Structural elucidation, antioxidant and hepatoprotective activities of chemical composition from Jinsi Huangju (Chrysanthemum morifolium) flowers

2.1 General experimental procedures

The high-resolution-mass-spectrometry (HR-ESI-MS) data were acquired using a UPLC/xevo G2 QToF spectrometer (Waters Corporation, Milford, MA, USA). The optical rotations were measured on a Perking-Elmer 341-MC digital polarimeter at room temperature (PerkinElmer, Waltham, MA, USA). The infrared (IR) data were acquired utilizing a Hitachi 260–30 spectrometer (Hitachi High-Technologies Co., Japan). The electronic circular dichroism (ECD) data were determined on a Jasco J-810 Circular Dichroism spectropolarimeter at room temperature (Jasco, Mary's Court Easton, MD, USA). The nuclear magnetic resonance (NMR) data were obtained on a Bruker ARX-600 spectrometer (Bruker Technology Co., Ltd., Karlsruhe, Germany), using solvent peaks as references. Compounds were purified on an Agilent 1100 liquid chromatography (Agilent Technologies, Santa Clara, CA, USA) semi-preparative HPLC, equipped with an Agilent 1100 liquid chromatography (Agilent Technologies, Santa Clara, CA, USA) with an Agilent C18 (10 mm × 250 mm) column. Silica gel (80–100 mesh and 300–400 mesh) was got from Qingdao Haiyang Chemical Co. (Shandong, China). ABTS was purchased from Beijing Solarbio Science and Technology Co., Ltd (Beijing, China). HepG2 human liver cancer cell was got from American Type Culture Collection (ATCC) (USA). Three microbial strains, Staphylococcus aureus (ATCC 6538), Escherichia coli (ATCC 25922) and Pseudomonas aeruginosa (ATCC 27853) were got from ATCC (USA).

2.2 Plant material

The flowers (9.9 kg) of “Jinsi Huangju” (Chrysanthemum morifolium Ramat.) were collected from Ningyuan county, Hunan province, China in November 2018. The plant was identified by Prof. Wei Wang and the voucher sample (the flowers, no. JSHJ2018-11) was deposited in the School of Pharmacy, Hunan University of Chinese Medicine.

2.3 Extraction and isolation

Dried flowers of “Jinsi Huangju” (9.9 kg) were immersed in 95 % ethanol (2 × 50L) at room temperature (each time 5 days) to give a crude extract. The ethanolic extract was suspended in water and successively partitioned with petroleum ether (PE), ethyl acetate (EA) and n-butanol. The PE layer (693.76 g) was subjected to silica gel column chromatography (CC) and eluted with PE-EA (100/1 to 0/1) to get seven fractions (A-G). Fraction F (134.04 g) was separated by a silica gel CC was eluted with a gradient PE/EA solvent system (100:1 to 0:1) to afford seven sub-fractions (PF1-PF7). PF2 (2.13 g) was further chromatographed on a silica gel column (PE-EA, 30:1 to 0:1) to get compound 16 (5.2 mg).

