Int J Pharm Pharm Sci, Vol 6, Issue 11, 318-325Original Article

SAMARAS OF AUSTROPLENCKIA POPULNEA (CELASTRACEAE): NEW CONSTITUENTS AND EFFECT OF EXTRACTS AND FRIEDELIN ON GERMINATION OF BIDENS PILOSA (ASTERACEAE)

CAROLINA M CANESCHI¹, SHIARA M SOUZA¹, THAIS S CERTO¹, GUSTAVO H B SOUZA¹, MICHELE S TACCHI CAMPOS², LUCIENIR P DUARTE³, GRACIA D F SILVA³, MARCOS S GOMES⁴, SIDNEY A VIEIRA FILHO^1*

¹Department of Pharmacy, Universidade Federal de Ouro Preto, Ouro Preto, Minas Gerais, Brazil, ²Department of Health Sciences, Biological and Agronomy, Universidade Federal do Espírito Santo, São Mateus, Espírito Santo, Brazil, ³Department of Chemistry, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil, ⁴Department of Chemistry, Universidade Federal de Lavras, Lavras, Minas Gerais, Brazil. Email: bibo@ef.ufop.br

Received: 23 Sep 2014 Revised and Accepted: 25 Oct 2014

---

ABSTRACT

Objective: Evaluation of the impact of extracts and constituents from samaras of Austroplenckia populnea on percentage of seed germination (%SG), germination speed index (GSI), length of rootlets (LR), seedling length (SL), and on dry mass (DM) of Bidens pilosa L weed.

Methods: The powder of samaras was extracted with organic solvents providing the hexane (SAPEH), chloroform (SAPEC), ethyl acetate (SAPEAE) and ethanol (SAPEE) extracts. The terpene 1 was isolated from SAPEH by means of column and thin layer chromatography and identified through NMR spectroscopy. Each extract and 1 were subjected to growth inhibition assays evaluating the following parameters: %SG, GSI, LR, SL and DM, with five repetitions.

Results: The compounds Friedelin (1), 7-hydroxy-clerodan-3-en-16,15:18,20-diolide (2), 3,5,7,4'-tetrahydroxy-6-methoxy-8-prenylflavanone (3), tetradecanamide (4), and 4-hydroxy-1,6,15-acetyloxy-8,9-benzoyloxy-agarofurane (5) were isolated from hexane extract of samaras of A. populnea and identified by spectroscopic data. The compounds 2, 3 and 5 were not previously described as being chemical constituents from Celastraceae family. In addition, the novel compounds 3 and 5 were described here for the first time. Substantial effect on the germination of B. pilosa L. (picão-preto) was observed after treatment of seeds with nonpolar extracts from Samaras of A. populnea. Friedelin inhibited the seed germination in the tested concentrations showing toxic properties against picão-preto.

Conclusion: The germination inhibition of seeds was higher using nonpolar extracts than polar extract. Friedelin inhibited the seed germination in the tested concentrations showing toxic properties against B. pilosa.

Keywords: Austroplenckia populnea, Samaras, Growth inhibition, Bidens pilosa, Allelopathy.

---

INTRODUCTION

During the last decades, natural products represent an adequate source of biologically active compounds that have been used to preserve or regenerate the human health. Many of these compounds have chemical properties associated to enzyme activity with consequent effects on the biosynthesis of essential endogenous substances and their ability to perform inherent functions in producing organisms [1]. Many of these natural compounds have been isolated from Celastraceae family species. Some of them have shown imunossupressor activities [2] and insecticide properties [3]. The Celastraceae family includes 98 genus with approximately 1210 species [4].

Austroplenckia populnea (Reissek), Brazilian species of the Celastraceae family, is popularly known as “mangabeira-brava”, “mangabarana” and “marmelinho do campo”. This plant has been found in Brazilian “Cerrado” (savanna region), mainly in Minas Gerais, São Paulo and Goiás states [5]. Leaves and roots of A. populnea have been used in traditional medicine to treat ulcers, dysenteries and rheumatism [6]. For this plant were also demonstrated larvicidal and molluscicidal properties [7]. Compounds from leaves of A. populnea were isolated and identified such as pentacyclic triterpenes of the oleanane [8, 9], friedelane [10, 11, 5], agarofuran sesquiterpenes [7, 12] and the polar constituents, epigalocathechin and pro-antocianidin A [12, 13, 14]. Despite the huge diversity of properties attributed to extracts or constituents isolated from A. populnea, no seed germination activity has been reported until now. Allelopathy effects have been considered a biological process in which plants, fungi and other microorganisms synthesize secondary metabolites, known as allelochemicals, modifying the growth, survival and reproduction of other organisms. Potential allelopathic interactions by different plants have often invoked why some invasive plants have become dominant inside plant communities [15]. Allelochemicals could have beneficial or detrimental properties on the target organisms [16, 17] and may alter plant competition also indirectly through changes in ecosystem properties [15]. Thus, allelochemicals have important biological relevance in plant resistance against herbivory [16]. Allelopathic interference has been considered an important factor to determine the distribution of some plant species and their abundance inside plant communities suggesting to be essential in success of many dominant plants [16, 17].

Bidens pilosa L. (Asteraceae) or hairy beggar stick is widely scattered mainly in tropical regions of the world, being endemically found in Central and South America. B. pilosa has been considered an invasive plant, commonly known in Brazil as “picão-preto”, that is found in entire Brazilian territory, with major prevalence in the agricultural areas of the south-central region. In Brazil, B. pilosa is considered one of the most important unwanted plants in both annual and perennial crops [18]. In this work, it was described the preparation of extracts of different polarities from samaras of A. populnea. It was also performed the preparation of hexane extract to be submitted to phytochemical methods targeting to isolate friedelin, the main pentacyclic triterpene found in this type of extract. Aiming to contribute to control of B. pilosa, in parallel with chemical study, the extracts and friedelin isolated from Samaras of A. populnea and submitted to assays to evaluate their potential effect on percentage of seed germination (%SG), germination speed index (GSI), length of the rootlets (LR), seedling length (SL), and on dry mass (DM) of B. pilosa L. weed.

