Evaluation of In Vitro Solar Protection Factor (SPF), Antioxidant Activity, and Cell Viability of Mixed Vegetable Extracts from Dirmophandra mollis Benth, Ginkgo biloba L., Ruta graveolens L., and Vitis vinífera L.

The aim of this study was to validate a HPLC method for the assay of flavonoids in extracts obtained from natural sources, i.e., from Dirmophandra mollis Benth, Ginkgo biloba L., Ruta graveolens L., and Vitis vinífera L. The potential sun protecting effect, antioxidant activity, and cell viability of the extracts were also determined. Individual extracts (obtained from each individual species) and a mixed extract (containing the four extracts) were analyzed by the validated HPLC method for the identification of flavonoids and quantification of rutin and quercetin. An in vitro cell viability study was carried out using the neutral red method. The in vitro sun protection factor was determined by spectral transmittance and in vitro antioxidant efficacy was evaluated against DPPH, ABTS, and AAPH radicals. The HPLC method used for the identification and quantification of flavonoids in extracts exhibited linearity, precision, accuracy, and robustness. Detection and quantification limits were, respectively, 2.881 ± 0.9 μg·mL−1 and 0.864 ± 0.9 μg·mL−1 for quercetin, and 30.09 ± 1 μg·mL−1 and 9.027 ± 1.1 μg·mL−1 for rutin. All extracts did not affect cell viability at the evaluated concentration range and exhibited a sun protection effect and antioxidant activity. Among the evaluated extracts, Ginkgo biloba L. and the mixed extract depicted the most expressive antioxidant activity. The mixed extract exhibited sunscreen protection against ultraviolet A (UVA) and ultraviolet B (UVB) and a critical wavelength of 372.7 ± 0.1. Our results translate the enhanced flavonoids’ composition of the mixed extract, which may be a potential alternative over sunscreens and antioxidants in pharmaceutic/cosmetic formulations.

The precision assay was performed using intra-day and inter-day repeatability [30,31]. Six samples (analytical standard) with a concentration of 50 µg·mL −1 to quercetin and 500 µg·mL −1 to rutin were analyzed on the same day and on two consecutive days. Areas of the standards' peaks were obtained and variation coefficient percentage was calculated. The mixed sample was also subjected to precision assay for reliability of results.
For robustness determination, six concentrations of mixed sample were analyzed and the analyzer was varied for data comparison [30].
Detection and quantification limits (DL and QL, respectively) were evaluated to determine and quantify the lowest acceptable concentration of quercetin and rutin in extracts [30]. Therefore, rutin and quercetin solution (analytical standards) were prepared containing low concentration, and DL and QL were calculated using Equations (1) and (2).
where SD is standard deviation of intercept with the Y axis of at least three analytical curves constructed, and SC is the slope of analytical curve.
To determine the method selectivity, chromatograms of extracts were compared with analytical standards chromatograms to determine impurities in the extracts [30].

Cell Viability Assay
The in vitro cell viability assay was performed as described by Stokes et al. [32] and OECD [33]. The HaCaT (4 × 104 cells·mL −1 ), in 96-well plates (100 µL cells·well −1 ), were exposed to the samples' final concentrations, in triplicate, for 48 h. The final DMSO concentration (≤0.25%) did not affect cell viability. Doxorubicin chloride (0.5 µg·mL −1 ) was used as positive control. After 48 h of exposure, medium was removed and replaced by neutral red work solution (200 µL·well −1 ). Cells were incubated for 3 h, followed by supernatant removal and the addition of ethanol/acetic acid solution (1.0%; v/v) (100 µL·well −1 ). Absorbance values were read at 540 nm using a spectrophotometer (Versamax, Molecular Devices, São Paulo, Brazil). Concentration-response curves for each sample were plotted using GraphPad Prism (version 5.02) software for Windows (GraphPad Software, San Diego, CA, USA).

In Vitro Sun Protection Factor Evaluation
The in vitro sun protection factor was determined by the ultraviolet-visible spectrophotometry method described by Mansur et al. [34]. Spectrophotometric readings were obtained for each extract (100 µg·mL −1 ) at 290-320 nm and SPF values were determined using Equation (3): where SPF stands for solar protection factor; CF for correction factor; EE (λ) is the erythemogenic effect of wavelength radiation (λ) nm, which was previously calculated by Sayre et al. [35]; I (λ) is the intensity of solar radiation in the wavelength (λ) nm; and Abs (λ) is the spectrophotometry reading of the absorbance of sunscreen solution in the wavelength (λ) nm.

