Diethylalkylsulfonamido(4-methoxyphenyl)methyl)phosphonate/phosphonic acid derivatives act as acid phosphatase inhibitors: synthesis accompanied by experimental and molecular modeling assessments

Abstract

Purple acid phosphatases (PAPs) are binuclear metallo-hydrolases that have been isolated from various mammals, plants, fungi and bacteria. In mammals, PAP activity is associated with bone resorption and can lead to bone metabolic disorders such as osteoporosis; thus human PAP is an attractive target to develop anti-osteoporotic drugs. The aim of the present study was to investigate inhibitory effect of synthesized diethylalkylsulfonamido(4-methoxyphenyl)methyl)phosphonate/phosphonic acid derivatives as potential red kidney bean PAP (rkbPAP) inhibitors accompanied by experimental and molecular modeling assessments. Enzyme kinetic data showed that they are good rkbPAP inhibitors whose potencies improve with increasing alkyl chain length. Hexadecyl derivatives, as most potent compounds (K_i = 1.1 µM), inhibit rkbPAP in the mixed manner, while dodecyl derivatives act as efficient noncompetitive inhibitor. Also, analysis by molecular modeling of the structure of the rkbPAP–inhibitor complexes reveals factors, which may be important for the determination of inhibition specificity.

Keywords:

• Acid phosphatase
• enzyme inhibition
• phosphonic acid
• synthesis

Introduction

Purple acid phosphatase (PAP; E.C.3.1.3.2) is a binuclear hydrolase that has been isolated from various mammals and plants¹. Mammalian PAP is also known as tartrate-resistant acid phosphatase² or osteoclastic acid phosphatase³. These enzymes are distinguished from other acid phosphatases by their characteristic purple color in concentrated solution⁴. PAPs catalyze the hydrolysis of phosphorylated substrates under neutral to acidic conditions in vivo¹. Several crystal structures of PAPs have been described, including those from plants (red kidney bean and a sweet potato PAP) as well as from mammals (pig, human and rat)⁵ and references 90–93 therein. The enzyme structures from the two plant organisms are homodimers consisting of 55 kDa subunits, while the mammalian PAPs are 35 kDa monomers^6–9. Despite the apparent difference between the structures of plant and animal PAPs, and the low degree of overall amino acid sequence homology between enzymes from different kingdoms, their catalytic sites display a remarkable similarity⁵ and references 2, 89, 107 therein (Figure 1) and include two metal centers, which are coordinated by seven invariant ligands (one aspartate, one tyrosine and one histidine for Fe³⁺and two histidines and one asparagine for Zn²⁺), with an aspartate residue bridging the two ions¹⁰. Due to the structural similarity in their catalytically relevant active sites, all PAPs are believed to employ highly conserved mechanistic strategies¹.

Figure 1. (Left) Active site of all known PAPs has an Fe³⁺in the active site, which coordinates with an invariant tyrosine ligand, producing a ligand to metal CT transition giving rise to the characteristic purple color of the enzyme. PAPs active site also contains a divalent metal ion which may be either Zn²⁺, Mn²⁺or Fe²⁺. The seven metal ion-coordinating residues are invariant among PAPs from different sources. Residue labels refer to the sequences of red kidney bean PAP and pig PAP (in brackets or parentheses). Taken from Schenk et al.⁵ (Right) SDS–PAGE illustrates PAP purity (lane 1). M, molecular size markers.

The biological roles of PAPs are diverse and depend on the host organism. Plant PAPs are believed to assist in mobilizing organic phosphate esters in soil during germination¹¹, and mammalian PAPs have been linked with iron transport during pregnancy¹². Both may be involved in bacterial killing since they are able, at least in vitro, to generate reactive oxygen species via Fenton chemistry as a response against pathogens^9,13,14. The association of mammalian PAP activity with osteoporosis^2,15–17 has prompted investigators, around the world, to examine its potential as a therapeutic target for the treatment of this disease, and to develop potent inhibitors of this enzyme^17,18. It is known that PAP is secreted into the bone-resorptive space in mammals², and PAP activity has long been associated with bone resorption¹⁹. In addition, transgenic mice in which the PAP enzyme was over-expressed developed osteoporosis due to increased bone resorption²⁰, whereas the corresponding knockout mice exhibited the opposite condition (osteopetrosis), further implicating the role of PAP in this process⁴, and establishing PAP as a drug target in the treatment of osteoporosis²¹. Although clinical treatments are available for osteoporosis, notably bisphophonates, which inhibit farnesylpyrophosphate synthase in osteoclast²², but these drugs have significant side effects and compliance issues. Also, a number of PAPs inhibitors including fluoride²³, and several tetrahedral inorganic oxyanions, including phosphate, arsenate, vanadate, tungstate and molybdate^7,24 have been discovered, but most of these do not possess drug-like properties and so are not suitable for further modifications that may convert them into chemotherapeutic agents. Myers et al.²⁵ have described a number of simple phosphonates with pendant metal-binding groups, such as thiol, phosphonate and carboxylate, and these have shown inhibitory activity against red kidney PAP (rkbPAP), in the range 80–3000 µM. Moreover, several modified phosphotyrosine containing tripeptides have also shown inhibitory activity toward several mammalian and plant PAPs, with IC₅₀ values in the mid-micromolar range²¹. More recently, α-alkoxynaphthylmethylphosphonic acids and acyl derivatives of α-aminonaphthylmethylphosphonic acid have also been shown to be PAP inhibitors with K_i and IC₅₀ values in the low micromolar range^18,26. These molecules were designed as the derivatives of 1-naphthylmethylphosphonic acid, a PAP inhibitor previously reported by Schwender et al.³ A number of phosphonate- and carboxylate-containing derivatives of these substrates have been synthesized, and these displayed mid- to low-micromolar inhibitory potency against several mammalian and plant PAPs²¹.

