Inhibition of Bacillus anthracis metallo-β-lactamase by compounds with hydroxamic acid functionality

Abstract

Metallo-β-lactamases (MBLs) that catalyze hydrolysis of β-lactam antibiotics are an emerging threat due to their rapid spread. A strain of the bacterium Bacillus anthracis has its ability to produce and secrete a MBL, referred to Bla2. To address this challenge, novel hydroxamic acid-containing compounds such as 3-(heptyloxy)-N-hydroxybenzamide (compound 4) and N-hydroxy-3-((6-(hydroxyamino)-6-oxohexyl)oxy)benzamide (compound 7) were synthesized. Kinetic analysis of microbial inhibition indicated that the both sides of hydroxamic acids containing compound 7 revealed a reversible, competitive inhibition with a K_i value of 0.18 ± 0.06 μM. The result has reflected that the both sides of dihydroxamic acids in a molecule play a crucial role in the binding affinity rather than monohydroxamic containing compound 4 which was unable to inhibit Bla2. In addition, in silico analysis suggested that compound 7 was coordinated with a zinc ion in the active site of enzyme. These observations suggest that the dihydroxamic acid-containing compound may be a promising drug candidate, and a further implication for designing new inhibitors of Bla2.

Keywords:

• Enzyme kinetics
• inhibition
• lactamase

Introduction

The bacterium Bacillus anthracis is responsible for causing anthrax infection, which is often fatal1. Because symptoms arising from the bacterial infection are similar to a common cold, misdiagnosis in the early stage is possible and frequent1. In addition, increasing prevalence of antibiotic resistant B. anthracis is extending our inability to effectively resolve these infections1,2. The primary cause of antibiotic resistance is the emergence of β-lactamases that catalyze the hydrolysis of β-lactam antibiotics such as penicillins, cephalosporins, and carbapenems1. The β-lactamases are categorized into four classes, i.e., A through D3. Unlike class A, C, and D β-lactamases, class B β-lactamase requires Zn²⁺ions in its active site for activity and are referred to as metallo-β-lactamases (MBLs)3. Class B is further categorized into three subclasses: B1, B2, and B3. The Sterne strain of B. anthracis is capable of producing a MBL which is a B1 β-lactamase, called Bla2, containing two zinc ions referred to as Zn₁ and Zn₂4,5. β-Lactamase inhibitors such as clavulanic acid, sulbactam, and tazobactam are commercially available and can be used with existing antibiotics to cure some antibiotic-resistant infections, but these are not effective against MBLs. Although some potential inhibitors have the ability to inhibit MBLs6,7, there are no commercially available inhibitors of MBLs. Considering the time-consuming process to develop a drug, it is urgent to find effective inhibitors of these enzymes.

Due to the crucial role that the Bla2 active site zinc ions play in reaction catalysis, one method of enzyme inhibition is to chelate these metal ions and render them inactive. In the past, effort has been placed on carboxylic acid-containing inhibitors and thiol-containing compounds, which inactivate the nucleophilic hydroxide group responsible for hydrolysis of the substrate8,9. These types of inhibitors are typically very specific for particular enzymes and do not have promising potential as broad-spectrum inhibitors. One potential approach to designing broad-spectrum inhibitors is to chelate the Zn²⁺ions that MBLs employ and, therefore, render the enzyme inactive. The hydroxamic acid functional group has been extensively investigated with matrix metalloproteinase targets, and when linked to a peptide backbone, inactivate the catalytically essential zinc ion of these enzymes10. Hydroxamic acid-containing compounds have also been investigated against MBLs, as targets for Aeromanas hydrophila MBL and various other MBLs, whose K_i values are within the micromolar range11,12. Here, we investigate if small molecules that contain hydroxamic acid functional groups would be able to bind to the zinc ions present in Bla2, thus inhibiting the enzyme’s activity. Novel hydroxamic acid-containing compounds were synthesized, and the effectiveness of Bla2 inhibition was tested using kinetic analysis. In addition, an in silico analysis in the binding interaction between Bla2 and the hydroxamic acid-containing compounds provide insight for the modification of the future study.

