Laxiflorol A, the first example of 7,8:15,16-di-seco-15-nor-21-homo-ent-kauranoid from Isodon eriocalyx var. laxiflora

Abstract

Laxiflorol A (1), an unprecedented 7,8:15,16-di-seco-15-nor-21-homo-ent-kauranoid, and its precursor analogue, laxiflorol B (2), were isolated from the leaves of Isodon eriocalyx var. laxiflora. The absolute configuration of 1 was determined by spectral methods and quantum chemical calculations. Compound 2 exhibited weak cytotoxicity.

---

1. Introduction

The Natural Products Library provided a diverse and unique source of bioactive lead compounds for drug discovery.¹ The ent-kaurane diterpenoids library has been constructed and maintained by our group since 1976 for the development of therapeutic agents to treat cancer, and more than 1000 pure ent-kauranoids, including more than 700 novel ones, have been identified from the Isodon genus.²

Ent-kaurane-type diterpenoids represent one of the most excellent examples of natural products with diverse structural scaffolds and important pharmaceutical activities.^2,3 Among these compounds in our library, many unique structures, including ternifolide A,⁴ neoadenoloside A,⁴ neolaxiflorin A,⁵ nervonin A,⁶ maoecrystal Z,⁷ maoecrystal V,⁸ xindongnin M,⁹ neoangustifolin¹⁰ and epinodosino¹⁰, have been reported by Natural Product Reports as hot off the press. Some compounds, such as oridonin,¹¹ eriocalyxin B,¹² adenanthin,^3a pharicin B,^3b and other compounds, have brought great attention to their potential antitumor application. In order to enrich and improve our ent-kaurane diterpenoids library, chemical constituents of the leaves of Isodon eriocalyx var. laxiflora was further investigated.^2,13 A trace amount of an ent-kauranoid (0.00003%) with a novel 7,8:15,16-di-seco-15-nor-21-homo-ent-kaurane skeleton, designated as laxiflorol A (1) (3.0 mg), together with its biogenetic precursor analogue, laxiflorol B (2), were obtained from this plant (Fig. 1).

Fig. 1 Chemical structures of compounds 1, 2, and 6,8-epi-1, and the X-ray crystallographic structure of 2.

Laxiflorol A (1) was the first example of 7,8:15,16-di-seco-ent-kauranoid, which was biosynthetically formed from multiple ring cleavages, adding (C-21) and losing (C-15) carbons, and sequential reactions on an ent-kauranoid (Scheme 1). Nevertheless, the 1D and 2D NMR experiments showed that compound 1 has two possible structures, 1 and 6,8-epi-1 (Fig. 1). After multiple attempts made by us, the definite structure of 1 was shown via NMR data coupled with quantum chemical calculations, including ¹³C NMR chemical shifts and ECD spectra. Herein, we report the isolation, structure elucidation, including absolute stereochemistry, and cytotoxic activities of compounds 1 and 2.

Scheme 1 Hypothetical biogenetic pathway of 1.

2. Results and discussion

As the precursor analogue of laxiflorol A (1), laxiflorol B (2) was obtained as colorless needles. On the basis of careful 1D NMR, 2D NMR analyses (Table S1†), and single-crystal X-ray diffraction using anomalous scattering of CuK_α radiation data (Fig. 1),¹⁴ the absolute configuration of compound 2 was assigned and could be described according to the following nomenclature: (3R,8S,9S,10S,13R,16S)-3,20-epoxy-6-hydroxy-17-ethoxy-5(6)-en-ent-kaur-1,7,15-trione.§

Laxiflorol A (1) was obtained as white, amorphous powder. The molecular formula, C₂₀H₂₆O₇, with eight degrees of unsaturation, was established based on HRESIMS ([M + Na]⁺, 401.1571; calcd for C₂₀H₂₆O₇Na, 401.1576) and NMR spectroscopy (Table 1). The analysis of the ¹³C NMR and DEPT spectra revealed the presence of 20 carbons, which were assigned as two methyl, seven methylene (two oxygenated), five methine (two oxygenated), and six quaternary carbons (one oxygenated, one ester and two carbonyls), which suggested that 1 is a highly oxygenated diterpenoid with a C₂₀ skeleton quite different from the ent-kaurane skeleton reported previously.²

