Characterization of Four New Compounds from Protea cynaroides Leaves and Their Tyrosinase Inhibitory Potential

Protea cynaroides (king protea) is a flowering plant that belongs to the Proteaceae family. This multi-stemmed shrub is the national flower of South Africa and has important economic and medicinal values. Traditionally, the main therapeutic benefits of this plant species include the treatment of cancer, bladder, and kidney ailments. There are very limited reports on the isolation of phytochemicals and their biological evaluation from P. cynaroides. In this study, the leaves of P. cynaroides were air-dried at room temperature, powdered, and extracted with 80% methanol followed by solvent fractionation (hexane, dichloromethane, ethyl acetate, and butanol). The ethyl acetate and butanol extracts were chromatographed and afforded four new (1–4) and four known (5–8) compounds, whose structures were characterized accordingly as 3,4-bis(4-hydroxybenzoyl)-1,5-anhydro-D-glucitol (1), 4-hydroxybenzoyl-1,5-anhydro-D-glucitol (2), 2-(hydroxymethyl)-4-oxo-4H-pyran-3-yl-6-O-benzoate-β-D-glucopyranoside (3), 3-hydroxy-7,8-dihydro-β-ionone 3-O-β-D-glucopyranoside (4), 4-hydroxybenzoic acid (5), 1,5-anhydro-D-glucitol (6), 3,4-dihydroxybenzoic acid (7), and 3-hydroxykojic acid (8). The structural elucidation of the isolated compounds was determined based on 1D and 2D NMR, FTIR, and HRMS spectroscopy, as well as compared with the available literature data. The tyrosinase inhibitory activities of the extracts and isolated compounds were also determined. According to the results, compounds 7 and 8 exhibited potent competitive tyrosinase inhibitory activity against L-tyrosine substrates with IC50 values of 0.8776 ± 0.012 and 0.7215 ± 0.090 µg/mL compared to the control (kojic acid, IC50 = 0.8347 ± 0.093), respectively. This study is the first chemical investigation of compounds 1–4 from a natural source and the first report of the biological evaluation of compounds 1–5 against the tyrosinase enzyme. The potent anti-tyrosinase activity exhibited by P. cynaroides constituents will support future exploration of the plant in the cosmetic field upon further biological and clinical investigations.