Materials and methods

General procedures

All microwave-assisted reactions were performed in a Milestone Start Microwave Labstation using the MonoPREP device and a single vessel reactor (50 mL of volume) covered in PTFE with normal pressure, 1.5, or 15 bar pressure caps. Melting points were determined using a Meltemp-II laboratory device and are uncorrected. All reagents and solvents were used without further purification from commercial sources. Organic extracts were dried over anhydrous MgSO₄ or Na₂SO₄. Reactions were monitored by analytical thin-layer chromatography (TLC) commercially prepared plates coated with 0.25 mm of self-indicating Merck Kieselgel 60 F254 and visualized by UV light at 254 nm. Preparative-scale silica gel flash chromatography (for purification of analytical samples only) was carried out by standard procedures using Merck Kieselgel 60 (230–400 mesh). Where not stated otherwise, assume standard practices have been applied.

The ¹H and ¹³C NMR spectra were obtained and recorded on a Varian 500 NMR system, at 500 MHz and 125 MHz for ¹³C, and were referred to solvent signals13. High-resolution mass spectra were obtained in the Baylor University Mass Spectrometry Core Facility on a Thermo Scientific LTQ Orbitrap Discovery using + ESI. IR data were recorded on a Nicolet 380 FT-IR Spectrometer using a diamond ATR (attenuated total reflectant) unit.

Synthesis of 3-(heptyloxy)-N-hydroxybenzamide (compound 4)

The alkylation reaction between phenol and alkyl halide was made in potassium carbonate (see the scheme in Figure 4A). Methyl 3-hydroxybenzoate (5.048 g, 33.18 mmol) was mixed with 1-iodoheptane (10.80 mL, 65.87 mmol), potassium carbonate (9.084 g, 65.73 mmol), and acetone (50 mL) in a reflux apparatus. The reaction mixture was refluxed for 3 h. The excess potassium carbonate from the mixture was removed using gravity filtration. The solvent was removed under reduced pressure. The crude was purified using column chromatography with ethyl acetate:hexanes (1:9) to give colorless oil. Yield: 2.805 g (33.77%). Data analysis suggested a successful synthesis of methyl 3-(heptyloxy)benzoate14, which serves as a precursor for 3-(heptyloxy)-N-hydroxybenzamide.

Following the successful precursor synthesis, methyl 3-(heptyloxy)benzoate (0.422 g, 1.69 mmol) was mixed with hydroxylamine hydrochloride (0.344 g, 5.10 mmol), potassium hydroxide (0.568 g, 10.14 mmol), and methanol (15 mL) in a microwave reactor vessel with a 1.5 bar pressure cap. The solution was placed in the microwave for three cycles consisting of a 6 min heating time to a temperature of 55 °C and maintained at temperature for additional 15 min. The solvent was removed under reduced pressure. The water was added and neutralized with 1 M hydrochloric acid to a pH of 6.8. After removing solvent on vacuum, the crude was purified using column chromatography (4:6:2 chloroform:hexanes:methanol) to give a white solid. Yield: 72.0 mg (16.95%); 1H NMR (DMSO-d₆): δ 11.21 (s, 1H), 8.98 (s, 1H), 7.43–7.26 (m, 4H), 3.97 (t, 2H), 1.62–1.78 (m, 2H), 1.19–1.42 (m, 8H), 0.82 (t, J = 5.75, 3H); IR (cm ^− 1) 3181.36, 2922.17, 2851.84, 2538.25, 2161.01, 2024.85, 1975.59, 1601.24, 1547.62, 1462.47; HRMS Calcd C14H21NO3 (M+) 251.15, found (M+) 251.145.

