Tandem mass spectrometry of aqueous extract from Ficus dubia sap and its cell-based assessments for use as a skin antioxidant

Since 2006, Ficus dubia has been reported as a new Ficus species in Thailand. As per our recent report, the red-brown aqueous extract of F. dubia sap (FDS) has been determined to strongly exhibit in vitro anti-radicals. However, the phytochemicals in the FDS extract related to health-promoting antioxidation have not been explored. Thus, in this study, we aimed to investigate the chemical components of the F. dubia sap extract by liquid chromatography-electrospray ionization quadrupole time-of-flight mass spectrometry (LC–ESI–MS/QTOF-MS) and its potential use in cosmetics in terms of cellular antioxidation on keratinocytes (HaCaT), phototoxicity, and irritation on 3D skin cell models following standard tests suggested by the Organization for Economic Cooperation and Development (OECD). It was found that the sap extract was composed of quinic acid and caffeoyl derivatives (e.g., syringoylquinic acid, 3-O-caffeoylquinic acid, 4-O-caffeoylquinic acid, and dimeric forms of caffeoylquinic acids). The extract has significantly exhibited antioxidant activity against H2O2-induced oxidative stress in HaCaT cells. The cellular antioxidative effect of the FDS extract was remarkably dependent on the presence of 3- and 4-O-caffeoylquinic acid in the extract. Furthermore, the FDS extract showed negative results on skin phototoxicity and irritation. Overall, the results reveal that the FDS extract could be developed as a new antioxidant candidate for a skin healthcare product.

www.nature.com/scientificreports/ a vital role in skin aging by deteriorating the structural proteins like collagen and elastin with activated matrix metalloproteinases 4 . Therefore, the prevention of ROS generation or reduction or delay of ROS degenerative effect can be promising ways for skin protection. Antioxidants are defined as substances counteracting oxidative radicals to terminate the redox chain reaction. Thus, antioxidant agents may help prevent skin damage and premature aging associated with oxidative stress. Ascorbic acid and tocopherols are effective antioxidants primarily used in skin care products 5 . Recently, natural antioxidants are being considered more economical, safe, and effective in treating or preventing ROS-induced skin problems 6 . Natural antioxidants from plant extracts are more abundant with phenolic compounds, such as phenolic acids, flavonoids, stilbenoids, ellagitannins, and phlorotannins. Quercetin and resveratrol, for instance, are in a group of flavonoids, which are well perceived by consumers 7 . The phenolic compounds can effectively scavenge and/or inhibit ROS production, resulting in reduced cellular ROS level 8 . Moreover, phenolic compounds have been shown to suppress intracellular inflammatory reactions 9,10 . Thus, in the cosmetic industry, finding a new natural antioxidant ingredient is a topic of high interest.
However, not much notice has been given to sap, a sticky exudate of tearing plant tissue. Among the sapbearing plant families, Moraceae, especially the genus Ficus, is reported as a source of sap of medicinal importance for its anticancer 11,12 and antihelminthic properties 13 and inhibition of α-amylase and α-glucosidase 14 . One of the most interesting species is Ficus dubia, whose sap turns bloody red after exposure to air. This makes it a costly material like Dragon's blood, which is a captivating crude drug characterized with red resin or sap obtained from Croton spp., Daemonorops spp., Dracaena spp., and Pterocarpus spp. 15,16 .
Ficus dubia is an indigenous plant distributed in the southern part of Thailand, Malaysia, Indonesia, and Brunei 17 . It is believed that the consumption of F. dubia sap would have health benefits. In fact, Suttisansanee et al. 18 found that the extract of the F. dubia sap was rich in phenolics and flavonoids. The extract also promoted in vitro antioxidant activities determined by DPPH, ABTS, ferric reducing ability of plasma (FRAP), and oxygen radical absorbance capacity (ORAC) assays. Further, it was determined to also exhibit an antidiabetic mechanism via α-glucosidase inhibitory activity similar to the action of acarbose, which is a standard antidiabetic drug used to treat type 2 diabetes mellitus and prediabetes. Interestingly, both root and sap of the F. dubia were not genotoxic, as they did not present mutagenicity on a Drosophila wing-spot test. Antioxidant activities of the F. dubia sap seemed to be associated with its phenolic contents. However, only caffeic acid was found in the hydrolyzed sap. Thus, caffeic acid derivatives could be key components representing antioxidant activities in the F. dubia sap.
