Coumarin Derivatives from Esenbeckia grandiflora. and its Larvicidal Activity Against Aedes aegypti.

Abstract

From a hexane fraction prepared from the roots of Esenbeckia grandiflora. Mart., effective against fourth instar larvae of Aedes aegypti. L. (LC₅₀ 108.23 ppm), eight coumarin derivatives [daphnetin 7-methyl-8-(3,3-dimethylallyl) ether, a mixture containing pimpinellin eight one of its photodimerization product, 5-senecioyl-xanthotoxin, 3-(1′,1′)-dimethylallylcolumbianetin, swietenocoumarin B, isopimpinellin, xanthotoxin, and 5-(1′-hydroxyisopentenyl)bergapten] were isolated in addition to sitosterol and sitostenone. The mixture containing pimpinellin and its photodimerization product was effective against fourth instar larvae (LC₅₀ at 45.77 ppm), and a mixture of pimpinellin and swietenocoumarin B showed a LC₅₀ value of 62.23 ppm. All isolated compounds were identified on the basis of the spectral data including 2D experiments. Furthermore, previously unreported spectral data for some compounds are given.

Keywords:

• Aedes aegypti
• coumarin derivatives
• Esenbeckia grandiflora
• larvicidal activity

Introduction

Mosquitoes of several genera transmit a wide range of tropical diseases. The search for new methods to destroy the vector of these diseases is of major importance, as this is one of the strategies of the World Health Organization (WHO) in combating tropical diseases. Interest in Aedes aegypti. L. (Culicidae) lies in the fact that it acts as a vector for an arbovirus responsible for yellow fever in Central and South America and in West Africa and also as a vector of dengue hemorragic fever, which is endemic to Southeast Asia, the Pacific Islands area, Africa, and the Americas (Ciccia et al., 2000; Mongelli et al., 2002). Although yellow fever has been reasonably brought under control with a vaccine, no vaccine is available for dengue. Therefore, the only way of decreasing the incidence of this disease is the eradication of A. aegypti.. For this reason, we have included in our routine screening of crude plant extracts and isolated compounds the evaluation of larvicidal activity against A. aegypti..

As part of our work on isolation and structural identification of constituents of plants with larvicidal activity, we describe the isolation of nine coumarin derivatives [daphnetin 7-methyl-8-(3,3-dimethylallyl) ether (1), pimpinellin (2) and its photodimerization product (3), swietenocoumarin B (4), isopimpinellin (6), xanthotoxin (7), 5-senecioylxanthotoxin (8), 3-(1′,1′)-dimethylallylcolumbianetin (9), and 5-(1′-hydroxy-isopentenyl)-bergapten (10) (Fig. 1)], in addition to two phytosteroids, sitosterol (5) and sitostenone (11), from the hexane fraction of Esenbeckia grandiflora. Mart. (Rutaceae) roots, active against fourth instar larvae of Aedes aegypti. (LC₅₀ 108.23 ppm).

Figure 1 Coumarin derivatives from E. grandiflora..

Materials and Methods

General experimental procedures

The ¹H (400 and 500 MHz) and ¹³C (100 and 125 MHz) spectral data of the compounds were obtained in CDCl₃ with the solvent as reference, employing Bruker Avance 400 (Chemistry Department of Queen Mary University, London, UK) and Bruker DRX-500 (CENAUREMN da Universidada Federal do Ceará, Brazil) spectrometers, respectively. IR spectra were recorded on a Perkin-Elmer FTIR 1750 spectrophotometer (Departamento de Química da Universidade Federal de Alagoas, Brazil). The column chromatography and thin-layer chromatography (TLC) were performed on silica gel 60 (70-230 mesh) and silica gel PF₂₅₄ (both Merck), respectively. Gel permeation chromatography was carried out with Sephadex LH-20 (Pharmacia Departamento de Química da Universidade Federal de Alagoas, Brazil).

Plant material

The roots of Esenbeckia grandiflora. Mart. were collected in November 1999 in Área de Proteção Ambiental de Santa Rita (Mucuri), Marechal Deodoro/AL, Brazil, and identified by Rosangela P. L. Lemos of Instituto do Meio Ambiente do Estado de Alagoas, Brazil, where a voucher specimen was deposited (MAC-10.754).

Extraction and isolation

The dried pulverized roots (2300 g) were extracted by maceration with 90% ethanol. After evaporation of solvent to dryness under vacuum, the crude extract obtained (46.2 g) was dissolved in methanol/water (3:2) and extracted successively with hexane, chloroform, and ethyl acetate. After removal of solvents, crude extract and fractions of hexane, chloroform, ethyl acetate, and methanol/water were taken for larvicidal assays on fourth larval instar of A. aegypti..