The EA extract (413.03 g) was then subjected to silica gel CC and eluted with CH₂Cl₂-MeOH (50:1 to 0:1) solvent system to get eleven fractions (EF1-EF11). Fraction EF5 (9.82 g) was separated by a YMC column, eluted successively with MeOH/H₂O (60:40 to 100:0) to obtain ten sub-fractions (SF5.1–5.10). SF5.3 (5.2 g) was re-subjected to a YMC column with MeOH-H₂O (10:90 to 100:0) to get further nine subfractions (5.3.1–5.3.9). SF5.3.4 (1.29 g) was then separated by silica gel column using gradient of PE/EA (10:1 → 0:1) and further purified by semi-preparative HPLC (flow rate: 3.0 ml/min) to obtain 33 (3.0 mg, t_R = 13.2 min; CH₃CN/H₂O, 38:62), 34 (9.3 mg, t_R = 20.1 min; CH₃CN/H₂O, 37:63), 35 (1.9 mg, t_R = 15.0 min; CH₃CN/H₂O, 34:66), 28 (2.0 mg, t_R = 16.2 min; CH₃CN/H₂O, 34:66), 24 (13.6 mg, t_R = 14.1 min; CH₃CN/H₂O, 31:69), 20 (13.6 mg, t_R = 15.2 min; CH₃CN/H₂O, 31:69), 22 (90.8 mg, t_R = 17.2 min; CH₃CN/H₂O, 31:69), 21 (20.4 mg, t_R = 20.4 min; CH₃CN/H₂O, 31:69), 23 (7.0 mg, t_R = 15.6 min; CH₃CN/H₂O, 30:70), 38 (2.5 mg, t_R = 17.5 min; CH₃CN/H₂O, 30:70), and 25 (3.1 mg, t_R = 21.2 min; CH₃CN/H₂O, 25:75). SF5.4 (1.80 g) was loaded on a YMC column (MeOH-H₂O, 50:50 to 100:0) to obtain six sub-fractions (SF5.4.1–5.4.6). SF5.4.3 (160.3 mg) was loaded on Sephadex LH-20CC (CH₂Cl₂/MeOH, 1:1) and further purified by semi-preparative HPLC (2.0 ml/min; CH₃CN/H₂O, 58:42) to give compounds 19 (1.0 mg, t_R = 13.2 min), 17 (7.1 mg, t_R = 17.5 min), 18 (1.5 mg, t_R = 18.4 min), and 22 (1.7 mg, t_R = 13.1 min; CH₃CN/H₂O, 55:45; 2.0 ml/min). EF7 (30.46 g) was chromatographed on YMC column (MeOH-H₂O, 0:100 to 100:0) to afford eight sub-fractions (EF7.1–7.8). EF7.2 (2.90 g) was applied on a Sephadex LH-20 column (100 % MeOH) and further separated by semi-preparative HPLC (3.0 ml/min; CH₃CN/H₂O, 20:80) to get 42 (4.4 mg, t_R = 7.7 min) and 43 (4.1 mg, t_R = 10.8 min).

Fraction EF7.3 (8.15 g) was subjected to silica gel CC (PE/EA, 5:1 → 0:1) and then further purified by a semipreparative HPLC (3.0 ml/min) to obtain compounds 13 (5.1 mg, t_R = 7.9 min; CH₃CN/H₂O, 32:68), 31 (4.2 mg, t_R = 9.3 min; CH₃CN/H₂O, 30:70), 32 (4.1 mg, t_R = 9.9 min; CH₃CN/H₂O, 28:72), 14 (2.1 mg, t_R = 13.9 min; CH₃CN/H₂O, 21:79), 39 (28.9 mg, t_R = 15.0 min; CH₃CN/H₂O, 21:79), 40 (12.1 mg, t_R = 16.0 min; CH₃CN/H₂O, 21:79), 41 (3.0 mg, t_R = 18.2 min; CH₃CN/H₂O, 21:79), 37 (7.0 mg, t_R = 22.7 min; CH₃CN/H₂O, 21:79), 36 (4.0 mg, t_R = 22.7 min; CH₃CN/H₂O, 21:79), 29 (11.7 mg, t_R = 13.6 min; CH₃CN/H₂O, 20:80), and 30 (20.4 mg, t_R = 13.5 min; CH₃CN/H₂O, 18:82).

Fraction EF7.4 (3.65 g) was separated by silica gel CC (PE/EA, 5:1 → 0:1) and then subjectedto a semi-preparative HPLC (3.0 ml/min) to obtain 1 (8.3 mg, t_R = 15.1 min; CH₃CN/H₂O, 32:68), 8 (10.4 mg, t_R = 21.2 min; CH₃CN/H₂O, 29:71), 7 (20.8 mg, t_R = 26.4 min; CH₃CN/H₂O, 30:70), 44 (10.0 mg, t_R = 26.4 min; CH₃CN/H₂O, 30:70), and 6 (20.8 mg, t_R = 27.8 min; CH₃CN/H₂O, 30:70).

