This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy
Phytochemistry 71 (2010) 294–300
Contents lists available at ScienceDirect
Phytochemistry journal homepage: www.elsevier.com/locate/phytochem
Flavones and phenylpropanoids from a sedative extract of Lantana trifolia L. Lisieux de Santana Julião a, Suzana Guimarães Leitão b,*, Cinzia Lotti c, Anna Lisa Picinelli c, Luca Rastrelli c, Patricia D. Fernandes d, François Noël d, Jean-Pierre Barros Thibaut d, Gilda Guimarães Leitão e a
Universidade Federal do Rio de Janeiro, Programa de Biotecnologia Vegetal, CCS, Bl. K, Rio de Janeiro, RJ, Brazil Universidade Federal do Rio de Janeiro, Faculdade de Farmácia, CCS, Bl. A, Ilha do Fundão, 21.941-590 Rio de Janeiro, RJ, Brazil c Dipartimento di Scienze Farmaceutiche, Università degli Studi di Salerno, Via Ponte Don Melillo, 84084 Fisciano (SA), Italy d Universidade Federal do Rio de Janeiro, Instituto de Ciências Biomédicas, CCS, Rio de Janeiro, Brazil e Universidade Federal do Rio de Janeiro, Núcleo de Pesquisas de Produtos Naturais, CCS, Bl. H, Rio de Janeiro, Brazil b
a r t i c l e
i n f o
Article history: Received 2 April 2008 Received in revised form 10 July 2009 Available online 16 November 2009 Keywords: Lantana trifolia Verbenaceae HSCCC Scutellarein-7-O-b-D-apiofuranoside Apigenin-7-O-b-D-apiofuranosyl-(1 ? 2)-bD-apiofuranoside Celtidifoline 1D and 2D NMR spectroscopy Sedative activity Benzodiazepine receptor
a b s t r a c t The flavone glycosides, named scutellarein-7-O-b-D-apiofuranoside and apigenin-7-O-b-D-apiofuranosyl(1 ? 2)-b-D-apiofuranoside, and the flavone celtidifoline (5,6,40 ,50 -tetrahydroxy-7,30 -dimethoxyflavone), along with other 11 known compounds, were isolated from leaves of the ethyl acetate extract of Lantana trifolia L. using step gradient High Speed Countercurrent Chromatography (HSCCC) and High Performance Liquid Chromatography (HPLC), respectively. Their structures were elucidated by spectroscopic methods, including 2D NMR and mass spectrometry (ESI-MS) techniques. The ethanolic and ethyl acetate extracts produced an intense sedative effect in mice, one hour after oral administration of 1 mg/kg. This effect was neither due to a benzodiazepine-like effect of the three flavone derivatives neither of the phenylpropanoids, betonyoside F and verbascoside, that were tested for their affinity for the [3H] flunitrazepam binding sites. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Lantana is a neotropical genus with approximately 150 species, many of them occurring in Brazil. Phytochemical investigations have established the presence of phenylpropanoids, flavonoids and terpenoids, as the main class components with relevant biological activities (Ghisalberti, 2000). Lantana trifolia L. (syn. Lantana celtidifolia Kunth.) is a small shrub that occurs in all regions of Brazil and is extensively used in folk medicine in the form of infusions and syrups for the treatment of respiratory disorders and as sedative (Lorenzi, 2000; Correa, 1926). Few studies have been carried out on either the chemistry or pharmacology (Achola and Munenge, 1996; Silva et al., 2005; Katuura et al., 2007) of this species. Umuhengerin, a pentamethoxyflavone isolated from its leaves, exhibited antibacterial and anti-fungal activities (Rwangabo, 1988), whereas germacrene D and caryophyllene are described as major components of its essential oil (Muhayimana et al., 1998). Iridoid glycosides, which are frequently common in the Verbenaceae, were not detected by Rimpler and Sauerbier (1986). In the course of a preliminary pharmacological investigation of this plant * Corresponding author. Tel.: +55 21 25626413; fax: +55 21 25626425. E-mail address:
[email protected] (S.G. Leitão). 0031-9422/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2009.10.007
(Silva et al., 2005), we reported that the crude ethanolic extract presented anti-inflammatory, antinociceptive and sedative activity in mice. The ethyl acetate extract resulting from the liquid–liquid partition of the ethanolic extract showed a similar pharmacological profile (data not shown) and, therefore was chosen for study. Here we report the isolation, using a HSCCC step gradient elution approach combined with HPLC, and structural determination of two new flavone glycosides scutellarein-7-O-b-D-apiofuranoside (1), and apigenin-7-O-b-D-apiofuranosyl-(1 ? 2)-b-D-apiofuranoside (3) and one new flavone celtidifoline (5,6,40 ,50 -tetrahydroxy-7,30 dimethoxyflavone) (2), as well as eleven known compounds betonyoside F (4), verbascoside (5), vanillic acid (6), protocatechuic acid (7), 4-hydroxybenzoic acid (8), caffeic acid (9), coumaric acid (10), martynoside (11), scutellarein-7-O-b-D-glucopyranoside (12), sorbifolin (13) and samioside (14). We also performed experiments in order to assess the sedative effect and affinity for the benzodiazepine receptor of these three new flavones and of the phenylpropanoids betonyoside F and verbascoside. 2. Results and discussion The ethyl acetate extract of the L. trifolia was separated by highspeed counter-current chromatography (HSCCC) using a gradient
Author's personal copy
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
elution approach, with a biphasic solvent mixture composed of nhexane–ethyl acetate–n-butanol–water (0.4:1:X:1, v/v/v/v), where X = 0.1, 0.3, 0.5 and 0.7 (solvent systems A–D). The choice of an appropriate solvent system for a gradient elution in CCC can be a difficult task, in part because any change in the composition of one liquid phase of a biphasic liquid system may induce changes in the other liquid phase (Berthod et al., 2002). In a previous work, we used the gradient elution approach for separation of phenylpropanoids and iridoids by HSCCC, using a solvent system composed of ethyl acetate–n-butanol–water (Leitão et al., 2005). The same strategy of gradient elution was used here, with different ratios of ethyl acetate and n-butanol being tested, but the system proved too polar for the effective separation of compounds in Lantana extract. This could be due to the complexity of the extract and to the presence of phenylpropanoids and flavonoids instead of phenylpropanoids and iridoids. Therefore, hexane was introduced as a fourth solvent. The optimization of the number of steps of the gradient composed of hexane–ethyl acetate–n-butanol–water 0.4:1:X:1 (v:v:v:v) was done with a series of test tube experiments (see Section 4), where X varied from 0.1 to 0.7. The aim of these experiments (monitored by simple TLC experiments) is to set two extreme situations: one in which we have all (or at least the major part of) compounds in the upper phase, and the other where we have the opposite situation (all compounds in the lower phase). The majority of the compounds on the extracts