Hexane fraction (12.79 g), effective against fourth larval instar of A. aegypti. (LC₅₀ of 108.23 ppm), was chromatographed on a silica gel column using benzene containing increasing amounts of EtOAc as eluent. TLC monitored the fractions, and the chromatographically identical fractions were combined. A (1.18 g) and B (1.60 g) fractions after successive chromatographic fractionation on silica gel column with hexane containing increasing amounts of EtOAc, preparative TLC [silica gel, C₆H₁₄/EtOAc/AcOH (95:5:0.1) for fraction A and (9:1:0.1) and (97:3:0.1) for fraction B], and recrystallization from MeOH afforded 1 (22 mg) from fraction A, and a mixture containing 2 and 3 (32 mg), 4 (18 mg), and sitosterol 5 (73 mg) from fraction B. The remaining fractions [C (1.34 g) and D (0.42 g)] after chromatographic fractionation on silica gel column with hexane containing increasing amounts of EtOAc, gel permeation (Sephadex LH-20, MeOH), and recrystallization from MeOH yielded a mixture containing 6 and 7 (30 mg) and 10 (15 mg) from fraction C and substances 8 (15 mg), 9 (9 mg), and 11 (33 mg) from fraction D. Only the previously unreported spectral data for compounds 1, 3, 4, and 10 are given.

Daphnetin 7-methyl-8-(3,3-dimethylallyl)ether (1)

¹³C NMR (125 MHz, CDCl₃): δ 161.85 (C-2), 114.48 (C-3), 144.35 (C-4), 120.33 (C-4a), 125.87 (C-5), 108.12 (C-6), 146.62 (C-7), 138.87 (C-8), 139.16 (C-8a), 72.31 (C-1′), 123.49 (C-2′), 132.34 (C-3′), 27.32 (C-4′), 18.25 (C-5′), 56.25 (MeO-7).

Pimpinellin dimer (3)

IR (KBr) cm⁻¹: 2835, 1732, 1623, 1510, 1480, 1385, 1335, 1117, 1059, 819, 745. ¹H NMR (500 MHz, CDCl₃): δ 4.05 (d, J. = 8 Hz, H-3), 4.41 (d, J. = 8 Hz, H-4), 6.76 (d, J. = 2.1 Hz, H-3′), 7.48 (d, J. = 2.1 Hz, H-4′). ¹³C NMR (125 MHz, CDCl₃): δ 38.64 (C-3), 40.21 (C-4), 104.37 (C-3′), 106.83 (C-4a), 114.21 (C-8), 134.96 (C-6), 139.93 (C-8a), 144.97 (C-2′), 147.94 (C-5), 148.59 (C-7), 165.05 (C-2), 61.15 (MeO-6), 61.17 (MeO-5).

Swietenocoumarin B (4)

¹³C NMR (125 MHz, CDCl₃): δ 160.85 (C-2), 114.38 (C-3), 141.55 (C-4), 125.99* (C-4a), 131.58 (C-5), 126.05* (C-6), 147.93 (C-7), 114.12 (C-8), 144.43 (C-8a), 146.42 (C-2′), 106.12 (C-3′), 28.51 (C-1″), 122.49 (C-2″), 133.29 (C-3″), 25.97 (C-4″), 18.55 (C-5″), 61.83 (MeO-5). HMQC correlations: [δ 6.38 (114.38), 7.98 (141.55), 7.68 (146.42), 6.86 (106.12), 5.14 (122.49), 4.25 (61.83), 3.71 (28.51), 1.86 (18.55), 1.72 (25.97)]; main HMBC correlations: [H-3 (C-2); H-4 (C-2, C-3, C-8a, C-5, and C-4a), H-2′ (C-6, C-7, C-8, and C-3′), H-3′ (C-6 and C-2′), H-1″ (C-8 and C-2″), H-2″ (C-1″, C-3″, C-4″, and C-5″), MeO-5 (C-5)]. (Note:. *may be interchanged.)

5-(1′-Hydroxy-isopentenyl)bergapten (10)

IR (KBr) cm⁻¹: 3375, 2835, 1727, 1623, 1512, 1483, 1388, 1334, 1119, 1060, 819, 745. ¹H NMR (500 MHz, CDCl₃): δ 6.44 (H-3, d, J. = 9.9 Hz), 8.31 (H-4, d, J. = 9.9 Hz), 7.75 (H-2′, d, J. = 2.2 Hz), 6.98 (H-3′, d, J. = 2.2 Hz), 4.19 (H-1′, d, J. = 3.8 Hz), 6.53 (H-2″, m.), 2.08, 2.32 (H-4″ and H-5″, s. each), 4.39 (MeO-5, s.).