MATERIALS AND METHODS

General

Column chromatography (CC) was performed on silica gel (Merck 60, 70-230 Mesh). Thin layer chromatography (TLC) was carried out using silica gel 60 F₂₅₄ (Merck) plates. Melting points were recorded on a Metler FP82 apparatus with processor Metler FP800. Gas chromatography coupled to mass spectroscopy (GC-MS) was realized using a GC-MSD 5975 Agilent equipment, with the following conditions: 240 °C injector temperature, 1.0 µL injected sample, 100:1 split, 1.4 psi helium pressure, column temperature program: 200 °C (1.0 min isotherm), 20 °C/min until 280 °C (40 min isotherm), with mass analysis range from 25 to 750 Da. ¹H and ¹³C NMR spectra were measured on Bruker AVANCE DRX400 and AVANCE DPX200 MHz spectrometers in CDCl₃, with tetramethylsilane (TMS) as an internal standard.

Plant material

Samaras of A. populnea were collected closest to Miguelão Lake at Nova Lima, Minas Gerais, Brazil. The material was compared and identified with a voucher specimen (No 10473) deposited at the Herbarium of the Museum of Historia Natural of Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil. Each plant material was dried over paper at room temperature (r. t.) And then fragmented using a knife mill.

Extract obtainment

The powder of samaras (286.97 g) was subjected to exhaustive extraction in a Soxhlet apparatus with organic solvents providing the hexane (SAPEH), chloroform (SAPEC), ethyl acetate (SAPEAE), and finally ethanol (SAPEE) extracts. Each solvent extractor was respectively recovered using a rotatory evaporator at temperature ≤ 40 ^oC and the respective extracts were dried in a vacuum desiccator. Targeting the isolation of friedelin, only the hexane extract was submitted to silica gel CC, TLC and other phytochemical methods phytochemical methods, as suggested by Matos (1980) [19] and Wagner & Bladt (1996) [20].

Phytochemical assay

The hexane extract of samaras (44.91g) was obtained as oil mixed with solid material, which was carefully separated under vacuum on a Buchner filter. Friedelin (1) (150.2 mg), melting range 260-263 ⁰C, was isolated from this solid material in accordance to method previously described [19, 20]. The remaining of hexane extract was treated with ethyl acetate that induced the formation of another solid material (35.21 mg) which was separated, under vacuum on Büchner filter and dried on a dessicator. By TLC analysis, it was observed the presence of one constituent (Rf 0.8, CHCl₃) which was isolated using preparative TLC as a whitish solid material [2 (9.28 mg)], melting range 189-192°C. The constituent 2 was subjected to ¹H and ¹³C NMR spectroscopy, including DEPT experiment.

The oily fraction was subjected to fractionation on silica CC, eluted with hexane, chloroform, ethyl acetate and methanol, providing 100 fractions (EH-O) of 25 mL each. The eluent was recuperated using a rotary evaporator. The fractions were grouped in accordance to the TLC similar profile. By this process, the majority of fractions were considered as the complex mixture and not submitted to other assays. Two fractions, EH-O 55 and EH-O 59 were purified aiming to isolate pure constituents. The fraction EH-O 55 (152.2 mg) was subjected to preparative TLC eluted with a mixture of hexane, chloroform and ethyl acetate [35:50:15]. After treatment with iodine vapor, were observed three distinct spots. The constituent of greater concentration (Rf 0.6) was isolated by preparative TLC. After isolation, constituent 3 (11.5 mg) present a single spot when analyzed by TLC, even using eluents of different polarities. Due to this result, constituent 3 was submitted to ¹H and ¹³C NMR spectroscopy including DEPT-135 experiment. The other two constituents were isolated in amounts ≤ 2 mg and disregarded. EH-O 59 (425.7 mg) was subjected to silica gel CC eluted with an isocratic mixture of chloroform and ethyl acetate [6:4], providing 26 fractions of 30 mL. After dried, the fraction 17, provides a pasty material 4 (15.4 mg). After evaporation of the eluent, the residue of fraction 18 was re-crystalized with chloroform-ethanol (8:2) to obtain compound 5 (12.08 mg) as a white amorphous solid, melting range 148-152 °C. Constituents 4 and 5 present a single spot when subjected to TLC, with eluent of different polarities, and then were submitted to ¹H and ¹³C NMR spectroscopy including DEPT-135 experiment.

Allelopathy assays

Seeds of B. pilosa L. (picão-preto) were collected at the experimental plantation of the Centro Universitário Norte do Espírito Santo (CEUNES), São Mateus, Brazil, and sterilized by immersion for 10 minutes in 2 % sodium hypochlorite aqueous solution.

The germination, development and growth of seedlings of B. pilosa were carried out treating seeds with the extracts and friedelin, both isolated from Samaras of A. populnea. The samples were dissolved in corresponding solvent to reach final concentration of 31.25; 62.5; 125; 250; 500; 750 and 1000 mg/L of extract (SAPEH, SAPEC, SAPEAE or SAPEE), and 6.25; 12.50; 25; 50; 100; 200 and 400 mg/L of fridelin (1). Due to the small isolated quantities, compounds 2 to 4 isolated from Samaras of A. populnea were not subjected to seed growth inhibition assays. Seeds of B. pilosa (picão-preto) were distributed in Petri plates on two filter paper discs, previously humidified with the extract or Fridelin (1), both in adequate concentration. Later the plates were conditioned in germination camera at 25 °C, equipped with fluorescent light (8 x 40 W) photoperiod of 12 hours, for twelve days.

