillin and 0.05 mM ZnSO4 at 310 K. At an optical density (Ab600) of 0.6, protein expression was induced for 3 h by the addition of isopropyl b-D-1-thiogalactopyranoside (IPTG) to a final concentration of 1 mM in the presence of 0.5 mM ZnSO4 (to

Figure 1. (A) Active site geometry of the type I b-CA, Rv1284 (PDB ID: 1YLK) isolated from Mycobacterium tuberculosis exhibiting the ‘open’ configuration. (B) Active site geometry of the type II b-CA, Rv3588c (PDB ID: 1YM3) isolated from Mycobacterium tuberculosis exhibiting the ‘closed’ configuration. Zinc-bound solvent (red sphere). Both structures were determined at pH 7.5. Figure made using PyMOL.37

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

prevent Zn2+ depletion). Cells were harvested by centrifugation, resuspended in a buffer of 20 mM Tris–HCl pH 7.9, 5 mM imidazole, 150 mM NaCl, and lysed by sonication. After centrifugation, the cell-free supernatant was loaded onto Ni2+-charged IMAC column and washed with 20 mM Tris–HCl pH 7.9, 60 mM imidazole, 150 mM NaCl. psCA3 was eluted with 20 mM Tris–HCl pH 7.9, 300 mM imidazole, 150 mM NaCl. To confirm protein purification, the elution fractions were resolved using SDS-PAGE followed by Coomassie Blue R-250 staining. These fractions were then dialyzed against 20 mM Tris–HCl pH 7.9, 150 mM imidazole, 100 mM NaCl, 10% glycerol for 2 h and then against 20 mM Tris–HCl pH 7.9, 50 mM imidazole, 50 mM NaCl, 5% glycerol for 1 h. Finally, psCA3 was dialyzed against 20 mM Tris–HCl, pH 8.3 three times (1 h each) and stored in this buffer. psCA3 was concentrated using an Amicon Ultra-15 Centrifugal Filter Unit (Millipore, Billerica, Massachusetts, USA) and its concentration was determined by UV–vis spectroscopy at 280 nm using a molar extinction coefficient of 36440 M1 cm1. 2.2. Crystallization Crystals of psCA3 were obtained using the Gryphon crystallization robot by the sitting-drop vapor diffusion method. Drops were prepared by mixing protein solution at 10 mg ml1 and precipitant solution at two ratios (1:1 and 2:3 protein: precipitant solution) and then equilibrated at 290 K against a 60 ll reservoir containing precipitant solution. The precipitant solution consisted of 2.2 M ammonium sulfate, 0.1 M malic acid, and 0.1 M imidazole pH 7.5. Crystals were observed within 14 days. One crystal was cryoprotected by immersion in a 20% (w/v) glycerol precipitant solution and flash cooled by exposure to liquid nitrogen at 100 K for data collection. 2.3. Diffraction data collection Diffraction data for psCA3 were collected at the F1 station at Cornell High Energy Synchrotron Source (CHESS F1; k = 0.9177 Å) on an ADSC Q-270 detector using the microfocused beam. Data were collected with a crystal-to-detector distance of 200 mm and a 1.0° oscillation angle with an exposure time of 20 s per image over 107 frames. The data sets were integrated, merged and scaled using HKL-2000.24 Phasing was carried out with the PHENIX25 suite of programs using the auto molecular-replacement procedure to obtain the initial phases. An unrefined psCA3 structure determined at a lower resolution with no solvent was used to calculate the initial phases.23 2.4. Structure determination The graphics program Coot26 was used to view electron density maps, and the psCA3 structure was manually adjusted. Refinement was continued until the Rwork and Rfree converged. The geometry of the final psCA3 model was verified using PROCHECK and MOLPROBITY.27,28 2.5. Enzymatic activity and measurements Kinetic parameters for the CO2 hydration reaction catalyzed by psCA3 were assayed using the modified colorimetric stopped-flow method.29 All measurements were made at 293 K, pH 8.3 in 20 mM Tris buffer and 20 mM NaClO4. An Applied Photophysics stoppedflow instrument has been used for assaying the CA catalysed CO2 hydration activity. Bromothymol blue (at a concentration of 0.2 mM) has been used as indicator, working at the absorbance maximum of 557 nm, with 20 mM Tris (pH 8.3) buffer, and 20 mM NaClO4 (for maintaining constant the ionic strength.

