Article

pubs.acs.org/jnp

Cytotoxic ent-Kaurane Diterpenoids from cavaleriei † ‡ † ‡ † † † † ‡ Heng Zheng, , Qiong Chen, , Mengke Zhang, Yongji Lai, Liang Lei, Penghua Shu, Jinwen Zhang, † † § † † Yongbo Xue, Zengwei Luo, Yan Li, Guangmin Yao,*, and Yonghui Zhang*, † Key Laboratory of Natural Medicinal Chemistry and Resource Evaluation, School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, People’s Republic of China ‡ Tongji Hospital Affiliated to Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, People’s Republic of China § State Key Laboratory of Phytochemistry and Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, People’s Republic of China

*S Supporting Information

ABSTRACT: Fifteen new ent-kaurane diterpenoids, compounds 1−15, and two known analogues, 4-epi-henryine A (16) and leukamenin E (17), were isolated from the whole of Salvia cavaleriei. The structures of the new compounds were established by spectroscopic methods, and their absolute configurations were determined by electronic circular dichroism and single- crystal X-ray diffraction analyses with Cu Kα radiation. Compounds 1−15 were evaluated for their cytotoxicity against five human cancer cell lines, HL-60, SMMC-7721, A-549, MCF-7, and SW480, as well as the noncancerous Beas-2B cell line. Compounds 1−10, 12, 14,and15 showed broad-spectrum cytotoxicity, with compounds 1, 3, 6−10, 12, and 15 exhibiting more potent cytotoxicity than the positive control, cis-platin, with IC50 values ranging from 0.65 to 6.4 μM.

Salvia L., the largest genus of the family , comprises active components. In the process, 15 new ent-kaurane over 1000 species, which are widely distributed in tropical and diterpenoids, compounds 1−15, and two known ent-kaurane temperate regions of the world,1 and some species have been diterpenoids, 4-epi-henryine A (16)14 and leukamenin E (17),15 cultivated for use as herbal medicines and ornamentals.2,3 were isolated. Here, we describe the isolation, structure Studies on the chemical constituents of Salvia revealed the elucidation, and cytotoxicity of compounds 1−15. presence of polyphenolics2 and terpenoids.3,4 The identified diterpenoids mainly belong to seven skeletal types, abietane, ■ RESULTS AND DISCUSSION clerodane, pimarane, labdane, ent-kaurane, icetexane, and 11α-Hydroxyleukamenin E (1) was obtained as colorless 3,4 apianane. Collectively, these compounds possess antibacte- needles, mp 195−197 °C. The molecular formula of 1 was rial, antileishmanial, antimicrobial, antioxidant, antispasmolytic, − assigned as C22H32O6 by the HRESIMS of the pseudomolecular 3 6 + antituberculosis, and antitumor activities. Interestingly, ion [M + Na] at m/z 415.2067 (calcd for C22H32O6Na, although ent-kaurane diterpenoids are common in plants of 415.2097), indicating seven indices of hydrogen deficiency. The the genus Isodon (Lamiaceae),7 only one ent-kaurane UV absorption maximum at 234 nm was consistent with the diterpenoid has been previously reported from the Salvia presence of an α,β-unsaturated carbonyl group. The IR − genus.8 spectrum suggested the presence of hydroxy (3394 cm 1), an − − Salvia cavaleriei Levĺ is endemic to China and is distributed in ester carbonyl (1725 cm 1), a double bond (1647 cm 1), and − the , , , , Hubei, , conjugated carbonyl (1705 cm 1) functionalities. The 1H NMR 9 δ , , and provinces. The whole plant of S. spectrum (Table 1) showed resonances for three methyls ( H δ cavaleriei is used as a Chinese herbal medicine to treat 0.92, H3-18; 0.99, H3-19; 1.34 H3-20), an acetyl ( H 2.07), four δ β β hematemesis, metrorrhagia, dysentery with bloody stools, and oxymethines ( H 3.98, ddd, H-11 ; 4.07, dd, H-7 ; 4.62, t, H- traumatic hemorrhage.10 Previous phytochemical investigations 3α; 5.10, s, H-14α), and two olefinic protons arising from an 11 δ on S. cavaleriei var. cavaleriei and S. cavaleriei var. exocyclic double bond ( H 6.04, s, H-17a; 5.42, s, H-17b). The 13 simplicifolia12 have resulted in the isolation of depsides, C NMR and DEPT spectra (Table 4) of 1 displayed a total of δ phenolic glycosides, triterpenoids, and steroids. In our search 22 carbon resonances assignable to a ketocarbonyl ( C 209.0, 13 δ δ for anticancer agents from Chinese herbal medicines, we C-15), an acetyl ( C 172.7, 21.3), an exocyclic double bond ( C found that the 95% EtOH extract of the whole plants of S. cavaleriei showed significant cytotoxicity against the HL-60 cell Received: July 24, 2013 line, which encouraged us to pursue the characterization of the Published: November 20, 2013

© 2013 American Chemical Society and American Society of Pharmacognosy 2253 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article

dichroism (ECD) spectrum.17 The NOESY correlations of H- 11 to H-9β and H-12β and the large coupling constants J = δ 13.0, 8.0, and 6.0 Hz of H-11 ( H 3.98, ddd) with H-9 and H- 12 established the β-orientation of H-11. Analyses of the 2D NMR data, including HSQC, 1H−1H COSY, HMBC, and NOESY (Figure 1), defined compound 1 as 7α,11α,14β- trihydroxy-3β-acetoxy-ent-kaur-16-en-15-one. The structure was confirmed by single-crystal X-ray diffraction analysis, and its absolute confi guration was assigned as 3S,5S,7R,8R,9S,10R,11R,13S,14R based on a Flack parameter of 0.19(14)18 (Figure 2). 11β-Hydroxyleukamenin E (2) exhibited a pseudomolecular ion at m/z 415.2065 [M + Na]+ in the HRESIMS, which is in agreement with the molecular formula C22H32O6 (calcd for C22H32O6Na, 415.2097). Compound 2 has the same molecular formula as 1, and the NMR resonances for 2 (Tables 1 and 4) resemble those of 1. The obvious difference is that the multiplicity and coupling constants of H-11 are a doublet (J = 4.5 Hz) in 2 instead of a doublet of doublets of doublets (J = δ 13.0, 8.0, 6.0 Hz) in 1. Correspondingly, C-11 ( C 65.5) in 2 δ was shifted upfield compared to the corresponding signal in 1 150.2, C-16; 117.1, C-17), three methyls ( C 29.2, C-18; 22.5, δ (δ 69.5). These differences suggest that 2 is the 11-epimer of C-19; 19.3, C-20), four methylenes ( C 41.1, C-12; 36.2, C-1; C δ 1. The α-orientation of H-11 in 2 was determined by the 29.3, C-6; 24.0, C-2), four oxymethines ( C 79.2, C-3; 76.3, C- δ NOESY correlations between H-11 and both H-1α and H - 14; 76.0, C-7; 69.5, C-11), three methines ( C 61.2, C-9; 49.3, 3 δ 20α. Detailed 2D NMR analyses confirmed the identification of C-5; 47.4, C-13), and three quaternary carbons ( C 62.8, C-8; 42.6, C-10; 38.1, C-4). The 13C NMR spectrum of 1 is similar 2 as 7α,11β,14β-trihydroxy-3β-acetoxy-ent-kaur-16-en-15-one. to that of the co-occurring known compound 17 (leukamenin The absolute configuration of 2 was determined as E),15 except for a C-11 oxymethine group in 1 replacing a 3S,5S,7R,8R,9S,10R,11S,13S,14R by single-crystal X-ray diffrac- methylene in 17. Thus, 1 is the 11-OH derivative of 17. 1H−1H tion with Cu Kα radiation (Figure 3). COSY correlations from H-11 to both H-9 and H-12, together The molecular formula of 11-oxoleukamenin E (3) was with HMBC correlations of H-11 to C-8 and C-10, supported determined to be C22H30O6 by the HRESIMS ion at m/z + the above assignment. The NOESY correlations (Figure 1) of 413.1923 [M + Na] (calcd for C22H30O6Na, 413.1940). H-14/H3-20 and H-5/H-9 allowed assignment of 1 as an ent- Comparison of the NMR data of 3 (Tables 1 and 4) with those 16 δ kaurane diterpenoid, which was further supported by the of 2 indicated that a C-11 oxo ( C 206.8) group in 3 replaced ff λ δ negative Cotton e ect at max 342 nm in the electronic circular the oxymethine function in 2 ( C 65.5). This suggestion was

