Aryldihydronaphthalene-type lignans from Bursera fagaroides var. fagaroides and their antimitotic mechanism of action

Abstract

Three new aryldihydronaphthalene-type lignans, namely, 7′,8′-dehydropodophyllotoxin (1); 7′,8′-dehydro acetylpodophyllotoxin (2); and 7′,8′-dehydro trans-p-cumaroylpodophyllotoxin (3), were isolated from the stem bark of Bursera fagaroides var. fagaroides (Burseraceae), together with six known lignans, podophyllotoxin (4), acetylpodophyllotoxin (5), 5′-desmethoxy-β-peltatin A methylether (6), acetylpicropodophyllotoxin (7), burseranin (8), and hinokinin (9). The coumarin scopoletin (10) was also isolated from this bark. The chemical structure of all these compounds was determined by spectroscopic analyses including 2D NMR. We demonstrated that compounds 1–3 show different degrees of cytotoxic activity against human nasopharyngeal (KB), colon (HF-6), breast (MCF-7) and prostate (PC-3) cancer cell lines, with IC₅₀ values ranging from 1.49 to 1.0 × 10⁻⁵ μM. In vivo studies of the effect of these natural lignans on the cell cycle, cell migration and microtubule cytoskeleton of developing zebrafish embryos, demonstrated their antimitotic molecular activity by disturbing tubulin. This is the first report on the occurrence of aryldihydronaphthalene lignans in the genus Bursera of the Burseraceae family, as well as on the determination of their cytotoxic activity and mechanism of action.

---

1. Introduction

Bursera fagaroides (Kunth) Engl, which is a small tree belonging to the genus Bursera of the Buseraceae family B. fagaroides, is distributed from the United States to south Mexico, and is locally known as “aceitillo”, “copal” and “sarzafrás”.¹ Its stem bark and exudates are both used in folk medicine to treat cuts and tumors. Previous studies have demonstrated its effects on mouse and human spermatozoa immobilization,² as well as activities as an amoebicide,³ immunomodulator and antitumoral medication.^4,5 Previous phytochemical studies showed the presence of podophyllotoxin related lignans in its bark.^5,6 Recently, a bioassay-guided phytochemical investigation revealed the presence of seven aryltetralin-type lignans as the cytotoxic components of the hydroalcoholic extract obtained from the stem bark of this plant.⁷ In addition, their antimitotic activity by disturbing tubulin was demonstrated using the zebrafish embryo model.⁸

In view of the biological importance of compounds obtained from Bursera, and in continuation of our search for plant cytotoxic natural compounds,^9–11 we conducted a detailed investigation of the dichloromethane extract from the stem bark of B. fagaroides. Herein, we describe the isolation, and structure elucidation of nine podophyllotoxin-type lignans (1–9), together with the coumarin scopoletin (10). Compounds 1–3 are new natural products with an aryldihydronaphthalene skeleton, which are reported for the first time in the genus Bursera. The isolated pure compounds (1–3) were assayed for their cytotoxic activity, and their effect on the cell cycle using the in vivo model of zebrafish embryos was determined. The obtained results suggest the necessity of further investigations with these compounds, to develop moieties of therapeutic use.

2. Results and discussion

In a previous study, the CH₂Cl₂ extract from the stem bark of Bursera fagaroides revealed significant cytotoxic activity against a panel of four cancer cell lines with IC₅₀ values ranging from 0.96 to 3.71 μg mL⁻¹.¹² In this study, repeated column chromatography resulted in the isolation of three new aryldihydronaphthalene-type lignans, namely, 7′,8′-dehydropodophyllotoxin (1), 7′,8′-dehydro acetyl podophyllotoxin (2), and 7′,8′-dehydro-trans-p-coumaroyl podophyllotoxin (3), together with seven known compounds.

Compound 1 was obtained as an amorphous powder, mp 271–273 °C [α]²⁵_D −95.0 (c 0.090, CHCl₃). The HRFABMS of 1 showed the quasimolecular ion peak at m/z 413.1301 [M + H]⁺. Mass and NMR data (Table 1) provided its molecular formula as C₂₂H₂₁O₈ [M + H]⁺ (calcd for C₂₂H₂₁O₈ [M + H]⁺ 413.1191), which indicates 13 degrees of unsaturation. The IR spectrum of compound 1 showed that it contained absorption bands for hydroxyl (3440.0 cm⁻¹) and ester (1757.5, 1709.3 cm⁻¹) groups, as well as those for aromatic rings (1633.4, 1608.6, and 1588.7 cm⁻¹).

