Screening of Panamanian Plants for Cosmetic Properties, and HPLC-Based Identification of Constituents with Antioxidant and UV-B Protecting Activities

A library of 600 taxonomically diverse Panamanian plant extracts was screened for DPPH scavenging and UV-B protective activities, and the methanolic extracts of Mosquitoxylum jamaicense, Combretum cacoucia, and Casearia commersionia were submitted to HPLC-based activity profiling. The compounds located in the active time windows were isolated and identified as gallic acid derivatives and flavonoids. Gallic acid methyl ester (3) and digallic acid derivatives (2, 6) showed the highest DPPH scavenging activity (<10 μg/mL), while protocatechuic acid (7) and isoquercitrin (10) exhibited the highest UV-B protective properties.


Introduction
The skin is the largest organ of the human body, functioning as an effective barrier against the harmful effects of the environment [1]. Several factors affect skin health and promote skin aging, such as ionizing radiation, severe physical and psychological stress, alcohol intake, poor nutrition, overeating, environmental pollution, and exposure to UV radiation. The latter is believed to contribute up to 80% of extrinsic skin damage [2].
In cosmetics, natural products play a major role as active ingredients given that they are considered by many as safer alternatives to synthetic products and, therefore, possess higher consumer acceptance. Numerous cosmetic products for dry skin, skin protection (ROS, radicals, and UV light), prevention or alleviation of skin inflammation, hyperpigmentation, and anti-aging products are commercially available [3][4][5].
Free radical formation can induce skin damage through a series of mechanisms leading to cell death and ultimately, to skin aging. In a search for new active ingredients for skin care products, compounds and extracts of natural origin are of significant interest [6]. The potential of purified plant compounds in skin protection is generally recognized, but plant extracts also show significant potential due to their complex composition [7].
In the field of cosmetic ingredients, relatively few studies on novel plant extracts or pure natural products have been published in recent years, and the majority of these studies were linked to ethnobotanical sources [3]. Screening of taxonomically diverse and unique plant collections is an alternative strategy to an ethnobotany-driven approach, and it has been successfully applied in the drug discovery field [8]. A diversity-oriented approach is the most successful if plants from regions of high biodiversity can be accessed. Panama is located in one of the 25 biodiversity hotspots worldwide. Despite the small surface of the country, its flora comprises 9,893 vascular plant species including 1,327 (13.4%) endemic plants [9][10][11]. The flora of Panama is a rich source of bioactive molecules and represents a largely untapped source for new compounds with promising activities for pharmaceutical, agrochemical, and cosmetic industries [12][13][14].
In an FP7 framework project aiming at the discovery of new natural products for cosmetic use, we screened a library of 600 extracts generated from a set of taxonomically diverse Panamanian plants. The focus was on the identification of plants with promising UVprotective and anti-aging properties. The best extracts were submitted to a process termed HPLC-based activity profiling [15], whereby physicochemical data recorded online are combined with bioassay data of HPLC microfractions.

Results and Discussion
A library of 600 extracts prepared from Panamanian plants was screened for antioxidant capacity and the ability to protect human skin fibroblasts against UV-B-induced cell death. The screening results of the selected extracts are given in Table 1, and a flow chart for the further progression of samples is shown in Fig. 1. A total of 19 extracts were found to possess considerable radical scavenging activity, i.e. IC 50 ≤ 30 μg/ml in the DPPH assay. These extracts were screened for their ability to protect human skin fibroblasts against UV-B-induced cytotoxicity, and three extracts were found to reduce UV-B-induced cell death to ≤15% of the control.