was retained in the lower phase (aqueous) when 0.1 n-butanol was used, whereas a gradual inversion of the partition coefficients of the target compounds occurred when the n-butanol ratio was raised to 0.7. In the step gradient HSCCC separation, seven fractions were obtained. This fractionation afforded the isolation of the new flavone indentified as scutellarein-7-O-b-apiofuranoside (1) in fraction II and of betonyoside F (4) in fraction VII. Gradient elution was not effective for isolation of verbascoside (5), the major constituent of this extract. Therefore, fraction V was further purified by isocratic elution with the solvent system hexane–ethyl acetate–n-butanol–water, 0.4:1:0.6:1. The concentration of n-butanol chosen for the isocratic system was between the two steps of the gradients where the fraction was collected. This procedure allowed separation of verbascoside (5) from 1 in this fraction. This strategy is useful since all previous isolation and purification procedures described in the literature for these compounds involved more than one HSCCC passage (Leitão et al., 2005; Li et al., 2005). Fraction I was submitted to preparative HPLC to yield vanillic acid (6), protocatechuic acid (7), 4-hydroxybenzoic acid (8), caffeic acid (9), coumaric acid (10) and another new flavone named celtidifoline, 5,6,40 ,50 -tetrahydroxy-7,30 -dimethoxyflavone (2). Final purification of the other isolated compounds was achieved by column chromatography and HPLC (see Section 4). Fraction III afforded the new flavone apigenin-7-O-b-D-apiofuranosyl-(1 ? 2)-b-Dapiofuranoside (3), whereas fraction IV yielded martynoside (11), scutellarein-7-O-b-D-glucopyranoside (12) and sorbifolin (13), and fraction VI afforded samioside (14) (Fig. 1). The structures of new compounds 1, 2 and 3 were elucidated by the use of 600 MHz NMR techniques (DQF-COSY, HMBC, and HSQC), and ESI-mass spectrometry. The positive HR-ESI-MS of compound 1 showed a molecular ion peak at m/z 419.0897 [M+H]+, in accordance with an empirical molecular formula of C20H18O10. The negative ESI-MS spectrum showed a pseudomolecular ion peak at 417 [MH] indicating a molecular weight of 418 for compound 1. ESI-MS-MS experiments showed further fragment ions at m/z 285 [MH132], suggesting the presence of a pentosyl moiety (C5H7O4) and at m/z 168 and m/z 118 ascribable to the retro-Diels–Alder fragments of ring A and B of a flavone skeleton. The 1H NMR spectrum of 1 (Table 1) showed two doublets at d 7.95 (2H, J = 8.8 Hz) and d 6.90 (2H, J = 8.8 Hz), due to the B ring of a
295
40 -oxygenated flavonoid. Two singlets at d 6.88 and d 6.80 were assigned to the H-8 proton of the A ring and H-3 proton of the C ring, respectively. The 13C NMR spectrum displayed 20 carbons (Table 1), whereas the HSQC spectrum gave a correlation between H-8 with the signal at d 94.62 (C-8). The HMBC spectrum gave correlations between H-8 and C-6, C-7 and C-10. These set of experiments led to identification of scutellarein as the aglycone (Xia et al., 2007). The apiose moiety was confirmed by presence of an anomeric proton at d 5.63 (1H, J = 3.2 Hz), as well as from four other signals between d 4.28 and d 3.42 in the 1H NMR spectrum. The DEPT spectrum showed signals for CH2 at d 76.56 and 62.84 and lack of any resonance at d 79.28 in agreement with values for C4, C-5 and C-3 of a b-apiosyl unit (Table 1). The configuration of the b-apiofuranosyl moiety was assigned after hydrolysis of 1 with 1 N HCl. The hydrolysate was trimethylsilylated, and GC retention times of each sugar were compared with those of authentic samples prepared in the same manner. In this way, the sugar unit of 1 was determined to be D-apiose. The apiofuranosyl ring configuration was also confirmed by comparing 1H–1H scalar coupling constants with those reported for methyl apiofuranosides and DLapioses and by NOE observations (Ishii and Yanagisawa, 1998). The 2D ROESY spectrum of 1 showed cross-peaks between H-2 and the protons of the hydroxymethyl group, and H-2 and H-4b, indicating that H-2, the hydroxymethyl group, and H-4b are found on the same face of the ring for this sugar and confirming its structure as b-D-apiose. The HMBC spectrum confirmed the location of the apiose at C-7 by showing connectivity between the anomeric hydrogen of apiose and the carbon at d 151.89 (C-7). The HSQC spectrum furnished all the direct correlations between protons and carbons (Table 1). Comparison with literature data also confirmed the 13C values for a scutellarein-7-O-glycoside (Emam et al., 1998). Thus, the structure of 1 could be deduced as being the new flavone glycoside scutellarein-7-O-b-D-apiofuranoside (Fig. 1). The HR-ESI-MS (positive-ion mode) of compound 2 exhibited a pseudomolecular ion peak at m/z 369.0597 [M+Na]+, ascribable to a molecular formula of C17H14O8. The ESI-MS experiment gave (negative-ion mode) a quasi-molecular ion peak [MH] at m/z 345 indicating a molecular weight of 346 for 2. Further fragment ion peaks in the ESI-MS-MS spectrum were observed at m/z 330 [MH15] and m/z 315 [MH30] corresponding to successive loss of two CH3 groups. The fragment ions at m/z 182 (C8H6O5) and at m/z 164 (C9H8O3) were the retro-Diels–Alder fragment of rings A and B, respectively, and indicated the presence of a methoxyl group in both rings A and B of a flavone skeleton. The 1H NMR spectrum of 2 (Table 1) had a signal at d 7.14 (2H), attributed to two-non equivalent protons (H-20 , H-60 ) of the B ring indicating oxygenation at C-30 , C-40 , e C-50 . Two singlets at d 6.86 (1H) and at d 6.65, were assigned to the H-8 and H-3 protons, respectively. These data as well as the intense signals at d 3.95 and 4.05 (both 3H, s), relative to two OCH3 groups, suggested presence of a tetrahydroxyflavone with two additional methoxyl group substitutions (Marco et al., 1988). The 13C NMR spectrum of 2 had 17 carbons (Table 1) with typical C-5, C-6 and C-7 values for a trioxygenated A ring (Vilegas et al., 1999). In the HMBC spectrum, cross-peaks disclosing the bonding site of each methoxyl were observed: dH 3.95 correlated with dC 149.52 (C-30 ), and dH 4.05 correlated with dC 155.72 (C-7). Correlations were also observed between H-3/C-10 and C-10 , H-8/C-6, C-7 and C-10, H-20 and H-60 / C-2, C-20 , C-40 , C-60 . Consequently, 2 was determined to be the new 5,6,40 ,50 -tetrahydroxy-7,30 -dimethoxyflavone, named celtidifoline (Fig. 1). The HR-ESI-MS (positive-ion mode) of compound 3 exhibited a pseudomolecular ion peak at m/z 535.1462 [M+H]+, ascribable to a molecular formula of C25H26O13. The ESI-MS experiment gave (negative-ion mode) a quasi-molecular ion peak at m/z 533 [MH]
Author's personal copy
296
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
Fig. 1. Structures of the major compounds from Lantana trifolia.
