Article pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. 2018, 140, 2485−2492
Structural Elucidation and Bioinspired Total Syntheses of Ascorbylated Diterpenoid Hongkonoids A−D Jin-Xin Zhao,†,§ Yan-Yan Yu,†,‡,§ Sha-Sha Wang,† Su-Ling Huang,† Yu Shen,† Xin-Hua Gao,†,‡ Li Sheng,† Jing-Ya Li,† Ying Leng,† Jia Li,† and Jian-Min Yue*,† †
State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zuchongzhi Road, Shanghai 201203, People’s Republic of China ‡ University of Chinese Academy of Sciences, No. 19A Yuquan Road, Beijing 100049, People’s Republic of China S Supporting Information *
ABSTRACT: Hongkonoids A−D (1−4), the first example of ascorbylated terpenoids featuring a unique 5,5,5-fused tricyclic spiroketal butyrolactone moiety and diterpenoid-derived long chain, were isolated from Dysoxylum hongkongense. Their structures were unambiguously assigned by a combination of spectroscopic data, chemical degradation, Xray crystallography, CD analysis, and total synthesis. The total syntheses of compounds 1−4 were effectively accomplished by a convergent strategy with the longest linear sequences of 12−14 steps and overall yields of 5.4− 9.6%. Notably, we exploited a bioinspired one-pot method to construct the key intermediate 14 from an easily made compound 12 by involving the cascade reactions of an elaborate Claisen rearrangement, deprotections, and a 5-exo-trig cyclization. The desired major epimer 14a was then transformed to the main building block 21. Assembly of 21 and the long chain vinyl iodide 7 was made by an NHK coupling reaction to furnish the framework of 1−4. Some of the hongkonoids and/or synthetic analogs showed significant to moderate inhibitory activities against NF-κB, 11β-HSD1, and sterol synthesis. The most active NF-κB inhibitor 34 exhibited distinct inhibition on the LPS-induced inflammatory responses in RAW 246.7 and primary BMDM cells.
■
INTRODUCTION Natural products from plants are a very important source for drug discovery.1 The plant Dysoxylum hongkongense has been used in Chinese herbal medicine2 and is mainly distributed in South China.3 Only a few triterpenoids have been identified from it previously.2 In continuing our studies on Dysoxylum plants for structurally interesting and biologically important compounds,4 four ascorbylated phytane-type diterpenoids, named hongkonoids A−D (1−4) featuring a unique tricyclic 5,5,5-fused spiroketal butyrolactone moiety and diterpenoidderived long chain (Figure 1), were isolated and characterized from D. hongkongense. Hongkonoids A−D are the first example of ascorbylated terpenoids. Although ascorbic acid (vitamin C) is ubiquitous in plants, only a limited number of ascorbylated natural products, mainly ascorbylated phenolics, have been identified hitherto.5 Some of them exhibited a wide range of important biological activities, such as modulation of cytochrome P4501A1 activity, anticancer, and prolyl endopeptidase and metalloproteinase inhibitions.5a Furthermore, we initiated the total syntheses of compounds 1−4 by a convergent strategy from commercially available materials. A bioinspired one-pot method involving several important chemical reactions to furnish the key intermediate 14 was developed, in which the transition states of the rearrangement and the stereoselectivity of the ring formation together © 2018 American Chemical Society
Figure 1. Structures of compounds 1−4.
controlled the cascade reactions to yield the desired product 14a as the major epimer. This synthetic work not only confirmed the structures of 1−4 but also provided a sufficient quantity of samples for further biological study. The inhibitory activities of compounds 1−4 and the synthetic analogs against NF-κB, 11β-HSD1, and sterol synthesis were evaluated. Received: September 26, 2017 Published: February 2, 2018 2485
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society
■
to ozonolysis6 (Scheme 1) to obtain compounds 5 and 6. Compound 5 was identified by NMR and MS data as one of
