Introduction

Medicinal plants have been used as folk medicines in Indonesia, and the bioactive metabolites of these traditional medicines have attracted much interest as new drug candidates [1, 2]. The genus Artocarpus is distributed in the tropical and subtropical areas of Asia and used as traditional medicines to treat inflammation, diabetes, malarial fever, abscesses, and diarrhea [3] as well as a food in South-East Asia [4]. The constituents of Artocarpus species are well-studied and a variety of phenolic compounds including flavonoids, arylbenzofurans, and stilbenoids were reported as the major bioactive components [5]. Flavonoids isolated from Artocarpus species show a wide range of biological activities such as cytotoxicity [6, 7], antibacterial and antifungal [8], antimalarial [9], anti-inflammatory [10, 11], and tyrosinase inhibitory activities [9]. Continuing research on the constituents of the traditional medicinal plants is necessary in order to utilize the traditional ethnopharmacological knowledge for the development of new drugs. In this study, we investigated the biologically active constituents of the leaves of A. communis (family Moraceae) collected in Indonesia and obtained four new geranyl flavonoids along with four known congeners.

Results and Discussion

Separation of eight geranyl flavonoids (1–8)

The leaves of A. communis were extracted with EtOH. The extract was concentrated and partitioned between EtOAc and H2O, and then the organic layer was concentrated and partitioned between n-haxane and 90% MeOH–H2O. The concentrated aqueous MeOH fraction was subjected to silica gel column chromatography followed by HPLC to afford eight compounds. The structures of four new geranyl flavonoids were determined to be (S,E)-2-(3,4-dihydroxyphenyl)-8-(3,7-dimethylocta-2,6-dien-1-yl)-5,7-dihydroxychroman-4-one (1), (S)-5,7-dihydroxy-8-((2E,5E)-7-hydroxy-3,7-dimethylocta-2,5-dien-1-yl)-2-(4-hydroxyphenyl)chroman-4-one (2), (2S)-5,7-dihydroxy-8-((E)-6-hydroxy-3,7-dimethylocta-2,7-dien-1-yl)-2-(4-hydroxyphenyl)chroman-4-one (3), and 1-(2,4-dihydroxyphenyl)-3-(8-hydroxy-2-(3-hydroxy-4-methylpent-4-en-1-yl)-2-methyl-2H-chromen-5-yl)propan-1-one (4) (Fig. 1) by detailed analyses of the spectroscopic data. In addition to 1–4, four known flavonoids sophoraflavanone A [12, 13], 2-geranyl-2′,3,4,4′-tetrahydroxydihydrochalcone [14], 1-(2,4-dihydroxyphenyl)-3-[8-hydroxy-2-methyl-2-(4-methyl-3-pentenyl)-2H-1-benzopyranyl]-1-propanone [14], and 1-(2,4-dihydroxyphenyl)-3-[8-hydroxy-2-methyl-2-(4-hydroxy-4-methyl-2-pentenyl)-2H-1-benzopyran-5-yl]-1-propanone (8) [15] were identified (Fig. 1).

Fig. 1
figure 1

Chemical structures of geranyl flavonoids 1–8 from the leaves of A. communis

Full size image

Structure elucidation of four new geranyl flavonoids (1–4)