3

Perchlorate is not inhibiting the enzyme at concentrations up to 100 mM, data not shown, as for many other CAs investigated earlier), following the initial rates of the CA-catalyzed CO2 hydration reaction for a period of 10–100 s.29 The CO2 concentrations ranged from 1.7 to 17 mM for the determination of the kinetic parameters and inhibition constants. Lineweaver–Burk plots were used for determining the kinetic parameters. For each inhibitor at least six traces of the initial 5–10% of the reaction have been used for determining the initial velocity. The uncatalyzed rates were determined in the same manner and subtracted from the total observed rates. Stock solutions of inhibitors (10 mM) were prepared in distilled-deionized water and dilutions up to 0.01 lM were done thereafter with the assay buffer. Inhibitor and enzyme solutions were preincubated together for 15 min at room temperature prior to assay, in order to allow for the formation of the E-I complex. The inhibition constants were obtained by non-linear least-squares methods using PRISM 3 and the Cheng–Prusoff equation, as reported earlier, and represent the mean from at least three different determinations. The concentration of psCA3 used in the experiments reported in this work was of 12.1 nM. 2.6. Bacterial strains, media, and growth conditions P. aeruginosa PAO1 strain, originally isolated from a wound, was used in this study. The PAO1 transposon mutant 8872 with the genotype: PA4676-B07::ISlacZ/hah was obtained from the University of Washington Two-Allele library. Both PAO1 and PW8872 cells were grown at 310 K, 200 rpm in either 50% brain heart infusion broth (BHI, Teknova) or biofilm minimal medium (BMM) medium, which contained (per liter): 9.0 mM sodium glutamate, 50 mM glycerol, 0.02 mM MgSO4, 0.15 mM NaH2PO4, 0.34 mM K2HPO4, and 145 mM NaCl, 20 ll trace metals, 1 ml vitamin solution. Trace metal solution (per liter of 0.83 M HCl): 5.0 g CuSO4.5H2O, 5.0 g ZnSO4.7H2O, 5.0 g FeSO4.7H2O, 2.0 g MnCl24H2O. Vitamins solution (per liter): 0.5 g thiamine, 1 mg biotin. The pH of BMM was adjusted to 7.0. The cultures were grown in ambient air, in the presence of 100 or 200 lM of 3-aminobenzenesulfonamide (ABS), 1-amino,2-chloro-benzenedisulfonamides (ACB), sulfamide, indisulam (IND), iminodisulfonate (IDS), or AAZ (Supplementary Fig. 2A–F). AAZ, ABS, ACB, IND, and IDS were dissolved in 50% DMSO. For these compounds, the control cells were grown in the presence of 6.4 mM of DMSO. For quantitative growth assays, PAO1 and PW8872 cells were first grown in tubes containing 5 ml 50% BHI or BMM to mid-log phase (8 h and 16 h, respectively). The mid-log cultures were diluted in the corresponding media to obtain absorbance at 600 nm (Ab600) of 0.3. Dilutions 1:100 of these standardized cultures were inoculated into a sterile flat bottom 96-well polystyrene microtiter plate (Sigma Aldrich, St. Louis, Missouri, USA) with fresh media with or without inhibitors or DMSO. Ab600 of the cultures was measured using a Bio-Tek Synergy HT microtiter plate reader (Tecan Instruments Inc.) at 310 K for a total of 15 h (50% BHI) and 26 h (BMM) with slow shaking. Non-inoculated wells were measured as blanks. The results represent the mean values of three biological replicates, each of which contained three technical replicates. 3. Results 3.1. Structural analysis The psCA3 crystals diffracted to 1.9 Å resolution in the orthorhombic space group I222 with an Rsym of 8.0% and unit cell parameters a = 71.2, b = 77.9, c = 87.7 Å. The initial phases were calculated using a preliminary unrefined structure determined to 3.0 Å resolution.23 The psCA3 crystallized as a monomer in the

