Introduction

Ultraviolet rays (UVRs) are the extrinsic factor responsible for the extensive skin damage caused by sunlight. Excessive UVR irradiation, especially UVB, damages the epidermal layer (epidermal cells) and induces inflammation accompanied by erythema (sunburn) and edema in the acute phase [1]. Chronic UVR irradiation over the course of years prolongs inflammation, causing rough skin, spots, dryness, reduced skin elasticity, wrinkles, and peripheral vasodilation [2]. The inflammatory response is an important physiological function in biological defense, but needs to be regulated when excessive or chronic.

When exposed to UVR irradiation, epidermal cells produce large amounts of inflammatory cytokines such as interleukin-1 (IL-1) [3, 4], interleukin-6 (IL-6) [5], tumor necrosis factor-α (TNF-α) [6, 7], and interleukin-8 (IL-8) [8]. These cytokines induce vasodilation in the dermis, which in turn reddens the skin [9] and promotes vascular permeability and the ensuing processes of edema and neutrophil infiltration [10, 11]. Bernard et al. [12] identified non-coding RNAs such as U1-RNA, a molecule with double-stranded domains released by UVR-damaged cells, as an important source of the inflammatory cytokines that lead to erythema. They also found that U1-RNA induced the production of inflammatory cytokines from epidermal cells, a cell type not severely damaged by UVR, via toll-like receptor 3 (TLR-3) [12]. Their experimental results showed that an intradermal injection of U1-RNA induced redness, swelling, and the production of TNF-α and IL-6 in the skin of wild-type mice, but not the skin of TLR-3−/− mice. The production of inflammatory cytokines can be disrupted by inhibiting the uptake of self-RNAs (which act as damage-associated molecular patterns (DAMPs)) via the endocytosis of epidermal cells, as most TLR-3 in human epidermal cells is expressed in the endosomes [13]. It can also be disrupted by inhibiting the binding between RNA and TLR-3, or by inhibiting downstream signaling after the RNA binds to TLR-3. Disrupting the production of inflammatory cytokines is thus thought to be effective in preventing skin disorders caused by UVR irradiation.

Botanicals are a rich source of biologically efficacious compounds, and botanical-derived products are widely used all over the world as herbal medicines, external preparations, and health foods. Among them, polyphenols containing ellagitannin have attracted great attention for their antioxidant [14], vasorelaxant [15], anti-inflammatory [15, 16], anti-hyperlipidemic [17], and anti-diabetic [18] actions. Ellagitannins are found in only a few fruits, such as strawberry, raspberry, pomegranate, and muscadine grape [19]. Fumagalli et al. [20] reported that two ellagitannins derived from strawberry, casuarictin and agrimoniin, suppressed IL-8 production in TNF-α-stimulated gastric adenocarcinoma cells (AGS) in an in vitro model of human gastritis induced by Helicobacter pylori infection.

Rosa roxburghii fruit is rich in tannins, which gives the fruit an astringent character and exerts hemostatic, analgesic, and antiseptic effects that have been applied in folk medicines and cosmetics [21]. Having found that R. roxburghii fruit contains ellagitannins such as alnusiin, casuarictin, and tellimagrandin II [22], we have been investigating whether a R. roxburghii fruit extract (RRFE) reduces skin inflammation induced by UVR irradiation.

In this study we (1) sought to evaluate whether RRFE reduces the inflammation caused by self-RNAs released from epidermal cells that have been significantly damaged by UVR. We did so by evaluating the effect of the RRFE in suppressing the gene expression of IL-8 using normal human epidermal keratinocytes (NHEKs) stimulated with polyinosinic:polycytidylic acid (poly(I:C)), a ligand of TLR-3 [23]. We also examined whether dipotassium glycyrrhizinate (GK2) [24], a known suppressor of the erythema induced by UVR irradiation, and allantoin [25], a suppressor of eczema, have similar effects. GK2 and allantoin are plant-derived anti-inflammatory agents that have been conventionally formulated in external preparations. (2) To identify the ellagitannins contained in the RRFE, and to show that they were active ingredients, we evaluated the inhibitory effects of the RRFE on IL-8 gene expression and IL-8 protein release.