RESULTS AND DISCUSSION Structural Elucidation and Biosynthetic Consideration for Compounds 1−4. Hongkonoid A (1), a colorless gum, possessed the molecular formula C29H44O10 as assigned by the 13 C NMR data and (−)-HRESIMS ion peak at m/z 597.2917 [M + HCO2]− (calcd 597.2911), requiring 8 double bond equivalents (DBEs). The IR spectrum showed absorptions for hydroxy (3462 cm−1) and carbonyl (1790, 1736 cm−1) groups. Analysis of the NMR data [Table S1, Supporting Information (SI)] with the aid of DEPT and HSQC experiments revealed the presence of an acetyl group, two trisubstituted double bonds, two ester carbonyls (δC 173.7 and 174.3), six methyls, seven methylenes, five methines, one ketal carbon (δC 115.8), and two oxygenated tertiary carbons. The two double bonds and three ester carbonyls accounted for five out of eight DBEs, and the remaining three thus required the presence of three rings in 1. Comprehensive analysis of its 2D NMR spectra allowed the construction of the planar structure for 1 (Figure S4, SI). Four proton-bearing structural fragments a−d as depicted in bold bonds were readily established by interpretation of the HSQC and 1H−1H COSY spectra. In the HMBC spectrum, the Δ6 double bond was located by the correlation networks of H-6/C8, H-19/C-6 and C-7, and H-8/C-7, which also attached C-8 and Me-19 to C-7. Similarly, the Δ10 double bond was placed, and C-12 and Me-18 were connected to C-11 by the correlations of H-10/C-12, H-18/C-10, and H2-13/C-11. The key HMBC correlations of H2-2 and OCH3/C-1 fixed the methoxycarbonyl group at C-2. The HMBC correlations of H16(17)/C-14 and H2-13/C-15 readily attached Me-16, Me-17, and C-14 to C-15. The key HMBC correlation from H-12 to the acetyl carbonyl carbon placed the acetoxy group at C-12. The above assignment clearly delineated the partial structure of a phytane-type diterpenoid for 1. Two broad singlet proton resonances showing no correlations with any carbons in the HSQC spectrum were assigned to OH-2′ (δH 3.27, br s) and OH-5′ (δH 3.35, br s) by the HMBCs of OH-2′/C-2′ and OH5′/C-5′, respectively. The multiple HMBCs of OH-2′/C-1′, C2′, and C-3′; H-4′/C-1′ and C-3′; H-5′/C-3′; and H2-6′/C-3′ and C-4′ suggested the presence of an ascorbyl moiety. The key HMBC correlation from OH-2′ to C-14 (δC 51.9) connected the ascorbyl moiety and the diterpenoid part via the C-14 and C-2′ bond. A five-membered lactone ring was furnished by the chemical shift of H-4′ (δH 4.59) and the HMBC correlation of H-4′/C-1′. A ketal group at C-3′ was assigned by the chemical shifts of C-15 (δC 89.9), C-3′ (δC 115.8), and C-6′ (δC 74.1) and the HMBC correlation of H-6′/C-3′ to satisfy the remaining two DBEs requirement, which finally furnished the planar structure for 1. In the ROESY spectrum (Figures S4 and S13, SI), the crosspeaks of H-6/H-8 and H-10/H-12 revealed that both Δ6 and Δ10 double bonds had an E-geometry. The ROESY correlations of H-4′/H-14, H-4′/H-16, and H-14/H-16 indicated that H-4′, H-14, and Me-16 were cofacial and arbitrarily assigned in a βorientation. The Me-17 and OH-2′ were subsequently assigned as α-oriented by the ROESY correlations between H-17 and OH-2′. The small coupling constant between H-4′ and H-5′ (J = 1.7 Hz) suggested that H-5′ was α-oriented. The assignment of relative configurations at C-3, C-12, and C-3′ was challenging due to the freely rotating long chain and/or the absence of reliable ROESY correlations. After the failed recrystallization of 1 in many solvent systems, it was subjected
Scheme 1. Ozonolysis of Compound 1a
Reagents and conditions: (a) O3, CH2Cl2/MeOH (20:1), − 78 °C, 0.5 min; (b) Me2S (excess), rt, 12 h.
a
the ozonolysis products of 1. An X-ray crystallographic study using Cu Kα radiation revealed the absolute configuration of 5 as depicted [the absolute structure parameter: −0.01(12)]7 (Scheme 1). Compound 6 was assigned as methyl (R)-(+)-6oxo-3-methylhexanoate8 by the 1H NMR data (Figure S46, SI) and specific rotation ([α]25 D +10.6). Taken together, the structure of 1 (2′S, 3′R, 4′R, 5′S, 3R, 12S, and 14R) was thus completely determined. Hongkonoid B (2) had the molecular formula C27H42O9 based on the 13C NMR data and (+)-HRESIMS ion peak at m/ z 533.2723 [M + Na]+ (calcd 533.2727), which is 42 mass units less than that of 1. The NMR data (Table S2, SI) of 2 highly resembled those of 1 except for the shielded H-12 (ΔδH − 0.86) and the concomitant absence of the proton and carbon resonances for the acetyl group, indicating that 2 was the deacetylated derivative of 1. As the result, the C-11 (ΔδC 6.2) and C-13 (ΔδC 3.6) of 2 were deshielded as compared with those of 1 due to the lack of γ-gauche effects from the acetyl group. This deduction was verified by 2D NMR data (Figure S5, SI). The absolute configuration of 2 was assigned as shown by the compatible CD curves with those of 1 (Figure 2).
Figure 2. CD spectra of compounds 1−4.