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

4

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

Table 1 Data collection and refinement statistics for psCA3 PDB accession code Beamline Detector Crystal-to-detector distance (mm) Wavelength (Å) Temperature (K) Space group Unit-cell parameters (Å) Resolution (Å) Total number of reflections Individual reflections Redundancy Completeness I/r Rsyma Rcryst/Rfreeb r.m.s.d. for bond lengths/angles (Å, °) Average B-factors (Å2) Main/side Water molecules No. of protein atoms No. water molecules

4RXY CHESS F1 ADSC Q270 CCD 150 0.9177 100 I222 a = 71.2, b = 77.9, c = 87.7 1.90 (1.97–1.90)* 49728 16312 4.3 98.2 (99.8) 14.3 (2.94) 0.080 (0.526) 0.167/0.199 0.008/1.040 30.9/36.3 38.5 1696 72

P P Rsym = ( |I  |/ ) P P Rcryst = ( |Fo  Fc|/ |Fo|). Rfree is calculated in the same way as Rcryst except it is for data omitted from refinement (5% of reflections for all data sets). * Values in parentheses represent highest resolution bin. a

b

crystallographic asymmetric unit with all 215 amino acids ordered. The refined model converged to Rcrys of 16.7 and Rfree of 19.9%. The data-collection parameters and final refinement statistics are summarized in Table 1. In addition, the Ramachandran plot generated by PROCHECK27 showed that over 90% of the residues had dihedral angles within the most favored region, while the rest were in the allowed region. Both r.m.s.d values for bond lengths and bond angles were within acceptable limits. The active biological unit of psCA3, like many other b-CAs is a fundamental dimer and was confirmed by MALDI-TOF—based protein identification.16 The fundamental dimer and the active site at dimer interface is shown in Figure 2A and B respectively. b-CAs possess a b-sheet core comprised of 4 or 5 strands with four or more a-helices surrounding this core4,9 (Fig. 3). At pH 8.3, the active site of psCA3 is in an ‘open’ configuration with the Asp-Arg dyad, consisting of two salt bridges between Arg46 and Asp44 (2.8 and 3.0 Å). The zinc is coordinated by His98, Cys42, Cys101 and a H2O/OH- molecule (zinc bound solvent, Zn-OH/H2O) 2.3 Å away in a pseudo-tetrahedral arrangement (Fig. 4). The structure of psCA3’s active site also reveals a plausible ordered water network comprised of at least four water molecules (Fig. 5). The water wire

Figure 3. Ribbon diagram of psCA3. The five b-strands are in 2-1-3-4-5 parallel arrangement (with the exception strand b5 is antiparallel to b4). Figure made using PyMOL.37

consists of the Zn-OH/H2O (B-factor: 24.6 Å2), a second water molecule (W1) (B-factor: 17.8 Å2) approximately 2.4 Å away from the Zn-OH/H2O, a third water (W2) (B-factor: 23.4 Å2) hydrogen bonds with the main chain nitrogen of His98 (3.0 Å) and side chain oxygen atom of Asn68 (2.7 Å), and a fourth water (W3) (B-factor: 22.3 Å2) 2.9 Å away from W2 forms hydrogen bonds with the side chain nitrogen atom of Asn151 and the main chain oxygen of Cys148. This water network could be a potential route for proton shuttling in and out of the active site, allowing for regeneration of the active form of the enzyme (Zn-OH). The psCA3 enzyme is shown to have the highest sequence identity (53%) to the type II b-CA ECCA (Supplementary Fig. 1). As with all the b-CAs, two cysteines (Cys42 and Cys101) and histidine (His98) as well as aspartate (Asp44) and arginine (Arg46), constituting the Asp-Arg dyad, were conserved in the active site. The

Figure 2. (A) psCA3 fundamental dimer; subunit A (olive), subunit B (light orange). The active sites are also shown at the dimer interface. (B) psCA3 dimer interface showing active site residues coordinating zinc as well as Gln33 which is critical in dimerization. Figure made using PyMOL.37

