Pogocablanols A–G, a new class of sesquiterpenoids with an unprecedented bicyclic sesquiterpenoid skeleton from Pogostemon cablin (patchouli) and their activities

2.1 General experimental procedures

Optical rotations were measured on a Rudolph Autopol-I automatic polarimeter (Rudolph Research Analytical, USA). UV and CD spectra were recorded using an Applied Photophysics Chirascan and Chirascan-plus circular dichroism spectrometer (Applied Photophysics Ltd., Leatherhead, England). An Agilent Cary 600 FT-IR microscope instrument (Agilent Technologies Inc., CA, USA) was used to measure IR spectra. NMR data were obtained by a Bruker-AVIIIHD-600 spectrometer (Bruker Corporation, MA, USA) with the solvent peaks used as references. HR-ESI-MS data were acquired on a Waters Synapt G₂ HDMS instrument (Waters Corporation Milford, MA, USA). X-ray crystallographic analysis was performed on a Bruker D8 Quest diffractometer with Cu Kα radiation. Column chromatography was performed using silica gel (200–300 mesh; Yantai Institute of Chemical Technology, Yantai, China), and Toyopearl HW-40F (Tosoh Corp., Japan). TLC was performed using glass precoated silica gel GF₂₅₄ plates (Qingdao Marine Chemical Inc., Shandong, China). MPLC separations were performed using a BÜCHI Gradient Former B-687 instrument (BÜCHI, Flawil, Switzerland). HPLC separation was carried out using an Agilent 1100 instrument with a Kromasil semi-preparative C₁₈ column (250 × 10 mm², 5 μm) for reversed-phase semi-preparative HPLC. Fetal bovine serum (FBS) was purchased from Zhejiang Tianhang Biotechnology Co., Ltd, and dulbecco's modified Eagle's medium (DMEM) was purchased from Hyclone (Logan, UT, USA). Assays for relaxant activity were conducted with a Radnoti 8 Channel Tissue Bath System (Radnoti Glass Tech., Monrovia, CA).

2.2 Plant material

Pogostemon cablin (Blanco) Benth. was collected in Yangchun City, Guangdong Province, China, in December 2012. The sample was identified by Dr. Fei Long and deposited in the School of Pharmacy in Chengdu University of TCM, Chengdu, China (voucher specimen: SGHX-20121224).

2.3 Extraction and isolation

The air-dried leaves and stems of P. cablin (40 kg) were subjected to hydrodistillation for 10 h using a big modified Clevenger-type apparatus with a water-cooled oil receiver to obtain the essential oil (215 g). After dried with anhydrous sodium sulfate, the oil was subjected to silica gel column chromatography using a gradient elution of increasing EtOAc (0–100%) in petroleum ether to give 31 fractions (F₁–F₃₁) on the basis of TLC analysis. F₁₈ (15.5 g) was further separated by reversed-phase (RP) MPLC with a gradient of increasing MeOH (40–100%) in water to afford 14 subfractions (F_18-1–F_18-14). The subfraction F_18-2 was isolated via silica gel column chromatography with a gradient solvent system (100:1–5:1, CHCl₃–Me₂CO) to obtain 11 subfractions (F_18-2-1–F_18-2-11). Further purification of F_18-2-3 by preparative TLC (CH₂Cl₂–n-hexane–Me₂CO, 3:1:0.1) and semipreparative RP HPLC (45% MeOH in H₂O) successively afforded compounds 2 (4.6 mg, t_R = 45 min) and 3 (3.5 mg, t_R = 56 min). The subfraction F_18-2-6 was purified via preparative TLC (CH₂Cl₂–n-hexane–Me₂CO, 2:1:0.1) to obtain compound 1 (5.6 mg). The subfraction F_18-2-7 was further divided into four parts (F_18-2-7-1–F_18-2-7-4) using a Toyopearl HW-40F column (85% MeOH in H₂O, v/v). F_18-2-7-2 was applied to preparative TLC (CH₂Cl₂–n-hexane–Me₂CO, 4:1:0.1) to yield compound 5 (4.7 mg). F_18-2-7-3 was further separated by semipreparative RP HPLC (50% MeOH in H₂O) to yield compound 6 (6.3 mg, t_R = 42 min). Compounds 4 (5.1 mg) and 7 (6.2 mg) were obtained from F_18-2-8 via preparative TLC (CH₂Cl₂–n-hexane–Me₂CO, 2:1:0.1).