Synthesis of N-hydroxy-3–(6-(hydroxyamino)-6-oxohexyloxy)benzamide (compound 7)

Methyl 3-hydroxybenzoate (2.015 g, 13.24 mmol) was mixed with ethyl 6-bromohexanoate (4.80 mL, 27.0 mmol), potassium carbonate (3.666 g, 26.53 mmol), and acetone (25 mL) in a microwave reactor vessel with a 15 bar pressure cap. The mixture was microwaved for two cycles 6 min at 164 °C, followed by 15 min at 164 °C. The solvent was removed on a rotary evaporator, and the residue was purified by column chromatography using ethyl acetate:hexanes (1:9) to give a light yellow oil. Methyl 3-(6-ethoxy-6-oxohexyloxy)benzoate 6 (0.640 g, 2.17 mmol) was mixed with hydroxylamine hydrochloride (0.890 g, 13.19 mmol), potassium hydroxide (1.472 g, 26.28), and methanol (20 mL) in a microwave reactor vessel with a 15 bar pressure cap. The reaction mixture was microwaved for 6 min at 80 °C and for 15 min at 80 °C. The reaction mixture was neutralized with 6 M hydrochloric acid. The aqueous phase was extracted several times with ethyl acetate. The organic fraction was concentrated on the rotary evaporator. The crude product was recrystallized in acetone to provide a white crystalline product. Yield: 70.0 mg (11.4%); m.p. 153 °C; ¹H NMR (dimethylsulfoxide (DMSO)-d₆): δ 11.19 (s, 1H), 10.37 (s, 1H), 9.04 (s, 1H), 8.69 (s, 1H), 7.27–7.36 (m, 3H), 7.05 (dq, J = 3.63, 1H), 3.97 (t, J = 12.9, 2H), 1.97 (t, J = 14.6, 2H), 1.38–1.71 (m, 6H); ¹³C NMR: 169.10, 163.98, 158.58, 134.14, 129.60, 119.04, 117.48, 112.54, 67.52, 32.26, 28.40, 25.19, 24.93. IR (cm⁻¹) 3169.70, 3030.29, 2870.84, 2535.67, 2160.29, 2029.23, 1976.57, 1626.95, 1605.38, 1573.76, 1536.29, 1484.85; HRMS Calcd C₁₃H₁₈N₂O₅ (M+) 282.1216, found (M+) 282.111.

Inhibition studies

The enzyme assay method with penicillin G as a substrate was described previously5. The assay for IC₅₀, the concentration of inhibitor necessary to inhibit 50% of the enzymatic activity, was performed under the previous enzyme assay conditions after incubating the reaction mixture for 10 min with varying concentrations of compounds 4 and 7 before initiating the reaction with a fixed concentration of penicillin G (0.78 mM). The enzyme concentration used was 0.13 μg/mL (4.8 nM). A separate series of assays for each inhibitor was run at fixed inhibitor concentrations (0–0.8 μM for compound 4 and 0–1.0 μM for compound 7) in which, at each inhibitor concentration, the concentration of the substrate penicillin G was varied from 0.4 to 1.6 μM. The resulting data were analyzed for mode of inhibition by Lineweaver–Burke plots and were also fit using non-linear regression through SigmaPlot version 11.0 (Sigma, St. Louis, MO), using the competitive inhibition equations: ν = V_maxS/[K_m·(1 + 1/K_i) + S].

Mass spectrometry

Mass spectrometry studies were completed to test if compounds 4 and 7 covalently bind to the enzyme or not. For the analysis, an Accela liquid chromatograph coupled to an LTQ Orbitrap Discovery mass spectrometer (Thermo Electron, Bremen, Germany) was used with positive electrospray ionization (+ESI). Protein samples (150 μM) were injected (10 μL) into the LC system consisting of a 15 cm × 2.1 mm (5 μm, 80 Å) Exteded-C18 column (Agilent Technologies, Palo Alto, CA) with a binary mobile phase gradient containing 0.1% (v/v) formic acid in water and acetonitrile. Additional chromatographic parameters were as follows: column temperature 30 °C; flow rate 400 μL/min. Full-scan mass spectra (m/z range: 200–4000) of eluting compounds were obtained on the Orbitrap mass analyzer and processed using Xcalibur v.2.0.7 software0 (Sigma, St. Louis, MO). Electrospray source conditions were the following: sheath and auxiliary gas flow 60 and 10 arbitrary units (a.u.), respectively; heated capillary temperature 275 °C; electrospray voltage 4.5 kV; capillary voltage 27 V; tube lens voltage 240 V.