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

5

Figure 4. The psCA3 structure showing zoomed in active site (inset) of psCA3. The zinc (grey sphere) coordinated by: Cys42, His98and Cys101and a solvent molecule (red sphere). Also shown is the pseudo-tetrahedral coordination of the zinc and the conserved salt bridge (dashed lines) between Asp44 and Arg46 dyad (inset). All distances in Å. Figure made using PyMOL.37

Figure 5. Hydrogen bond network in psCA3 active site. Indicated hydrogen bonds (dashed lines) and solvent molecules (red spheres). All distances in Å. Figure made using PyMOL.37

crystal structures of ECCA and psCA3 both consist of an N-terminus dominated by two a-helices. Superposition of psCA3 onto a monomeric unit of ECCA (PDB ID: 1I6O) gives an r.m.s.d of 1.5 Å. However, in ECCA the Asp-Arg dyad is disrupted due to reorganization of the 44-48 loop and Asp44 points towards the zinc ion since its crystal structure was solved under acidic conditions (Fig. 6). 3.2. Kinetic studies Carbonic anhydrase activity for psCA3 was measured, and at pH 7.5 no catalytic activity was detected for the enzyme. A kcat of 1.4  105 and a kcat/Km of 1.0  107 M1 s1 was measured at pH 8.3 showing that psCA3 activity is pH dependent. Being a b-class enzyme, this is not unexpected. In fact, it has been proven by Jones’ group that in some types of such enzymes the active site is closed at pH values under 8.3, as the Zn(II) ion is coordinated

Figure 6. Overlay of ECCA (PDB ID: 1I6P, pH 7.0) (pink) onto psCA3 (pH, 8.3 olive). Highlighted the pH-induced reorganization of the 44–48 loop (open circle).

by four conserved residues, with no water present bound to the metal ion.30 This is the so-called ‘closed active site’. When the pH becomes more alkaline, towards pH values of 8.3 or higher, the coordinated aspartate dissociates from the metal ion as the metal-bound water becomes deprotonated and generates a hydroxide ion which alters the Asp-solvent binding equilibrium toward the metal-solvent form. The liberated Asp residue forms a strong interaction with the Arg residue from the catalytic dyad conserved in all b-CAs investigated so far. In this way an incoming water molecule becomes coordinated to zinc and subsequently acts as nucleophile for the hydration of CO2. Apparently psCA3 belongs to the b-CAs with a closed active site at pH values <8.3,

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

6

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

and we were able to measure catalytic activity only at pH value 8.3 (Table 2). The catalytic rates of other b-CAs are also included for comparison in Table 2.

-

Zn

B

a b c d e f g h

Activity level

kcat (s1)

Kcat/KM (M1  s1)

KI (AAZ) (nM)

hCA I⁄a hCA II⁄b Can2§c CalCA§d SceCA§e FbiCA1§f CahB1§g psCA3§h

Moderate Very high Moderate High High Low Moderate Low

2.0  105 1.4  106 3.9  105 8.0  105 9.4  105 1.2  105 2.4  105 (1.4  105 ± 0.12)

5.0  107 1.5  108 4.3  107 9.7  107 9.8  107 7.5  106 6.3  107 (1.0 ± 0.1)  107

250 12 10.5 132 82 27 76 75.9 ± 3.2

Human carbonic anhydrase I. Human carbonic anhydrase II. Cryptococcus neoformans carbonic anhydrase 2. Candida albicans carbonic anhydrase. Saccharomyces cerevisiae carbonic anhydrase. Flaveria bidentis carbonic anhydrase 1. Coleofasciculus chthonoplastes carbonic anhydrase. Pseudomonas aeruginosa carbonic anhydrase 3.

