Identification of Bioactive Compounds in Polar and Nonpolar Extracts of Araujia sericifera

Key words

Araujia sericifera - Asclepiadaceae - metabolite profile - cancer cell lines - GC-MS - NMR spectroscopy

Introduction

Araujia sericifera Brot. (Asclepiadaceae), also called the bladder flower, is a native perennial, climbing laticiferous shrub from South America that was introduced to other countries as an exotic ornamental plant. Currently, it is naturalized in Europe, South Africa, North America, Australia, and New Zealand. In the Mediterranean coastline, it competes with different crops, such as citrus trees, for water and nutrients [1] [2]. A. sericifera was reported to possess emetic, analgesic, antihistaminic, and anti-inflammatory properties [3] [4] [5]. Its seeds were described to exert toxicity on the central nervous system [3]. Its fruits contain luteolin-7-glucoside, serotonin [3], lupeol-3-cinnamate, and germanicol-3-acetate [4]. Nevertheless, despite the promising properties of A. sericifera, there is no systematic study of its chemical composition.

In a plant, different classes of secondary metabolites exert individual biological functions. They accumulate in specific tissues playing different roles in physiological processes or ecological interactions [6]. Metabolic profiling of different plant organs (root, shoot, leaves, etc.) has been a tool to achieve better authentication and chemotaxonomic analyses of plants [7]. In line with this, the metabolic composition of the different organs of A. sericifera, with a broad tolerance for environmental conditions, has attracted our attention. Several methods exist to detect and quantify plant metabolites, such as NMR spectroscopy- and MS-based techniques, the major analytical tools used for metabolite profile studies. Both analytical techniques are considered complementary to obtain optimal results, each with its own particular advantages [8]. The aim of this study was to analyze the composition and cytotoxic properties of both nonpolar (hexane) and polar (methanol) extracts of the three aerial parts of A. sericifera (leaves, fruits, and stems) using GC-MS and NMR spectroscopy methods in conjunction with multivariate statistical analyses in order to explore this poorly studied plant.

Results and Discussion

Polar and nonpolar metabolites were extracted by an automated Soxhlet extraction (Soxtec) of each aerial plant part with n-hexane and, subsequently, methanol to give the hexanic and methanolic extract, respectively. The leaves, stems, and fruits extracted with n-hexane gave extract yields (w/w), which were 5.6% for both leaves and fruits, and 2.2% for stems. The subsequent extraction with methanol gave higher extract yields for leaves and stems (21.0% and 16.4%, respectively) than for fruits (11.0%). In the hexanic extracts, 20 different metabolites were identified by a combination of GC-MS (after trimethylsilyl derivatization) and 1D and 2D NMR spectroscopy. These nonpolar metabolites consisted of fatty esters (FEs), fatty acids [FAs: palmitic (1), linoleic acid (2), oleic acid (3), stearic acid (4)], squalene (5), n-alkanes (6-8), β-sitosterol (9), β-amyrin-3-acetate (12, oleanane type), and chiefly triterpen-3-ols and/or their esters, such as lupeol (lupane type) and germanicol esters (oleanane type), and compounds 10, 11, and 13-19. For an unequivocal characterization of the lupeol and germanicol esters, an aliquot of the fruit hexanic extract was subjected to silica gel column chromatography and semipreparative reversed-phase HPLC purification. Several pentacyclic triterpenes could be isolated, including to new compounds, 3-O-tigloyl germanicol (18, oleanane type) and 3-O-tigloyl lupeol (19, lupane type) ([Fig. 1]). The structure elucidation of 18 and 19 was carried out by GC-MS, 1D and 2D NMR spectroscopy, and UPLC-Q-TOF analysis. The ¹H and ¹³C NMR spectra of compound 18 and 19 ([Table 1] and Fig. 3S, 4S, and 5S, Supporting Information) exhibited the typical signals of an (E)-tigloyl group; e.g., for compound 19 at δ _H (ppm) 6.83 (q, J=7.1 Hz, H-33), 1.83 (s, H₃-35) and 1.78 (d, J=7.1 Hz, H₃-34) and δ _C (ppm) 167.9 (CO), 136.5 (CH-33), 129.3 (C-32), 14.3 (CH₃-34) and 12.1 (CH₃-35). The HMBC spectrum of these compounds also showed a correlation from H-3 (δ _H 4.51 ppm) to the carbonyl carbon (CO) of the tigloyl group at δ _C 167.9 ppm. The GC-MS data also displayed a characteristic MS fragmentation of oleanane and lupane skeletons for compounds 18 and 19, respectively. Intense fragment ions were observed at m/z 204, 189, and 177 for compound 18 (Δ ¹⁸-oleanene with a methyl group at C-17). However, compound 19 with a C-20-29 double bond and a methyl group at C-17 presented a main peak at m/z 207 ([C₁₄H₂₃O]⁺, [A+B]⁺ rings) and at m/z 189 ([C₁₄H₂₁]⁺, [A+B]⁺ rings-H₂O). Fragmentation peaks at m/z 508, 493, and 408 corresponded to the molecular ion peaks [M]⁺, [M-CH₃]⁺, and [M-tiglic acid]⁺, respectively. The HRESIMS analysis for compounds 18 and 19 displayed an ion peak at m/z 409.3832 or 409.3829 [(M-C₅H₇O-H₂O) + H]⁺, respectively, which suggests the molecular formula C₃₀H₄₈ (calcd. 409.3829) for both compounds corresponding to the loss of the tigloyl group and one water molecule.