Table 1 ¹H-NMR (400 MHz) and ¹³C-NMR (100 MHz) spectral data for compounds 1–3^c

Position	1 ^a		2 ^b		3 ^a
	δH	δC	δH	δC	δH	δC
a NMR of compounds 1 and 2 were obtained in CDCl3-DMSOd6.b NMR of compound 3 was obtained in CDCl3.c Chemical shift values are in ppm and J values (in Hz) are presented in parentheses. Assignments are based on HSQC, HMBC and 1H–1H TOCSY experiments.
1		12.2		130.6		129.8
2		129.5		132.0		129.1
3	6.52	110.0	6.47, s	110.0	6.54, s	110.2
4		150.1		150.5		150.0
5		147.5		148.4		147.9
6	7.2	104.8	6.91, s	105.8	6.83, d (1.2)	105.1
7	4.89 (d, 14)	74.2	6.17, dd (14, 1.2)	75.2	6.31, d (13.2)	74.6
8	3.37, dt (14, 8.4)	43.9	3.48, dt (14.0, 8.8)	42.2	3.64, dt (14.0, 8.8)	41.7
9a	4.76, dd (9.2, 9)	70.1	4.62, t (8.8)	70.0	4.63, t (8.8)	69.6
9b	4.26, dd (9.6, 8.4)		4.22, t (8.8)		4.33, t (8.8)
1′		129.1		130.5		131.2
2′	6.52, s	110.0	6.44, s	109.6	6.54, s	110.3
3′		153.0		153.7		153.1
4′		135.7		139.5		138.8
5′		153.0		154.0		153.1
6′	6.52, s	110.0	6.44, s	109.6	6.54, s	110.3
7′		147.9		147.2		147.7
8′		118.4		119.7		118.0
9′		168.0		167.7		167.7
O–CH2–O	5.991, d (0.8)	102.0	6.05, d (0.8)	103.1	5.99, d (1.2)	102.1
	5.998, d (0.8)
C-3′-OMe	3.75, s	58.6	3.79	56.6	3.85, s	56.5
C-4′-OMe	3.92, s	61.2	3.70	60.7	3.92, s	61.2
C-5′-OMe	3.75, s	58.6	3.79	56.6	3.85, s	56.5
C̲H̲3CO			2.3, s	21.0
CH3C̲O̲				171.5
C-1′′						127.1
C-2′′					7.51, d (8.8)	130.6
C-3′′					6.88, d (8.8)	116.3
C-4′′						158.5
C-5′′					6.88, d (8.8)	116.3
C-6′′					7.51, d (8.8)	130.6
C-7′′					7.8, d (15.6)	146.8
C-8′′					6.47, d (15.6)	114.2
C-9′′						167.1

---

Its ¹³C NMR spectrum (Table 1) displayed occurrence of 21 carbon resonances ascribable to one carboxyl group (δ_C 168.0), one tetrasubstitued double bound (δ_C 147.9, and 118.4), one methylenedioxy group (δ_C 102.0), three methoxyl groups (δ_C 58.6 (×2) and 61.2), one aliphatic methylene group (δ_C 70.1), two aliphatic methine groups (δ_C 74.2, and 43.9), and 12 aromatic carbons (δ_C 102.0–153.0) due to two benzene rings.

Its ¹H NMR spectrum (Table 1) showed signals of one unsymmetrical 1,2,4,5-tetrasubstituted [δ_H 7.2, 1H, s, and 6.52, 1H, s] aromatic ring, one symmetrical 1,3,4,5-tetrasubstituted (δ_H 6.52, 2H, s) aromatic ring, one oxygenated methine [δ_H 4.89, 1H, d, J = 14.0 Hz], one oxygenated methylene [δ_H 4.76 (1H, dd, J = 9.2, 9.0 Hz) and 4.26 (1H, dd, J = 9.6, 8.4 Hz)], three methoxyls [δ_H 3.92 (3H, s) and 3.75 (6H, s)], and one methylenedioxy [δ_H 5.991 (1H, d, J = 0.8 Hz) and 5.998 (1H, d, J = 0.8 Hz)]. ¹H-NMR (Table 1) and ¹H/¹HCOSY spectra revealed the presence of a 7′,8′-dihydronaphthalene-type lignan skeleton by the four proton spin system [δ_H 4.89 (1H, d, J = 14.0 Hz, H-7), 4.76 (1H, dd, J = 9.2, 9.0 Hz, H-9a), 4.26 (1H, dd, J = 9.6, 8.4 Hz, H-9b), and 3.37 (1H, dt, J = 14.0, and 8.4 Hz, H-8)] (Fig. 1), which was supported by the three aliphatic resonances in the ¹³C NMR spectrum [δ_C 74.2 (CH, C-7), 43.9 (CH, C-8), and 70.1 (CH₂, C-9)], as well as by the presence of a tetrasubstituted double bond [δ_C 147.9 (C, C-7′) and 118.4 (C, C-8′)].

Fig. 1 Structures of 1–10 from the stem bark of B. fagaroides.

In addition, the absorption bands at λ_max = 247 (1.89) and 345 (0.77) nm in the UV spectrum accounted for the dihydronaphthalene skeleton.¹³

In the HMBC of compound 1 (Fig. 2), the aromatic proton signal at δ_H 6.52 and two methoxy proton signals (δ_H 3.92 and 3.75) showed correlations with carbon resonances at δ_C 135.7 and 153.0, which indicate the presence of methoxy groups at C-3′, C-4′ and C-5′. The signals for methoxy groups at δ_H 3.75 (6H) were attributed to C-3′ and C-5′, while the signal for methoxy group at δ_H 3.92 (3H) was attributed to C-4′ because of the correlations observed between those protons and the carbon resonances at δ_C 153.0 (C-3′ and C-5′) and δ_C 135.7 (C-4′). The position of the hydroxyl group at C-7 was confirmed by the HMBC correlations δ_H 7.2 (H-6)/δ_C 74.2 (C-7) and δ_H 4.76 and 4.26 (H-9)/δ_C 74.2 (C-7). Furthermore, the tetrasubstituted double bond was placed on C-7′/C-8′ based on the correlations δ_H 3.37 (H-8)/δ_C 147.8 (C-7′), δ_H 4.76 (H-9)/δ_C 118.4 (C-8′), and δ_H 6.52 (H-3)/δ_C 147.9 (C-7′) (Fig. 1). The large coupling value (J = 14.0 Hz) between H-7 and H-8 is consistent with a trans diaxial relationship between these protons. The CD spectrum of compound 1 showed a positive Cotton effect at Δε 358 (1.68) nm, and a negative Cotton effect at Δε 251 (−3.49) nm, which suggest an 8R configuration.¹⁴ This is supported also by the fact that the other aryltetralin lignans (4–6) that were isolated in this study have the 7R,8R configuration. Based on these findings, the structure of this compound was determined to be (7R,8R)-7′,8′-dehydro-podophyllotoxin (1), which is a new natural product.

Fig. 2 Selected HMBC (H → C) and COSY (H → H) correlations of compounds 1–3.