Compound 1 was obtained as pale yellow solid. The molecular formula of C25H28O6 was confirmed by HRESIMS. Interpretation of 1H and 13C NMR spectra together with HSQC spectrum revealed the presence of a carbonyl carbon at δC 196.6 (C-4), five oxygenated aromatic carbons at δC 161.1 (C-5), 164.4 (C-7), 159.7 (C-8a), 145.2 (C-3′), and 145.5 (C-4′), eleven sp2 carbons at δC 101.8 (C, C-4a), 95.2 (CH, C-6), 107.0 (C, C-8), 129.8 (C, C-1′), 114.2 (CH, C-2′), 115.3 (CH, C-5′), 117.6 (CH, C-6′), 122.4 (CH, C-2″), 133.8 (C, C-3″), 124.1 (CH, C-6″), and 130.6 (C, C-7″), four methylenes at δC 42.1 (C-3), 21.2 (C-1″), 39.3 (C-4″), and 26.2 (C-5″), one methine at δC 78.2 (C-2), and three methyls at δC 25.5 (C-8″), 17.5 (C-9″), and 15.8 (C-10″) (Table 1). The COSY correlations indicated the presence of three spin systems, H-2 (δH 5.33)/H-3 (δH 3.10), H-1″ (δH 3.08)/H-2″ (δH 5.09), and H-4″ (δH 1.86)/H-5″ (δH 1.96)/H-6″ (δH 5.01) (Fig. 2a). The HMBC correlations from two methyls at δH 1.58 (Me-8″) and δH 1.51 (Me-9″) to C-7″ and C-6″, and from a methyl proton at δH 1.55 (H-10″) to C-2″, C-3″, and C-4″, together with a NOE correlation between H-1″ (δH 3.08) and H-10″ assigned the structure of a geranyl moiety (Fig. 2a). The flavanone skeleton was established by the HMBC correlations from an aromatic proton H-6 (δH 5.95) to C-4a, C-5, C-7, and C-8, from hydroxyl proton 5-OH (δH 12.09) to C-4a, C-5, and C-6, from H-2′ to C-2, C-4′, and C-6′, and from H-3 (δH 2.69) to C-4. The geranyl side chain was attached to C-8, which was established by the HMBC correlations from H-1″ to C-7, C-8, and C-8a. The structure of ring B was determined as 3′,4′-dihydroxyphenyl on the basis of the comparison of the 13C NMR chemical shifts reported in the literature [16]. Although the proton signals for H-5′ (δH 6.72) and H-6′ (δH 6.72) were completely overlapped in DMSO-d6, they were slightly differentiated in CD3OD, and the ortho coupling constants (8.3 Hz) of H-5′ (δH 6.78)/H-6′ (δH 6.80) and the meta coupling constants (1.5 Hz) of H-2′ (δH 6.92)/H-6′ confirmed the 3′,4′-dihydroxyphenyl structure. The absolute configuration at C-2 was unambiguously determined to be S by the negative Cotton effect around 290 nm [17] (Fig. 2b). Yang et al. [16] previously pointed out that the assignment of the substitution pattern in the dihydroxy aromatic ring requires careful analyses of 1H and 13C NMR data and revised the structures of 28 compounds with 3,5-dioxygenated aromatic rings. During this study, we found that the reported 3′,5′-dihydroxyphenyl structures of three flavanones, 5,7,3′,5′-tetrahydroxyflavanone [18], schizolaenone C [19], and xeractinol [20], should be corrected to the 3′,4′-dihydroxyphenyl structures (Fig. 3), which were ultimately identified to be (–)-eriodictyol [21], diplacone [22], and taxifolin 6-C-glucoside [23], respectively.

Table 1 13C NMR data for 1–3 (150 MHz)
Full size table
Fig. 2
figure 2

a COSY (bold line), key HMBC (solid arrow), and key NOE (dashed arrow) correlations of 1. b ECD spectrum of 1 in MeOH

Full size image
Fig. 3
figure 3

Proposed (upper) and corrected (lower) structures of three previously reported flavanones

Full size image

The molecular formula of 2 was determined to be C25H28O6. Comparison of the 1H NMR spectra of 1 and 2 indicated their structural similarity (Table 2). The 1H NMR spectrum of 2 showed mutually coupled aromatic protons at δH 7.28 (H-2′ and H-6′) and δH 6.86 (H-3′ and H-5′), indicating the presence of a p-substituted phenyl group. The HMBC correlations from H-2′/H-6′ to a methine carbon at δC 78.6 (C-2) and an oxygenated aromatic carbon at δC 156.3 (C-4′) and from H-3′/H-5′ to an aromatic carbon at δC 130.6 (C-1′) assigned a hydroxyl group substituted at C-4′ in the aromatic ring B of the flavanone. Detailed analysis of the 1D and 2D NMR spectra of 2 indicated the 5,7,4′-trihydroxy flavanone with a modified geranyl side chain bound to C-8. The 1H NMR spectrum showed the presence of an E-olefin (JH-5″,H-6″ 15.0 Hz) at C-5″ (δH 5.54, δC 124.8)/C-6″ (δH 5.60, δC 139.5). The HMBC correlation from H-8″/H-9″ (δH 1.28) to C-6″ and C-7″ (δC 70.9) confirmed the position of a hydroxyl group at C-7″. Finally, the planar structure of 2 was determined and the absolute configuration at C-2 was determined to be S by the negative Cotton effect at 290 nm in ECD spectroscopy (Fig. 4) [17].