Table 1 1 H NMR and
13
C NMR spectroscopic data of compounds 1, 2 and 3 in CD3OD at 600 and at 150 MHz.
1
2 1
H
2 3 4 5 6 7 8 9 10 10 20 30 40 50 60
Apiose 7 10 0 20 0 30 0 40 0 50 0
6.8 (s)
6.8 (s)
7.9 (d, 8.8) 6.9 (d, 8.8) 6.9 (d, 8.8) 7.9 (d, 8.8)
5.6 (d, 3.2) 4.3 (d, 3.2) 3.7, 4.0 (d, 9.6) 3.4 (m)
13
164.5 102.8 182.7 147.2 131.1 151.9 94.6 149.4 105.9 121.7 128.9 116.4 161.6 116.4 128.9
3 1
C
H
2 3 4 5 6 7 8 9 10 10 20 30 40 50 60 OCH3 (30 ) OCH3 (7)
6.6 (s)
6.8 (s)
7.1 (s)
7.1 (s) 3.9 4.0
108.08 76.56 79.28 75.16 62.84
Chemical shifts are in ppm from TMS and J values in Hz are presented in parentheses. All signals were assigned by 1D-TOCSY, DQF-COSY, HSQC and HMBC experiments. a Interchangeable values.
13
1
C
166.5 104.0 184.1 149.7 131.4 155.7 91.89 152.1 106.5 122.4 103.1 149.5 139.1 147.1 108.8 56.9 57.1
H
2 3 4 5 6 7 8 9 10 10 20 30 40 50 60
Apiose 7 10 0 20 0 30 0 40 0 50 0 Apiose 20 0 10 0 0 20 0 0 30 0 0 40 0 0 50 0 0
6.6 (s)
6.4 (d, 1.5) 6.7 (d, 1.5)
7.8 (d, 8.3) 6.8 (d, 8.3) 6.8 (d, 8.3) 7.8 (d, 8.3)
5.8 (d, 2.6) 4.3 (d, 2.6) 4.0, 4.1 (d, 9.6) 3.6, 3.7a (m) 5.2 (d, 2.6) 4.1 (d, 2.6) 3.8, 3.9a (d, 10.1) 3.6, 3.7a (m)
13
C
167.4 103.3 184.4 163.4 101.5 164.7 96.2 159.2 107.5 122.9 129.9 118.1 164.7 118.1 129.9
107.4 84.18 83.01 75.41 64.51 111.11 78.13 81.5 75.45 64.03
Author's personal copy
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
indicating a molecular weight of 534 for compound 3. Further fragment ion peaks in the ESI-MS-MS spectrum were observed at m/z 401 [MH132] and m/z 269 [MH132132] corresponding to the successive loss of two pentosyl moieties. The 1H NMR spectrum of 3 (Table 1) showed signals attributable to an apigenin skeleton: d 6.43 (1H, d, J = 1.5 Hz, H-6), d 6.57 (1H, s, H-3), d 6.76 (1H, d, J = 1.5 Hz, H-8), d 6.85 (2H, d, J = 8.3 Hz, H-30 and H-50 ), d 7.80 (2H, d, J = 8.3 Hz, H-20 and H-60 ). The 13C NMR shifts of the aglycone part of 3 (Table 1) corresponded well with the shifts for apigenin, with the only significant difference being those corresponding to C-6, C7 and C-8. These shifts are analogous to those reported when the 7hydroxy group is glycosylated in a flavone glycoside (Agrawal, 1989). Two anomeric protons, assigned to the C-1 protons of two apiofuranosyl units, were easily identified in the spectra of 3. They resonated at d 5.21 (d, J = 2.63 Hz) and d 5.84 (d, J = 2.63 Hz), and they correlated to carbons at d 111.11 and d 107.4, respectively. From the assigned aglycon and sugar values, deduced from 1 H–1H COSY and HSQC experiments (Table 1), it was apparent that a disaccharide unit was attached to C-7 of the aglycone. The chemical shifts of all the individual protons of the two sugar units were attributed on the basis of 1D-TOCSY and DQF-COSY spectroscopic analysis, and the 13C chemical shifts of their relative attached carbons were clearly assigned from the HSQC spectrum (Table 1). These data showed the presence of a terminal b-D-apiofuranosyl (d 5.21) and a 2-substituted b-apiofuranosyl (d 5.84) as indicated by the downfield shift of its C-2 (d 84.18) signal. An unambiguous determination of the linkage site was obtained from the HMBC spectrum, which showed key correlation peaks between the anomeric proton of the inner apiose (d 5.84) and the C-7 of the apigenin (d 164.68), and between the anomeric proton of the outer apiose (d 5.21) and the C-2 of the inner apiofuranosyl unit (d 84.18). The b-configuration at the anomeric position for the apiofuranosyl units was determined from their coupling constants and by ROE observations as described for compound 1. Also in this case, the D configuration of both apiose units was determined by acid hydrolysis of 3 followed by GC analysis. Therefore, the structure of 3 was determined as the new flavone glycoside apigenin7-O-b-D-apiofuranosyl-(1 ? 2)-b-D-apiofuranoside, (Fig. 1). The other major compounds from L. trifolia were identified as phenolic acids vanillic acid (6), protocatechuic acid (7), 4-hydroxybenzoic acid (8), caffeic acid (9), coumaric acid (10), phenylpropanoid glycosides betonyoside F (4), verbascoside (5) (Wu et al., 2004), martynoside (11) and samioside (14), and 6-hydroxy flavones scutellarein-7-O-b-D-glucopyranoside (12) (Emam et al., 1998) and sorbifolin (13) (Alam et al., 1986) by 1H and 13C NMR spectroscopic analyses and comparison of data with those in the literature. Betonyoside F (4) and samioside (14) were reported, respectively, from aerial parts of Stachys officinalis (Miyase et al., 1996) and Phlomis samia (Kyriakopoulou et al., 2001), two genus of the family Lamiaceae (Labiatae), which is very close to the Verbenaceae. The occurrence of 6-hydroxy and methoxy flavones in L. trifolia is of taxonomic