Fraction 7.5 (1.73 g) was subjected on a silica gel CC (PE/EA, 5:1 → 0:1) to afford compound 12 (4.5 mg) and sub-fractions 7.5.1 to 7.5.7. Fr. 7.5.3 (430.4 mg) was separated by Sephadex LH-20 column (100 % MeOH) and then processed on semi-preparative HPLC (3.0 ml/min) to obtain 2 (5.1 mg, t_R = 22.3 min; CH₃CN/H₂O, 28:72), 3 (3.1 mg, t_R = 24.1 min; CH₃CN/H₂O, 28:72), 4 (3.0 mg, t_R = 30.6 min; CH₃CN/H₂O, 30:70), 5 (119.7 mg, t_R = 22.6 min; CH₃CN/H₂O, 32:68), 9 (4.8 mg, t_R = 23.1 min; CH₃CN/H₂O, 29:71), 10 (10.4 mg, t_R = 28.7 min; CH₃CN/H₂O, 32:68), 11 (3.1 mg, t_R = 22.4 min; CH₃CN/H₂O, 30:70), 26 (20.1 mg, t_R = 18.1 min; CH₃CN/H₂O, 28:72), 27 (18.6 mg, t_R = 18.1 min; CH₃CN/H₂O, 28:72). Fraction 11 (28.32 g) was separated on a YMC column (MeOH/H₂O, 0:100 → 100:0) to give compound 15 (27.5 mg).

2.4 ABTS radical scavenging assay

Antioxidant capacity of the samples was evaluated through the ABTS assay was evaluated as previously published literature (Yang et al., 2016). The ABTS radical scavenging ability (RSA) of the compounds was determined as follows: $$RSA\left(\%\right)=\left[A_{\mathit{blank}}-\left(A_{\mathit{sample}}-A_{\mathit{control}}\right)\right]/A_{\mathit{blank}}\times 100\%$$

_l-Ascorbic acid (V_C) was used as a reference standard. All results show the average values of experiments performed in triplicate.

2.5 Hepatoprotective activity

HepG2 cells were used to evaluate the hepatoprotective capacity of the compounds of JH, employing the method performed as described previously (Hao et al., 2012). The cells were cultured in 96-well plates (1 × 10⁵ cells/well) overnight. 10 μM test samples, 10 μM bicyclol (positive control), and 6 mM APAP were added to the wells and incubated for another 48 h. The cell effect was evaluated with the help of 3-(4,5)-dimethylthiahiazo(-z-y1)-3,5-diphenytetrazoliumromide (MTT) assay (Hao et al., 2012). The optical density (OD) of the formazan solution was assessed using a microplate reader at 490 nm. Bicyclol was used as a reference standard.

2.6 Antibacterial activity evaluation

The extracts and compounds were tested against one gram-positive bacteria (S. aureus ATCC 6538) and two gram-negative bacteria (E. coli ATCC 25,922 and P. aeruginosa ATCC 27853) by a twofold serial dilution method in a 96-well microtiter plate, as reported previously (Jiang et al., 2019; Wei et al., 2019). Oxacillin was used as a reference standard. The minimum inhibitory concentration (MIC) values were calculated.

2.7 Network pharmacology analysis

2.8 Molecular docking study

The crystal structure of vascular endothelial growth factor A (VEGFA) (PDB ID: 1MKK) was downloaded from the RCSB Protein Data Bank (Gong et al., 2021). The 3D structures of compounds 1 and 13 were drawn by Chem3D program. Docking was done using AutoDockTools (version 1.5.6) and structure visualization was analyzed by Pymol (v. 2.3.2) and Discovery Studio client 2020.

2.9 Statistical analysis

The tests were performed in triplicate, and the results shown as mean ± standard deviation (SD). The data was calculated using Graph prism 8.0 software with analysis of variance; at the same time, p < 0.05 and p < 0.01 was identified statistically significant.

3.1 Structure elucidation

Chromatographic separation of the PE and EtOAc layers of JH flowers resulted in the isolation of six new compounds (1–4, 11 and 31) along with thirty-eight known compounds (5–10, 12–30 and 32–44) (Fig. 1).

Close

Fig. 1

The structures of compounds 1–44.

Export to PPT

Huangjusu A (1), was yielded as a light brown solid. Its molecular formula was established to be C₂₂H₂₄O₉ from the (+)-HR-ESI-MS analysis, showing a signal at m/z 455.1314 [M + Na]⁺ and 887.2770 [2 M + Na]⁺ (calcd for C₂₂H₂₄O₉Na⁺, 455.1318), indicating 11 degrees of unsaturation. The ¹H NMR spectra of 1 displayed the presence of a benzyl moiety [δ_H 7.44 (2H, d, J = 7.8 Hz), 7.34 (2H, t, J = 7.8 Hz),7.28 (1H, t, J = 7.8 Hz) and δ_H 4.96 (1H, d, J = 12.0 Hz), 4.70 (1H, d, J = 12.0 Hz)] and an E-caffeoyl moiety [δ_H 7.05 (1H, d, J = 1.8 Hz), 6.96 (1H, dd, J = 8.4, 1.8 Hz), 6.78.