Fig. 1 Nonpolar secondary metabolites identified in hexanic extracts, germanicol type and lupeol type, and polar metabolites from methanolic extracts of A. sericifera.

Table 1 ¹H and ¹³C NMR data for 18 and 19 in CDCl₃ (δ in ppm, J in Hz).

	18		19
Position	δC (mult.)	δH (J in Hz)	δC (mult.)a	δH (J in Hz)
1a	37.3 (CH2)	1.50, m	38.4 (CH2)	1.69, m
1b				1.00, m
2a	23.4 (CH2)	1.65, m	23.7 (CH2)	1.47, m
2b		1.14, m		1.08, m
3	80.7 (CH)	4.52, d (6.4)	80.9 (CH)	4.51, dd (5.9)
4	37.7 (C)	−	38.1 (C)	−
5	55.2 (CH)	0.8, m	55.4 (CH)	0.80, m
6a	17.9 (CH2)	1.60, m	18.2 (CH2)	1.5, m
6b		1.38, m		1.45, m
7a	34.4 (CH2)	1.55, m	34.2 (CH2)	1.4, m
7b		1.41, m
8	43.1 (C)	−	40.9 (C)	−
9	51.0 (CH)	1.33, m	50.3 (CH)	1.35, m
10	37.2 (C)	−	37.1 (C)	−
11a	20.9 (CH2)	1.55, m	21.0 (CH2)	1.5, m
11b		1.40, m
12a	26.0 (CH2)	1.64, m	25.1 (CH2)	1.60, m
12b		1.25, m
13	38.2 (CH)	2.26, d (11.0)	38.1 (CH)	1.66, s
14	40.5 (C)	−	43.0 (C)	−
15a	27.2 (CH2)	1.72, m	27.4 (CH2)	1.46, m
15b		1.25, m		1.00, m
16	38.2 (CH2)	1.70, m	35.6 (CH2)	1.5, m
17	34.2 (C)	−	42.8 (C)	−
18	167.5 (C)	−	48.3 (CH)	1.35, m
19	129.7 (CH)	4.86, s	48.0 (CH)	2.38, ddd (10.0)
20	32.2 (C)		151.0 (C)	−
21a	33.3 (CH2)	1.60, m	29.8 (CH2)	1.92, m
21b		1.38, m		1.30, m
22a	37.3 (CH2)	1.49, m	40.0 (CH2)	1.45, m
22b				1.22, m
23	28.9 (CH3)	0.86, s	28.1 (CH2)	0.86, s
24	16.6 (CH3)	0.87, s	16.7 (CH3)	0.89, s
25	16.4 (CH3)	0.91, s	16.2 (CH3)	0.87, s
26	15.8 (CH3)	1.08, s	16.0 (CH3)	1.02, s
27	14.3 (CH3)	0.74, s	14.5 (CH3)	0.98, s
28	25.0 (CH3)	1.02, s	18.0 (CH3)	0.80, s
29a	28.8 (CH3)	0.97, s	109.3 (CH2)	4.69, d (5.8)
29b				4.57, s
30	31.2 (CH3)	0.94, s	19.3 (CH3)	1.69, s
31	167.7 (CO)	−	167.9 (CO)	−
32	129.1 (C)	−	129.3 (C)	−
33	136.1 (CH)	6.84, q (5.8)	136.5 (CH)	6.83, q (7.1)
34	14.1 (CH3)	1.79, d (5.8)	14.3 (CH3)	1.78, d (7.1)
35	11.9 (CH3)	1.83, s	12.1 (CH3)	1.83, s