The current study is particularly novel and significant because to the best of our knowledge, no previous similar studies have been reported on possible inhibitory potential of diethylalkylsulfonamido(4-methoxyphenyl)methyl)phosphonate/phosphonic acid derivatives on the PAP enzyme. Based on previous lead compound(s) and rational drug design, different sulfonyl derivatives of alkylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid with varying lengths of alkyl chains were synthesized and evaluated as PAP inhibitors. The sulfonyl and alkyl groups were introduced for two reasons: (i) the sulfonyl group was expected to improve the water solubility of these compounds; (ii) McGeary et al. have previously reported that PAP inhibitors bearing an alkoxy group with long alkyl chain were superior to those that did not, suggesting effective hydrophobic interactions between the alkyl group of the inhibitors and nonpolar surface residue near the enzyme’s active site¹⁸; (iii) a nonclassical bioisosteric replacement was made in which the acyl moiety was changed to sulfonamide one. Therefore, inhibitors of general structures in Scheme 1 with C₆, C₈, C₁₂ and C₁₆ alkyl chains were examined. Moreover, to evaluation of coordination of phosphonate moiety to the bimetal center, the ethyl groups were attached to the oxygen atoms of phosphonate. The aim of the current study was also to expand this working framework through suggesting a molecular modeling component to advance the mechanistic understanding of PAP–drug interactions.

Scheme 1. Synthesis of diethylalkylsulfonamido(4-methoxyphenyl)methyl)phosphonate/phosphonic acid derivatives.

Materials and methods

Chemistry

All starting materials, reagents, and solvents were purchased from Merck Finnigan (Germany). The purity of the synthesized compounds was confirmed by thin layer chromatography (TLC) using various solvents of different polarities. Merck silica gel 60 F254 plates were applied for analytical TLC. Column chromatography was performed on Merck silica gel (70–230 mesh) for purification of the intermediate and final compounds. Melting points were determined on a Kofler hot stage apparatus (Vienna, Austria) and are uncorrected. ¹H NMR spectra were recorded using a Bruker 500 spectrometer (Bruker, Rheinstatten, Germany), and chemical shifts are expressed as δ (ppm) with tetramethylsilane (TMS) as an internal standard. Coupling constants (J) are reported in hertz and peak multiplicities described as singlet (s), doublet (d), triplet (t), multiplet (m) or broad (br). The IR spectra were obtained on a Shimadzu 470 (Shimadzu, Tokyo, Japan) spectrophotometer (potassium bromide disks). The mass spectra were run on a Finnigan TSQ-70 spectrometer (Finnigan) at 70 eV.

General procedure for the syntheses and spectral data of the synthesized compounds

General procedure for the synthesis of diethyl(amino(4-methoxyphenyl)methyl)phosphonate (2)

This compound was prepared according to the general procedure described in the literature^3,18,26,27. NH₄OAc (7.70 g, 0.10 mol) was added to EtOH (20 mL) in the round-bottomed flask containing the activated molecular sieves, followed by the addition of p-methoxybenzaldehyde (1) (13.60 mL, 0.10 mol) and diethyl phosphite (12.90 mL, 0.10 mol). The reaction mixture was stirred at 60 °C under nitrogen atmosphere for 48 h. Once it has cooled to room temperature, conc. HCl was added to the resulting mixture to acidify it to pH 1, and then was washed with Et₂O (50 mL). Addition of 2 M NaOH to the aqueous layer changed the pH to 11, and the product was extracted with dichloromethane (100 mL) to give the free base of diethyl (amino(naphthalen-1-yl)methyl)phosphonate²⁷ as yellow oil. The product was later treated with HCl gas in EtOH (10 mL)–Et₂O (10 mL). Et₂O (20 mL) was further added to the resulting solution, and the solvent was evaporated to obtain the hydrochloride salt of diethyl compound 2 as white solid.

General procedure for the preparation of diethyl (alkylamido(4-methoxyphenyl)methyl)phosphonates (3a–3d)

This compound was prepared according to the general procedure described in the literature²⁷. A solution of DIPEA (2.60 mL, 15.2 mmol) and the hydrochloride salt 2 (0.50 g, 1.52 mmol) in CH₂Cl₂ (5 mL) were then added dropwise to the sulfonyl chloride and the reaction mixture was stirred under argon N₂ for 24 h. The crude product was purified using silica flash column chromatography (10–60% EtOAc in light petroleum) to give 3a–3h.

General procedure for the preparation of (alkylamido(4-methoxyphenyl)methyl)phosphonic acids (4a–4d)

This compound was prepared according to the general procedure described in the literature²⁷. NaI (5 equiv) and distilled MeCN (5 mL) were added to the corresponding diethyl ester 3a–3d (1 equiv). Distilled TMSCl (5 equiv) was then added to the mixture, which was then stirred at 40 °C under N₂ for 18 h. MeCN was removed in vacuo. The precipitate was then extracted with EtOAc (3 × 10 mL), washed with NaS₂O₃ (10 mL), water and NaCl (2 × 10 mL), followed by drying with NaSO₄. The product was then co-evaporated with toluene (3 × 5 mL), and purified under high vacuum to give the corresponding phosphonic acids 4a–4d.

