Toxicity–structure activity evaluation of limonoids from Swietenia species on Artemia salina

Abstract

Context: Many plant extracts and compounds are being investigated for their cytotoxicity and hence their medicinal or therapeutic properties. Reports of toxicity studies with limonoid analogs have been sparse and have involved mainly crude extracts. In this study, individual natural limonoids have been isolated and their toxicity manipulated via semisynthesis.

Objective: The lethality of limonoid analogs from Swietenia macrophylla King and Swietenia aubrevilleana Stehlé & Cusin (Meliaceae) against Artemia salina Leach was determined.

Materials and methods: Four known natural limonoids were isolated from the dry ground seeds of S. macrophylla and S. aubrevilleana, modified using acylation and hydrolysis reactions and tested in A. salina lethality assays at 1–400 ppm. A 50% lethal concentration (LC₅₀) was determined by probit analysis.

Results: Higher levels of toxicity were achieved in most of the prepared analogs compared with the parent natural limonoids. The compound showing the highest toxicity with LC₅₀ 3.9 ppm was 3-O-benzoyl-3-detigloylisoswietenine (20). Other analogs with high toxicity were 6-O-benzoylswietenolide (7), 6-O-benzoylswietenine (17), and 3,6-O,O-dipropionylswietenolide (9), which showed LC₅₀ values of 4.3, 7.5, and 28.5 ppm, respectively.

Discussion and conclusions: Toxicity can be improved via semisynthesis. The compounds exhibiting high toxicity (low LC₅₀) may be good candidates for cytotoxicity studies.

Keywords::

• Limonoids
• semisynthesis
• bioassay
• LC₅₀
• swietenine
• swietenolide
• swietemahonin G
• Meliaceae
• mahogany
• SAR studies

Introduction

Fifty percent or more of organic compounds introduced as pharmaceutical drugs are derived from natural products (El Sayed, 2002; Ghisalberti, 2008) and some have been discovered because of traditional or folk remedies (Alves and Rosa, 2007). Among the plants being investigated for their medicinal properties are species of Meliacae and Rutaceae. For example, the fruit concentrate of Swietenia macrophylla King (Meliaceae) has been used to improve blood circulation and skin condition in Southern Mexico, Central America, and Bolivia (Tan et al., 2009); extracts of Swietenia mahogani (L.) Jacq. (Meliaceae), a parent species (Pennington et al., 1981) for Swietenia aubrevilleana Stehlé & Cusin (Meliaceae) are used in Indonesia for the treatment of hypertension and diabetes (Chen et al., 2007) and Citrus seeds (Rutaceae) are used in traditional Chinese medicine to treat liver diseases and hernia (Tian et al., 2001). Phytochemical investigations have revealed limonoids and triterpenoids, in general, to be the major compounds of interest in these Meliaceae and Rutaceae folk medicines, and more recently there is evidence that these compounds may be effective anticancer agents (Setzer and Setzer, 2003; Kumar et al., 2010).

The Artemia salina lethality assay is regarded as a useful screening tool for preliminary assessment of toxicity and hence potential cytotoxicity of compounds, which may suggest medicinal or therapeutic effects (Carballo et al., 2002; Moshafi et al., 2009). Some reports of A. salina lethality studies using limonoid analogs have involved extracts from Melia azedarach L. (Pisutthanan et al., 2004; Fukuyama et al., 2006); nimbolide (from Azadirachta indica A. Juss. (neem) seeds) with an ED₅₀ of 568.7 ppm (Genupur et al., 2006); capensolactone 1 and 2 showing moderate biological activity (Mulholland and Lourine, 1998) and the chloroform, methanol, and aqueous extracts of A. indica giving LC₅₀ of 7.5 ppm (Bhat et al., 2009). The aim of this research was to document the toxicity of our isolated limonoids from S. macrophylla and S. aubrevilleana using A. salina lethality assays and alter their toxicity by semisynthesis.

Materials and methods

Isolation, semisynthesis, and chemical data

Fruits of S. macrophylla and S. aubrevilleana were collected from the grounds of The University of the West Indies St. Augustine campus in February 1999 (identified by Winston Johnson, National Herbarium of Trinidad and Tobago, 2nd Floor Sir Frank Stockdale Building, Faculty of Science and Agriculture, University of the West Indies, St. Augustine, Trinidad). Voucher specimens are secured by the National Herbarium of Trinidad and Tobago (S. macrophylla 33274 and S. aubrevilleana 33272). Swietenolide (1), 3,6-O,O-diacetylswietenolide (2), swietenine (3), swietemahonin G (4), and 2-hydroxyswietenine (5) (Figure 1) were isolated from the EtOAC extract of the dried seeds (Mootoo et al., 1999) using column and plate chromatography, utilizing solvent systems CHCl₃–EtOAc, petroleum ether–EtOAc, or petroleum ether–CHCl₃–MeOH.