Compounds 2 and 3 were obtained as amorphous powders. Their HRFABMS spectra showed quasimolecular ion peaks at m/z 455.1324 [M + H]⁺ (C₂₄H₂₃O₉) for 2, and at m/z 559.1586 [M + H]⁺ (C₃₁H₂₇O₁₀) for 3. Analysis of the NMR data (Table 1) indicated that 2 and 3 share many structural features with 7′,8′-dehydropodophyllotoxin (1). The only significant differences between the NMR spectra were the presence of an additional acetate group in 2 and a trans-p-cumaroate group in 3. This was evidenced by the downfield shift of the oxymethine proton H-7 at δ_H 6.17 (dd, J = 14.0, 1.2 Hz) in 2 and at δ_H 6.31 (d, J = 13.2 Hz) in 3. This signal appears in 1 at δ_H 4.89 (1H, d, J = 14.0 Hz), which indicates that acyl substitution occurred at C-7 in 2 and 3. The presence of a methyl signal at δ_H 2.3 (3H, s) in the ¹H NMR spectra and acetate carbon signals at δ_C 171.5 and 21.0 in the ¹³C NMR spectra of 2 support this assumption. In addition, the signals due to an A₂B₂ system [δ_H 7.51 (1H, d, J = 8.8, Hz), 6.88 (1H, d, J = 8.8 Hz)], corresponding to a 1,4-disubstituted aromatic ring, and one trans-double bond at δ_H 7.8 (1H, d, J = 15.6 Hz) and 6.47 (1H, d, J = 15.6 Hz) were also revealed in compound 3, thus indicating the presence of an additional trans-p-coumaroyl group with respect to 1.

Detailed analysis of the 1D and 2D NMR data of 2 and 3 enabled the assignment of all the ¹H and ¹³C NMR signals. Their ¹H–¹H COSY spectra confirmed the presence of the aliphatic spin system formed by the protons H-9/H-8/H-7 in both compounds (Fig. 1). HMBC correlations between H-7 and C-1′′ and that between H-6 and C-7, allowed us to place the ester group at C-7, and correlations between H-8/C-7′, H-7/C-8′, and H-9/C-8′ confirmed the dihydronaphthalene skeleton (Fig. 2). In addition, the absorption bands at λ_max = 266 (2.19) and 348 (1.61) nm in the UV spectrum of compound 2 are in accordance with a dihydronaphthalene skeleton.¹³

As performed for 1, the relative stereochemistry of 2 and 3 was also determined by the analysis of the J values obtained from the ¹H NMR spectrum, and the absolute configuration was determined to be 7(R),8(R) because of the positive Cotton effect at 354 (26.0) and the negative Cotton effect at 252 (−44.45) nm observed in the CD spectra of compound 2. In addition, mild basic hydrolysis (Na₂CO₃ : MeOH : H₂O) of 2 and 3 afforded 1; the reactions were identical in all aspects with the natural product, thus confirming their identities. Based on these findings, the structure of these compounds was determined to be (7R,8R)-7′,8′-dehydro acetyl podophyllotoxin (2) and (7R,8R)-7′,8′-dehydro trans-p-cumaroyl podophyllotoxin (3), which represent new natural products.

The chemical structures of the known compounds were identified as podophyllotoxin (4),⁷ acetylpodophyllotoxin (5),⁷ 5′-desmethoxy-β-peltatin A methylether (6),⁷ acetylpicropodophyllotoxin (7),¹⁵ burseranin (8),⁷ hinokinin (9),¹⁶ and the coumarin scopoletin (10),¹⁷ by comparison of their physicochemical and spectroscopic data (IR, MS, 1D and 2D NMR) with that reported in the literature.

Podophyllotoxin (4), acetylpodophyllotoxin (5), 5′-desmethoxy-β-peltatin A methylether (6), and burseranin (8) were previously isolated from this plant species, and their cytotoxic activity has been described.⁷ Furthermore, in vivo effects over the cell cycle, cell migration and microtubules on zebrafish embryos demonstrated that these natural lignans act by disturbing tubulin.⁸ Acetylpicropodophyllotoxin (7) is a constituent of Hernandia ovigera and it is a potent, selective inhibitor of the type I insulin-like growth factor receptor (IGF-IR).¹⁸ Its desacetyl derivative is currently used with notable success in clinical trials that include patients with aggressive types of epithelial tumors.¹⁹ Hinokinin (9) has been isolated from several plant species and its cytotoxic activity against several cell lines has been documented.²⁰ Scopoletin (10) has been isolated from several plant species, and this coumarin induces cell cycle arrest and increases apoptosis in human prostate tumor cells and human leukemia cell line via the activation of caspase-3.²¹

Evaluation of the cytotoxic activity of compounds 1–3 against the cancer cell lines KB, PC-3, MCF-7, and HF-6 showed that they all display good activity against KB, PC-3 and HF-6, but were not active against the MCF-7 cell line (Table 2). When compared with podophyllotoxin (4) (IC₅₀ = 2.10 × 10⁻⁴ μM), compounds 1 and 3 were the most active against the PC-3 cell line, displaying similar toxicities (IC₅₀ = 2.4 × 10⁻⁵ and 2.42 × 10⁻⁵ μM, respectively), while compound 2 was the least active one (IC₅₀ = 0.06 μM). Lignans 1–3 showed moderate activity against the KB and HF-6 cell lines when compared to 4.