Table 1 NMR spectroscopic data of compound 1 and its calculated ¹³C NMR data (δ in ppm, J in Hz)

No.	δH	δC
	Exp^a	Exp^a	Cal^b	Cal^c
a Data were recorded in C5D5N on a 600 MHz spectrometer.b Calculated ^13C NMR data of 1.c Calculated ^13C NMR data of 6,8-epi-1.
1	—	207.3 s	206.9 s	207.3 s
2	2.83 br d (2.6)	41.4 t	41.8 t	43.5 t
3	3.68 br t (2.6)	77.2 d	75.6 d	77.2 d
4	—	35.3 s	37.3 s	39.3 s
5	2.28 br s	45.2 d	48.1 d	42.6 d
6	5.07 d (3.6)	71.9 d	72.0 d	73.6 d
7	—	172.5 s	168.9 s	169.7 s
8	—	111.4 s	110.8 s	112.3 s
9	2.65 dd (13.1, 3.8)	39.3 d	39.3 d	40.6 d
10	—	47.4 s	49.5 s	51.4 s
11	1.70 m; 1.17 m	24.4 t	25.4 t	22.5 t
12	2.03 m; 1.41 m	27.7 t	29.7 t	20.6 t
13	2.90 overlap	47.8 d	45.4 d	45.2 d
14	2.27 br d (13.2); 2.14 br d (13.2)	35.9 t	34.3 t	32.4 t
16	—	210.3 s	214.3 s	213.9 s
17	2.90 overlap	44.5 t	43.5 t	42.0 t
18	1.00 s	28.8 q	27.6 q	27.8 q
19	1.54 s	24.4 q	23.6 q	25.0 q
20	4.51 d (10.8); 3.88 d (10.8)	61.4 t	60.2 t	62.2 t
21	4.20 t (6.0)	57.6 t	57.8 t	60.1 t

---

The HMBC spectrum of 1 showed correlations from the geminal methyls Me-18 (δ_H 1.00) and Me-19 (δ_H 1.54) to C-3, C-4, and C-5. Furthermore, the AB spin system of methylene H₂-20 showed HMBC correlations with C-1, C-3, C-5, C-9, and C-10. Other HMBC correlations were noted between methylene H₂-2 (δ_H 2.83) and C-1, C-3, C-4, and C-10, between oxygenated methine H-3 (δ_H 3.68) and C-1, C-5, and C-20, between methine H-5 (δ_H 2.28, 1H, br s) and C-1, C-4, C-6, C-7, C-9, C-10, C-18, C-19, and C-20, and between oxygenated methine H-6 (δ_H 5.07) and C-7 and C-10. These observed HMBC correlations, coupled with two spin systems (CH₂CH, H₂-2/H-3 and CHCH, H-5/H-6) established by ¹H–¹H COSY correlations and the HSQC spectra, gave rise to partial structure part a (Fig. 2).

Fig. 2 ¹H–¹H COSY (bold), selected HMBC (arrow) correlations of compound 1, and key ROESY (full arrow) correlations of conformer 1g.

The HMBC spectrum showed that oxygenated methylene group H₂-21 (δ_H 4.20) correlated with C-16 (δ_C 210.3, s) and C-17, H-13 correlated with C-16, and that H-17 correlated with C-21 (δ_C 57.6, t). This evidence, along with two proton spin systems deduced from the ¹H–¹H COSY correlations, H₂-17/H₂-21, and H-9/H₂-11/H₂-12/H-13/H₂-14 suggested the partial structure, part b (Fig. 2).

Moreover, the key HMBC correlations of H-6 with C-8 (δ_C 111.4, s) and of H-9 (δ_H 2.65, 1H, dd, 13.1, 3.8 Hz) with C-1, C-5, C-8, C-10, C-11, and C-20 permitted the partial structures, part a and part b, to be connected through a carbon–carbon connection between C-9 and C-10 and an oxo bridge between C-6 and C-8. The molecular formula, C₂₀H₂₆O₇, of 1 demonstrated eight degrees of unsaturation, indicating the existence of a lactone group between C-7 and C-8, and this assignment was also supported by the similar chemical shift of C-7 (δ_C 172.5, s) and C-8 (δ_C 111.4, s) compared to those of norstaminolactone A¹⁵ (a norstaminane-type diterpenoid bearing similarly structural unit with its C-14 at δ_C 172.6 and C-8 at δ_C 107.0). In the ROESY spectrum of 1, the NOE correlations of Me-19/H₂-20 and H-11α/H₂-20/H-13 suggested that H-13, Me-19, and C-20 all adopted an α-orientation. The cross-peaks between H-3/H-5, H-5/H-9, and H-5/Me-18 in the ROESY spectrum demonstrated that H-3, H-5, H-9, and Me-18 were β-oriented (Fig. 2).