Statistical Analysis
All assays were performed in triplicate. Statistical analysis was performed using an ANOVA test (p < 0.05) for independent variables, Origin version 8 and Graph Pad Prism (version 5.02) software for Windows (GraphPad Software, San Diego, CA, USA).

Results and Discussion
After collecting plant material, plants were dried, milled, and presented physical chemical parameters according to Table 1. All samples presented acceptable plant material parameters [28], ensuring quality.  Dry extracts obtained from plants were subjected to identification reaction assays following the guidelines of the Brazilian Pharmacopoeia [28]. All extracts showed rose color in Shinoda and Pew reactions and yellow fluorescence in Taubock reaction. They also exhibited fluorescence under UV light and brown color when subjected to aluminum and ferric chloride, respectively, highlighting the presence of flavonoids in the extracts [40].
The HPLC method was validated for the quantification of flavonoids, for which the parameters of specificity/selectivity, linearity, precision, sensitivity (detection limit and quantification limit), accuracy, and robustness were determined.
The linearity assay was determined for rutin (y = 19862x + 26548; R 2 = 0.9997; Figure S1) and quercetin (y = 35758x − 58195; R 2 = 0.9996; Figure S2) analytical curves. For the precision test, the area of the peaks obtained by HPLC of quercetin and rutin standards and these flavonoids in the mixed sample are displayed in Table 2 (intra-day and inter-day assay). Variation coefficients less than 5.0% were obtained, showing the precision of the method [30]. During precision assay, robustness of the method was determined by changing the analyzer. Since a variation coefficient less than 5% was obtained, the quantification method was considered robust.
A recovery assay was performed to determine the accuracy of the method [30]. The percentage of quercetin and rutin recovered (R%) was calculated and results demonstrate the accuracy of the method, as the mean recovery rate was close to 100.0% and variation coefficient less than 5.0% (Table 3). Detection and quantification limits (DL and QL) were calculated from Equations (2) and (3), achieving the following values: DL = 0.86 ± 0.91 µg·mL −1 (quercetin) and 9.02 ± 1.12 µg·mL −1 (rutin); QL = 2.88 ± 0.92 µg·mL −1 (quercetin) and 30.09 ± 1.01 µg·mL −1 (rutin). To determined selectivity of the method, the extracts' chromatograms were compared to standard chromatograms [30,41]. The extracts and mixed sample did not exhibit impurities or other compounds capable to interfere in identification of quercetin and rutin peaks.
According to Stokes et al. [32], the evaluation of neutral red uptake is directly proportional to living cell number [33]. In our study, in all tested concentration, the extracts and the mixed sample reduced HaCat cell viability to less than 50% ( Figure 6). Thus, the concentration required to reduce by 50% the cell viability (IC 50 ) was higher than 200 µg·mL −1 . We did not test higher concentrations of the selected extracts to avoid the production of artefacts in culture medium, as already described for many phenolic compounds [27]. Spectrophotometry in the ultraviolet region is an adjuvant and preliminary in vitro method to evaluate sun protection factor of compounds, especially from vegetal sources [34,44,45]. Thus, the four extracts (100 μg.ml −1 ) and mixed sample were subjected to spectrophotometric analysis. D. mollis Benth presented an SPF value of 5.04 ± 0.2, G. biloba L. of 8.31 ± 0.5, R. graveolens L. of 7.08 ± 0.4, V. vinifera L. of 3.71 ± 0.5, and the mixed sample (1:1:1:1) of 7.72 ± 0.4.
As determined by cell viability assay, both SPF assays were performed in concentration up to 200 μg.ml −1 that was determined as the highest non-cytotoxic concentration tested in our work. These preliminary results show that G. biloba and R. graveolens extracts were the most promising extracts, besides the mixed sample. To confirm the SPF results, all samples were then evaluated using spectral transmittance [46,47]. All individual extracts, the mixed sample, and the positive control (Tinosorb S TM ) presented absorption in UVA and UVB regions (Table 4) in different ways. While D. mollis and G. biloba extracts absorbed in the 320-400 nm range, corresponding to UVA radiation, R. graveolens and V. vinifera extracts absorbed around 310 nm, corresponding to UVB radiation. Moreover, the mixed sample and Tinosorb S TM absorbed in a higher range, configuring