Construction of the Bacillus subtilis strain containing bla2 gene

The gene mbla2 for mature Bla2 was obtained by PCR with the primer set bla2-F2 (caatttcttctgtacaagcagaacgaaaggtagagcataa) and bla2-R2 (atacccgggttattttaacaaatccaatg) from the plasmid pET24a-bla24,5. DNA fragment for N-terminal signal sequence of the Bla2 was amplified by PCR with the primer set bla2-F1 (attggatccatgaaaaatacattattaaaattaggggtatgtgttagtttactag) and bla2-R1 (tgcttgtacagaagaaattgtactaacaaatggagttattcctagtaaactaacacatacc) and fused with the mbla2 by a fusion PCR. The resulting PCR product was digested with BamHI/XmaI and cloned into the corresponding site of a plasmid pHT0115 to construct plasmid pHT01-bla2. The plasmid pHT01-bla2 was introduced into B. subtilis 168 using a previously reported method16.

Antibacterial activity assay: minimum protective concentration (MPC)

The MPC of β-mercaptoethanol (BME) and compounds 4 and 7 was determined for B. subtilis cells with or without the Bla2 plasmid. Bacterial cells were incubated 24 h in Luria–Bertani (LB) media with or without chloramphenicol (5 μg/mL) as a selection agent. This culture was then spread on agar plates with or without chloramphenicol (30 μg/mL) and incubated at 37 °C for 12 h to determine the colony forming unit (CFU) of the 24 h culture. The 24 h culture was stored at room temperature while the CFU was being determined.

After determining the CFU, the 24 h culture was able to be diluted with sterile LB media to give an approximately 10⁶ bacteria per mL. Ampicillin was prepared fresh to a final concentration of 1.5 μg/mL in Mueller–Hinton (MH) broth. BME solution was also freshly prepared to a stock concentration of 17.2 mg/mL and then serially diluted twofold in MH broth. Solutions of compounds 4 and 7 were prepared by dissolving the compounds in 30% (v/v) DMSO in water, followed by diluting to the final stock concentration in MH broth without additional DMSO, and then immediately adding to the microplate. The final concentration of DMSO in the microplate was 3%. A volume of 50 μL of ampicillin was added to the wells of a 96-well microplate with either 50 μL BME for the MPC determination of BME, 50 μL MH broth containing compound 4 for the MPC determination of compound 4, or 50 μL MH broth containing compound 7 for the MPC determination of compound 7. After the addition of 50 μL of the bacterial cell samples, the wells were mixed and incubated at 37 °C for 24 h. The plates were measured every 4 h at 620 nm with a Thermo Scientific Multiskan MCC/340 microplate reader (Thermo Fisher Scientific, Waltham, MA) after 20 s of shaking at 1200 rpm.

In silico analysis

The program modeler17 was used to calculate the homology model of Bla2. A high-resolution crystal structure (1.35 Å resolution) of BcII (1mqoA) was used as a template to compute the model of Bla2, and the PROCHECK and QMEAN programs were used for verification of the homology model18,19. The docking studies were performed using AutoDock, which employs the optimization algorithm in its scoring function20. For the 3D structures of compounds 4 and 7, the MM2 force field in the ChemBio3D ver14 program was used for energy minimization. All the possible torsion angles in the compounds were set free to carry out flexible docking. Hydrogens were added to the protein using AutoDock Tools, and Gasteiger partial charges were assigned to the protein.

A grid box of 60 Å × 60 Å × 60 Å with 0.375 Å grid spacing centered on x−y−z coordinates (12.619, 10.184, and 30.051) covering the potential binding pocket of Bla2 was set for configuration in Autodock 4.2, which uses Autodock force field modified from Amber force field. The results produced by the Autodock 4.2 showed the binding energies and hydrogen bonding interactions between the protein and compounds 4 and 7. The structure with lowest energy is chosen for computing intermolecular binding energies.