Introduction
Tyrosinase is a multifunctional copper-containing enzyme; it plays diverse physiological roles in different organisms [1]. Tyrosinase catalyzes the initial step in the formation of the pigment melanin from tyrosine, it is thought to be involved in wound healing and possibly sclerotization of the cuticle [2] in plants. In addition, tyrosinase is known to be involved in the molting process of insects and the adhesion of marine organisms [3][4][5][6]. The various phenolic compounds involve the physiological substrates of tyrosinase and oxidize in the observed browning pathway when the tissues are damaged [7]. In mammals, tyrosinase overactivity causes the accumulation of melanin pigments [8]. Excessive amounts of melanin darken the skin and cause the formation of dark spots [9]. The skin-lightening NMR experiments (HSQC and HMBC), FTIR, and LC-MS data, while the know pounds 5-8 were confirmed by comparing the spectroscopic data to the literature
The combined spectroscopic analysis of the COSY, HSQC, and HMBC data allowed the establishment of the structure of compound 1 (Figure 2). The HSQC spectra allowed the assignment of all the protons attached to their corresponding carbons. Inspection of 1 H-1 H COSY was observed between the two aromatic doublets at H-2 /2 , H-6 /6 , and H-3 /3 , H-5 /5 , which also confirmed the presence of a para-substituted benzene ring. In addition, the attachment of the acyl groups to the sugar moiety was confirmed by the HMBC crosspeak correlation between H-3/C-7 and H-4/H-7 ; thus, the structure of compound 1 was established as 3,4-bis(4-hydroxybenzoyl)-1,5-anhydro-D-glucitol. A Sci-Finder database search provided no evidence for compound 1 as having been previously reported; therefore, it was proposed as a new compound. Similar structures have been isolated before but with a pyrogallol moiety instead of a 4-hydroxybenzoate. These kinds of compounds are known to show different bioactivities such as alpha-glucosidase, alpha-amylase inhibition, and antioxidant activities [32]. The combined spectroscopic analysis of the COSY, HSQC, and HMBC the establishment of the structure of compound 1 (Figure 2). The HSQC sp the assignment of all the protons attached to their corresponding carbons. 1 H-1 H COSY was observed between the two aromatic doublets at H-2′/2″, H 3′/3″, H-5′/5″, which also confirmed the presence of a para-substituted be addition, the attachment of the acyl groups to the sugar moiety was con HMBC cross-peak correlation between H-3/C-7′ and H-4/H-7″; thus, the stru pound 1 was established as 3,4-bis(4-hydroxybenzoyl)-1,5-anhydro-D-gl Finder database search provided no evidence for compound 1 as having be reported; therefore, it was proposed as a new compound. Similar structu isolated before but with a pyrogallol moiety instead of a 4-hydroxybenzoat of compounds are known to show different bioactivities such as alpha-gluco amylase inhibition, and antioxidant activities [32].   167.7 (C-7 ), 163.8 (C-4 ), 133.2 (C-3 /5 ), 122.1 (C-1 ), and 116.3 (C-2 /6 ). Apart from the 4-hydroxybenzoyl carbon signals, six oxygenated carbon signals at δ C 81.0 (C-5), 77.9 (C-3), 73.0 (C-4), 71.7 (C-2), 71.1 (C-1), and 62.9 (C-6) were observed in the 13 C NMR spectrum, which also supported the presence of a sugar moiety. A further combined analysis of the COSY, HSQC, and HMBC spectra (Supplementary Material) allowed the establishment of the structure of compound 2. The HSQC spectrum allowed the assignment of all the protons attached to their corresponding carbons. The 1 H-1 H COSY spectrum ( Figure 2) revealed a connection between the doublet appearing at δ H 7.83 ppm (H-2 /6 ) and the doublet at δ H 6.75 ppm (H3 /5 ), which confirmed the presence of the disubstituted benzene moiety. Other 1 H-1 H correlations between δ H 3.16 (assigned as H-1a), δ H 3.89 (H-1b), and H-2, δ H 3.38 (H-5) with 4.81 ppm, already assigned as H-4, confirmed the presence of the