Table 1. 1H NMR [δ, mult (J in Hz)] Data for Compounds 1−5 (400 MHz) position 1a 2a 3b 4a 5a 1α 2.52 ddd (14.0, 3.5, 3.3) 1.64 overlap 1.32 ddd (13.8, 3.3, 3.3) 2.57 ddd (14.5, 8.5, 6.6) 1.58 overlap 1β 1.37 ddd (14.0, 13.3, 3.8) 1.40 ddd (14.0, 13.5, 3.0) 1.41 ddd (13.8, 12.1, 3.0) 1.81 ddd (14.5, 10.5, 4.8) 1.24 m 2α 1.97 overlap 2.00 m 1.74 m 2.25 ddd (15.5, 8.5, 4.8) 1.90 overlap 2β 1.54 dddd (15.3, 3.4, 3.4, 3.3) 1.64 overlap 1.59 dddd (15.3, 3.3, 3.0, 3.0) 2.70 ddd (15.5, 10.5, 6.6) 1.53 overlap 3α 4.62 t (2.6) 4.66 t (2.8) 4.61 t (2.6) 3.36 t (3.5) 5β 1.47 dd (12.2, 1.5) 1.48 dd (12.5, 1.4) 1.43 dd (12.9, 1.4) 1.96 dd (13.0, 2.0) 1.49 dd (12.4, 1.9) 6α 1.70 ddd (12.8, 12.2, 12.0) 1.70 ddd (12.7, 12.5, 12.0) 1.82 ddd (12.9, 12.4, 11.9) 1.76 ddd (13.0, 12.8, 11.9) 1.84 overlap 6β 1.88 ddd (12.8, 3.9, 1.5) 1.90 ddd (12.7, 4.2, 1.4) 1.93 dd (12.4, 2.6) 1.89 ddd (12.8, 3.4, 2.0) 1.92 overlap 7β 4.07 dd (12.0, 4.0) 4.21 dd (12.0, 4.2) 4.37 dd (11.9, 4.3) 4.09 dd (11.9, 3.4) 4.22 dd (11.9, 4.6) 9β 1.69 d (8.0) 1.48 br s 1.81 br s 1.84 d (8.0) 1.75 br s 11α 3.96 d (4.5) 11β 3.98 ddd (13.0, 8.0, 6.0) 3.98 ddd (12.8, 8.0, 5.6) 12α 2.24 ddd (13.0, 12.8, 2.5) 2.22 ddd (14.7, 4.5, 3.1) 2.74 dd (16.5, 3.2) 2.16 ddd (12.8, 12.8, 1.8) 2.94 dd (16.5, 3.5) 12β 1.98 overlap 2.05 ddd (14.7, 3.8, 2.2) 2.57 br d (16.5) 1.99 ddd (12.8, 5.2, 5.6) 2.53 ddd (16.5, 3.0, 1.8) 13α 3.05 br s 2.96 br s 3.19 br s 3.06 br s 3.22 br s 14α 5.10 s 4.90 s 5.29 s 5.01 s 5.39 br s 17a 6.04 s 5.95 s 6.20 s 6.05 s 6.19 s 17b 5.42 s 5.34 s 5.53 s 5.43 s 5.61 s 18 0.92 s 0.94 s 0.87 s 1.15 s 0.99 s 19 0.99 s 0.98 s 0.91 s 1.10 s 0.91 s 20 1.34 s 1.08 s 1.11 s 1.21 s 1.17 s OAc 2.07 s 2.06 s 2.01 s a b Recorded in methanol-d4. Recorded in CDCl3.

2254 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article

Table 2. 1H NMR [δ, mult (J in Hz)] Data for Compounds 6−10 (400 MHz)

position 6a 7a 8a 9a 10b 1α 1.84 dd (14.1, 5.8) 1.51 m 1.77 ddd (13.0, 3.8, 2.6) 1.54 m 1.40 m 1β 1.71 dd (14.1, 9.0) 1.33 ddd (13.5, 13.5, 3.6) 1.53 ddd (13.5, 13.0, 3.7) 1.07 ddd (13.4, 12.7, 2.8) 1.40 m 2α 2.48 dd (9.0, 5.8) 1.98 m 2.01 m 1.62 m 1.75 m 2β 2.48 dd (9.0, 5.8) 1.62 dddd (15.2, 3.3, 3.2, 3.2) 1.67 dddd (15.0, 3.5, 3.3, 3.2) 1.49 m 1.58 dddd (15.2, 3.2, 3.1, 3.0) 3α 4.66 t (2.6) 4.69 t (2.7) 1.43 m 4.60 dd (2.7, 2.3) 3β 1.36 m 5β 1.75 dd (12.4, 7.0) 1.51 dd (12.4, 1.2) 1.39 dd (9.8, 1.8) 1.34 dd (11.6, 2.4) 1.43 dd (11.6, 1.5) 6α 1.92 ddd (13.0, 12.4, 9.5) 1.71 ddd (12.4, 12.4, 12.1) 1.75 ddd (12.7, 12.2, 9.8) 1.85 ddd (12.4, 11.6, 11.6) 1.78 ddd (12.6, 11.9, 11.6) 6β 1.98 ddd (13.0, 7.0, 6.8) 1.91 ddd (12.4, 4.6, 1.2) 1.86 ddd (12.7, 5.3, 1.8) 1.92 ddd (12.4, 4.8, 2.4) 1.89 ddd (12.6, 4.3, 1.5) 7β 4.22 dd (9.5, 6.8) 4.32 dd (12.1, 4.6) 4.39 dd (12.2, 5.3) 4.16 dd (11.6, 4.8) 4.23 dd (11.9, 4.3) 9β 1.84 br s 1.84 dd (3.6, 1.1) 1.41 d (4.2) 1.71 br s 1.71 br s 11α 5.47 dd (9.6, 3.6) 11β 3.10 dd (4.2, 4.1) 12α 2.95 dd (16.6, 3.6) 6.05 ddd (9.6, 6.9, 1.4) 2.93 dd (16.5, 3.6) 2.49 dd (17.6, 3.9) 12β 2.55 ddd (16.6, 3.5, 2.0) 3.14 dd (3.8, 3.8) 2.53 ddd (16.5, 3.2, 1.8) 2.66 ddd (17.6, 2.4, 2.0) 13α 3.24 ddd (3.6, 3.5, 1.1) 3.28 br d (6.9) 3.39 dd (3.8, 1.3) 3.21 dd (3.6, 3.2) 2.79 dd (3.9, 2.4) 14α 5.37 d (1.1) 5.03 d (1.4) 5.04 d (1.3) 5.38 s 5.34 s 16α 3.23 dd (9.4, 4.7) 17a 6.18 s 5.77 s 6.11 s 6.18 s 3.59 dd (9.9, 4.7) 17b 5.61 s 5.18 s 5.56 s 5.60 s 3.36 dd (9.9, 9.4) 18a 1.15 s 0.95 s 0.94 s 3.97 d (11.1) 0.85 s 18b 3.65 d (11.1) 19 1.11 s 0.98 s 1.01 s 0.91 s 0.89 s 20 1.20 s 1.07 s 1.40 s 1.21 s 1.08 s OAc 2.07 s 2.08 s 2.09 s 2.00 s OMe 3.25 s a b Recorded in methanol-d4. Recorded in CDCl3.

1 δ J − d Table 3. H NMR [ , mult ( in Hz)] Data for Compounds 11 15 (400 MHz) in Methanol- 4 position 11 12 13 14 15 1α 1.35 ddd (13.6, 3.4, 3.2) 1.42 m 1.85 overlap 1.56 overlap 1.64 overlap 1β 1.44 overlap 1.42 m 1.72 overlap 1.08 m 1.06 ddd (14.2, 13.3, 4.2) 2α 1.91 overlap 1.91 overlap 2.47 dd (9.0, 5.8) 1.63 overlap 1.49 overlap 2β 1.59 dddd (15.2, 3.4, 3.4, 3.2) 1.59 dddd (15.4, 3.3, 3.2, 3.2) 2.47 dd (9.0, 5.8) 1.51 overlap 1.61 overlap 3α 4.64 t (2.6) 4.64 t (2.7) 1.36 overlap 1.42 overlap 3β 1.43 overlap 1.35 overlap 5β 1.43 overlap 1.46 dd (12.6, 1.6) 1.71 overlap 1.30 dd (12.0, 1.8) 1.32 overlap 6α 1.82 ddd (12.6, 12.4, 12.0) 1.82 ddd (12.6, 12.5, 12.0) 1.93 overlap 1.80 ddd (12.8, 12.0, 11.8) 1.80 ddd (12.5, 12.3, 11.8) 6β 1.92 overlap 1.93 overlap 1.93 overlap 1.89 ddd (12.8, 4.7, 1.8) 1.88 ddd (12.5, 4.5, 2.3) 7β 4.11 dd (12.0, 4.6) 4.07 dd (12.0, 4.4) 4.11 dd (10.3, 6.1) 4.05 dd (11.8, 4.7) 4.01 dd (11.8, 4.5) 9β 1.70 d (1.4) 1.65 s 1.81 d (1.4) 1.68 br s 1.63 br s 12α 2.88 dd (16.6, 3.6) 2.69 d (3.5) 2.89 dd (16.6, 3.6) 2.87 dd (16.5, 3.8) 2.66 m 12β 2.53 ddd (16.6, 3.3, 1.4) 2.69 d (3.5) 2.55 ddd (16.6, 3.3, 1.4) 2.52 ddd (16.5, 3.2, 1.9) 2.62 m 13α 2.69 dd (3.6, 3.3) 2.78 m 2.70 dd (3.6, 3.3) 2.67 dd (3.8, 3.2) 2.76 m 14α 5.37 s 5.44 br s 5.35 s 5.36 s 5.44 br s 16α 3.22 m 3.22 m 16β 2.32 t (7.0) 2.34 t (7.1) 2.31 t (7.1) 17a 3.74 d (7.0) 3.71 dd (10.0, 4.6) 3.69 d (7.1) 3.68 d (7.1) 3.63 d (10.1) 17b 3.74 d (7.0) 3.42 overlap 3.69 d (7.1) 3.68 d (7.1) 3.37 d (10.1) 18a 0.93 s 0.93 s 1.14 s 3.95 d (11.2) 3.95 d (11.1) 18b 3.64 d (11.2) 3.64 d (11.1) 19 0.99 s 0.99 s 1.10 s 0.91 s 0.91 s 20 1.19 s 1.17 s 1.19 s 1.20 s 1.18 s OAc 2.06 s 2.07 s 2.08 s 2.08 s OMe 3.34 s 3.33 s 3.29 s OEt 3.50 m 3.48 m 1.18 t (7.0) 1.15 t (7.0)

