Isolation of tyrosinase and melanogenesis inhibitory flavonoids from Juniperus chinensis fruits

ABSTRACT

A new biflavonoid, amentoflavone-7-O-β-D-glucoside, and thirteen known flavonoids were isolated from the fruits of Juniperus chinensis using a bioactivity-guided method and their tyrosinase inhibitory effects were tested using a mushroom tyrosinase bioassay. Two isolates, hypolaetin-7-O-β-D-glucoside and quercetin-7-O-α-L-rhamnoside, were found to reduce tyrosinase activity at a concentration of 50 μM. Quercetin-7-O-α-L-rhamnoside attenuated cellular tyrosinase activity and melanogenesis in α-MSH plus IBMX-stimulated B16F10 melanoma cells. Molecular docking simulation revealed that quercetin-7-O-α-L-rhamnoside inhibits tyrosinase activity by hydrogen bonding with residues His85, His244, Thr261, and Gly281 of tyrosinase.

Abbreviations: EtOH, ethanol; CH2Cl2, dichloromethane; EtOAc, ethylacetate; n-BuOH, n-butanol; MeOH, metanol; CHCl3,chloroform; DMSO, dimethylsulfoxide; DMEM, Dulbecco’s modified Eagle’s medium; FBS, fetal bovine serum; α-MSH, α-melanocyte stimulating hormone; L-DOPA, L-3, 4-dihydroxyphenylalanine

GRAPHICAL ABSTRACT

A new biflavonoid, amentoflavone-7-O--D-glucoside, and a tyrosinase inhibitory flavonoid, quercetin-7-O--L-rhamnoside, were isolated from J. chinensis fruits.

KEYWORDS:

• Juniperus chinensis fruit
• amentoflavone-7-O-β-D-glucoside
• quercetin-7-O-α-L-rhamnoside
• tyrosinase
• melanogenesis

Visible skin pigmentation primarily depends on the functions of melanocytes, the pigment melanin-producing cells that specialize in the synthesis and distribution of melanin [1,2]. Melanin production in melanocytes is triggered by a variety of physiological factors, such as, melanosomal enzymes and melanogenic regulators. Pigment production in murine melanoma cells is mainly regulated by the enzymatic components of melanosomes, such as, tyrosinase, tyrosinase-related protein 1, and dopachrome tautomerase [3]. Tyrosinase is a copper-dependent, rate-limiting factor that catalyzes melanin synthesis in melanocytes [4], which when excessive can cause pathologic and esthetic problems, such as, freckles, age spots, and melisma [5]. Diverse skin-whitening agents like kojic acid and hydroquinone, act as strong tyrosinase inhibitors, but their topical toxicities and storage stabilities are topics of concern [6,7]. Natural substances are widely used to treatment skin problems because of their proven safeties and therapeutic properties, which including anti-inflammatory, antimicrobial, and cell-stimulating effects [8]. Thus, plant resources including extracts, pure compounds, and phytochemical combinations continue to attract the interests of those seeking to find new ways of preventing and treating conditions related to skin hyperpigmentation.

Juniperus chinensis L. (Cupressaceae, common name Chinese juniper) is an evergreen coniferous tree that is widely distributed in Korea, China, and Japan [9,10]. The heartwood of J. chinensis has been traditionally used to treat the common cold and inflammatory diseases, such as, urinary infections and rheumatic arthritis, whereas its fruit is known to treat convulsion, hyperhidrosis, and hepatitis [11,12]. Analyses of the chemical compositions of Juniperus fruits have shown that phenolics, including phenolic acids and flavonoids, constitute the main class of non-volatile metabolites [13,14], and natural phenolics that effectively inhibit tyrosinase are of interest to those designing and developing skin-whitening cosmetics [15,16]. On the other hand, the flavonoids are naturally occurring antioxidants, and are also considered potentially useful skin-depigmenting agents and tyrosinase-inhibitors [17,18]. In the present study, we undertook to isolate and identify flavonoids from the fruits of J. chinensis with tyrosinase inhibitory activities using cellular and non-cellular systems and to investigate the mechanisms involved by molecular docking.