Known triterpenes were also identified by making a direct comparison of their NMR spectroscopic and GC-MS data with those reported in the literature, including germanicol (10) [9], lupeol (11) [10], β-amyrin-3-acetate (12) [11], germanicol-3-acetate (13) [12], lupeol-3-acetate (14) [9], lupeol-3-cinnamate (15) [13], germanicol-3-propionate (16), and germanicol-3-butyrate (17) [14] ([Fig. 1]). Among them, lupeol-3-acetate (14), germanicol-3-propionate (16), and germinacol-3-butyrate (17) were identified in A. sericifera for the first time. A representative GC-MS chromatogram section of the three plant aerial parts containing principally triterpen-3-ols and/or their esters is drawn in [Fig. 2a]. In [Fig. 2b], the ¹³C NMR spectra of the hexanic extracts of leaves, stems, and fruits are depicted. The acquisition of ¹³C spectra seemed an optimal choice for these nonpolar fractions since the ¹H NMR spectra were highly overlapped due to the high amounts of alkanes and enough sample amount was available to obtain a good signal/noise ratio in the less sensitive ¹³C spectroscopy. It is worth mentioning that β-amyrin-3-acetate (12), present in very small amounts in leaves and stems, was only detected by the most sensitive technique GC-MS but not in the NMR spectra. In addition, only general signals of alkanes and FAs were assigned in the NMR spectra due to overlapping. On the other hand, NMR spectroscopy allowed for the quantification of several nonvolatile compounds such as long-chain FEs that could not be analyzed by GC-MS. Similarly, the quantification of lupeol-3-cinnamate (15) was only feasible by NMR spectroscopy since its peak overlapped with germanicol-3-propionate (16) in the GC-MS chromatogram (t _R=26.09 min). Relative percentages of metabolites from the hexanic extracts determined by GC-MS and ¹³C NMR in leaves, stems, and fruits are shown in Fig. 6S and 7S, Supporting Information. Thus, we found that the combination of NMR and GC-MS is optimal for the detection of nonpolar metabolites in A. sericifera.

Fig. 2 Identification of nonpolar metabolites from leaves, stems, and fruits of A. sericifera containing principally triterpen-3-ols and/or their esters. a A section of total ion GC-MS chromatograms of the TMS-derivatized hexanic extracts. b ¹³C NMR spectra of hexanic extracts at 600 MHz. Assignments: 8, tritiacontane; 9, β-sitosterol; 10, germanicol; 11, lupeol; 12, β-amyrin-3-acetate; 13, germanicol-3-acetate; 14, lupeol-3-acetate; 15, lupeol-3-cinnamate; 16, germanicol-3-propionate; 17, germanicol-3-butyrate; 18, 3-O-tigloyl germanicol; 19, 3-O-tigloyl lupeol; 20, cis-polyisoprene.

Concerning the polar fraction, 25 metabolites were identified in the methanolic extracts using GC-MS (after methoximation and trimethylsilyl derivatization) and NMR spectroscopy. In order to unambiguously identify several polar metabolites, an aliquot of the fruit methanolic extract was subjected to a solid-phase extraction (SPE) C₁₈ cartridge and was eluted firstly with 100% H₂O, followed by 100% MeOH. The methanolic fraction (F-2) was further purified by reversed-phase HPLC to give viburnitol (21, SF_meth2-1) [15], trigonelline (22, SF_meth2-2) [16], and conduritol F (23, SF_meth2-3, also known as L-leuchanthemitol) [17] ([Fig. 1]). The GC-MS analysis of the whole extracts also detected viburnitol (21), malate (37), L-asparagine (31), sucrose (38), glucose (40), fructose, and myo-inositol. A representative GC-MS chromatogram of the fruit methanolic extract is shown in Fig. 8S, Supporting Information. NMR spectroscopy confirmed the presence of metabolites detected by MS and further allowed for the identification of luteolin-7-glucoside (24), serotonin (25), allantoin (26), choline (28), and malate (37). In fact, luteolin-7-glucoside (24) and serotonin (25) were previously identified from the leaves and stems of A. sericifera [3], whereas virbutinol (21), trigonelline (22), conduritol F (23), and allantoin (26) were found in this plant for the first time. Representative ¹H NMR spectra of the leaf, stem, and fruit methanolic extracts are depicted in [Fig. 3], along with an assignment of the most significant metabolites.

Fig. 3 Representative ¹H NMR spectra of leaf, stem, and fruit methanolic extracts of A. sericifera. Metabolite keys are given in Table 1S, Supporting Information. Assignments: 22, trigonelline; 23, conduritol F; 24, luteolin-7-glucoside; 25, serotonin; 26, allantoin; 28, choline; 29, formate; 30, fumarate; 31, L-asparagine; 32, L-isoleucine; 33, L-leucine; 34, L-threonine/lactate; 35, L-tryptophan; 36, L-valine; 37, malate; 38, sucrose; 39, γ-butyric acid; 40, glucose; 41, L-alanine; 42, adenosine; 43, succinate; FA, fatty acids; FE, fatty esters. ■Please remove ,,a“ from figure, as there is no ,,b“■

In order to gain insight about how the identified compounds were distributed in the different plant parts, they were quantified from leaves, stems, and fruits (Fig. 9S, Supporting Information), and an unsupervised multivariate analysis in the form of principal component analysis (PCA) was performed. The resulting score and loading plots from the PCA of the GC-MS and ¹³C NMR data from nonpolar extracts are represented in [Fig. 4]. While the score plot ([Fig. 4a and c]) gives information about how the different samples cluster based on their metabolic composition, the loading plot ([Fig. 4b and d]) indicates which compounds are more present in each sample group. As we can see, a similar clustering was obtained by both analytical techniques. In both cases, large amounts of alkanes, saturated fatty acids/esters were associated with the stem extracts, while larger quantities of unsaturated fatty acids/esters, linoleic and linolenic acids and derivatives were found in the leaves. Cis-polyisoprene (20) and squalene (5) were identified only in the leaves, which also had greater amounts of β-sitosterol (9) and lupeol (11) than the other aerial parts. However, fruit extracts stood out for possessing germanicol (10) and large quantities of triterpene esters including germanicol-3-acetate (13), lupeol-3-acetate (14), lupeol-3-cinnamate (15), germanicol-3-propionate (16), 3-O-tigloyl germanicol (18), and 3-O-tigloyl lupeol (19).