The Nuclear Overhauser Effect (NOE) is commonly recognized as one of the best approaches for structural and conformational analyses. However, if the internuclear distance was less than 3 Å, even for the two spin systems bearing opposite orientations, the NOE correlations could also be observed. Then, the NOE experiment often fails to predict the structures of these types of compounds. For example, the structure of rubescensin S, an ent-kauranoid, was revised to account for an error in assigning the configuration at C-13 by using NOE.¹⁶ Therefore, more evidence should be provided when determining the configuration of structures, especially when assigning their stereochemistry.

Certainly, there was only one possible orientation for H-6, α or β. However, both NOE correlations between H-6 and Me-18β and between H-6 and Me-19α could be observed, which suggested that H-6 of 1 might have two possible orientations. This evidence implied that compound 1 had two possible structures, 1 and 6,8-epi-1 (Fig. 1). X-ray diffraction or chemical transformation of 1 was scarcely finished. Under the circumstances, calculation of the ¹³C NMR chemical shifts of 1 and 6,8-epi-1 were performed in order to confirm the definite structure. After all the conformers of 1 and 6,8-epi-1 were optimized at B3LYP/6-31G(d,p), the ¹³C NMR chemical shifts of all the conformers were calculated with the GIAO method at mPW1PW91/6-31G(d,p) level, which has been reported to be applicable for highly oxygenated diterpenoids.¹⁷

As shown in Table 1, the largest error and the mean average error of the Boltzmann-averaged ¹³C NMR chemical shifts were 7.1 (C-12) and 1.5 ppm, respectively. Overall, the calculated ¹³C NMR chemical shifts of 1 and 6,8-epi-1 were both in good agreement with the structure elucidated by the NMR data (Table 1). Since the 7,8:15,16-di-seco-15-nor-21-homo-ent-kaurane skeleton of 1 was definite, the orientation of H-6 was still not assigned due to the weak errors of the calculated ¹³C NMR data between 1 and 6,8-epi-1 in C-5, C-6, C-7, C-8, C-9, and C-10 (Fig. 3). In addition, from the calculations given in the ESI† and Fig. 2, the distance between H-6 to H3–18 and H3–19 in the most stable conformer, 1g, are 2.90 and 2.43 Å, respectively. These distances are 3.00 and 2.14 Å in the most stable conformer, 6,8-epi-1, respectively. This evidence shows why the NOE could not be used to analyse the relative configuration of C-6 of 1.

Fig. 3 Partial comparison of calculated chemical shifts for two possible structures, 1 (red) and 6,8-epi-1 (blue), with experimentally observed shifts.

Therefore, the calculated electronic circular dichroism (ECD) spectra for (3R,5R,6S,8S,9S,10R,13R)-1 and (3R,5R,6R,8R,9S,10R,13R)-6,8-epi-1 were performed using time-dependent density-functional theory (TDDFD) method20 at B3LYP-SCRF/6-31G++(d,p) level with PCM in methanol,¹⁸ in order to determine its absolute configuration.

The calculated ECD spectra of (3R,5R,6S,8S,9S,10R,13R)-1 were compatible with the experimental ECD curve (Fig. 4). Molecular orbital (MO) analysis of the predominant conformer, 1g, at B3LYP/6-31G++(d,p) level with PCM in methanol, gave us the experimental ECD spectra (Fig. 5). The curve peak at 293 nm correlated to two positive rotatory strengths at 289.9 and 280.8 nm, which resulted from the electronic transitions from MO101 to MO102, involving an n → π* transition in the carbonyl group of ring A, and from MO100 to MO103, involving an n → π* transition in the side chain, respectively. The curve trough at 195 nm might arise from the electronic transitions from MO97 to MO104, involving an n → π* transition in the γ-lactone group of ring D, which gave rise to the negative rotatory strengths at 194.8 nm. As a result, the absolute configuration of 1 was established as shown.