Materials and methods

General

¹H and ¹³C NMR, COSY, HMQC, HMBC, and NOESY spectral data were obtained using an Agilent Superconducting FT-NMR 400 or 500 MHz Spectrometer. HR-ESI mass spectra were recorded on an Agilent Technologies, 6530 Accurate-Mass Q-TOF LC/MS. The HPLC unit (Shimadzu, Japan) was equipped with a pump (Model LC-20AT), an UV detector (Model SPD-20A) at 254 nm, and a data station (Model CBM-20A). Silica gel (230–400 mesh, Merck, Germany) and Sephadex LH-20 gel (25–100 μM mesh, Pharmacia, Sweden) were used for column chromatography. L-tyrosine, mushroom tyrosinase, kojic acid, α-MSH (α-melanocyte stimulating hormone), and MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) were purchased from Sigma (St. Louis, MO, USA).

Plant material

Fresh fruits of J. chinensis were collected from Muju (Jeollanam-Do) in South Korea and identified by Prof. Eun Ju Jeong at the Department of Agronomy and Medicinal Plant Resources, Gyeongnam National University of Science and Technology. A voucher specimen (PNU-0022) was deposited at the Medicinal Herb Garden, Pusan National University.

Extraction and isolation

Dried fruits of J. chinensis (3.9 kg) were extracted with 95% EtOH (30 L × 3) and the extract obtained was then evaporated under reduced pressure. This extract (824.2 g) was then suspended in distilled water (1 L) and successively partitioned with CH₂Cl₂ (6 L), EtOAc (6 L), and n-BuOH (6 L). The active EtOAc soluble fraction (14.6 g) was chromatographed on a Silica gel CC using CH₂Cl₂-MeOH (10:1 → 100% MeOH, gradient system) as eluent to yield ten fractions (JChE1~ JChE10). Fraction JChE3 (329.7 mg) was applied to Sephadex LH-20 (MeOH), to afford compound 14 (2.4 mg), fraction JChE4 (CH₂Cl₂:MeOH = 10:1, 544.7 mg) yielded compound 8 (7.0 mg), and fraction JChE5 (CH₂Cl₂:MeOH = 10:1, 977.3 mg) yielded compound 10 (22.1 mg) and compound 7 (22.6 mg), sequentially. Fraction JChE6 (CH₂Cl₂:MeOH = 5:1, 626.8 mg) was suspended in MeOH and compound 13 (31.4 mg) was obtained by filtration. The resulting filtrate of JChE6 was fractionated into six subfractions (Frs. JChE6-1~ JChE6-6) by Sephadex LH-20 (MeOH). Subfractions JChE6-4 and JChE6-5 contained compound 3 (5.9 mg) and 11 (4.1 mg), respectively. Subfraction JChE6-3 (150.0 mg) was subjected to RP HPLC (Watchers 120 ODS-BP, S-10 μm, 150 × 10 mm; detection, UV at 254 nm; flow rate, 2 mL/min) using a MeOH-H₂O gradient (5:5 → 9:1 over 40 min) to yield compound 6 (53.2 mg, t_R 20 min). Compound 4 (30.4 mg) was obtained as pure powder by filtering fraction JChE7 (CH₂Cl₂:MeOH = 5:1, 1306.6 mg) and compound 5 (36.5 mg) by recrystallizing JChE7 filtrate from MeOH.. Fraction JChE8 (CH₂Cl₂:MeOH = 5:1, 798.1 mg) was fractionated into thirteen subfractions (Frs. JChE8-1~ JChEN8-13) by Sephadex LH-20 (MeOH). Subfraction JChE8-6 (178.3 mg) was separated by RP HPLC (Watchers 120 ODS-BP, S-10 μm, 150 × 10 mm; detection, UV at 254 nm; flow rate, 2 mL/min) and eluted with a MeOH-H₂O gradient (4:6 → 9:1, 40 min) to yield compound 2 (1.2 mg, t_R 19 min) and compound 1 (12.5 mg, t_R 20 min). Subfraction JChE8-8 (14.1 mg) was subjected to RP HPLC (Watchers 120 ODS-BP, S-10 μm, 150 × 10 mm; detection, UV at 254 nm; flow rate, 2 mL/min) and eluted with a MeOH-H₂O gradient (3:7 → 9:1, 40 min) to yield compound 12 (3.6 mg, t_R 29 min). The subfraction JChE8-9 (4.9 mg) was also subjected to RP HPLC (Watchers 120 ODS-BP, S-10 μm, 150 × 10 mm; detection, UV at 254 nm; flow rate, 2 mL/min) and eluted with a MeOH-H₂O gradient (3:7 → 9:1, 40 min) to yield compound 9 (1.8 mg, t_R 32 min).