At the end of an experimental period the following parameters were evaluated: percentage of seed germination (%SG), germination speed index (GSI), length of rootlets (LR), the seedling length (SL) and dry mass (DM) of B. pilosa L., with five repetitions [21]. The GSI was calculated as suggested by Maguire et al. (1962) [22], using the number of germinated seeds, divided by the day of germination and adding until the last day of germination according to the equation: N₁/ D₁ + N₂/ D₂ + N₃/ D₃ +…. + N_n/ D_n. The terms N₁, N₂, N₃ and N_n represent the number of germinated seeds until enesimo day. And, D₁, D₂, D₃ e D_n represent the number of days in which the seed germination was evaluated. The inhibition percent were calculated basing on the experimental data obtained for control, realized without samples, maintaining constant the other parameters.

Statistical analysis

The experimental results were subjected to analysis of variance (ANOVA) and the significant variables established by F test were submitted to regression analysis and the test of averages, compared according Scott & Knott (1974) [23]. A completely randomized design was adopted, with the treatments arranged in factorial scheme (4x7), being four extracts (SAPEH, SAPEC, SAPEAE and SAPEE) and seven concentrations (0.00; 31.25; 62.50; 125.00; 250.00; 500.00; 750.00 and 1000.00 mg/L), with five repetitions. The data were analyzed using the System of Analysis of Variance for balanced data software (SISVAR) [24]. P values < 0.05 were considered to be statistically significant.

RESULTS AND DISCUSSION

Phytochemical assay

It was observed Friedelin (1) (fig. 1) was a principal pentacyclic triterpene found in the solid material isolated from hexane extract of samaras of A. populnea. The chemical structure of terpene 1 was identified by melting point, TLC analysis in comparison with authentic sample and NMR spectral data. The chemical shift assignments of 1 were in accordance with the NMR spectral data of friedelin previously reported [6, 25, 26].

The constituent 2 was subjected to ¹H and ¹³C NMR spectroscopy, including DEPT experiment. Thus, it was possible to establish the chemical shift assignments correspondent to 2 methyl, 8 methylene, 5 methynic, being one bonded to hydroxyl (δ_c 72.39), and 5 non-hydrogenated carbon atoms. The signals at δ_c 169.91 and δ_c 174.74 were associated to carboxyl groups [27]. The presence of the double bond [δ_c 134.45 (CH) and δ_c 139.23 (C)], and the ¹H and ¹³C NMR profiles suggested a structure of the clerodane diterpene class. For this reason, it was performed a comparison of the NMR data of compound 2 with published data for clerodanes (Table 1). The NMR spectral data of 2 were similar to the clerodane BT-CD isolated from Bacharis trimera (Asteraceae) [28, 29]. Based on the spectral data was attributed to 2 the structure of BT-CD [29] or 7-hydroxy-clerodan-3-en-16,15:18,20-diolide (2) (fig. 1).

Fig. 1: Chemical structures of friedelin (1), 7-hydroxy-clerodan-3-ene-16,15:18,20-diolide (2), 3,5,7,4’-tetra-hydroxy-6-metoxy-8-prenylflavanone (3), tetradecanamide (4) and 4-hydroxy-1,6,15-tri-acetyloxi-8,9-di-benzyloxi-agarofurane (5).

The presence of clerodanes diterpenes was also found in species of the family Flacourtiaceae [30], Asteraceae [29] and Labiatae [31]. Properties such as antifungal, antibacterial, insect repellence and others were attributed for compounds of the clerodane class by bio-monitored fractionations [30, 31].

The constituent 3 was isolated as an amorphous solid, melting point 85-88 °C. By the ¹³C NMR and DEPT-135 spectral data were observed three signals correspondents to methyl, one of them attributed to methoxyl, one methylene, seven methine groups, and ten signals associated to non-hydrogenated carbon. The signal at δ_C 195.91 was attributed to ketone carbonyl and the signals at δ_C 121.30 and δ_C 132.54 were associated to double bond. In the ¹H NMR spectrum were observed signals correspondent to aromatic hydrogen atoms in the region between δ_H 6.00 and δ_H 8.00. The signal centered at δ_H 3.97 was attributed to hydroxyl linked to CH (δc 73.33) (Table 2).

A comparison of the NMR spectral data of 3 with those available in the literature was performed. Our results demonstrated the high similarity with the flavonoid 3,5,7,2'-tetrahydroxy-6-methoxy-8-prenylflavanone isolated from Dioclea grandiflora by Lemos et al. (2002) [32]. Based on this comparison it was possible to propose for constituent 3 the chemical structure of 3,5,7,4'-tetrahydroxy-6-methoxy-8-prenylflavanone (3) (C₂₁H₂₂O₇, MW 386 g/mol) (fig. 1), an unpublished flavanoid.

Table 1: 1D NMR spectral data (CDCl₃, 200 MHz) of 7-hydroxy-clerodan-3-en-16,15:18,20-diolide (2) and comparison of its δ_C with reported data.

Carbon Number	7-hydroxy-clerodan-3-ene-16,15:18,20-diolide (2)			Reported dC
	dC	Attribution	dH [J (Hz)]	(Werner et al., 1977)	(Januário et al., 2004)
1	19.35	CH3	0.86 s	19.4	19.3
2	27.65	CH2	-	27.6	27.6
3	134.45	CH	3.9	139.2	135.1
4	139.23	C	6.72 d	134.7	139.0
5	44.90	C	-	44.9	44.9
6	40.44	CH	-	40.5	40.4
7	72.39	CH	1.63 s	72.21	72.3
8	40.47	CH2	-	40.5	40.4
9	38.56	C	-	38.4	38.3
10	48.19	CH	-	48.3	48.1
11	36.22	CH2	-	36.2	32.2
12	26.56	CH2	-	26.5	26.5
13	36.13	CH	-	36.1	36.1
14	34.18	CH2	-	34.6	34.7
15	176.74	C	-	176.6	176.9
16	73.23	CH2	3.9-5.3 m	73.1	73.3
17	11.63	CH3	1.03 d [7.2]	11.9	12.0
18	19.26	CH2	1.00 d	19.1	19.3
19	72.71	CH2	4.11 s	72.7	72.8
20	169.91	C	-	169.8	170.1

Through the analysis of ¹H-NNR spectrum of constituent 4, the signal at δ_H 0.80 was attributed to hydrogen of methyl group and the intense signal between δ_H 1.20 and δ_H 1.90 was correlated to hydrogen atoms of long chain methylene groups. In the ¹³C NMR spectrum, signals were predominant in the region between the δ_C 10.00 and δ_C 35.00. These signals were associated to long chain containing methylene and methyl groups [27]. These observations were proved by the sub-spectrum DEPT-135, in which the signal at δ_C 14.75 was attributed to methyl group and the intense signals observed between δ_C 28.16 and δ_C 33.02 was associated to a long-chain methylene groups. The signal at δ_C 175.10 was attributed to a carboxyl-terminal amide [27]. Based on the correlation between the number of hydrogen and carbon atoms was possible to establish the presence of 14 carbon atoms and suggested for compound 4 the structure of tetradecanamide (fig. 1).