Table 2 IC₅₀ values (μM) of compounds 1–3 isolated from B. fagaroides^a

Compound	KB	PC-3	MCF-7	HF-6
a KB: nasopharyngeal cancer, PC-3: prostate cancer, MCF-7: breast cancer, and HF-6: colon cancer.
1	0.25 ± 0.0025	2.42 × 10^−5 ± 0.004	>9.7	0.012 ± 0.008
2	0.297 ± 0.0006	0.061 ± 0.0089	>8.8	0.066 ± 0.055
3	3.61 ± 0.08	2.42 × 10^−5 ± 0.0036	>7.2	1.27 ± 0.027
Camptothecin	4.54 × 10^−3 ± 0.009	2.758 ± 0.006	3.67 × 10^−4 ± 0.015	1.58 × 10^−5 ± 0.007
Podophyllotoxin	2.10 × 10^−4 ± 0.003	2.05 ± 0.009	2.30 × 10^−4 ± 0.005	0.018 ± 0.05

---

To measure the effect of natural lignans 1–3 over the cell cycle and morphology on the zebrafish (Danio rerio) model, we used the serine 10-phosphorylated modification of histone H3 (H3S10ph) as a mitotic marker that was detected by indirect immunofluorescence on whole zebrafish embryos as described previously.⁸ Natural lignans (1–3) were assayed at a standard concentration of 200 μg mL⁻¹ in whole embryos. Nocodazole and aphidicolin were used as controls at the same concentration. Fig. 3 shows the effect over the number of marked nuclei (H3S10ph) on the treated embryos, and the quantification of the fold change is presented in Fig. 4. The results indicate that the embryos treated with the natural lignans 1–3 showed a marked increase in H3S10ph positive nuclei by 1.92, 2.41 and 2.57 times, respectively, with respect to the control (DMSO).

Fig. 3 Histone H3 phosphorylated at serine 10 (H3S10ph) whole mount fluorescence immunolocalization in 30 h post fertilization (hpf) zebrafish embryos. 24 hpf embryos were treated for 6 h with the corresponding compounds and afterwards were immunostained against the mitotic marker, H3S10ph. Embryos were imaged by conventional fluorescence microscopy. (A) Dimethyl sulfoxide 1% (control). (B) Aphidicolin. (C) Nocodazole. (D) 7′,8′-Dehydropodophyllotoxin (1). (E) 7′,8′-Dehydro acetyl podophyllotoxin (2). (F) 7′,8′-Dehydro-trans-p-coumaroyl podophyllotoxin (3). Scale bar, 500 μm.

Fig. 4 Results of H3S10ph-based evaluation of the cell-cycle effect of natural lignans 1–3 using zebrafish embryos. The fold changes were determined using the values for the treated embryos and compared against the control treated embryos. *Significant differences (p < 0.01). These results are also presented in ESI Table 1.†

Moreover, these compounds caused morphological changes in the embryos by modifying the circularity by 0.78, 0.73 and 0.77, respectively (Fig. 5), with respect to the embryos treated with the vehicle. This effect was observed previously for the compounds that interfere with the stability of microtubules, such as podophyllotoxin.⁸ These results indicate that compounds 1–3 promote mitotic arrest and induction of morphological changes in a similar manner to that of the antimitotic drugs, nocodazole and podophyllotoxin.

Fig. 5 Morphological effect of pure compounds of B. fagaroides var. fagaroides evaluated by measuring the circularity of zebrafish embryos. *Significant difference in circularity values (p < 0.01). These results are also presented in ESI Table 2.†

Epiboly migration on zebrafish embryos is used to screen compounds that destabilize microtubules.²² Using this approach, we analyzed the effect of the pure lignans (1–3) on epiboly migration in zebrafish embryos. Embryo treatment started at the sphere stage and afterwards all treatment groups were fixed at the same time point when the control DMSO treated embryos reached 90% of epiboly. As shown in Fig. 6, compounds 1–3 induced epiboly delay and larger blastoderm cells and nuclei, as evidenced by phalloidin alexa and SYTOX green staining of the embryos. These effects show similarities to the treatment with nocodazol, which caused more penetrant phenotypes. In contrast, embryos treated only with vehicle or aphidicolin showed normal epiboly migration movement because the yolk was fully covered. This result suggests that treatment of the embryos with compounds 1–3 affected microtubule cytoskeleton.

Fig. 6 Whole mount localization of actin filaments and nuclei in zebrafish embryos after treatment with compounds 1–3 to determine their effect on cell migration. Sphere stage zebrafish embryos were treated with different compounds until the control embryos reached 90% epiboly, then fixed and processed for actin and nuclei staining and visualized by confocal microscopy. (A) Dimethyl sulfoxide (control). (B) Aphidicolin. (C) Nocodazole. (D) 7′,8′-Dehydropodophyllotoxin (1). (E) 7′,8′-Dehydro acetyl podophyllotoxin (2). (F) 7′,8′-Dehydro-trans-p-coumaroyl podophyllotoxin (3). Scale bar, 250 μm.

To corroborate that lignans 1–3 effectively act directly over the microtubule cytoskeleton, we performed a tubulin immunolocalization on zebrafish embryos. For this assay, embryos were treated with the pure lignans (1–3) starting at the sphere stage and the embryos of all treatment groups were fixed when they reached 50% of epiboly, therefore they were stage matched. This stage was chosen because the yolk cell presents large and conspicuous microtubules that are readily visualized by confocal microscopy as previously reported.⁸ Fig. 7 shows embryos treated with each of the pure lignans 1–3 at a concentration of 200 μg mL⁻¹. The absence of microtubule arrays in the embryos treated with lignans 1–3 is notable, in comparison with the control (DMSO 1%) and aphidicolin treatments, which indicate that these compounds promote destabilization of the microtubules. This effect is comparable to that observed by the nocodazole treatment. This result confirms that the in vitro cytotoxic activity, as well as the in vivo effects, over cell cycle and morphological changes observed for these three natural lignans is due to their effect over microtubules.