To develop the F. dubia sap as one of the cosmeceutical ingredients, in this work, we further investigated a chemical profile of the sap extract by using liquid chromatography-electrospray ionization quadrupole time-offlight mass spectrometry (LC-ESI-MS/QTOF-MS); we further studied a cellular antioxidant activity of F. dubia sap extract on keratinocytes. Furthermore, phenolic acids in the extract were quantified and tested for antioxidant effects on keratinocytes. Additionally, we confirmed the safety of the extract using 3D skin models. We hope that this experiment can guide us to understand how F. dubia sap extract can be used as a skin antioxidant ingredient.

Results
Aqueous extract of F. dubia sap and its total phenolic content. In total, around 7.67% w/w of F. dubia sap (FDS) extract was obtained from the F. dubia sap. Its appearance was observed as a red-brown solid and freely soluble in water. The FDS extract contained a high number of phenolics, about 300.9 ± 9.6 mg gallic acid equivalence per gram extract (mg GAE/g extract).
The MS fragmentation and proposed structure of all compounds in the FDS extract were shown in Supplementary Data (Figures S1 and S2). www.nature.com/scientificreports/ Quantification of quinic acid and 3-and 4-caffeoylquinic acids in F. dubia sap extract. Quinic acid was quantified using LC-ESI-MS (Fig. 4), while 3-and 5-O-caffeoylquinic acids were determined using the HPLC method (Fig. 5). The validation parameters are presented in Table S1 (see Supplementary Material). It was found that the FDS extract in this study contained 1.32 ± 0.15% w/w quinic acid (QA), 0.39 ± 0.01% w/w 3-O-caffeoylquinic acid (3-CQA), and 0.34 ± 0.00% w/w 4-O-caffeoylquinic acid (4-CQA). Unfortunately, this  www.nature.com/scientificreports/   www.nature.com/scientificreports/ and dimeric forms of caffeoylquinic acid (compounds 9 and 11). Therefore, isolation technique is needed to further separate those compounds. The concentrations of QA, 3-CQA, and 4-CQA presented in the FDS extract were used to study the cellular antioxidation in keratinocytes in the next experiment. The FDS extract at 0-500 µg/mL, QA, and chlorogenic acids at equivalent concentrations presented in the FDS extract significantly restored cell viability upon oxidative stress (Fig. 6A). Moreover, they rather scavenged the ROS in keratinocytes, as shown in Fig. 6B. The extract at concentrations of 7.81 and 15.83 μg/mL has significantly decreased intracellular ROS to a level similar to the control group. An increase in the extract concentration up to 500 μg/mL further dramatically decreased the intracellular ROS.

Cellular antioxidation activities of
The cytoprotective effects of chlorogenic acid derivatives were also observed at concentrations existing in the extract at 0-500 µg/mL. Based on chemical quantification, the derivatives were observed in ranges of 0-6.60 μg/ mL QA, 0-1.95 μg/mL 3-CQA, and 0-1.70 μg/mL 4-CQA, respectively. The ROS scavenging activity of the FDS extract was likely related to its phytochemical compositions, such as 3-CQA and 4-CQA, rather than QA. The cytoprotective dose of the extract was 125-500 µg/mL, which contained about 0.49-1.95 µg/mL of 3-CQA and 0.43-1.70 µg/mL of 4-CQA. At these concentrations, 3-CQA and 4-CQA were noted to have significantly exhibited the depleting effect on intracellular ROS without toxicity on keratinocytes. Among the test compounds, the protective effect was marked on 3-CQA at 1.95 µg/mL, which exhibited the highest decrease in intracellular ROS to 43.49 ± 8.61%. Accompanied by 3-CQA and 4-CQA, our results suggest that 3-CQA and 4-CQA can influence the intracellular decreasing manner of the extract in higher concentrations. www.nature.com/scientificreports/ 3T3 neutral red uptake phototoxicity test of F. dubia sap extract. The phototoxicity potential of the FDS extract using the 3T3 neutral red uptake test is presented in Table 3. The half-maximal inhibitory concentration (IC 50 ) of chlorpromazine hydrochloride (CPZ) upon non-UV-and UV-irradiated conditions was accepted, regarding the recommendation of the OECD TG 432 (see Table 3). The IC 50 of CPZ should be in ranges of 7.0-90.0 μg/mL without UV irradiation and 0.1-2.0 μg/mL after UV-irradiated condition. The mean photo effect (MPE) and photo-irritation-factor (PIF) values of CPZ also showed phototoxicity. It suggested that the experimental procedure used in this study was validated. The phototoxicity analysis of the FDS extract showed no cytotoxicity whether the cells were irradiated or non-irradiated by UV. Furthermore, the IC 50 of the FDS extract was observed to be above 1000 µg/mL. Thus, the IC 50 was unable to determine the phototoxicity of the FDS extract. In this case, the MPE value can be used instead for predicting the phototoxicity of the FDS extract. In this study, the MPE value of the FDS extract was 0.02 ± 0.01 (see Table 3). Thus, it can be considered that the FDS extract has no phototoxicity.  . The 3D skin model remained 100.00 ± 4.20% viability when the tissue was treated with a buffer. In comparison, the tissue showed irritation effect by a dramatic decrease in cell viability to nearly 0 after 5%sodium dodecyl sulfate (SDS) was introduced. Interestingly, the FDS extract at a concentration of 830 mg/mL did not cause skin irritation as the tissue viability was 92.94 ± 4.03% (Fig. 7).