Table 2 1H NMR data for 1–3 (600 MHz)
Full size table
Fig. 4
figure 4

ECD spectrum of 2 in MeOH

Full size image

Compound 3 had the molecular formula C25H28O6. The 1H and 13C NMR spectra of 2 and 3 implied a difference in the modified geranyl side chain. HMBC correlations from a methyl at δH 1.69 (Me-9″) to an oxymethine carbon at δC 75.8 (C-6″) and olefin carbons at δC 147.2 (C-7″) and δC 111.1 (C-8″) assigned the planar structure of 3. The 2S configuration of 3 was determined by the ECD spectrum (Fig. 5a) [17]. Detailed analysis of the 13C NMR spectrum of 3 showed two sets of slightly shifted signals at δC 35.82/35.87 (C-4″) and δC 75.76/75.80 (C-6″), indicating that 3 was an epimeric mixture at C-6″. Purification by HPLC with a chiral-phase column furnished two equivalent peaks (Fig. 5b), and thereby 3 was determined to be an epimeric mixture of 6″R- and 6″S-3 in the ratio 1:1.

Fig. 5
figure 5

a ECD spectrum of 3 in MeOH. b HPLC chromatogram of 3 (0.8 mg) with a chiral-phase column

Full size image

The molecular formula of 4 was determined to be C25H28O6 from the HRESIMS. In the 13C NMR spectrum of 4, most carbons showed sets of slightly shifted chemical shifts in the ratio 1:1 (Table 3), and 4 was indicated to be a mixture of stereoisomers. 1D and 2D NMR spectroscopic data showed one carbonyl carbon at δC 203.74/203.78, four oxygenated aromatic carbons at δC 139.40/139.45 (C-3), δC 143.06/143.07 (C-4), δC 165.3 (C-2′), and δC 163.27/163.28 (C-4′), 12 sp2 carbons at δC 127.96/127.98 (C, C-1), δC 118.88/118.93 (CH, C-2), δC 114.69/114.71 (CH, C-5), δC 121.26/121.29 (CH, C-6), δC 113.62/113.64 (C, C-1′), δC 103.6 (CH, C-3′), δC 107.9 (CH, C-5′), δC 132.21/132.25 (CH, C-6′), δC 119.7 (CH, C-1″), δC 129.64/129.71 (CH, C-2″), δC 146.99/147.07 (C, C-7″), and δC 111.44/111.57 (CH2, C-8″), two methyls at δC 29.15/29.29 (C-9″) and δC 26.25/26.41 (C-10″), four methylenes at δC 39.64/39.69 (C-α), 26.58/26.63 (C-β), δC 36.68/36.88 (C-4″), and δC 17.51/17.59 (C-5″), one oxymethine at δC 75.87/75.90 (C-6″), and one quaternary carbon at δC 78.51/78.63 (C-3″). COSY correlations revealed five spin systems H-5 (δH 6.71)/H-6 (δH 6.60), H-α (δH 3.07)/H-β (δH 2.96), H-5′ (δH 6.30)/H-6′ (δH 7.48), H-1″ (δH 6.50)/H-2″ (δH 5.58), and H-5″ (δH 1.61)/H-6″ (δH 4.05) (Fig. 6a). A dihydrochalcone skeleton was confirmed by the HMBC correlations from H-5 to C-1 and C-3, from H-6 to C-2 and C-4, from H-5′ to C-1′ and C-3′, from H-6′ to C-2′, C-4′, and a carbonyl carbon, from H-β to C-1, C-2, and C-6, from H-α to a carbonyl carbon (Fig. 6a). The remaining structure was elucidated from the 1D and 2D NMR spectroscopic data. A coupling constant of 10.0 Hz between olefin protons H-1″ and H-2″ confirmed the Z-olefin configuration. HMBC correlations from H-1″ to C-1, C-3, and C-3″, from H-2″ to C-2 and C-3″, from H-6″ to C-4″, C-5″, C-7″, and C-8″, from Me-10″ (δH 1.36) to C-2″, C-3″, and C-4″, from Me-9″ (δH 1.68) to C-6″, C-7″, and C-8″ established the planar structure of 4. As explained above, the 13C NMR spectrum showed 4 to be a mixture of two stereoisomers. In addition, no Cotton effect in the ECD spectroscopy indicated that 4 had a racemic nature. Taken together, 4 appeared to be a mixture of four stereoisomers. HPLC analysis with a chiral-phase column showed four peaks with nearly equal peak areas. Thus, 4 was confirmed to be a mixture of four stereoisomers at C-3″ and C-6″ (Fig. 6b).