Fig. 7 Whole mount immunolocalization of yolk cell microtubules in zebrafish embryos after treatment with B. fagaroides compounds. Sphere stage zebrafish embryos were treated with different compounds until all treatment groups embryos reached 50% epiboly, then fixed and processed for microtubule fluorescent immunolocalization and visualized by confocal microscopy. (A) Dimethyl sulfoxide (control). (B) Aphidicolin. (C) Nocodazole. (D) 7′,8′-Dehydropodophyllotoxin (1). (E) 7′,8′-Dehydro acetyl podophyllotoxin (2). (F) 7′,8′-Dehydro-trans-p-coumaroyl podophyllotoxin (3). Scale bar, 25 μm.

3. Conclusions

Three new aryldihydronaphthalene-type lignans (1–3) were isolated from the stem bark of Bursera fagaroides, along with six known podophyllotoxin-type lignans (4–9) and one coumarin (10). Compounds 1–3 were identified for the first time in the genus Bursera. Compounds 1–3 showed marked cytotoxic activity against the human nasopharyngeal (KB), colon (HF-6) and prostate (PC-3) cancer cell lines. Evaluation of the effect of compounds 1–3 over the developing zebrafish embryo model showed that these compounds promote mitotic arrest and a delay in cell migration by disturbing the microtubule cytoskeleton. This in vivo activity suggests that these natural lignans affect tubulin in a similar manner as that of podophyllotoxin.

4. Experimental section

General procedures

NMR spectra were acquired on a Varian Unity NMR spectrometer operating at 400 MHz for ¹H and 100 MHz for ¹³C nuclei. Chemical shifts are listed in parts per million (ppm), referenced to CDCl₃ and were assigned on the basis of ¹H–¹H gCOSY, NOESY, gHSQC and gHMBC spectral analysis as required. NMR experiments performed in CDCl₃ are referenced to Me₄Si (0 ppm). FABMS spectra in a matrix of m-nitrobenzyl alcohol or glycerol were recorded on a JEOL JMX-AX 505 HA mass spectrometer. All the reagents and solvents used were of analytical grade. Optical rotations were acquired with a Perkin-Elmer 241MC polarimeter (10 cm, 1 mL cell) at the sodium D line. IR spectra were obtained on a Vectro 22, Bruker spectrometer, and UV spectra were obtained on an Agilent 8453 (1 cm, 3 mm cell) spectrometer. CD spectra were recorded on a CD spectropolarimeter (model J-715, Jasco Analytical Instruments) using a 0.1 cm path length cell over the range of 190–260 nm. The purity of the lignans was confirmed by reverse-phase HPLC, using Agilent Infinity 1260, equipped with an autosampler, photodiode array detector, a quaternary pump and an Agilent C-P18 column (2.7 μm, 4.6 × 50 mm) with MeOH–H₂O (1 : 1) as the isocratic eluent system, UV detection at 215 nm and a flow rate of 1 mL min⁻¹. 7′,8′-Dehydro-podophyllotoxin (1) (t_R = 0.93 min), 7′,8′ dehydro-acetylpodophyllotoxin (2, t_R = 1.80 min), podophyllotoxin (4, t_R = 0.87 min), acetyl podophyllotoxin (5, t_R = 1.62 min), 5′-desmethoxy-β-peltatin-A-methylether (6, t_R = 1.97 min), acetyl picropodophyllotoxin (7, t_R = 2.45 min), burseranin (8, t_R = 2.29 min), and hinokinin (9, t_R = 1.38 min).

Plant material

The bark of B. fagaroides var. fagaroides (H.B.K.) Engl. was collected from the village of Capula between Zacapu and Quiroga, Michoacán, México. Its identification was made at the Herbarium of the Instituto Mexicano del Seguro Social (registration number-12 051 IMSSM) and at the Institute of Botany, University of Guadalajara (IBUG-140 748), México.

Extraction and isolation

The stem bark of Bursera fagaroides var. fagaroides (1.140 g) was extracted by maceration at room temperature thrice with CH₂Cl₂, (3 L) and was concentrated to dryness under reduced pressure. The three extracts were combined (33 g) and fractionated by column chromatography on 990 g of silica gel (70–230 mesh) eluting with n-hexane-EtOAc mixtures of increasing polarity to yield five fractions: F-1 (1.2 g, 100 : 00 to 9 : 1), F-2 (2.72 g, 4 : 1), F-3 (5.1 g, 4 : 1), F-4 (8.9 g, 7 : 3), and F-5 (1.6 g, 1 : 1).

F-2 was subjected to CC (90 : 10 → 70 : 30, n-hexane/CH₂Cl₂), and a total of 46 fractions were collected, each of about 50 mL. Similar fractions were combined according to their TLC properties to yield two main sub-fractions. Fractions eluted with n-hexane-CH₂Cl₂ (9 : 1) were chromatographed on silica gel to yield 131 mg of β-sitosterol and 36.5 mg of burseranin (8). Fractions eluted with n-hexane-CH₂Cl₂ (8 : 2) were combined and the residue (1.21 g) was purified by column chromatography (90 : 10 → 00 : 100, n-hexane/EtOAc) to afford 278.5 mg of acetylpodophyllotoxin (5), 36.5 mg of burseranin (8) and 145 mg of 7′,8′-dehydro-acetylpodophyllotoxin (2).

F-3 was chromatographed on a silica gel column (9 : 1 → 7 : 3) with n-hexane/EtOAc. Fractionation resulted in three fractions. The sub-fraction F-3-2 (649.3 mg), which was eluted with a solvent ratio of 8 : 2, was then purified by preparative TLC and eluted with 95 : 05 benzene-EtOAc to yield 28.4 mg of hinokinin (9) and 48.3 mg of burseranin (8). The subtraction F-3-3, eluted with 7 : 3 n-hexane-EtOAc (2.8 g), was purified by silica gel column chromatography, eluting with a gradient of n-hexane/CH₂Cl₂ (8 : 2 → 6 : 4) to yield 117 mg of 5′-desmethoxi β-peltatin A methyl ether (6) and 15 mg of acetylpicropodophyllotoxin (7).