Materials and methods

Materials

Poly(I:C) was purchased from Sigma-Aldrich (St Louis, MO). Dipotassium glycyrrhizinate (GK2) and allantoin were purchased from Fujifilm Wako Pure Chemical Co., Ltd (Tokyo, Japan). Rosa roxburghii fruits harvested in Hubei Province, China in 2017 (Lot No. 79Q4) were purchased from Matsuura Yakugyo Co., Ltd. (Nagoya, Japan).

Preparation of Rosa roxburghii fruit extract and fractionation, isolation, and identification of pure compounds from the extract

The R. roxburghii fruit extract (RRFE) used for the in vitro testing was prepared by the following method. The fruits were soaked in 50% ethanol (30 times their weight) at room temperature, filtered, and concentrated in an evaporator. The RRFE was prepared as a solution by adjusting the solid content to 11.6 mg/ml with 50% ethanol.

The active compounds were isolated and identified by the following steps. Rosa roxburghii fruits were soaked in 100% methanol (MeOH) (10 times their weight). The methanol extract (16.9 g) was separated by Sephadex LH-20 gel column (80 mm φ × 450 mm) chromatography and eluted with solvent (MeOH: water = 9: 1 v/v) to obtain 8 fractions (Fr. 1–8). Fr. 8 (675.7 mg) was separated by Sephadex LH-20 gel column (35 mm φ × 450 mm) chromatography and eluted with solvent (MeOH: water = 9: 1 → 8: 2 → 7: 3 → 6: 4 v/v) to obtain 10 fractions (Fr. 8–1 to Fr. 8–10) (Fig. 1).

Fig. 1
figure 1

Scheme for the fractionation and identification of active ingredients in the Rosa roxburghii fruit methanol extract

Full size image

Strictinin (CID 73330) (5.1 mg) and casuarictin (CID 73644) (27.6 mg) were isolated from Fr. 8–5 and Fr. 8–10, respectively, using a preparative HPLC (p-HPLC) method [ODS-3 (4.6 mm φ × 250 mm) (MeOH: 0.05% TFA aq. = 5%: 95% → 50%: 50% (40 min)].

The isolated compounds were identified by 1H–NMR, 13C–NMR, 1H–1H COSY, HMQC, HMBC (JEOL ECA 600 NMR spectrometer), and MALDI-TOF–MS (Shimadzu Biotech Axima Resonance 2.9.1.20100121).

Strictinin (1–O–galloyl–4,6–O–hexahydroxydiphenoyl–β–d–glucopyranose): λmax (H2O: MeOH = 71: 29) nm (log ε): 218 (2.94), 269 (2.61). The molecular weight was determined by MALDI-TOF–MS as 657.0975 m/z [M + Na]+ (Calcd for C27H22NaO18: 657.0698) (Shimadzu Biotech Axima Resonance: Mode positive, Low 100 + , power:120). 1H-NMR (Methanol-d4, JEOL ECA-600, 600 MHz) δ: Glc, 3.59 (1H, t, J = 8.28, H-2), 3.70 (1H, t, J = 9.66, H-3), 3.80 (1H, d, J = 12.4, H-3), 4.02 (1H, dd, J = 5.52, 8.94, H-5), 4.77 (1H, t, J = 15.5, H-4), 5.21 (1H, dd, J = 6.9, 13.7, H-6α), 5.65 (1H, d, J = 8.22, H-1); 4,6-HHDP, 6.53 (1H, s, H-3′′′), 6.67 (1H, s, H-3′′); galloyl, 7.12 (2H, s, H-2′,6′). 13C NMR ((Methanol-d4, 150 MHz): δ Glc, 62.9 (C-6α), 71.9 (C-4), 72.3 (C-5), 73.4 (C-2), 74.7 (C-3), 94.9 (C-1); galloyl, 109.2 (C-2′, 6′), 119.2 (C-1′), 139.1 (C-4′), 145.2 (C-3, 5), 165.5 (C-7′); 4,6-HHDP, 106.9 (C-3′′′), 107.3 (C-3′′), 115.3(C-1′′′), 115.5 (C-1′′), 125.0, 125.2 (C-2′′ or 2′′′), 136.0 (C-5′′′), 136.3 (C-5′′), 143.4, 143.5 (C-4′′ or 4′′′), 144.5, 144.6 (C-6′′ or 6′′′), 168.3 (C-7′′), 168.5 (C-7′′′) (Fig. 2a).