Table 3 1H and 13C NMR data for 4 in CDCl3
Full size table
Fig. 6
figure 6

a COSY (bold line) and key HMBC (arrow) correlations of 4. b HPLC chromatogram of 4 (10 μg) with a chiral-phase column

Full size image

Study on the absolute configurations of four known geranyl flavonoids (5–8)

Four known geranyl flavonoids 5–8 were obtained from A. communis in this study. However, as there were no previous descriptions of the stereochemistries of 7 and 8 [14, 15], we investigated their absolute configurations. The specific rotations of 7 and 8 were nearly zero (7, − 0.7; 8, + 0.4), and no obvious Cotton effects were observed in the ECD spectra. Resolution of 7 by HPLC with a chiral-phase column successfully afforded equivalent amounts of (+)-7 \( ([\alpha ]^{ 20}_{\text{D}} + 40) \) and (–)-7 \( ([\alpha ]^{ 20}_{\text{D}} - 4 5) \), and thus 7 was concluded to be a racemic mixture (Fig. 7a). The ECD spectrum of (+)-7 showed positive Cotton effects around 220 and 275 nm, and (−)-7 showed opposite Cotton effects to (+)-isomer (Fig. 7b). A Cotton effect due to the styrene chromophore is expected to be observed around 270 nm [24, 25]. Thus, the positive Cotton effect around 270 nm of (+)-7 indicated a 3″S configuration. To confirm this result, the ECD spectrum of S-chromenone (9) was calculated and a positive Cotton effect was observed at 270 nm (Fig. 7b). Compound 8 was separated by HPLC with a chiral-phase column to afford (+)- and (–)-isomers in the ratio 1:1 (Fig. 8a). The absolute configurations of (+)-and (−)-isomers were determined to be 3″S and 3″R, respectively, from their ECD spectra (Fig. 8b).

Fig. 7
figure 7

a HPLC chromatogram of 7 (1.2 mg) with a chiral-phase column. b Experimental ECD spectra of (+)- and (–)-7 and calculated ECD spectrum of a model compound S-9

Full size image
Fig. 8
figure 8

a HPLC chromatogram of 8 (1.5 mg) with a chiral-phase column. b Experimental ECD spectra of (+)- and (−)-8

Full size image

Biological activities of 1–8

The biological activities of 1–6, (+)- and (–)-7, and (+)- and (–)-8 were evaluated with our in-house screening including cytotoxicity, antimicrobial activities, inhibitory activity of the cholesterol ester accumulation in macrophages, inhibitory activity of the RANKL-induced formation of multinuclear osteoclasts, and inhibitory activities of the ubiquitin–proteasome system (proteasome, E1, Ubc13 (E2)–Uev1A interaction, p53–Mdm2 (E3) interaction, and USP7). Among them, 1, 4, 5, and 6 inhibited USP7 activity with 76, 92, 42, and 94% inhibition, respectively, at 1 μM, and the IC50 values of 1, 5, and 6 were 0.26, 1.2, and 0.094 μM, respectively. In addition, 1, 5, and 6 inhibited the chymotrypsin-like activity of the proteasome by 77, 24, and 67%, respectively, at 1 μM. This study showed that these geranyl flavonoids inhibited the enzymatic activity of USP7 and the proteasome at relatively low concentrations. Flavonoids are highly reactive and are known to inhibit various enzymatic activities including the proteasome [26, 27]. Further study is needed to reveal the relationship between structures of geranyl flavonoids and inhibitory activities of USP7 and the proteasome.