F-4 was subjected to CC and eluted with an isocratic mixture of 65 : 35 n-hexane/EtOAc, which produced 85 fractions of 100 mL each. Fractions 20–36 were a mixture of two lignans, as determined from their NMR spectra. This mixture was purified by semi-preparative RP HPLC eluting with CH₃OH–H₂O (52 : 48) to obtain 254 mg of 5′-desmethoxy-β-peltatin-A-methyl-ether (6, t_R = 1.97 min) and 40.5 mg of 7′,8′-dehydro-acetyl-podophyllotoxin (2, t_R = 1.80 min). The yields were based on the peak areas of the HPLC chromatogram.

Fractions 40–65 were combined and the residue (1.6 g) was adsorbed on reverse phase silica gel and subjected to RP column chromatography and eluted with a gradient of MeOH : H₂O (1 : 1 → 6 : 4) to yield 39 fractions of 50 mL each, which were combined in three main fractions: F4-1 (42 mg, 50 : 50), F4-2 (185 mg, 55 : 45), and F4-3 (290 mg). Fraction F4-1 constituted mainly of one compound and was purified by crystallization from acetone to yield 8 mg of scopoletin (10). F4-2 was purified by silica gel column chromatography, eluting with an isocratic mixture of n-hexane/EtOAc (6 : 4), to obtain two main fractions. The less polar fraction was purified by preparative RPTLC eluted with CH₃CN–H₂O (72 : 25) to yield 13 mg of 7′,8′-dehydro trans-p-coumaroylpodophyllotoxin (3). The most polar fraction was purified by preparative TLC eluted with benzene/EtOAc (55 : 45) (three developments) to afford 8 mg of podophyllotoxin (4) and 42 mg of 7′,8′-dehydropodophyllotoxin (1).

7′,8′-Dehydro-podophyllotoxin (1). White amorphous powder, mp 271–273 °C; purity = 98%; [α]²⁴_D −95° (c 0.090, CHCl₃); UV (CHCl₃) λ_max, nm (log ε): 247 (1.89), 345 (0.77); CD (c 1.33 × 10⁻⁴ M, CHCl₃) nm (Δε): 3.58 (1.68), 314 (0.166), 278 (0.438), 251 (−3.49); IR (KBr) ν_max cm⁻¹: 3440.9, 2921.6, 1757.5, 1608.6, 1503.3, 1480.5, 1331.6, 1280.9, 1161.3 cm⁻¹; ¹H-NMR (400 MHz, CDCl₃-DMSOd₆) and ¹³C NMR (100 MHz, CDCl₃-DMSOd₆) data see Table 1. FABHRMS m/z 413.1301 [M + H]⁺ (calcd for C₂₂H₂₁O₈, [M + H]⁺ 413.1191).

7′,8′-Dehydro-acetylpodophyllotoxin (2). White amorphous powder, mp 262–264 °C; purity = 95%; [α]²⁴_D −87° (c 0.015, CHCl₃); UV (CHCl₃) λ_max, nm (log ε): 266 (2.19), 287 (0.472), 348 (1.61); CD (c 1.33 × 10⁻⁴ M, CHCl₃) λ_max(Δε): 354 (26.0), 313 (4.65), 280 (6.08), 252 (−44.45) nm; IR (KBr) ν_max cm⁻¹: 2921.1, 2848.9, 1757.8, 1608.2, 1516.2, 1480.5, 1349.6, 1280.9, 1129.4 cm⁻¹; ¹H-NMR (400 MHz, CDCl₃-DMSOd₆) and ¹³C NMR (100 MHz, CDCl₃-DMSOd₆) data see Table 1. FABHRMS m/z 455.1324 [M + H]⁺ (calcd for C₂₄H₂₃O₉ [M + H]⁺ 455.1342).

7′,8′-Dehydro-trans-p-coumaroylpodophyllotoxin (3). White amorphous powder, mp 287–288 °C; purity = 95%; [α]²⁴_D −75.0° (c 0.014, CHCl₃); IR (KBr) ν_max cm⁻¹: 2920.2, 1757.2, 1608.6, 1503.6, 1483.6, 1330.9, 1280.6, 1161.1 cm⁻¹; ¹H-NMR (400 MHz, CDCl₃) and ¹³C NMR (100 MHz, CDCl₃) data see Table 1. FABHRMS m/z 559.1586 [M + H]⁺ (calcd for C₃₁H₂₇O₁₀ [M + H]⁺ 559.1604).

Basic hydrolysis of 2 and 3

7 mg of 2 and 4.2 mg of 3 were separately dissolved in a solution of 7 : 3 : 1 THF, MeOH, and H₂O, and then stirred with a 10% K₂CO₃ (2 equiv.) solution for 9 hours at room temperature. The product was purified by preparative TLC and eluted with 1 : 1 hexane-EtOAc to yield 2 mg and 0.75 mg, respectively, of compound 1.

Cytotoxicity assay

In vitro cytotoxicity was determined by the sulphorhodamine B (SRB) (MP Biomedicals, LLC) protein staining assay^23,24 using KB (nasopharyngeal), HF-6 (colon), MCF-7 (breast), and PC-3 (prostate) cancer cell lines. The cell cultures were maintained in an RPMI-1640 medium supplemented with 10% fetal bovine serum, 5000 units per mL penicillin, 5 mg mL⁻¹ streptomycin, 7.5% NaHCO₃, and cultured in a 96-well Microtiter plate (10⁴ cells per mL, 190 μL per well) at 37 °C in a 5% CO₂-air atmosphere (100% humidity). The cells at the log phase of growth were treated in triplicate (n = 3) with different concentrations of the test compounds (0.16, 0.8, 4 and 20 μg mL⁻¹), and incubated for 72 h. The cell concentration was determined by protein analysis. The optical density was measured at 590 nm with an ELISA-Reader (Molecular Devices, SpectraMax Plus 384). Results are expressed as the concentration that inhibits 50% of control growth after the incubation period (IC₅₀). The values were estimated from the semi-log plot of the extract concentration (μg mL⁻¹) against the percentage of viable cells. Camptothecin and podophyllotoxin were included as positive standards. Pure products with ED₅₀ ≤ 4 μg mL⁻¹ were considered active according to the National Cancer Institute (NCI) guidelines described in the literature.²⁵

Statistical analysis

The results were analyzed using one-way ANOVA followed by Kaplan–Meier estimation of survival and Cox's regression using the statistical package SPSS V.15.