Fig. 4 PCA of metabolites in the hexanic extract of leaves, stems, and fruits of A. sericifera analyzed by GC-MS and ¹³C NMR spectroscopy. a Score plot of GC-MS analysis, 2 components R2X (cum)=0.89, Q2 (cum)=0.79. b Loading plot of GC-MS analysis. c Score plot of ¹³C NMR analysis 2 components R2X (cum)=0.87, Q2 (cum)=0.77. d Loading plot of ¹³C NMR analysis.

For the polar metabolites, we quantified the identified compounds by ¹H NMR in the same set of samples of the three plant aerial parts and also performed PCA to compare the metabolite contents. The resulting score and loading plots are represented in [Fig. 5]. We observed that the stem extracts stand out for their high content in sugars such as sucrose (38) and glucose (40). Larger amounts of branched amino acids such as isoleucine (32), leucine (33). and valine (36) proved to be characteristic of the leaf extracts. Flavonoid luteolin-7-glucoside (24) was also significantly increased in the leaves. The metabolic composition of fruits differed vastly from that of the stems and leaves, standing out for high concentrations of viburtinol (21), trigonelline (22), conduritol F (23), serotonin (25), choline (28), succinate (43), or fumarate (30).

Fig. 5 Multivariate analysis of ¹H NMR spectra of polar methanolic extracts of A. sericifera, an unsupervised PCA model. a Score plot: two components R2X (cum)=0.63, Q2 (cum)=0.46. b Loading plot.

Thus, our study lays bare that A. sericifera, a climbing weed that competes with crops, contains several bioactive compounds with beneficial applications. Accordingly, nonpolar extracts possess high levels of triterpen-3-ol and their esters, especially in leaves and fruits, respectively. Naturally occurring and synthetic pentacyclic triterpenes exhibit a variety of unique biological activities, including antitumor, antiviral, antidiabetic, anti-inflammatory, antimicrobial, antiparasitic, cardio-, hepato- and gastro-protection, and analgesic and wound healing effects, among others. In fact, they are receiving ever-increasing interest as therapeutic agents in pharmacological research [18]. Lupeol (11) has the ability to inhibit α-amylases [19] and α-glucosidases [20] and possesses hypoglycemic and antidiabetic properties. In addition, germanicol (10) and lupeol (11) display anti-inflammatory [21] and antidyslipidemic effects [19]. Among the polar secondary metabolites of A. sericifera, trigonelline (22), which is abundant in leaves, has shown hypoglycemic, hypocholesterolemic, antitumoral, and antiseptic properties besides playing an essential role in the resistance process of plants against several pathogens [22]. Concerning the two natural cyclitols viburnitol (21) and conduritol F (23), compound 21 acts as a glycosidase inhibitor [23] and compound 23 has potential in the treatment of metastatic cancer and diabetes because of its ability to inhibit type I α-glucosidase [24]. Luteolin-7-O-glucoside (24) has an important nutraceutical application value thanks to its numerous biological properties, including antioxidant activity [25].

In order to evaluate the cytotoxic activity of A. sericifera plant extracts and the pentacyclic triterpenes, MTS assays were performed on human breast carcinoma (MDA-MB-453 and MCF-7) and human colon carcinoma (HCT-116) cell lines. As a result, the fruit hexanic extract of A. sericifera showed a significant cell growth reduction of HCT-116 and MCF-7 cells, while the effect produced by leaf extracts was more modest but with a higher impact against the aggressive cell line MDA-MB-453 ([Fig. 6a and b]). These results are coherent with the fact that the most bioactive compounds reported from nonpolar extracts occurred in these two plant organs. The cytotoxic activity for fruit hexanic extracts may be attributed to germanicol derivatives [26]. In leaves, high amounts of lupeol, a known inhibitor of proliferation [27], can induce cytotoxicity. In order to find out if 3-O-tigloyl lupeol (19) contributes to the cytotoxic activity of the nonpolar fruit and leaf extracts, it was semisynthesized by esterification between lupeol and tiglic acid. Nevertheless, compounds 15 and 19 did not show significant cell growth reduction at the tested concentrations, which was only obtained for lupeol (11) against the three cell lines ([Fig. 6c]). 3-O-Tigloyl germanicol (18) could not be assayed as not enough quantity was available. In the case of polar extracts of A. sericifera, a significant growth reduction was only detected in leaf extracts ([Fig. 6d]). This effect may be mainly due to the presence of luteolin-7-O-glucoside, which was previously described to inhibit proliferation of cancer cells [28].