Diethyl(amino(4-methoxyphenyl)methyl)phosphonate (2)

Molecular weight: 204 (g/mol), FT-IR (KBr, cm⁻¹): 3429 (stretch NH), 2935 (stretch CH, aromatic), 2904 (stretch CH, aliphatic), 1608 (stretch C=C, aromatic), 1519 (bending NH), 1462 (bending CH₂), 1388 (bending CH₃), 1242 (stretch C–O), 1026 (stretch C–O, OCH₃), 948 (stretch CH, oop), 547 (bending NH).

Diethyl(hexylsulfonamido(4-methoxyphenyl)methyl)phosphonate (3a)

Molecular weight: 421 (g/mol), FT-IR (KBr, cm⁻¹): 3437 (stretch NH), 2931 (stretch CH, aromatic), 2873 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1516 (bending NH), 1465 (bending CH₂), 1392 (bending CH₃), 1311, 1188 (stretch O=S=O), 1249 (stretch C–O), 1041 (stretch C–O, OCH₃), 952 (stretch CH, oop), 551 (bending NH).

¹H NMR (CDCl₃, 300 MHz): δ (ppm): 7.68–7.66 (d, J = 6 Hz, 1H, NH), 7.39–7.36 (dd, J = 1.5, 1.5 Hz, 2H, C_(2,6)H aromatic), 6.80–6.78 (d, J = 6.6 Hz, 2H, C_(3,5)H aromatic), 5.52–5.44 (dd, J = 7.2, 7.2 Hz, H, CH–P), 4.15–4.04 (m, 2H, C_(1,1*)H–OP), 3.88–3.79 (m, 1H, C₍₁₎H–OP), 3.72 (s, 3H, CH₃–O), (m, 1H, C_(1*)H–OP), 2.16 (t, J = 5.7 Hz, 2H, CH₂–S), 1.53 (t, J = 5.4 Hz, 2H, CH₂–CH₂S), 1.28 (t, J = 5.4 Hz, 3H, C₍₂₎H₃–CH₂O), 1.15 (bro, J = 7.5 Hz, 10H, CH aliphatic), 1.03 (t, J = 5.7 Hz, 3H, C_(2*)H₃–CH₂O), 0.80 (t, J = 5.4 Hz, 3H, CH₃–RS).

Diethyl((4-methoxyphenyl)(octylsulfonamido)methyl)phosphonate (3b)

Molecular weight: 449 (g/mol), FT-IR (KBr, cm⁻¹): 3437 (stretch NH), 2927 (stretch CH, aromatic), 2858 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1512 (bending NH), 1465 (bending CH₂), 1396 (bending CH₃), 1303, 1215 (stretch O=S=O), 1246 (stretch C–O), 1037(stretch C–O, OCH₃), 948(stretch CH, oop), 721 (rocking CH₂), 551 (bending NH).

¹H NMR (CDCl₃, 300 MHz): δ (ppm): 7.54–7.49 (dd, J = 3.6, 3.6 Hz, 1H, NH), 7.39–7.35 (dd, J = 2.1, 2.1 Hz, 2H, C_(2,6)H aromatic), 6.81–6.78 (dd, J = 9 Hz, 2H, C_(3,5)H aromatic), 5.53–5.46 (dd, J = 20.7 Hz, H, CH–P), 4.14–4.04 (m, 2H, C_(1,1*)H–OP), 3.88–3.80 (m, 1H, C₍₁₎H–OP), 3.73 (s, 3H, CH₃–O), 3.68–3.59 (m, 1H, C_(1*)H–OP), 2.17 (t, J = 7.8 Hz, 2H, CH₂–S), 1.54 (t, J = 6.9 Hz, 2H, CH₂–CH₂S), 1.28 (t, J = 7.2 Hz, 3H, C₍₂₎H₃–CH₂O), 1.16 (bro, J = 7.5 Hz, 10H, CH aliphatic), 1.04 (t, J = 7.2 Hz, 3H, C_(2*)H₃-CH₂O), 0.81 (t, J = 6.9 Hz, 3H, CH₃–RS).

¹³C NMR (CDCl₃, 75 MHz): δ (ppm): 13.9 (CH3–RS), 16.1 (C_(2*)H3 ethoxy), 16.3 (C₍₂₎H3 ethoxy), 22.5 & 22.6 & 29.1 & 29.2 & 31.7 & 36.1 (CH₂ alkyl), 48.1(C_(1*)H₂ ethoxy), 49.7 (C₍₁₎H₂ ethoxy), 55.0 (P–CH–N), 62.8 (S–CH₂), 63.1 (CH₃ methoxy), 113.7 (C_3,5 phenyl), 127.3 (C_2,6 phenyl), 129.4 (C₁ phenyl), 159.2 (C₄ phenyl).

Diethyl(dodecylsulfonamido(4-methoxyphenyl)methyl)phosphonate (3c)

Molecular weight: 505 (g/mol), FT-IR (KBr, cm⁻¹): 3444 (stretch NH), 2924 (stretch CH, aromatic), 2854 (stretch CH, aliphatic), 1651 (stretch C=C, aromatic), 1512 (bending NH), 1458 (bending CH₂), 1396 (bending CH₃), 1311, 1176 (stretch O=S=O), 1211(stretch C–O), 1041 (stretch C–O, OCH₃), 948 (stretch CH, oop), 729 (rocking CH₂), 547 (bending NH).