sugar moiety. The HMBC correlations between H-4 and the ester carbonyl (C-7 ) indicated that the 4-hydroxybenzoate group was linked at C-4 of the 1,5-anhydro-glucitol moiety; thus, the structure of compound 2 was established as 4-hydroxybenzoyl-1,5-anhydro-D-glucitol. A SciFinder database search provided no evidence for compound 2 as having been previously reported; therefore, it was proposed as a new compound. The oxymethylene signal appeared at δ C 57.7 (C-7). The combined spectroscopic analysis of the 1 H-1 H COSY, 1 H-13 C HSQC, and 1 H-13 C HMBC data allowed the establishment of the structure of compound 3. In the COSY spectra, correlations between δ H 6.31 (H-5) and δ H 7.97 (H-6) confirmed the presence of a 1,2-disubstituted 4-pyrone unit. In addition, a strong correlation was observed between δ H 7.84 (H-2 /6 ) and δ H . 6.83 (H-3 /5 ), which also confirmed the presence of a para-substituted benzene ring. The observed correlation of H-6 from the sugar unit with C-7" in the HMBC spectrum indicated that the 4-hydroxybenzoyl moiety was attached at C-6 of the glucosyl. Furthermore, the HMBC experiment revealed a strong correlation of the anomeric proton H-1 with the carbon signal at δ C 142.3, which was attributed to C-3 of the pyranone unit. In addition, the oxymethylene was located at C-2 of the pyranone unit and was confirmed through HMBC correlations between H-7 and C-2/C-3 ( Figure 3). Therefore, compound 3 was determined as 2-(hydroxymethyl)-4-oxo-4H-pyran-3-yl-6-O-benzoate-β-D-glcopyranoside. A SciFinder database search provided no evidence for compound 3 as having been previously reported; therefore, it is hereby proposed as a new compound. was attached at C-6′ of the glucosyl. Furthermore, the HMBC exp strong correlation of the anomeric proton H-1′ with the carbon signa was attributed to C-3 of the pyranone unit. In addition, the oxymethy C-2 of the pyranone unit and was confirmed through HMBC correla and C-2/C-3 ( Figure 3). Therefore, compound 3 was determined as 2oxo-4H-pyran-3-yl-6-O-benzoate-β-D-glcopyranoside. A SciFinder d vided no evidence for compound 3 as having been previously repo hereby proposed as a new compound.   including four methyl groups at δ C 19.5 (C-13), 22.0 (C-12), 25.6 (C-11), and 30.1 (C-10), three methylene at δ C 22.2 (C-7), 37.9 (C-4), and 43.9 (C-8), two oxymethine at δ C 75.7 (C-3) and 76.9 (C-2), and four fully substituted carbon resonances at δ C 42.2 (C-1), 123.9 (C-5), 136.0 (C-6), including one carbonyl at δ C 208.9 (C-9) for the aglycone group. The 13 C NMR displayed a characteristic anomeric carbon resonance at δ C 101.4 (C-1 ), which corresponded to a typical anomeric proton resonance at δ H 4.28 (d, J = 7.8 Hz) in the HSQC spectrum. This suggested that compound 4 had one sugar group. The remaining sugar moiety signals appeared at δ C 61.4 (C-6 ), 70.4 (C-4 ), 73.7 (C-2 ), 77.2 (C-3 ), and 76.8 (C-5 ). The coupling of H-3 with H-2 and H-4, as well as H-7 with H-8, were observed in the 1 H-1 H COSY spectrum. In the HMBC spectra, there was a strong correlation of the anomeric proton H-1 with the carbon signal at δ C 75.7 attributed to C-3 of the aglycone unit, which confirmed the attachment position of the sugar moiety. Furthermore, the HMBC experiment exhibited a strong correlation of the oxymethine proton placed at H-2 with C-11, C-12, and C-3. In addition, a correlation between H-13 with C-4, C-5, and C-6 confirmed the position of the methylene at position 4 of the structure. Therefore, based on the above information, compound 4 was determined as 2-hydroxy-7,8-dihydro-ionone-3-O-β-D-glucopyranoside. Similarly, a SciFinder database search provided no evidence for compound 4 as having been previously reported; therefore, it is proposed as a new compound. This compound is an isomer of a known compound called 3,4-dihydroxy-7,8-dihydro-ionone-3-O-β-D-glucopyranoside (icariside B 8 ), which was first isolated from Epimedium diphyllum [33].

