The genus Hypericum (Guttiferae) comprises about 460 species globally distributed in temperate and subtropical regions [1]. There are 64 species in China, and some of them have antibacterial, anti-inflammatory, and hemostatic efficacy [2]. The Hypericum plants have triggered interests of natural product chemists owing to an array of meroterpenoids they generated. These metabolites not only have diverse and complex structures derived from polyprenylated acylphloroglucinol core with prenyl, geranyl, and acyl groups, but also possess intriguing bioactivities including antidepressant, neuroprotective, anti-inflammatory, antitumor, antibacterial, and antiviral activities [3-6].

The plant Hypericum patulum is a shrub widespread in China and can be used as a folk medicine for the treatment of gonorrhea, hepatitis, cold, tonsillitis, and bruises [2]. Some novel prenylated acylphloroglucinol meroterpenoids from this plant have been reported previously [7-9]. As a continuing exploration to seek structurally diverse and biologically interesting meroterpenoids from higher plants [10-12], phytochemical investigation on the aerial parts of H. patulum afforded two novel tetracyclic meroterpenoids, hyperinoids A (1) and B (2) (Fig. 1). Both structures with absolute configurations were elucidated by extensive analysis of NMR, MS, and X-ray crystallographic data. Compounds 1 and 2 possess two unprecedented tetracyclic systems of 11, 12-dioxatetracyclo[5.4.3.0^{1, 7}.0^{4, 14}]tetradecane and 10, 11-dioxatetracyclo[5.3.3.0^{1, 7}.0^{4, 13}]tridecane, respectively.

Nuclear factor-kB (NF-kB) is a key regulator of inflammation, and activation of NF-kB promotes inflammation-associated metabolic disorders such as obesity, type 2 diabetes, and atherosclerosis [13]. Macrophages, as innate immune cells, have been recognized as essential effector cells in the initiation and development of inflammation and insulin resistance [14]. For the discovery of bioactive natural products against inflammation-associated metabolic diseases, compounds 1 and 2 were investi-gated for the inhibitory activities in NF-kB pathway luciferase assay and the effects on the LPS-induced inflammatory responses in macrophages. In this paper, the structural identification and bioactivity evaluation of 1 and 2 are discussed.

Hyperinoid A (1) was isolated as colorless crystals with a specific rotation of + 54.0. Its molecular formula was determined to be C₃₂H₄₂O₆ by the HR-MS (ESI-TOF) ion at m/z 523.3057 [M+H]⁺ (calcd. for C₃₂H₄₃O₆: 523.3054), indicating 12 indices of hydrogen deficiency. The IR spectrum of 1 showed absorption bands for OH (3442 cm^-1), carbonyl (1780 and 1737 cm^-1), and aromatic (1654 and 1465 cm^-1) functionalities. The¹H NMR spectrum displayed characteristic signals for a monosubstituted benzene ring (δ_H 7.88, H-18/H-22; δ_H 7.51, H-20; δ_H 7.42, H-19/H-21), two olefinic protons (δ_H 4.78, H-24; δ_H 5.12, H-29), and seven singlet methyls (δ_H 1.15, H₃-10; 1.09, H₃-14; 1.05, H₃-15; 1.34, H₃-26; 1.18, H₃-27; 1.73, H₃-31; 1.66, H₃-32) (Table 1). The¹³C NMR, DEPT, and HSQC spectra revealed 32 carbon resonances (Table 1), assignable to a benzoyl moiety, an ester carbonyl (δ_C 174.0, C-2), one ketal carbon (δ_H 109.1, C-9), two prenyl (3-methylbut-2-enyl) groups, three methyls, two methylenes, four methines (one oxygenated at δ_C 70.9, C-12), two oxygenated tertiary carbons (δ_C 91.2, C-3; δ_C 79.0, C-13), and two quaternary carbons. These data accounted for eight indices of hydrogen deficiency, and the remaining ones required the existence of four additional rings in 1. Therefore, 1 was speculated to have a tetracyclic system decorated with a benzoyl and two prenyl groups.

The planar structure of 1 was established by extensive analysis of its¹H-¹H COSY and HMBC spectra (Fig. 2A). The¹H-¹H COSY correlations of H-1/H-8 and H-5/H-6 and the HMBC cross-peaks of H₃-10/C-3, C-6, C-8; H₂-5/C-3, C-4, C-9; and H-1/C-9 established a cyclopentane (ring A) and a cyclohexane (ring B). The¹H-¹H COSY correlations of H-11/H-12 and the HMBC networks of H-11/C-4, C-5, C-9, C-13; H-12/C-13; and H₃-14, 15/C-12 defined a 3-hydroxy- 2, 2-dimethyltetrahydropyran ring (ring C). The remaining ring (ring D) was formed by a lactone bridge between C-3 and C-9, as suggested by the diagnostic chemical shifts of C-2, C-3, and C-9. Furthermore, the HMBC correlations of H-29/C-6 and H-24/C-8 and the¹H-¹H COSY correlations of H-6/H₂-28/H-29 and H-8/H₂-23/H- 24 showed that the two prenyl groups were connected to C-6 and C-8, respectively. The benzoyl group was connected to C-1, as revealed by the HMBC correlations of H-1/C-16 and H-18, H-22/C- 16. Thus, the gross structure of 1 was deduced as a meroterpenoid with an unprecedented 11, 12-dioxatetracyclo[5.4.3.0^{1, 7}.0^{4, 14}]tetra- decane ring system.

