Euphorane C, an unusual C17-norabietane diterpenoid from Euphorbia dracunculoides induces cell cycle arrest and apoptosis in human leukemia K562 cells

1 Introduction

Natural products are important sources of “privileged structures” in drug development (Cragg et al., 2009). For anti-cancer agents, about 70 % of FDA-approved antitumor drugs are either natural products or their derivatives (Newman et al., 2020). Among them, the nonspecific “higher diterpenes” (also known as polycyclic diterpenoids) with a common 6/6/6 ring system from Euphorbia species are of particular significance due to their extensive physiological activities. Polycyclic diterpenoids are well-known for their structural diversity with many different skeletal types, such as rosanes, abietanes, atisanes, and kauranes (Shi et al., 2008; Vasas et al., 2014; Xu et al., 2021). Abietane diterpenoids from Euphorbia are usually substituted by an α,β-unsaturated γ-lactone ring at C-12 and C-13. Kauranes comprise a perhydrophenanthrene unit (A, B, and C rings) fused with a cyclopentane (D ring), while in atisanes, the perhydrophenanthrene unit is fused with a cyclohexane (Vasas et al., 2014). Recent studies have shown that polycyclic diterpenoids may be potential chemical entities in future antitumor drug development. For example, jolkinolide B, an abietane diterpene, could induce apoptosis and sensitize bladder cancer to mTOR inhibitors (Sang et al., 2021; Sang et al., 2022). 8,9-Seco-ent-kaurene, a kaurane diterpene, could effectively suppress triple-negative breast cancer cells via Akt signaling pathway (Fan et al., 2021).

Euphorbia dracunculoides, a perennial herb of the genus Euphorbia, is mainly distributed in North Africa, South Europe, and Southwest Asia (Wang et al., 2015a). Previous phytochemical studies showed that the plant is rich in diterpenes with different skeletal types, such as lathyranes, myrsinanes, and tiglianes, etc. (Chi et al., 2016; Dai et al., 2016; Wang et al., 2015a, 2015b, 2016, 2017). In our ongoing efforts to discover antitumor diterpenoids from Euphorbia plants (Chen et al., 2022; Yan et al., 2019), four new polycyclic diterpenoids (1–4) and 12 known compounds were isolated from the EtOAc fraction of E. dracunculoides. Compounds 2 and 3 represent unusual examples of aromatic abietane norditerpenoids in Euphorbia species. All the compounds were tested for antiproliferative activities against five human cancer cell lines, and 3 showed significant activity against the human leukemia K562 cells with micromolar IC₅₀. Herein, we reported the isolation, structural identification, and cytotoxicity of these compounds.

2 Material and methods

3 Results and discussion

The whole plants of E. dracunculoides were extracted with 95 % EtOH and then partitioned with EtOAc. The obtained EtOAc extract was separated by various column chromatographic methods to afford compounds 1–16 (Fig. 1).

Close

Fig. 1

Structures of compounds 1–16.

Export to PPT

Compound 1 was isolated as colorless needles and showed a molecular formula of C₂₀H₂₈O₃ according to the HRESMS at m/z 339.1920 [M + Na]⁺ (calcd for C₂₀H₂₈O₃Na, 339.1931). Its IR spectrum displayed the absorptions for hydroxyl (3421 cm⁻¹) and carbonyl (1742 cm⁻¹) groups. The ¹H NMR data (Table 1) of 1 showed signals for four methyls (δ_H 0.86, 0.88, 0.91, and 1.97), two oxygenated methines [δ_H 4.58 (1H, dd, J = 11.8, 6.2 Hz) and 5.05 (1H, s)], an olefinic proton [δ_H 5.99 (1H, s)] and a series of multiplets for aliphatic protons. The ¹³C NMR data of 1 (Table 2) showed 20 carbon resonances, which was classified as an ester carbonyl (δ_C 174.9), two double bonds (δ_C 123.7, 129.6, 135.8, and 162.3), four sp³ methines (two oxyganated at δ_C 70.5 and 78.3), five sp³ methylenes, four methyls, and two sp³ quaternary carbons (δ_C 32.8 and 36.2). The above NMR data were very similar to those of a reported abietane diterpenoid, fischeriolide A (Lee et al., 2016), except for the replacement of the oxygenated methine (δ_C 65.7, C-11) in fischeriolide A by a methylene (δ_C 27.0) in 1. This was supported by ¹H–¹H COSY correlations (Fig. 2) from the methylene protons (H₂-11) to H-9 and H-12. Further HMBC and ¹H–¹H COSY analysis confirmed the gross structure of 1.