Figure 1.  Structure of compounds used in A. salina Leach toxicity assays.

The analogs 6–15, 17, 20–22, 25, and 26 (Figure 1) were prepared from the acylation reactions of compounds 1, 3, 19, and 24 (Figure 1), respectively. Each compound was dissolved in a minimum of 60 equivalents of pyridine, followed by the addition of minimum 6.3 equivalents of the selected acylating reagent (benzoyl chloride, isovaleryl chloride, isobutyryl chloride, propionic anhydride, or tert-butylacetyl chloride). The reaction mixture was stirred for 18–24 h at 25°C. Reactions with benzoyl chloride, propionyl anhydride, and isobutyryl chloride were refluxed at 40–60°C for an additional 1 h. The reaction mixture was quenched with ice and the suspension extracted with EtOAc. The EtOAc layer was washed with NaHCO₃ (1 M) followed by distilled H₂O, HCl (0.1 M) and finally distilled H₂O. The EtOAc layer was then dried with anhydrous Na₂SO₄, filtered, evaporated in vacuo to dryness and purified using column and preparative thin layer chromatography and solvent systems CHCl₃–EtOAc, petroleum ether–EtOAc, petroleum ether–CHCl₃–MeOH, or hexane–EtOAC (Fowles et al., 2010).

Compound 16 was prepared by heating compound 1 (0.10 mmol) at 60–70°C for 1.5–2 h then stirring overnight at 25°C in MeOH (10 mL), methanol KOH (50 mL, 0.9 M), and distilled water (50 mL). Compounds 18 and 23 (0.1 mmol) were prepared by refluxing at 60–70°C 3 and 5 in methanol KOH (3: 50 mL, 0.5 M and 5: 50 mL, 0.9 M methanol KOH) and distilled water (50 mL). Compound 4 was stirred in methanol KOH for 3 h at 25°C to give its hydrolyzed product (not isolated). Work up involved cooling the alkaline solutions then acidifying with concentrated HCl to pH 1 (1 and 5), pH 2.5 (3) and pH 4 (4). The acidic solutions were extracted with EtOAC, dried with anhydrous Na₂SO₄, filtered, and evaporated in vacuo to dryness to give compounds 16, 18, and 23 (Figure 1). Unreacted starting compounds were removed by dissolving the mixture in CHCl₃ and extracting the solution with NaHCO₃ (1 M). Methylation of compounds 18, 23, and hydrolyzed product of compound 4 using an ethereal solution of diazomethane gave compounds 19, 24, and 27 (Figure 1). Analogs were identified using one- and two-dimensional (COSY, HMQC, and HMBC) ¹H and ¹³C NMR, IR, and MS (Fowles et al., 2010).

Bioassay

Eggs of A. salina were placed in a vessel partially filled with artificial seawater and incubated at 29°C for 48 h until they hatched. The test compounds (1 mg each) were dissolved in dimethyl sulfoxide (DMSO) (50 μL) and made up to 1 mL with artificial seawater. Serial dilutions (50, 100, 200, 400, 600, and 800 ppm) of each test compound were prepared. The diluted samples (100 μL) were placed in microplate wells (Corning Glass Works, Corning, New York 14831, USA) in triplicates. Control wells comprised DMSO diluted with artificial seawater. Artificial seawater (100 μL) containing 10–15 A. salina nauplii was added to each well and the plate was incubated at 29°C for 24 h (the effective concentrations were therefore 25, 50, 100, 200, 300, and 400 ppm). Plates were examined under a microscope and the number of dead (non-mobile) nauplii in each well counted. Where 100% mortality was observed at 25 ppm, the experiment was repeated with serial dilutions from 25 to 1 ppm. Methanol (100 μL) was then added to each well and the total number of organism in each well counted after 15 min. LC₅₀ values were calculated using probit analysis (Finney, 1971). Statistical analysis was done using the Tukey–Kramer multiple comparison test at a 5% significance level (P = 0.05).

Results

All the synthetic swietenolide (1) derivatives were found to be more toxic than the parent compound (LC₅₀ > 500 ppm). Noted among the derivatives were 6-O-benzoylswietenolide (7) and 3,6-O,O-dipropionylswietenolide (9), which showed low LC₅₀ values (high toxicity) of 4.3 and 28.5 ppm, respectively. Addition of one methyl group to compound 10 caused a decrease in toxicity from LC₅₀ 85.0 to 120.9 ppm as seen in 3-O-isobutyrylswietenolide (14), whereas further addition of one methylene group to give 3-O-isovalerylswietenolide (13) increased the toxicity to LC₅₀ 73.8 ppm (Table 1).

Table 1.  Toxicity (LC₅₀) of compounds 1–27 against A. salina Leach.