2.4 X-ray crystallographic analysis

2.5 Effects of compounds 1–7 on acetylcholine-induced contractions of rat jejunum smooth muscle

Male Sprague-Dawley rats (200–220 g) were obtained from Da Shuo Biotechnology Co., Ltd (Chengdu, Sichuan, China). All rats were maintained on a 12 h light/dark cycle under a controlled temperature at 25 ± 1 °C and a relative humidity of 50 ± 5%. All the experimental procedures were performed in accordance with the guidelines of the Management Committee for Experimental Animals, China. Animals were fasted for 12 h but free access to water before used. The animal was sacrificed by cervical dislocation, and then the jejunum smooth muscle was carefully dissected and immediately transferred into oxygenated Tyrode’s solution [composition (mM): NaCl, 136.9; KCl, 2.7; NaHCO₃, 11.9; NaH₂PO₄, 0.4; MgCl₂, 1.1; glucose, 5.6; CaCl₂, 1.8] at 4 °C (Ghayur et al., 2005). The jejunum was cleaned of surrounding fat and connective tissues and cut into approximately 1–2 cm in length, and then suspended in Tyrode’s solution at a constant temperature of 37 °C and bubbled with a gas mixture of 95% O₂ and 5% CO₂. All the muscles were equilibrated under an initial tension of 1 g in normal Tyrode’s solution before the start of the experiment (Bian et al., 2015). The jejunum smooth muscles were pre-contracted with 1 mg/mL acetylcholine (Ach) solution. When the muscles were stably contracted, the cumulative concentrations of the test compounds (0.5, 1.5, 5, 15, and 50 μM) and dimethyl sulfoxide (DMSO) were added into the organ bath, and the effects were recorded. The relaxation and E_max (maximal relaxation) values of the test compounds were calculated.

2.6 Effects of compounds 1–7 on ox-LDL-induced RAW264.7 macrophage foam cells formation

3.1 Structure elucidation

Pogocablanol A (1) was obtained in the form of colorless crystals. Its molecular formula was established to be C₁₅H₂₄O₄ by HR-ESI-MS (m/z 291.1574 [M + Na]⁺, calcd 291.1572), corresponding to four degrees of unsaturation. The IR spectrum displayed characteristic absorptions for hydroxy and carbonyl functionalities (3445 and 1737 cm⁻¹). The ¹H NMR spectrum of compound 1 (Table 1) showed signals of three tertiary methyls (δ_H 1.19, 1.22, and 1.27), two exchangeable hydroxy protons (δ_H 3.12 and 3.81), and several aliphatic methylenes and methines between δ_H 1.44 and 2.25 ppm. The ¹³C NMR and DEPT data (Table 2) revealed the presence of 15 carbons, including three methyls, six methylenes, one methine, four oxygenated quaternary carbons, and one carbonyl carbon. The above functionalities account for one out of the four degrees of unsaturation, and the remaining three degrees suggest compound 1 to be tricyclic. A comprehensive analysis of the 2D NMR spectra was conducted to establish the planar structure of compound 1. The ¹H–¹H COSY correlations of H₂-3/H₂-4, as well as the HMBC correlations of OH-5 and H₃-15 with C-1, C-4, and C-5; of H₂-4 with C-1, C-2, and C-5; and of H₂-3 with C-2, C-4, and C-5, established the existence of a five-membered carbocycle (ring A) possessing a hydroxy and a methyl group substituted at C-5 and a carbonyl group at C-2 in compound 1 (Fig. 2). Additionally, the ¹H–¹H COSY correlations of H-7/H₂-12/H₂-11, combined with the HMBC correlations from H₃-13 and OH-10 to C-9, C-10, and C-11, from H₃-14 to C-7, C-8, and C-9, constructed the substructure of a six-membered carbocycle (ring B) with a hydroxy group at C-10 and two methyl groups attached to C-8 and C-10, respectively. Regarding the linkage of rings A and B, the ¹H–¹H COSY correlation of H₂-6/H-7, together with the HMBC correlations of H₂-6 with C-1, C-2, C-5, C-7, and C-8, revealed that rings A and B were connected by a methylene bridge (C-6). Finally, based on the molecular formula and the ¹³C chemical shifts of C-1 (δ_C 88.9) and C-8 (δ_C 84.7), a 1-O-8 ether linkage was deduced to form an interesting 1-oxaspiro[4,4]nonane unit in compound 1.