Larvicidal assays

Larvicidal assays with substances 2 and 3, 2 and 4 as mixtures and 10 were evaluated as recommended by WHO (1975) with some modification. The larvae were maintained in laboratory conditions of a temperature controlled at 27 ± 4° C with relative humidity at 80 ± 4°C. Substances were prepared dissolving a weighed amount in aqueous dimethylsulfoxide as a 1% solution. Three replicates of 20 larvae were used along with controls that were placed by a 1% aqueous solution of dimethylsulfoxide. Mortality was determined after exposure of larvae at concentrations 10, 30, 60, and 90 ppm at time intervals ranging from 1 to 48 h. The LC₅₀ was determined using linear regression analysis.

Results and Discussion

Compounds 1 (Barua et al., 1980; Borges-Del-Castillo et al., 1984), 2 (Steck & Mazurek, 1972; Murray & Jorge, 1984), 4 (Bhide et al., 1977), 5 (Macari et al., 1990), 11 (Wehrli & Nishida, 1979; Macari et al., 1990), 6 (Steck & Mazurek, 1972; Guilhon et al., 1994), 7 (Elgamal et al., 1979), 8 (Trani et al., 2004), 9 (Oliveira et al., 1996), and 10 (Trani et al., 2004; Torres et al., 2004) were identified on the basis of their spectral data and by direct comparison with reported data of the analogous or model compounds.

Compound 3 was isolated in mixture with 2 (1:2) and identified on the basis of their spectral data and by comparison with data of models in the literature (Zdero et al., 1990; Rojas-Lima et al., 1999). The IR spectrum showed absorptions at 1732 (C = O), 1623, 1510, and 1480 (C = C), 2850 and 1385 (OMe) cm⁻¹. The ¹H NMR spectrum showed doublets at δ 7.48 and 6.76 (J. = 2.1 Hz) corresponding to a furane unit, singlets at δ 3.86 and 3.70 for two methoxyl groups, and a pair of doublets at δ 4.42 and 4.06 (J. = 8.0 Hz) for two methine hydrogens. Thus, a dimer of pimpinellin (2) was proposed. Further evidence for the dimeric structure was the absence of the hydrogens coumarin double bond present in 2 [δ 8.09 and δ 6.38 (d, J. = 9.7 Hz). The ¹³C NMR spectra (BB and DEPT, 125 MHz) showed only signals for four methine, seven nonprotonated and two methyl carbons. Among which, signals for a carbonyl group (δ 165.05), two saturated carbons (δ 40.21 and 38.64), and two methoxyl groups (δ 61.17 and 61.15) were observed. 2D NMR experiments supported all the attributions of the ¹H and ¹³C chemical shifts. In the Correlation Spectroscopy (COSY) spectrum, the signal at δ 7.48 (H-2′) correlates with δ 6.76 (H-3′), and δ 4.05 (H-3) with δ 4.41 (H-4). In the Hydrogen Multiple Quantum Connectivity (HMQC), these hydrogens correlates with δ 144.97, δ 104.37, δ 38.64 and δ 40.21, respectively. The Hydrogen Multiple Bond Connectivity (HMBC) spectrum shows that the signal of H-4 (δ 4.41) correlates with C-2 (δ 165.05), C-3 (δ 38.64), C-5 (δ 147.94), C-4a (δ 106.83), and C-8a (δ 139.93), while H-3 (δ 4.05) correlates with C-2 (δ 165.05), C-4 (δ 40.21), and C-4a (δ 106.83). Dimeric coumarins, psoralen (Rojas-Lima et al., 1990) and umbelliferone-3,3-dimethylallyl ether dimers (Zdero et al., 1990), have previously been reported from natural source (Zdero et al., 1990; Rojas-Lima et al., 1999).

As photodimerization of coumarins can lead to different cyclobutane derivatives (Krauch et al., 1966; Zdero et al., 1990), unfortunately the stereochemistry of the dimer could not be deduced directly from NMR spectral data. However, psoralen and bergapten dimers with head-to-head cis-syn. stereochemistry have been obtained by UV irradiation of the corresponding furocoumarins as the major product (Rojas-Lima et al., 1999).

Among substances submitted to larvicidal assays, the most active was the mixture containing pimpinellin (2) and its photodimerization product (3) (LC₅₀ 45.77 ppm). The mixture containing pimpinellin (2) and swietenocoumarin B (4) showed moderate activity (LC₅₀ 62.23 ppm). Plants may be an alternative source of mosquito larval control agents because their constituents are a rich source of bioactive chemicals (Sharma et al., 1998; Ciccia et al., 2000; Siddiqui et al., 2000). Plants of the family Rutaceae have drawn attention because they already contain larvicidal principle (Rahuman et al., 2000). This is the first time that Esenbeckia. species have been screened against A. aegypti..

Acknowledgments

The authors are grateful to FAPEAL, CNPq, and IMSEAR-MCT for financial support and scholarships and to Dr. Edilberto R. Silveira (CENAUREMN da Universidade Federal do Ceará) and Dr. Edson de Souza Bento for NMR spectra.