Tab. 1.
Activity data of selected extracts in DPPH and UV-B protection assays Active extracts were then submitted to HPLC-based activity profiling [15] in order to track the active constituents in the extract. Time-based microfractions were collected and submitted to screens. HPLC traces and activity profiles are shown in Fig. 2. Extracts were then prioritized on the basis of HPLC traces and activity profiles. In the case of the MeOH extract of Casearia commersionia (Salicaceae) ( Fig. 2A), a broad window of activity corresponded to a broad hump in the baseline of the HPLC chromatogram. This was a strong indicator for the presence of tannins, and the extract was therefore excluded from the follow-up. In contrast, for the MeOH extracts of Mosquitoxylum jamaicense (Anacardiaceae) (Fig. 2B) and Combretum cacoucia (Combretaceae) (Fig. 2C), the activity profile correlated with discrete peaks in the chromatograms, even though broad humps in the baseline were also indicative of tannins. These two extracts were selected for characterization of the active constituents. were submitted to further purification by HPLC. Peak 1 was identified as gallic acid (Fig.  3), by spiking with a commercial reference and by NMR spectroscopy. Given that the radical scavenging and antioxidant properties of gallic acid are known [20], the compound was not pursued further. The other two early-eluting peaks were identified as a 7:3-mixture of meta-and para-digallic acid (2) [21] and a gallic acid methyl ester (3) [22]. Both compounds were found to possess good radical scavenging activity (Table 2), which was in accordance with the well-known radical scavenging properties of gallic acid [20]. In addition, compounds 2 and 3 showed protective capacity against UV-B radiation.   (6), protocatechuic acid (7), (9), isoquercitrin (10), guaijaverin (11), and quercetin (12) The major peak at t R 10 min in the HPLC chromatogram consisted of two co-eluting flavonol glycosides 4 and 5. Compound 4 was purified and identified as quercetin [24] was identified from a fraction containing 4 and 5. Compound 4 was found to be a slightly weaker antioxidant and photoprotectant than 2 and 3 ( Table 2). Peak a consisted of several co-eluting compounds and was not pursued further. Peak 6 was enriched by filtration over polyamide, and HPLC purification afforded a 7:3-mixture of meta-and paradigallic acid methyl ester (6) [25]. Compound 6 showed radical scavenging activity comparable to 2 and 3, but no protective capacity against UV-B (Table 2).
The activity profile recorded for the methanolic leaf extract of Combretum cacoucia showed a zone of radical scavenging capacity between t R 5 and 15 min (Fig. 2C). Filtration over polyamide afforded five tannin-depleted fractions (Fig. 2S, Supporting Information) from which peaks b and c had disappeared. The main peak at t R 10 min consisted of three compounds which were further purified and identified as hyperoside (8) [26], 3,3'-dimethylellagic acid 4-O-β-glucopyranoside (9) [27], and isoquercitrin (10) [26]. Compound 9 could not be obtained at a purity of ≥95% required for further testing. Compounds 8 and 10 were found to possess DPPH scavenging activity and the capacity to protect human skin fibroblasts from UV-B radiation ( Table 2). Isoquercitrin (10) almost completely blocked cell death. Three minor peaks (7, 11, and 12) in the active time window were enriched in fraction PA2, and were identified as protocatechuic acid (7) [28], guaijaverin (11) [29], and quercetin (12) [30]. Compound 7 could only be obtained at 90% purity, with traces of phenolic glycosides as contaminants. The compound was nevertheless tested and exhibited weak antioxidant activity, but very good UV-B protection. Quercetin (12) and its glycosides 8, 10, and 11 showed significant free radical scavenging properties as previously reported [31][32][33], but only the glycosides 8, 10 and 11 showed significant UVprotective activity (Table 2). However, the lower test concentrations for compounds 11 and 12 had to be taken into account.
The screening of a taxonomically diverse library of Panamanian plant extracts followed by an activity-driven identification of radical scavenging and UV-B protecting properties led to the identification of a series of known polyphenols. The example shows that the profiling approach can be efficiently used not only for the discovery of bioactive compounds of pharmaceutical, but also of cosmetic interest.
Preparative HPLC was carried out on an LC 8A preparative liquid chromatograph equipped with a SPD-M10A VP PDA detector (all Shimadzu). A SunFire C 18 column (150 x 30 mm i.d., 5 μm; Waters) connected to a pre-column (10 x 10 mm) was used, at a flow rate of 20 mL/min. HPLC-based activity profiling was performed on an Agilent 1100 system equipped with a PDA detector. A SunFire C 18 column (150 x 10 mm i.d., 5 μm; Waters) connected to a pre-column (10 x 10 mm) was used. The flow rate was 4 mL/min. Timebased fractions were collected with a Gilson FC204 fraction collector. ESI-MS spectra were obtained on an Esquire 3000 Plus ion trap mass spectrometer (Bruker Daltonics). NMR spectra were recorded on an Avance III 500 MHz spectrometer (Bruker BioSpin) equipped with a 1-mm TXI microprobe.