Fig. 6 Cell growth inhibition obtained by the MTS assay against three different cell lines at two different concentrations.

In summary, our study of different aerial parts of A. sericifera by GC-MS and NMR spectroscopy allowed for the identification of interesting compounds with potential pharmaceutical and/or industrial uses. Several secondary metabolites have been identified in this plant for the first time, and two new triterpen-3-ol esters, 3-O-tigloyl germanicol (18) and 3-O-tigloyl lupeol (19), were characterized from the hexanic extracts. A systematic analysis of an array of plants showed that the major bioactive compounds were distributed in fruits and leaves. Accordingly, the new compound 3-O-tigloyl lupeol (19) did not show any cytotoxicity against the cancer cell lines at the tested concentrations. Therefore, the moderate cytotoxicity exhibited for both fruit and leaf hexanic extracts against the breast cancer model MCF-7 and the human colon carcinoma HCT-116 may be attributed to lupeol (11).

Materials and Methods

Cell lines, chemicals, and biochemicals

MDA-MB-453, MCF-7 (human breast carcinomas), and HCT-116 (human colon carcinoma) cell lines were provided by M. Orzáez (CIPF) [29], and grown at 37+°C in a humidified 5% CO₂, 95% air incubator. All the standard compounds and reagents (purities ≥94%) were purchased from Sigma-Aldrich with the exception of N,O-bis(trimethylsilyl)-trifluoroacetamide (BSTFA) with 1% trimethylchlorosilane (TMCS) (purity >98%), which was purchased from Acros. Hexane, dichloromethane, methanol, and pyridine were analytical grade and purchased from Scharlab SL. Chloroform-d with 0.05% v/v trimethylsilane (TMS), methanol-d ₄, and trimethylsilyl-2,2,3,3-tetradeuteropropionic acid (TSP; purities >99%)were purchased from Deutero GmbH.

Plant material

Aerial parts of A. sericifera (leaves, stems, and fruits) were harvested in September 2014 from adult plants that grew in an open field located in Moncófar (Castellón, eastern Spain). The botanical characterization of the plant was carried out by Prof. Herminio Boira from the Mediterranean Agroforesty Institute of the Polytechnic University of Valencia, and a voucher specimen was deposited at the VALA herbarium of the University Polytechnic of Valencia (Spain) with the registration number 6647.

Extraction and isolation

Oven-dried leaves, stems, and fruits were separated from each plant and, in addition, the fruits were cut and seeds removed. These aerial organs were finely powdered in a mechanical grinder and extracted by an automatic Soxhlet in a Foss Tecator Soxtec^TM system with a 2043 extraction unit and a 2046 Soxtec Foss Control unit. Each sample (5 g) was firstly extracted with n-hexane (50 mL) for 1 h 30 min of boiling time and 20 min of rising time, followed by MeOH (50 mL) under the same operating conditions [30]. The solvents were evaporated under reduced pressure to obtain the hexanic and methanolic extracts.

Dry leaf hexanic extracts (100 mg) were dissolved in MeOH (4 mL) and centrifuged at 3000 rpm for 8 min at 22+°C in order to partially remove the rubber content. The supernatant was separated and the pellet was dissolved in MeOH (4 mL) and then centrifuged again. This MeOH centrifugation procedure was performed three times for each sample [10]. The supernatants were combined and dried by nitrogen stream to give a residue of leaf hexanic extract (100 mg) that was subjected to flash column chromatography on silica gel 60 (40–63 μm) using a stepwise gradient soln. solution from 100% hexane to 100% ethyl acetate and yielding five fractions. The major fraction F-2 (hexane/EtOAc, 9:1, 13 mg) was purified by a semipreparative reversed-phase HPLC instrument (Waters 600E system) equipped with a solvent delivery pump unit (Waters 600E) coupled to a photodiode array detector (Waters 2996 PDA) and an evaporative light scattering detector (Waters 2420 ELSD). The separation of metabolites was carried out using a Phenomenex Luna C18 (2) column (25.0×1 cm, 5 µm) and 100% methanol as the mobile phase (flow of 3 mL/min) to afford SF_hex2-1 (t _R=30 min), SF_hex2-2 (t _R=35 min), SF_hex2-3 (t _R=42 min), SF_hex2-4 (t _R=45 min), SF_hex2-5 (t _R=48 min), and SF_hex2-6 (t _R=52 min). On the other hand, a residue of the fruit methanolic extract (85 mg) was subjected to the solid-phase extraction (SPE) C₁₈ cartridge (2 g, 12 ml) model ExtraBond (Scharlab) with 100% H₂O followed by 100% MeOH. Fraction F-2 (100% MeOH) was purified by semipreparative reversed-phase HPLC using a Phenomenex Luna C18 (2) column (25.0×1 cm, 5 µm) and 100% H₂O as the mobile phase (flow of 3 mL/min) to afford the SF_meth2-1 (t _R=6 min), SF_meth2-2 (t _R=9 min), and SF_meth2-3 (t _R=11 min). All purified fractions from both hexanic and methanolic extracts were dried, derivatized, and analyzed by GC-MS and NMR spectroscopy.