In the ¹H-NMR spectrum of 5 it was observed a set of hydrogen signals between δ_H 1.26 to δ_H 1.67 and δ_H 2.08 to δ_H 2.11, which was associated to aliphatic carbon chain. The signals observed between δ_H 7.38 to δ_H 7.67, and δ_H 7.93 to δ_H 8.15, was correlated to hydrogen atoms bonded to aromatic carbon. Singlet signals were observed at δ_H 1.47; 1.52; 2.11; 2.08 and δ_H 2.26. Through integration of this hydrogen signals it was possible to establish the presence of methyl groups bonded to non-hydrogenated carbons. The integration of the signal at δ_H 1.56 was correlated with hydrogen atoms of two methyl groups and the signal at δ_H 4.01 was attributed to a hydroxyl bonded to CH. By the ¹³C NMR spectrum and DEPT-135 of 5 it was established the presence of signals associated to 6 methyls, 3 methylenes, 15 methynes, and 11 non-hydrogenated carbon atoms, and the signal at δ_C 170.55 was attributed to C=O of acetyl group bonded to methylene carbon. Through the analysis of NMR data it was possible to establish the presence of hydroxyl, methyl, acetyl and benzoyl groups bonded to the basic structure of 5 (Table 3).

Similarities were found between NMR spectral data of 5 with published data of 4-hydroxy-1,2,6,15-tetrahydro-9-acetyloxy-benzoyloxy-agarofurane or populnane isolated from leaves of A. populnea [26]. ¹³C NMR data of 5 (Table 3) were also similar to carbon chemical shifts of agarofurans known as reissantines FH1, FH2 and FH3, isolated from Reissantia buchananii, species of the Celastraceae family [33].

Table 2: ¹H and ¹³C NMR (400 MHz, CDCl3) data of 3,5,7,4'-tetrahydroxy-6-methoxy-8-trocar prenil por prenyl (3) and comparison with previously reported ¹³C data of an isomer^*

Carbon Number	3,5,7,4’-Tetra-hydroxy-6-metoxy-8-prenylflavone (3)			dc of isomer*
	dc	Attribution	dH
2	79.00	CH	6.99	79.2
3	73.33	CH	3.97	72.6
4	195.91	C	-	199.6
5	151.09	C	11.19 (OH)	153.9
6	128.72	C	-	129.9
7	156.82	C	-	158.7
8	108.24	C	-	108.4
9	155.59	C	-	156.4
10	100.38	C	-	101.4
1´	124.10	C	-	124.9
2´	126.94	CH	7.26	157.3
3´	130.09	CH	7.0	116.3
4´	153.96	C	-	129.9
5´	121.46	CH	7.03	119.5
6´	118.53	CH	7.55	129.0
1”	21.094	CH2	3.27	22.4
2”	121.30	CH	5.43	123.4
3”	132.54	C	-	130.8
4”	17.76	CH3	1.66	17.7
5”	25.76	CH3	1.25	25.7
OCH3	61.09	CH3	3.30	60.2

* 3,5,7,2'-Tetrahydroxy-6-methoxy-8-prenylflavanone (Lemos et al., 2002).

Table 3: 1D NMR spectral data (CDCl₃, 200 MHz) of 4-hydroxy-1,6,15-tri-acetyloxi-8,9-di-benzyloxi-agarofurane (5) and comparison of its δ_C with the published data for reissantines [26]

4-hydroxy-1,6,15-tri-acetyloxi-8,9-di-benzyloxi-agarofurane (5)			dc Reissantines (CDCl3, 125 MHz) (Chang et al., 2006)
C	Attribution	dc	FH1	FH2	FH3
1	CH	75.23	76.39	77.32	-
2	CH2	22.69	25.06	27.49	-
3	CH2	37.79	37.99	39.11	39.84
4	C	70.58	70.22	70.95	70.16
5	C	91.95	92.51	91.32	93.87
6	CH	78.25	75.11	78.54	74.87
7	CH	53.45	52.07	53.71	53.74
8	CH	74.02	75.50	68.74	70.88
9	CH	72.85	75.04	72.68	-
10	C	52.03	50.71	51.08	-
11	C	82.65	83.97	84.47	83.82
12	CH3	24.40	25.79	30.41	24.00
13	CH3	29.59	29.62	26.51	29.37
14	CH3	22.48	23.41	23.78	23.29
15	CH2	61.04	61.03	-	-
CH3COO-C1
C=O		169.74	169.82	169.24
CH3		21.53	21.43	20.80
CH3COO-C6			-	-	-
C=O		169.92	-	-	-
CH3		21.28	-	-	-
FCOO-C8
C=O		164.81	-	-	-
C (C1)		129.98	-	-	-
CH (C2-C6)		128.60	-	-	-
CH (C3-C5)		128.54	-	-	-
CH (C4)		133,24	-	-	-
FCOO-C9
C=O		166.08	165.62	165.69	166.27
C (C1)		129.64	128.38	128.83	129.28
CH (C2-C6)		129.81	129.02	129.90	130.16
CH (C3-C5)		129.27	128.38	128.67	128.82
CH (C4)		133.38	133.35	133.37	133.82
CH3COO-C15
C=O		170.55	169.82	169.24	-
CH3		20.70	21.43	20.80	-

