Human neutrophil elastase inhibitory dihydrobenzoxanthones and alkylated flavones from the Artocarpus elasticus root barks

Abstract

Neutrophil elastases are deposited in azurophilic granules interspace of neutrophils and tightly associated with inflammatory ailments. The root barks of Artocarpus elasticus had a strong inhibitory potential against human neutrophil elastase (HNE). The responsible components for HNE inhibition were confirmed as alkylated flavones (2–4, IC₅₀ = 14.8 ~ 18.1 μM) and dihydrobenzoxanthones (5–8, IC₅₀ = 9.8 ~ 28.7 μM). Alkyl groups on flavone were found to be crucial functionalities for HNE inhibition. For instance, alkylated flavone 2 (IC₅₀ = 14.8 μM) was 20-fold potent than mother compound norartocarpetin (1, IC₅₀ > 300 μM). The kinetic analysis showed that alkylated flavones (2–4) were noncompetitive inhibition, while dihydrobenzoxanthones (5–8) were a mixed type I (K_I < K_IS) inhibitors, which usually binds with free enzyme better than to complex of enzyme–substrate. Inhibitors and HNE enzyme binding affinities were examined by fluorescence quenching effect. In the result, the binding affinity constants (K_SV) had a significant correlation with inhibitory potencies (IC₅₀).

Introduction

Serine proteases are the largest and most important unit among the enzyme groups found in eukaryotes and prokaryotes, which distinctively increases the reaction of the peptide bonds hydrolysis [1]. One of them the neutrophil elastase (NE, EC 3.4.21.37) 29 kDa chymotripsis family protease, largely expressed by neutrophils precursors inside the bone marrow, then in its mature active form they placed in primary (azurophilic) granules [2]. Throughout inflammation spread into the cell interspace NE activity closely contained by an abundant blood plasma inhibitor—α1-antytrypsin (AAT) [3]. The stoppage of releasing AAT in hepatocytes expressively decreases its levels, leading to emphysema because of deficient defence of the lower respiratory tract from human neutrophil elastase (HNE), and further alveoli damage [4]. This leads to the development of the chronic obstructive pulmonary disease, emphysema and lung cancer, also in the liver, it causes benign neonatal hepatitis syndrome, fibrosis which evolves to cirrhosis and even to hepatocellular carcinoma [5]. Augmentation therapy has been accepted as the best therapy for AAT deficiency, and less costly methods such as low molecular weight NE inhibitors have failed clinically, despite diligent efforts over the past three decades [6]. On the other hand, the release of granular contents leads to the recruitment of inflammatory cells through cytokines (TNF-α, IL-1β, IL-6, IL-18, IL-10), adipokines (adiponectin, resistin and leptin), and chemokines (IL-8), which are the main intermediaries and progressors of renal tissue damage [7]. Thus, the invention of novel protease inhibitors is an invaluable therapeutic tool. There have been many attempts to develop lead structures for inhibiting HNE from plant sources [8, 9].

Artocarpus elasticus, which common name is Terap, from the Moraceae family, distributed in the Asian tropics. The plant is widely used in food: peeled shoot tips and fleshy perianth can even be eaten raw, and the seeds are eaten fried. The ripe fruits are edible, but they often have an unpleasant smell. Leaves were used for nursing mothers, young leaves to treat vomiting, diseases related to blood, the bark inner side was used to treat sores, and latex also used in dysentery, fever, hypertension and diabetes [10, 11]. The therapeutic effect is associated with the abundance of phenolics, such as flavonoids, stilbenoids, dihydrochalcones, dihydrobenzoxanthones and prenylated flavonoids [12]. Single metabolites reported in vivo antimalarial [13], inhibition of LPS-induced inflammation [14], α-glucosidase inhibitory [15], TRAIL-resistance overcoming with antiviral [16], and anticancer [17] activities. Alkylated flavonoid Artonin E from A. elasticus also revealed considerable cytotoxic effect in MCF-7 human breast cancer cells by ROS mediated mitochondrial pathway [18, 19]. In particular, metabolites in A. elasticus has not been reported to have HNE inhibitory potential capacity.