Results and discussion

Synthesis strategy

We designed and synthesized small molecules based on metal ion chelation which have mono- and di-hydroxamic acid moiety at the terminal sides of a molecule in order to target zinc metalloenzymes. The rational of designing the molecules is that (1) the hydroxamate moiety previously showed the possibility of binding the functional zinc ion in MBL11,12 and that (2) the zinc ions buried in the hydrophobic active site may be reached by a benzene-attached polymethylene chain where the meta-position was used for the attachment because it was better than the para-position in a similar case14. We are also curious about the binding affinity based on the direction of between the head and the tail of hydroxamic acid moieties at the active site of a zinc metalloenzyme. First, a mono-hydroxamic acid containing-compound, 3-(heptyloxy)-N-hydroxybenzamide 4 was synthesized. As starting materials, commercially available compounds 1 and 2 were used. Methyl 3-(heptyloxy)benzoate 3 was prepared under microwave-assisted reaction in the presence of the compounds 1 and 2 in the solution of potassium carbonate and acetone. The crude 3 was chromatographed, eluting with 4:6:2 chloroform:hexanes:methanol. The subsequent microwave-assisted reaction was performed in the presence of potassium hydroxide and hydroxyl amine in methanol to provide of 3-(heptyloxy)-N-hydroxybenzamide 4 in a (16.9%) 17% yield (Figure 1A).

Figure 1. Schemes of the syntheses of hydroxymate-containing compounds. (A) The scheme of the synthesis of 3-(heptyloxy)-N-hydroxybenzamide (compound 4). (B) The scheme of the synthesis of N-hydroxy-3-((6-(hydroxyamino)-6-oxohexyl)oxy)benzamide (compound 7).

For a dihydroxamic acid-containing compound, N-hydroxy-3-((6-(hydroxyamino)-6-oxohexyl)oxy)benzamide 7 was synthesized using commercially available compounds 1 and 5, as starting materials. Methyl 3-((6-methoxy-6-oxohexyl)oxy)benzoate 6 was prepared under microwave-assisted reaction in the presence of compounds 1 and 5 with potassium carbonate in acetone to provide 6 in a moderate yield (70%). The subsequent microwave-assisted reaction of 6 was performed in the presence of potassium hydroxide and hydroxyl amine hydrochloride in methanol to provide the N-hydroxy-3-((6-(hydroxyamino)-6-oxohexyl)oxy)benzamide 7 in an 11% yield (Figure 1B).

Inhibition tests

To dissolve the two compounds 4 and 7, 15% (v/v) DMSO was used. Although the concentration of DMSO was low (less than 1%) in the inhibition test conditions, the effect of DMSO concentration on the activity of Bla2 was determined (Supplementary Figure S1). No significant effect was observed at up to 2% DMSO in the incubation with Bla2, and then a sigmoidal decrease in activity was observed. Thus, the particular concentration of DMSO used in this assay has no effect on enzyme inhibition.

To examine the possibility of inhibition of Bla2 activity with compounds 4 and 7, IC₅₀ values were obtained. Concentrations of compound 4 were used from a range of 1.0 nM to 100 μM and compound 7 from a range of 0.34 nM to 100 μM to determine IC₅₀ values. The data in the presence of the compounds were fit to a concentration–response plot with the equation v_i/v_o = 1/(1 + ([I]/IC₅₀)^h), where I is an inhibitor and h is the Hill coefficient (the Hill coefficients used were between 0.5 and 1). Although compound 4 did not inhibit the activity of Bla2 up to 100 μM (data not shown), Figure 2 shows the result of inhibition by compound 7 with an IC₅₀ value of 0.48 ± 0.2 μM, where the data were fit to the concentration-response plot, and the IC₅₀ value was calculated from the plot at 50% inhibition. In an effort to understand the mode of inhibition, inhibitory enzyme assay was performed using various concentrations of substrate (i.e., penicillin G) and inhibitor (i.e., compound 7). In order to obtain the mode of inhibition, kinetic analysis was performed with addition of compound 7. As shown in Figure 3, the Lineweaver–Burke plots with a nest of lines that intersect at the y-axis are diagnostic of competitive inhibition. K_i values were determined on the basis of the plots with an average K_i value of 0.18 ± 0.06 μM for compound 7, where K_i values were calculated from the slope based on K_m and apparent K_m values were obtained from Figure 3. In an effort to confirm competitive inhibition, 10 μM ZnSO₄ was added to the assay solution. The comparison of K_i values in the presence or absence of additional ZnSO₄ would suggest whether compound 7 binds to the active site of the enzyme or simply binds to zinc ions. The inhibition assay in the presence of additional free zinc ion resulted in a slightly increased IC₅₀ value (0.52 ± 0.3 μM), but within the error range. The IC₅₀ values in the presence and absence of additional free zinc ions are almost identical. These observations would lead us to conclude that the mode of inhibition follows a competitive inhibition.

