5-Hydroxytryptamine_1A receptor binding activity of bisbenzyltetrahydroisoquinoline alkaloids from Popowia odoardi

Abstract

A 96-well filtration-based 5-HT_1A receptor-radioligand binding assay was optimized and adapted to carry out a bioassay-guided fractionation of the methanol extract of the bark of Popowia odoardi Diels (Annonaceae). This purification led to the isolation of two compounds identified as O-methyldauricine (1) and popisidine (2), which bind competitively to 5-HT_1A receptors at high concentrations with K_i values of 4.5 and 5.8 μM, with [³H]8-OH-DPAT as the radioligand. These results suggest that both of these bisbenzyltetrahydroisoquinoline alkaloids influence the activity of 5-HT_1A receptors through competitive binding as agonists or antagonists and may possess anxiolytic and anti-depressive effects.

Keywords::

• Alkaloid
• Annonaceae
• 5-hydroxytryptamine receptor
• O-methyldauricine
• Popowia odoardi
• popisidine

Introduction

5-Hydroxytryptamine (5-HT, serotonin) mainly functions as a neurotransmitter for tryptaminergic neurons and its corresponding receptors in the brain, mediating sleep, feeding, mood, temperature regulation, and sexual behavior. To date, seven 5-HT receptor families, designated 5-HT₁ to 5-HT₇, have been identified and classified into at least 14 subtypes (Hoyer et al., 2002).

The 5-HT_1A receptor is an important therapeutic target for the treatment of CNS disorders such as anxiety, depression, and schizophrenia, and potentially Alzheimer’s disease (Hensler, 2003; Hoyer et al., 2002; Schechter et al., 2002). For example, the psychoactive agents buspirone and gepirone exert both anxiolytic and anti-depressant effects. In addition to depression and anxiety, 5-HT_1A receptor ligands have therapeutic utility in schizophrenia. This has led to the development of the antipsychotic drugs ziprasidone and aripiprazole (Gunasekara et al., 2002). Neurochemical and behavioral data suggest that receptor blockage by a 5-HT_1A antagonist appears to enhance activation and signaling through heterosynaptic neuronal circuits known to be involved in cognition processes. This suggests that the 5-HT_1A receptor antagonists may have therapeutic potential in Alzheimer’s disease and other diseases associated with cognition dysfunction (Schechter et al., 2002).

In the search for plant extracts with G-protein-coupled receptor and antimicrobial activities, more than 200 Malaysian plants were screened; the plants were chosen based on availability and similarity to species with ethno-botanical uses in medicine (Chung et al., 2004, 2005a, 2005b, 2006, 2008). The crude methanol extract of Popowia odoardi Diels (Annonaceae) showed pronounced competitive 5-HT receptor_1A binding activity. The competitive inhibition of GABA_B receptor, dopamine receptor (D₂S), muscarinic receptor, chemokine (CXCR2) receptor, and chemotactic C5a-anaphylatoxin receptor binding activity was low or minimal. P. odoardi, a small tree that grows widely in the lowland or hill forests of Borneo, was therefore selected for further bioactivity (5-HT1A receptor binding) guided fractionation studies.

The genus is known as “pisang-pisang” in Peninsula Malaysia, and species such as P. nervifolia Maing., P. nervosa Ridl. and P. ramosissima Hook. f. were used in house construction (Burkill, 1935). Although P. odoardi is not known in indigenous medicine, related species have been used in other parts of the world. For example, P. fornicata Baill. (synonym, Monanthotaxis fornicata Baill.) is indicated for the treatments of mental diseases and snake bites in East Africa (Hedberg et al., 1982), while P. herenthas is used in Madagascar as febrifuge (Rasoanaivo et al., 1992). A phytochemical screening survey reported that P. odoardi contain several alkaloids which have not been isolated and identified (Teo et al., 1990).

However, isoquinoline-derived alkaloids (Chuah et al., 2001; Josang et al., 1986a,b) and flavonoids (Waterman & Pootakahm, 1979) were isolated from other members of the Popowia genus. We now report the isolation of two bisbenzyltetrahydroisoquinoline alkaloids from the crude methanol extract of P. odoardi and their 5-HT1A receptor binding activities.