(1H, d, J = 8.4 Hz), δ_H 7.60 (1H, d, J = 15.6 Hz) and 6.30 (1H, d, J = 15.6 Hz)]. Moreover, the ¹H NMR data of 1 displayed resonance signals for a glucosyl moiety [δ_H 4.86 (1H, overlapped), 4.43 (1H, d, J = 7.8 Hz), 3.62 (1H, m), 3.51 (1H, m), 3.37 (1H, m) and 3.64 (1H, m), 3.58 (1H, m)]. The ¹³C NMR spectra showed 22 carbon signals, including a carbonyl carbon (δ_c 168.6), 12 aromatic carbons (2 benzene rings), 2 olefinic carbons, 5 oxygenated methine carbons (δ_C 103.3, 76.2, 75.8, 75.3 and 72.6), and 2 oxygenated methylene carbons (δ_C 71.8 and 62.6). The 1D NMR spectrum (Table 1 and 2) inferred that the structure of 1 was similar to 1-O-benzyl-6-O-E- caffeoyl-β-d-glucopyranoside (Janggyoo et al., 2015), except for the E-caffeoyl group in place of glucosyl group. The result was confirmed by the ¹H–¹H COSY connectivity of H-1′ (δ_H 4.86)/H-2′ (δ_H 3.37)/H-3′ (δ_H 3.62)/H-4′ (δ_H 4.86)/H-5′ (δ_H 3.51)/H-6′ (δ_H 3.64, 3.65) and the HMBC correlations from H-4′ and H-7′' (δ_H 7.60) to C-9′' (δ_C 168.6) (Fig. 2). Furthermore, the sugar group was identified as d-glucoside (t_R = 17.19 min) by acid hydrolysis and HPLC analysis (Fig. S1). Hence, the structure of 1 was confirmed to be 1′-O-benzyl-4′-O-E-caffeoyl-β-d-glucopyranoside and named huangjusu A.

Close

Fig. 2

The key ¹H−¹H COSY and HMBC correlations of compounds 1–4, 11 and 31.

Export to PPT

Compound 2, a light brown solid, possessed the molecular formula C₂₃H₂₆O₉ according to its (+)-HR-ESI-MS (m/z: 469.1481 [M + Na]⁺ and 915.3050 [2 M + Na]⁺, calcd. for C₂₃H₂₆O₉Na, 469.1475). The 1D NMR spectra of 2 showed high similarities to that of compound 1. An additional methylene group was elucidated to be present between C-1 and C-7. This was confirmed by the HMBC correlations of H-7 (δ_H 2.96) with C-2,6 (δ_C 129.4), C-1 (δ_C 140.0) and C-8 (δ_C 71.8). The molecular formula of 2 further confirmed that it had an extra methylene moiety. Hence, the structure of 2 was confirmed to be 1′-O-phenethyl-4′-O-E-caffeoyl-β-d-glucopyranoside, named as huangjusu B.

Compound 3, a light yellow solid, possessed the molecular formula C₂₁H₂₈O₉ according to its (+)-HR-ESI-MS (m/z: 447.1636 [M + Na]⁺, calcd. for C₂₁H₂₈O₉Na, 447.1631). The 1D NMR spectra of 3 exhibited similarities to that of compound 1, except for the benzyl moiety replaced by cis-3-hexenyl moiety at C-1′. This was confirmed by the 1D NMR and 2D NMR data. The 1D NMR spectra displayed signals for the cis-3-hexenyl moiety [δ_H 3.90 (H-1a), δ_H 3.58 (H-1b)/δ_C 70.6 (C-1), δ_H 2.40 (H-2)/δ_C 28.8 (C-2), δ_H 5.40 (1H, dt, J = 10.8, 1.8 Hz, H-3)/δ_C 125.8 (C-3), δ_H 5.48 (1H, dt, J = 10.8, 1.8 Hz, H-4)/δ_C 134.6 (C-4), δ_H 2.08 (H-5)/δ_C 21.5 (C-5) and δ_H 0.98 (H-6)/δ_C 14.6 (C-6)]. Furthermore, this was further confirmed by the HMBC correlations of H-1′ (δ_H 4.35) with C-1 (δ_C 70.6). Hence, the structure of 3 was confirmed and named as huangjusu C.