Inhibitory Activities of Isolated Compounds on Tyrosinase
The average tyrosinase inhibition percentages of the isolated compounds screened at 0.2 and 0.1 mg/mL revealed that compounds 7 and 8 had inhibitions of 100%, while 5 and 6 showed moderate (25%) and weak (55%) inhibitions, respectively. The rest of the compounds were not effective inhibitors of mushroom tyrosinase with their inhibition percentages less than 10%. The compounds with the highest inhibition percentages were investigated further to determine their IC 50 values. Compounds 7 and 8 (Table 3, Figure 4) showed IC 50 values of 0.8776 ± 0.12 and 0.7215 ± 0.09 µg/mL, respectively, and the two compounds showed similar inhibition activities compared to kojic acid, a well-known tyrosinase inhibitor. Indeed, benzoic acid and its derivatives were reported as good tyrosinase inhibitors [38], as benzoic acid can chelate the copper at the active site of the enzyme. be required for comparison with kojic acid as a potential inhibitor of tyrosinase. Worthy to mention, compound 3, the glycosylated derivative of kojic acid displayed a low inhibitory activity against tyrosinase, which may reflect the importance of having a free 3-OH group (or 5-OH) to provide extra stability for the inhibition of the tyrosinase enzyme.

Plant Material
The leaves of Protea cynaroides were collected at Kirstenbosch National Botanical Gardens, South Africa, Cape Town (−33°59′13.19″ S, 18°25′29.39″ E) on 31 August, 2018. The identity of the plant species (voucher number: MY-2018-3) was confirmed by the curator of the Compton Herbarium, Kirstenbosch.

Equipment and Chemical Reagents
The 1D ( 1 H, 13 C, and DEPT-135) and 2D NMR (COSY, HSQC, HMBC) spectra were recorded on the Avance 400 MHz NMR spectrometer (Bruker, Rheinstetten, Germany) at 400 (proton, 1 H) and 100 (carbon, 13 C) MHz. Chemical shifts were reported in parts per million (ppm) and coupling constants (J) in Hz. The 1 H and 13 C NMR values were relative to the internal standard TMS and were acquired in CD3OD, CDCl3, or DMSO-d6. HRESI-MS were obtained on a Waters Synapt G2 mass spectrometer (Cone Voltage 15 V), which operated in the negative and/or positive ion modes using direct injection. ATR-FTIR (PerkinElmer Spectrum 100, Llantrisant, Wales, UK) at a transmission mode of 400-4000 Kojic acid is a well-known tyrosinase inhibitor by forming a chelate with the copper ion in the tyrosinase active site through the 5-hydroxyl and 4-carbonyl groups; interestingly, compound 8 has an extra hydroxyl group in position 3, which may have a role in increasing the chelation power. Interestingly, the additional 3-OH group in compound 8 does not appear to have major effects on the inhibitory activity compared to kojic acid. However, further studies on the stability, safety, and toxicity of compounds 7 and 8 will be required for comparison with kojic acid as a potential inhibitor of tyrosinase. Worthy to mention, compound 3, the glycosylated derivative of kojic acid displayed a low inhibitory activity against tyrosinase, which may reflect the importance of having a free 3-OH group (or 5-OH) to provide extra stability for the inhibition of the tyrosinase enzyme.

Equipment and Chemical Reagents
The 1D ( 1 H, 13 C, and DEPT-135) and 2D NMR (COSY, HSQC, HMBC) spectra were recorded on the Avance 400 MHz NMR spectrometer (Bruker, Rheinstetten, Germany) at 400 (proton, 1 H) and 100 (carbon, 13 C) MHz. Chemical shifts were reported in parts per million (ppm) and coupling constants (J) in Hz. The 1 H and 13 C NMR values were relative to the internal standard TMS and were acquired in CD 3 OD, CDCl 3 , or DMSOd 6 . HRESI-MS were obtained on a Waters Synapt G2 mass spectrometer (Cone Voltage 15 V), which operated in the negative and/or positive ion modes using direct injection. ATR-FTIR (PerkinElmer Spectrum 100, Llantrisant, Wales, UK) at a transmission mode of 400-4000 cm −1 column chromatography was performed using Sephadex (LH-20, Sigma-Aldrich, St. Louis, MO, USA), and normal-phase silica gel 60 (70-230 mesh ASTM, Merck, Readington Township, NJ, USA). TLC was performed on silica gel aluminum sheets (Silica gel 60 F254, Merck) to monitor the fractions. Visualization was achieved with 10% H 2 SO 4 and detection with the vanillin sulfuric acid reagent and heating to 105 • C.