Materials and methods

Materials

8-Hydroxy-2-[2,3-³H]di-n-(propylamino)tetralin ([³H]8- OH-DPAT; 220 Ci/mmol) was obtained from GE Healthcare Life Sciences (Little Chalfont, Buckinghamshire). Tris(hydroxygethyl)aminomethane (Tris base), Bradford reagent, bovine serum albumin type V, 5-hydroxytryptamine hydrochloride (5-HT.HCl), (±)-8-hydroxy-2-(di-n-propylamino)-tetralin ([±]-8-OH-DPAT), buspirone hydrochloride, dl-propanolol hydrochloride, and (–)scopolamine methyl bromide were obtained from Sigma-Aldrich (St. Louis, MO).

Ninety-six-well flat-bottom Falcon plates were obtained from Becton Dickinson Labware (Franklin Lakes, NJ). Microscint-O scintillation cocktail, TopSeal A, and UniFilter GF/C plates were supplied by Perkin Elmer (Wellesley, MA). Silica gel 60 (mesh 230–400) for column chromatography and silica gel 60 F²⁵⁴ (0.5 mm thick, glass plate) for TLC were obtained from Merck (Darmstadt, Germany). All solvents were of analytical grade and purchased from Fisher Scientific (Loughborough, Leicestershire).

General experimental procedures

¹H and ¹³C Nuclear magnetic resonance (NMR) spectra were obtained on a 500 MHz Varian Unity INOVA spectrometer (Palo Alto, CA). Correlation spectroscopy (COSY), Heteronuclear multiple-bond correlation (HMBC), Heteronuclear multiple quantum correlation (HMQC), and Nuclear overhauser effect spectroscopy (NOESY) NMR experiments were also conducted. Deuterated chloroform (CDCl₃₎ was used as the solvent and tetramethylsilane (TMS) as the internal reference. Mass spectrometry (MS) data were recorded on a Finnigan Polaris Q (Wattham, MA) through direct injection.

Plant material

The bark of P. odoardi Diels (Annonaceae), previously reported as P. odoardoi Diels, was collected from Sabah, Malaysia (Chung et al., 2005a, 2005b, 2006), and identified at the Sandakan (SAN) Herbarium of the Forest Research Center, Sepilok, Sandakan, Malaysia (voucher no.: SAN 143395).

Extraction and isolation

The air-dried bark (200 g) was powdered and extracted exhaustively with methanol, and after removal of the solvent in vacuo, yielded 19.8 g (9.9%) (5-HT_1A receptor binding inhibition at 10 μg/well, 71.1% ± 5.9%). The crude extract (10 g) was dissolved in CHCl₃/MeOH (1:1) and partitioned with 5% aqueous hydrochloric acid. The aqueous acid phase obtained was basified with 30% ammonium hydroxide to pH 10, followed by partitioning with CHCl₃. The chloroform fraction was then dried in vacuo to give a crude alkaloid extract (2.1 g) (5-HT_1A receptor binding inhibition at 10 μg/well, 102.1% ± 1.2%; 1 μg/well, 63.5% ± 1.3%). The CHCl₃/MeOH fraction containing non-alkaloid materials showed low inhibition on 5-HT_1A receptor binding (1 μg/well, 14.6% ± 4.4%).

The CHCl₃ extract (600 mg) was fractionated on a silica gel chromatography column (silica gel 60, mesh 230–400) using a CHCl₃/MeOH step gradient system (100% CHCl₃, 98:2, 96:4, 94:6, 92:8, 90:10, 80:20, 50:50), and eight fractions (F-1 to F-8) were collected. Fraction F-4 (47.3 mg) and fraction F-7 (74.3 mg) showed 84.4% ± 1.1% and 52% ± 7.5% inhibition (1 μg/well), respectively. 15 mg of fraction F-4 was further separated by preparatory scale TLC. Purification by TLC on a silica gel 60 F²⁵⁴ plate using CHCl₃/MeOH (95:5) saturated with ammonium hydroxide gave compound 1 (R_f 0.4; 6.6 mg; 0.22% of the powdered bark). For fraction F-7, 28 mg was subjected to further separation as described above. Purification by TLC using CHCl₃/MeOH (97:3) saturated with ammonium hydroxide yielded compound 2 (R_f 0.4; 13.5 mg; 0.12% of the powdered bark).