Discussion
Ficus dubia is a Ficus plant distributed in a limited area in the southern forest of Thailand. A study conducted by Suttisansanee et al. 18 18 . The number of total phenolics controlled the quality of the FDS. In this study, total phenolic content (TPC) found in the FDS extract was closed as previously reported 18 . The F. dubia sap was determined to be abundant with phenolic compounds as commonly found in Ficus species and reported as major phytochemicals contributing to antioxidant activities. Most studies on TPC and antioxidant activities were reported on the fruit and latex of F. carica (common fig) [19][20][21][22] . The TPCs of figs were determined, depending on cultivars and extraction methods 22 . In cases of aqueous and alcoholic extractions, the skin of fresh fig, cultivar "VB1, " contained the total phenolic content of about 140 mg/100 g fresh weight 20 . The "White Genoa" F. carica cultivar showed about 300 μg GAE/mL. It revealed that the sap of Ficus is a good source of phenolics. The high content of phenolics found in FDS extract was further investigated for its identified components.
The phenolic compounds in the FDS extract were determined in our study. It mainly comprised of QA and its acyl derivatives, that is, chlorogenic acids. Such a high amount of QA found in the extract suggested the phenolic derivatives are from the shikimate pathway. It is also possible that QA could be a metabolite used for biosynthesis of chlorogenic acids. QA was found in the leaves of Portuguese F. carica varieties "Pingo de mel" and "Branca Tradicional", containing 10,502.5 ± 97.1 mg/kg and 5183.3 ± 116.5 mg/kg of lyophilized extracts, respectively. In addition, citric acid (m/z 191.0196) was found in the FDS extract, a similar component in F. carica latex 23   www.nature.com/scientificreports/ the phenolic substances presented in the latex; however, they were not exactly identified. Two research works by Oliveira et al. 23,25 revealed the latex metabolites of F. carica, which are phytosterols (especially β-sitosterol), terpenoids, fatty acids, amino acids, and organic acids. It was reported on an antioxidant against DPPH · at IC 25 of 1049 μg/mL. Additionally, it is interesting to note that QA and its acyl derivatives (chlorogenic acids) in Ficus dubia sap were found in high amounts as similar to those in Lecokia cretica and Inula graveolens extracts, which are known for their antioxidant activities 26,27 . The epidermis is the outer skin layer. Keratinocytes, which are the major cells in the epidermis, are often deteriorated by oxidants resulting from physiological processes and harmful stimuli, such as UV radiation and particulate matter pollution 4 . Thus, keratinocytes are a target for skin delivery of antioxidants to nourish and maintain their integrity. Several natural compounds, such as dendropachol 28 and ellagic acid 29 , are known to exhibit protective effects against oxidative stress. This study found that the FDS extract demonstrated a dosedependent manner in decreasing intracellular ROS and maintained almost 90% cell survival compared to the control. The FDS extract at concentrations of 7.80 and 15.63 μg/mL recovered the cell viability (up to 95%) and decreased the intracellular ROS to the normal condition (nearly 100%). The ROS reduction and cell survival patterns of FDS extract reflected the cell protective tendency of the extract influenced by QA, 3-CQA, and 4-CQA. Compared to 3-CQA, QA exhibited weak antioxidant activities on in vitro antioxidant assays underlying mechanism of single electron transfer (SET) such as Trolox equivalent antioxidant capacity (TEAC) and FRAP because of its lack of extended π-system to stabilize the radical cation 30 . Results demonstrated that 3-CQA and 4-CQA were superior to QA in terms of reducing intracellular ROS (p < 0.05). In concordance with the results studied by Gul et al. 31 , chlorogenic acid at a concentration of 0.11 ± 0.01 μM exhibited 50% inhibition of ROS induced by H 2 O 2 in rat cortical slices. Moreover, many studies also claimed the protective effect of chlorogenic acid on cells such as osteoblast (MC3T3-E1) and lens epithelial (HLE-B3) cells from H 2 O 2 -induced oxidative stress 32,33 .