¹H NMR (CDCl₃, 300 MHz): δ (ppm): 7.55–7.53 (dd, J = 2.1, 2.1 Hz, 1H, NH), 7.38–7.36 (dd, J = 1.2 Hz, 2H, C_(2,6)H aromatic), 6.81–6.79 (d, J = 5.1 Hz, 2H, C_(3,5)H aromatic), 5.51–5.45 (dd, J = 5.7, 5.7 Hz, H, CH–P), 4.13–4.07 (m, 2H, C_(1,1*)H–OP), 3.85–3.82 (m, 1H, C₍₁₎H–OP), 3.73 (s, 3H, CH₃–O), 3.66–3.61 (m, 1H, C_(1*)H–OP), 2.17 (t, J = 4.2 Hz, 2H, CH₂–S), 1.54 (t, J = 6.9 Hz, 2H, CH₂–CH₂S), 1.28 (t, J = 4.2 Hz, 3H, C₍₂₎H₃–CH₂O), 1.16 (bro, J = 7.5 Hz, 10H, CH aliphatic), 1.04 (t, J = 4.2 Hz, 3H, C_(2*)H₃–CH₂O), 0.83 (t, J = 6.9 Hz, 3H, CH₃–RS).

¹³C NMR (CDCl₃, 75 MHz): δ (ppm): 13.9 (CH3–RS), 16.1 (C_(2*)H3 ethoxy), 16.3 (C₍₂₎H3 ethoxy), 22.5 & 22.6 & 29.1 & 29.2 & 29.4 & 31.7 & 36.2 (CH₂ alkyl), 48.4(C_(1*)H₂ ethoxy), 49.6 (C₍₁₎H₂ ethoxy), 55.1 (P–CH–N), 62.8 (S–CH₂), 63.1 (CH₃ methoxy), 113.8 (C_3,5 phenyl), 127.3 (C_2,6 phenyl), 129.4 (C₁ phenyl), 159.2 (C₄ phenyl).

Diethyl(hexadecylsulfonamido(4-methoxyphenyl)methyl)phosphonate (3d)

Molecular weight: 561 (g/mol), FT-IR (KBr, cm⁻¹): 3433 (stretch NH), 2920 (stretch CH, aromatic), 2850 (stretch CH, aliphatic), 1608 (stretch C=C, aromatic), 1512 (bending NH), 1465 (bending CH₂), 1400 (bending CH₃), 1307, 1184 (stretch O=S=O), 1249 (stretch C–O), 1053 (stretch C–O, OCH₃), 968 (stretch CH, oop), 721 (rocking CH₂), 536 (bending NH).

¹H NMR (CDCl₃, 300 MHz): δ (ppm): 7.36–7.33 (dd, J = 1.5, 1.8 Hz, 2H, C_(2,6)H aromatic), 6.85–6.83 (d, J = 6.3 Hz, 2H, C_(3,5)H aromatic), 6.60–6.56 (dd, J = 3.3, 3.3 Hz, 1H, NH), 5.47–5.40 (dd, J = 7.2, 7.2 Hz, H, CH–P),4.13–4.07 (m, 2H, C_(1,1*)H–OP), 3.92–3.86 (m, 1H, C₍₁₎H–OP), 3.70 (s, 3H, CH₃–O), 3.68–3.64 (m, 1H, C_(1*)H–OP), 2.19 (t, J = 6 Hz, 2H, CH₂–S), 1.58 (t, J = 5.7 Hz, 2H, CH₂–CH₂S), 1.29 (t, J = 5.1 Hz, 3H, C₍₂₎H₃–CH₂O), 1.23 (bro, J = 7.5 Hz, 26H, CH aliphatic), 1.07 (m, J = 5.1 Hz, 3H, C_(2*)H₃–CH₂O), 0.85 (t, J = 5.4 Hz, 3H, CH₃–RS).

(Hexylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid (4a)

Molecular weight: 365 (g/mol), FT-IR (KBr, cm⁻¹): 3433 (stretch NH), 2931 (stretch CH, aromatic), 2866 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1516 (bending NH), 1462 (bending CH₂), 1392 (bending CH₃), 1303, 1188 (stretch O=S=O), 1045 (stretch C–O, OCH₃), 952 (stretch CH, oop), 775 (rocking CH₂), 559 (bending NH).

¹H NMR (CH₃OH-d₄, 300 MHz): δ (ppm): 7.38–7.35 (d, J = 8.4 Hz, 2H, C_(2,6)H aromatic), 6.90–6.86 (d, J = 4.2 Hz, 2H, C_(3,5)H aromatic), 5.46–5.29 (dd, J = 21, 21 Hz, H, CH-P), 3.77 (s, 3H, CH₃–O), 2.28 (t, J = 7.2 Hz, 2H, CH₂–S), 1.59 (t, J = 5.7 Hz, 2H, CH₂–CH₂S), 1.25 (bro, 6H, CH aliphatic), 0.85 (t, J = 5.4 Hz, 3H, CH₃–RS).

MS (m/z, %): 365 (M), 364 (M-1, 4), 336 (M-29, 1), 310(M-57, 2), 284 (M-81, 9), 216 (M-148, 2), 163 (M-201, 18), 148 (M-216, 11), 136 (M-229, 26), 106 (M-259, 9), 91 (M-274, 13), 83 (M-282, 49), 72 (M-293, 78), 56 (M-311, 76), 43 (M-292, 97).

((4-Methoxyphenyl)(octylsulfonamido)methyl)phosphonic acid (4b)

Molecular weight: 393 (g/mol), FT-IR (KBr, cm⁻¹): 3606 (stretch NH), 2927 (stretch CH, aromatic), 2858 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1516 (bending NH), 1462 (bending CH₂), 1396 (bending CH₃), 1307, 1184 (stretch O=S=O), 1041 (stretch C–O, OCH₃), 960 (stretch CH, oop), 729 (rocking CH₂), 547 (bending NH).