Amentoflavone-7-o-β-d-glucoside (9)

Yellowish powder; UV (MeOH) λ_max (log ε) at 336 nm (4.06), 269 nm (4.17); IR (neat) ν_max 3357, 1653, 1608, 1494 cm^{−1; 1}H NMR (DMSO, 600 MHz) and ¹³C NMR (DMSO, 150 MHz) see Table 1; HRESIMS m/z 699.1386 (calcd. 700.1428 for C₃₆H₂₈O₁₅).

Table 1. ¹H (600 MHz) and ¹³C (150 MHz) NMR spectral data of compound 9 in DMSO-d_6.

	Compound 9
Position	δH (J in Hz)	δC
2		163.54
3	6.87 (s)	103.18
4		182.66
5		161.93
6	6.41 (s)	99.76
7		165.11
8	6.72 (s)	95.48
9		157.62
10		103.66
1’		121.34
2’	8.12 (s)	128.27
3’		122.08
4’		161.59
5’	7.01 (d, 8.0)	116.45
6’	7.97 (d, 8.0)	121.34
2”		164.13
3”	6.72 (s)	103.18
4”		182.60
5”		161.59
6”	6.41 (s)	99.37
7”		161.76
8”		105.93
9”		155.16
10”		103.66
1’’’		122.08
2’’’ (6’’’)	7.60 (d, 8.1)	128.83
3’’’ (5’’’)	6.64 (d, 8.2)	116.35
4’’’		161.16
Glc 1	5.02 (d, 7.3)	100.13
2	3.33	73.71
3	3.33	77.02
4	3.15	70.06
5	3.33	77.70
6	3.64,	61.13

Acid hydrolysis of compound 9

Acid hydrolysis was determined as described previously [19]. Briefly, compound 9 (1 mg) in 50% MeOH was hydrolyzed with 2 M HCl for 3 h and concentrated under reduced pressure. The residue was dissolved in H₂O and extracted with CHCl₃. The collected aqueous layer was concentrated and glucose was identified by the TLC with an authentic sample of amentoflavone using CHCl₃–MeOH (10:1).

In vitro mushroom tyrosinase assay

A mixed solution containing 170 μL of 1 mM L-tyrosine in 50 mM phosphate buffer (pH 6.5) and 10 μL of test sample was prepared for in vitro mushroom (Agaricus bisporus) tyrosinase experiment. The reaction conducted by adding 20 μL of 1000 U/mL of tyrosinase and incubating the mixture at 25°C for 30 min. DMSO (0.5%) was added as a negative control and kojic acid (50 μM) was used as a positive control. The amount of dopachrome produced was monitored against a blank (no tyrosinase addition) at 492 nm using a microplate reader (GENios; Tecan Instruments, Salzburg, Austria).

Cell culture. Mouse melanoma B16F10 cells were obtained from the Korea Cell Line Bank, cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS and 1% penicillin/streptomycin, and then incubated at 37°C in a 5% CO₂ atmosphere.

Cell viability

Cell viabilities were assessed using the EZ-Cytox assay. Cells were seeded at 5 × 10³ cells/well in DMEM containing 10% FBS and incubated for 24 h. Cells were then treated with 1–100 μM one of the 14 compounds for 24 h or 48 h. Cells treated with 0.1% DMSO were used as controls. After treatments, 10 μL of EZ-Cytox solution was added to each well, and cells were incubated for additional 2 h. Well absorbance were measured at 450 nm using an ELISA reader. All determinations were performed in triplicate and results were averaged.