Close

Fig. 2

Key ¹H−¹H COSY and HMBC correlations of 1–7.

Export to PPT

In the NOESY spectrum, cross peaks of H₃-13/H-7 and H₃-14/OH-10 revealed the spatial orientations of these protons in 1 (Fig. 3). However, it was difficult to elucidate the stereochemistry of the spiro carbon C-1 due to a lack of available NOESY data. Fortunately, a single crystal of compound 1 was obtained and successfully analyzed by X-ray diffraction using Cu Kα radiation (Fig. 4) with a Flack parameter of −0.14(13). Therefore, the absolute configuration of compound 1 was unambiguously determined to be 1S,5R,7R,8S,10R.

Close

Fig. 3

Key NOESY correlations of 1–6.

Export to PPT

Close

Fig. 4

X-ray crystallographic structures of 1–7.

Export to PPT

Pogocablanol B (2) appears as colorless crystals, with the same molecular formula, C₁₅H₂₄O₄, as compound 1. Comparison of the ¹H, ¹³C, and 2D NMR data of 2 and 1 (Tables 1 and 2 and Fig. 2) suggested that the two compounds have the same planar structure. Their ¹³C NMR data showed minor differences in the same deuterated solvent. The C-7, C-9, C-10, C-11, C-12, and C-13 resonances in compound 2 were shielded by Δδ −7.3, −3.3, −1.5, −8.5, −2.0, and −1.1 ppm, respectively, while the C-6 and C-14 resonances in compound 2 were deshielded by Δδ +3.3 and +8.3 ppm, respectively. Therefore, compound 2 is a stereoisomer of compound 1. The relative stereochemistry of H₃-14, H-7, and OH-10 in 2 was deduced from the NOESY data of H₃-14 with H-7 and OH-10 (Fig. 3). An X-ray diffraction experiment verified the planar structure of 2 and confirmed its absolute configuration to be 1R,5S,7R,8R,10S [Flack parameter = −0.01(12)] (Fig. 4).

Pogocablanol C (3) is an isomer of 1, as indicated by HR-ESI-MS, ¹H NMR, and ¹³C NMR data. Detailed analysis of the ¹H−¹H COSY, HSQC, and HMBC data confirmed that they have the same planar structure. The NOESY data and X-ray diffraction analysis [Flack parameter = 0.07(8)] further confirmed the structure and absolute configuration (1R,5S,7R,8S,10R) of compound 3.