¹H NMR (CH₃OH-d₄, 300 MHz): δ (ppm): 7.38–7.36 (dd, J = 1.2 Hz, 2H, C_(2,6)H aromatic), 6.9–6.87 (dd, J = 6.6 Hz, 2H, C_(3,5)H aromatic), 5.45–5.30 (dd, J = 15.9, 5.9 Hz, H, CH–P), 3.76 (s, 3H, CH₃–O), 2.29–2.26 (m, 2H, CH₂–S), 1.59 (t, J = 5.1 Hz, 2H, CH₂–CH₂S), 1.25 (bro, 10H, CH aliphatic), 0.88 (t, J = 2.4 Hz, 3H, CH₃–RS).

MS (m/z, %): 393 (M), 321 (M-71, 12), 313 (M-80, 39), 299 (M-94, 17), 204 (M-189, 5), 176 (M-217, 19), 164 (M-229, 18), 148 (M-245, 25), 135 (M-258, 100).

(Dodecylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid (4c)

Molecular weight: 449 (g/mol), FT-IR (KBr, cm⁻¹): 3410 (stretch NH), 2920 (stretch CH, aromatic), 2850 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1512 (bending NH), 1462 (bending CH₂), 1392 (bending CH₃), 1307, 1184 (stretch O=S=O), 1053 (stretch C–O, OCH₃), 952 (stretch CH, oop), 725 (rocking CH₂), 559 (bending NH).

¹H NMR (CH₃OH-d₄, 300 MHz): δ (ppm): 7.37–7.36 (dd, J = 4.8 Hz, 2H, C_(2,6)H aromatic), 6.88 (t, J = 5.1 Hz, 2H, C_(3,5)H aromatic), 5.45–5.31 (dd, J = 12.51, 12.57 Hz, H, CH–P), 3.76 (s, 3H, CH₃–O), 2.27 (t, J = 1.8 Hz, 2H, CH₂–S), 1.58 (t, J = 5.7 Hz, 2H, CH₂–CH₂S), 1.25 (bro, 18H, CH aliphatic), 0.88 (t, J = 5.4 Hz, 3H, CH₃–RS).

MS (m/z, %): 450 (M + 1), 368(M − 81, 25), 281 (M − 169, 12), 233 (M − 217, 17), 218 (M − 232, 13), 163 (M − 286, 33), 148 (M-301, 40), 135 (M − 314, 100).

(Hexadecylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid (4d)

Molecular weight: 505 (g/mol), FT-IR (KBr, cm⁻¹): 3603 (stretch OH), 3356 (stretch NH), 2916 (stretch CH, aromatic), 2850 (stretch CH, aliphatic), 1612 (stretch C=C, aromatic), 1512 (bending NH), 1465 (bending CH₂), 1392 (bending CH₃), 1311, 1180 (stretch O=S=O), 1037 (stretch C–O, OCH₃), 952 (stretch CH, oop), 721 (rocking CH₂), 555 (bending NH).

¹H NMR (CH₃OH-d₄, 300 MHz): δ (ppm): 7.36–7.34 (d, J = 4.8 Hz, 2H, C_(2,6)H aromatic), 6.86–6.84 (d, J = 4.8 Hz, 2H, C_(3,5)H aromatic), 3.74 (s, 3H, CH₃–O), 2.27 (s, 2H, CH₂–S), 1.57 (s, 2H, CH₂–CH₂S), 1.25 (s, Hz, 26H, CH aliphatic), 0.86 (t, J = 5.4 Hz, 3H, CH₃–RS).

MS (m/z, %): 505 (M), 425 (M − 80, 27), 338 (M − 167, 2), 316 (M − 189, 12), 302 (M − 203, 17), 289 (M − 217, 9), 282 (M − 223, 6), 224 (M − 281, 77), 212 (M − 293, 8), 204 (M − 301, 25), 190 (M − 315, 100).

Enzyme preparation and purification

PAP from red kidney bean (rkbPAP) was purified following a previously published protocol⁸. Briefly, red kidney beans (Phaseolus vulgaris), provided from local market, were ground and suspended in 0.5 M sodium chloride solution. The suspension was filtered through a muslin cloth, followed by ethanol fractionation and ammonium sulfate precipitation. The purple enzyme solution was dialyzed against several changes of 0.25 M NaCl over a period of several days. The purple precipitate, which formed in the dialysis bag was collected by centrifugation. The resulting preparation stored at 4 °C in 0.5 M NaCl. The protein identity and purity was tested by SDS–polyacrylamide gel electrophoresis (SDS–PAGE) and by comparing with molecular weight protein markers, by loading the protein samples on a 12% slab gel. Protein concentration was determined by Lowry method²⁸.

Enzyme assay

Inhibition assays for rkbPAP was performed in 96-well multi-titer plates using a microplate reader. Kinetic measurements were taken at pH 4.9 (0.1 M sodium citrate buffer with 5% DMSO) at 25 °C using p-nitrophenyl phosphate (p-NPP) as the substrate at concentrations of 1, 2, 4, 6, 8, 10, 12, 14, 18 and 24 mM²⁹. The hydrolysis of p-NPP by PAP produces p-nitrophenol (p-NP), which has a characteristic intense yellow color at 405 nm with ɛ = 343 M⁻¹ cm^−1 18,26. The concentration of enzyme used was ∼60 nM, while at least eight concentrations of each compound ranged from 50 to 2000 µM. The IC₅₀ values obtained by “Dose–response inhibition/variable slopes” equation that was performed using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego CA, www.graphpad.com). The experimental data are representative example of three independent experiments. In most cases, the standard deviations were almost within 5% of the experimental values, except for some inhibition data points where standard errors of means were within 10% of the measured values. p < .05 was considered statistically significant.