Hongkonoid C (3) possessed the molecular formula C27H40O9 as assigned by the (+)-HRESIMS ion peak at m/z 531.2562 [M + Na]+ (calcd 531.2570). Its 1H and 13C NMR spectra (Table S3, SI) showed high similarities to those of 2 with the major differences being the presence of a keto carbonyl (δC 201.3) and the deshielded H2-13 (ΔδH 1.22 and 1.26) and shielded C-10 (ΔδC 19.0) resonances, suggesting that a keto group was located at the C-12 of 3. This assignment was supported by 2D NMR data (Figure S6, SI). 2486
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society
biosynthetic pathway for 1−4 was proposed as shown in Scheme 2. Oxidation of vitamin C by ROS would produce a
Hongkonoid D (4), a colorless gum, showed the molecular formula C26H38O9 as deduced from the (−)-HRESIMS ion peak at m/z 493.2442 [M − H]− (calcd 493.2438), which is 14 mass units less than that of 3. The NMR data (Table S4, SI) of 4 highly resembled those of 3 except for the absence of proton and carbon signals for the methyl group, indicating that it was the hydrolysis product of 3. This conclusion was corroborated by 2D NMR data (Figure S7, SI). Compounds 3 and 4 are structurally closely related to 2 and were presumably biosynthesized from the latter by simple oxidation and/or methylation. Their absolute configurations were thus tentatively assigned as depicted on biosynthetic considerations. The CD curves of 3 and 4 were well matched, but obviously different from those of 1 and 2 (Figure 2), which was likely caused by the presence of the chromophore of the α,β-unsaturated keto system. It was interesting that many 13C NMR signals for compounds 2−4 acquired in CDCl3 (Tables S2−S4, SI) were observed as apparent doublets. However, when the 13C NMR spectra of 2− 4 were recorded in CD3OD, all the doublet signals became singlets, clearly indicating that they were pure compounds. These observations suggested the formation of two different intramolecular H-bonds (Figure 3) for compounds 2−4 in
Scheme 2. Plausible Biosynthetic Pathway for Compounds 1−4
free radical i, which would be captured by GGPP to give an allyl ether ii. Subsequent Claisen rearrangement of ii would provide the intermediate iii, which would then undergo a spontaneous reaction cascade of a ketal formation and an intramolecular 5exo-trig cyclization to yield the key intermediate iv of a 5,5,5fused tricyclic spiroketal lactone derivative. Finally, a series of enzyme-involved modifications on iv including dephosphorylation, redox process, and methylation would afford compounds 1−4. Total Syntheses of Compounds 1−4. Our retrosynthetic analysis is illustrated in Scheme 3. Strategically, we envisioned that hongkonoids A−D (1−4) could be accessible from the intermediate A, which could be readily made by assembly of the aldehyde B and vinyl iodine 7 via Nozaki−Hiyama−Kishi (NHK) reaction. It was envisaged that B could be generated from C by simple protection and deprotection of the hydroxy groups, followed by oxidation of the primary alcohol. Construction of the unique 5,5,5-fused tricyclic spiroketal butyrolactone block C is the most crucial and challenging issue for our synthetic strategy owing to the highly oxygenated and congested carbon skeleton with multiple chiral centers. Inspired by our biosynthetic proposal for this compound class (Scheme 2), we envisioned an acid-catalyzed cascade reaction sequence involving deprotection of D and formation of the tetrahydrofuran ring of the hemiketal intermediate, which then transformed to the key building block C by an intramolecular 5-exo-trig cyclization. The gulono-γ-lactone derivative D could be made by a thermal Claisen rearrangement of the corresponding lactone E, which could be traced back to the readily available materials of 5,6-O-isopropylidene-L-ascorbic acid 8 and allyl bromide derivative F. The vinyl iodine 7 could be derived from the known compound (R)-3,7-dimethyloct-6en-1-yl benzoate (9) in several steps. According to our retrosynthetic analysis, the asymmetric synthesis of the key building block B (21) commenced with two known chemical entities 8 and 10 (Schemes 4 and 5). Compound 8 was prepared by isopropylidene protection of Lascorbic acid,16 and allyl bromide 10 was made from 3benzyloxy-1-propanol in four steps according to a reported procedure.17 Reaction of 8 with 10 at the presence of Na2CO3
Figure 3. H-bonds for compounds 2 (A), 3 (B), and 4 (B).
CDCl3, which were demolished in the polar solvent CD3OD. For compound 2 (Figure 3A), two preferred intramolecular Hbonds were likely formed between the lactone carbonyl group and two hydroxy groups furnishing eight- (OH-12) and fivemembered (OH-2′) rings, respectively. Similarly, compounds 3 and 4 favored to form two intramolecular H-bonds (Figure 3B) of OH-2′/keto and OH-2′/lactone, which shaped seven- and five-membered rings, respectively. For compound 1, only one intramolecular H-bond was likely formed between OH-2′ and lactone due to the acetylation of OH-12. The production of compounds 1−4 in plants may represent a “one stone two birds” strategy of biosynthetic, catabolic, and conjugation pathways that together control the homeostasis of the pools of antioxidants and other geranylgeranyl pyrophosphate (GGPP) associated physiologically active substances.9 Vitamin C is one of the major antioxidant components in plants and provides protection for plant cells from the injury of unnecessary and excessive reactive oxygen species (ROS) to reduce oxidative stress caused by environmental factors, such as pollution, UV light, drought, heat and cold, and pathogen infection.10 Phytane-type diterpenoids are biosynthesized from GGPP via simple redox processes. GGPP is the biosynthetic precursor of many physiologically active substances, such as chlorophyll,11 vitamins E12 and K1,13 and in particular plant hormones gibberellins14 and abscisic acid.15 A plausible 2487
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society Scheme 3. Retrosynthetic Analysis
Scheme 4. Syntheses of the Key Intermediate 14a/ba
Scheme 5. Synthesis of Compound 21a
a Reagents and conditions: (a) 8 (1.1 equiv), Na2CO3 (1.2 equiv), rt, 30 min, DMSO/THF (9:8), then 10 (1.0 equiv), rt, 4 h; (b) EOMCl (1.2 equiv), NEt3 (1.3 equiv), DMAP (0.075 equiv), THF, rt, 12 h; (c) toluene, reflux, 5 h; (d) TsOH·H2O (4.0 equiv), THF, reflux, 8 h, then CH2Cl2, reflux, 12 h.