interest because they are relatively uncommon flavonoids in Angiosperms, and it has been reported that they occur in some abundance in those families where they are present (Harborne and Baxter, 1999). The sedative properties of L. trifolia extracts were evaluated by the open field method (Whimbey and Denenberg, 1967). One hour after their administration, the LA and LE extracts produced an intense sedative effect in all animals, reported as reduction of walked squares, with similar pattern and intensity at 1 and 10 mg/kg (Table 2), indicating that the maximal effect had probably been attained. It was noteworthy that the effect increased with time, at least during the two first hours after administration. In order to identify a possible mechanism of action and/or the active substances in the extract, we tested the isolated compounds 1–5 in a competition assay in rat brain, using [3H] flunit-
297
razepan as a classical marker of the benzodiazepine receptor. The most active compound was the new scutellarein-7-O-b-Dapiofuranoside (1) with an half maximal inhibitory concentration (IC50) of 187 lM (Table 3). This value is very high when compared with classical benzodiazepines, such as clonazepam, flunitrazepam and diazepam, which exhibit IC50 of 1, 3 and 11 nM, respectively, when tested in the same conditions (Noël et al., 2007). These results indicate that compounds 1–5 have a low affinity for the benzodiazepine site present in the GABAA receptor complex, as already reported for other flavonoids (Hennebelle et al., 2007; Fernández et al., 2006; Coleta et al., 2008). These authors suggest that flavonoids could induce CNS effects via different mechanisms, such as inhibition of the NMDA (Losi et al., 2004) or 5-HT (Coleta et al., 2008; Marder et al., 2003) receptors. The type of sugar linkage with the aglycone should be an important factor for the sedative activity of these flavonoids (Fernández et al., 2006). On the other hand, the glycosidic part of the flavonoid derivatives could cause a steric hindrance responsible for the weakness of the binding to the benzodiazepine receptor (Fernández et al., 2006). With respect to the methoxyflavones, anxiolytic and sedative effects have been demonstrated by different authors (Huen et al., 2003; Hui et al., 2002; Marder et al., 2003) who showed that, for this class of compounds, a 20 -hydroxyl group enhanced the interaction between the ligand and the benzodiazepine receptor (Huen et al., 2003). Marder et al. (2001) also demonstrated that carbons 30 and 6 are the most effective positions to place substituents that enhance the affinity for the receptor, while other pattern of substitution do not have the same effect. As far as phenylpropanoids usually found in Verbenaceae are concerned, verbascoside, 5, has already been tested on benzodiazepine receptors (Daels-Rakotoarison et al., 2000) and its IC50 (1.9 mM) was compatible with the lack of effect at 300 lM, observed in the present study (Table 3). For the other phenylpropanoid tested here (compound 4) there is no previous data in the literature. Although verbascoside had no affinity for the [3H] flunitrazepam binding sites, its effectiveness in the open field test (Table 2) indicates that this compound could be one of the active principles of the extract.
3. Concluding remarks In the present study, phytochemical investigation of the ethyl acetate extract of L. trifolia led to the isolation and identification of flavonoids and phenylpropanoids, which are compounds reported to have sedative properties (Losi et al., 2004; Marder et al., 2003; Zétola et al., 2002; Viola et al., 1995). Flavonoids have already been reported to be efficacious at different receptor systems in the CNS (Hui et al., 2002). They are a relatively new class of ligands of the benzodiazepine site of the GABAA receptors, and their glycosides are considered to form the newest group within the growing family of flavonoids with CNS activity (Fernández et al., 2006). Fernández et al. (2006) recently demonstrated that a series of flavonoid glycosides were depressant of the CNS in several in vivo assays (including locomotor activity measured in the open field method) but had low affinity for the [3H] flunitrazepan binding sites. The authors suggested that the flavonoid glycosides are easily metabolized by the organism and that secondary metabolites could activate the GABAA receptors and mediate the sedative effects. As a conclusion, we showed that the sedative effect of L. trifolia extracts demonstrated in mice cannot be attributed to the direct activation of the central benzodiazepine site by the new flavones 1–3 and the phenylpropanoids betonyoside F, 4 and verbascoside, 5.