The coupling constant of H-1/H-8 (J =11.4 Hz) and the H-1/H- 23b ROESY cross-peak (Fig. 2B) indicated that H-1 and H-8 were axially trans-oriented, and they were then assigned an α-orientation and a β-orientation, respectively. The H₃-10/H-8 ROESY correlation suggested an equatorial β-orientation for H₃-10. The H-1/H-28b, H-1/H-5α, and H-5α/H-12 ROESY correlations revealed that the prenyl group at C-6 and H-12 were α-oriented. The relative configurations at C-3, C-4, and C-9 could not be elucidated by ROESY experiment. But owing to the restriction by the rigid tetradecane ring system, they could be assigned as 3R*, 4S*, and 9R*. A single-crystal X-ray diffraction experiment for 1 was conducted by Ga Kα radiation with an absolute structure parameter of 0.04(6) (Fig. 3). Thus, the absolute configuration of 1 was established as 1S, 3R, 4S, 6S, 7R, 8R, 9R, and 12S.

Hyperinoid B (2) was obtained as colorless crystals and had optical activity (+ 61.0). The molecular formula C₃₂H₄₂O₆ was established by the HR-MS (ESI-TOF) ion at m/z 523.3056 [M+H]⁺ (calcd. for C₃₂H₄₃O₆: 523.3054). Interpretation of its NMR data (Table 1) including¹H-¹H COSY and HMBC spectra (Fig. 4A) indicated that 2 was also a tetracyclic meroterpenoid with the same rings A, B, and D moieties as 1. The only difference between them was the ring C. Compound 2 contained a 2-(1-hydroxy-1-methylethyl)tetrahydrofuran ring rather than the 3-hydroxy-2, 2-dimethyltetrahydropyran moiety in 1, as supported by the obviously downfield-shifted C-12 (δ_C 91.6), the oxygenated C-13 (δ_C 70.2), and the¹H-¹H COSY and HMBC correlations shown in Fig. 4A.

The relative configurations at C-1, C-6, C-7, and C-8 were consistent with those of 1, as verified by the coupling constant of H-1/H-8 (J =10.8 Hz) and the ROESY correlations of H-1/H-23b, H₃-10/H-8, and H-1/H-28b (Fig. 4B). Similar to 1, the 3R*, 4S*, and 9R* configurations were also deduced by the rigidity of the skeleton. The configuration at C-12 could not be defined from the available data. Fortunately, the absolute configuration of 2 was finally assigned as 1S, 3R, 4S, 6S, 7R, 8R, 9R, and 12R by an X-ray crystallographic study (Ga Kα radiation) with an absolute structure parameter of 0.02(9) (Fig. 5). Compound 2 is the first meroterpe-noid possessing a novel 10, 11-dioxatetracyclo[5.3.3.0^{1, 7}.0^{4, 13}]tride-cane ring system.

Compounds 1 and 2 were proposed to biogenetically originate from the prenylated acylphloroglucinol precursor i (Scheme 1). It would undergo intramolecular cyclization to form ii, which after oxidative ring cleavage would give iii. Through decarboxylation, intermediate iv would be formed [7]. After epoxidation and intramolecular nucleophilic addition, the hemiketal intermediate v would be produced. Then, 1 and 2 would be formed via two different ring opening pathways of epoxide.

In order to discover natural products against inflammation-associated metabolic diseases, compounds 1 and 2 were evaluated for their anti-inflammatory effects in vitro. Compounds 1 and 2 exhibited significant inhibitory activities in NF-kB pathway luciferase assay with IC₅₀ values of 0.75 0.17 and 1.19 0.48 μmol/L, respectively. In this test, bortezomib (PS-341) was used as the positive control (IC₅₀ = 0.07 0.01 μmol/L). They were further evaluated for influences on the LPS-induced inflammatory responses in RAW 246.7 macrophages and primary mouse BMDM cells. The MTS assay showed no obvious cytotoxicity against RAW 246.7 cells at the tested concentrations (Fig. S1 in Supporting information). The mRNA levels of some pro-inflamma-tory genes, such as IL-1β, IL-6, and iNOS, were downregulated by 1 and 2 (Fig. 6).

In summary, hyperinoids A (1) and B (2), two meroterpenoids biogenetically related to prenylated acylphloroglucinols, were obtained from H. patulum. The majority of meroterpenoids from the genus Hypericum belong to the bicyclic polyprenylated acylphloroglucinols featuring a bicyclo[3.3.1]nonane-2, 4, 9-trione core, the adamantane-type with tricyclo[3.3.1.1]decane skeleton, and the homoadamantane-type with tricyclo[4.3.1.1] undecane motif [3]. Different from them, compounds 1 and 2 possess two unprecedented tetracyclic systems of 11, 12-dioxate-tracyclo[5.4.3.0^{1, 7}.0^{4, 14}]tetradecane and 10, 11-dioxatetracyclo [5.3.3.0^{1, 7}.04,¹³]tridecane, respectively, derived from a rare bicyclo [3.2.1]octane framework [7]. Both of them showed significant anti-inflammatory effects in vitro. They are potent NF-kB pathway inhibitors and can effectively decrease the inflammatory response of macrophages. The novel structures and significant bioactivities of 1 and 2 may provide useful reference for discovering and designing drug leads against inflammation-related metabolic diseases.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

This research was financially supported by Fudan-SIMM Joint Research Fund (No. FU-SIMM20181011).

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.cclet.2019.10.014.