The aim of the present study was the investigation of HNE inhibitory potential of the root barks of A. elasticus, and to isolate the responsible components for HNE inhibition. The inhibitory mechanisms were fully characterized by double reciprocal plots from the Michaelis–Menten equation. We also tried to determine the binding affinities of the isolated inhibitors to HNE enzyme by fluorescence (FS) quenching experiment.

Materials and methods

Instruments and chemicals

Bruker AM500 spectrometer (Karlsruhe, Germany) used for measurement of proton and carbon-13 NMR spectra. By mass spectrometer JEOL JMS-700 (Tokyo, Japan) obtained all mass data. Forte/R 100 (YMC, Kyoto, Japan) recycling HPLC and MPLC with Triart C18 column (YMC, Japan) used for isolation of compounds. Methanol, acetonitrile, water and acetic acid of the analytical grade for HPLC were purchased from Fisher Scientific (Korea). For performing enzyme assays SpectraMax M3 Multi-Mode Microplate Reader (Molecular device, USA) was used. HNE (EC 3.4.21.37) was ordered from Sigma Aldrich (St. Louis, USA).

Plant material

Collection in December 2013 and storage until extraction of Malaysian the A. elasticus barks done by associated professor Dr. Mohd Azlan Nafiah. The voucher of the specimen (TM1016) was given in the Universiti Pendidikan Sultan Idris, Malaysia.

Extraction and isolation

Extraction of A. elasticus barks (250 g) was done in methanol (10L), at approximately 25 °C, resulting to crude reddish-brown colour gum (27 g). The methanol extract was dried well and dissolved in 500 mL of water to be suspended for further fractionation into dark-red chloroform extract (dry weight 14 g). Then chloroform extract was chromatographed by MCI GEL CHP20P (30 × 5 cm, 75–150 μm, 500 g) column with step by step changing of water/methanol system (8:2 → 0:1) resulting to A1-15 fractions. Nonpolar A8-12 fractions (total 4.6 g) were refractionated by MPLC (using ODS column 25 × 3 cm, S-10 μm, 12 nm, YMC) with a slow changing of solvents from water to MeOH (0 → 100%) to give B1-80 subfractions. The subfractions B26-35, totally 1.8 g, were further subjected to recycling HPLC (using ODS column 25 × 3 cm, S-5 μm, 12 nm, YMC), which resulted to the isolation of compounds 1 (38 mg), 3 (21 mg), and 8 (35 mg). Likewise, the subfractions B36-43 total 1.4 g, were put to recycling HPLC to isolate dihydrobenzoxanthone 5 (14 mg), 6 (22 mg), and alkylated flavone 4 (19 mg). In the same way, the subfractions B44-56 (1.5 g) were also recycled on HPLC to purify compounds 2 (12 mg) and 7 (18 mg). Purification of compounds done on Sephadex-LH20.

Measurement of HNE inhibitory activity

The inhibitory activity of HNE (EC 3.4.21.37) was analysed by measuring of the hydrolysis of N-methoxysuccinyl-Ala–Ala-Pro-Cal-p-nitroanilide at 405 nm in optimal pH of 8.0 (0.02 mM, Tris–HCl buffer) and at 37 °C [8]. In brief, to 96-well plates 10 μL the inhibitors or caffeic acid as a control (dissolved in DMSO), 40 μL substrate (MeOSuc-AAPV-pNA, 1.5 mM), 130 μL of buffer and 20 μL enzyme (stock concentration 0.2 unit/mL) mixed. The concentration of 50% inhibition of the enzyme (IC₅₀) was expressed as compounds activity. Inhibition rate (%) calculated by next Eq. (1):

$$ {\text{Activity }}\left( \% \right) \, = {\text{ [1 + ([I]/IC}}_{50} )] \, \times { 1}00 $$