Pogocablanol D (4) was assigned a molecular formula of C₁₅H₂₄O₂ by an ion peak at m/z 259.1676 [M + Na]⁺ in the HR-ESI-MS (calcd for C₁₅H₂₄O₂Na, 259.1674). The ¹H NMR data (Table 1) of compound 4 showed signals of three methyl groups, one exchangeable proton, and several aliphatic methylenes and methines. A total of 15 carbon signals, corresponding to three methyl groups, six methylenes, two methines, and four quaternary carbons (two olefinic), were revealed by its ¹³C NMR (Table 2) and DEPT spectra. These data indicated that compound 4 was a pogocablane sesquiterpenoid similar to 3. Ring A in 4 was confirmed by ¹H−¹H COSY correlations of H₂-3/H₂-4/H-5/H₃-15 and HMBC correlations of H₃-15 with C-1, C-4, and C-5; of H₂-3 with C-1 and C-5; and of H₂-4 with C-1, C-2, and C-15 (Fig. 2). Ring B in 4 was the same as that in 3, as indicated by the ¹H−¹H COSY correlations of H₂-6/H-7/H₂-12/H₂-11, together with HMBC correlations from H₃-13 and OH-10 to C-9, C-10, and C-11; from H₃-14 to C-7, C-8, and C-9; and from H₂-12 to C-8 and C-10. Based on the HMBC correlations of H₂-6 with C-1, C-2, C-7, and C-8, the molecular formula, and the chemical shifts of C-2 (δ_C 149.7) and C-8 (δ_C 78.9), it was deduced that a pyran ring fused with rings A and B. A series of cross peaks of H₃-15/H-6a/H-7/H₃-13 and H₃-14/OH-10 observed in the NOESY spectrum of 4 revealed its relative stereochemistry (Fig. 3). A single crystal of 4 was obtained in MeOH for X-ray diffraction analysis [Flack parameter = 0.01(7)], which verified its absolute configuration to be 5S,7R,8S,10R (Fig. 4).

Pogocablanol E (5) has the molecular formula C₁₅H₂₄O₃ (HR-ESI-MS). The spectroscopic data of 5 resembled those of 4 (Tables 1 and 2), with the differences in the migration of the double bond from Δ¹⁽²⁾ in 4 to Δ¹⁽⁵⁾ in 5 and replacement of the oxygenated olefinic quaternary carbon (C-2) in 4 by a carbonyl (δ_C 209.1) in 5. In addition, an extra exchangeable proton signal observed in the ¹H NMR spectrum of 5 suggested that a hydroxy group is attached to C-8 in 5 to match the molecular composition. Thus, rings A and B are connected only through a methylene group. The planar structure of 5 was further supported by ¹H−¹H COSY and HMBC data (Fig. 2), especially the ¹H−¹H COSY cross peaks of H₂-3/H₂-4 and H₂-6/H-7/H₂-12/H₂-11 and HMBC correlations from H₃-14 and OH-8 to C-7, C-8, and C-9; from H₂-12 to C-6, C-7, and C-8; from H₃-15 to C-1, C-4, and C-5; and from H₂-6 to C-1, C-2, C-5, C-7, C-8, and C-12. The NOESY correlations of H₃-14 with OH-10 and H-7 with OH-8 revealed the anti-orientation of OH-10 to H-7 and OH-8. Unambiguously, a single crystal of 5 was successfully analyzed by X-ray diffraction using Cu Kα radiation (Fig. 4) with a Flack parameter of –0.01(5). Therefore, the absolute configuration of 5 was determined as 7R,8S,10R.

Pogocablanol F (6) had similar NMR data to 5. The molecular formula of compound 6 was determined as C₁₅H₂₆O₃ with two more hydrogen atoms than compound 5. By comparing the ¹H and ¹³C NMR data of 6 and 5, the only difference between them was that the Δ¹⁽⁵⁾ double bond in 5 was reduced in 6. The planar that the Δ¹⁽⁵⁾ double bond in 5 was reduced in 6. The planar structure of 6 was confirmed by ¹H–¹H COSY and HMBC data analysis (Fig. 2), especially the ¹H–¹H COSY correlations of H₂-3/H₂-4/H-5/H-1/H₂-6/H-7/H₂-12/H₂-11. In the NOESY spectrum, only two available correlations were observed (H₃-13/OH-8 and H₃-15/H-1). Fortunately, the X-ray diffraction data of compound 6 indicated that its absolute configuration is 1R,5S,7R,8S,10R [Flack parameter = –0.07(7)].