Determination of the inhibition mode and constants

To determination of inhibition mode, the Michaelis–Menten constant (K_m) and maximum velocity (V_max) and inhibition constants (K_i and K_I) values were obtained using Equation (1)(1) ^29,30 in the presence of at least four concentrations of each compound which were selected from concentration–activity curves. The data were analyzed by nonlinear regression using the program GraphPad Prism (1) where K_i and K_I are the equilibrium dissociation constants for competitive and uncompetitive inhibitor binding, respectively, while, v, [S] and [I] represent the rate of product formation, the substrate and the inhibitor concentration, respectively²⁹. It has been found that there are two inhibitor constants: K_i, which defined as the dissociation constant of the enzyme–inhibitor complex, and K_I, which defined as the dissociation constant of the enzyme–substrate–inhibitor complex. If K_i<K_I, the type of enzyme inhibition was defined as competitive-noncompetitive and if K_i>K_I, the type of enzyme inhibition was defined as uncompetitive–noncompetitive. The type of enzyme inhibition was defined as pure noncompetitive inhibition when K_i=K_I^30–32. For better visual presentation of data of inhibition mode, the Lineweaver–Burk plots were also provided.

Docking studies

The structures of molecules were drawn and optimized using HyperChem (HyperCube Inc., Gainesville, FL). Semi-empirical AM1 method with Polak–Ribiere algorithm until the root mean square gradient of 0.01 kcal mol⁻¹ was used as the optimization method. The crystal structure of rkbPAP retrieved from the Protein Data Bank (PDB code: 2QFR) was used. Then, after determining Kollman united atom charges³³ and merging nonpolar hydrogens, rotatable bonds were assigned. Using autoGrid tool, the grid maps (one for each atom type in the ligand, and one for electrostatic interactions) were constructed adequately large to include the active site of enzyme as well as significant regions of the surrounding surface. In all the cases, a grid map of 60 points in each Cartesian direction besides a grid-point spacing of 0.375 Å (a quarter of the carbon–carbon single bond) were generated. By the ligand location in the complex of ligand–enzyme, the maps were centered on the binding site (Zn atom of the enzyme), searching effective interactions with the functional groups. Based on Lamarckian Genetic Algorithm³⁴, using the pseudo-Solis and Wets local search method³⁵, AutoDock Tools was employed to produce both grid and docking parameter files, that is .gpf and .dpf files. Applying 0.5 Å clustering tolerance to construct clusters of the closest compounds, the initial coordinates of the ligand were used as the reference structure. Finally, docking results (enzyme–inhibitor complex) were visualized using Viewer Lite 4.2.

Results and discussion

Chemistry

The conversion of 4-methoxybenzaldehyde (1) into diethyl (amino(4-metoxybenzyl)-methyl)phosphonate was achieved by heating 1 with a mixture of ammonium acetate and diethyl phosphite over activated molecular sieves in ethanol, according to a literature procedure reported for the synthesis of diethyl α-aminobenzylphosphonate from benzaldehyde (Scheme 1)³⁶. Introducing hydrogen chloride gas to the free amine product gave its corresponding hydrochloride salt 2 in 23% overall yield³⁶. Alkylation of 2 to give the sulfonamides 3a–d was achieved with adding appropriate sulfonyl chlorides and DIPEA to 2 in chloroform. The reaction mixture was stirred under nitrogen atmosphere for 24 h. The final synthetic step to produce the newly designed inhibitors 4a–d required the cleavage of the phosphonate ester bonds of 3a–d to give their corresponding free phosphonic acids. This was achieved by stirring 3a–d with a mixture of potassium iodide and trimethylsilyl chloride in chloroform at 25 °C for 24 h that giving the desired products in moderate yields.

Enzyme purification and biological activity measurements

Although human PAP (hPAP) would be the best choice for our purposes, hPAP can only be obtained in minute quantities using a baculoviral recombinant expression system³⁷. However, as stated earlier, rkbPAP is a valid model of hPAP (see Figure 1), as indicated by the highly conserved active sites across these different enzymes, the similarities in substrate binding pockets, and the similarities of inhibition constants for a range of structurally different inhibitors reported for several animal (including human) and plant PAPs^18,26. The rkbPAP was purified from red kidney bean according to the method described by Beck et al.⁸ The SDS–PAGE analysis is shown in Figure 1, which clearly indicates the reasonable purity of the isolated PAP. The densitometry of bands indicates that the mass fraction of the purified enzyme is around 0.90 (∼90% purity). It is noteworthy that the plant enzymes are mostly dimeric glycoproteins, and have subunit molecular masses of 50–60 kD. As shown in the figure, differential glycosylation of the subunits, leading to a heterogeneity in the molecular mass of the subunits, has been reported for kidney bean PAP, which is a dimeric glycoprotein (molecular mass 110 kD) with an inter-subunit disulfide linkage¹¹. Activity staining on PAGE gel also showed that only the ∼50 kDa bands had esterase activity, confirming that there is no contaminant phosphatase enzymes other than desired PAP (data not shown).