a
Reagents and conditions: (a) TESCl (4.0 equiv), imidazole (5.0 equiv), DMF, rt, 16 h; (b) Pd/C (10 wt %), H2, MeOH, 60 °C, 7 h; (c) DMP (4.0 equiv), CH2Cl2, rt, 2 h; (d) Ac2O (5.0 equiv), DMAP (1.2 equiv), pyridine, 110 °C, 4 h; (e) KHCO3 (2.0 equiv), MeOH, rt, 2.5 h; (f) TESCl (1.4 equiv), imidazole (1.5 equiv), DMF, rt, 7 h; (g) Pd/C (10 wt %), H2, MeOH, 60 °C, 7 h; (h) DMP (2.0 equiv), CH2Cl2, rt, 1.5 h.
instead of the conventionally used K2CO318 gave 3′-O-allyl product 11 (according to the biosynthetic numbering for hongkonoids), which was unstable during the workup and purification procedures due likely to the exposure of OH-2′ and was thereby immediately protected with chloromethyl ethyl ether (EOMCl) after a quick workup to afford 12 in overall 41% isolated yield. At the beginning we tried the direct thermal Claisen rearrangement of the reaction product 11, but only got a complex mixture. Subsequent thermal Claisen rearrangement18a,c of 12 produced C-2′ alkylated ascorbic acid 13 as a mixture of three diastereomers (d.r. ≈ 2.5:1:1 as determined by 1 H NMR). After removal of solvent toluene, this was straightforwardly subjected to an elaborate acid-catalyzed cascade sequence of reactions involving the deprotections of isopropylidene and EOM groups, followed by a spontaneous
hemiketal formation and an intramolecular 5-exo-trig cyclization.19 This afforded the desired 5,5,5-fused tricyclic spiroketal lactone 14a and its epimer 14b (d.r. ≈ 4:1) as determined by the NOESY spectra (Figures S8, S55, and S58; SI) in appreciable yield of 43% over two steps. The diastereomeric ratios of 13 and 14 were inconsistent and attracted our great interest in seeking the inherent mechanism. Four plausible chairlike transition states a−d for the Claisen rearrangement were proposed as depicted in Figure 4, in which the transition state a was preferred to b due to the equatorial position of the bulky group R1 in the state a, while for the transition states c and d, the latter was forbidden owing to the mutual repulsion 2488
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society
chemistry, which allowed resolution of the epimers. Debenzylation of 16 with H2 catalyzed by palladium on activated charcoal provided alcohol 17, which after oxidation with Dess− Martin periodinane (DMP)21 gave hemiacetal 18 in excellent yield (91%). However, our attempts to implement coupling reaction of 18 with vinyl iodide 7 and its analogs failed under the general NHK reaction conditions or even under more acute condition (n-butyllithium),22 suggesting that the protection of OH-2′ was necessary. To our delight, after the failure of trimethylsilylation,23 acetylation of 16 on heating generated the desired product 20 and an unanticipated diacetyl protected product 19, and the latter was then transformed to 20 by selective deacetylation with KHCO3, followed by TES protection. Debenzylation of 20 followed by Dess−Martin oxidation afforded the desired aldehyde 21 in high yield. We then moved forward to synthesize the long chain vinyl iodide 7 (Scheme 6). (R)-3,7-Dimethyloct-6-en-1-yl benzoate
Figure 4. Transition states of the Claisen rearrangement for 12.
of two spatially close bulky groups R1 and R2. The above analysis clearly demonstrated that only three diastereomers 13a (major), 13b, and 13c were formed via the transition states a− c, respectively. The Newman projections for the intermediates 13a−c are illustrated in Figure 5, from which compound 13c
Scheme 6. Synthesis of Compound 7a
Figure 5. Newman projections along the C14/C2′ bond for the three diastereomers of 13a−c (the configurations for the C14 and C2′ were determined by correlating with 14 and the reaction mechanisms).
