Mangrove Tirucallane- and Apotirucallane-Type Triterpenoids: Structure Diversity of the C-17 Side-Chain and Natural Agonists of Human Farnesoid/Pregnane–X–Receptor

Ten new triterpenoid compounds with structure diversity of the C-17 side-chain, including nine tirucallanes, named xylocarpols A–E (1–5) and agallochols A–D (6–9), and an apotirucallane, named 25-dehydroxy protoxylogranatin B (10), were isolated from the mangrove plants Xylocarpus granatum, Xylocarpus moluccensis, and Excoecaria agallocha. The structures of these compounds were established by HR-ESIMS and extensive one-dimensional (1D) and two-dimensional (2D) NMR investigations. The absolute configurations of 1 and 2 were unequivocally determined by single-crystal X-ray diffraction analyses, conducted with Cu Kα radiation; whereas those of 4, 6–8 were assigned by a modified Mosher’s method and the comparison of experimental electronic circular dichroism (ECD) spectra. Most notably, 5, 6, 7, and 9 displayed potent activation effects on farnesoid–X–receptor (FXR) at the concentration of 10.0 μM; 10 exhibited very significant agonistic effects on pregnane–X–receptor (PXR) at the concentration of 10.0 nM.


Introduction
Cholestasis, a clinical syndrome of hepatobiliary diseases, is usually caused by accumulation of bile acids in the liver and systemic circulation [1]. Long-term cholestasis can lead to primary biliary cirrhosis, primary sclerosing cholangitis, and hepatic failure. In clinical practice, abnormal metabolism of bile acids is deemed to be a crucial risk factor that induces cholestasis and cholestatic liver injury [1,2]. Farnesoid-X-receptor (FXR) and pregnane-X-receptor (PXR) are two members of the nuclear receptor family. Due to the regulation capability of a suite of genes involved in the metabolism, transport, and elimination of bile acids, FXR and PXR are considered to be the key target proteins for the treatment of cholestasis and liver injury [1,[3][4][5].
Xylocarpus granatum and Xylocarpus moluccensis, true-mangrove species of the family Meliaceae, have been used in folk medicine, particularly in South and Southeast Asian countries, for the treatment

Results and Discussion
Compound 1 was obtained as colorless crystals. Its molecular formula C30H50O5 with six indices of hydrogen deficiency was established by the negative HR-ESIMS quasi molecular ion peak at m/z 525.3356 ([M + Cl] − , calculated for 525.3352). According to the 1 H and 13 C NMR spectroscopic data of 1 (Tables 1 and 2), two elements of unsaturation were due to a carbon-carbon double bond and a carbonyl group. Thus, the molecule was tetracyclic. The 13 C NMR spectroscopic data and the

H-1 H COSY correlations between H-24/H 2 -23 (Supplementary Materials
S178-S180) and HMBC interactions between H-24/C-23, H-24/C-25, H 3 -26/C-24, and H 3 -27/C-24 (Supplementary Materials S184-S189).    rings-A, B, C, and D), was established to be the same as that of 7, except for the deficiency of the chiral C-24, by the accurate fit of their experimental ECD spectra ( Figure 10). Therefore, the structure of 8, named agallochol C, was determined to be (3S,5R,9R,10R,13S,14S,17S,20S)-3,25-dihydroxytirucalla-7,23-diene-6-one. The relative configuration of the tetracyclic tirucallane core (rings-A, B, C, and D) of 8 was established as the same as that of 7 by NOE interactions. Those between H-5/H-3, H-5/H-9, H-5/H3-29, H-9/H3-18, H3-18/H-20, H3-19/H3-28, H3-19/H3-30, and H3-30/H-17 assigned the same relative configuration of 8 as that of 7. Moreover, the absolute configuration of 8, particularly that of the tetracyclic tirucallane core (rings-A, B, C, and D), was established to be the same as that of 7, except for the deficiency of the chiral C-24, by the accurate fit of their experimental ECD spectra ( Figure 10). Therefore, the structure of 8, named agallochol C, was determined to be (3S,5R,9R,10R,13S,14S,17S,20S)-3,25-dihydroxytirucalla-7,23-diene-6-one.  resembled those of a trinortirucalla-7-ene, i.e., sikkimenoid F [27], except for the replacement of the C-24 aldehyde group in sikkimenoid F by a C-24 carboxyl group in 9. The above deduction was corroborated by the upshifted C-24 (δ H 9.75 (br s), δ C 203.1 CH in sikkimenoid F; whereas δ C 178.5 qC in 9) and HMBC correlations from H 2 -22 and H 2 -23 to the carbonyl carbon (C-24) of this carboxyl group. The relative configuration of the tetracyclic tirucallane core (rings-A, B, C, and D)   In order to search for natural agonists of human FXR and PXR, most of the above isolated compounds were screened for their agonistic effects on these nuclear receptors. Chenodeoxycholic acid (CDCA) or rifampicin was used as the positive control at the concentration of 80.0 μM or 10.0 μM, respectively (Figures 12 and 13). The results showed that 6 and 7 displayed significant agonistic effects on FXR at the concentration of 1.0 μM; while 5, 6, 7, and 9 exhibited significant agonistic effects on FXR at the concentration of 10.0 μM. Moreover, 1 displayed a moderate significant agonistic effect on FXR at the concentration of 10.0 μM ( Figure 12). Compound 10 exhibited a significant agonistic effect on PXR at the concentration of 10.0 nM, and even a higher agonistic effect on PXR as compared to that of the positive control, rifampicin, at the same concentration of 10.0 μM ( Figure 13). In order to search for natural agonists of human FXR and PXR, most of the above isolated compounds were screened for their agonistic effects on these nuclear receptors. Chenodeoxycholic acid (CDCA) or rifampicin was used as the positive control at the concentration of 80.0 µM or 10.0 µM, respectively (Figures 12 and 13). The results showed that 6 and 7 displayed significant agonistic effects on FXR at the concentration of 1.0 µM; while 5, 6, 7, and 9 exhibited significant agonistic effects on FXR at the concentration of 10.0 µM. Moreover, 1 displayed a moderate significant agonistic effect on FXR at the concentration of 10.0 µM (Figure 12). Compound 10 exhibited a significant agonistic effect on PXR at the concentration of 10.0 nM, and even a higher agonistic effect on PXR as compared to that of the positive control, rifampicin, at the same concentration of 10.0 µM (Figure 13).
In order to search for natural agonists of human FXR and PXR, most of the above isolated compounds were screened for their agonistic effects on these nuclear receptors. Chenodeoxycholic acid (CDCA) or rifampicin was used as the positive control at the concentration of 80.0 μM or 10.0 μM, respectively (Figures 12 and 13). The results showed that 6 and 7 displayed significant agonistic effects on FXR at the concentration of 1.0 μM; while 5, 6, 7, and 9 exhibited significant agonistic effects on FXR at the concentration of 10.0 μM. Moreover, 1 displayed a moderate significant agonistic effect on FXR at the concentration of 10.0 μM ( Figure 12). Compound 10 exhibited a significant agonistic effect on PXR at the concentration of 10.0 nM, and even a higher agonistic effect on PXR as compared to that of the positive control, rifampicin, at the same concentration of 10.0 μM ( Figure 13).