Compounds	LC50 (ppm)*	95% CI
Swietenolide (1)	>500	—
3,6-O,O-Diacetylswietenolide (2)	>500	—
Swietenine (3)	>500	—
Swietemahonin G (4)	220.1^a	200.0–243.5
2-Hydroxyswietenine (5)	>500	—
3,6-O,O-Dibenzoylswietenolide (6)	73.1^be	50.8–105.3
6-O-Benzoylswietenolide (7)	4.3^c	2.8–6.7
3-O-Benzoylswietenolide (8)	117.8^ab	66.1–209.8
3,6-O,O-Dipropionylswietenolide (9)	28.5^bd	15.9–51.0
3-O-Propionylswietenolide (10)	85.0^bd	69.3–104.3
6-O-Propionylswietenolide (11)	63.8^bde	50.1–81.1
3-O-Acetoxyacetylswietenolide (12)	145.6^d	123.9–171.2
3-O-Isovalerylswietenolide (13)	73.78^bd	54.2–100.5
3-O-Isobutyrylswietenolide (14)	120.9^bd	105.1–139.2
3,6-O,O-Di-tert-butylacetylswietenolide (15)	>500	—
Demethylisoswietenolide (16)	497.7^f	484.5–511.3
6-O-Benzoylswietenine (17)	7.5^c	3.9–14.7
Demethyl-3-detigloylisoswietenine (18)	>500	—
3-Detigloylisoswietenine (19)	144.0^d	122.8–168.8
3-O-Benzoyl-3-detigloylisoswietenine (20)	3.9^c	1.7–9.0
3,6-O,O-Dibenzoyl-3-detigloylisoswietenine (21)	213.7^ad	172.5–264.8
3-O-Propionyl-3-detigloylisoswietenine (22)	60.0^bde	50.1–71.9
Demethyl-3-detigloyl-2-hydroxyisoswietenine (23)	>500	—
3-Detigloyl-2-hydroxyisoswietenine (24)	436.5^g	415.8–458.2
3,6-O,O-Dipropionyl-3-detigloyl-2-hydroxyisoswietenine (25)	336.9^h	291.4–389.6
3,6-O,O-Dibenzoyl-3-detigloyl-2-hydroxyisoswietenine (26)	210.1^d	179.3–246.1
3-Detigloylisoswietemahonin G (27)	>500	—

*Values followed by the same letter are not significantly different from each other based on Tukey–Kramer multiple comparisons test (P = 0.05).

All swietenine (3) derivatives except demethyl-3-detigloylisoswietenine (18) were more toxic than 3, which was not toxic up to 500 ppm (Table 1). The introduction of a benzoyl group to the swietenine skeleton to form 6-O-benzoylswietenine (17) increased toxicity significantly to LC₅₀ 7.5 ppm. The compound 3-O-benzoyl-3-detigloylisoswietenine (20) was the most toxic with an LC₅₀ of 3.9 ppm although this was not significantly more toxic than either compound 7 or 17. The intermediates in the preparation of 20 were the hydrolysis and methylation compounds demethyl-3-detigloylisoswietenine (18) and 3-detigloylisoswietenine (19). Compound 18 was not toxic up to 500 ppm, whereas 19 gave an LC₅₀ 144.0 ppm (Table 1).

Compound 5 was not toxic up to 500 ppm and its derivatives generally showed low levels of toxicity. The most toxic derivative was 3,6-O,O-dibenzoyl-3-detigloyl-2-hydroxy isoswietenine (26) with an LC₅₀ of 210.1 ppm (Table 1).

Hydrolysis followed by methylation of 4 (LC₅₀ 220.1 ppm) resulted in 3-detigloylisoswietemahonin G (27), which was not toxic up to 500 ppm (Table 1).

Discussion

Among all the compounds investigated, 3-O-benzoyl-3-detigloylisoswietenine (20) had the lowest LC₅₀ value (3.9 ppm) and was therefore the most toxic compound. Other compounds that showed high toxicity were 6-O-benzoylswietenolide (7) (LC₅₀ 4.3 ppm), 6-O-benzoylswietenine (17) (LC₅₀ 7.5 ppm) and 3,6-O,O-dipropionylswietenolide (9) (LC₅₀ 28.5 ppm). The demethylated compounds 16, 18, and 23 were not toxic up to 500 ppm, similar to their respective starting materials.

Addition of acyl groups, in particular benzoyl groups, generally resulted in higher toxicity than the starting materials with the exception of 3,6-O,O-di-tert-butylacetylswietenolide (15) and 3,6-O,O-diacetylswietenolide (2).

Conclusions

Twenty-seven compounds including natural limonoids were studied in A. salina lethality assays. Most of the semisynthesized analogs showed increased toxicity over their parent compounds. Despite the very high toxicity values in some of the compounds tested against A. salina, the results may indicate that some limonoids are suitable candidates for studies in cytotoxic assays, which will lead to preliminary understanding of the medicinal value of these compounds.

Acknowledgement

We thank Marie Mohammed and Anissa Pierre for assistance with this project.

Declaration of interest

Financial support for this work was provided by the University of the West Indies, St. Augustine Campus.