2255 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article

Table 4. 13C NMR Data for Compounds 1−15 (100 MHz)

position 1a 2a 3b 4a 5a 6a 7a 8a 9a 10b 11a 12a 13a 14a 15a 1 36.2 34.2 32.9 41.9 33.7 38.6 34.1 34.0 40.2 32.7 34.4 34.2 38.6 40.3 40.1 2 24.0 23.9 22.6 34.9 26.2 34.9 23.7 23.5 18.7 22.6 23.7 23.6 34.9 18.7 18.7 3 79.2 79.2 77.0 221.2 75.9 218.4 79.1 79.0 36.3 77.0 78.7 78.7 218.5 36.3 36.3 4 38.1 37.9 36.9 48.3 38.8 48.3 37.8 37.5 37.8 36.9 38.0 38.0 48.3 37.8 37.8 5 49.3 48.8 47.1 53.1 47.1 52.6 47.7 48.1 47.9 47.2 48.5 48.7 52.6 47.9 48.2 6 29.3 29.4 27.5 31.2 29.4 30.9 29.5 29.9 29.7 27.9 29.5 29.4 31.1 29.9 29.8 7 76.0 75.7 73.3 75.0 74.2 73.5 75.6 75.5 73.8 73.7 73.9 74.6 73.3 73.6 74.4 8 62.8 60.5 60.3 62.5 61.2 60.8 60.4 58.8 61.2 60.7 62.6 61.5 62.3 62.7 61.6 9 61.2 66.9 67.3 59.8 69.7 68.0 59.9 53.3 69.9 66.8 69.1 69.1 67.5 69.4 69.4 10 42.6 39.6 39.8 41.0 41.0 39.9 39.8 41.8 41.0 39.8 40.7 40.8 39.8 40.9 41.0 11 69.5 65.5 206.8 68.6 209.1 209.0 126.4 50.9 209.1 207.5 209.7 209.8 209.8 209.8 209.8 12 41.1 40.7 49.8 40.9 51.0 50.8 133.7 55.6 51.0 44.6 50.8 45.7 50.6 50.7 45.7 13 47.4 46.7 44.5 47.3 46.2 46.1 48.8 49.2 46.2 39.7 41.2 41.3 41.1 41.1 41.3 14 76.3 76.8 74.0 76.1 75.4 75.4 74.3 73.7 75.4 74.6 77.7 75.9 77.7 77.8 75.9 15 209.0 208.2 204.7 208.5 205.6 205.4 209.8 208.9 205.7 216.9 217.3 217.6 217.2 217.3 217.4 16 150.2 150.4 146.0 150.1 148.6 148.4 150.5 145.7 148.4 50.0 54.2 51.4 54.1 54.0 51.4 17 117.1 116.0 122.4 117.3 121.5 121.7 113.0 119.1 121.8 68.3 72.6 67.2 74.7 74.8 69.3 18 29.2 28.7 28.2 28.8 29.3 27.5 28.5 28.7 73.3 28.2 28.7 28.6 27.6 73.3 73.3 19 22.5 22.3 21.9 21.0 22.8 22.1 22.4 22.8 18.2 21.8 22.3 22.3 22.1 18.2 18.1 20 19.3 18.4 19.1 20.3 19.5 19.1 17.8 16.8 20.0 19.0 19.2 19.3 19.0 19.9 20.0 OAc 172.7 172.5 170.8 172.6 172.6 173.1 170.8 172.5 172.5 173.1 172.9 21.3 21.2 21.4 21.2 21.2 20.9 21.4 21.2 21.2 20.9 20.9 OMe 59.0 59.0 58.9 59.2 OEt 67.5 67.8 15.5 15.5 a b Recorded in methanol-d4. Recorded in CDCl3

Figure 1. 1H−1H COSY, key HMBC, and NOESY correlations of compound 1.

Figure 3. X-ray ORTEP drawing of compound 2. Figure 2. X-ray ORTEP drawing of compound 1. tal X-ray diffraction analysis by Cu Kα radiation (Figure 4) fi δ fi veri ed by the HMBC correlations from H-9 ( H 1.81, br s), established the absolute con guration of 3 as δ δ H2-12 ( H 2.74, dd; 2.57, br d), and H-13 ( H 3.19, br s) to the 3S,5S,7R,8R,9S,10R,13S,14R. C-11 carbonyl. Consequently, 3 was characterized as 7α,14β- 3-Oxo-11α-hydroxyleukamenin E (4) was obtained as β dihydroxy-3 -acetoxy-ent-kaur-16-ene-11,15-dione. Single-crys- colorless needles. Its molecular formula, C20H28O5, was inferred

2256 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article

the HRESIMS determined the molecular formula of 6 as 13 C20H26O5 (calcd for C20H26O5Na, 369.1678). The C NMR spectrum (Table 4) of 6 differed from that of 5 in the presence δ of a carbonyl ( C 218.4) at C-3 in 6, instead of an oxymethine δ in 5 ( C 75.9). The assignment of the C-3 carbonyl group was δ further supported by the HMBC correlations from H2-1 ( H δ δ δ 1.84 and 1.71), H2-2 ( H 2.48), H3-18 ( H 1.15), and H3-19 ( H δ fi 1.11) to the carbonyl ( C 218.4). Compound 6 was identi ed as 7α,14β-dihydroxy-ent-kaur-16-ene-3,11,15-trione by detailed 2D NMR analyses. The similarity of the ECD spectra of 6 and 5 indicates that 6 possesses the same absolute configuration as that of 5, except at C-3. 11,12-Didehydroleukamenin E (7) had the molecular Figure 4. X-ray ORTEP drawing of compound 3. + formula C22H30O5 by HRESIMS (m/z 397.1976 [M + Na] , calcd for C22H30O5Na, 397.1991), implying eight indices of from the pseudomolecular ion peak at m/z 371.1822 [M + hydrogen deficiency. The NMR data of 7 (Tables 2 and 4) + Na] in the HRESIMS (calcd for C20H28O5Na, 371.1834). resembled that of 1, except for the presence of a double bond Comparison of the NMR spectra of 4 (Tables 1 and 4) with δ δ ( H 6.05, ddd, H-12; 5.47, dd, H-11; C 133.7, C-12; 126.4, C- those of 1 indicated that the acetyl signals in 1 were absent in 4, δ δ 11) in 7, instead of an oxymethine ( H 3.98, ddd, H-11; C and the C-3 oxymethine function in 1 was replaced by a 69.5, C-11) and a methylene (δ 2.24, H-12α; 1.98, H-12β; δ δ δ H C carbonyl ( C 221.2) in 4. HMBC correlations from H2-1 ( H 41.1, C-12) in 1. Correlations from the olefinic protons to H-9 δ δ 2.57 and 1.81), H2-2 ( H 2.25 and 2.70), H3-18 ( H 1.15), and and H-13 in the 1H−1H COSY spectrum indicated that the δ H3-19 ( H 1.10) to the carbonyl established the carbonyl double bond was located at C-11. This was confirmed by the placement at C-3. Comprehensive 2D NMR analyses supported HMBC correlations from H-9, H-13, and H-14 to C-11 and C- α α β the assignment of the structure of 4 as 7 ,11 ,14 -trihydroxy- 12. Two-dimensional NMR analyses allowed the unambiguous fi ent-kaur-16-ene-3,15-dione. The absolute con guration of 4 identification of 7 as 7α,14β-dihydroxy-3β-acetoxy-ent-kaur- was illustrated as 5S,7R,8R,9S,10R,11R,13S,14R by single- 11(12),16(17)-dien-15-one. In consideration of its biosynthetic ff α crystal X-ray di raction analysis with Cu K radiation (Figure origin, the absolute configuration of compound 7, except at C- 5). 11, is the same as that of compound 1. The molecular formula of 11α,12α-epoxyleukamenin E (8) was assigned as C22H30O6 based on HRESIMS (m/z 413.1926 + [M + Na] , calcd for C22H30O6Na, 413.1940), indicating eight indices of hydrogen deficiency. The NMR spectra (Tables 2 and 4) of 8 were similar to those of 1. The significant ff δ di erences were H-11 ( H 3.10, dd, J = 4.2, 4.1 Hz) and C-11 δ fi δ ( C 50.9) in 8 were shifted up eld compared to 1 ( H 3.98, δ ddd, J = 13.0, 8.0, 6.0 Hz, H-11; C 69.5, C-11), and an δ δ oxymethine ( H 3.14, H-12; C 55.6, C-12) in 8 replaces the C- δ 12 methylene ( C 41.1) in 1. Compound 8 possesses one more index of hydrogen deficiency than 1, which, in view of the chemical shift differences, is consistent with the presence of an 11,12-epoxide. This was confirmed by the HMBC cross-peaks δ δ δ for 8 from H-11 ( H 3.10) to C-8 ( C 58.8) and C-9 ( C 53.3) δ δ α and from H-12 ( H 3.14) to C-14 ( C 73.7). The -orientation of the 11,12-epoxy moiety was determined on the basis of the Figure 5. X-ray ORTEP drawing of compound 4. β NOESY correlations between H-11 and H-9 and H2-1 and between H-12 and H2-17. Thus, the structure of 8 was 3-Deacetyl-11-oxoleukamenin E (5) possessed the molecular established as 7α,14β-dihydroxy-3β-acetoxy-11α,12α-epoxy-ent- fi fi formula C20H28O5, con rmed by the HRESIMS ion at m/z kaur-16-en-15-one. Its structure was con rmed and its absolute + fi fi 371.1821 [M + Na] (calcd for C20H28O5Na, 371.1834). con guration de ned as 3S,5S,7R,8R,9S,10R,11S,12R,13S,14R Analysis of its NMR spectra (Tables 1 and 4) and comparison by single-crystal X-ray diffraction analysis (Figure 6). with those of 3 showed that 5 lacked an acetyl group compared 18-Deacetyl-4-epi-henryine A (9) was obtained as a white, fi δ with 3. The up eld shift of H-3 in 5 ( H 3.36, t) compared with amorphous powder. Its molecular formula was determined as δ + that of 3 ( H 4.61, t) suggested that the 3-acetoxy group in 3 C22H30O6 by HRESIMS at m/z 413.1923 [M + Na] (calcd for was replaced with a hydroxy group in 5. The cross-peaks from C22H30O6Na, 413.1940). Compound 9 has the same molecular ff H-3 to H2-2, H3-18, and H3-19 in the NOESY spectrum of 5,as formula as that of 3, but shows several di erences in its NMR well as the smaller coupling constant J = 2.6 Hz between H-3 data (Tables 2 and 4). First, 9 shows signals for two methyls α δ δ and H2-2, indicated that H-3 is -oriented. Subsequently, the ( H 1.21, s, H3-20; 0.91, s, H3-19; C 20.0, C-20; 18.2, C-19), β α β δ structure of 5 was determined as 3 ,7 ,14 -trihydroxy-ent-kaur- one fewer than in 3. Second, signals for an oxymethylene ( H fi δ δ 16-ene-11,15-dione. The absolute con guration of 5 is the same 3.97 and 3.65, d, H2-18; C 73.3, C-18) and a methylene ( H δ as that of 3, based on similar ECD spectra. 1.43 and 1.36, H2-3; C 36.6, C-3) are observed in 9, replacing a δ δ 3,11-Dioxoleukamenin E (6) was isolated as a white, methyl ( H 0.87, s, H3-18; C 28.2, C-18) and an oxymethine + δ δ amorphous powder. The [M + Na] ion at m/z 369.1665 in ( H 4.61, t, H-3; C 77.0, C-3) in 3. HMBC correlations of H2-