Inh

+ InhH

Zn

2+

Cys101 Cys42 His98

Table 2 Kinetic parameters for the CO2 hydration reaction catalyzed by the human cytosolic isozymes a-class CAs⁄ and the b-CAs§. Inhibition data with the clinically used sulfonamide AAZ are also provided (stopped-flow assay) Isozyme

OH

-H2O

2+

Cys101 His98

Cys42

E

A

- BH +

-

+ CO2

OH2 Zn Cys42

2+

Cys1013 His980

Zn

D

H2O

Cys42

- HCO 3

Inh

Zn

Cys42

+ H2O

-H+

InhH

OH

-

2+

Cys101 His98

O O

2+

Cys101 His98 B

O

H

O Zn

Cys42

O

2+

-

Cys101 His980 C

F

Figure 7. Schematic representation of the catalytic cycle and inhibition mechanisms of psCA3. Table 3 Inhibition constants of anionic inhibitors against the a-CA (human) isoform hCA II, and Pseudomonas aeruginosa psCA3, for the CO2 hydration reaction, at 293 K and pH 8.3 KIb (mM)

Inhibitora

F Cl Br I CNO SCN CN N 3 HCO 3 NO 3  NO2 HS HSO 3 SO24 SnO2 3 SeO2 4 2 TeO4 P2O4 7 V2O4 7 B4O2 7 ReO 4 RuO 4 S2O2 8 SeCN CS2 3 Et2NCS 2 ClO 4  BF4 FSO 3 NH(SO3)2 2 H2NSO2NH2 H2NSO3H Ph-B(OH)2 Ph-AsO3H2 PF 6 ABS ACB a

hCA II

psCA3

>200 200 63 26 0.03 1.6 0.02 1.5 85 35 63 0.04 89 >200 0.83 112 0.92 48.50 0.57 0.95 0.75 0.69 0.084 0.086 0.0088 3.1 >100 >100 0.46 0.76 1.13 0.39 23.1 49.2

>100 >100 >100 >100 0.65 0.61 0.53 0.77 3.2 14.2 0.94 0.62 7.0 >100 1.4 23.5 7.9 8.2 4.9 9.1 0.83 0.94 0.72 0.68 0.84 0.33 >100 >100 2.4 4.7 0.008 0.038 0.008 0.009 0.92 0.000032 0.000095

0.0003 0.00024

As sodium salt; nt = not tested. Errors were in the range of 3–5% of the reported values, from three different assays. b

3.3. Inhibition studies The antibacterial effects of CA inhibition by sulfonamides, anions and small molecules has been studied in numerous pathogenic bacteria and has been reported by the Supuran group in detail.31–35 A set of inorganic anions and other small molecules known to interact with CAs, such as sulfamide, sulfamic acid, phenylboronic acid, phenylarsonic acid and diethyldithiocarbamate were assayed as inhibitors of psCA3 (Table 3). Anion inhibitors are important to be investigated as some of them (chloride, bicarbonate, sulfate, etc.) may be present at elevated concentrations under the physiologically relevant conditions and interfere with the catalytic activity of CAs. Otherwise, finding small molecule—active site binders may lead to the development of new classes of effective organic inhibitors by considering inorganic anions as leads. The inhibition data showed that all the halides, as well as sulfate, perchlorate and tetrafluoroborate were not inhibitors of psCA3 up to concentrations of 100 mM. This is to be expected for weak nucleophiles such as perchlorate and tetrafluoroborate, but not for the halides, which normally bind the Zn(II) ion from other CAs such as a-class CA II (Table 3). In addition, weak inhibition (KI’s in the range of 15–30 mM) was observed for carbonate, nitrate and selenite (Table 3). Other anions, among which bicarbonate, hydrogensulfite, stannate, tellurate, diphosphate, divanadate, tetraborate, fluorosulfonate and iminodisulfonate were slightly better inhibitors, with KI’s in the range of 1.0–1.0 mM. More effective inhibitors like the pseudohalides, nitrite, hydrogensulfide, perrhenate, perruthenate, peroxydisulfate, selenocyanide, trithiocarbonate, diethyldithiocarbamate and hexafluorophosphate displayed submillimolar KI values range (0.3– 1.0 mM). However, the most potent psCA3 inhibitors were shown to be: sulfamide, sulfamic acid, phenylboronic acid and phenylarsonic acid, which showed KI’s in the range of 10–40 lM. The potent inhibitory activity of these small molecules is of great interest as they contain fragments amenable to medicinal chemistry modifications, which could eventually lead to much more potent, hopefully nanomolar psCA3 inhibitors. The inhibition mechanism of this enzyme is depicted in Figure 7.