Close

Fig. 2

Key ¹H−¹H COSY ( ) and HMBC (arrows) correlations of 1–4.

Export to PPT

The relative configuration of 1 was partially determined by a NOESY experiment. As shown in Fig. 3, the correlations of H-5/H-9 and H₃-20/H-6α indicated that these protons were pseudoaxially oriented on the half-chair conformational B ring. H-5 and H-9 were arbitrarily designated as β-orientation, while CH₃-20 was determined as α-orientation. Thus, the NOESY correlation from H₃-20 to H-12 assigned H-12 to be α-orientation. As the absence of NOESY correlation from H-14 to H-9, the remaining configuration of H-14 was difficult to confirm. Fortunately, a qualified crystal of compound 1 was obtained via recrystallization in EtOH, and subsequent X-ray diffraction analysis accurately assigned the absolute configuration of 1 as 5R,9S,10R,12R,14S (Fig. 4). Thus, compound 1 was assigned as 14β-hydroxy-ent-abieta-13(15)-ene-12α,16-olide and was given the trivial name euphorane A.

Close

Fig. 3

Key NOESY correlations ( ) of 1–4.

Export to PPT

Close

Fig. 4

X-ray structures of compounds 1, 2, 4, and 11.

Export to PPT

Compound 2 had a molecular formula of C₁₇H₂₄O₂ as determined by the HRESIMS at m/z 259.1707 [M − H]⁻ (calcd C₁₇H₂₃O₂, 259.1704) and ¹³C NMR data. The 1D NMR data of 2 were similar to those of a known podocarpane-like trinorditerpene, 7-deoxynimbidiol (Alvarez-manzaneda et al., 2007; Xiong et al., 2006), with the major differences focusing on the ¹³C chemical shifts in the aromatic ring C (e.g. δ_C 141.2, 141.5, and 143.6 in 7-deoxynimbidiol and δ_C 153.3, 156.4, and 156.4 in 2), suggesting that the two phenolic hydroxyls in 2 might be at C-12 and C-14 respectively. This was supported by HMBC correlations from an aromatic proton (δ_C 6.25, H-11) to C-10 and C-8, and from another aromatic proton (δ_C 6.10, H-13) to C-8 (Fig. 2). The relative configuration was determined to be the same as that of 7-deoxynimbidiol based on their similar ¹³C NMR data around rings A and B. This assignment was also supported by the NOESY correlations of H₃-20/H₃-19 and H₃-18/H-5 (Fig. 3). To determine the absolute configuration of 2, a single crystal X-ray diffraction analysis was conducted, and the results showed that 2 possessed a 5R,10R configuration (Fig. 4), contrary to that of 7-deoxynimbidiol (5S,10S). Thus, compound 2 was assigned as an unusual aromatic 15,16,17-trinorabietane diterpenoid and was named euphorane B.

Compound 3 (euphorane C) had a molecular formula of C₁₉H₂₆O₃ as determined by its HRESIMS and ¹³C NMR data. The 1D NMR data of 3 were very similar to those of 2, except for the absence of signals for an aromatic methine (δ_C 100.3, δ_H 6.10) and the presence of additional signals for an aromatic quaternary carbon (δ_C 109.3) and an ethanone fragment (δ_C 205.9 and 33.9, δ_H 2.95) in 3, suggesting that 3 was a C-13 acetylated derivative of 2. Observation of a four-bond W-type HMBC correlation from the aromatic proton (δ_H 6.70, H-11) to the ketone carbonyl (δ_C 205.9) confirmed this assignment. Similar ¹³C NMR data around the aromatic ring C of 3 and a synthesized analogue S1 (Fig. 5A) (Shockley et al., 2014) also supported this conclusion. Further computational calculation of the 1D NMR data of two positional isomers 3a and 3b (Fig. 5A) by gauge-independent atomic orbital (GIAO) method and analyzed by the improved probability DP4 + method confirmed this structure. As indicated in Fig. 5B, Tables S4 and S5, a final DP4 + probability score of 100.00 % and 0.00 % was obtained for 3b and 3a respectively, and the experimental ¹³C NMR data around ring C in 3 matched well with those calculated for isomer 3b (Fig. 5A, Table S2, S4). These results indicated that compound 3 had a structure as depicted and was a rare C17-norabietane diterpenoid.