(1)

HNE inhibitory kinetic assay

By the same way with activity experiments, the enzyme kinetic behaviours were identified using MeOSuc-AAPV-pNA substrate concentrations of 0, 0.75, 1.5, and 3 mM with different concentrations of inhibitors [8]. Data analysis on Sigma Plot (Chicago, USA) used to determine the variables of curves. By Lineweaver–Burk plots, kinetic values such as Michaelis–Menten (K_m) and maximum velocity (V_max) were determined. From Dixon plots, enzyme and inhibitors (K_i) dissociation constants calculated. Constants of inhibitions when compounds binding whether free enzyme or else enzyme–substrate complex, were calculated from plots of the slopes of the straight lines (K_I) or vertical intercept (K_IS) verse the concentration of inhibitors by Eqs. (2)–(4).

$$ 1 /V{ = }K_{\text{m}} /V_{\text{max} } ( 1 {\text{ + [I]/}}K_{\text{I}} ) {\text{ 1/S + 1/}}V_{ \text{max} } $$

(2)

$$ {\text{Slope = }}K_{\text{m}} /K_{\text{I}} V_{ \text{max} } [ {\text{I] + }}K_{\text{m}} /V_{ \text{max} } $$

(3)

$$ {\text{Intercept = 1/}}K_{\text{IS}} V_{ \text{max} } [ {\text{I] + }}K_{\text{m}} /V_{ \text{max} } . $$

(4)

Measurement of binding affinity to the enzyme

In the 96-well black immuno-plates the reaction mixture of 180 µl of 0.02 mM Tris–HCl buffer (pH 8.0) with 10 µl of HNE (0.2 unit/mL) were added into the 10 µl inhibitors with concentrations of 0, 3.125, 6.25, 12.5, 25 and 50 µM [8]. FS quenching spectra were obtained using a spectrophotometer (SpectraMax M3) at the emission of from 300 to 400 nm (excitation 295 nm) with slits of 2.0 nm.

Statistical analysis

Experiments were repeated thrice, then analyzed on SigmaPlot (v. 10.0), p < 0.05 were determined as a noteworthy alteration.

Results and discussions

Isolation of HNE inhibitors from A. elasticus root barks

Artocarpus elasticus has been proven to possess anti-inflammatory activities that might be associated with HNE enzyme. In the primary experiment, the root barks methanol extract of target plant presented a significant potential (50 µg/mL, 80% inhibition) against HNE. This encourages us to examine HNE inhibitory metabolites from the root part of this species. The methanol extracts of the root barks were fractionated by solvents with different polarities (hexane, chloroform, ethyl acetate, water) for further isolation based on HNE inhibitory potencies. The chloroform fractions were purified by chromatography over octadecyl-functionalized silicagel and Sephadex LH-20 to yield eight compounds. As shown in Fig. 1, compounds were identified as norartocarpetin (1), artoflavone B (2), KB 2 (3), artonin E (4), artobiloxanthone (5), artoindonesianin W (6), cycloartobiloxanthone (7), and artoindonesianin P (8) by obtained spectra (see Additional file 1), with the comparison to the published data [20, 21].

Fig. 1

The chemical structures of compounds 1–8 isolated from the root barks of A. elasticus

Among the isolated compounds we elucidated the most active inhibitor dihydrobenzoxanthone 5 chemical structure. The compound 5 has a molecular formula C₂₅H₂₂O₇ as established by the [M]⁺ ion at 434.1363 (Calcd for 434.1366) in HREIMS analysis. A five cyclic skeleton of this compound was deduced by extra 5 unsaturation degrees from calculation of double bonds. The D-ring was deduced by the unique of endomethylene H9_a/b (δ_H 2.51 and 3.47) and propenyl functionality (δ_H 1.86, 4.07, 4.38, and 4.07). The presence of C10 propenyl moiety was obtained by HMBC correlation of H10 (δ_H 4.07) to C11 (δ_C 150.9). The pyran group was established from proton coupling of H14 (δ_H 6.99, d) to H15 (δ_H 5.73, d) and HMBC correlation between H14a and oxygenated carbon C15 (δ_C 129.88). The locations of methyls were deduced from HMBC of both H17 (δ_H 1.51) and H18 (δ_H 1.54) to C15. From this the compound 5 was identified as 8,9-dihydro-6,10,11,13-tetrahydroxy-3,3-dimethyl-9-(1-methylethenyl)-3H,7H-benzo[c]pyrano[3,2-h]xanthen-7-one and named as artobiloxanthone.