Melanin content assay

B16F10 cells were seeded in 6-well culture plates at 2.5 × 10⁴ cells/well in a 5% CO₂ atmosphere and cultured for 24 h. To determine the inhibitory effect of MHY773 on melanogenesis, fresh medium was replaced with medium containing one of the 14 compounds at 10, 20, or 50 µM or with 10 μM kojic acid (positive control), and 0.5 μM α-MSH and 200 μM IBMX for 48 h. Cells were then washed twice with PBS, centrifuged, disrupted in 100 μL of 1 N NaOH, incubated at 60°C for 1 h, and mixed to solubilize the melanin. Absorbance at 405 nm converted to concentrations using a standard curve prepared with synthetic melanin. Relative melanin levels were calculated with respect to protein contents, which were determined using BCA protein assay reagent (Thermo Scientific, Rockford, IL, USA). Results are presented as means of three individual experiments.

Cellular tyrosinase activity assay

B16F10 cells were plated in 6-well dishes at a density of 2.5 × 10⁴ cells/well, incubated with or without 0.5 μM α-MSH and 200 μM IBMX, and then treated for 48h with compounds 1 to 14 at concentrations of 0 to 50 μM. Cells were then washed and lysed in 100 μL of 50 mM sodium phosphate buffer (pH 6.5) containing 1% Triton X-100 and 0.1mM PMSF (phenylmethylsulfonyl fluoride), frozen at – 80°C for 30 min, thawed, mixed, and cellular extracts were the clarified by centrifuging at 12,000 rpm for 30min at 4°C. In each case an 8 μL sample of supernatant and 20 μL of l-DOPA (2 mg/mL) were placed in a 96-well plate, and the absorbance was read at 492 nm every 10 min for 1h at 37°C using an ELISA reader. Final activities were expressed as Δ OD/min.

Docking simulation of tyrosinase and compound 5

AutoDock 4.2 was used for the in silico protein-ligand docking simulation. Successful binding was obtained between a protein and a ligand. The 3D structure of tyrosinase used was as previously described for the tyrosinase of Agaricus bisporus (PDB ID: 2Y9X), and the defined binding site of tyrosine was used as the docking pocket. Docking simulations were performed between tyrosinase and compound 5 or kojic acid. To prepare compounds for docking simulation, (1) 2D structures were converted into 3D structures, (2) charges were calculated, and (3) hydrogen atoms were added using the ChemOffice program (http://www.cambridgesoft.com). LigandScout 3.0 was used to predict possible residues responsible for hydrogen bonding between the 14 compounds and tyrosinase.

Results and discussion

Preliminary screening of plant materials for anti-tyrosinase activity showed the 95% EtOH extract of J. chinensis fruits inhibited mushroom tyrosinase activity. The 95% EtOH extract of J. chinensis fruit was suspended in distilled water and partitioned sequentially against CH₂Cl₂, EtOAc, and n-BuOH. The active EtOAc soluble fraction (14.6 g) was subjected to silica gel column chromatography, Sephadex LH-20 column chromatography, and RP HPLC to yield a novel biflavonoid, amentoflavone 7-O-β-D-glucoside (9) and 13 known flavonoids; apigenin 7-O-β-D-glucoside (1) [20], luteolin 7-O-β-D-glucoside (2) [21], hypolaetin 7-O-β-D-xyloside (3) [22], isoscutellarein 7-O-β-D-xyloside (4) [22], quercetin 7-O-α-L-rhamnoside (5) [23], kaempferol 3-O-α-L-rhamnoside (6) [24], amentoflavone (7) [25], 7-O-methylamentoflavone (8) [26], cupressuflavone (10) [27], moghatin (11) [27], cupressuflavone 4′-O-β-D-glucoside (12) [28], hinokiflavone (13) [25], and isocryptomerin (14) [25]^, (Figure 1). The structures of these 14 compounds were elucidated by 2D-NMR and HR-MS.

Figure 1. Structures of flavonoids isolated from J. chinensis fruit extract.

Compound 9 was obtained as a yellowish powder, [α]_D – 66.9 (c 0.1, MeOH) of molecular formula C₃₆H₂₈O₁₅, as determined by a negative-ion HRESIMS peak at m/z 699.1386. The MS fragmentation pattern of 9 showed fragments characteristic of amentoflavone [29]. Acid hydrolysis of 9 resulted in the isolation of amentoflavone (7) and glucose, both of which were identified by direct TLC comparisons with authentic materials. The IR spectrum of compound 9 displayed three significant absorption bands at 3357, 1653, and 1494 cm⁻¹, which were indicative of hydroxyl and carbonyl groups and an aromatic ring, respectively. UV spectrum maxima at 336 nm (log ε 4.06) and 269 nm (log ε 4.17) indicated that 9 was a flavone-like compound. MS, IR, and UV data indicated 9 contained a biflavone monoglycoside moiety containing a hexosyl residue.