Author's personal copy
298
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
Table 2 Reduction of spontaneous locomotor activity by Lantana trifolia extracts and verbascoside, 5, in the open field test. Values are expressed as mean ± S.D. Extract
Dose (mg/kg)
Hour after administration
Reduction of walked squares (%)
LE
1
1st 2nd 1st 2nd
33.4 ± 4.6* 61.3 ± 5.8* 38.6 ± 9.8* 62.6 ± 6.9*
1st 2nd 1st 2nd
51.1 ± 6.1* 67.4 ± 6.8* 59.9 ± 6.5* 69.2 ± 3.5*
1st 2nd 1st 2nd
38.8 ± 5.9* 86.7 ± 7.1* 44.7 ± 5.2* 73.9 ± 6.8*
10 LA
1 10
Verbascoside, 5
1 3
LE, ethanolic extract; LA, ethyl acetate extract. * P < 0.05, ANOVA followed by Bonferroni0 s test.
Table 3 Inhibition of [3H] flunitrazepam binding in rat brain synaptosomes: inhibitory concentrations (IC50) of some major compounds of Lantana trifolia ethyl acetate extract. Compound
IC50 (lM)
Scutellarein-7-O-b-D-apiofuranoside, 1 Celtidifoline, 2 Apigenin-7-O-b-D-apiofuranosyl-(1 ? 2)-b-D-apiofuranoside, 3 Betonyoside F, 4 Verbascoside, 5
187 440 670 550 »300a
a No inhibition at 300 lM indicating that the IC50 should be much higher than this concentration.
4. Experimental
(Waters) instrument. HSCCC was performed on a P.C. Inc. (Potomac, MD) apparatus, equipped with an interchangeable multilayer triple coil of PTFE tubing, 1.68 mm internal diameter (15 + 80 + 240 mL). The HSCCC system was connected to a solvent pump Rainin SD-200 Dynamax, a manual injection valve Rheodyne 5020A with 5 mL loop, and a Dynamax FC-1 fraction collector. All separation were performed at 850 rpm. HPLC separations were carried out on a Waters 590 system equipped with a Waters R401 refractive index detector, a Phenomenex C-18 column, 10 lm (10 250 mm, flow rate 3 mL/min), U6K injector and on an Agilent 1100 series chromatograph, equipped with a G-1312 binary pump, a G-1328A Rheodyne injector, a G-1322A degasser, and a G-1315A photodiode array detector (Waters Corp., Milford, MA) using the same column (flow rate 2.5 mL/min. TLC analyses were performed with Macherey–Nagel precoated silica gel 60 F254 plates.
4.1. General Optical rotations were determined on a Jasco DIP-1000 polarimeter equipped with a sodium lamp (589 nm) and a 10 mm microcell. UV spectra were obtained with a Beckman DU 670 spectrophotometer in MeOH, (c = 1). A Bruker DRX-600 NMR spectrometer, operating at 599.19 MHz for 1H and at 150.86 MHz for 13C, using the UXNMR software package was used for NMR experiments; chemical shifts are expressed in d (parts per million) referring to the solvent peaks dH 3.34 and dC 49.0 for CD3OD, coupling constants, J, are in Hertz. Distortionless enhancement by polarization transfer (DEPT) 13C, 1H–1H double quantum filtered correlation spectroscopy (DQF-COSY), 1H–13C heteronuclear single quantum coherence (HSQC), heteronuclear multiple bond coherence (HMBC) and ROESY (Rotating-frame Overhauser Enhancement Spectroscopy) were obtained by employing the conventional pulse sequences. The selective excitation spectra, 1D-TOCSY, were acquired using waveform generator-based GAUSS shaped pulses, mixing time ranging from 100 to 120 ms and a MLEV-17 spin-lock field of 10 kHz preceded by a 2.5 ms trim pulse NMR experiments. Electrospray ionization mass spectrometry (ESI-MS) was performed using a Finnigan LC-Q Advantage instrument from Thermoquest (San Jose, CA) equipped with Excalibur software. Samples were dissolved in MeOH and infused in the ESI source by using a syringe pump; the flow rate was 3 ll/min. The capillary voltage was 5 V, the spray voltage 5 kV, and the tube lens offset 35 V. The capillary temperature was 220 °C and the data were acquired in the MS1 and MS/MS scanning modes. The scan range was m/z 200–1000 and for the MS/MS scanning mode, the percentage of collision energy was 30%. Exact masses were measured by an ESI/Q-TOF
4.2. Plant material and extracion Leaves of L. trifolia were collected in Mendes, Rio de Janeiro State, Brazil in January 2004, and identified by Prof. Dr. Fatima Regina G. Salimena, from the Universidade Federal de Juiz de Fora, Minas Gerais, Brazil, where a voucher specimen was deposited (L. trifolia L. CESJ 30801). Dried and pulverized leaves (900 g) were exhaustively extracted with EtOH at room temperature for one week, and the extract was concentrated under reduced pressure to afford a brown syrup (LE). This residue was partitioned between water and organic solvents of increasing polarities, to afford the new extracts: hexane, CH2Cl2, EtOAc (LA), and n-BuOH, in this order.
4.3. Test tube experiments for the choice of the solvent system Small amounts of the EtOAc extract of L. trifolia were dissolved in separate test tubes containing the solvent system EtOAc–nBuOH–H2O in the ratios 0.8:0.2:1; 0.7:0.3:1; 0.6:0.4:1; 0.5:0.5:1; 0.3:0.7:1 and 0.2:0.8:1. The test tubes were shaken and the compounds allowed to partition between the two liquid phases. Equal aliquots of each phase were spotted beside each other, separately, on silica gel TLC plates developed with the solvent system EtOAc– acetone–H2O 25:8:5 (organic phase). The results were visualized under UV light (365 nm). Another set of test tube experiments was performed by introducing hexane into the EtOAc–n-BuOH– H2O system. The resulting biphasic solvent mixture was hexane– EtOAc–n-BuOH–H2O, in the ratios 0.4:1:X:1 v:v:v:v (solvent systems A–D), where X = 0.1, 0.3, 0.5 and 0.7).