Fig. 4 Experimental ECD of 1 (black), calculated ECD of 1 in methanol (red), and calculated ECD of 6,8-epi-1 in methanol (blue).

Fig. 5 ECD of the important MOs of the optimized conformer, 1g.

Three 15,16-seco-ent-kauranoids, compounds 3–5, have been isolated from the title plant, and their possible biogenetic route had been discussed in the literature.^13f The hypothetical biogenetic pathway of 1 was biogenetically inspired from 3–5, and could be plausibly traced back to laxiflorol B (2) and rabdonervosin J (6).¹⁹ The formation of intermediate C from intermediate A and intermediate B, by rearrangement involving the ring-cleavage reactions, decarboxylation, and aldol condensation route, is the key step to form compound 1.

Compounds 1 and 2 were tested for in vitro cytotoxicity against A-549, MCF-7, SMMC-7721, SW-480 and HL-60 human cancer cell lines using the MTT method;²⁰ cis-Platin was used as the positive control. Only compound 2 showed weak cytotoxic activity against the above cell lines with IC₅₀ values of 26.06, 16.63, 35.64, 24.29, and 17.92 μM, respectively.

3. Conclusions

In summary, this paper describes the isolation and structure elucidation of laxiflorol A (1), which is the first example of 7,8:15,16-di-seco-ent-kauranoid, together with its biogenetic precursor analogue, laxiflorol B (2), obtained from I. eriocalyx var. laxiflora. The structure of laxiflorol A (1) might be biosynthetically formed from multiple ring cleavages and adding (C-21) and losing carbons (C-15) on an ent-kauranoid. Although more than 1000 natural ent-kauranoids have been reported in the literature to date, laxiflorol A (1) is a novel compound, bearing a 7,8:15,16-di-seco-15-nor-21-homo-ent-kaurane skeleton.

Acknowledgements

This project was supported financially by the NSFC-Joint Foundation of Yunnan Province (U1302223), the National Natural Science Foundation of China (no. 21322204, 81172939, and 21402213), the reservation-talent project of Yunnan Province (2011CI043), and the Science and Technology Program of Yunan Province (no. 2008IF010).