2257 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article

Compound 11 was isolated as colorless crystals. The molecular formula C24H36O7 was determined from the HRESIMS ion at m/z 459.2324 [M + Na]+ (calcd for C24H36O7Na, 459.2359). The NMR data of 11 (Tables 3 and ff δ 4) were similar to those of 10,di ering by an ethoxy group ( H δ 1.18, t, 3H; 3.50, m, 2H; C 67.5, 15.5) in 11 replacing the δ methoxy group in 10. HMBC correlations of H2-17 ( H 3.74, δ d) to the oxymethylene ( C 67.5) of the ethoxy group, and the δ oxymethylene protons ( H 3.50, m) of the ethoxy group to C- δ fi 17 ( C 72.6), con rmed the ethoxy group was located at C-17. α δ The NOESY correlations between H-17 and H-13 ( H 2.69, δ β δ dd), between H-16 ( H 2.32, t) and H-12 ( H 2.53, ddd), and α δ δ between H-12 ( H 2.88, dd) and H3-20 ( H 1.19, s) Figure 6. X-ray ORTEP drawing of compound 8. α established the -orientation of CH2-17. Detailed 2D NMR analyses established the structure of 11 as 7α,14β-dihydroxy- δ δ δ 17α-ethoxymethyl-3β-acetoxy-ent-kaur-11,15-dione. The abso- 18 with C-3 ( C 36.3), C-4 ( C 37.8), C-5 ( C 47.9), and C-19 δ lute confi guration was determined as ( C 18.2) indicated that the oxymethylene is located at C-18. In δ 3S,5S,7R,8R,9S,10R,13S,14R,16R from single-crystal X-ray fact, further correlations of H2-18 to the carbonyl ( C 173.1) ff α place an acetoxy group at C-18. The C-18 acetoxy methylene di raction analysis by Cu K radiation (Figure 8). group is β-oriented according to NOESY correlations between β β fi H2-18 and both H-5 and H-6 . As con rmed by 2D NMR data analyses, 9 was identified as 7α,14β-dihydroxy-18β- acetoxy-ent-kaur-16-ene-11,15-dione. The absolute configura- tion of 9 was assigned as 5S,7R,8R,9S,10R,13S,14R, based on the similarity of the ECD spectra of 9 and 3. The HRESIMS ion at m/z 445.2185 [M + Na]+ indicates a C23H34O7 (calcd for C23H34O7Na, 445.2202) molecular formula for compound 10. The 1H and 13C NMR data (Tables 2 and 4) of 10 resemble those of 3, except for an additional δ δ δ methoxy group ( H 3.25, s, 3H; C 59.0), an oxymethylene ( H δ Figure 8. X-ray ORTEP drawing of compound 11. 3.59, dd, H-17a; 3.36, dd, H-17b; C 68.3, C-17), and a methine (δ 3.23, dd, H-16; δ 50.0, C-16) in 10, replacing a C-16−C- H C − 17 exocyclic double bond in 3. The cross-peaks between H-13/ The molecular formulas of compounds 12 15 were H-16/H-17 in the 1H−1H COSY spectrum and the HMBC determined as C24H36O7,C21H30O6,C23H34O7,and δ C H O , respectively, according to their HRESIMS data. correlations from H2-17 and H-16 to C-13 ( C 39.7) and C-15 23 34 7 δ Comparison of their NMR data with those of compounds 3, 6, ( C 216.9) supported the assignment of the oxymethylene at C- 17 and the methine at C-16 in the five-membered ring. The and 9, respectively, suggested that compound 12 was an EtOH methoxy group explains the oxygenation at C-17, and its addition product of compound 3, compound 13 was a MeOH addition product of compound 6, and compounds 14 and 15 position is assigned based on the HMBC correlations from H2- 17 to the OCH carbon and from the methoxy protons to C-17. were MeOH addition products of compound 9. Detailed 2D 3 − NOESY correlations between H-17b and H-12β, between H- NMR data and ECD analyses assigned compounds 12 15 as α β β β 12α and H -20, and between H-16 and H-13α suggested that 7 ,14 -dihydroxy-17 -ethyoxymethyl-3 -acetoxy-ent-kaur- 3 α β α the C-17 methoxymethyl group is β-oriented. Consequently, 11,15-dione, 7 ,14 -dihydroxy-17 -methoxymethyl-ent-kaur- α β α β the structure of 10 was elucidated as 7α,14β-dihydroxy-17β- 3,11,15-trione, 7 ,14 -dihydroxy-17 -methoxymethyl-18 -ace- α β β methoxymethyl-3β-acetoxy-ent-kaur-11,15-dione. Analysis of toxy-ent-kaur-11,15-dione, and 7 ,14 -dihydroxy-17 -methox- β single-crystal X-ray diffraction (Figure 7) confirmed the ymethyl-18 -acetoxy-ent-kaur-11,15-dione, respectively. structure and established the absolute configuration of 10 as Since EtOH and MeOH were used during the extraction and − 3S,5S,7R,8R,9S,10R,13S,14R,16S. isolation, compounds 10 15 may be artifacts of compounds 3, 6, and 9, respectively. In order to verify this, compounds 3, 6, and 9 were separately dissolved in MeOH and EtOH and stirred overnight. Co-TLC detected the presence of com- pounds 10−15 in the solution of compounds 3, 6, and 9, respectively. To date, seven diterpene skeletal types, abietane, clerodane, pimarane, labdane, ent-kaurane, icetexane, and apianane, have been isolated from Salvia species. Among them, abietane and clerodane diterpenoids are the most common. Interestingly, abietane diterpenoids are mainly reported from Asian and European Salvia species, while clerodane diterpenoids were found commonly in American Salvia species.3,4 However, the ent-kaurane diterpenoid skeleton is rare in Salvia species. Only one ent-kaurane diterpene, verbenacine, was reported from a Figure 7. X-ray ORTEP drawing of compound 10. Salvia species (S. verbenaca).8 The isolation of the ent-kaurane