Preparation of 5-hydroxytrptamine receptor crude membrane

Rat brain membranes (cerebral cortices) from male Sprague-Dawley rats (300–400 g) were prepared according to the methods described by Chung et al. (2008) with minor modifications (Soo, 2006), and in accordance to the guideline promulgated by the Ethics Committee on Animal Welfare and Experimentation, Faculty of Medicine, University of Malaya, Malaysia (Ref. No.: FAR/21/10/2005/MRM (R)). Rat cerebral cortices were removed on ice, weighed, and placed in 10 volumes (w/v) of ice-cold 50 mM Tris-HCl buffer, pH 7.4, The tissue was homogenized with an Ultra-Turrax homogenizer (IKA Labortechnik, Staufen, Germany) and glass-teflon homogenizer (Terre Haute, Indiana, USA). The homogenate was then centrifuged at 40,000 g for 15 min (Sorvall^® Ultra Pro 80, Kendro, Newtown, CT). The resulting supernatant was discarded, and the pellet was resuspended in the same volume of ice-cold buffer and centrifuged at 40,000 g for another 15 min. This procedure was repeated, and the resuspension was incubated in a water bath at 25°C for 20 min to destroy endogenous serotonin (Harms et al., 2000) before being centrifuged as above. All centrifugation steps were performed at 4°C. The final membrane preparation was suspended in 2-3 volumes of ice-cold 50 mM Tris-HCl buffer, pH 7.4, and stored at −86°C until use. Membrane protein concentration was determined using the Bradford protein reagent (Sigma-Aldrich) with bovine serum albumin as the standard, and the protein content corresponded to about 20 mg protein/mL.

96-Well microplate 5HT_1A receptor binding assay

Crude extract and partitioned fractions dissolved in DMSO (final concentration, 5% v/v) were tested at 50, 10, and/or 1 μg/well. The isolated compounds and standards (buspirone and (±)-8-OH-DPAT, serotonin, propanolol) were tested at 11–12 concentrations (10⁻²–10⁻¹³ M) in order to obtain K_i values. All assays were performed in triplicate and inhibition results shown are the mean ± SD (n = 3).

The 96-well microplate 5-HT_1A receptor assay was optimized and performed as described by Soo (2006). Around 200 μL of the membrane suspension per well (35 μg of protein per well) were added to [³H]8-OH-DPAT (final concentration, 0.4 nM) to determine total binding (TB). Non-specific binding (NSB) was determined by adding 25 μL of 5-HT (final concentration, 10 μM) to the reaction mixture. TB was subtracted by NSB to give specific binding (SB). To determine the inhibition of binding by [³H]8-OH-DPAT in the presence of plant extracts, isolated compounds, and known ligands, 25 μL of the respective compound was added instead of 5-HT. The reaction mixture was incubated at 25°C at pH 7.4 for 90 min. After incubation, the reaction mixtures were rapidly filtered through pre-wetted GF/C filter plates using the FilterMate Cell Harvester (Perkin Elmer) and washed with 200 μL/well of ice-cold 50 mM Tris-HCl buffer pH 7.4 (four times). The seal was then applied to the bottom of the plate, and 25 μL aliquots of [³H]8-OH-DPAT were spotted directly onto the GF/C filter plate in triplicate to determine the total radioactivity per well. The plate was then air-dried and 25 μL of scintillation cocktail were added to each well (MicroScint-O). The top side of the plate was then sealed with TopSeal A, and the plate was agitated at 400 rpm for 5 min before counting each well for 1 min with a TopCount NXT microplate scintillation counter (Perkin Elmer).

Data analysis on saturation experiments using the non-linear least-squares regression method was performed using PRISM^® software, version 3.03 (GraphPad Software, San Diego, CA), and the results were presented as the mean ± SD (n = 3). The K_i was calculated according to the Cheng-Prussof equation:

where D is the concentration of the radioligand and K_D is the dissociation constant of the radioligand (Cheng & Prussof, 1973). The quality of the assays was assessed by performing screening window factor (z’ factor) analysis according to the equation: 1-(3SD_TB+3SD_NSB)/(Mean_TB-Mean_NSB) as described by Zhang et al. (1999), where SD_TB is the standard deviation of total binding, SD_NSB is the standard deviation of non-specific binding, Mean_TB is the mean of total binding, and Mean_NSB is the mean of non-specific binding.

The percentage of inhibition of specific binding to 5-HT_1A receptors in the presence of the test compounds or extracts was calculated using a standard data reduction algorithm as shown below:

where TB is the total binding, B is the binding in the presence of test sample, and NSB is the non-specific binding in the presence of excess reference ligand. The data was expressed as the mean ± SD (n = 3) and analyzed statistically using Student’s t test.