3-O-Tigloyl germanicol (18): white amorphous powder; m.p. 244-249+°C; [α]_D ²⁵+27.2 (c 0.5, CHCl₃); IR (film): 2940, 1710, 1634, 1446 cm^-1; ¹H and ¹³C NMR, see [Table 1]; HRESIMS m/z 409.3832 [(M-C₅H₇O-H₂O)+H]⁺ (calcd. for C₃₀H₄₈, 409.3829).

3-O-Tigloyl lupeol ( 19): white amorphous powder; m.p. 233–237+°C; [α]_D ²⁵+42.5 (c 0.5, CHCl₃); IR (film): 2987, 1710, 1654, 1457 cm⁻¹; ¹H and ¹³C NMR, see [Table 1]; HRESIMS m/z 409.3829 [(M-C₅H₇O-H₂O)+H]⁺ (calcd. for C₃₀H₄₈, 409.3829).

Derivatization and GC-MS analysis

Samples of the hexanic extracts were derivatized according to Van Beek [31]. Trimethylsilyl derivatives of hexanic extract samples (5 mg) were prepared by the addition of 100 μL of dry CH₂Cl₂ and 100 µl of BSTFA with 1% TMCS. Next, the mixture was stirred overnight at room temperature. Samples of the methanolic extracts were derivatized according to Herebian et al. [32]. Derivatives of the methanolic extracts were prepared by a two-step procedure involving a methoximation-trimethylsilylation process. For methoximation, the extract samples (5 mg) were treated with methoxyamine hydrochloride soln. solution (20 mg mL^-1 in pyridine) (100 µL) and stirred overnight at room temperature. Next, BSTFA with 1% TMCS (100 µL) was added as a silylation reagent and stirred for 3 h at room temperature. Derivatized samples were directly analyzed by GC-MS on a PerkinElmer Clarus^® 500 gas chromatograph-mass spectrometer, operating in the electron impact mode (EI) and equipped with a ZB-5 MS (30 m×0.25 mm×0.25 μm particle size) capillary column (Phenomex Inc). GC-MS parameters, peaks identification, and quantification procedure are described in more detail in the Supporting Information.

UPLC-Q-TOF analysis

High-resolution mass (HRESIMS) spectra for the new compounds were obtained from a Triple TOF^TM 5600 hybrid quadrupole time-of-flight (TOF) LC-MS/MS system (AB SCIEX) and a Waters Acquity BEH C18 column (50×2.1 mm i.d., 1.7 μm). UPLC-Q-TOF parameters are described in more detail in the Supporting Information.

General procedure for the preparation of 3-tigloyl lupeol (19)

The esterification procedure was carried out according to Liu et al. [33]. Lupeol (6.0 mg, 1 equiv) was added to a soln. solution of trans-2-methyl-2-butenoic acid (1.4 mg, 1 equiv), DCC (5.8 mg, 2 equiv), and DMAP (3.4 mg, 2 equiv) in dry CH₂Cl₂ (2 ml), and the reaction mixture was shaken for 24 h. Next, the solvent was removed under reduced pressure to give a white solid that was redissolved in CH₂Cl₂, washed with 5% NaHCO₃, dried over Na₂SO₄, and concentrated under reduced pressure to obtain a crude residue. The residue was purified by semipreparative reverse-phase HPLC-PDA-ELSD using a Phenomenex Luna C18 (2) column (25.0×1 cm, 5 µm) and 100% methanol as the mobile phase with a flow rate of 3 mL min^-1. 3-O-Tigloyl lupeol (19) was obtained in a 63% yield as white powder.

NMR spectroscopy

NMR spectra were recorded at 25+°C on a Bruker AVII-600 using a 5-mm TCI cryoprobe (¹³C spectra) and a Bruker AVIII-500 using a 5-mm TBI probe (all the other experiments) and processed using Topspin 3.17 software (Bruker GmbH). Nonpolar extracts (20 mg) were dissolved in 500 µL of 99.8% chloroform-d with 0.05% v/v TMS. Polar extracts were dissolved in 500 µl of 50% methanol-d ₄ in buffer [90 mM KH₂PO₄, pH=6, 1 mM trimethylsilyl-2,2,3,3-tetradeuteropropionic acid (TSP)]. Chemical shifts (δ) of ¹H and ¹³C NMR are given in ppm. The acquisition parameters of NMR spectra and quantification procedure are described in more detail in the Supporting Information.

Principal component analysis

Variable sized bucketing was performed of the ¹H and ¹³C NMR spectra, excluding the regions corresponding to CDCl₃ (in nonpolar samples) and to D₂O and MeOD resonances (in polar samples). After normalization and univariate scaling, data were subjected to PCA, an unsupervised pattern recognition method [34], using the software Simca-P+ 12.0. PCA score plots show clustering trends between samples, how metabolite concentrations are related, and if there are any strong outliers. The discriminant metabolites were identified from the corresponding loading plots.

MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium]assay with cancer model cell lines

The effect of organic leaf, fruit, and stem extracts and lupeol (purity ≥94%; Sigma), lupeol-cinnamate, and 3-O-tigloyl lupeol compounds was evaluated by the standard procedure of the MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] cell viability assay [35] in HCT-116, MFC-7, and MDA-MB-453 cells lines. A stock solution of extracts and compounds (2 mg/mL) in DMSO was prepared. This stock solution was diluted in culture medium (DMEM for HCT-116 and MFC-7, or DMEM/F12 for MDA-MB-453) to obtain of 32, 64, 96, and 120 µg/µL solutions (DMSO, maximal concentration 0.1%). A solution of medium containing only DMSO at 0.1% was also prepared. Cells were harvested in their logarithmic phase and seeded at concentrations of 12500 (HCT-116), 5000 (MCF-7), and 10000 (MDA-MB-116) cells per well in 96-well microtiter plates and incubated for 18 h in 50 µL of their respective media at 37+°C and 5% CO₂. Cells were then observed under a light microscope (20×) to check if they were attached to the plates. Fifty µL of medium were added with the absence (only DMSO at a final concentration of 0.05%) or presence of compounds/extracts to yield final compound/extract concentrations of 16, 32, 48, and 60 μg/μL per well, and cells were incubated for 72 h. MTS was freshly prepared at 5 mg/mL in PBS, and 20 µL of a mixture of MTS solution and phenazine methosulfate (20:1) were added to each well and incubated at 37+°C for another 3 h. Finally, the absorbance was measured with a spectrophotometer (VICTOR2 1420 Multilabel HTS Counter) at 595 nm. Results are the mean of three independent experiments (n=3).