Compound 4 was obtained as a light yellow solid. Its molecular formula was assigned to be C₂₇H₂₈O₁₂ from its (+)-HR-ESI-MS data (m/z 545.1660 [M + H]⁺, calcd for C₂₇H₂₉O₁₂⁺, 545.1659). The 1D NMR data of 4 was similar to ethyl-3,5-di-O-caffeoylquinate (5) (Wang et al., 1992), the only difference being the replacement of the trans-olefinic protons with a large coupling constant [δ_H 7.55 (1H, d, J = 15.9 Hz, H-7′), 6.23 (1H, d, J = 15.9 Hz, H-8′)] by cis-olefinic protons with a smaller coupling constant [δ_H 6.81 (1H, d, J = 12.6 Hz, H-7′), 6.72 (1H, d, J = 12.6 Hz, H-8′)]. Thus, the structure of 4 was confirmed and named as huangjusu D.

Compound 11 was purified as a pale-yellow powder. Its molecular formula was assigned to be C₂₅H₂₆O₁₂ from its (+)-HR-ESI-MS data (m/z 541.1328 [M + Na]⁺ and 1059.2746 [2 M + Na]⁺, calcd for C₂₅H₂₆O₁₂Na⁺, 541.1322), indicating 13 degrees of unsaturation. The ¹H NMR spectra of 11 exhibited typical 5,7-substituted ring A resonances of flavonoids at δ_H 6.23 (1H, m), 6.20 (1H, m); 1′,3′,4′-substituted ring B resonances at δ_H 6.92 (1H, m), 6.89 (2H, m); an oxygenated methine [δ_H 5.34 (1H, m)], and a methylene [δ_H 3.13 (1H, d, J = 16.8, 12.6 Hz) and 2.76 (1H, d, J = 16.8, 8.0 Hz] of the ring C of flavonone. Furthermore, the ¹H NMR data of 11 also illustrated a sugar moiety [δ_H 5.04 (1H, m), 4.89 (1H, m), 3.69 (1H, m), 3.69 (1H, m), 3.53 (1H, m), and 3.61 (1H, m), 3.51 (1H, m)], and a 2-E-butenoate moiety [δ_H 7.06 (1H, m), 5.92 (1H, d, J = 15.6 Hz) and 1.91 (1H, d, J = 6.6 Hz)]. According to the above data, compound 11 revealed existence of a flavonone unit, a glucopyranosyl unit and a 2-E-butenoate group. Their linkages were established from the 2D NMR data. HMBC experiment revealed that H-1′' (δ_H 5.04), H-8 (δ_H 6.23) and H-6 (δ_H 6.20) all have correlations to C-7 (δ_C 166.9). This confirmed the attachment of the β-glucose unit with flavonone at C-7. Other key HMBC relationships from H-4′' (δ_H 4.89), and H-3′'' (δ_H 7.06) to C-1′'' (δ_C 167.3) deduced that the 2-E-butenoate moiety would be attached to C-4′' of the β-glucose moiety. Moreover, on the basis of the negative cotton effect at 290 nm and positive cotton effect at 334 nm in the experimental ECD spectrum, the absolute configuration at C-2 of the flavonone aglycone was determined to be S (Slade et al., 2005). Therefore, the structure of 11 was ascertained as (+)-eriodictyol 7-O-(4′'-O-2-E-butenoate)-β-glucopyranoside and named huangjusu E.

Compound 31, a colorless oil, possessed the molecular formula C₁₂H₁₆O₃ concluded by (+)-HR-ESI-MS (m/z: 231.0989 [M + Na]⁺, calcd for C₁₂H₁₆O₃Na, 231.0997). The 1D NMR data of 31 was similar to cis-jasmone (Birkett et al., 2000), except for the methyl ester moiety at C-11 in 31 as compared to methyl group in cis-jasmone. This outcome was confirmed by the HMBC correlations of H-10 (δ_H 3.25) and H-1′ (δ_H 3.68) with C-11 (δ_C 174.0) and of H-10 with C-9 (δ_C 130.2) and C-8 (δ_C 123.1) and the molecular weight of the compound. Thus, the structure of 31 was ascertained and named huangjusu F.