Fish maintenance and strains

Wild type zebrafish (Danio rerio) were obtained from natural crosses and raised at 28 °C. Embryo stages were determined by morphological criteria according to Kimmel and collaborators.²⁶ Zebrafish were handled in compliance with the local animal welfare regulations EU Directive 2010/63/EU indications²⁷ and approved by Comité de ética (Instituto de Biotecnología, UNAM).

Chemical treatment of zebrafish embryos

Treatments were performed as previously described in the literature.⁸ In brief, the compounds used for chemical treatments were diluted in anhydrous DMSO (276855, Sigma-Aldrich). Plant extracts for screening were tested at a standard final concentration of 200 μg mL⁻¹ in water. The control compounds were aphidicolin (10 μg mL⁻¹, A0781, Sigma-Aldrich) and nocodazole (10 μg mL⁻¹, M1404, Sigma-Aldrich). 10 zebrafish embryos of 24 hour post fertilization age for each treatment were incubated at 28 °C in egg water (60 μg mL⁻¹ of “Instant Ocean” sea salts in distilled water) for 6 hours in 48-well plates.²⁸ 3 μL of each compound stock solution was added to a total volume of 300 μL at the beginning of the incubation. Control embryos were treated with DMSO alone at 1% final concentration (v/v), and processed as described below.

Immunofluorescense

Whole-mount immunostaining in zebrafish embryos was used to determine the effect of pure compounds as previously described in the literature.²⁹ The primary antibody rabbit IgG anti phospho histone H3 (sc-8656-R, Santa Cruz Biotechnology) was diluted in the ratio of 1 : 100 in blocking buffer (final concentration 2 μg mL⁻¹), and as a secondary antibody a 1 : 100 goat anti-rabbit Alexa Fluor 488 (A11008, Molecular Probes) solution in blocking buffer was used. For the analysis of cell migration, embryos were stained with SYTOX orange (S11368, Molecular Probes), diluted 1 : 2000 in blocking buffer to visualize DNA and nuclei and counter-stained with phalloidin Alexa 488 (A12379, Molecular Probes). For tubulin immunolocalization, as the primary antibody, mouse anti α tubulin monoclonal antibody diluted 1 : 500 in blocking buffer was used (T9026, Sigma-Aldrich) and a goat anti-mouse coupled to Alexa 647 (A21235, Molecular Probes) diluted 1 : 1000 in blocking buffer was used as a secondary antibody. Embryos were mounted in 2.5% methyl cellulose in PBS for epifluorescence or in 1% low melting point agarose for confocal microscopy visualization.

Fluorescence microscopy and image analysis

Fluorescent signals corresponding to H3S10ph positive cells in whole zebrafish embryos were imaged by fluorescence microscopy with a 5× objective 0.15 N.A., PlanNeofluoar, under a Zeiss Axioscop Microscope. Image stacks (10 to 14 images per embryo) were merged into a single focused image with the public domain ImageJ software³⁰ with the stack focuser plugin. To quantify the phospho histone H3 positive nuclei in each embryo, focused images were made binary by thresholding to highlight in black the H3S10ph positive nuclei and in white the background, and automatically quantified by the analyze particles command in the ImageJ software. Embryo contour was delineated on each image and the circularity was measured in ImageJ. The Student's t-test statistical analysis was performed in Excel.

Confocal laser scanning microscopy

Zebrafish embryos stained with the specified fluorescent dyes were visualized on a FluoView FV1000 confocal microscope coupled to an inverted IX81 Olympus microscope with a UPlanSApo 10× (numeric aperture = 0.4) objective or a UPlanSApo 20× (numeric aperture = 0.75) objective. The pinhole aperture was maintained at 200. Serial optical sections were obtained with a z-step of 8 or 2 μm. Images were processed with the public domain software ImageJ³⁰ and Adobe Photoshop.

Conflict of interest

The authors declare that there are no conflicts of interest.

Acknowledgements

Mayra Yaneth Antúnez Mojica and Mario A. Mendieta-Serrano acknowledge fellowships 253315 and 323762, respectively, from CONACYT. Partial financial supports from CONACYT, México (Grants 82851, 240801 and 222714) and UNAM (Grants IX201110 and IN205612) are acknowledged. The authors thank Laboratorio Nacional de Estructura de Macromoléculas (Conacyt 251613) for the spectroscopic and mass analyses and Confocal Microscopy Service provided by Laboratorio Nacional de Microscopía Avanzada, Instituto de Biotecnología, Universidad Nacional Autónoma de México.