Evaluation of FDS extract on skin toxicity is necessary before using it as an active ingredient in antioxidant skin products. There are some case reports on phytophotodermatitis after getting in contact with Ficus sap. This skin toxicity can be attributed to the presence of photoactive furocoumarins (e.g., psoralen and bergapten) in the Ficus sap 34,35 . However, results demonstrated that FDS extract did not show phototoxicity and skin irritation regarding the 3T3-neutral red uptake and 3D skin tests, respectively. Our extraction process using DI water as an extraction solvent may contribute to the nontoxic properties of the extract. Moreover, the mass adducted of possible furocoumarins were not found in the extract analyzed via LC-ESI-MS/QTOF-MS. Thus, these furocoumarins, which are hydrophobic, were excluded from the extract. Hence, FDS extract was able to pass the preliminary tests, confirming its safety to the skin. Alternatively, amine compounds were found in FDS (Table 2). N-Methyltyramine (8) and hordenine (10), which are biogenic amines found in various plants, were reported as toxic compounds on animals 36 , but not specifically related to human skin.
Chlorogenic acids demonstrate direct antioxidant activity in human diets; these days, they are also used for skin protection in skincare and cosmetic products. Coffee beans are known as a source of chlorogenic acids comprising of 5-CQA as the main component 37 . In this paper, we suggest FDS extract as a new source of chlorogenic acid derivatives, which is enriched with 3-and 4-COA. Based on our results, the FDS extract could be a promising new antioxidant use in cosmetics. Briefly, the F. dubia sap is harvested by slicing 15-20 cm and 15°-25° angle of a groove into the tree's cambium with a hooked knife. The grooves were made until 1 L of the sap was collected. The sap (1000 g) was lyophilized for 48 h. Then, the dried sample was re-dissolved with deionized water to remove resinous and other water-insoluble substances. The solution was slowly stirred by a magnetic stirrer for an hour and then filtered through Whatman filter paper No. 1. The red-brown filtrate was then collected and lyophilized for 48 h. Then, the aqueous FDS extract powder (76.7 g) was obtained and kept in an aluminum foil bag and stored at − 20 °C for further studies. www.nature.com/scientificreports/ Phytochemical analysis. Determination of TPC in the F. dubia sap extract by spectrophotometer. The TPC of the FDS extract was determined using the Folin-Ciocalteu method 38 . The extract (20 µL) at different concentrations (12.5, 25, 50, 100, 200, and 400 µg/mL) was mixed with 100 µL of 10% Folin-Ciocalteu solution and 80 µL of 7.5% sodium bicarbonate solution in a 96-well plate. The mixture was gently mixed and stored in the dark for 30 min. Later, the absorbance of mixture solutions was measured at 765 nm by spectrophotometer (SpectraMax, Molecular Devices, USA).Total phenolic contents were calculated using the calibration curve of gallic acid. The values were expressed as mg gallic acid equivalent per gram of extract (mg GAE/g extract).