In addition, the structures of 38 known compounds were determined as ethyl 3,5-di-O-caffeoylquinate (5) (Wang et al., 1992), ethyl 3,4-di-O-caffeoylquinate (6) (Wang et al., 1992), ethyl 4,5-di-O-caffeoylquinate (7) (Guan et al., 2014), methyl 3,5-di-O-caffeoylquinate (8) (Liu et al., 2010), methyl 3,4-di-O-caffeoylquinate (9) (Liu et al., 2010), 3,5-di-O-caffeoylqunic acid (10) (Wang et al., 1992), (+)-eriodictyol (12) (Huvaere et al., 2012), (2R,3R)-taxifolin (13) (Huvaere et al., 2012), apigenin (14) (Wei et al., 2019), apigenin-7-O-β-d-glucopyranoside (15) (Wei et al., 2019), artemetin (16) (Hussain et al., 2009), jaceidin (17) (Hussain et al., 2009), jaceosidin (18) (Martínez et al., 1987), 6-methoxytricin (19) (Martínez et al., 1987), (3S,5R,6S,7E)-3,5,6-trihydroxy-7-megastigmen-9-one (20) (Sun et al., 2007), pubinernoid A (21) (Wan et al., 2017), isololiolide (22) (Sun et al., 2007), (1R)-3-oxo-4,6,6-trimethylcyclohex-4-ene-l-carboxylic acid (23) (Marco et al., 1991), crocusatin C (24) (Wook et al., 2016), (+)-trans-sobrerol (25) (Pellegata et al., 1987), (-)-cis-chrysanthenol O-β-d-glucopyranoside (26) (Miyakado et al., 1974), (-)-borneol 2-O-β-d-glucopyranoside (27) (Orihara and Furuya, 1993), 4-hydroxypiperitenone (28) (Thulasiram and Madyastha, 2005), geniposide (29) (Ono et al., 2005), (Z)-5′-hydroxyjasmone 5′-O-β-d-glucopyranoside (30) (Ding et al., 2009), tuberonic acid (32) (He et al., 2019), 4-propionylphenol (33) (Ma et al., 2021), 5,7-dihydroxy-4H-4-chromenone (34) (Kwon et al., 2016), umbelliferone (35) (Ahn et al., 2021), aeculetin (36) (Ahn et al., 2021), caffeic acid (37) (Ahn et al., 2021), 1H-​indole-​2-​carboxylic acid (38) (Silva et al., 2021), pinoresinol-4-O-glucoside (39) (Deyama et al., 1986), eucommin A (40) (Deyama et al., 1986), (+)-isoeucommin A (41) (Deyama et al., 1986), protocatechuic acid (42) (Ahn et al., 2021), p-hydroxybenzoic acid (43) (Ahn et al., 2021), benzyl-6-[(2E)-2-butenoate]-β-d-glucopyranoside (44) (Qi et al., 2011) by comparison of their 1D NMR data with those described in the literature (showed in Supplement data).

3.2 Antioxidant activity

Some of the isolated compounds were tested for their antioxidant potential by employing the ABTS radical cation scavenging assay according to the procedures as previously described. As shown in Table 3, the new compounds huangjusus A-E (1–4 and 11) exhibited strong scavenging effect with IC₅₀ values of 11.34 μM, 15.24 μM and 10.32 μM, 6.38 μM, and 10.23 μM, respectively. Eleven known compounds (5–9, 12, 13, 17, 39, 40, and 42) showed moderate to potent antioxidant effects with IC₅₀ values of 4.22–56.33 μM.

Structure and antioxidant activity relationship study (SARs) showed that the introduction of two adjacent phenolic hydroxyl moieties could obviously enhance the activity (comparing the activity of 42 and 43). Neither the alcoholic hydroxyl group nor a single phenolic hydroxyl group showed potent activity.

3.3 Hepatoprotective activity

The compounds were tested for hepatoprotective activity using MTT assay. As shown in Fig. 3, most of these compounds (except 8, 10, and 14) improved the survival rate of APAP-triggered toxicity in HepG2 cells compared with the model group. Among them, ten compounds (1, 11, 13, 19, 21, 29, 30, 31, 39, and 41) exhibited potent hepatoprotective activities against APAP-induced toxicity in HepG2 cells with the cell survival rates of 61.53 %, 63.55 %, 63.05 %, 59.75 %, 59.15 %, 61.07 %, 62.72 %, 60.01 %, 58.86 %, and 58.76 % (positive control bicyclol, 58.41 %) at 10 μM, respectively.