protection in UVA and UVB regions [6]. G. biloba L. extract, followed by the mixed sample, presented the highest SPF values (Table  4), attributed to the higher flavonoid concentration. Moreover, D. mollis Benth and V. vinifera L. extracts that showed the lowest quercetin concentration presented low values, corroborating literature data [48]. Spectrophotometry in the ultraviolet region is an adjuvant and preliminary in vitro method to evaluate sun protection factor of compounds, especially from vegetal sources [34,44,45]. Thus, the four extracts (100 µg·mL −1 ) and mixed sample were subjected to spectrophotometric analysis. D. mollis Benth presented an SPF value of 5.04 ± 0.2, G. biloba L. of 8.31 ± 0.5, R. graveolens L. of 7.08 ± 0.4, V. vinifera L. of 3.71 ± 0.5, and the mixed sample (1:1:1:1) of 7.72 ± 0.4.
As determined by cell viability assay, both SPF assays were performed in concentration up to 200 µg·mL −1 that was determined as the highest non-cytotoxic concentration tested in our work. These preliminary results show that G. biloba and R. graveolens extracts were the most promising extracts, besides the mixed sample. To confirm the SPF results, all samples were then evaluated using spectral transmittance [46,47]. All individual extracts, the mixed sample, and the positive control (Tinosorb S TM ) presented absorption in UVA and UVB regions (Table 4) in different ways. While D. mollis and G. biloba extracts absorbed in the 320-400 nm range, corresponding to UVA radiation, R. graveolens and V. vinifera extracts absorbed around 310 nm, corresponding to UVB radiation. Moreover, the mixed sample and Tinosorb S TM absorbed in a higher range, configuring protection in UVA and UVB regions [6]. G. biloba L. extract, followed by the mixed sample, presented the highest SPF values (Table 4), attributed to the higher flavonoid concentration. Moreover, D. mollis Benth and V. vinifera L. extracts that showed the lowest quercetin concentration presented low values, corroborating literature data [48]. In addition, according to the literature [16,49,50], the SPF values found in the extracts studied in our work were lower than chemical sun filters, such as Tinosorb S TM .
The colorimetric evaluation shows that all individual extracts have flavonoids and they can therefore be considered a promising plant sources to be used as sunscreen.
Antioxidant activity of phenolic compounds such as flavonoids is widely known, and thus, flavonoids are widely studied as ingredients in cosmetic formulations against early skin aging by scavenging reactive oxygen species produced by sun radiation [4,22,51].
All of the four extracts, together with the mixed sample, were subjected to in vitro antioxidant assays. From the DPPH and ABTS experiments, the results were expressed as the sample concentration required for 50% reduction of the radical concentration, while on ORAC protocol, the ability of scavenger peroxyl radicals was expressed as the equivalent concentration of Trolox [39]. Then, IC 50 values were calculated, and once again, the best results were seen for G. biloba L. extract and the mixed sample (Table 5), and this can be attributed to the higher flavonoid concentration in these samples. Table 5. In vitro antioxidant evaluation against DPPH, ABTS, and AAPH free radicals of some potential sunscreen natural products. Values are presented as an average of three measurements and standard deviation (±SD). Based on these results, the extracts and the mixed sample presented antioxidant activity against DPPH, ABTS, and AAPH radicals, which is indicative of premature aging protection. Then, considering the samples' potential as sunscreens, all extracts and mixed samples can be incorporated into a cosmetic formulation, aiming to develop a new sunscreen containing chemical sun filter from plant material.

Conclusions
This work is the first study about the sun protection action of Dimorphandra mollis Benth and Ruta graveolens L. All flavonoid-enriched extracts were not cytotoxic, and presented antioxidant activity and sun protection factor, as shown by in vitro methods. The mixed sample composed by the four studied plants presented promising results. The quantification method exhibited linearity, precision, accuracy, robustness, and did not exhibit impurities or other compounds capable of interfering in the identification peak of flavonoids. The mixed sample may be an alternative to treat deleterious effects from exposure to ultraviolet radiation and a promise as a potential sunscreen.