Author's personal copy
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
4.4. HSCCC separation by gradient elution Preparative fractionation of the EtOAc extract of L. trifolia was performed in the 310 mL coil (80 + 240 mL, medium + large coils) of the P.C. Inc. apparatus. Three grams of extract were dissolved in 15 mL of both phases of solvent system A and injected in the 310 mL coil (previously equilibrated with stationary and mobile phases of system A, VS = 277 mL, VM = 33 mL, SF = 89.35%). The aqueous phase was used as the stationary phase, while the organic phase was used as the mobile phase in the tail-to-head elution mode, at a flow rate of 3 mL/min. Fractions of 12 mL were collected, leading to final seven fractions as follows: solvent system A: tubes 1–27 (Fractions I and II); solvent system B: tubes 28–54 (Fractions II and III); solvent system C: tubes 55–81 (Fractions III–VI), and solvent system D: tubes 81–108 (Fractions VI and VII). At the end of the final step of the gradient, rotor rotation was stopped and the column content (organic and aqueous phases) was ‘‘washed-off”. This procedure led to purified compounds 1 in fraction II (58.7 mg, tubes 22–29) and 4 in fraction VII (118.2 mg, tubes 96–102).
299
see Table 1; (+)-HRESI-MS, m/z 419.0897 [M+H]+, calcd. for C20H19O10, 419.0978; ESI-MS in positive mode m/z 419 [M+H]+ and in negative mode m/z 417 [MH], 285 [MH132], 267 [MH18], 257 [MH28], 239 [MH46]. 4.8. 5,6,40 ,50 -Tetrahydroxy-7,30 -dimethoxyflavone (celtidifoline, 2) Amorphous powder; [a]D = 74.0 (MeOH; c 0.9), UV kmax (MeOH): 365, 285; for 1H NMR and 13C NMR (CD3OD, 600 MHz) see Table 1; (+)-HRESI-MS, m/z 369.0597 [M+Na]+, calcd. for C17H14O8 Na, 369.0586; ESI-MS in negative mode m/z 345 [MH], 330 [MH15], 315 [MH15]. 4.9. Apigenin-7-O–D-apiofuranosyl-(1 ? 2)–D-apiofuranoside (3) Amorphous powder: [a]D = 85.2 (c 1.0; MeOH); UV kmax (MeOH): 340, 270; for 1H NMR and 13C NMR (CD3OD, 600 MHz) see Table 1; (+)-HRESI-MS, m/z 535.1469 [M+H]+, calcd. for C25H27O13, 535.1452; ESI-MS in negative mode m/z 533 [MH], 401 [MH132], 269 [MH132].
4.5. HSCCC separation by isocratic elution
4.10. Acid hydrolysis of compounds 1 and 3
Fraction V (400 mg, tubes 74–79) from the preparative gradient elution (solvent C, X = 0.5 n-butanol) was further purified using an isocratic elution mode with the same solvent system in the ratios of 0.4:1:0.6:1. The fraction was dissolved in 5 mL of both phases of the solvent system and injected in the 80 mL coil. The aqueous phase was used as stationary while the organic phase was used as mobile (tail-to-head) at a flow rate of 2 mL/min. Fractions of 4 mL were collected, rotor rotation was stopped at tube 30 and the column content (organic and aqueous phases) was ‘‘washedoff” leading to final 70 tubes, combined by similarity after TLC analysis. These were obtained in total six fractions, with compounds 1 (27.8 mg, fraction 1, tubes 6–8) and 5 (418.5 mg, fraction 4, tubes 28–31) purified.
A solution of compounds 1 and 3 (2.0 mg each) in 1 N HCl (1 mL) was stirred at 80 °C in a stoppered reaction vial for 4 h. On cooling, the solution was evaporated under a stream of N2. Each residue was dissolved in 1-(trimethylsilyl)imidazole and pyridine (0.2 mL), and the solution was stirred at 60 °C for 5 min. After drying the solution with a stream of N2, the residue was separated by H2O and CHCl3 (1 mL, 1:1 v/v). The CHCl3 layer was analyzed by GC using an l-Chirasil-Val column (0.32 mm 25 m). Temperatures of the injector and detector were 200 °C for both. A temperature gradient system was used for the oven, starting at 100 °C for 1 min and increasing up to 180 °C at a rate of 5 °C/min. Peak of the hydrolysate of 1 and 3 was detected by comparison with retention times of authentic sample of D-apiose (Sigma–Aldrich, St. Louis, MO) after being treated with 1-(trimethylsilyl)imidazole in pyridine.
4.6. Isolation compounds 2–3, 6–14 4.11. Locomotor activity evaluation Fraction I from the gradient elution HSCCC (255 mg) was submitted to HPLC using MeOH ± H2O (38:62) as eluent to yield compounds 6 (10.7 mg, tR 5 min), 7 (2.6 mg, tR 7.5 min), 8 (2.5 mg, tR 8.5 min), 9 (2.2 mg, tR 9.75 min), 10 (5.5 mg, tR 12 min) and 2 (4.5 mg, tR 35 min). The fraction III (70 mg) from HSCCC was applied to a Sephadex LH-20 column (20 g), using MeOH as eluent at a flow rate 1 mL/min; 29 fractions of 3 mL was obtained. Fractions 17–18 afforded to pure compound 3 (3.7 mg) and fraction 19 compound 11 (5.8 mg). Fraction IV (185 mg) from HSCCC was purified by silica gel CC leading to 97 fractions of 8 mL (EtOAc:acetone:H2O, 25:8:2 from 1 to 54, 25:10:5 from 55 to 94 and MeOH from 95 to the end). Fractions were monitored by TLC (EtOAc:acetone:H2O 25:8:2) and combined by similarity. Fractions 15–31 (47 mg) containing a mixture of flavonoids and phenylpropanoids were submitted to HPLC using MeOH–H2O (70:30) as eluent to yield pure compounds 11 (1.7 mg, tR 6.55 min), 12 (9.4 mg, tR 5.48 min) and 13 (2.6 mg, tR 7.65 min). Fraction VI (350 mg) was subjected to Sephadex LH-20 CC (50 2.5 cm) using MeOH as mobile phase affording 80 fractions of 5 mL. Fractions 27–41 (110 mg) were further purified by gradient HPLC (flow rate 2.5 mL/min) of MeOH–H2O (from 40:60 to 48:52, stepwise) to yield compound 14 (5.9 mg, tR 36.32 min). 4.7. Scutellarein-7-O–D-apiofuranoside (1) Amorphous powder: [a]D = 71.2 (c 1.1; MeOH); UV kmax (MeOH): 335, 290; for 1H NMR and 13C NMR (CD3OD, 600 MHz),
4.11.1. Animals and extracts Male Swiss mice (18–20 g) were obtained from our own animal facility. Animals were maintained in a room with controlled temperature (22 ± 2 °C) for 12 h light/dark cycle with free access to food and water. Twelve hours before each experiment animals received only water, in order to avoid food interference with absorption of the compounds. Animal care and research protocols were in accordance with the principles and guidelines adopted by the Brazilian College of Animal Experimentation (COBEA) and approved by the Biomedical Science Institute/UFRJ – Ethical Committee for animal research. Extracts were dissolved in DMSO in order to prepare a stock solution at a concentration of 100 mg/ml and stored at 70 °C. On the day of the experiment the testing solutions were diluted in phosphate buffered solution (PBS). In all experiments, the final concentration of DMSO did not exceed 5% at which this solvent had no effect per se. 4.11.2. Open field test Spontaneous locomotor activity was determined according to Whimbey and Denenberg (1967). Briefly, the open field consisted of a box, which floor was divided into 50 squares. Animals received oral administration of extracts or verbascoside at doses of 1, 3, 10, or 30 mg/kg. After 1 and 2 h mice were placed individually at the centre of the box and allowed to freely explore the new environment. The number of squares by which each animal passed during 5 min was counted.