Close

Fig. 3

Activities of some compounds on the survival rates of APAP-induced HepG2 cells. Data are presented as the mean ± SD (n = 3). Values with the “^##”, “**” and “*” are significantly different (^## p < 0.01 vs blank; ** p < 0.01, * p < 0.05 vs model).

Export to PPT

3.4 Network pharmacology analysis

3.4.1

3.4.1 Screening of the common targets

According to database analysis, 65 targets of JH-1 ∼ 9 and 1234 targets of non-alcoholic liver injury were obtained. Duplicate results were removed. The Venn diagram was produced using the online Venn analysis tool, and a total of 52 common targets were found to be the potential common targets of JH-1 ∼ 9 and non-alcoholic liver injury (Fig. 4A and 4B).

Close

Fig. 4

Network pharmacology analysis of JH-1 ∼ 9 and non-alcoholic liver injury. (A) Chrysanthemum-component-target network diagram. (B) Venn diagram of the com-mon targets of JH-1 ∼ 9 and non-alcoholic liver injury. (C) GO enrichment analysis of the predicted targets in Biological Process, Molecular Function and Cell Composition. (D) Bubble plot of the potential targets obtained from KEGG enrichment analysis. (E) Protein-protein target interaction network of the common targets. (F) The core target interaction diagram.

Export to PPT

3.5 Molecular docking

According to the results of hepatoprotective activity and the potential core targets obtained from network pharmacology analysis, molecular docking was utilized to predict the possible binding ability between the evaluated huangjusu A (1) and 13 and the VEGFA (Gong et al., 2021). The computing results of molecular docking are presented in Fig. 5. Huangjusu A (1) formed hydrogen-bonding interaction with Leu-32, Leu-66, Thr-31, and Asn-62 residues and hydrophobic interactions with Leu-32 and Ile-29 (Fig. 5A). Compound 13 showed hydrogen-bonding interactions with Asp-34, Ala-61, Phe-47, and Ile-43, and the hydrophobic interaction with Asp-34 and Ile-46 (Fig. 5B). These results indicated that the hydroxy and benzene ring play a major role in the docking, moreover, pharmacological mechanisms need to be further explored (Hung et al., 2022).

Close

Fig. 5

Docking picture of compounds 1 (A) and 13 (B).

Export to PPT

3.6 Antibacterial activity

As an attempt to illustrate the antibacterial effect of compounds and crude extract of JH, a radial diffusion assay was firstly applied on a sensitive strain of S. aureus, E. coli, and P. aeruginosa. Compounds and extracts with anti-microbial activity were indicated by the inhibition zones observed on the agar plate.(Wei et al., 2019) To establish the MIC values, the samples were subjected to a series of liquid dilution techniques on a 96-well microtitration plate.

As shown in Supplementary Table S1, the crude extract of JH and some compounds exhibited moderate and weak anti-bacterial effects against all the selected strains. Compound 19 displayed significant inhibition of S. aureus with MIC values of 22.5 μg/mL. Moreover, ethanol and PE extracts exhibited weak activity with MIC values of 250.0 and 500 μg/mL, respectively. Compounds 5, 6, 8, 15, 13, 20, 21, and 35 showed selected activity against P. aeruginosa with MIC values of 34.0–121.0 μg/mL, compared to the positive control oxacillin (IC₅₀ = 100.0 μg/mL). Among these, compounds 5, 6, and 7 are ethyl dicaffeoylquinates which have the same molecular weight, but their antibacterial activities were different. Ethyl 3,5-di-O-caffeoylquinate (5) and ethyl 3,4-di-O-caffeoylquinate (6) exhibited weak to moderate antibacterial activities, but ethyl 4,5-di-O-caffeoylquinate (7) showed no activity, indicating that the caffeoyl group at C-3 of quinine ethyl carbonate is critical for antibacterial activity. Additionally, both ethyl 3,5-di-O-caffeoylquinate (5) and methyl 3,5-di-O-caffeoylquinate (8) showed weak to moderate antibacterial activities, while 3,5-di-O-caffeoylqunic acid (10) had no activity, demonstrating that the carboxyl substituent of quinic acid also has a great influence on the antibacterial activity. No compounds or extracts showed strong inhibitory effects against the E. coli strain.