Conclusion

Four new geranyl flavonoids were obtained from the leaves of A. communis collected in Indonesia along with four known congeners. The chemical structures were elucidated by analyses of NMR and MS spectroscopic data together with the calculated ECD spectra. During the structure elucidation, we found that the reported structures for 5,7,3′,5′-tetrahydroxyflavanone, schizolaenone C, and xeractinol were incorrect and the same as those of (–)-eriodictyol, diplacone, and taxifolin 6-C-glucoside, respectively. The geranyl flavonoids obtained in this study inhibited the activities of USP7 and the chymotrypsin-like activity of the proteasome. Although an inhibitory effect of flavonoids on the proteasome has already been reported, this is the first report of geranyl flavonoids that inhibited USP7 and the proteasome. Among the geranyl flavonoids tested in this experiment, the USP7 inhibitory activity of 6 (IC50 value, 0.094 μM) was 55 times more potent than the commercially available positive control, P5091 (IC50 value, 5.2 μM). Ubiquitin-dependent protein degradation is a promising target for cancer drug discovery, and our study showed the potential of geranyl flavonoids as inhibitors of the ubiquitin–proteasome system.

Experimental

General experimental procedures

Optical rotations were measured on a JASCO DIP-1000 polarimeter in MeOH. UV spectra were measured on a JASCO V-550 spectrophotometer in MeOH. ECD spectra were measured on a JASCO J-820 spectropolarimeter in MeOH. IR spectra were recorded on a Perkin Elmer Frontier FT-IR spectrophotometer. 1H and 13C NMR spectra were recorded on a Bruker Avance III 600 NMR spectrometer. HRESIMS spectra were measured on a Bruker BioTOF and impact II mass spectrometer. The preparative HPLC system comprised a Waters 515 HPLC pump, Waters 2489 UV/visible detector, and Pantos Unicorder U-228. Silica gel-precoated plates (TLC Silica gel 60F254, Merck) were used for TLC detection, and spots were visualized using UV light (254 and/or 365 nm) and p-anisaldehyde reagent.

Plant material

Leaves of A. communis (15P099, fresh weight 169 g) were collected at Sam Ratulangi University, Kampus Bahu, Indonesia, in 2015 and soaked in EtOH immediately.

Extraction and separation

Leaves of A. communis (fresh weight, 169 g) were extracted with EtOH. Half of the extract was partitioned between EtOAc and H2O. The EtOAc fraction was partitioned between n-hexane and 90% MeOH–H2O. The 90% MeOH–H2O layer (1.1 g) was concentrated in vacuo and subjected to silica gel column chromatography with a stepwise gradient elution using n-hexane/EtOAc (2:1), CH2Cl2/MeOH (19:1 and 9:1), and CH2Cl2/MeOH/H2O (6:4:1) to afford 10 fractions. A portion (30 mg) of Fr. 2 (130 mg), eluted with n-hexane/EtOAc (2:1), was applied to gel filtration HPLC (Asahipak GS-310P, Asahi Chemical Industry Co., Ltd., 21.5 × 500 mm) with n-hexane/EtOAc (2:1) and HPLC (Inertsil Diol, GL Sciences, 20 × 250 mm) with n-hexane/EtOAc (2:1) to afford 7 (6.9 mg). A portion (30 mg) of Fr. 3 (90 mg), eluted with n-hexane/EtOAc (2:1), was applied to gel filtration HPLC (Asahipak GS-310P, 21.5 × 500 mm) with CH2Cl2/MeOH (10:1) to yield 5 (10.4 mg). Fr. 4 (64 mg), eluted with CH2Cl2/MeOH (19:1), was subjected to gel filtration HPLC (Asahipak GS-310P, 21.5 × 500 mm) with CH2Cl2/MeOH (4:1) followed by HPLC (Inertsil Diol, 20 × 250 mm) with n-hexane/EtOAc (1:1) to afford 1 (2.5 mg) and 6 (2.2 mg). The remaining half of the EtOH extract was subjected to solvent partitioning followed by silica gel column chromatography in a similar manner to that described above and afforded 10 fractions. Fr. 1 (140 mg), eluted with CH2Cl2/MeOH (19:1), was applied to gel filtration HPLC (Asahipak GS-310P, 21.5 × 500 mm) with MeOH to afford nine fractions. Frs. 6 (13 mg) and 7 (14 mg) were combined and subjected to normal phase HPLC (COSMOSIL 5SL-II, Nacalai Tesque Inc., 10 × 250 mm) with n-hexane/CH2Cl2/MeOH (20:19:1) followed by ODS HPLC (COSMOSIL 5C18-AR-II, Nacalai Tesque Inc., 10 × 250 mm) eluted with 70% MeOH–H2O to afford 2 (0.5 mg), 3 (0.8 mg), and 4 (1.7 mg). Fr. 5 (12 mg) was purified by normal phase HPLC (COSMOSIL 5SL-II, 10 × 250 mm) with n-hexane/CH2Cl2/MeOH (20:19:1) and ODS HPLC (COSMOSIL π-Nap, Nacalai Tesque Inc., 10 × 250 mm) with 70% MeOH–H2O to afford 8 (1.5 mg). Separation of 3 (0.8 mg) by HPLC using a chiral-phase column (CHIRALCEL OJ-H, Daicel, 4.6 × 250 mm) with n-hexane/EtOH (4:1) afforded two equivalent amounts of diastereomers (0.4 mg each). Analysis of 4 (10 μg) under the same conditions revealed four peaks. Resolution of 7 (2.4 mg) and 8 (1.5 mg) was carried out by the same column with n-hexane/2-propanol (4:1) to yield (+)- (1.1 mg)/(–)-7 (1.1 mg) and (+)- (0.4 mg)/(–)-8 (0.6 mg), respectively.