Author's personal copy
300
Lisieux de Santana Julião et al. / Phytochemistry 71 (2010) 294–300
4.12. [3H] flunitrazepam binding Brain hemispheres from albino Wistar rats were homogenized in 15 volumes 0.32 M sucrose buffered to pH 7.4 with 5 mM Tris–HCl. The homogenate was centrifuged at 1000gmax for 10 min. After washing of the pellet, the combined supernatants were centrifuged at 48,000gav for 20 min to obtain crude synaptosomes that were resuspended in physiological solution (NaCl 118 mM, KCl 4.8 mM, MgCl2 1.2 mM and CaCl2 1.2 mM) buffered to pH 7.4 with 15 mM Tris/HCl and stored at 80 °C. Synaptosomes (100–200 lg protein) were incubated at 4 °C in 1 mL physiological solution containing 0.1 nM [3H] flunitrazepam (85 Ci/mmol, New England Nuclear, USA). After 90 min, the samples were diluted with 3 mL of ice-cold buffer and rapidly filtered under vacuum on GF/C glass fiber filters (GMF 3, Filtrak, Germany). Filters were further washed twice with the same washing buffer. The nonspecific binding measured in the presence of 1 lM unlabeled flunitrazepam was less than 3% of the total binding (Noël et al., 2001). Acknowledgements This work was supported by CNPq (MCT – CNPq/MS-SCTIEDECIT. 410475/2006-8), FAPERJ and CAPES (fellowship). We are indebted to Dr. Fatima Regina Gonçalves Salimena, from Universidade Federal de Juiz de Fora, Minas Gerais, Brazil, for plant identification. Authors wish to thank Claudio Jose da Silva Ferreira for plant collecting. We are also indebted to Centro Nacional de Ressonância Magnética Nuclear Jiri Jones, UFRJ, Rio de Janeiro, for some of the NMR experiments. References Achola, K.J., Munenge, R.W., 1996. Pharmacological activities of Lantana trifolia on isolated guinea pig trachea and rat phrenic nerve diaphragm. Int. J. Pharmacogn. 34, 273–276. Agrawal, P.K., 1989. Carbon-13 NMR of Flavonoids, vol. 39. Elsevier Science Publisher BV, Amsterdam, The Netherlands. p. 564. Berthod, A., Brown, L., Leitão, G.G., Leitão, S.G., 2002. In: Barceló, D., Berthod, A. (Eds.), Countercurrent Chromatography. The Support Free Liquid Stationary Phase Comprehensive Analytical Chemistry, vol. XXXVIII. Elsevier, New York, p. 37. Coleta, M., Campos, M.G., Cotrim, M.D., de Lima, T.C.M., da Cunha, A.P., 2008. Assessment of luteolin (30 ,40 ,5,7-tetrahydroxyflavone) neuropharmacological activity. Behav. Brain Res. 189 (1), 75–82. Correa, M.P., 1926. Dicionário das Plantas Úteis do Brasil e das Exóticas Cultivadas. Reimpressão pelo Ministério da Agricultura (1984) 1, 747 (IBDF). Daels-Rakotoarison, D.A., Seidel, V., Gressier, B., 2000. Neurosedative and antioxidant activities of phenylpropanoids from Ballota nigra. Arzneim. Forsch. 50, 16–23. Emam, A.M., Elias, R., Moussa, A.M., Faure, R., Debrauwer, L., Balansard, G., 1998. Two flavonoids triglycosides from Buddleja madagascariensis. Phytochemistry 48, 739–742. Fernández, S.P., Wasowski, C., Loscalzo, L.M., 2006. Central nervous system depressant action of flavonoid glycosides. Eur. J. Pharmacol. 539, 168–176. Ghisalberti, E.L., 2000. Lantana camara L. (Verbenaceae). Fitoterapia 71, 467–486. Harborne, J.B., Baxter, H., 1999. The Handbook of Natural Flavonoids, vol. I. John Wiley and Sons Ltd., Baffins Cane, Chinchester, England. p. 889. Hennebelle, T., Sahpaz, S., Gressier, B., Joseph, H., Bailleul, F., 2007. Antioxidant and neurosedative properties of polyphenols and iridoids from Lippia alba. Phytother. Res. 22 (2), 256–258. Huen, M.S.Y., Hui, K.M., Leung, J.W.C., Sigel, E., Baur, R., Wong, J.T.F., Xue, H., 2003. Naturally occurring 20 -hydroxyl-substituted flavonoids as high-affinity benzodiazepine site ligands. Biochem. Pharmacol. 66, 2397–2407. Hui, K.M., Huen, M.S.Y., Leung, J.W.C., Wang, H.Y., Zheng, H., Sigel, E., Baur, R., Ren, H., Li, Z.W., Wong, J.T.F., Xue, H., 2002. Anxiolytic effect of wogonin, a