Results and discussion

The agonist, 8-hydroxy-2-[2,3-³H]di-n-(propylamino)tetralin ([³H]8-OH-DPAT), was used as the radioligand as it exhibited saturation binding with high affinity to 5-HT_1A receptors. The binding data were fitted to the one-site and two-site binding models using the nonlinear regression analysis in Prism^® software. The statistical F test indicated that the two-site binding model gave the best fit (p <0.05) with a K_d(H) of 0.15 nM and a K_d(L) of 10.5 nM; B_max(H) was 143.4 fmol/mg protein and B_max(L) was 750.1 fmol/mg protein (Figure 1A). These values were similar to those of Carli et al. (1997) and Nénonéné et al. (1996). The inhibition constant (K_i) for the competitive reference ligand (5-HT) was 3.25 (K_d(H)) and 3940 (K_d(L)) nM, and the values were similar to those of Nénonéné et al. (1994). In the time-course experiment, association of [³H] 8-OH-DPAT was completed in less than 90 min, clearly demonstrating that the 90-min incubation time was adequate (Figure 1B). The quality of the assay was assessed using z’ factor analysis; A z’ factor of 0.91 for the 5-HT_1A receptor binding assay demonstrated the suitability and robustness of the data quality of the assay for screening (Zhang et al., 1999).

Figure 1.  (A) Saturation curve for the 5-hydroxytryptamine_1A receptor with [³H]8-OH-DPAT as the radioligand. Receptor membrane (35 μg protein/well) was incubated with increasing concentrations of [³H]8-OH-DPAT (0–10 nM) at 25°C for 90 min. The binding was fitted to the two-site binding model. The Rosenthal plot is shown as the inset of the saturation binding curve. (B) Time course of association of [³H]8-OH-DPAT to the 5-hydroxytryptamine_1A receptor. Receptor membrane (35 μg protein/well) was incubated with [³H]8-OH-DPAT (0.2 nM) at 25°C for different time periods (0–180 min). Each data point is expressed as mean ± SD (n = 3).

In the bioassay-guided fractionation of P. odoardi using the optimized 5-HT_1A receptor-radioligand binding assay, the crude chloroform extract gave a percentage of inhibition of 71.1% ± 5.9% (10 μg sample/well) (Student’s t-test, P <0.05 compared to the control), and the crude alkaloid extract gave a percentage of inhibition of 102.1% ± 1.2% (10 μg sample/well) and 63.5% ± 1.3% (1 μg sample/well) (Student’s t-test, P <0.05 compared to the control). The alkaloid-containing fractions with 5-HT_1A receptor binding activities were separated further to obtain pure compounds. By comparison of the spectral data (EI-MS, ¹H-, ¹³C-NMR, COSY, HMQC, HMBC, NOESY, DEPT) with those in the literature (Chuah et al., 2001; Jossang et al., 1986a), compounds 1 and 2 were identified to be the bisbenzyltetrahydroisoquinoline alkaloids O-methyldauricine (1) and popisidine (2), respectively (Figure 2).

Figure 2.  Chemical structures of compounds isolated from P. odoardi as competitive inhibitors on [³H]8-OH-DPAT binding to 5-hydroxytryptamine_1A receptor.

O-Methyldauricine (1) and popisidine (2) showed moderate inhibition with a K_i of 4.5 μM and 5.8 μM, respectively (Figure 3); the K_i values were calculated using the best fit one-site binding model. This is higher than the inhibition constants of (±)-8-hydroxy-2-(di-n-propylamino)-tetralin [(±)-8-OH-DPAT] buspirone, and 5-hydroxytryptamine (5-HT) (Table 1). The values are several-fold higher than that of propanolol (0.14 μM), a 5HT_1A/5HT₂ receptor antagonist, but are similar to those of other structurally related isoquinoline alkaloids, such as annonaine, nornuciferine and asimilobine (less than 10 μM) (Figure 4) (Hasrat et al., 1997a).

Figure 3.  Competitive inhibition on the binding of [³H]8-OH-DPAT to the 5-hydroxytryptamine_1A receptor by O-methyldauricine and popisidine. Receptor membrane (35 μg protein/well) was incubated with [³H]8-OH-DPAT (0.4 nM) in the presence of O-methyldauricine and popisidine at 25°C for 90 min. Each data point is expressed as mean ± SD (n = 3).