The IR, HR-ESI-MS, and NMR data of compound 7 indicated that it is an epimer of 6. The X-ray diffraction data of 7 (Fig. 4) confirmed its relative configuration; however, its absolute configuration could not be confirmed reliably [Flack parameter = 0.13(10)]. Comparison of the experimental and calculated ECD data of 7 revealed its configuration as 1R,5S,7R,8R,10R (Fig. 5). Therefore, the structure of pogocablanol G (7) was defined as (1R,5S,7R,8R,10R)-pogocabla-8,10-diol-2-one.

Close

Fig. 5

The experimental and calculated ECD spectra of compound 7.

Export to PPT

The biogenetic origin of pogocablanols A–G (1–7) could plausibly be traced back to 5-epi-α-bulnesene (Scheme 1), a guaiane-type sesquiterpene that has been found in the essential oils of plants (Moronkola et al., 2015; Cai et al., 2014). First, the Δ^1,10 bond of bulnesene is oxidatively cleaved to form intermediate i, which contains two characterized carbonyl units (Zhou et al., 2022). The Prins reaction is an acid-catalyzed condensation of alkenes with ketones or aldehydes, which is widely used in the fragrance industry (Doro et al., 2019). Due to the non-stereoselectivity, intermediate i is likely to undergo an intramolecular Prins reaction to construct two epimers (intermediates ii and iii) (Meinwald and Yankeelov, 1958; Doro et al., 2019). Then, intermediate ii is nucleophilically attacked by H₂O to form two sesquiterpenoids featuring a C5-C1-C6 skeleton system with three methyl groups (6 and 7). Subsequently, pogocablanol D (4) was generated through hemiketalation and dehydration from pogocablanol F (6) (Xiong et al., 2015). In addition, pogocablanol F (6) was oxidized to afford pogocablanol E (5). Successive oxidation, nucleophilic substitution, and etherification of 5 would afford pogocablanols A and C (1 and 3) substitution processes (Zhou et al., 2022). Pogocablanol B (2) could be yielded from intermediate iii through the same reaction processes used for the generation of pogocablanols A and C due to the stereoselectivity in the epoxidation and nucleophilic (1 and 3).

Close

Scheme 1

Possible biosynthetic pathways for pogocablanols A–G (1–7).

Export to PPT

3.2 Biological activities

3.2.1

3.2.1 Relaxing effects on rat jejunum smooth muscle

The aerial part of P. cablin (patchouli) is used as a traditional Chinese medicine for the treatment of abdominal pain, vomiting, and diarrhea. It is well known that gastrointestinal smooth muscle spasms can cause abdominal pain and diarrhea, which occur frequently in the clinic. Modern studies have reported that patchouli and its compound preparations (Huoxiang Zhengqi Oral Liquid and Huoxiang Zhengqi Capsule) have good effects on intestinal disorders such as diarrhea-predominant irritable bowel syndrome, intestinal mucositis, acute colitis, and diarrhea (Chen et al., 2021; Guo et al., 2021; Gan et al., 2020; Yu et al., 2017), while patchouli and its extract act as relaxants on intestinal smooth muscles (Chen et al., 1998; Zhou et al., 2018b). Therefore, the regulatory effects of pogocablanols A–G on intestinal functions were investigated using an in vitro model of Ach-induced contraction of isolated rat jejuna. As shown in Fig. 6, compounds 1–4, at concentrations of 0.5, 1.5, 5, 15, and 50 μM, significantly decreased the contraction tension of the rat jejuna smooth muscle induced by Ach compared to the control group (P < 0.01 or P < 0.05), compounds 5–7 had no significant effects. The E_max values of 1, 2, 3, and 4 against Ach-induced contractions were 57.19 ± 10.17%, 69.02 ± 7.54%, 63.29 ± 14.42%, and 48.87 ± 4.01%, respectively. After comparing the results, the preliminary structure–activity relationships of these pogocablane-type sesquiterpenoids were obtained. The only difference between 1–4 and 5–7 was the linkage of the five- and six-membered carbocyclic rings (A and B). Thus, it can be inferred that the ether linkage between rings A and B to form an oxygen heterocycle plays an important role in the intestinal regulatory function. In addition, comparison of the effects of compounds 1–3 and 4 indicated that the 1-oxaspiro[4,4]nonane moiety is more beneficial for this relaxing effect than the pyran ring in pogocablanol-type sesquiterpenoids.