was readily distinguished from 13a and 13b by the downfield shifted H-14 (δH 3.18) due to the deshielding effects of the gauche-positioned keto and lactone groups. The chemical shifts of H-14 for both 13a (δH 2.89) and 13b (δH 2.82) were very close and were differentiated from each other by correlating with the products 14a and 14b according to the reaction mechanism. The 5-exo-trig cyclization of the deprotected 13c was not allowed, because the Δ15 double bond and the hydroxy group of the hemiketal intermediate were not cofacial and far away. Therefore, only two epimers 14a and 14b were obtained from the one-pot reaction, which well matched the changes of diastereomeric ratios from 13 to 14. The strategy for construction of the 5,5,5-fused tricyclic spiroketal butyrolactone skeleton in 14 was initially simulated by an exploratory model experiment (see SI) involving 5-exo-trig cyclization of 2methallyl-3-ketohexulosonic acid lactone.20 We found that the cyclization proceeded smoothly in CH2Cl2, whereas only trace amounts of the product were detected in THF, and no product was obtained in other conventional solvents such as DMF, DMSO, methanol, and/or acetonitrile. In the one-pot synthesis of 14 from 12, the deprotections were carried out in THF due to the poor solubility of toluene-4-sulfonic acid monohydrate (TsOH·H2O) in CH2Cl2, while the cyclization was performed in CH2Cl2 as guided by the model experiment. Next, the mixture of 14a/b underwent TES protection to afford diprotected 15 and monoprotected 16 in pleasing yields (Scheme 5), and the latter possessed the desired stereo-
a
Reagents and conditions: (a) SeO2 (0.5 equiv), TBHP (70% in H2O) (2.0 equiv), CH2Cl2, 10 °C, 5 h; (b) NaBH4 (1.2 equiv), MeOH, 0 °C to rt, 1 h; (c) PBr3, pyridine (0.05 equiv), Et2O, 0 °C, 1.5 h; (d) 1trimethylsilyl-1-propyne (4.0 equiv), n-BuLi (4.0 equiv), THF, −20 °C, 30 min, then 24, 0 °C, 12 h; (e) TBSCl (1.5 equiv), imidazole (1.7 equiv), DMF, rt, 2 h; (f) K2CO3 (1.5 equiv), MeOH, rt, 12 h; (g) nBuLi (2.5 equiv), THF, −78 °C, 2 h, then MeI (3.0 equiv), rt, 1 h; (h) Cp2ZrCl2 (4.8 equiv), DIBAL-H (4.6 equiv), THF, 50 °C, 1.5 h, then I2 (3.0 equiv), THF, 0 °C, 15 min; (i) TBAF (2.0 equiv), THF, rt, 3 h; (j) DMP (2.0 equiv), CH2Cl2, rt, 15 min; (k) KOH (2.6 equiv), I2 (1.3 equiv), MeOH, 0 °C, 1 h.
(9) prepared from citronellol24 was oxidized to give the major alcohol 23 and a minor byproduct 22 that was then recycled by reduction with NaBH4.25 Bromination26 of 23 with PBr3 provided 24, which was subjected to a coupling reaction with 1-trimethylsilyl propargyl lithium accompanied by debenzoylation to produce alcohol 25.27 After tert-butyldimethylsilyl (TBS) protection, subsequent desilylation afforded the terminal alkyne 26, which was lithiated and then quenched with methyl iodide to furnish a one-carbon-more homologue 27.28 2489
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society Scheme 7. Syntheses of Compounds 1−4a
a
Reagents and conditions: (a) 7 (1.5 equiv), 21 (1.0 equiv), CrCl2 (12.0 equiv), Ni(dppe)Cl2 (0.1 equiv), DMSO, rt, 72 h; (b) K2CO3 (1.1 equiv), MeOH, rt, 1 h; (c) DMP (2.0 equiv), CH2Cl2, rt, 30 min; (d) K2CO3 (1.1 equiv), MeOH, rt, 1.5 h; (e) LiOH (15.0 equiv), THF/H2O (5:1), 50 °C, 3 h, then 1 M HCl (excess); (f) K2CO3 (1.1 equiv), MeOH, rt, 1.5 h; (g) K2CO3 (2.0 equiv), MeOH, rt, 2 h; (h) Ac2O (10.0 equiv), pyridine, 50 °C, 20 h; (i) KHCO3 (1.3 equiv), MeOH, rt, 3 h.