Figure 4.  Chemical structures of asimilobine, nornuciferine and annonaine (aporphine alkaloids).

Table 1.  K_i values obtained from the competition experiments with standard unlabelled ligands on 5-HT_1A receptor binding activity with [³H]8-OH-DPAT as the radioligand.

Unlabelled ligand	Pharmacological action	Ki (M)
8-OH-DPAT	5-HT1A receptor agonist	0.2 × 10^−9, 1.34 × 10^−8 *
Buspirone	5-HT1A receptor agonist	1.46 × 10^−9, 2.69 × 10^−8 *
5-HT	5-HT receptor agonist	3.25 × 10^−9, 3.94 × 10^−6 *
Propanolol	5-HT1A receptor antagonist	1.37 × 10^−7 **

The values shown were the mean of three independent experiments; (±)-8-OH-DPAT = (±)-8-hydroxy-2-(di-n-propylamino)-tetralin; 5- HT = 5-hydroxytryptamine; *The data were fitted to a two-site competition curve giving inhibition constants for high-affinity (K_i(H)) and low affinity (K_i(L)) sites; **The data were fitted to a one-site competition curve.

Indigenous medicine, related species have been used in other parts of the world. For example, P. fornicata Baill. (synonym, Monanthotaxis fornicata Baill.) is indicated for the treatments of mental diseases and snake bites in East Africa (Hedberg et al., 1982), while P. herenthas is used in Madagascar as febrifuge (Rasoanaivo et al., 1992). A phytochemical screening survey reported that P. odoardi contain several alkaloids which have not been isolated and identified (Teo et al., 1990).

Both O-methyldauricine (1) and popisidine (2) belong to the group of bisbenzyltetrahydroisoquinoline alkaloids, which are derived biogenetically via phenol-oxidative coupling of two units of benzyltetrahydroisoquinoline alkaloids. The bisbenzlyisoquinoline alkaloids constitute one of the best-studied classes of natural products. They are reported to have a variety of biological activities, including antimalarial (Frappier et al., 1996), cytotoxic (Angerhofer et al., 1999), anti-HIV (Kashiwada et al., 2005), antitumor (Jossang et al., 1996), dopamine inhibitory (Cortes et al., 1992; Protais et al., 1995), antihypertensive (Qian, 2002), antibacterial (Lohombo-Ekomba et al., 2004), and smooth muscle relaxant activity (Zafra-Polo et al., 1993). However, 5-HT_1A receptor binding activity by bisbenzylisoquinoline alkaloids has not been reported. Structurally related isoquinoline derivatives, such as annonaine, nornuciferine, and asimilobine, isolated from the fruit of Annona muricata, have been shown to inhibit the binding of [³H]rauwolscine to 5-HT_1A receptors in calf hippocampus (Figure 4). The IC₅₀ values for inhibition of receptor binding for annonaine, nornuciferine, and asimilobine were 3, 9, and 5 μM, respectively (Hasrat et al., 1997b), which are close to the K_i values of O-methyldauricine (1) (4.5 μM) and popisidine (2) (5.8 μM). In the functional assay, based on the inhibition of the accumulation of cAMP in NIH-3T3 cells stably transfected with 5-HT_1A receptor from human, the K_i values for annonaine, nornuciferine, and asimilobine were estimated to be less than 10 μM (Hasrat et al., 1997a). In another study, the IC₅₀ for the interaction of annonaine with the dopamine reuptake pump was reported to be 0.8 μM (Protais et al., 1995). Therefore, these aporphines were suggested to be non-selective 5-HT_1A receptor agonists. The IC₅₀ of O-methyldauricine in a ³H-dopamine uptake binding assay has been reported to be 5.8 μM (Protais et al., 1995), but O-methyldauricine is less active in the D₁ and D₂ dopamine receptor binding assays, with an IC₅₀ of 100 μM (Cortes et al., 1992).

In summary, the findings suggest that O-methyldauricine (1) and popisidine (2) have moderate affinity towards cerebral cortical 5-HT_1A receptors acting as agonists or antagonists, and may contribute to their CNS activities. Therefore, these compounds warrant further investigation for their potential therapeutic activity.

Acknowledgements

This project was partly supported by the Biotechnology Directorate, Ministry of Science, Technology and Innovation, Malaysia (IRPA 26-02-06-0127).

Declaration of interest: The authors alone are responsible for the content and writing of the paper.