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

Figure 8. The effect of AAZ on PAO1 growth. The cultures were grown in BMM with or without the addition of 100 or 200 lM AAZ dissolved in DMSO. The latter added alone served as a solvent control (CTR). Graphs represent the averages of three independent growth experiments.

3.4. Growth inhibition by sulfonamides, inorganic anions and small molecules Although the disruption of psCA3 did not affect growth of PAO1 in neither rich nor mineral media (data not shown), we investigated whether selected compounds shown to inhibit psCA3 activity (Tables 2 and 3) would have an inhibiting effect on growth of P. aeruginosa PAO1 and transposon insertion mutant PW8872 with disrupted psCA3. Compounds from different chemical groups were selected: sulfonamides, AAZ, IND, ABS, ACB, inorganic anions (IDS), and small molecules (sulfamide) (Supplementary Table 1). The most significant growth inhibiting effect was recorded in the presence of AAZ, which delayed growth of P. aeruginosa PAO1 by 4 h and reduced growth rate and yield (Fig. 8). Doubling the amount of the compound proportionally decreased the growth. Sulfamide, one of the most potent inhibitors tested, delayed the growth of the PW8872 mutant by 1 h, however, the increase in sulfamide concentration up to 1 mM did not increase the growth defect. There was no effect of sulfamide on PAO1 growth (data not shown). 4. Discussion For years, it has been hypothesized that the pH sensitive nature of the type II subclass and its ability to adopt a stable inactive form despite the presence of available water molecules in the protein environment suggests some sort of regulatory process.36 However, the mechanism for this regulation is still not fully understood. One argument is that bicarbonate may play a role in the allosteric regulation of some type II b-CAs and that its binding in a non-catalytic site results in the enzyme adopting the ‘closed’ active site configuration. Kinetic data of both Haemophilus influenzae b-CA: HICA (PDB ID: 2A8D) and ECCA show that bicarbonate modulates their activity and allows the cooperative transition between the active and inactive forms as seen by pH-rate profiles at both steady-state and chemical equilibrium (at high HCO 3 concentrations substrate inhibition is observed in HICA).36 This transition was observed at pH 8 during kinetic measurements of both CO2 hydration and HCO 3 dehydration. In addition, structural evidence for the presence of a bicarbonate binding site 8 Å from the zinc also exists. This binding site is characterized by a Trp-Arg-Tyr triad (Trp39, Arg64, Tyr181 in HICA) and is conserved in psCA3 (Trp39, Arg64, Tyr181). The Trp-Arg-Tyr triad is absent in all type I b-CAs. In both HICA and ECCA each residue donates one or more hydrogen bond to bicarbonate; in addition, two water molecules also form multiple hydrogen bonds with bicarbonate’s oxygen atoms. However, neither HICA nor ECCA have been crystallized in their alternative