Thus, it was proposed the basic structure of agarofuran for 5. To confirm this hypothesis, constituent 5 was subjected to GC-MS. In the mass spectra obtained for the main constituent [retention time 39.19 min (76.02 %)], it was observed a peak at m/z 592 associated with [M⁺ - acetic acid], m/z 105, concerning to the loss of C₇H₅O (ø–CO^.) and m/z 43 corresponding to CH₃-C=O. Based on the spectral data, it was suggested for 5 the structure of 4-hydroxy-1,6,15-tri-acetyloxi-8,9-di-benzyloxi-agarofurane [C₃₅H₄₀O₁₂, MW 652 g/mol] (fig. 1). Sesquiterpene esters with β-dihydro-agrofuran structure were suggested as a taxonomic marker of the species of the Celastraceae family [34]. From the hexane extract of samaras were isolated and identified by spectroscopy data the constituents: friedelin (1); 7-hydroxy-clerodan-3-en-16,15:18,20-diolide (2); 3,5,7,4'-tetrahydroxy-6-methoxy-8-prenylflavanone (3); tetradecanamide (4) and 4-hydroxy-1,6,15- -acetyloxy-8,9-benzoyloxy-agarofurane (5) (fig. 1). Compounds 2, 3 and 5 were not previously identified in species of the family Celastraceae. In addition, it was not found reports in the literature regarding the compounds 3 and 5.

Allelopathy assay

The triterpenoids have been shown one of the main groups of secondary compounds with allelopathic activity [35]. Pentacyclic triterpenes have been considered as a class of promising secondary metabolites which arising from cyclization of squalene [36]. Several biological activities have been attributed to this group of secondary metabolites, such as bactericidal, fungicidal, antiviral, cytotoxic, analgesic, anticancer, spermicidal, cardiovascular and antiallergic [37]. Waller (1999) [35] suggested triterpenoids as being the principal group of secondary compounds with allelopatic properties. Among the pentacyclic triterpenes, fridelin was cited as being potent allelochemical when it is used along with epifridelinol against weeds malice (Mimosa pudica) and obtusifolia (Senna obtusifolia) [38]. This pentacyclic triterpene was also isolated from leaves of A. populnea [26] and from other species of the Celastraceae family, such as M. gonoclada [38].

The potential allelochemical effect has been associated to the nature of secondary metabolite and the sensitivity of the type that assesses development [39]. The structural diversity of triterpenes with potential antiprotozoal activity involved in different mechanisms of action has stimulated the interest to identify other natural compounds that may provide new types of antimicrobials and also for weed control [40].

Then, the main objective associated to the studies of allelopathic properties has been based on the resistance or tolerance acquisition to secondary metabolites which act as allelochemicals being more or less specific [39]. The hexane (SAPEH), chloroform (SAPEC), ethyl acetate (SAPEAE), and ethanol (SAPEE) extracts from Samaras of A. populnea were subjected to assays to evaluate their allelochemicals properties through %SG, GSI, LR, SL and DM of B. pilosa L.

Percentage of seed germination (%SG)

In this work the most efficient extract was the SAPEH. At concentration of 1000 mg/L, this extract inhibited 50% of seed germination, while SAPEE, the least efficient, had no significant inhibition at the same concentration. These extracts from samaras displayed more phytotoxic effect against B. pilosa L comparing with the other extracts. Therefore, the extracts SAPEE and SAPEAE showed higher germination percentage and consequently lower inhibitory effect on seed germination. No statistical differences between these extracts were observed (Table 4). Similar results related to the reduction of the percentage of germination were observed in previous studies developed by Hoffmann et al. (2007) [41], Azambuja et al. (2010) [42] and Haida et al. (2010) [43] using seeds of B. pilosa.

Germination speed index (GSI)

SAPEH and SAPEAE extract displayed high inhibition comparing with the other extracts (Table 5).

Table 4: Effect of the concentration of extracts from samaras of A. populnea on percentage of seed germination (%SG) of Bidens pilosa L

Extract (mg/L)	SG % after treatment with extracts
	SAPEH	SAPEC	SAPEAE	SAPEE
0	83.32 Ba	88.32 Aa	81.68 Ba	90.02 Aa
31.25	73.36 Ba	86.66 Aa	80.00 Ba	88.36 Aa
62.50	71.66 Bb	83.34 Ab	70.00 Bb	85.00 Ab
125	70.00 Bb	80.00 Ab	68.32 Bb	83.36 Ab
250	66.66 Bb	76.66 Ab	66.66 Bb	81.66 Ab
500	61.68 Bc	73.34 Ac	60.00 Bc	79.98 Ac
750	51.66 Bc	70.00 Ac	58.34 Bc	78.34 Ac
1000	41.68 Bc	70.00 Ac	50.00 Bc	73.36 Ac

* S = Samaras, A = Austroplenckia, P = populnea, H = hexane, C = Chloroform, EA = ethyl acetate, and E = ethanol. ** Means followed by the same letter, being this tiny on the column, and capital in same line, do not differ by Scott-Knott test at 5 % significance.

Table 5: Effect of the concentration of extracts from samaras of A. populnea on germination speed index (GSI) of Bidens pilosa L

Extract (mg/L)	GSI after treatment with extracts
	SAPEH	SAPEC	SAPEAE	SAPEE
0	20.70 Ba	23.26 Aa	19.78 Ba	21.88 Aa
31.25	18.88 Bb	21.24 Ab	18.46 Bb	20.56 Ab
62.50	18.06 Bb	21.30 Ab	17.14 Bb	21.32 Ab
125	17.94 Bb	19.26 Ab	15.92 Bb	19.90 Ab
250	16.70 Bc	18.32 Ac	14.40 Bc	19.68 Ac
500	15.94 Bc	17.68 Ac	13.24 Bc	18.66 Ac
750	13.12 Bd	16.04 Ad	14.10 Bd	18.74 Ad
1000	10.60 Bd	17.34 Ad	12.20 Bd	16.78 Ad

* S = Samaras, A = Austroplenckia, P = populnea, H = hexane, C = Chloroform, EA = ethyl acetate, and E = ethanol. ** Means followed by the same letter, being this tiny on the column, and capital in same line, do not differ by Scott-Knott test at 5 % significance.