Material and methods
Phytochemical screening by liquid chromatography-electrospray ionization quadrupole time-of-flight mass spectrometry (LC-ESI-MS/QTOF-MS). A protocol of LC-ESI-MS/QTOF-MS was modified from research works of Nuengchamnong et al. 39 and Duangjai et al. 40 . Briefly, the LC-ESI-MS/QTOF-MS system consists of HPLC unit 1260 infinity series (Agilent Technologies, Waldbronn, Germany) coupled with an ultra-high-definition accurate mass spectrometer (Agilent Technologies, Singapore). Chromatographic separation of the FDS extract (concentration 20 mg/mL of water) was conducted on a Luna C18(2) column (150 × 4.6 mm, 5 μm, Phenomenex, USA) at a flow rate of 0.5 mL/min. The column temperature was maintained at 35 °C. The mobile phase consisted of A (0.1% formic acid in type I water, v/v) and B (0.1% formic acid in acetonitrile, v/v). The elution gradient was started from 5 to 20% B for 5 min, then changed to 50% B in 10 min, and increased to 80% B in 2 min, and held for 3 min. The total run was 15 min, and the post-run was for 5 min. The injection volume was 10 mL using an autosampler. The dual electrospray ionization (ESI) source was operated in both negative and positive modes. The ESI condition was as follows: drying gas (gas N 2 ) temperature, 350 °C; gas flow rate, 10 L/min; nebulizer pressure, 30 psig; mass range, 100-1000 m/z; scan rate, 4 spectra/s; capillary voltage, 3500 V; skimmer voltage, 65 V; octapole RFV, 750 V; and fragment voltage, 50 V. The automatic fragmentation pattern was set with collision energy at 10, 20, and 40 V using UHP N 2 gas. Accurate mass measurements (error < 5 ppm for analytes) were obtained through an automated calibrant delivery system on a daily basis using a dual-nebulizer ESI source Quantification of QA and its derivatives. A quantitative analysis of QA was conducted using an Agilent LC-MS system (Agilent 1290 Infinity Series, Waldbronn, Germany). The chromatogram of QA was conducted using Agilent Poroshell 120 (2.1 mm × 100 mm, 2.7 μm). The mobile phase consisted of (A) deionized water and (B) acetonitrile: 5-mM phosphate buffer (80:20). The flow rate of the mobile phase was kept at 0.5 mL/min. The gradient elution starting from the low content of A was as follows: 0-7 min, gradient mixture from 0% A:100% B to 6% A:94% B; 7-10 min, and the gradient mixture 6% A:94% B to 0% A:100% B. Later, it was held at 0% A for 5 min. The injection volume was 5 µL. The mass spectrometric analysis was performed using an Agilent 6495 Triple Quadrupole operating in negative ionization mode. The mass spectrometer condition was as follows: drying gas temperature, 225 °C; gas flow rate, 15.1 L/min; nebulizer pressure, 25 psi; mass range, 100-100 m/z; capillary voltage, 3000 V; skimmer voltage, 65 V; fragment voltage, 380 V; and collision energy 10, 20, 40, and 60 V. As for QA derivatives, a high-performance liquid chromatography system (Water Alliance 2695, USA) equipped with a photodiode array detector was used for quantification of 3-O-caffeoylquinic acid (3-CQA) and 4-O-caffeoylquinic acid (4-CQA). A reversed-phase column, Luna ® C18 100 Å (4.6 mm × 150 mm, 5 µm), Phenomenex ® (USA) was used as the stationary phase. The mobile phase consisted of 0.1% formic acid in acetonitrile (A) and 0.1% formic acid in deionized water (B), which was programmed as follows: 0-15 min, gradient elution from 5A:95B to 20A:80B; then constantly held at 20A:80B for 2 min; 17-20 min; gradient running from 20A:80B to 5A:95B; then constantly held at 5A:95B for 4 min. The chromatographic conditions were set as follows: flow rate, 0.5 mL/min; injection volume, 10 µL; column temperature, 35 °C; and detection wavelength, 330 nm.