Regioselective hydrozirconation of 27 was achieved smoothly to yield 28 (11:1 regioselectivity in the preference of 28) after iodine quench.29 Deprotection of 28 produced a primary alcohol 29, which was then oxidized to aldehyde 30 with DMP. Finally, the aldehyde 30 was treated with an alkaline solution of iodine in methanol to deliver the target vinyl iodide 7.30 With both crucial building blocks of aldehyde 21 and vinyl iodide 7 in hand, our attention was thus directed to the intermolecular NHK coupling reaction, which afforded the two epimers 31a and 31b (Scheme 7A) in good overall yield (70%). Deprotection of 31b gave 12-epi-2 as confirmed by comparing its NMR data with those of 2, indicating that the C-12 of 31b did not possess our desired configuration (Scheme 7B). The epimer 31b was subsequently oxidized to ketone 32, deprotection of which finally afforded hongkonoid C (3) in appreciable yield (72%). Hydrolysis of 3 then yielded hongkonoid D (4) in moderate yield (59%).31 Deprotection of the acetyl group of 31a was unexpectedly slower than the cases of 31b and 32 due likely to the effect from the β-oriented OH-12 in 31a, and the products were decomposed when the reaction was kept overnight. In order to avoid the decomposition, the reaction was terminated in about 1.5 h to give hongkonoid B (2) and intermediate 33 that was then deacetylated to yield compound 2. Selective acetylation of OH12 of compound 2 failed since OH-5′ was more reactive. Acetylation of 2 was thus conducted under a heating condition
to produce the diacetylated product 34 (Scheme 7C), which after selective deacetylation at OAc-5′ with KHCO3 by the aforementioned method yielded the last target hongkonoid A (1). The NMR data and optical rotations of the synthetic compounds 1−4 were all in good agreement with those of the corresponding natural hongkonoids A−D, respectively (see SI). Our synthetic work unequivocally confirmed the structural assignment for compounds 1−4. Bioassays. Nuclear factor-κB (NF-κB) is an important transcription factor that plays central roles in inflammation, oxidative stress, and carcinogenesis by regulating the expression of genes critically involved in the process of inflammation, immunity, aging, cell survival and apoptosis, and metabolic diseases.32 Compounds 1−4, 12-epi-2, 33, and 34 were evaluated in NF-κB pathway luciferase assay for the inhibitory effects,33 and compound 1 and two synthetic analogs 33 and 34 exhibited inhibition with IC50 values ranging from 2.87 to 16.3 μM (Table 1). The result suggested that acetylation of OH-2′ or OH-5′ of 1 and 2 will increase the inhibitory activity of this compound class. The most active compound 34 was further investigated for the effects on the LPS-induced inflammatory responses in RAW 246.7 macrophages. MTS test on 34 was performed to exclude the possible cytotoxicity, and it did not show obvious cytotoxicity at the indicated concentrations (Figure S1, SI). RAW 246.7 macrophages were treated with LPS and the 2490
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society
Compounds 1−4, 12-epi-2, 33, and 34 were also evaluated for the inhibitory activities against both human and mouse 11βhydroxysteroid dehydrogenase type 1 (11β-HSD1), which are NADPH-dependent enzymes that regulate interconversion of the active/inactive forms of glucocorticoids.35 Compounds 3 and 4 showed selective inhibition against mouse 11β-HSD1 with IC50 values of 0.86 ± 0.30 and 7.74 ± 1.63 μM, respectively, and glycyrrhetinic acid was used as the positive control (IC50 = 8.76 ± 3.44 nM) (Figure S2, SI). The inhibitory activities of compounds 1−4, 12-epi-2, 33, and 34 against the sterol synthesis in HepG2 cells were tested by the reported method with 5-aminoimidazole-4-carboxamide1-β-D-ribofuranoside (AICAR) as the positive control (ca. 40% inhibition at 500 μM) (Table S7, SI).36 Compounds 3 and 34 showed significant inhibition with IC50 values of 14.8 ± 3.9 and 21.8 ± 4.5 μM, respectively (Figure S3, SI). In conclusion, hongkonoids A−D (1−4) were discovered and characterized as the first example of ascorbylated terpenoids from the medicinal plant D. hongkongense. Moreover, asymmetric total syntheses of compounds 1−4 were achieved with the longest linear sequences of 12−14 steps and overall yields of 5.4−9.6% starting from readily available material 9, which not only corroborated the structural assignments for the compounds but also mimicked the biosynthetic proposal for hongkonoids A−D. In particular, a bioinspired one-pot method to construct the key structural fragment 5,5,5-fused tricyclic spiroketal butyrolactone 14 was exploited, which might pave a way for the synthesis of natural products and relevant congeners with such ring systems. The unprecedented structures and the aforementioned important biological activities of this compound class provided a new structural template of promising drug leads for the treatment of metabolic disorders and cardiovascular diseases.
Table 1. Inhibitory Effects of 1−4, 12-epi-2, 33, and 34 in NF-κB Pathway Luciferase Assay cmpd
IC50 (μM)
cmpd
IC50 (μM)
1 2 3 4
16.3 ± 0.7 NAa NAa NAa
12-epi-2 33 34 PS-341b
NAa 7.59 ± 3.43 2.87 ± 1.32 0.28 ± 0.08
IC50 > 50 μM was considered inactive (NA). bPS-341 was used as the positive control. a
expression of pro-inflammatory genes was investigated by qRTPCR at the presence or the absence of 34. The mRNA levels of pro-inflammatory genes, including TNF-α, IL-1β, IL-6, iNOS, COX2, and MCP-1, were decreased by the treatment with 34 at 5 and 10 μM as compared to the LPS-induced group (Figure 6A). Phosphorylation of IκB-α by IKK leads to the degradation
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b10135. NMR data for natural and synthetic 1−4, experimental details, X-ray crystal data for 5, and NMR spectra of 1−4 and all the intermidates (PDF) X-ray crystallographic data of 5 (CIF)
Figure 6. Compound 34 suppresses the LPS-induced inflammatory responses in macrophages. (A, C) Effect of 34 on LPS-induced target genes expression in RAW 246.7 cells and primary BMDMs, respectively. Cells were treated with LPS alone (100 ng/mL) or together with 34 for 24 h. BAY 11-7082 was set as a positive control, and the relative mRNA level of each gene was normalized by GAPDH. (B, D) Effect of 34 on LPS-induced NF-κB activation pathway in RAW 246.7 cells and primary BMDMs, respectively. Cells were pretreated with 34 for 6 h before LPS (100 ng/mL) treatment for 30 min. The results are shown as the mean ± SEM *P < 0.05, ** P < 0.01, ***P < 0.001.