Close

Fig. 5

1D NMR calculation of 3. (A) Structures of S1, 3, 3a, and 3b, and their experimental or calculated ¹³C NMR data around ring C. (B) DP4 + probability of 1D NMR data of 3a (isomer 1) and 3b (isomer 2).

Export to PPT

The relative configuration of 3 was determined to be the same as that of 2 based on their similar 1D NMR data around rings A and B. This assignment was also supported by the NOESY correlations of 3 (Fig. 3). Then, the absolute stereochemistry of 3 could be studied by comparison of its experimental and calculated ECD spectra. As shown in Fig. 6, the experimental ECD curve of 3 (3b) showed four Cotton effects around 200 (−), 230 (+), 275 (−), and 350 (+) nm respectively, which matched well with those calculated for the isomer 5R,10R-3b, indicating that 3 possessed the same absolute configuration (5R,10R).

Close

Fig. 6

The experimental and calculated ECD curves of 3.

Export to PPT

Euphorane C (3) is an unusual C17-norabietane diterpenoid, and its biosynthetic pathway was proposed in Scheme 1. Briefly, proton-initiated polyene cyclization of trans-geranylgeranyl diphosphate could successively produce (+)-copalyl diphosphate and pimarane-like diterpene. Then, 1,2-migration of methyl from C-13 to C-15 constructed the characteristic isopropyl group of the abietane-like precursor, which further underwent the aromatization to give an important intermediate i (Ravn et al., 2002; Yang et al., 2002). The hydroxylations of C-12 and C-14 followed by oxidation of CH₃-17 in i afforded intermediate ii. Finally, the C17-norabietane diterpenoid (3) might be generated from ii via decarboxylization and oxidation reactions (Reis et al., 2014; Wang et al., 2015; Yun et al., 2018).

Close

Scheme 1

Proposed Biosynthetic Pathway of 3.

Export to PPT

Compound 4 was isolated as colorless crystals. Its molecular formula was determined as C₂₂H₃₂O₄ by the HRESIMS at m/z 383.2205 [M + Na]⁺ (calcd for C₂₂H₃₂O₄Na, 383.2193). The 1D NMR data of 4 were very similar to those of a reported ent-atisane diterpenoid, ent-(3β,13S)-3,13-dihydroxyatis-16-en-14-one (Zhang et al., 2012), except for the presence of an additional acetyl in 4, suggesting that 4 was an O-acetylated derivative of the reported compound. This was supported by its 2D NMR data (Fig. 2), and the acetoxyl group was located at C-13 by the HMBC correlation from H-13 to the carbonyl (δ_C 170.5). The relative configuration of 4 was determined by the NOESY experiment. As shown in Fig. 3, the NOESY correlations of H-5/H-3, H-5/H₃-18, and H₃-20/H₃-19 suggested that H-5, H-3, CH₃-19, and CH₃-20 occupied the axial positions of the chair conformational A ring. H-3 and H-5 were arbitrarily designated as β-orientations, CH₃-19 and CH₃-20 were accordingly determined as α-orientations. Thus, the NOESY correlations of H-5/H-9 and H-9/H₂-15 established the β-orientation of H-9 and C-8 − C-15, while the NOESY correlation of H₃-20/H-13 assigned H-13 as α-orientation. Furthermore, the absolute stereochemistry of 4 was determined by single crystal X-ray diffraction experiment to be 3R,5S,8S,9S,10R,12S,13R (Fig. 4). Thus, compound 4 was assigned as 13β-acetoxy-3α-hydroxy-ent-atis-16-ene-14-one and was named euphorane D.