2258 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article − μ Table 5. Cytotoxicity of Compounds 1 15 (IC50 in M) against Five Human Cancer Cell Lines and the Noncancerous Human Beas-2B Cell Line

compound HL-60 SMMC-7721 A-549 MCF-7 SW480 BEAS-2B 1 2.5 1.1 3.5 3.4 1.4 2.8 2 9.7 3.8 >10 >10 5.5 >10 3 2.8 3.0 2.7 2.5 1.6 2.0 4 >10 >10 >10 >10 8.1 >10 5 >10 >10 >10 >10 7. 6 >10 6 3.3 4.9 6.4 3.1 2.2 3.1 7 2.1 2.8 3.1 1.7 2.8 0.88 8 0.65 1.3 1.1 3.0 1.8 0.73 9 0.71 0.96 0.94 1.8 1.7 0.87 10 3.7 4.0 2.9 2.5 2.0 2.5 11 >10 >10 >10 >10 >10 >10 12 3.3 3.5 3.5 3.1 3.0 2.8 13 >10 >10 >10 >10 >10 >10 14 >10 >10 >10 >10 6.7 >10 15 2.8 3.5 3.0 2.0 2.6 3.0 DDP (cis-platin) 1.2 6.4 9.2 15.9 13.4 11.1 paclitaxel <0.008 <0.008 <0.008 <0.008 <0.008 0.58 diterpenoids, compounds 1−17, suggests that S. cavaleriei showed more potent cytotocixity than compounds 11, 13, and expresses a nontypical chemical profile and may taxonomically 14, having a 16R configuration, thus suggesting that a 16S be related to S. verbenaca. configuration in ent-kaurane diterpenoids may enhance the Compounds 1−15 were evaluated for their cytotoxicity cytotocixity. As a result, a cyclopentanone conjugated with an against the human cancer cell lines HL-60, SMMC-7721, A- exomethylene group, a 3-acetoxy group, an 11α-OH, a Δ11(12) 549, MCF-7, and SW480 and an immortalized noncancerous double bond, an 11α,12α-epoxide, and 16S configuration in ent- cell line, Beas-2B, by the reported MTS method.13a,19 cis-Platin kaurane diterpenoids appear to enhance cytotoxicity. and paclitaxel were used as positive controls. As shown in Table 5, compounds 1−10, 12, 14, and 15 showed cytotoxicity, while ■ EXPERIMENTAL SECTION μ compounds 11 and 13 were inactive (IC50 >10 M). General Experimental Procedures. Melting points were Compounds 1, 3, 6−10, 12, and 15 exhibited more potent measured on an X-5 micromelting point apparatus without correction cytotoxicity than cis-platin. Among them, compound 9 showed (Beijing Tech Instrument Co. Ltd.). Optical rotations were determined in MeOH on a Perkin-Elmer 341 polarimeter. UV spectra the strongest activity, with IC50 values against HL-60, SMMC- 7721, A549, MCF-7, and SW480 cells lines ranging from 0.71 were recorded on a Varian Cary 50 spectrometer. ECD spectra were μ − obtained on a JASCO J-810 spectrometer. IR spectra were determined to 1.8 and 1.7 M. However, compounds 1, 3, 6 10, 12, and on a Bruker Vertex 70 instrument. NMR spectra were recorded on a 15 also showed strong cytotoxicity against the noncancerous Bruker AM-400 spectrometer, and the 1H and 13C NMR chemical δ Beas-2B cell line and did not exhibit selective cytotoxicity shifts were referenced to the solvent peaks for CDCl3 at H 7.24 and δ δ δ against the cancer cell lines tested. Compound 16, the 18- C 77.23 or for methanol-d4 at H 3.31 and C 49.15. HRESIMS were deacetyl derivative of 9, was reported to be noncytotoxic to the conducted in the positive-ion mode on a Thermo Fisher LC-LTQ- 20 Orbitrap XL spectrometer. The crystallographic data were obtained on A549 and MCF-7 cell lines, which suggested that the 18-O- ff acetyl group in 9 may be required for activity. Compounds 2, 4, a Bruker SMART APEX-II CCD di ractometer equipped with graphite-monochromatized Cu Kα radiation (λ = 1.54178 Å). Column 5, and 14 exhibited higher IC50 values against the noncancerous − μ chromatography (CC) was carried out on silica gel (100 200 mesh, Beas-2B cell line (IC50 >10 M) than against the SW480 cell 200−300 mesh, and 400 mesh, Qingdao Ocean Chemical Industry Co. μ ’ line, with IC50 values of 5.5, 8.1, 7.6, and 6.7 M, respectively, Ltd., People s Republic of China), RP C18 silica gel (ODS-A-HG, indicating at least a small selectivity in inhibiting SW480 cells YMC Co. Ltd., Japan), and Sephadex LH-20 (GE Healthcare Bio- (Table 5). In addition, compound 2 showed selective Sciences AB, Sweden). HPLC was conducted on an Agilent 1200 μ cytotoxicity against the SMMC-7721 cell line compared with instrument with detection at 210 or 230 nm using an RP C18 (5 m, × − μ 250 10 mm, YMC-pack ODS-A) column and MeOH H2Oor the Beas-2B cell line, with an IC50 value of 3.8 and >10 M, − fl MeCN H2O as the mobile phase at a ow rate of 1.5 mL/min. MPLC respectively. Comparison of the activities of 1 and 2 suggests fi α was carried out with an EZ Puri er III chromatography system. TLC that -orientation of the 11-OH increases cytotoxicity was performed with silica gel 60 F254 (Yantai Chemical Industry compared with the β-orientation. Although 4 possesses an Research Institute, Yantai, China) and RP-C18 F254 plates (Merck). 11α-OH, its lower cytotoxicity compared with that of 1 may be Plant Material. The whole plants of S. cavaleriei were collected at due to the lack of a 3-acetoxy group. In fact, compound 3 Enshi, Hubei Province, People’s Republic of China, in September possesses more potent cytotoxicity than 5, a further indication 2010. The plant material was authenticated by Dr. Jianping Wang at that the 3-acetoxy group is important for the cytotoxicity of the School of Pharmacy, Tongji Medical College, Huazhong these compounds. Compound 17 wasreportedtobe University of Science and Technology. A voucher specimen (No. noncytotoxic to the A-549 and MCF-7 cell lines;21 however, 20100901) has been deposited at Hubei Key Laboratory of Natural fi Medicinal Chemistry and Resource Evaluation, School of Pharmacy, compounds 1, 3, 7, and 8 had signi cant cytotoxicity against Tongji Medical College, Huazhong University of Science and Δ11(12) − these two cell lines. Thus, a double bond or a C-11 C- Technology. 12 epoxy group may be essential for activity. Interestingly, Extraction and Isolation. The dried whole plants of S. cavaleriei compounds 10, 12, and 15, which possess a 16S configuration, (20 kg) were extracted four times in 95% aqueous EtOH at room