■
AUTHOR INFORMATION
Corresponding Author
*
[email protected] ORCID
Jian-Min Yue: 0000-0002-4053-4870 Author Contributions §
These authors contributed equally.
of IκB-α, which results in the translocation of NF-κB into the nucleus to promote the transcription of target genes.34 We then examined the effects of 34 on the levels of IκB-α and the phosphorylation of IKK and NF-κB induced by LPS in RAW 246.7 cells. As was expected, compound 34 decreased LPSdependent IKK and NF-κB phosphorylation and enhanced IκBα level (Figure 6B). We further tested the effects of 34 on primary mouse BMDMs. The expression of pro-inflammatory genes was reduced, and the NF-κB related signaling pathways were changed accordingly in a dose-dependent manner (Figure 6C,D).
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation (no. 21532007) and National Key Research and Development (no. 2016YFC1305500) of the People’s Republic of China. We would like to give our truthful thanks to Prof. Zhi-Xiang Xie of Lanzhou University for his valuable discussions during the work. We also thank Prof. Shi-Man 2491
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
Article
Journal of the American Chemical Society
(27) Huang, J.; Wu, C.; Wulff, W. D. J. Am. Chem. Soc. 2007, 129, 13366. (28) (a) Qiu, W. W.; Surendra, K.; Yin, L.; Corey, E. J. Org. Lett. 2011, 13, 5893. (b) Surendra, K.; Rajendar, G.; Corey, E. J. J. Am. Chem. Soc. 2014, 136, 642. (29) (a) Valente, C.; Organ, M. G. Chem. - Eur. J. 2008, 14, 8239. (b) Takizawa, A.; Fujiwara, K.; Doi, E.; Murai, A.; Kawai, H.; Suzuki, T. Tetrahedron 2006, 62, 7408. (30) (a) Yamada, S.; Morizono, D.; Yamamoto, K. Tetrahedron Lett. 1992, 33, 4329. (b) Hoffmann, R. W.; Schlapbach, A. Tetrahedron Lett. 1993, 34, 7903. (c) Li, X. Q.; He, Q.; Tu, Y.; Zhang, F. M.; Zhang, S. Y. Chin. J. Chem. 2007, 25, 1357. (31) Meng, D. F.; Tan, Q.; Danishefsky, S. J. Angew. Chem., Int. Ed. 1999, 38, 3197. (32) (a) Sarkar, F. H.; Li, Y.; Wang, Z.; Kong, D. Int. Rev. Immunol. 2008, 27, 293. (b) Baker, R. G.; Hayden, M. S.; Ghosh, S. Cell Metab. 2011, 13, 11. (33) (a) Peng, Y. M.; Zheng, J. B.; Zhou, Y. B.; Li, J. Acta Pharmacol. Sin. 2013, 34, 939. (b) Xu, J. B.; Fan, Y. Y.; Gan, L. S.; Zhou, Y. B.; Li, J.; Yue, J. M. Chem. - Eur. J. 2016, 22, 14648. (34) Karin, M.; Greten, F. R. Nat. Rev. Immunol. 2005, 5, 749. (35) (a) Mundt, S.; Solly, K.; Thieringer, R.; Hermanowski-Vosatka, A. Assay Drug Dev. Technol. 2005, 3, 367. (b) Ye, Y. L.; Zhou, Z.; Zou, H. J.; Shen, Y.; Xu, T. F.; Tang, J.; Yin, H. Z.; Chen, M. L.; Leng, Y.; Shen, J. H. Bioorg. Med. Chem. 2009, 17, 5722. (36) (a) Alberts, A. W.; Ferguson, K.; Hennessy, S.; Vagelos, P. R. J. Biol. Chem. 1974, 249, 5241. (b) Huang, S. L.; Yu, R. T.; Gong, J.; Feng, Y.; Dai, Y. L.; Hu, F.; Hu, Y. H.; Tao, Y. D.; Leng, Y. Diabetologia 2012, 55, 1469.
Huang of Hainan University for the identification of the plant material.