The isolated rkbPAP was assayed using p-NPP as a substrate. The obtained K_m and V_max by fitting data to the Michaelis–Menten equation were 12.6 ± 1.4 and 48 ± 6, respectively. We also designed/synthesized and evaluated the catalytic effects of a number of new inhibitor candidates of rkbPAP, based on the alkylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid scaffold. The inhibitory effects of derivatives 3a–d and 4a–d as well as tartrate as the positive control were tested against rkbPAP at pH 4.9³⁸. In order to identifying the PAP inhibition mode of test compounds and determining apparent kinetic parameters, the kinetic data were analyzed using Equation (1)(1) . The IC₅₀ and inhibition constant (K_i and K_I) values of test compounds have been summarized in Table 1. As it was expected, PAP activity was weakly inhibited by tartrate, which is in good agreement with previous reports^11,38–40. Most of the synthesized compounds demonstrated their inhibitory potentials on PAP in a dose-dependent manner. Among the evaluated compounds, 3d (IC₅₀=92 µM, K_i= 13 ± 1 µM), 4c (IC₅₀=421 µM, K_i= 95 ± 9 µM) and 4d (IC₅₀=13 µM, K_i= 1.1 ± 0.2 µM), were the most potent inhibitors; of which 3d and 4d were recognized as potent inhibitors of PAP even stronger than the standard inhibitor^11,38–40. As can be seen, the K_i values are in reasonable agreement with the IC₅₀ values, comparable at least with the best of the naphthalenemethylphosphonic acid derivatives^3,18. Also, the 3a, 3b, 4a and 4b exhibited negligible inhibitions of PAP, in millimolar range. Thus, these compounds were excluded from further studies (determination of inhibition mode).

Table 1. Kinetic data for inhibitors against rkbPAP at pH 4.9.

Inhibitor name	IC50 (μM)	Ki (μM)	KI (μM)
3a and 3b	Very weak inhibitor	–	–
3c	759	202 ± 11	211 ± 8
3d	92	13 ± 1	36 ± 3
4a and 4b	Very weak inhibitor	–	–
4c	421	95 ± 9	103 ± 5
4d	13	1.1 ± 0.2	8.2 ± 1.4
Tartrate	3315	783 ± 41	–

Although an in depth SAR analysis is currently impossible, it appears that a major improve in rkbPAP inhibitory activity correlates with increasing the asymmetric hydrophobicity of the compounds. It appears that inhibitor potency increases with increase the length of alkyl chain in alkylsulfonamide moieties (Table 1 and Scheme 1), so that the largest effect on inhibition was achieved with the hexadecyl moiety. Substitution at the dodecyl (in 3c/4c) with hexadecyl (in 3d/4d) greatly decrease the inhibition constant of the inhibitor, suggesting an extremely large hydrophobic or sterically tolerant region near/not into the phosphate binding site of the enzyme as well as a possible preference for the long-chain acyl (≥ C₁₂) derivatives. McGeary et al. was reported a similar trend for the rkbPAP and pig PAP enzymes, that is, the longer the alkyl chain the more potent inhibitor of PAPs¹⁸.

The inhibition modes of the compounds can be elicited by comparing values of K_i and K_I and represented by Lineweaver–Burke plot. Typical Lineweaver–Burk plots for inhibitory activity of 3c and 4d as representative have been demonstrated in Figure 2. If the Lineweaver–Burke plot shows crisscross lines at a single point on the 1/V (y) axis, it is indicative of competitive inhibition. Also, the trend lines of Lineweaver–Burke plot for 3c and 4c compounds intersect each other on the 1/S (x) axis, which is characteristic of noncompetitive inhibition, so K_i≈K_I. Such noncompetitive mode inhibition has also been previously reported for hexyl-derivative of α-alkoxynaphthylmethylphosphonic acid¹⁸. For longer chain-length inhibitors (3d and 4d), the lines of the Lineweaver–Burke plot intersect at a single point between x- and y-axis, close to the 1/S axis, indicate that PAP activity was inhibited with mixed manner by these compounds. On the other hand, since K_i<K_I, the exact mechanism of inhibition is competitive–noncompetitive^18,23,26,31. In agreement with this mode change, McGeary et al. reported that longer alkyl chains of α-alkoxynaphthylmethylphosphonic acid derivatives inhibit rkbPAP and pPAP with mixed (competitive–noncompetitive) manner¹⁸. This behavior may reflect a stronger anchoring effect of the longer alkyl chains into the groove adjacent to the active site of the enzyme, which would favor partially competitive inhibition. Furthermore, the replacement of the diethyl phosphonate group of series 3 by phosphonate in series 4 has a minor decrement effect on the inhibitory effect of compound and does not alter the mode of inhibition, since probably this moiety is not bind/coordinate to bimetal/binuclear center.

Figure 2. Typical Lineweaver–Burk plots for inhibitory activity of synthetic compounds against rkbPAP. The data represent the average of 3–5 experiments. (A) Lineweaver–Burk plot of rkbPAP activity in the absence (•) and the presence of 300 (^), 600 (▴) and 1200 μM of 3c (▵). (B) Lineweaver–Burk plot of rkbPAP activity in the absence (•) and the presence of 10 (^), 20 (▴) and 40 μM of 4d (▵).

Molecular docking studies

Molecular docking studies on binding modes are essential to elucidate key structural characteristics and interactions and they provide helpful data for designing effective PAP inhibitors⁴¹. Hence, to make the rational design of novel and more selective PAP inhibitors possible, molecular docking was carried out on PAP binding pocket using a set of PAP inhibitors shown in Scheme 1. As well as RMSD cluster analysis, AutoDock also uses binding free energy assessment to assign the best binding conformation. Energies estimated by AutoDock are described by intermolecular energy (including van der Waals, hydrogen bonding, desolvation, and electrostatic energies), internal energy, and torsional free energy⁴². Among these calculated energies by AutoDock, the first two provide the docking energy, while the sum of the first and the third items account for the binding energy. Among all interactions occurring in the active site, the electrostatic interaction between the ligand and the enzyme is the most significant, because in most cases it can assign the strength of binding and the exact position of the inhibitor in the binding site energy⁴².