benzodiazepine receptor ligand isolated from Scutellaria baicalensis Georgi. Biochem. Pharmacol. 64, 1415–1424. Ishii, T., Yanagisawa, M., 1998. Synthesis, separation and NMR spectral analysis of methyl apiofuranosides. Carbohydr. Res. 313, 189–192. Katuura, E., Waako, P., Tabuti, J.R.S., Bukenya-Ziraba, R., Ogwal-Okeng, J., 2007. Antiplasmodial activity of extracts of selected medicinal plants used by local communities in western Uganda for treatment of malaria. African J. Ecol. 45 (Suppl. 3), 94–98. Kyriakopoulou, I., Magiatis, P., Skaltsounis, A.L., Aligiannis, N., Harvala, C., 2001. Samioside, a new phenylethanoid glycoside with free-radical scavenging and antimicrobial activities from Phlomis samia. J. Nat. Prod. 64, 1095– 1097. Leitão, G.G., de Souza, P.A., Moraes, A.A., Brown, L., 2005. Step-gradient CCC separation of phenylpropanoid and iridoid glycosides from roots of Stachytarpheta cayennensis (Rich.) vahl. J. Liq. Chromatogr. Relat. Technol. 28, 2053–2060. Li, L., Tsao, R., Liu, Z.Q., Liu, S.Y., Yang, R., Young, J.C., Zhu, H.H., Deng, Z.Y., Xie, M.Y., Fu, Z.H., 2005. Isolation and purification of acteoside and isoacteoside from Plantago psyllium L. by high-speed counter-current chromatography. J. Chromatogr. A 1063, 161–169. Lorenzi, H., 2000. Plantas Daninhas do Brasil, third ed. Instituto Plantarum de Estudos da Flora LTDA. p. 540. Losi, G., Puia, G., Garzon, G., 2004. Apigenin modulates GABAergic and glutamatergic transmission in cultured cortical neurons. Eur. J. Pharmacol. 502, 41–46. Marco, J.A., Barbera, O., Rodriguez, S.C., Adell, J., 1988. Flavonoids and other phenolics from Artemisia hispanica. Phytochemistry 27, 3155–3159. Marder, M., Estiu, G., Blanch, L.B., Viola, H., Wasowski, C., Medina, J.H., Paladini, A.C., 2001. Molecular modeling and QSAR analysis of the interaction of flavone derivatives with the benzodiazepine binding site of the GABA(A) receptor complex. Bioorg. Med. Chem. 9 (2), 323–335. Marder, M., Viola, H., Wasowski, C., Fernandez, S., Medina, J.H., Paladini, A.C., 2003. 6-Methylapigenin and hesperidin: new valeriana flavonoids with activity on the CNS. Pharmacol. Biochem. Behav. 75, 537–545. Miyase, T., Yamamoto, R., Ueno, A., 1996. Phenylethanoid glycosides from Stachys officinalis. Phytochemistry 43, 475–479. Muhayimana, A., Chalchat, J.C., Garry, R.P., 1998. Chemical composition of essential oils of Lantana trifolia L. from Rwanda. J. Essent. Oil Res. 10, 547–549. Noël, F., Mendonça-Silva, D.L., Quintas, L.E.M., 2001. Radioligand binding assays in the drug discovery process: high throughput screening and the risk of false negative. Arzneim. Forsch. Drug Res. 51, 169–173. Noël, F., Mendonça-Silva, D.L., Thibaut, J.P.B., Lopes, D.V.S., 2007. Characterization of two classes of benzodiazepine binding sites in Schistosoma mansoni. Parasitology 134 (7), 1003–1012. Rimpler, H., Sauerbier, H., 1986. Iridoid Glucosides as taxonomic markers in the genera Lantana, Lippia, Aloysia and Phyla. Biochem. Syst. Ecol. 14, 307–310. Rwangabo, P.C., 1988. Umuhengerin, a new antimicrobially active flavonoid from Lantana trifolia. J. Nat. Prod. 51, 966–968. Alam, M.S., Qasim, M.A., Kamil, M., Ilyas, M., 1986. Sorbifolin 6-galactoside from Garcinia andamanica. Phytochemistry 25, 2900–2901. Silva, G.N., Martins, F.R., Matheus, M.E., Leitao, S.G., Fernandes, P.D., 2005. Investigation of anti-inflammatory and antinociceptive activities of Lantana trifolia. J. Ethnopharmacol. 100, 254–259. Vilegas, W., Nehme, C.J., Dokkeddal, A.L., Piacente, S., Rastrelli, L., Pizza, C., 1999. Quercetagetin 7-methyl ether glycosides from Paepalanthus vellozioides and Paepalanthus latipes. Phytochemistry 51, 403–409. Viola, H., Wasowski, C., Destein, M.L., Wolfman, C., Silveira, R., Dajas, F., Medina, J.H., Paladini, A.C., 1995. Apigenin, a component of Matricaria-Recutita flowers, is a Central Benzodiazepine receptors-ligand with anxiolytic effects. Planta Med. 61, 213–216. Whimbey, A.E., Denenberg, V.H., 1967. Two independent behavioral dimensions in open-field performance. J. Compar. Physiol. Psychol. 63, 500–504. Wu, J., Huang, J.S., Xiao, Q., Zhang, S., Xiao, Z.H., Li, Q.X., Long, L.J., Huang, L.M., 2004. Spectral assignments and reference data – complete assignments of H-1 and C13 NMR data for 10 phenylethanoid glycosides. Magn. Reson. Chem. 42, 659– 662. Xia, H.J., Qiu, F., Zhu, S., Zhang, T.Y., Qu, G.X., Yao, X.S., 2007. Isolation and identification of ten metabolites of breviscapine in rat urine. Biol. Pharm. Bull. 30, 1308–1316. Zétola, M., De Lima, T.C.M., Sonaglio, D., Gonzalez-Ortega, G., Limberger, R.P., Petrovick, P.R., Bassani, V.L., 2002. CNS activities of liquid and spray-dried extracts from Lippia alba – Verbenaceae (Brazilian false melissa). J. Ethnopharmacol. 82, 207–215.