Close

Fig. 6

Relaxing effects of compounds 1, 2, 3, and 4 on isolated rat jejunum pre-contracted with Ach. Data are shown as the mean ± SD (n ≥ 3). *P < 0.05, **P < 0.01 vs. control group.

Export to PPT

3.2.2

3.2.2 Inhibitory activity against ox-LDL-induced RAW264.7 macrophage foam cells formation

Structurally diverse sesquiterpenoids, such as pachouli alcohol, a major component of pachouli oil, have been reported to have inhibitory effects on atherosclerosis (Liu et al., 2015; He et al., 2021; Wang et al., 2016). To investigate the potential roles of these novel pogocablane-type compounds in atherosclerosis progression, Nile red staining, Oil red O staining, and a cholesterol assay were performed to evaluate the inhibitory activity against ox-LDL-induced foam cell formation. First, compounds 1–7 were tested for their cytotoxicity in RAW264.7 macrophages, and the results showed no cytotoxicity at concentrations of 12.5, 25, and 50 μM. Subsequently, adipogenesis was analyzed by Nile red staining in RAW264.7 macrophages after treatment with ox-LDL and different concentrations of compounds 1–7. As shown in Fig. 7, at concentrations of 12.5, 25, and 50 μM, compound 2 significantly attenuated ox-LDL-induced adipogenesis in RAW264.7 macrophages in a concentration-dependent manner, while the other compounds showed no such activity.

Close

Fig. 7

Inhibitory activity of compound 2 against ox-LDL-induced adipogenesis in RAW264.7 macrophages. The lipid content was measured by Nile red staining. Data are shown as the mean ± SD (n = 3). **P < 0.01 vs. model group.

Export to PPT

Oil red O staining was further used to confirm the effect of compound 2 on lipid accumulation in RAW264.7 macrophages. As shown in Fig. 8A and 8B, there was a large amount of red staining, indicating more lipid accumulation in the model group than the control group. The addition of compound 2 to ox-LDL-induced foam cells significantly decreased the lipid accumulation. These results suggest that compound 2 could decrease intracellular lipid accumulation in ox-LDL-induced macrophages.

Close

Fig. 8

Inhibitory activity of compound 2 against ox-LDL-induced lipid accumulation in foam macrophages, and cholesterol content in normal cells and foam cells. (A) The extent of lipid loading was assessed by Oil red O staining. The magnification of each panel is × 400. (B) Intracellular lipid deposition. (C) Content of total cholesterol (TC). (D) Content of free cholesterol (FC). (E) Content of cholesteryl ester (CE). Data are shown as the mean ± SD (n = 3). *P < 0.05, **P < 0.01 vs. model group.

Export to PPT

The levels of cellular total cholesterol (TC) and free cholesterol (FC) were determined by ELISA, and the level of cholesteryl ester (CE) was calculated as CE = TC – FC. Compared to the model group, the levels of TC, FC, and CE in RAW264.7 macrophages were significantly decreased by treatment with compound 2 (all P < 0.01) (Fig. 8C–8E). These results suggest that compound 2 inhibits ox-LDL-induced lipid accumulation in macrophages.