2259 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article fi λ Δε − temperature. The ltrate was combined and concentrated under ECD (MeOH) max ( ) 352 ( 0.1), 311 (+1.5)228 (+2.2) nm; IR ff ν reduced pressure to a ord a residue (1.5 kg). The residue was (KBr) max 3381, 2955, 2877, 1733, 1698, 1645, 1460, 1388, 1142, −1 1 13 suspended in H2O and extracted with petroleum ether, CHCl3, EtOAc, 1090, 1071, 992, 775 cm ; H NMR data see Table 1; C NMR data + and n-BuOH, successively. The active CHCl3 extract (252.5 g) was see Table 4; positive HRESIMS m/z 371.1821 [M + Na] (calcd for − subjected to CC over silica gel (100 200 mesh) eluting with a C20H28O5Na, 371.1834). − 6 α 20 petroleum ether acetone step gradient system (8:1, 5:1, 3:1, 2:1, 1:1, 3,11-Dioxoleukamenin E ( ): white, amorphous powder; [ ] D +4 − λ ε 0:1) to give fractions A F. Fraction B was separated on an RP C18 (c 0.1, MeOH); UV (MeOH) max (log ) 232 (3.93) nm; ECD − ff λ Δε − column (from 30:70 to 80:20 MeOH H2O) by MPLC to a ord (MeOH) max ( ) 352 ( 0.1), 313 (+1.2), 224 (+2.4) nm; IR (KBr) − ν subfractions B1 B4. Fraction B3 was subjected to CC on silica gel max 3360, 2985, 2963, 1727, 1704, 1672, 1646, 1457, 1386, 1234, using a gradient system of petroleum ether−acetone (5:1) providing 1198, 1134, 1079, 1042, 968, 777, 654 cm−1; 1H NMR data see Table − 13 three subfractions, B3A B3C. Compounds 7 (6.0 mg, tR 46.2 min) 2; C NMR data see Table 4; positive HRESIMS m/z 369.1665 [M + fi and 17 (12.0 mg, tR 50.6 min) were puri ed from subfraction B3B by Na]+ (calcd for C H O Na, 369.1678). − 20 26 5 HPLC (60:40 MeOH H2O). Separation of fraction C by MPLC RP 11,12-Didehydroleukamenin E (7): colorless needles (MeOH); mp C eluted with a MeOH−H O (20:80−100:0) gradient system − ° α 20 − λ ε 18 2 212 214 C; [ ] D 182 (c 0.2, MeOH); UV (MeOH) max (log ) yielded five main subfractions, C1−C5. Subfraction C2 was chromato- λ Δε − 228 (3.75) nm; ECD (MeOH) max ( ) 244 ( 1.1), 220 (+2.7), 205 graphed on Sephadex LH-20 (MeOH) and then purified by silica gel − ν ( 13.2) nm; IR (KBr) max 3435, 2959, 2869, 1728, 1652, 1460, 1381, − − CC, eluted with CHCl3 acetone (300:1), to yield compounds 3 (25.0 1250, 1180, 1161, 1094, 1027, 984, 804, 671 cm 1; 1H NMR data see mg) and 10 (20.0 mg). Subfraction C3 was subjected to repeated Table 2; 13C NMR data see Table 4; positive HRESIMS m/z 397.1976 chromatography over silica gel (petroleum ether−acetone, from 5:1 to + − ff [M + Na] (calcd for C22H30O5Na, 397.1991). 0:1) to give subfractions C3A C3C. Subfraction C3A a orded 11α,12α-Epoxyleukamenin E (8): colorless needles (MeOH); mp compounds 8 (13.0 mg, tR 32.7 min) and 11 (4.0 mg, tR 45.2 min) − ° α 20 − λ ε − 197 198 C; [ ] D 87 (c 0.4, MeOH); UV (MeOH) max (log ) by HPLC (40:60 MeCN H2O) and 12 (29.0 mg) by silica gel column λ Δε − − − 227 (3.84) nm; ECD (MeOH) max ( ) 335 ( 0.1), 226 ( 2.2), 200 chromatography (CH2Cl2 acetone, 5:1). Compound 14 (3.0 mg) was ν (+0.8) nm; IR (KBr) max 3262, 2944, 2871, 1722, 1650, 1376, 1251, separated from the subfraction C3C by recrystallization from − 1210, 1186, 1114, 1077, 1017, 978, 869, 660, 563 cm 1; 1H NMR data petroleum ether−acetone (3:1). HPLC (60:40 MeOH−H O) was 2 see Table 2; 13C NMR data see Table 4; positive HRESIMS m/z applied to furnish compounds 9 (6.0 mg, tR 46.4 min) and 15 (5.0 mg, + 413.1926 [M + Na] (calcd for C22H30O6Na, 413.1940). tR 61.1 min) from the remainder of subfraction C3C. Fraction D was − 18-Deacetyl-4-epi-henryine A (9): white, amorphous powder; divided into four subfractions (D1 D4) by Sephadex LH-20 α 20 λ ε [ ] D +8 (c 0.1, MeOH); UV (MeOH) max (log ) 228 (3.71) (MeOH). Compounds 6 (7.0 mg) and 13 (3.0 mg) were crystallized λ Δε − nm; ECD (MeOH) max ( ) 351 ( 0.1), 312 (+1.4), 226 (+2.4) nm; from subfraction D2. The remainder of fraction D2 was separated ν IR (KBr) max 3450, 2952, 2853, 1729, 1706, 1648, 1466, 1384, 1254, further by silica gel (petroleum ether−acetone, 3:1), followed by − − 1138, 1094, 1041, 676 cm 1; 1H NMR data see Table 2; 13C NMR HPLC (25:75 MeCN H2O), to obtain compound 5 (9.0 mg, tR 38.9 + min). Separation of subfraction D3 by silica gel column chromatog- data see Table 4; positive HRESIMS m/z 413.1923 [M + Na] (calcd raphy (petroleum ether−acetone, from 6:1 to 4:1) afforded for C22H30O6Na, 413.1940). 7α,14β-Dihydroxy-17β-methoxymethyl-3β-acetoxy-ent-kaur- compounds 1 (49.0 mg) and 2 (10.0 mg). Compound 16 (6.0 mg, 10 − ° t 105.5 min) was obtained by HPLC (40:60 MeOH/H O) from the 11,15-dione ( ): colorless needles (MeOH); mp 165 167 C; R 2 α 20 λ Δε 4 [ ] D +12 (c 1.1, MeOH); ECD (MeOH) max ( ) 306 (+0.7) nm; remainder of subfraction D3. Fraction E gave compound (12.0 mg) ν IR (KBr) max 3381, 2951, 1734, 1704, 1476, 1392, 1313, 1244, 1153, by Sephadex LH-20 (MeOH) and silica gel column chromatography − − 1087, 1058, 969, 776 cm 1; 1H NMR data see Table 2; 13C NMR data (petroleum ether acetone, 3:1). + 11α-Hydroxyleukamenin E (1): colorless needles (MeOH); mp see Table 4; positive HRESIMS m/z 445.2185 [M + Na] (calcd for − ° α 20 − λ ε C H O Na, 445.2202). 195 197 C; [ ] D 36 (c 1.6, MeOH); UV (MeOH) max (log ) 23 34 7 λ Δε − 7α,14β-Dihydroxy-17α-ethoxymethyl-3β-acetoxy-ent-kaur- 234 (3.92) nm; ECD (MeOH) max ( ) 344 ( 0.3), 242 (+1.1), 209 − ν 11,15-dione (11): colorless needles (MeOH); mp 198−200 °C; ( 4.8) nm; IR (KBr) max 3394, 2937, 2873, 1725, 1705, 1647, 1450, − α 20 λ Δε 1375, 1248, 1182, 1069, 979, 843, 713 cm 1; 1H NMR data see Table [ ] D +71 (c 0.1, MeOH); ECD (MeOH) max ( ) 308 (+2.0), 218 − ν 1; 13C NMR data see Table 4; positive HRESIMS m/z 415.2067 [M + ( 0.3) nm; IR (KBr) max 3283, 2968, 2866, 1736, 1703, 1461, 1372, −1 1 13 Na]+ (calcd for C H O Na, 415.2097). 1243, 1103, 1061, 979, 879, 505 cm ; H NMR data see Table 3; C 22 32 6 + 11β-Hydroxyleukamenin E (2): colorless needles (MeOH); mp NMR data see Table 4; positive HRESIMS m/z 459.2324 [M + Na] − ° α 20 − λ ε 201 202 C; [ ] D 48 (c 0.3, MeOH); UV (MeOH) max (log ) (calcd for C24H36O7Na, 459.2359). λ Δε − − 7α,14β-Dihydroxy-17β-ethyoxymethyl-3β-acetoxy-ent-kaur- 236 (3.90) nm; ECD (MeOH) max ( ) 327 ( 0.3), 246 ( 1.0), 203 − ν 11,15-dione (12): white, amorphous powder; [α]20 −2(c 1.0, ( 2.1) nm; IR (KBr) max 3380, 2939, 2863, 1732, 1694, 1654, 1471, D −1 1 λ Δε ν 1380, 1273, 1078, 1029, 993, 965, 594 cm ; H NMR data see Table MeOH); ECD (MeOH) max ( ) 306 (+0.6) nm; IR (KBr) max 13 3377, 2958, 2877, 1733, 1706, 1451, 1376, 1248, 1095, 1025, 982, 886, 1; C NMR data see Table 4; positive HRESIMS m/z 415.2065 [M + − + 1 1 13 Na] (calcd for C22H32O6Na, 415.2097). 844, 776, 608, 510 cm ; H NMR data see Table 3; C NMR data 11-Oxoleukamenin E (3): colorless needles (MeOH); mp 170−172 see Table 4; positive HRESIMS m/z 459.2324 [M + Na]+ (calcd for ° α 20 λ ε C; [ ] D +38 (c 0.2, MeOH); UV (MeOH) max (log ) 230 (3.77) C24H36O7Na, 459.2359). λ Δε − α β α nm; ECD (MeOH) max ( ) 351 ( 0.1), 312 (+1.1), 226 (+2.2) nm; 7 ,14 -Dihydroxy-17 -methoxymethyl-ent-kaur-3,11,15-trione ν (13): − ° α 20 − IR (KBr) max 3370, 2957, 1727, 1706, 1646, 1438, 1380, 1250, 1138, colorless needles (MeOH); mp 180 182 C; [ ] D 5(c 0.1, −1 1 13 λ Δε 1093, 1022, 982, 658 cm ; H NMR data see Table 1; C NMR data MeOH); ECD (MeOH) max ( ) 310 (+2.1), 206 (+0.8) nm; IR + ν see Table 4; positive HRESIMS m/z 413.1923 [M + Na] (calcd for (KBr) max 3227, 2930, 2853, 1732, 1702, 1456, 1381, 1204, 1160, −1 1 13 C22H30O6Na, 413.1940). 1111, 1067, 950, 754, 555 cm ; H NMR data see Table 3; C NMR 3-Oxo-11α-hydroxyleukamenin E (4): colorless needles (MeOH); data see Table 4; positive HRESIMS m/z 401.1913 [M + Na]+ (calcd − ° α 20 − λ mp 256 258 C; [ ] D 184 (c 0.4, MeOH); UV (MeOH) max (log for C21H30O6Na, 401.1940). ε λ Δε − − α β α β ) 234 (3.88) nm; ECD (MeOH) max ( ) 340 ( 0.4), 290 ( 1.8), 7 ,14 -Dihydroxy-17 -methoxymethyl-18 -acetoxy-ent-kaur- − ν 14 − ° 242 (+0.9), 200 ( 4.4), nm; IR (KBr) max 3259, 2966, 2938, 2895, 11,15-dione ( ): colorless needles (MeOH); mp 199 200 C; α 20 λ Δε 2865, 1731, 1704, 1650, 1486, 1460, 1366, 1242, 1130, 1062, 1022, [ ] D +32 (c 0.1, MeOH); ECD (MeOH) max ( ) 308 (+2.7), 220 −1 1 13 − ν 948 cm ; H NMR data see Table 1; C NMR data see Table 4; ( 0.3), 204 (+0.6) nm; IR (KBr) max 3464, 3396, 2952, 2934, 2898, positive HRESIMS m/z 371.1822 [M + Na]+ (calcd for C H O Na, 1740, 1707, 1468, 1440, 1382, 1229, 1116, 1075, 1041, 965, 905, 755, 20 28 5 − 371.1834). 585 cm 1; 1H NMR data see Table 3; 13C NMR data see Table 4; 5 + 3-Deacetyl-11-oxoleukamenin E ( ): white, amorphous powder; positive HRESIMS m/z 445.2169 [M + Na] (calcd for C23H34O7Na, α 20 λ ε [ ] D +22 (c 0.3, MeOH); UV (MeOH) max (log ) 227 (3.50) nm; 445.2202).