Plant Material
The

HPLC-Based Activity Profiling
Extract solutions dissolved in DMSO (50 mg/mL) were separated by semi-preparative HPLC. Two aliquots of 200 μL corresponding to 10 mg of the extract were injected. A gradient of 5-100% MeCN in 30 min in 0.1% aqueous formic acid, followed by 100% MeCN over 5 min was used. Fractions of 0.75 min were collected from t = 3 min to t = 33 min. Fractions were transferred into 96-deepwell plates, evaporated, and submitted to screening.

Extraction and Isolation
Powdered leaves of M. jamaicense (704.6 g) were percolated with 12 L MeOH to afford 198.5 g of the extract. A portion (20. Compounds were identified with the aid of 1 H-and 2D-NMR, and ESI-MS spectroscopy, and by comparison with the literature data. The purity of the isolated compounds was >95% as determined by NMR except for compounds 7 (90%) and 9 (<70%).

DPPH Radical Scavenging Assay
The antioxidant potential of the test samples was monitored by the change in optical density of the DPPH radical. A stock solution of 0.314 mM DPPH in EtOH was prepared. This stock solution was prepared fresh every day. Extracts were initially tested at 200 μg/ml. Samples that exhibited a strong DPPH scavenging activity, i.e > 80% scavenging, were further evaluated at lower concentrations.
Dry microfractions of the selected extracts in 96-deepwell plates were dissolved in DMSO and tested directly against DPPH scavenging. When a large number of active microfractions appeared for one extract, the most active fractions were tested at a 5-fold dilution. In a 96-well plate, 10 μl of the sample (extract/ fraction/ compound) in DMSO and 190 μl of DPPH solution were mixed and incubated in the dark for 30 min at ambient temperature. Absorbance was measured at 517 nm using an Infinite M200Pro plate reader (Tecan, Männedorf, Switzerland). Measurements were done in triplicate. Blanks for every sample without DPPH were also measured. Gallic acid was used as the positive control. The percentage of DPPH scavenging was estimated by the following equation:

Cell Protection Against UV-B Irradiation
A human skin fibroblast cell line (AG01523; Coriell Institute for Medical Research, Camden, NJ, USA) was used for the assessment. Cells were routinely cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with antibiotics (100 IU/ml penicillin; 100 μg/ml streptomycin) and 15% Fetal Bovine Serum (FBS) in an environment of 5% CO 2 , 85% humidity, at 37°C, and subcultured once a week at a 1:2 split ratio, using a trypsin-citrate solution (0.25%-0.3%, respectively). Cell counting after trypsinization was performed using a Coulter counter.
For assessing the possible cytotoxicity of the samples (extracts, fractions, or isolated compounds), cells were plated in flat-bottom, tissue culture-treated 96-well plates at a density of 5,000 cells/well. After 48 hours of growth, the medium was changed to serumfree, phenol red-free DMEM, and serial dilutions of the test samples were added. The corresponding dilutions of dimethylsulfoxide (DMSO) served as negative controls.
Following incubation with the test samples for 72 hours, the medium was changed to serum-free, phenol red-free DMEM containing 1 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) as described by Kostakis et al. [34]. After incubation with MTT for 4 hours, the medium was discarded, and the MTT-formazan crystals were dissolved in isopropanol. Absorbance was measured at 550 nm (reference wavelength; 690 nm) in an Infinite M200 microplate reader (Tecan) using Magellan TM software.
The highest non-cytotoxic concentration of each sample (extract, fraction, or isolated compound) was tested for the ability to protect human skin fibroblasts against toxicity of UV-B irradiation. Cells were plated in 96-well-plates and left to grow as described above. Then samples were added at the test concentrations determined as described above, along with serum-free, phenol red-free DMEM. After incubation for 18 hours, cells were subjected to UV-B irradiation for 10 min (corresponding to 726 mJ/cm 2 ) using a black box equipped with a closely spaced array of four Sankyo Denki UV-B lamps (Zhe Jiang, China) emitting between 280 nm and 360 nm (peak at 306 nm). Following further incubation for 72 hours, cytotoxicity was estimated using the MTT-method, as described in the previous paragraph. The plates treated in an identical manner, except for the UV-B irradiation, were used as the controls. The UV-B-protective capacity of the samples was calculated using the following equation