• References

• 1 Spellman DL, Gunn CR. Morrenia odorata and Araujia sericofera (Asclepiadaceae): weeds in citrus groves. Castanea 1976; 41: 139-148
• 2 Forster PI, Bruyns PV. Clarification of synonomy for the common moth-vine Araujia sericifera (Asclepiadaceae). Taxon 1992; 41: 746-749
• 3 Federici E, Galeffi C, Nicoletti M. Constituents of Araujia sericofera . J Nat Prod 1988; 51: 189-190
• 4 López de Medrano-Villar MJ, Bello R, Esplugues J. et al. Triterpenoid compounds from Araujia sericofera B. Effects on the isolated guinea pig ileum. Methods Find Exp Clin Pharmacol 1997; 19: 515-520
• 5 Bello R, Barrachina MD, Esplugues J. et al. Analgesic activity and effects on isolated smooth muscle of different fractions of hexane extract from Araujia sericofera Brot. Phytother Res 1996; 10: 337-339
• 6 Matsuda F, Yonekura-Sakakibara K, Niida R. et al. MS/MS spectral data tag-based annotation of non-targeted profile of plant secondary metabolites. Plant J 2009; 57: 555-577
• 7 Frederich M, Choi YH, Angenot L. et al. Metabolomic analysis of Strychnos nux-vomica, Strychnos icaja and Strychnos ignatii extracts by 1H nuclear magnetic resonance spectrometry and multivariate analysis techniques. Phytochem 2004; 65: 1993-2001
• 8 Barding GA, Béni S, Fukao T. et al. Comparison of GC-MS and NMR for metabolite profiling of rice subjected to submergence stress. J Proteome Res 2013; 12: 898-909
• 9 Sholichin M, Yamasaki K, Kasai R. et al. 13C Nuclear magnetic resonance of lupane-type triterpenes, lupeol, botulin and betulinic acid. Chem Pharm Bull 1980; 28: 1006-1008
• 10 Palomino-Schätzlein M, Escrig PV, Boira H. et al. Evaluation of nonpolar metabolites in plant extracts by 13C NMR spectroscopy. J Agric Food Chem 2011; 59: 11407-11416
• 11 Ebajo Jr. VD, Shen CC, Ragasa CY. Terpenoids and Sterols from Hoya multiflora Blume. J Appl Pharm Sci 2015; 5: 33-39
• 12 Dat NT, Cai XF, Bae K. et al. Terpenoid constituents from Youngia koidzumiana. Nat Prod Sci 2002; 8: 55-57
• 13 Wood CA, Lee K, Vaisberg AJ. et al. A bioactive spirolactone iridoid and triterpenoids from Himatanthus sucuuba. Chem Pharm Bull 2001; 49: 1477-1478
• 14 Rojas LB, Grignon-Dubois M, Rezzonico B. et al. Pentacyclic triterpenes from Sarcostemma clausum. Chem Nat Comp 2004; 40: 565-568
• 15 Boutaghane N, Voutquenne-Nazabadioko L, Simon A. et al. A new triterpenic diester from the aerial parts of Chrysanthemun macrocarpum. Phytochem Lett 2013; 6: 519-525
• 16 Le Gall GL, Colquhoun IJ, Davis AL. et al. Metabolite profiling of tomato (Lycopersicon esculentum) using 1H NMR spectroscopy as a tool to detect potential unintended effects following a genetic modification. J Agric Food Chem 2003; 51: 2447-2456
• 17 Abe F, Yamauchi T, Honda K. et al. Conduritol F glucosides and terpenoid glucosides from Cynanchum liukiuense and distribution of conduritol F glucosides in several Asclepiadaceous plants. Chem Pharm Bull 2000; 48: 1090-1092
• 18 Sheng H, Sun H. Synthesis, biology and clinical significance of pentacyclic triterpenes: a multi-target approach to prevention and treatment of metabolic and vascular diseases. Nat Prod Rep 2011; 28: 543-593
• 19 Papi Reddy K, Singh AB, Puri A. et al. Synthesis of novel triterpenoid (lupeol) derivatives and their in vivo antihyperglycemic and antidyslipidemic activity. Bioorg Med Chem Lett 2009; 19: 4463-4466
• 20 Ortiz-Andrade RR, García-Jiménez S, Castillo-España P. et al. α-Glucosidase inhibitory activity of the methanolic extract from Tournefortia hartwegiana: an anti-hyperglycemic agent. J Ethnopharmacol 2007; 109: 48-53
• 21 Padilha MM, Vilela FC, Rocha CQ. et al. Antiinflammatory properties of Morus nigra leaves. Phytother Res 2010; 24: 1496-1500
• 22 Machado ART, Lage GA, Medeiros FS. et al. Quantitative analysis of trigonelline in some Annona species by proton NMR spectroscopy. Nat Prod Bioprospect 2013; 3: 158-160
• 23 Angelaud R, Babot O, Charvat T. et al. Desymmetrization of cyclohexadienylsilanes. Regio-, diastereo-, and enantioselective access to sugar mimics. J Org Chem 1999; 64: 9613-9624
• 24 Worawalai W, Rattanangkool E, Vanitcha A. et al. Concise synthesis of (+)-conduritol F and inositol analogues from naturally available (+)-proto-quercitol and their glucosidase inhibitory activity. Bioorg Med Chem Lett 2012; 22: 1538-1540
• 25 Benavente-García O, Castillo J, Lorente J. et al. Antioxidant activity of phenolics extracted from Olea europaea L. leaves. Food Chem 2000; 68: 457-462
• 26 Gutiérrez-Nicolás F, Gordillo-Román B, Oberti JC. et al. Synthesis and anti-HIV activity of lupane and olean-18-ene derivatives. Absolute configuration of 19,20-epoxylupanes by VCD. J Nat Prod 2012; 75: 669-676
• 27 Liu Y, Tingting B, Wang G. et al. Lupeol inhibits proliferation and induces apoptosis of human pancreatic cancer PCNA-1 cells through AKT/ERK pathways. Naunyn-Schmiedeberg's Arch Pharmacol 2015; 388: 295-304
• 28 Hwang YJ, Lee EJ, Kim HR. et al. Molecular mechanisms of luteolin 7-O-glucoside-induced growth inhibition on human liver cancer cells: G2/M cell cycle arrest and caspase-independent apoptotic signaling pathways. BMB Rep 2013; 46: 611-616
• 29 Andreu-Fernández V, Sancho M, Genovés A. et al. Bax transmembrane domain interacts with prosurvival Bcl-2 proteins in biological membranes. PNAS 2007; 114: 313-315
• 30 Nyström L, Lampi AM, Rita H. et al. Effects of processing on availability of total plant sterols, steryl ferulates and steryl glycosides from wheat and rye bran. J Agric Food Chem 2007; 55: 9059-9065
• 31 Van Beek TA. Chemical analysis of Ginkgo bioloba leaves and extracts. J Chromatogr A 2002; 967: 21-55
• 32 Herebian D, Hanisch B, Marner FJ. Strategies for gathering structural information on unknown peaks in the GC/MS analysis of Corynebacterium glutamicum cell extracts. Metabolomics 2005; 1: 317-324
• 33 Liu XK, Ye BJ, Wu Y. et al. Synthesis and anti-tumor evaluation of panaxadiol derivatives. Eur J Med Chem 2011; 46: 1997-2002
• 34 Jackson JE. A user’s guide to principal components. New York: Wiley; 1991
• 35 Barltrop JA, Owen TC, Cory AH. et al. 5-(3-Carbomethoxyphenyl)-2-(4,5-dimethylthiazolyl)-3-(4-sulfophenyl)tetrazolium, inner salt (MTS) and related analogs of 3-(4,5-dimethylthiazolyl)-2,5-diphenylterazolium bromide (MTT) reducing to purple water-soluble formazans as cell viability indicators. Bioorg Med Chem Lett 1991; 1: 611-614