7

‘open’ configuration but the kinetic data suggests these enzymes adopt this conformation in order to function in a highly cooperative manner.36 In psCA3, a network of specific hydrogen bonds involving residues: Cys42, Arg46, Val47, Cys96, and His98 may also contribute to binding bicarbonate and this site appears to be uniquely designed for bicarbonate recognition. Binding of bicarbonate expels the side chain of Val47, which causes reorganization of the 44–48 loop and results in the disruption of the Asp44-Arg46 dyad, allowing Asp44 to displace the zinc-bound water and bind directly to the zinc ion. However, bicarbonate is not seen in psCA3’s proposed bicarbonate binding site as the enzyme is in the ‘open’ configuration but the presence of this conserved site suggests bicarbonate may also play a similar role in the allosteric regulation of this b-CA as well. 5. Conclusions Both the kinetic data at pH 8.3 showing catalytic activity and the ‘open’ active site configuration seen in the crystal structure confirms that psCA3, one of the three b-CAs expressed in a human pathogen P. aeruginosa, belongs to the type II subclass of b-CAs. In addition, the inhibition studies of psCA3 by sulfamide, sulfamic acid, phenylboronic acid and phenylarsonic acid displayed KI’s in the micromolar range proving to be the most effective inhibitors among the many others tested. While the halides, along with sulfate, perchlorate and tetrafluoroborate proved to be less potent inhibitors displaying KI’s in the millimolar range. Growth studies showed that AAZ is the only compound among tested with a significant dose-dependent growth inhibiting effect on P. aeruginosa PAO1, suggesting a potential alternative drug treatment. It is not clear, however, whether the AAZ-mediated growth defect was due to its ability to inhibit the activity of all three functional b-CAs (psCA1, psCA2, and psCA3) and possibly three predicted c-CAs16 or due to its toxicity to P. aeruginosa cells. Further studies will aim to characterize the effect of AAZ on the other CAs of P. aeruginosa and identify the mechanism of AAZ growth inhibition. Considering limited permeability of P. aeruginosa membrane, studies of the inhibitors that would penetrate pseudomonal membrane and disrupt the activity of multiple CAs may be of interest for developing an alternative to current therapeutics for efficient treatments of P. aeruginosa infections. 6. Funding This work was supported in part by NIH – United States grant GM25154. Acknowledgements The authors would like to thank the staff at the Cornell High Energy Synchrotron Source, for assistance during X-ray diffraction data collection and the Center of Structural Biology for support of the X-ray facility at UF. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmc.2015.05.029. References and notes 1. Aggarwal, M.; Boone, C. D.; Kondeti, B.; McKenna, R. J. Enzyme Inhib. Med. Chem. 2013, 28, 267. 2. Smith, K. S.; Ferry, J. G. FEMS Microbiol. Rev. 2000, 24, 335. 3. Smith, K. S.; Ferry, J. G. J. Bacteriol. 1999, 181, 6247.

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029

8

M. A. Pinard et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx

4. Rowlett, R. S. Biochim. Biophys. Acta 1804, 2010, 362. 5. Guilloton, M. B.; Korte, J. J.; Lamblin, A. F.; Fuchs, J. A.; Anderson, P. M. J. Biol. Chem. 1992, 267, 3731. 6. Bury-Moné, S.; Mendz, G. L.; Ball, G. E.; Thibonnier, M.; Stingl, K.; Ecobichon, C.; Avé, P.; Huerre, M.; Labigne, A.; Thiberge, J.-M.; De Reuse, H. Infect. Immun. 2008, 76, 497. 7. Valdivia, R. H.; Falkow, S. Science 1997, 277, 2007. 8. Burghout, P.; Cron, L. E.; Gradstedt, H.; Quintero, B.; Simonetti, E.; Bijlsma, J. J. E.; Bootsma, H. J.; Hermans, P. W. M. J. Bacteriol. 2010, 192, 4054. 9. Covarrubias; Larsson, A. M.; Högbom, M.; Lindberg, J.; Bergfors, T.; Björkelid, C.; Mowbray, S. L.; Unge, T.; Jones, T. A. J. Biol. Chem. 2005, 280, 18782. 10. Majeau, N.; Coleman, J. R. Plant Physiol. 1996, 112, 569. 11. Syrjänen, L.; Tolvanen, M.; Hilvo, M.; Olatubosun, A.; Innocenti, A.; Scozzafava, A.; Leppiniemi, J.; Niederhauser, B.; Hytönen, V. P.; Gorr, T. A.; Parkkila, S.; Supuran, C. T. BMC Biochem. 2010, 11, 28. 12. Cronk; Endrizzi, J. A.; Cronk, M. R.; O’neill, J. W.; Zhang, K. Y. Protein Sci. Publ. Protein Soc. 2001, 10, 911. 13. Covarrubias, A. S.; Bergfors, T.; Jones, T. A.; Högbom, M. J. Biol. Chem. 2006, 281, 4993. 14. Holm, J. P. Y.; Hilberg, O.; Noerskov-Lauritsen, N.; Bendstrup, E. Dan. Med. J. 2013, 60, A4636. 15. Mesaros, N.; Nordmann, P.; Plésiat, P.; Roussel-Delvallez, M.; Van Eldere, J.; Glupczynski, Y.; Van Laethem, Y.; Jacobs, F.; Lebecque, P.; Malfroot, A.; Tulkens, P. M.; Van Bambeke, F. Clin. Microbiol. Infect. Off. Publ. Eur. Soc. Clin. Microbiol. Infect. Dis. 2007, 13, 560. 16. Lotlikar, S. R.; Hnatusko, S. B.; Dickenson, N. E.; Choudhari, S. P.; Picking, W. L.; Patrauchan, M. A. Microbiol. Read. Engl. 2013. 17. Merlin, C.; Masters, M.; McAteer, S.; Coulson, A. J. Bacteriol. 2003, 185, 6415. 18. Ogrinc, G.; Kampalath, B.; Tomashefski, J. F., Jr Hum. Pathol. 1998, 29, 65. 19. Kennedy, M. P.; Noone, P. G.; Carson, J.; Molina, P. L.; Ghio, A.; Zariwala, M. A.; Minnix, S. L.; Knowles, M. R. Respir. Med. 2007, 101, 76. 20. Vullo, D.; Del Prete, S.; Osman, S. M.; Scozzafava, A.; Alothman, Z.; Supuran, C. T.; Capasso, C. Bioorg. Med. Chem. Lett. 2014, 24, 4402.