Length of rootlets (LR)

The hexane extract of A. populnea samaras, at concentrations ≥ 25 mg/L induced strong allelopathic effect on the length of the rootlets of B. pilosa L., in accordance with the quadratic model. It was observed that SAPEH behaved significantly different from other extracts and their activity increased with higher concentrations. Regarding extracts SAPEC and SAPEE were not observed significant differences related to length of rootlets. On the other hand, SAPEAE induced a significant difference on length of rootlets (Table 6).

Table 6: Effect of the concentration of extracts from samaras of A. populnea on length of rootlets (LR) (mm) of Bidens pilosa L

Extract (mg/L)	LR (mm) after treatment with extracts
	SAPEH	SAPEC	SAPEAE	SAPEE
0	28.54 Ca	37.96 Aa	30.96 Ba	36.12 Aa
31.25	28.08 Ca	37.14 Aa	30.54 Ba	35.68 Aa
62.50	23.86 Cb	36.82 Ab	29.26 Bb	34.52 Ab
125	21.14 Cb	35.98 Ab	29.26 Bb	33.22 Ab
250	21.04 Cb	35.34 Ab	29.20 Bb	32.64 Ab
500	20.70 Cc	32.06 Ac	26.88 Bc	31.44 Ac
750	19.42 Cc	31.92 Ac	26.32 Bc	29.84 Ac
1000	18.76 Cc	30.32 Ac	25.76 Bc	29.56 Ac

* S = Samaras, A = Austroplenckia, P = populnea, H = hexane, C = Chloroform, EA = ethyl acetate, and E = ethanol. ** Means followed by the same letter, being this tiny on the column, and capital in same line, do not differ by Scott-Knott test at 5 % significance.

According Gussman et al. (1994) [44] and Hoffman et al. (2007) [41], the elongation of shoots and roots are dependent on the intensity of cellular division, training and exchange xilematic vessels. These structures are dependent on the nutrient uptake and nutrient partitioning by seedling. As a result, the root tissues of plants are the most sensitive to the action of allelochemicals. Thus, the lower germination speed index showed more difficult for the plant to stretch, as observed after treatment with SAPEE. Our results were in accordance with previous studies developed by Alves et al. (2004) [45], related to length of rootlets of lettuce seedlings observed after the treatment of seeds with oils isolated from cinnamon, rosemary, peppermint and citronella grass.

Seedling length (SL)

In relation to the seedlings length, the hexane (SAPEH) and ethyl acetate (SAPEAE) extracts from samaras it was not observed significant differences among them. These extracts reduced the shoot length of hairy beggartick seeds (B. pilosa) at low concentrations ≤ 62.50 mg/L. The extracts SAPEC and SAPEE exhibited similar behavior and also did not differ statistically, but induced less allelopathic effect comparing with other extracts from samaras (Table 7).

Dry mass (DM)

The inhibition of germination of “picão-preto” (B. pilosa L.) seeds was higher after treatment with nonpolar extracts SAPEH and SAPEC. At 1000 mg/L, SAPEH inhibited 50 % germination of the seeds, while SAPEE, the least efficient, had no significant inhibition at the same concentration. These results were also reflected through the results of dry mass (Table 8).

Fridelin (1) (fig. 1), isolated from hexane extract of Samaras reduced the percentage of germination (%SG) even at lower concentrations (12.50 mg/L) and also reduce the germination speed index (GSI) in concentration higher than 25 mg/L (Table 9). This triterpene did not induce differences in the length of rootlets (LR) and induced a small reduction in the seedlings length (SL) in concentrations higher than 50.0 mg/L (Table 9).

Table 7: Effect of the concentration of extracts from samaras of A. populnea on seedling length (SL) (mm) of Bidens pilosa L

Extract (mg/L)	SL (mm) after treatment with extracts
	SAPEH	SAPEC	SAPEAE	SAPEE
0	58.06 Ba	65.72 Aa	50.10 Ba	63.16 Aa
31.25	41.00 Bb	63.54 Ab	44.32 Bb	61.74 Ab
62.50	37.00 Bc	62.70 Ac	40.94 Bc	60.48 Ac
125	36.64 Bc	60.82 Ac	40.80 Bc	59.08 Ac
250	36.12 Bc	60.54 Ac	39.66 Bc	58.10 Ac
500	35.56 Bc	60.76 Ac	37.56 Bc	57.28 Ac
750	35.22 Bc	58.54 Ac	37.44 Bc	56.92 Ac
1000	33.22 Bd	50.54 Ad	36.02 Bd	48.66 Ad

* S = Samaras, A = Austroplenckia, P = populnea, H = hexane, C = Chloroform, EA = ethyl acetate, and E = ethanol. ** Means followed by the same letter, being this tiny on the column, and capital in same line, do not differ by Scott-Knott test at 5 % significance.

Table 8: Effect of the extract concentration of extracts from samaras of A. populnea on dry mass (mg) of Bidens pilosa L

Extract (mg/L)	DM (mg) after treatment with extracts
	SAPEH	SAPEC	SAPEAE	SAPEE
0	59,48 Ba	62,98 Aa	58,58 Ba	61,00 Ba
31.25	56,42 Bb	61,82 Ab	56,34 Bb	56,74 Bb
62.50	54,94 Bb	61,26 Ab	56,00 Bb	55,32 Bb
125	53,88 Bb	59,26 Ab	55,00 Bb	55,20 Bb
250	53,40 Bb	58,76 Ab	52,94 Bb	54,12 Bb
500	50,72 Ba	57,34 Aa	49,38 Ba	52,40 Ba
750	48,26 Ba	56,02 Aa	47,08 Ba	51,66 Ba
1000	45,68 Ba	55,46 Aa	42,62 Ba	51,00 Ba

* S = Samaras, A = Austroplenckia, P = populnea, H = hexane, C = Chloroform, EA = ethyl acetate, and E = ethanol. ** Means followed by the same letter, being this tiny on the column, and capital in same line, do not differ by Scott-Knott test at 5 % significance.