The ¹H NMR spectrum of compound 9 (Table 1) revealed the presence of 1,4,5-trisubstituted benzene ring [δ 8.12 (1H, s, H-2′), 7.01 (1H, d, J = 8.6 Hz, H-5′), and 7.97 (1H, d, J = 8.6 Hz, H-6′)], ortho-coupled protons were assignable to two p-substituted phenyl moieties [δ 7.60 (2H, d, J = 8.8 Hz, H-2′′′ and H-6′′′) and 6.64 (2H, d, J = 8.8 Hz, H-3′′′ and H-5′′′)], and three aromatic singlets [δ 6.87 (1H, H-3), 6.41 (2H, H-6 and 6′′), and 6.72 (2H, H-3′′ and H-8)]. Detailed analyses of HMQC and HMBC spectra allowed all ¹H- and ¹³C-NMR signals to be assigned to glucosyl and aglycone moieties. HMBC correlations between H-8 and C-6, H-2′ and C-6′, H-6′′ and C-5′′/C-7′′′, and H-3′′/C-3′′′ and C-1′′′ were consistent with 3′,8″-coupled dimers of apigenin (5,7,4′-trihydroxyflavone), indicating amentoflavone was the common aglycone in compound 9. Furthermore, NOESY correlations between anomeric H and H-6 and H-8 (Figure 2) confirmed a β-D-glucopyranosyl moiety at OH-7 in amentoflavone. Therefore, compound 9 was designated amentoflavone-7-O-β-D-glucoside.

Figure 2. Selected HMBC and NOESY correlations of amentoflavone-7-O-β-D-glucoside (9).

The tyrosinase inhibitory effects of all compounds from J. chinensis fruits were investigated using non-cellular mushroom tyrosinase. Of the 14 flavonoids, hypolaetin-7-O-β-D-glucoside (3) and quercetin-7-O-α-L-rhamnoside (5) effectively reduced mushroom tyrosinase activity to < 50% at a concentration of 50 μM (Table 2). Compound 3 was isolated from J. chinensis for the first time and most potently inhibited mushroom tyrosinase, which concurs with our previous report on the activity of hypolaetin-7-O-β-D-glucoside isolated from J. communis [30]. Quercetin-7-O-α-L-rhamnoside (5) effectively inhibited mushroom tyrosinase with an IC₅₀ value of 56.75 μM using tyrosine as a substrate (Figure 3). It has been already reported that quercetin, an aglycone form of compound 5, also exhibited a potent inhibitory effect on dopa oxidase activity of mushroom tyrosinase [31]. This study suggested both quercetin and its 7-rhamnoside derivate can be utilized in skin-lightening attempts as strong tyrosinase inhibitors. The depigmenting effect of compound 5 was investigated using B16F10 melanoma cells. Both tyrosinase activity and melanin content in B16F10 cells were markedly increased by 8.2-fold and 1.7-fold of the vehicle control by co-treating cells with compound 5 and α-MSH/IBMX. As shown in Figure 4(a), 5 down-regulated tyrosinase activity in both cell types at 50 μM in vitro. α-MSH and IBMX-induced melanin production was also reduced in B16F10 cells by 92.8% by 5 at 10 μM (Figure 4(b)). Structural and molecular docking simulations showed the best ranked energy docking result with lowest binding energy (−6.4 kcal/mol) (Figure 5(a)). The tyrosinase docking abilities of compound 5 and kojic acid (−5.7 kcal/mol, positive control) were similar. Compound 5 was found to bind to His85 near the active site of tyrosinase mainly by hydrogen bonding (Table 3 and Figure 5(b)). In silico docking results suggested that compound 5 (quercetin-7-O-α-L-rhamnoside) from J. chinensis competitively bound to the active site of tyrosinase.