HNE inhibitory activity of isolated compounds and their kinetics

HNE inhibitory activities of the isolated flavones at different concentrations (1–8) screened using a UV assay according to a previously reported method [8, 22]. All isolates inhibited in a dose-dependent manner the HNE enzyme, but alkylated flavones (2–4) and dihydrobenzoxanthones (5–8) showed much higher inhibitions than mother skeleton, norartocarpetin (1), their IC₅₀s given in Table 1. As shown in Fig. 2a, both compounds 1 and 2 showed a discriminatory dose-dependence curve. The alkylated flavone 2 (IC₅₀ = 14.8 µM) was 20-fold active than mother compound 1 (IC₅₀ > 300 µM). The similar inhibitory tendency was also observed from other alkylated flavones 3 and 4, which indicated that the prenyl group on C3 was a crucial functionality to HNE inhibition. Dihydrobenzoxanthones (5-8) inhibited HNE significantly with IC₅₀ values of 9.8 ~ 28.7 μM, compared to the mother compound. The number of the hydroxyl group of B-ring affected inhibitory capacities as follow: compounds 5 (IC₅₀ = 9.8 μM) versus 8 (IC₅₀ = 28.7 μM).

Table 1 Inhibitory effects of compounds on HNE

Fig. 2

Enzyme inhibition activity of compounds 1 and 2 against human neutrophil elastase (HNE). a Dose-dependent effect of the compounds 1 and 2 on HNE. b The catalytic activity of HNE as a function of enzyme concentration at different concentrations of compound 2. c Lineweaver–Burk plots of compound 2. (Inset) Tendencies for the maximal velocity (V_max) and Michaelis–Menten constants (K_m) for compound 2. d Dixon plots of compound 2

Reversibility between inhibitor and HNE enzyme was confirmed by the plotting activity of residual enzyme inhibited with different concentrations of inhibitor versus the enzyme concentration. Inhibitors showed a comparable plotting tendency between enzyme activity and concentration as like in Fig. 2b. As shown in Fig. 2b, increasing of representative inhibitors 2 concentrations caused a reduction of the slopes of lines, of which the straight lines family was passed through the beginning of coordinates. This phenomenon indicated that inhibitor 2 had reversibility to the enzyme.

The characterization of the inhibitory mechanism was showed off using Michaelis–Menten double reciprocal plots. First of all alkylated flavones (2–4) were confirmed as noncompetitive inhibitors, because of no change of K_m and the decrease of V_max as shown in Figs. 2c and 3a insets. Figures 2c and 3a elucidated Lineweaver–Burk plots of alkylated flavones 2 and 3, which have a common intercept on the x-axis at different inhibitor concentrations. The K_i values of 2 and 3 was calculated to be 12.6 μM and 15.4 μM, respectively by Dixon plots (Figs. 2d and 3b and Table 1). In the case of dihydrobenzoxanthones (5-8), they were determined as mixed type inhibitors. Since increasing of the inhibitor concentrations resulted in a common intercept of the family of lines on the left of the y-axis and above the x-axis as like Fig. 3c and e. The mixed type inhibitors usually have a dissimilar affinity to the substrate (mixed type I) and to the free enzyme (mixed type II). The results of the measured velocity of compounds 5 and 6 were fitted to the Eqs. (3) and (4) for the calculation of K_I and K_IS by secondary plots of K_m/V_max and 1/V_max versus compound concentrations. In particular, compound 5 was found to be mixed type I (K_I= 6.5 μM < K_IS= 31.8 μM), as like compound 6 (K_I= 4.7 μM < K_IS= 19.9 μM). At last, the K_i values of compounds 5 and 6 were determined to be 6.4 and 4.7 μM respectively, by Dixon plots (Fig. 3d and f, Table 1).