The inhibition of germination of “picão-preto” (B. pilosa) seeds was higher using nonpolar extracts. Friedelin inhibited the seed germination in the tested concentrations showing toxic properties against “picão-preto”. This study contributed to the researches of allelopathic effects and open perspectives for the production of natural substances that can be used in weeds control. And also it stimulates the accomplishment of new experiments using other invasive plants, others constituents isolated from other species of the Celastraceae family and respective derivatives in order to establish chemical structure versus allelopatic activity relationships.

Table 9: Effect of fridelin concentration on percentage of seed germination (%SG), germination speed index (GSI), length of rootlets (LR) and seedling length (SL) of Bidens pilosa L

Friedelin (mg/L)	Parameter evaluated
	SG (%)	GSI	LR (mm)	SL (mm)
0	85.00 A	17.00 A	45.60 A	44.14 A
6.25	76.66 A	17.04 A	44.32 A	42.10 A
12.50	58.34 B	15.62 A	41.90 A	42.10 A
25	58.32 B	11.60 B	38.96 A	39.76 A
50	58.34 B	15.62 A	38.92 A	35.50 B
100	56.64 B	12.02 B	33.86 A	33.00 B
200	53.34 B	12.42 B	33.70 A	31.96 B
400	41.66 B	9.90 B	33.02 A	31.54 B

* Means followed by the same letter, being this tiny on the column, and capital in the same line, do not differ by Scott-Knott test at 5 % significance.

ACKNOWLEDGMENT

The authors thank the Fundação de Amparo a Pesquisa de Minas Gerais (FAPEMIG) for financial support, and to Dr. Valdenir José Belinelo (In memoriam), CEUNES, Universidade Federal do Espírto Santo, Brazil, for allelopathy assays supervision.