General Methods
Optical rotations were recorded at room temperature on a MCP200 modular circular polarimeter (Anton Paar GmbH, Seelze, Germany). A GENESYS 10S UV-Vis spectrophotometer (Thermo Scientific, Shanghai, China) was used to obtain UV spectra. The NMR spectroscopic data were measured on a Bruker AV-400 NMR spectrometer (Bruker Scientific Technology Co. Ltd., Karlsruhe, Germany) using TMS as the internal standard. Single-crystal X-ray diffraction analyses were carried out on an Agilent Xcalibur Atlas Gemini Ultra-diffractometer with mirror monochromated Cu Kα radiation (λ = 1.54184 Å) at 100 K. An LC-ESI (Bruker Daltonics, Bremen, Germany) and an LC-ESI-QTOF mass spectrometer (SYNAPTTM G2 HDMS, Waters, Manchester, UK) were used to acquire HR-ESIMS data. For electronic circular dichroism (ECD) spectra, a Jasco 810 spectropolarimeter (JASCO Corporation, Tokyo, Japan) was applied with the solvent of acetonitrile. Semi-preparative HPLC was carried out on a Waters 2535 pump equipped with a 2489 UV detector (Waters Corporation, Milford, NY, USA) and an ODS column (YMC, 250 × 10 mm inner diameter, 5 µ m).

General Methods
Optical rotations were recorded at room temperature on a MCP200 modular circular polarimeter (Anton Paar GmbH, Seelze, Germany). A GENESYS 10S UV-Vis spectrophotometer (Thermo Scientific, Shanghai, China) was used to obtain UV spectra. The NMR spectroscopic data were measured on a Bruker AV-400 NMR spectrometer (Bruker Scientific Technology Co. Ltd., Karlsruhe, Germany) using TMS as the internal standard. Single-crystal X-ray diffraction analyses were carried out on an Agilent Xcalibur Atlas Gemini Ultra-diffractometer with mirror monochromated Cu Kα radiation (λ = 1.54184 Å) at 100 K. An LC-ESI (Bruker Daltonics, Bremen, Germany) and an LC-ESI-QTOF mass spectrometer (SYNAPTTM G2 HDMS, Waters, Manchester, UK) were used to acquire HR-ESIMS data. For electronic circular dichroism (ECD) spectra, a Jasco 810 spectropolarimeter (JASCO Corporation, Tokyo, Japan) was applied with the solvent of acetonitrile. Semi-preparative HPLC was carried out on a Waters 2535 pump equipped with a 2489 UV detector (Waters Corporation, Milford, NY, USA) and an ODS column (YMC, 250 × 10 mm inner diameter, 5 µm). Silica gel (100-200 mesh, Qingdao Mar. Chem. Ind. Co. Ltd., Qingdao, China) and ODS silica gel (A-HG 12 nm, 50 mm, YMC Co. Ltd., Kyoto, Japan) were used for column chromatography.

Plant Material
The seeds of two mangrove plants, Xylocarpus granatum and Xylocarpus moluccensis,