Extraction and Fractionation of the Plant Material
The air-dried leaves of the fresh plant material (445 g) were blended and extracted by maceration with 80% methanol (1.5 L × 24 h × 3) at room temperature (±25 • C). The methanol extract was filtered and evaporated to dryness under reduced pressure at 40 • C. The total extract was concentrated under a vacuum to remove the methanol for freeze- drying. The freeze-dried material was suspended in water and partitioned successively with n-hexane, DCM, EtOAc, and BuOH. Each extract was concentrated to dryness under reduced pressure. Purification and isolation of natural products were achieved through one or a combination of chromatographic techniques.
The EtOAc extract (3.27 g) was pre-adsorbed on silica gel and fractionated on a column by gravity elution using the mixture of DCM elution as follows: 1L of 100%, then 1L volumes of mixtures with EtOAc in the following ratios ( were each pre-adsorbed on silica gel and loaded onto a column for further separation. Fractions F 6 -F 9 were subjected to repeated silica gel columns and eluted with EtOAc: MeOH (95:5) isocratically to afford compound 5 as white crystals (36.0 mg). Fractions F 12 -F 14 were chromatographed on silica gel using 100% EtOAc, isocratically. This resulted in eleven subfractions, from which the combined subfractions five-seven were further purified on a silica gel column eluting isocratically with DCM-MeOH (95:5) to give compound 7 (45.8 mg). Fractions F 16 , F 18 -F 20 , and F 22 were each purified in Sephadex LH-20 using 95% ethanol as the eluent, isocratically. Fraction F 16 yielded compound 2 (20.5 mg), F 18 -F 20 afforded compound 8 (202.6 mg), and compound 3 (18.0 mg) was obtained from F 22 . Fraction F 24 was chromatographed on a silica gel column using the DCM: MeOH (90:10, 85:15) gradient and a total of eight subfractions were obtained, from which four-six were further purified on silica gel to give compound 4 (12.2 mg).

Antityrosinase Inhibition Assay
The skin enzymatic inhibitory assay was executed during the study following the approach used by Curto et al. [39] and Nerya et al. [40]. Extracts and isolates were dissolved in dimethyl sulfoxide (DMSO) to a final concentration of 20 mg/mL. The stock solutions were then diluted to 100 and 200 µg/mL in 50 mM of potassium phosphate buffer (pH 6.5). 'Kojic acid' was used as a control drug [41]. In the wells of a 96-well plate, 70 µL of each sample dilution was combined with 30 µL of tyrosinase (500 units/mL in phosphate buffer) in triplicate. After incubation at room temperature for 5 min, 110 µL of the substrate (2 mM L-tyrosine) was added to each well. Final concentrations of the extracts and isolated samples and positive controls ranged from 0.2 to 1000 µg/mL. Incubation commenced for 30 min at room temperature by measuring the absorbance at 490 nm with the AccuReader M965 Metertech (V1.11). Equation 1 was employed in determining the percentage of tyrosinase inhibition.
where A control is the absorbance of the control with the enzyme, A blank 1 is the absorbance of the control without the enzyme, A sample is the absorbance of the test sample with the enzyme, and A blank 2 is the absorbance of the test sample without the enzyme.

Conclusions
The phytochemical investigation of Protea cynaroides afforded eight compounds, from which compounds 1, 2, 3, and 4 were reported for the first time from a natural source, while 8 was isolated for the first time from Protea cynaroides. Compounds 7 and 8 exhibited potent competitive tyrosinase inhibition against the L-tyrosine substrate, while 6 and 5 demonstrated weak activity. Good anti-tyrosinase activity exhibited by two of these compounds suggests the potential exploration of P. cynaroides in the cosmetic and pharmaceutical industries upon further biological and clinical investigations. This study highlights the importance of P. cynaroides, the national plant of South Africa, as a medicinal plant with therapeutic potential. This is the first scientific report on the bio-evaluation of tyrosinase inhibitory activities of Protea cynaroides.

Data Availability Statement:
The raw data presented in this study are available on request from the corresponding author(s).