Cellular antioxidation capacities of FDS extract on keratinocytes. Cell viability of human keratino-
cytes (HaCaT). The HaCaT cells were seeded in a 96-well plate (20,000 cells per well) and cultured in 100-µL DMEM containing 10% FBS and 1% penicillin-streptomycin under 37 °C in a 5% CO 2 incubator for 24 h. Next, the cells were pretreated with various concentrations of FDS extract (0-500 µg/mL), QA (0.10-6.60 µg/mL), 3-CQA (0.03-1.95 µg/mL), and 4-CQA (0.03-1.70 µg/mL) and incubated for the next 24 h. After that, the cells were challenged with or without 100-µM hydrogen peroxide (H 2 O 2 ) in the incubator for 30 min. The cells were then incubated with an MTS reagent for 2 h. Finally, the absorbance was measured at the wavelength of 490 nm using a microplate reader (Spectramax, Molecular Devices, USA). The experiments were conducted in triplicate.
Determination of intracellular ROS. Intracellular ROS was measured using DCFH-DA, as previously described by Chansriniyom et al. 41 . The DCFH-DA was hydrolyzed to 2′,7′-dichlorodihydrofluorescein (DCFH) by cellular esterase. Then, DCFH reacts with intracellular ROS, leading to the formation of the fluorescent 2′,7′-dichlorofluorescein (DCF). HaCaT cells were seeded in a 96-well black plate containing 100-µL DMEM at a density of 20,000 cells per well for 24 h. Then, the cells were treated with the samples for 24 h. Before adding 10 µL of 10 µM DCFH-DA, the cell medium was removed, and the cells were rinsed with PBS. After 30 min of incubation, the  42 . Briefly, the cells were seeded in 96-well plates at 10 4 cells per well and cultured at 37 °C, 5% CO 2 for 24 h. Then, the cells were pretreated with eight concentrations of FDS extract in 1X Hanks'Balanced Salt Solution (HBSS) in a range of 7.8-1000 µg/mL, and chlorpromazine hydrochloride (CPZ) was used as a positive control. After 60 min, the cells were irradiated with 5 J/cm 2 of UVA, while the control plate was kept in the dark. After irradiation, the cells were washed twice with 1X HBSS and added a fresh medium to each well before being incubated for 18 h. Then, cell viability was determined using the neutral red uptake assay by measuring the absorbance at 540 nm. The half-maximal inhibitory concentration (IC 50 ) of crude extract and CPZ in irradiated and non-UV-irradiated conditions, MPE, and photo-irritation-factor (PIF) were calculated using Phototox version 2.0 software, as per the OECD guidelines.
Skin irritation. The acute skin irritation caused by FDS extract was investigated on the reconstructed human epidermis (EpiDerm™, MatTek, USA) according to OECD (TG No. 439) (OECD, 2020) 43 . Briefly, the 3D skin model of the human epidermis was constructed by incubating the cells in the culture medium overnight at 37 °C with 5% CO 2 and 90% relative humidity. Then, 30 µL of 830 mg/mL of the FDS extract was applied to the EpiDerm™ and incubated for an hour. The 5% SDS and sterile DPBS were used as positive and negative controls, respectively. After that, the 3D skin model was washed with PBS to remove the crude extract and further cultured in the culture medium for 18 h. The tissue viability was determined by MTT assay. MTT solution was incubated in the 3D skin model for 3 h; then, formazan was extracted, and measured the absorption at wavelengths of 570 and 650 nm. The extract solution is determined as a blank. Percentage tissue viability was calculated using the following formula: Statistical analyses. The experimental results were given as mean ± standard deviation (SD) of the three independent experiments. The data were analyzed via one-way ANOVA, followed by Tukey's test (IBM, SPSS statistic version 22, USA). A significant difference is accepted when the p-value is less than 0.05.

Conclusion
Phytochemicals in the aqueous extract of F. dubia sap were identified by LC-ESI-MS/QTOF-MS, revealing QA, syringoylquinic acid, 3-O-caffeoylquinic acid, 4-O-caffeoylquinic acid, and dimeric forms of caffeoylquinic acids. In the extract, the amount of QA, 3-O-caffeoylquinic acid, and 4-O-caffeoylquinic acid was 1.32 ± 0.15% w/w, 0.39 ± 0.01% w/w, and 0.34 ± 0.00% w/w, respectively. Furthermore, the radical scavenging activity of the F. dubia sap extract on keratinocytes was relatively correlated with 3-O-caffeoylquinic acid and 4-O-caffeoylquinic acid rather than QA. Furthermore, the F. dubia sap extract was determined to be safe for skin cells. In summary, the FDS extract is a promising extract that can be further developed as a new antioxidant ingredient for a skin healthcare product.