■
REFERENCES
(1) (a) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2016, 79, 629. (b) Cragg, G. M.; Grothaus, P. G.; Newman, D. J. Chem. Rev. 2009, 109, 3012. (2) Zhang, Q.; Luo, S.; Wang, H. Acta Bot. Yunnanica 1998, 20, 362. (3) Chen, S. K.; Li, H.; Chen, P. Y. In Zhongguo Zhiwu Zhi; Science Press: Beijing, 1997; Vol. 43, p 89. (4) (a) He, X. F.; Wang, X. N.; Gan, L. S.; Yin, S.; Dong, L.; Yue, J. M. Org. Lett. 2008, 10, 4327. (b) Han, M. L.; Zhao, J. X.; Liu, H. C.; Ni, G.; Ding, J.; Yang, S. P.; Yue, J. M. J. Nat. Prod. 2015, 78, 754. (c) Zhou, B.; Shen, Y.; Wu, Y.; Leng, Y.; Yue, J. M. J. Nat. Prod. 2015, 78, 2116. (5) (a) Kesinger, N. G.; Stevens, J. F. Phytochemistry 2009, 70, 1930. (b) Zhang, Z.; Liu, X.; Zhang, X.; Liu, J.; Hao, Y.; Yang, X.; Wang, Y. Arch. Pharmacal Res. 2011, 34, 801. (c) Sun, J.; Zhang, P.; Wei, Q.; Xun, H.; Tang, F.; Yue, Y.; Li, L.; Guo, X.; Zhang, R. Tetrahedron Lett. 2014, 55, 4529. (6) (a) Pappas, J. J.; Keaveney, W. P.; Gancher, E.; Berger, M. Tetrahedron Lett. 1966, 7, 4273. (b) Liu, J.; He, X. F.; Wang, G. H.; Merino, E. F.; Yang, S. P.; Zhu, R. X.; Gan, L. S.; Zhang, H.; Cassera, M. B.; Wang, H. Y.; Kingston, D. G.; Yue, J. M. J. Org. Chem. 2014, 79, 599. (7) Flack, H. D. Acta Crystallogr., Sect. A: Found. Crystallogr. 1983, 39, 876. (8) Mori, K. Tetrahedron 1977, 33, 289. (9) (a) Buchanan, B. B.; Gruissem, W.; Jones, R. L. Biochemistry & molecular biology of plants; American Society of Plant Physiologists: Rockville, MD, 2000. (b) Piotrowska, A.; Bajguz, A. Phytochemistry 2011, 72, 2097. (10) (a) Traber, M. G.; Stevens, J. F. Free Radical Biol. Med. 2011, 51, 1000. (b) Wheeler, G. L.; Jones, M. A.; Smirnoff, N. Nature 1998, 393, 365. (11) Rüdiger, W. Phytochemistry 1997, 46, 1151. (12) Collakova, E.; DellaPenna, D. Plant Physiol. 2001, 127, 1113. (13) Keller, Y.; Bouvier, F.; d’Harlingue, A.; Camara, B. Eur. J. Biochem. 1998, 251, 413. (14) Hedden, P.; Phillips, A. L. Trends Plant Sci. 2000, 5, 523. (15) Nambara, E.; Marion-Poll, A. Annu. Rev. Plant Biol. 2005, 56, 165. (16) Cho, B. H.; Kim, J. H.; Jeon, H. B.; Kim, K. S. Tetrahedron 2005, 61, 4341. (17) Kodama, M.; Shiobara, Y.; Sumitomo, H.; Fukuzumi, K.; Minami, H.; Miyamoto, Y. J. Org. Chem. 1988, 53, 1437. (18) (a) Wimalasena, K.; Mahindaratne, M. P. D. J. Org. Chem. 1994, 59, 3427. (b) Olabisi, A. O.; Wimalasena, K. J. Org. Chem. 2004, 69, 7026. (c) Olabisi, A. O.; Mahindaratne, M. P.; Wimalasena, K. J. Org. Chem. 2005, 70, 6782. (19) Shi, Y.; Jiang, X. L.; Tian, W. S. J. Org. Chem. 2017, 82, 269. (20) Poss, A. J.; Belter, R. K. Synth. Commun. 1988, 18, 417. (21) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155. (22) McDonald, F. E.; Ishida, K.; Hurtak, J. A. Tetrahedron 2013, 69, 7746. (23) (a) Cheng, X.; Micalizio, G. C. J. Am. Chem. Soc. 2016, 138, 1150. (b) Trost, B. M.; Krische, M. J. J. Am. Chem. Soc. 1996, 118, 233. (c) Zhang, J. J.; You, L.; Wang, Y. F.; Li, Y. H.; Liang, X. T.; Zhang, B.; Yang, S. L.; Su, Q.; Chen, J. H.; Yang, Z. Chem. - Asian J. 2016, 11, 1414. (d) Kuramochi, K.; Saito, F.; Takeuchi, R.; Era, T.; Takemura, M.; Kobayashi, J.; Sakaguchi, K.; Kobayashi, S.; Sugawara, F. Tetrahedron 2006, 62, 8006. (24) García, C.; León, L. G.; Pungitore, C. R.; Ríos-Luci, C.; Daranas, A. H.; Montero, J. C.; Pandiella, A.; Tonn, C. E.; Martín, V. S.; Padrón, J. M. Bioorg. Med. Chem. 2010, 18, 2515. (25) Sprung, I.; Carmès, L.; Watt, G. M.; Flitsch, S. L. ChemBioChem 2003, 4, 319. (26) Armstrong, R. J.; Weiler, L. Can. J. Chem. 1983, 61, 2530. 2492
DOI: 10.1021/jacs.7b10135 J. Am. Chem. Soc. 2018, 140, 2485−2492