The docking results show that all of the studied compounds occupy an almost similar space in the binding site. Also, the calculated binding affinities using computational modeling correlate well with measured inhibition constants (results not shown). Unexpectedly, modeling suggests that the phosphonate moiety of the 3d inhibitor does not bind to the dimetal center in the active site of rkbPAP. Furthermore, the alkyl chain of the inhibitor binds to the groove on the surface of the enzyme. Other part of molecules including polar atoms and aromatic portions are placed on the hydrophilic patch around the active site, which is in good agreement with the obtained inhibition modes. The best possible binding mode of 3d as the most biologically active compound in the rkbPAP binding site is illustrated in a three-dimensional (Figure 3) and two-dimensional space (Figure 4).

Figure 3. The three-dimensional representation of docking result of 3d in the close vicinity of rkbPAP binding pocket (left). Surface and stick representation of the predicted binding mode of 3d to the enzyme based on docking simulations (right).

Figure 4. The two-dimensional representation of docking result of 3d compound with binding pocket of rkbPAP.

On the other hand, to obtain a detailed description of the effects of individual residues on binding affinity, a per-residue decomposition of the total energy was carried out to evaluate the energetic influences of critical residues on the binding. The residues were separately studied, and the Van der Waals, electrostatic and interaction energies of these residues and 3d were determined. These include Phe-206, Glu-299, Gly-200, Asp-169, Tyr-256, Ser-287, Met-285 and Leu-289 (Table 2). Energy of interaction between individual residues and ligand was calculated with the “Calculate Interaction Energy” protocol encoded in Discovery Studio. The complexes were typed with CHARMM force field. All parameters used were kept at their default settings. The residue-based decomposition of interaction energies identified several critical residues of PAP. Table 2 lists the calculated energies contributions of these key residues of interest. As can be seen, major favorable energy contributions originate predominantly from Phe-206 and Glu-299. The methoxy phenyl core of 3d is anchored by an aromatic stacking interaction with Phe-206, whereas Glu-299 is one of major contributor of van der Waals interactions.

Table 2. Interaction, van der Waals and electrostatic energies for selected residues of PAP and 3d.

Residue	Interactionenergy (kcal/mol)	VDW interactionenergy (kcal/mol)	Electrostatic interactionenergy (kcal/mol)
PHE206	−12.1	−3.78	−8.30
GLU299	−11.3	−3.86	−7.47
GLY200	−7.45	−0.18	−7.27
ASP169	−7.32	0.02	−7.35
TYR256	−3.67	−1.66	−2.01
SER287	−2.60	15.2	−17.8
MET285	−1.22	−0.62	−0.60
LEU289	−0.27	−0.27	0.0

As it can be concluded from Figures 3, 4 and Table 2, the major part of the binding cavity of rkbPAP is hydrophilic, which allows for the tight binding of polar substrates and inhibitors. A long apolar groove adjacent to the active site was identified as an additional potential point of interaction of hydrophobic aliphatic chain and is required for recognition and directionality of the substrate aliphatic chain functionality¹⁸. Also, hydrogens of NH₂ group of Asn-201 side chain and sulfonamide group of ligand, also hydrogens of NH₂ group of Phe-206 backbone and oxygen of phosphonate group of ligand form hydrogen bonds (3.12 and 2.90 Å, respectively). It is was interesting to note that the aliphatic chain orients inside the binding cavity and interacts by both hydrophobic as well as hydrophilic interactions (Table 2) with amino acids such as Leu-289 and Met-303. With respect to the docking results (Figures 3 and 4) Ser-252, Tyr-256, Ser-287, Met-303 and Leu-289 assemble one hydrophobic cluster, which accommodates the aliphatic hydrophobic chain moiety of the ligand. Taking literature²⁶ into account, similar binding orientations of inhibitors in the active sites of pig, red kidney bean and hPAP, may support the contention that the inhibitory potencies obtained using rkbPAP can be extrapolated to hPAP.

Conclusions

To the best of our knowledge, there are no published reports in the literature on the application of alkylsulfonamidophenyl)methyl)phosphonic acid derivatives as potential PAP inhibitors. In the current work, we identified these new derivatives of alkylsulfonamido(4-methoxyphenyl)methyl)phosphonic acid, as potent inhibitors of rkbPAP. The hexadecyl derivative (4d, with K_i value of 1.1 µM) was found to be the most active compared to compounds contain shorter alkyl chains. In addition, it was illustrated that the inhibitor bound in the vicinity of the enzyme active site showed mixed mode inhibition due to partial competition whit substrate. Considering the crucial role(s) of mammalian PAPs in bone metabolism, such inhibitors are anticipated to be promising lead compounds in the development of new anti-osteoporotic chemotherapeutics.

Acknowledgement

This work was performed in partial fulfillment of the requirements for Pharm. D. of N. Alimoradi, in Faculty of Pharmacy, Kermanshah University of Medical Sciences, Kermanshah, Iran.

Disclosure statement

The authors declare that they have no conflict of interests.

Funding

The financial support [Grant No. 92047] from the Research Council of Kermanshah University of Medical Sciences is gratefully acknowledged.