Compound 1: a yellow amorphous solid. \( [\alpha ]^{ 2 1}_{\text{D}} {-} 1 9 \) (c = 2.1, MeOH). UV λ max (MeOH) nm (log ε): 292 (3.9) nm. IR (film) υmax 3198, 2923, 2854, 1634, 1602, 1520, 1436, 1378, 1342, 1264, 1180, 1072, 1114, 1072, 1022, 1000, 820, 781, 550 cm−1. 1H and 13C NMR data (DMSO-d6), see Tables 1 and 2. HRESIMS m/z 423.1837 [M–H]− (calcd for C25H27O6, 423.1813).

Compound 2: a yellow amorphous solid. \( [\alpha ]^{ 2 5}_{\text{D}} {-} 20 \) (c = 0.4, MeOH). UV λmax (MeOH) nm (log ε): 292 (4.1) nm. IR (film) υmax 3210, 2925, 1635, 1600, 1519, 1436, 1343, 1267, 1170, 1073, 1023, 834 cm−1. 1H and 13C NMR data (CDCl3), see Tables 1 and 2. HRESIMS m/z 447.1773 [M + Na]+ (calcd for C25H28NaO6, 447.1778).

Compound 3: a yellow amorphous solid. \( [\alpha ]^{ 2 5}_{\text{D}} {-} 1 4 \) (c = 0.4, MeOH). UV λmax (MeOH) nm (log ε): 294 (3.9) nm. IR (film) υmax 3210, 2924, 1634, 1600, 1518, 1435, 1342, 1267, 1170, 1073, 1023, 833 cm−1. 1H and 13C NMR data (CDCl3), see Tables 1 and 2. HRESIMS m/z 447.1774 [M + Na]+ (calcd for C25H28NaO6, 447.1778).

Compound 4: a yellow amorphous solid. \( [\alpha ]^{ 2 1}_{\text{D}} + 7. 2 \) (c = 1.9, MeOH). UV λ max (MeOH) nm (log ε): 326 (3.6), 276 (4.1), 218 (4.3) nm. IR (film) υmax 3252, 2926, 1719, 1627, 1494, 1443, 1365, 1239, 1207, 1133, 1022, 987, 902, 802, 724, 614, 563 cm−1. 1H and 13C NMR data (CDCl3), see Table 3. HRESIMS m/z 447.1777 [M + Na]+ (calcd for C25H28NaO6, 447.1778).

Conformational analysis and ECD calculation of S-9

This experiment was conducted as previously reported [28]. Conformational analysis was performed with Spartan’16 (Ver. 1.0.0; Wavefunction Inc., Irvine, CA), and ECD calculation was performed with Gaussian09 (Revision D.01; Gaussian, Wallingford, CT, USA) [29] at the B3LYP/TZVP level. No wavelength correction was needed.

USP7 inhibition assay

This assay was performed as previously reported [30]. P5091 (Abcam, Inc.) was used as positive control and its IC50 value was 5.2 μM.

Proteasome inhibition assay

This experiment was performed as previously reported [31]. MG132 (Peptide Institute, Inc.) was used as positive control and its IC50 value was 0.037 μM.