CONFLICT OF INTERESTS

Declared None

REFERENCES

1. Li JW-H, Vederas JC. Drug discovery and natural products: end of an era or an endless frontier? Sci 2009;325:161-5.
2. Shirota O, Morita H, Takeya K, Itokawa H. Revised structures of cangorosins, triterpene dimers from Maytenus ilicifolia. J Natl Prod 1997;60:111-15.
3. Avilla J, Teixido A, Velázquez C, Alvarenga N, Ferro E, Canela R. Insecticidal activity of Maytenus species (Celastraceae) nortriterpene quinone methides against codling moth, Cydia pomonella (L) (Lepidoptera: tortricidae). J Agric Food Chem 2000;48:88-92.
4. Simmons MP, Cappa JJ, Archer RH, Ford AJ, Eichstedt D, Clevinger CC. Phylogeny of the Celastreae (Celastraceae) and the relationships of Catha edulis (qat) inferred from morphological characters and nuclear and plastid genes. Mol Phyogenetics Evolution 2008;48:745-57.
5. Vieira-Filho SA, Duarte LP, Silva GDF, Howart OW, Lula IS. 3β-Stearyloxy-olean-12-ene from Austroplenckia populnea: structure elucidation by 2D-NMR and quantitative [13]C NMR spectroscopy. Helvetica Chimica Acta 2003;86:3445–9.
6. Miranda RRS, Duarte LP, Silva GDF, Vieira Filho SA, Carvalho PBD, Messas AC. Evaluation of antibacterial activity of “Mangabarana” Austroplenckia populnea Reissek (Celastraceae). Rev Bras Farmacogn 2009;19:370-5.
7. Vichnewski W, Prasad JS, Herz W. Polyhydroxyagarofuran derivatives from Austroplenckia polpunea. Phytochem 1984;23:1655-7.
8. Souza JR, Silva GDF, Pedersoli JL, Alves RJ. Friedelane and olenane triterpenoids from bark wood of Austroplenckia polpunea. Phytochem 1990;29, 3259-61.
9. Duarte LP, Vieira Filho SA, Silva GDF, Sousa JR, Pinto AS. Anti-trypanosomal activity of pentacyclic triterpenes isolated from Austroplenckia populnea (Celastraceae). Rev Inst Med Trop Sao Paulo 2002;44:109-12.
10. Cotta AB, Mascarenhas YP, Silva GDF, Souza JR. Structure of a triterpene extracted from Austroplenckia polpunea (Celastraceae), Methyl-3-oxo-friedelan-20α-oate. Acta Crystallogr 1990;46:326-7.
11. Vieira Filho SA, Lanna MCS, Paes HCS, Silva GDF, Sousa JR. Antibacterial activity of pentacyclic triterpenes from Austroplenckia populnea. Acta Hortic 1999;501:199-203.
12. Vieira Filho SA, Duarte LP, Silva GDF, Lula IS, dos Santos MH. Complete assigment of the ¹H and [13]C NMR spectra of a new polyester sesquiterpene from Austroplenckia populnea. Magn Reson Chem 2000;38:1023-6.
13. Vieira Filho SA, Duarte LP, Silva GDF, Lula IS, Santos MH. Total assignment of ¹H and [13]C NMR spectra of two 3, 4-secofriedelanes from Austroplenckia polpunea. Magn Reson Chem 2001;39:746-8.
14. Silva FC, Rodrigues VG, Duarte LP, Silva GDF, Miranda RRS, Vieira Filho SA. A new friedelane triterpenoid from the branches of Maytenus gonoclada (Celastraceae). J Chem Res 2011;35(10):955-7.
15. Prati D, Bossdorf O. Allelopathic inhibition of germination by Alliaria petiolata (Brassicaceae). Am J Bot 2004;91(2):285–8.
16. Stamp N. Out of the quagmire of plant defense hypotheses. Q Rev Biol 2003;78 (1):23–55.
17. Saffari M, Saffari VR, Torabi-Sirchi MH. Allelopathic appraisal effects of straw extract wheat varieties on the growth of corn. Afr J Plant Sci 2010;4:427-32.
18. Adegas FS, Voll E, Prete CEC. Inhibition and germination of hairy beggartick seeds (Bidens pilosa). Planta Daninha 2003;21:21-5.
19. Matos FJA. Introdução a Fitoquimica Experimental, Editora UFC, Fortaleza, Ceará; 1980. p. 150.
20. Wagner H, Bladt S. Plant Drug Analysis, Springer, Berlin; 1996. P. 384.
21. Oliveira LGA, Belinelo VJ, Almeida MS, Aguilar EB, Vieira Filho SA. Alelopatia de Emilia sonchifolia (L) Dc. (Asteraceae) na germinação e crescimento inicial de sorgo, pepino e picão preto. Enciclopédia Biosfera, Centro Científico Conhecer-Goiânia 2011;7(12):1-10.
22. Maguire JD. Speed of germination-aid in selection evaluation for seedling emergence and vigour. Crop Sci 1962;2:176-7.
23. Scott AJ, Knott M. A cluster analysis method for grouping means in the analysis of variance. Biometrics 1974;30(3):507-12.
24. Ferreira DF. Sistema Para Análise de Variância Para Dados Balanceados (SISVAR). Universidade Federal de Lavras. Lavras, Minas Gerais, Brazil; 1999.
25. Mahato SB, Kundu AP. ^[13]C-NMR spectra of pentacyclic triterpenoids–a compilation and some salient features. Phytochem 1994;37(6):1517-75.
26. Vieira Filho SA, Silva GDF, Duarte LP, Mazaro R, Di Stazi LC. Constituintes químicos e atividade antiespermatogênica em folhas de Austroplenckia populnea (Celastraceae). Brasilian J Pharmacogn 2002;12:123-4.
27. Silverstein RM, Webster FX, Kiemle DJ. Spectrometric Identification of Organic Compounds. 7th Edition. John Wiley & Sons. Hoboken, NJ. USA; 2005. p. 502.
28. Werner H, Pilotti AM, Soderholm AM, Shuhama IK, Vichnewski W. New ent-Clerodane-Type Diterpenoids from Baccharis trimera. J Org Chem 1977;42(24):3913-7.
29. Januário AH, Santos SL, Marcussi S, Mazzi MV, Pietro RCLR, Sato DN, et al. Neo-clerodane diterpenoid, a new metalloprotease snake venom inhibitor from Baccharis trimera (Asteraceae): anti-proteolytic and anti-hemorrhagic properties. Chem Biol Interact 2004;150:243-51.
30. Oberlies NH, Burgess JP, Navarro HN, Pinos RE, Fairchild CGR, Peterson RW, et al. Novel bioactive clerodane diterpenoids from the leaves and twigs of Casearia sylvestris. J Nat Prod 2002;65:95-9.
31. Viegas Jr C. Terpenes with insecticidal activity: an alternative to chemical control of insects. Química Nova 2003;26(3):390-400.
32. Lemos VS, Santos MH, Rabelo LA, Côrtes SF. Total assignments of ¹H and [13]C NMR spectra of a new prenylated flavanone from Dioclea grandiflora. Magn Reson Chem 2002;40:793-4.
33. Chang FR, Cheng I, Liao SC, Issa HH, Hayashi KI, Nosako, et al. Planta Med 2006;72:92-6.
34. González AG, Jiménez LA, Nuñez MP, Ravelo AG, Bazzocchi LL, Muñoz O M, et al. New sesquiterpenes from Maytenus species (Celastraceae). Taxonomic and chemotaxonomic considerations concerning Chitean Maytenus. J Chem Ecol 1994;20:823-30.
35. Waller GR, In: Macias FA, Galindo JCG, Molinillo JMG, Cutler HG. Recent Advances in Allelopathy. Cadiz, Serv Pub Univ Cadiz; 1999.
36. Laszczyk MN. Pentacyclic triterpenes of the lupane, oleanane and ursane group as tools in cancer therapy. Planta Med 2009;15:1549-60.
37. Patocka J. Biollogically active pentacyclic triterpenes and their current medicine signification. J Appl Biomed 2003;1(1):7-12.
38. Santos LS, Santos JCL, Souza Filho APS, Corrêa MJC, Veiga TAM, Freitas VCM, et al. Allelopathic activity of chemical substances isolated from Brachiaria brizantha cv. Marandu and their variations in function of pH. Planta Daninha 2008;26(3):531-8.
39. Ferreira AG, Aquila MEA. Alelopatia: uma área emergente da ecofisiologia. Rev Bras Fisiol Veg 2000;12:175-204.
40. Petroski RJ, Stanley DW. Natural compounds for pest and weed control. J Agric Food Chem 2009;57:8171-9.
41. Hoffmann CEF, Neves LAS, Bastos CF, Wallau GL. Atividade alelopática de Nerium Oleander L e Dieffenbachia picta schott em sementes de Lactuca Sativa L e Bidens pilosa L. Rev Ciências Agroveterinárias 2007;6:11-21.
42. Azambuja N, Hoffmann CEF, Neves LAS, Goulart EPL. Potencial alelopático de Plectranthus barbatus Andrews na germinação de sementes de Lactuca sativa L e de Bidens pilosa L. Rev Ciências Agroveterinárias 2010;9:66-73.
43. Haida KS, Coelho SRM, Haas-Costa J, Viecelli CA, Alekcevetch JC, Barth EF. Efeito alelopático de Achillea millefolium L sobre sementes de Lactuca sativa L. Rev Agronegócios Meio Ambiente 2010;3:101-9.
44. Gussman AB, Pitelli RA, Dias SM. Efeito do citronelol sobre a germinação e desenvolvimento do amendoim bravo (Euphorbia heterophila L.). II Semina: Ciências Agrícola 1994;15:14-22.
45. Alves MCS, Filho SM, Innecco R, Torres SB. Alelopatia de extratos voláteis na germinação de sementes e no comprimento da raiz de alface. Pesquisa Agropecuária Brasileira 2004;39:1083-6.