Notes and references

1. (a) J. W. Li and J. C. Vederas, Science, 2009, 325, 161; (b) S. Dandapani and L. A. Marcaurelle, Nat. Chem. Biol., 2010, 6, 861; (c) J. Medina-Franco, Drug Des., 2012, 1, e105.
2. H. D. Sun, S. X. Huang and Q. B. Han, Nat. Prod. Rep., 2006, 23, 673.
3. (a) C. X. Liu, Q. Q. Yin, H. C. Zhou, Y. L. Wu, J. X. Pu, L. Xia, W. Liu, X. Huang, T. Jiang, M. X. Wu, L. C. He, Y. X. Zhao, X. L. Wang, W. L. Xiao, H. Z. Chen, Q. Zhao, A. W. Zhou, L. S. Wang, H. D. Sun and G. Q. Chen, Nat. Chem. Biol., 2012, 8, 486; (b) Z. M. Gu, Y. L. Wu, M. Y. Zhou, C. X. Liu, H. Z. Xu, H. Yan, Y. Zhao, Y. Huang, H. D. Sun and G. Q. Chen, Blood, 2010, 116, 5289.
4. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2012, 29(10), 1033.
5. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2012, 29, 435.
6. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2008, 25, 647.
7. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2006, 23, 841.
8. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2005, 22, 11–14.
9. R. A. Hill and A. Sutherland, Nat. Prod. Rep., 2004, 21, H9.
10. R. A. Hill and A. R. Pitt, Nat. Prod. Rep., 1997, 14, iii.
11. T. Zhen, C. F. Wu, P. Liu, H. Y. Wu, G. B. Zhou, Y. Lu, J. X. Liu, Y. Liang, K. K. Li, Y. Y. Wang, Y. Y. Xie, M. M. He, H. M. Cao, W. N. Zhang, L. M. Chen, K. Petrie, S. J. Chen and Z. Chen, Sci. Transl. Med., 2012, 4, 127ra38.
12. Y. Lu, B. Chen, J. H. Song, T. Zhen, B. Y. Wang, X. Li, P. Liu, X. Yang, Q. L. Zhang, X. D. Xi, S. D. Chen, J. P. Zuo, Z. Chen and S. J. Chen, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 2258.
13. (a) W. G. Wang, X. N. Li, X. Du, H. Y. Wu, X. Liu, J. Su, Y. Li, J. X. Pu and H. D. Sun, J. Nat. Prod., 2012, 75, 1102; (b) W. G. Wang, X. Du, X. N. Li, H. Y. Wu, X. Liu, S. Z. Shang, R. Zhan, C. Q. Liang, L. M. Kong, Y. Li, J. X. Pu and H. D. Sun, Org. Lett., 2012, 14, 302; (c) W. G. Wang, H. Y. Wu, X. Du, J. M. Yan, Y. Li, J. X. Pu and H. D. Sun, Chin. J. Chem., 2012, 30, 1226; (d) W. G. Wang, X. Du, X. N. Li, B. C. Yan, M. Zhou, H. Y. Wu, R. Zhan, K. Dong, J. X. Pu and H. D. Sun, Nat. Prod. Bioprospect., 2013, 3, 1; (e) W. G. Wang, B. C. Yan, X. N. Li, X. Du, H. Y. Wu, R. Zhan, Y. Li, J. X. Pu and H. D. Sun, Tetrahedron, 2014, 70, 7445; (f) W. G. Wang, X. N. Li, X. Du, K. Dong, W. Zhao, H. Y. Wu, L. M. Kong, Y. Li, J. X. Pu and H. D. Sun, Tetrahedron Lett., 2012, 53, 2777.
14. H. Flack and G. Bernardinelli, Acta Crystallogr., Sect. A: Found. Crystallogr., 1999, 55, 908.
15. S. Awale, Y. Tezuka, A. H. Banskota, S. Shimoji, K. Taira and S. Kadota, Tetrahedron, 2002, 58, 5503.
16. (a) Q. B. Han, R. T. Li, J. X. Zhang and H. D. Sun, Helv. Chim. Acta, 2004, 87, 1119; (b) M. Zhang, Y. Zhang, W. Lu and F. J. Nan, Org. Biomol. Chem., 2011, 9, 4436.
17. S. D. Rychnovsky, Org. Lett., 2006, 8, 2895.
18. (a) G. Bringmann, T. Bruhn, K. Maksimenka and Y. Hemberger, Eur. J. Org. Chem., 2009, 2717; (b) N. Berova, L. Di Bari and G. Pescitelli, Chem. Soc. Rev., 2007, 36, 914.
19. Y. H. Gao, Z. X. Wei, Y. Cheng, T. H. Wang, L. Ni, X. H. Zhou and L. Y. Zhang, Chem. Biodiversity, 2013, 10, 1487.
20. M. C. Alley, D. A. Scudiero, A. Monks, M. L. Hursey, M. J. Czerwinski, D. L. Fine, B. J. Abbott, J. G. Mayo, R. H. Shoemaker and M. R. Boyd, Cancer Res., 1988, 48, 589.

---

Footnotes

† Electronic supplementary information (ESI) available: Detailed experimental procedures, method of cytotoxicity test, physico-chemical properties, 1D and 2D NMR, MS, UV, ORD spectra of compounds 1 and 2, ECD spectra of compound 1, and X-ray crystal structure of 2. CCDC 970056. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra12704h

‡ Both authors contributed equally to this work.

§ X-ray crystallographic data of laxiflorol B (2): C₂₂H₂₈O₆, M = 388.44, orthorhombic, a = 9.4749(3) Å, b = 11.6421(3) Å, c = 17.1244(5) Å, α = 90.00°, β = 90.00°, γ = 90.00°, V = 1888.95(9) ų, T = 100(2) K, space group P2₁2₁2₁, Z = 4, 9900 reflections measured, 3262 independent reflections (R_int = 0.0350). The final R₁ values were 0.0354 (I > 2σ(I)). The final wR(F²) values were 0.0987 (I > 2σ(I)). The final R₁ values were 0.0354 (all data). The final wR(F²) values were 0.0988 (all data). Flack parameter = −0.03(15).

---