Figure 2. Concentration–response plots for Bla2 inhibition with compound 7.

Figure 3. Lineweaver–Burke plots of inhibition of Bla2 y compound 7. The concentrations of compound 7 are 0 (circle), 0.5 (square), and 1.0 μM (triangle). Assays were performed in 50 mM MOPS, pH 7.0. Kinetic constants were determined by fitting of the data to the equation v = V_maxS/[K_m·(1 + 1/K_i) + S] which indicates competitive inhibition.

Previous studies have identified various compounds with low micromolar to nanomolar IC₅₀ values; for example, mercaptoacetic acid thiol ester compounds inhibit various MBLs with IC₅₀ values as low as 2 μM, while disubstituted succinic acids with various hydrophobic substituents have shown inhibition as low as 2.7 nM against IMP-1 from P. aeruginosa8,21,22. To our knowledge, this is the first report of inhibitors specifically against Bla2 from B. anthracis, and sub-micromolar range has not yet been achieved by the previously reported inhibitors of Bla2. Thus, compound 7 shows great promise for potential use against antibiotic β-lactam-resistant anthrax infections.

To have a complete understanding of the interaction between the compounds of interest and Bla2, it is important to investigate whether or not pre-incubation time had an effect on overall inhibition efficiency. Incubation of Bla2 with the compounds was varied with time at 0, 10, 20, and 30 min, but no significant difference in inhibition was detected, indicating that the inhibition does not depend on time. In order to assess the possibility of covalent bindings by the two compounds, mass spectrometry was used after appropriate procedures (see Method and materials section). ESI-mass spectrometry displays the same molecular mass in the absence and presence of either compound, suggesting that no covalent bonds between Bla2 and the compounds exist (Supplementary Figure S2). In combination, these results strongly support the idea that inhibition by compound 7 is reversible.

Furthermore, the minimum protective concentration (MPC) was determined, which is the lowest concentration at which antibiotics applied in cell growth are protected. We used compound 4 or 7 in combination with 0.5 μg/mL ampicillin to acquire MPC values. At 0.5 μg/mL ampicillin, there is no significant growth on B. subtilis harboring bla2 gene up to 24 h (Supplementary Figure S3). It should also be mentioned here that 2% DMSO used should insignificantly affect the MBL activity, and no cell growth effect by 2% DMSO was observed. It should also be noted that we used penicillin G for in vitro inhibition assays, but for MPC value determination we instead used ampicillin as a β-lactam antibiotic because instability of penicillin G is often observed for prolonged period of time23 while ampicillin is widely used for co-administration tests. It was determined that 16 h was the best incubation time and gave the most reproducible results for B. subtilis. For compound 4, there was no detectable MPC value of either the wild-type or Bla2 carrier, while compound 7 exhibited the MPC of 2.0 mg/mL after 16 h against wild-type B. subtilis and an MPC of 1.0 mg/mL after 16 h against the Bla2 carrier, both in the presence of 0.5 mg/mL ampicillin. This observation is consistent with the fact that compound 7 shows more effective inhibition in vitro than does compound 4. We also probed the possibility of inhibition by only compounds 4 and 7 and found that no minimal inhibition concentration was detected up to 4 mg/mL of each compound. This observation provides additional evidence that compound 7 shows synergic effect with ampicillin. One considerable reason for the inflated MPC value for compound 7 may be that it binds to the B. subtilis cell wall non-specifically, which depletes the free concentration of compound 7 available to interact with the MBL. Also, the determination of IC₅₀ values was performed in vitro with a fixed enzyme concentration while the determination of MPC values was carried out in the presence of an increased MBL concentration caused by continuous MBL secretion from the Gram-positive bacteria B. subtilis.