References

1. R. Medina-Lemus, Flora del valle de Tehuacán-Cuicatlán, Instituto de Biología, Universidad Nacional Autónoma de México, 2008, vol. 66, pp. 1–76.
2. R. L. Huacuja, N. M. Delgado, L. A. Carranco, L. R. Reyes and G. A. Rosado, Arch. Invest. Med., 1990, 21, 393–398.
3. P. Rosas-Arreguin, P. Arteaga-Nieto, R. Reynoso-Orozco, J. C. Villagomez-Castro, M. Sabanero-Lopez, A. M. Puebla-Perez and C. Calvo-Mendez, Exp. Parasitol., 2008, 119, 398–402.
4. A. M. Puebla-Pérez, L. Huacuja, G. Rodríguez, X. Lozoya, G. Zaitseva-Petrovna and M. Villaseñor-García, Phytother. Res., 1998, 12, 545–548.
5. E. Bianchi, K. Sheth and J. R. Cole, Tetrahedron Lett., 1969, 10, 2759–2762.
6. R. Velazquez-Jimenez, J. M. Torres-Valencia, C. M. Cerda-Garcia-Rojas, J. D. Hernandez-Hernandez, L. U. Roman-Marin, J. J. Manriquez-Torres, M. A. Gomez-Hurtado, A. Valdez-Calderon, V. Motilva and S. Garcia-Maurino, Phytochemistry, 2011, 72, 2237–2243.
7. A. M. Rojas-Sepúlveda, M. Mendieta-Serrano, M. Y. Antúnez, E. Salas-Vidal, S. Marquina, M. L. Villarreal, A. M. Puebla, J. I. Delgado, L. Alvarez, Molecules, 2012, 17, 9506–9519.
8. M. Y. Antúnez, A. M. Rojas-Sepúlveda, M. A. Mendieta-Serrano, L. González, S. Marquina-Bahena, E. Salas-Vidal and L. Alvarez, submitted.
9. M. A. Leyva-Peralta, R. E. Robles-Zepeda, A. Garibay-Escobar, E. Ruiz-Bustos, L. Alvarez and J. C. Gálvez-Ruiz, BMC Complementary Altern. Med., 2015, 15(13), 1–7.
10. I. Alejandre-García, L. Alvarez, A. Cardoso-Taketa, L. González-Maya, M. Antúnez, E. Salas-Vidal, F. Díaz, S. Marquina-Bahena and M. L. Villarreal, J. Evidence-Based Complementary Altern. Med., 2015, 298463, 11.
11. J. A. Moreno-Escobar, L. Alvarez, V. Rodríguez-López and S. Marquina, Phytochem. Lett., 2013, 6, 610–613.
12. M. Acevedo, P. Nuñez, L. Gónzález-Maya, A. Cardoso and M. L. Villarreal, J. Clin. Toxicol., 2015, 5, 232,  DOI:10.4172/2161-0495.1000232.
13. T. J. Schmidt, S. Vößing, M. Klaes and S. Grimme, Planta Med., 2007, 73, 1574–1580.
14. H. Otsuka, H. Kuwabara and H. Hoshiyama, J. Nat. Prod., 2008, 71, 1178–1181.
15. J.-Q. Gu, E. J. Park, S. Totura, S. Riswan, H. H. S. Fong, J. M. Pezzuto and A. D. Kinghorn, J. Nat. Prod., 2002, 65, 1065–1068.
16. V. D. Gangan and S. S. Hussain, J. Pharm. Res., 2011, 4, 4265–4267.
17. T.-S. Wu, L.-S. Shi, J.-J. Wang, S.-C. Iou, H.-C. Chang, Y.-P. Chen, Y.-H. Kuo, Y.-L. Chang and C.-M. Teng, J. Chin. Chem. Soc., 2003, 50, 171–178.
18. Q. G. Jian, P. E. Jung, T. Stephan, R. Seodarsono, H. S. Harry, J. M. Pezzuto and D. A. Kinghorn, J. Nat. Prod., 2002, 65, 1065–1068.
19. F. Z. Zhiwei-Huang, H. Zhen, L. Zhou, H. M. Amin and P. Shi, Leuk. Lymphoma, 2014, 55, 1876–1883.
20. M. C. Marcotullio, A. Pelosi and M. Curini, Molecules, 2014, 19, 14862–14878.
21. W. Liu, J. Hua, J. Zhou, H. Zhang, H. Zhu, Y. Cheng and R. Gust, Med. Chem. Lett., 2012, 22, 5008–5012.
22. H. S. Moon, E. M. Jacobson, S. M. Khersonsky, M. R. Luzung, D. P. Walsh, W. Xiong, J. W. Lee, P. B. Parikh, J. C. Lam, T. W. Kang, G. R. Rosania, A. F. Schier and Y. T. Chang, J. Am. Chem. Soc., 2002, 124, 11608–11609.
23. P. Houghton, R. Fang, I. Techatanawat, G. Steventon, P. J. Hylands and C. C. Lee, Methods, 2007, 42, 377–387.
24. P. Shekan, R. Storeng, D. Scudiero, A. Monks, J. McMahon, D. Vistica, J. T. Warren, H. Bokesch, S. Kenney and M. R. Boyd, JNCI, J. Natl. Cancer Inst., 1990, 82, 1107–1112.
25. M. Suffness and J. M. Pezzuto, Assays related to cancer drug discovery, Academic Press, London, 1991, vol. 6, pp. 71–133.
26. C. B. Kimmel, W. W. Ballard, S. R. Kimmel, B. Ullmann and T. F. Schilling, Dev. Dyn., 1995, 203, 253–310.
27. U. Strahle, S. Scholz, R. Geisler, P. Greiner, H. Hollert, S. Rastegar, A. Schumacher, I. Selderslaghs, C. Weiss, H. Witters and T. Braunbeck, Reprod. Toxicol., 2012, 33, 128–132.
28. M. Westerfield, The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio), Univ. of Oregon Press, Eugene, 4th edn, 2000.
29. M. A. Mendieta-Serrano, D. Schnabel-Peraza, H. Lomelí and E. Salas-Vidal, Gene Expression Patterns, 2015, 19, 98–107,  DOI:10.1016/j.gep.2015.08.003.
30. M. D. Abramoff, P. J. Magelhaes and S. J. Ram, Biophotonics Int., 2004, 11, 36–42.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra23516b

---