Table 2. Inhibitory effects of flavonoids isolated from J. chinensis fruit on mushroom tyrosinase.

Compounds	Relative enzyme activity (%)^a 50 μM
Apigenin-7-O-β-D-glucoside (1)	75.86 ± 1.07***
Luteolin-7-O-β-D-glucoside (2)	71.96 ± 2.27***
Hypolaetin-7-O-β-D-xyloside (3)	26.71 ± 0.71***
Isoscutellarein-7-O-β-D-xyloside (4)	74.58 ± 3.70***
Quercetin-7-O-α-L-rhamnoside (5)	46.81 ± 2.47***
Kaempferol-3-O-α-L-rhamnoside (6)	81.59 ± 6.68*
Amentoflavone (7)	83.40 ± 5.21*
7-O-Methylamentoflavone (8)	74.46 ± 1.84***
Amentoflavone-7-O-β-D-glucoside (9)	85.86 ± 6.98*
Cupressuflavone (10)	75.69 ± 6.01**
Moghatin (11)	88.81 ± 0.12**
Cupressuflavone-4′-O-β-D-glucoside (12)	79.54 ± 5.38**
Hinokiflavone (13)	65.46 ± 0.10***
Isocryptomerin (14)	76.80 ± 2.15***
Kojic acid^b	35.18 ± 2.38***

^a Relative enzyme activities were calculated as percentages of the activity of the control.

^b Positive control.

Table 3. Interaction and Autodock scores of quercetin-7-O-α-L-rhamnoside for tyrosinase.

Compound	Docking score kcal/mol	Interacting residue
Kojic acid	−5.7	HIS259, HIS263, and MET280
Quercetin-7-O-α-L-rhamnoside	−6.4	HIS85, HIS244, GLY281, and VAL283

Figure 3. Dose-dependent inhibition of mushroom tyrosinase activity by quercetin-7-O-α-L-rhamnoside.

Results are presented as the means ± standard errors of triplicate assays.

Figure 4. Inhibitory effects of quercetin-7-O-α-L-rhamnoside from J. chinensis fruit on cellular tyrosinase activities (a) and melanin contents (b).

Results are presented as means± SDs of triplicate experiments; #p < 0.001 vs. the control and *p < 0.05 and ***p < 0.001 vs. α-MSH and IBMX-treated cells.

Figure 5. In silico molecular docking with tyrosinase.

(a) Computational docking simulation images for tyrosinase with quercetin-7-O-α-L-rhamnoside and kojic acid. The geometry of whole mushroom tyrosinase is shown on the left. The active site of tyrosinase is magnified on the right. The two brown spheres indicate copper ions at the active site. Cyan denotes tyrosine binding sites predicted by Autodock, and magenta and yellow represent kojic acid and quercetin-7-O-α-L-rhamnoside binding sites, respectively, as predicted by Autodock. Amino acids in the cavity are shown as stick models. (b) Possible hydrogen-bonding interactions of compound 5.

In conclusion, bioactivity-guided isolation of the 95% ethanol extract of the fruits of J. chinensis afforded a new biflavonoid amentoflavone-7-O-β-D-glucoside and thirteen known compounds. Cell-free and cell-based assay systems and molecular docking indicated quercetin-7-O-α-L-rhamnoside from J. chinensis is a promising candidate inhibitor of tyrosinase and melanin formation for cosmetic applications, and that J. chinensis fruit extracts or quercetin-7-O-α-L-rhamnoside might be useful for the treatment of skin pigmentary disorders.

Authors contribution

S. Park, J. Jegal, and M.H. Yang conceived and designed the research. S. Park, J. Jegal, K.W. Chung, H.J. Jung, S.G. Noh, and J. Ahn performed experiments. S. Park, J. Jegal, H.Y. Chung, J. Kim and M.H. Yang drafted manuscript. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental data

Supplemental data for this article can be accessed here.

Additional information

Funding

This work was supported by the Korean National Research Foundation (NRF) funded by the Korean government (MSIP) (grant nos. NRF-2016R1C1B2007694 and NRF-2016K1A1A8A01938595) and the “Cooperative Research Program for Agriculture Science and Technology Development (Project No. PJ01282301)” Rural Development Administration, Republic of Korea.