In silico analysis

In an effort to understand the interaction between compound 7 and the target enzyme Bla2 in detail, we attempted to use a docking program, Autodock 4.224, where Autodock 4.2 uses an optimization algorithm and gradient optimization method to dock ligands into the binding site of a target macromolecule. The docking result with the lowest binding interaction energy was selected. The interaction energy of binding between Bla2 and the compound 7 was −7.8 kcal/mol. The affinity between the compound and the active site of Bla2 is primarily due to coordination with a zinc ion (Zn₁) and a number of amino acid residues in the pocket of the active site of Bla2, which are involved in non-covalent interactions including hydrogen bonds (Figure 4A and B). The docking result showed that the hydroxamic acid moiety extended from the hydrocarbon tail coordinates to Zn₁ with a 1.8 Å distance from oxygen of the hydroxyl group. In addition, three hydrogen bonds are present in the hydroxamic acid moiety near the benzene ring; the hydrogen atom of the hydroxyl group of Ser201 and the oxygen of hydroxamic acid moiety, the nitrogen of the amide group of His239, and the hydrogen atom of the hydroxamic acid moiety. The compound 7 clearly showed a binding interaction with Zn₁ (this zinc is coordinated to His117 and His178, and Cys197). To compare the binding interactions between compounds 7 and 4, the molecular docking for compound 4 was performed under the same conditions that were applied to compound 7. The binding energy value of compound 4 was significantly higher (−5.82 kcal/mol) than that of compound 7. This might explain the reason why the compound 4 was unable to inhibit Bla2 effectively. Although the binding affinity was lower, the monohydroxamic acid moiety of compound 4 was coordinated to Zn₁ and Zn₂ with a distance of 2.1 and 1.7 Å, respectively (Figure 4C and D). In the docking results, three hydrogen bonds formed between Bla2 and compound 4, where Asp119 and Asn209 and His239 are involved in the bindings. It is interesting to note that all the bindings engaged in the hydroxamic acid moiety and the other portion of the compound do not participate in any interaction. This reflects that the other portion of the compound would be imperative for inhibition of the enzyme.

Figure 4. Molecular docking between Bla2 and the compounds (4 and 7). (A and B) A snapshot of binding mode of compound 7 and Bla2. (C and D) A snapshot of binding mode of compound 4 and Bla2.

Conclusion

We have described the successful syntheses of two novel compounds containing mono and dihydroxamic acids functional groups, compounds 4 and 7, for the purpose of examining inhibition of Bla2’s β-lactamase activity. While kinetic analysis of microbial inhibition showed that compound 4 was not successful, compound 7 proved to be an extremely effective inhibitor of Bla2’s penicillinase activity via competitive inhibition. In silico analysis further confirms the inhibition type by demonstrating key interactions between compound 7 and amino acids at the active site of the enzyme. Taken together, these observations show that compound 7 may be a useful drug candidate as well as a lead compound for further development. To further understand and develop the compound, various modifications with different functionalities such as carboxylic acid, oxime, aldehyde or sulfonamide could be considered with structure–activity relationship analysis. Such modifications of the compound are underway in our laboratories. It should be noted that these promising hydroxamic acid-containing compounds should be tested further to understand the specificity and cytotoxicity prior to clinical use.

Declaration of interest

The authors report no declaration of interest. This work was funded by FRC at Northeastern State University (to S-K.K.), and Welch Departmental Grant (Z-0036) and Sam Taylor Fellowship at McMurry University (to H.S.).

Supplementary material available online

Supplementary Figures S1–S3.

Acknowledgements

Authors also thank Prof. Charles Garner for data acquisition of ¹H, ¹³C NMR on 500 MHz supported by the NSF (Award #CHE-042802) and Dr. Alejandro Ramirez for data analysis of High-Resolution Mass Spectroscopy using + ESI at Baylor University. H.S. participated in writing this manuscript as a corresponding author.