21. Prete, S. D.; Vullo, D.; Osman, S. M.; Scozzafava, A.; AlOthman, Z.; Capasso, C.; Supuran, C. T. Bioorg. Med. Chem. 2014, 22, 4537. 22. Larkin, M. A.; Blackshields, G.; Brown, N. P.; Chenna, R.; McGettigan, P. A.; McWilliam, H.; Valentin, F.; Wallace, I. M.; Wilm, A.; Lopez, R.; Thompson, J. D.; Gibson, T. J.; Higgins, D. G. Bioinformatics 2007, 23, 2947. 23. Pinard, M.; Lotlikar, S.; Patrauchan, M. A.; McKenna, R. Sect. F Struct. Biol. Cryst. Commun. 2013, 69, 891. 24. Otwinowski, Z.; Minor, W. In Methods in Enzymology; Charles, W., Carter, J., Eds.; Academic Press, 1997; Vol. 276, pp 307–326. 25. Adams, P. D.; Afonine, P. V.; Bunkóczi, G.; Chen, V. B.; Davis, I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R. J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H. Acta Crystallogr. D Biol. Crystallogr. 2010, 66, 213. 26. Emsley, P.; Cowtan, K. Acta Crystallogr. D Biol. Crystallogr. 2004, 60, 2126. 27. Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J. Appl. Crystallogr. 1993, 26, 283. 28. Davis, I. W.; Murray, L. W.; Richardson, J. S.; Richardson, D. C. Nucleic Acids Res. 2004, 32, W615. 29. Khalifa, R. G. J. Biol. Chem. 1971, 246, 2561. 30. Carta, F.; Maresca, A.; Covarrubias, A. S.; Mowbray, S. L.; Jones, T. A.; Supuran, C. T. Bioorg. Med. Chem. Lett. 2009, 19, 6649. 31. Capasso, C.; Supuran, C. T. J. Enzyme Inhib. Med. Chem. 2014, 29, 379. 32. Capasso, C.; Supuran, C. T. Expert Opin. Ther. Pat. 2013, 23, 693. 33. Capasso, C.; Supuran, C. T. In Novel Antimicrobial Agents and Strategies; Phoenix, D. A., Harris, F., Dennison, S. R., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA, 2014; pp 31–46. 34. Supuran, C. T. Front. Pharmacol. 2011, 2, 34. 35. Capasso, C.; Supuran, C. T. J. Enzyme Inhib. Med. Chem. 2015, 30, 325. 36. Cronk, J. D.; Rowlett, R. S.; Zhang, K. Y. J.; Tu, C.; Endrizzi, J. A.; Lee, J.; Gareiss, P. C.; Preiss, J. R. Biochemistry (Mosc.) 2006, 45, 4351. 37. The PyMOL Molecular Graphics System, Version 1.7.4 Schrödinger, LLC.

Please cite this article in press as: Pinard, M. A.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.05.029