2260 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article α β β β fl fi σ 7 ,14 -Dihydroxy-17 -methoxymethyl-18 -acetoxy-ent-kaur- re ections [R(int) = 0.0205]; nal R indices [I >2 (I)] R1 = 0.0331, 15 α 20 − 11,15-dione ( ): white, amorphous powder; [ ] D 13 (c 0.1, wR = 0.1081; R indices (all data) R = 0.0348, wR = 0.1177; Flack λ Δε ν 2 1 2 MeOH); ECD (MeOH) max ( ) 308 (+0.6) nm; IR (KBr) max parameter 0.1(3). 3406, 2938, 2910, 2852, 1744, 1721, 1467, 1389, 1367, 1229, 1145, Crystallographic data of 11: C H O , formula weight 436.53, − 24 36 7 1102, 1038, 963, 782, 526 cm 1; 1H NMR data see Table 3; 13C NMR crystal size 0.26 × 0.20 × 0.20 mm3, crystal system orthorhombic + data see Table 4; positive HRESIMS m/z 445.2180 [M + Na] (calcd space group P21, a = 12.0609(2) Å, b = 6.82630(10) Å, c = 29.7481(5) α γ ° β ° 3 for C23H34O7Na, 445.2202). Å, = =90, = 98.0820(10) , V = 2424.87(7) Å , Z =4,Dc = Single-Crystal X-ray Diffraction Analysis and Crystallo- 1.196 Mg/m3, μ(Cu Kα) = 0.733 mm−1, F(000) = 944, absorption graphic Data of Compounds 1−4, 8, 10, and 11. Diffraction coefficient 0.711 mm−1; total 18 261 reflections, 6586 independent intensity data for compounds 1−4, 8, 10, and 11 were acquired on a reflections [R = 0.0230]; final R indices [I >2σ (I)] R = 0.0370, ff (int) 1 Bruker APEX-II di ractometer employing graphite-monochromatized wR2 = 0.1032; R indices (all data) R1 = 0.0373, wR2 = 0.1037; Flack Cu Kα radiation (λ = 1.54178 Å) at 298(2) K. Data were collected by parameter −0.03(15). Bruker APEX2 software. Data reduction was conducted with Bruker Crystallographic data for the structures of compounds 1−4, 8, 10, SAINT. Structure solution and refinement were performed with the and 11 reported in this paper have been deposited with the Cambridge SHELXTL program package. All non-hydrogen atoms were refined Crystallographic Data Centre (deposit numbers CCDC 915525− anisotropically. The hydrogen atom positions were geometrically 915531, respectively). Copies of these data can be obtained, free of idealized and allowed to ride on their parent atoms. The crystal charge, on application to the Director, CCDC, 12 Union Road, structures of compounds 1−4, 8, 10, and 11 were drawn by ORTEP 3 Cambridge CB2 1EZ, UK (fax: +44-(0)1223-336033 or e-mail: for windows (version 2.02) and are shown in Figures 2−8, [email protected]). respectively. Cytotoxicity Assay. The cytotoxicity assay was performed 1 · Crystallographic data of : C22H32O6 2H2O, formula weight according to the MTS method in 96-well microplates, as reported 428.51, crystal size 0.15 × 0.12 × 0.11 mm3, crystal system monoclinic previously.13a,19 The assay is based on reduction of the tetrazolium salt space group P2(1), a = 8.5553(2) Å, b = 7.0580(2) Å, c = 18.5846(5) 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfo- α γ ° β ° 3 Å, = =90, = 94.6680(10) , V = 1118.48(5) Å , Z =2,Dc = pheny)-2H-tetrazolium (MTS) in metabolically active cells to produce − 1.272 Mg/m3, μ(Cu Kα) = 0.733 mm 1, F(000) = 464, absorption the corresponding formazan product, accomplished by NADPH- or − coefficient 0.793 mm 1; total 12 021 reflections, 3535 independent NADH-dependent dehydrogenase enzymes. After incubation with test fl fi σ re ections [R(int) = 0.0210]; nal R indices [I >2 (I)] R1 = 0.0310, substances and controls, this is a rapid colorimetric assay to measure wR2 = 0.0872; R indices (all data) R1 = 0.0312, wR2 = 0.0875; Flack cellular proliferation and cytotoxicity. Five human cancer cell lines, parameter 0.19(14). HL-60 human myeloid leukemia, SMMC-7721 human hepatocellular 2 · · Crystallographic data of : C22H32O6 CH3OH H2O, formula carcinoma, A-549 lung cancer, MCF-7 breast cancer, and SW480 weight 442.53, crystal size 0.12 × 0.12 × 0.11 mm3, crystal system human colon cancer, together with one noncancerous human orthorhombic, space group P2(1)2(1)2(1), a = 8.4307(2) Å, b = pulmonary epithelial cell line, BEAS-2B, were assayed. Cells were 11.7779(3) Å, c = 23.1333(5) Å, α = β = γ =90°, V = 2297.04(9) Å3, cultured in RPMI-1640 or in DMEM medium (Hyclone, Logan, UT, 3 μ α −1 Z =4,Dc = 1.280 Mg/m , (Cu K ) = 0.733 mm , F(000) = 960, USA) and supplemented with 10% fetal bovine serum (Hyclone, −1 ° μ absorption coefficient 0.788 mm ; total 11 826 reflections, 3364 Logan, UT, USA) in 5% CO2 at 37 C. A volume of 100 Lof fl fi σ independent re ections [R(int) = 0.0260]; nal R indices [I >2 (I)] R1 adherent cells was seeded into each well of the 96-well culture plates × 5 = 0.0291, wR2 = 0.0793; R indices (all data) R1 = 0.0294, wR2 = with initial density of 1 10 cells/mL, and cells were allowed to 0.0795; Flack parameter −0.04(16). adhere for 12 h before addition of test compounds. Each cancer cell 3 · Crystallographic data of : C22H30O6 H2O, formula weight 408.48, line was exposed to test compounds at concentrations of 0.064, 0.32, crystal size 0.32 × 0.20 × 0.20 mm3, crystal system orthorhombic 1.6, 8, and 40 μM in triplicate for 48 h at 37 °C in 200 μL of media. space group P2(1)2(1)2(1), a = 7.57840(10) Å, b = 10.6565(2) Å, c = Wells with DDP (cis-platin, Sigma, St. Louis, MO, USA) and paclitaxel α β γ ° 3 26.4145(5) Å, = = =90, V = 2133.21(6) Å , Z =4,Dc = 1.272 (Sigma, St. Louis, MO, USA) were used as positive controls. Then, Mg/m3, μ(Cu Kα) = 0.733 mm−1, F(000) = 880, absorption 100 μL of media and 20 μL of MTS (Sigma, St. Louis, MO, USA) coefficient 0.774 mm−1; total 13 794 reflections, 3390 independent were added to each well and cultured for 4 h. After compound fl fi σ λ re ections [R(int) = 0.0296]; nal R indices [I >2 (I)] R1 = 0.0335, treatment, cell viability was detected by a Bio-Rad 680 at = 595 nm, wR = 0.0923; R indices (all data) R = 0.0340, wR = 0.0928; Flack and a cell growth curve was graphed. IC50 values were calculated by 2 1 2 22 parameter −0.1(2). Reed and Muench’s method. 4 Crystallographic data of : C20H28O5, formula weight 348.42, crystal size 0.15 × 0.12 × 0.10 mm3, crystal system orthorhombic ■ ASSOCIATED CONTENT α space group P2(1)2(1)2(1), a = 6.506 Å, b = 13.700 Å, c = 20.624 Å, *S β γ ° 3 3 μ α Supporting Information = = =90, V = 1838.3 Å , Z =4,Dc = 1.259 Mg/m , (Cu K )= 0.733 mm−1, F(000) = 752, absorption coefficient 0.089 mm−1; total (+)-HRESIMS, UV, IR, ECD, and 1D and 2D NMR spectra for fl fl fi compounds 1−15 and X-ray crystallographic data (CIF file) for 8460 re ections, 1821 independent re ections [R(int) = 0.1119]; nal R σ compounds 1−4, 8, 10, and 11. This material is available free of indices [I >2 (I)] R1 = 0.0388, wR2 = 0.1037; R indices (all data) R1 = 0.0413, wR2 = 0.1043; Flack parameter 0(10). charge via the Internet at http://pubs.acs.org. 8 · Crystallographic data of : C22H30O6 H2O, formula weight 408.48, crystal size 0.23 × 0.10 × 0.10 mm3, crystal system orthorhombic ■ AUTHOR INFORMATION space group P2(1)2(1)2(1), a = 7.5766(2) Å, b = 10.6290(2) Å, c = Corresponding Authors α β γ ° 3 26.1402(5) Å, = = =90, V = 2105.11(8) Å , Z =4,Dc = 1.289 − *(G.Y.) Tel: 86-27-83692311. Fax: 86-27-83692762. E-mail: Mg/m3, μ(Cu Kα) = 0.733 mm 1, F(000) = 880, absorption coefficient 0.784 mm−1; total 11 433 reflections, 3047 independent [email protected]. fl fi σ *(Y.Z.) E-mail: [email protected]. re ections [R(int) = 0.0215]; nal R indices [I >2 (I)] R1 = 0.0336, wR2 = 0.0944; R indices (all data) R1 = 0.0339, wR2 = 0.0949; Flack Notes parameter 0.03(19). The authors declare no competing financial interest. 10 Crystallographic data of : C23H34O7, formula weight 422.50, 3 crystal size 0.20 × 0.10 × 0.10 mm , crystal system orthorhombic ■ ACKNOWLEDGMENTS space group P2(1)2(1)2(1), a = 6.71180(10) Å, b = 13.4388(2) Å, c = α β γ ° 3 fi 23.6639(4) Å, = = =90, V = 2134.45(6) Å , Z =4,Dc = 1.315 We thank Professor F. D. Horgen at Hawaii Paci c University Mg/m3, μ(Cu Kα) = 0.733 mm−1, F(000) = 912, absorption for editing the manuscript. We are grateful to Dr. J.-P. Wang at coefficient 0.790 mm−1; total 7137 reflections, 1998 independent Huazhong University of Science and Technology for the