Fig. 3

Enzymatic kinetics of compounds 3, 5, and 6. Lineweaver–Burk plots for HNE inhibition of a compounds 3, c compound 5, and e compound 6. (Inset) a Tendencies for the maximal velocity (V_max) and Michaelis–Menten constants (K_m) for compound 3 and The secondary plots of the straight lines of slope and intercept versus the concentration of c compound 5 and e compound 6. Dixon plots of b compounds 3, d compound 5, and f compound 6

Binding affinities between HNE and compounds

Intrinsic fluorescence (FS) intensity of protein usually changes as a function of ligand concentration when protein acts together with another ligands [23]. Generally, intrinsic FS of protein originates from phenylalanine (Phe), tryptophan (Trp) and tyrosine (Tyr) residues. The most sensitive fluorophore residue is a Trp [24]. HNE enzyme has a suitable FS intensity due to three residues of Trp (Trp-12, Trp-127, and Trp-212) [25]. The binding affinity of compounds with HNE was determined by FS quenching effects. There is no measurable emission from other ingredients of the mixture solution in the current experiment conditions (em. 300 ~ 400 nm). Figure 4 showed the effects of FS quenching of the representative compounds 1, 2 and 8 at the same concentrations (0 → 50 µM). FS intensity was decreased by inhibitor concentrations, but quenching affinities allowed to their HNE inhibitory activities. As shown in Fig. 4c, one of the most active compound 2 (IC₅₀ = 14.8 µM) quenched FS intensity dose-dependently, to be almost complete quenched at 50 µM concentration. FS quenching effect was not observed from mother compound 1 (IC₅₀ > 300 µM) at 50 µM concentration. The moderate compound 8 showed the average FS quenching effect also correlated with its inhibitory potency (IC₅₀ = 28.7 µM). The parameters regarding binding affinity can be calculated by the following Eqs. (5)–(6):

Fig. 4

The fluorescence spectra a-c of α-glucosidase at different concentrations (0, 3.125, 6.25, 12.5, 25.0, and 50.0 μM for curves a → f, T = 37 °C, a For compound 1, b For compound 8, c For compound 2. (Inset) Areas under curves of fluorescence emission spectra for HNE are shown for compounds 1, 8, and 2, respectively. d The correlation curve between the inhibitory potencies (IC₅₀s) and the Stern–Volmer constant (K_SV)

$$ F_{0} - \, F{ = 1 + }K_{\text{SV}} [ {\text{Q}}_{\text{f}} ] $$

(5)

$$ {\text{log [(}}F_{ 0} - F)/F] = \log K_{\text{A}} + n\log [{\text{Q}}_{\text{f}} ] $$

(6)

F₀ and F were the FS emission intensities in the absence or presence of a quencher, respectively. [Q_f] is a concentration of compounds, n is a binding sites number, and K_A is a binding constant. Binding parameters (K_SV, K_A, and n) of the compounds were calculated. Stern–Volmer constants (K_SV) shown in Table 2 had a significant correlation with their inhibitory potencies (IC₅₀). For example, compound 2 (IC₅₀ = 14.8 µM, K_SV = 0.0954 × 10⁵ L mol⁻¹) had 6-fold higher value of K_SV than 8 (IC₅₀ = 28.7 µM, K_SV = 0.0158 × 10⁵ L mol⁻¹). As well, all compounds Stern–Volmer constants strongly correlated (R² = 0.9862) to IC₅₀. These results designate a single binding site of inhibitors to HNE enzyme.

Table 2 Fluorescence quenching effect of compounds 1–8 on HNE