Compound 11 was identified as (23E)-25-methoxycycloart-23-en-3β-ol, a known cycloartane triterpenoid, by comparing its 1D NMR data with those reported (Cabrera et al., 1996; Zhang et al., 2005). However, the absolute stereochemistry of 11 remains to be determined. Thus, a recrystallization experiment was carried out, and a qualified crystal of 11 was obtained in MeOH. The crystallographic data exactly assigned the absolute stereochemistry of 11 as 3S,5R,8S,9S,10R,13R,14S,17R,20R (Fig. 4).

Other known compounds, 13-acetyloxy-atis-16-ene-3,14-dione (5) (Jia et al., 1990), ent-3,14-dioxo-16-atisene (6) (Lal et al., 1989), 17-hydroxyjolkinolide A (7) (Che et al., 1999), jolkinolide E (8) (Lal et al., 1990), jolkinolide A (9) (Lal et al., 1990), jolkinolide B (10) (Che et al., 1999), spiroinonotsuoxodiol (12) (Handa et al., 2010), agnosterol (13) (Emmons et al., 1989), kansenone (14) (Wang et al., 2003), 3β-hydroxy-25-methyloxylanosta-8,23-diene (15) (Fang et al., 2015), and cornusalterin A (16) (Kim et al., 2011) were identified by comparison of their NMR data with those reported in the literature.

The antiproliferative activities of compounds 1–16 were evaluated on five human cancer cell lines. As shown in Table 3, most of the tested compounds only showed mild to none inhibitory activity against the cancer cells. Compounds 10 and 13 imparted cytotoxicity against the chronic myelogenous leukemia cell line (K562) with IC₅₀ values of 7.23 and 5.63 μM respectively. Notably, compound 3 demonstrated significant antiproliferative activity against four of the cancer cell lines with micromolar IC₅₀, and had the highest activity against K562 cells (IC₅₀ = 3.59 μM). Thus, the preliminary mechanism of 3 on inhibiting K562 cells was further investigated by using flow cytometry. As shown in Fig. 7, the treatment of 3 at 4 and 8 μM dose- and time-dependently induced apoptosis of K562 cells as compared to the DMSO treated group. Furthermore, we tested whether 3 could arrest cell cycle. The results showed that 3 treatments (4 and 8 μM) changed cell cycle progression as shown by increased G0/G1 and reduced G2/M cell population (Fig. 8), suggesting that 3 arrested K562 cell cycle at G0/G1 phase. Taken together, compound 3 could arrest cell cycle at G0/G1 phase, finally leading to the apoptosis of K562 cells.

Close

Fig. 7

Compound 3 induced apoptosis of K562 cells. The corresponding quantitative analysis was included. Data represented the mean ± SD of at least three independent experiments. ****P < 0.0001 vs DMSO treated group.

Export to PPT

Close

Fig. 8

Compound 3 arrested K562 cell cycle at G0/G1 phase. Quantitative analysis was included. Data represented the mean ± SD of at least three independent experiments. **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs DMSO treated group.

Export to PPT

4 Conclusions

Polycyclic diterpenoids, such as abietanes (6/6/6 ring system) with 20α-CH₃ and 5β-H relative configuration, were common natural products in Euphorbia species. However, abietane trinor or norditerpenoids were rarely discovered in Euphorbia plants (Shi et al., 2008; Vasas et al., 2014; Xu et al., 2021). In the study, chemical investigation of E. dracunculoides led to the isolation of four new polycyclic diterpenoids (1–4) and 12 known terpenoids. Among them, euphorane B (2) represents the first example of aromatic 15,16,17-trinorabietane diterpenoid in Euphorbia species, while 3 is an unusual 17-norabietane diterpenoid. Biological assay indicated that 3 significantly inhibited the proliferation of K562 cells with micromolar IC₅₀. Preliminary mechanism investigation revealed that 3 could induce cell cycle arrest at G0/G1 phase, causing apoptosis of K562 cells. These findings not only expanded the structural diversity of abietane diterpenoids, but also provided potential chemical scaffold for the development of anti-leukemia drugs. Studies toward the in vivo efficacy and in-depth mechanism of 3 are still in progress.