2261 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262 Journal of Natural Products Article authentication of the plant material, and Dr. X.-G. Meng at (21) Zhao, Y.; Pu, J. X.; Huang, S. X.; Ding, L. S.; Wu, Y. L.; Li, X.; Central China Normal University for single-crystal X-ray Yang, L. B.; Xiao, W. L.; Chen, G. Q.; Sun, H. D. J. Nat. Prod. 2009, ff fi 72, 988−993. di raction analysis. This project was nancially supported by − the Hubei Key Laboratory Foundation of Natural Medicinal (22) Reed, L. J.; Muench, H. Am. J. Hyg. 1938, 27, 493 497. Chemistry and Resource Evaluation, Huazhong University of Science and Technology (2010-3, to G.Y.), the Fundamental Research Funds for the Central Universities (HUST: 2012QN003, to G.Y.), Scientific Research Foundation for the Returned Oversea Chinese Scholars, State Education Ministry of China (2010-1561, 40th, to G.Y.), Program for Youth Chutian Scholar of Hubei Province of China (to G.Y.), and Program for New Century Excellent Talents in University, State Education Ministry of China (NCET-2008-0224, to Y.Z).

■ REFERENCES (1) Li, M. H.; Chen, J. M.; Peng, Y.; Xiao, P. G. Mode. Tradit. Chin. Med. Mater. Med. 2008, 10,46−52. (2) Lu, Y.; Foo, Y. L. Phytochemistry 2002, 59, 117−140. (3) Wu, Y. B.; Ni, Z. Y.; Shi, Q. W.; Dong, M.; Kiyota, H.; Gu, Y. C.; Cong, B. Chem. Rev. 2012, 112, 5967−6026. (4) Pan, Z. H.; Xu, G.; Zhao, Q. S. Guangxi Zhiwu 2010, 30, 781− 790. (5) Kabouche, A.; Kabouche, Z. Stud. Nat. Prod. Chem. 2008, 35, 753−833. (6) Öztekin, N.; Baskan,̧ S.; Evrim Kepekci,̧ S.; Erim, F. B.; Topcu,̧ G. J. Pharmaceut. Biomed. Anal. 2010, 51, 439−442. (7) Zhao, W.; Pu, J. X.; Du, X.; Su, J.; Li, X. N.; Yang, J. H.; Xue, Y. B.; Li, Y.; Xiao, W. L; Sun, H. D. J. Nat. Prod. 2011, 74, 1213−1220. (8) Ahmed, B.; AI-Howiriny, T. A.; Al-Rehaily, A. J.; Mossa, J. S. Z. Naturforsch. C 2004, 59,9−14. (9) Li, X. W.; Ian, C. H. Lamiaceae. In Flora of China; Wu, Z. Y., Raven, P. H., Hong, D. Y., Eds.; Science Press: Beijing, China, and Missouri Botanical Garden Press: St. Louis, MO, USA, 1994; Vol. 17, pp 50−299. (10) Fang, Z. X., Liao, C. L. Medicinal Flora of Enshi, Hubei Province; Hubei Science and Technology Press: Wuhan, 2006. pp 288−289. (11) Wang, H. Y.; Hu, D. Y.; Xue, W.; Yang, S.; Song, B. A. Nat. Prod. Res. Dev. 2011, 23,63−65. (12) (a) Zhao, L. M.; Liang, X. T.; Li, L. N. J. Chin. Pharm. Sci. 1997, 6, 111−112. (b) Zhang, H. J.; Li, L. N. Planta Med. 1994, 60,70−72. (c) Li, L. N. Chin. J. Org. Chem. 1993, 13, 303−304. (13) (a) Luo, Z. W.; Wang, F. Q.; Zhang, J. W.; Li, X. Y.; Zhang, M. K.; Hao, X. C.; Xue, Y. B.; Li, Y.; Horgen, F. D.; Yao, G. M.; Zhang, Y. H. J. Nat. Prod. 2012, 75, 2113−2120. (b) Shu, P. H.; Wei, X. L.; Xue, Y. B.; Li, W. J.; Zhang, J. W.; Xiang, M.; Zhang, M. K.; Luo, Z. W.; Li, Y.; Yao, G. M.; Zhang, Y. H. J. Nat. Prod. 2013, 76, 1303−1312. (c) Guo, J. R.; Zhang, J. W.; Shu, P. H.; Kong, L. M.; Hao, X. C.; Xue, Y. B.; Luo, Z. W.; Li, G.; Li, Y.; Yao, G. M.; Zhang, Y. H. Molecules 2012, 17, 6424−6433. (14) Zhou, B. N.; Chen, J.-B.; Wang, C. Y.; Blasko, G.; Cordell, G. A. Phytochemisty 1989, 28, 3536−3538. (15) Takeda, Y.; Fujita, T.; Ueno, A. Chem. Lett. 1981, 10, 1229− 1232. (16) Sun, H. D.; Lin, Z. W.; Shen, X. Y.; Takeda, Y.; Fujita, T. Phytochemistry 1991, 30, 603−606. (17) (a) Hong, S. S.; Lee, S. A.; Han, X. H.; Jin, H. Z.; Lee, J. H.; Lee, D.; Lee, J. J.; Hong, J. T.; Kim, Y.; Ro, J. S.; Hwang, B. Y. J. Nat. Prod. 2007, 70, 632−636. (b) Li, L. M.; Weng, Z. Y.; Huang, S. X.; Pu, J. X.; Li, S. H.; Huang, H.; Yang, B. B.; Han, Y.; Xiao, W. L.; Li, M. L.; Han, Q. B.; Sun, H. D. J. Nat. Prod. 2007, 70, 1295−1301. (18) (a) Flack, H. D. Acta Crystallogr. 1983, A39, 876−881. (b) Flack, H. D.; Bernardinelli, G. Chirality 2008, 20, 681−690. (19) Cory, A. H.; Owen, T. C.; Barltrop, J. A.; Cory, J. G. Cancer Commun. 1991, 3, 207−212. (20) Zhao, Y.; Pu, J. X.; Huang, S. X.; Wu, Y. L.; Yang, L. B.; Xiao, W. L.; Han, Q. B.; Chen, G. Q.; Sun, H. D. J. Nat. Prod. 2009, 72, 125− 129.

2262 dx.doi.org/10.1021/np400600c | J. Nat. Prod. 2013, 76, 2253−2262