Fig. 2
figure 2

Structures of the monomeric ellagitannins, a strictinin and b casuarictin

Full size image

Casuarictin (1–O–galloyl–2,3–4,6–bis–O–hexahydroxydiphenoyl–β–d–Glucopyranose): λmax (H2O: MeOH = 66: 34) nm (log ε): 218 (3.11), 265 (2.86); The molecular weight was determined by MALDI-TOF–MS as 959.3599 m/z [M + Na]+ (Calcd for C41H28NaO26: 959.0761). 1H NMR (Acetone-d6, 600 MHz) δ Glc, 3.85 (1H, d, J = 13.1, H-6), 4.48 (1H, dd, J = 6.2, 9.7, H-5), 5.15 (m, H-4), 5.16 (1H, m, H-2), 5.35 (1H, dd, J = 6.9, 13.1, H-6), 5.39 (1H, t, J = 9.7, H-3), 6.19 (1H, d, J = 8.9, H-1); galloyl, 7.15 (2H, s, H-2, 6); 2, 3-HHDP, 6.34 (1H, s, H-3′), 6.43 (1H, s, H-3); 4,6-HHDP, 6.52 (1H, s, H-3), 6.65 (1H, s, H-3′); 13C NMR (Acetone-d6, JEOL ECA-600, 150 MHz): δ ppm Glc, 62.2 (C-6), 68.4 (C-4), 72.7 (C-5), 75.1 (C-2),76.4 (C-3), 91.4 (C-1); galloyl, 109.5 (C-2,6), 119.1 (C-1), 139.1 (C-4), 145.5 (C-3, 5), 164.2 (C-7); 2, 3-HHDP, 106.5 (C-3′′, 3′′′), 167.7 (C-7′′), 168.4 (C-7′′′); 4, 6-HHDP, 106.8 (C-3′′′′), 107.5 (C-3′′′′′), 167.0 (C-7′′′′), 167.2 (C-7′′′′′); HHDP, 113.3–115.2 (C-1′′, 1′′′, 1′′′′, 1′′′′′), 125.1–125.7 (C-2′′, 2′′′, 2′′′′, 2′′′′′), 135.4–135.8 (C-5′′, 5′′′, 5′′′′, 5′′′′′), 143.7–143.8 (C-6′′, 6′′′, 6′′′′, 6′′′′′), 144.3–144.5 (C-4′′, 4′′′, 4′′′′, 4′′′′′) (Fig. 2b).

Table 1 shows the amounts of strictinin and casuarictin contained in the RRFE and Fr. 8.

Table 1 The amounts of strictinin and casuarictin contained in 1 mg of solid component of the RRFE and Fr. 8
Full size table

Cell culture

NHEKs purchased from Kurabo (Osaka, Japan) were grown in serum-free keratinocyte growth medium KBM-2 (Lonza, Walkerville, MD) at a low calcium concentration (0.06 mM) together with bovine pituitary extract (BPE), human recombinant epidermal growth factor (hEGF), insulin, transferrin, hydrocortisone, epinephrine, and gentamicin/amphotericin-B, at 37 ℃ in a humidified atmosphere of 5% CO2. The medium was changed daily. When near confluence (70–90%), the cells were subcultured with trypsin/ethylenediaminetetraacetic acid (EDTA). Cells from passages 3–6 were used for the experiments.

Treatment of the cells

When the NHEKs reached 80% confluence, the medium was replaced with KBM medium without hydrocortisone and epinephrine. After further culture for 24 h, the cells were treated with the RRFE at the indicated concentrations for 1 h and then exposed to poly(I:C) (1 μg/ml) for 4 h for gene expression analysis or for 24 h for protein expression analysis.

RNA isolation and reverse transcription–polymerase chain reaction (RT-PCR)

After the cultured NHEKs were rinsed, total RNA was extracted with an RNeasy kit (QIAGEN) and quantified using NanoDrop software (Thermo Fisher Scientific, USA). cDNA was prepared from 50 ng of total RNA using a Prime Script RT Reagent Kit (Takara Bio, Japan) for RT-PCR. The RT-PCR procedure consisted of a reverse transcription reaction at 37 °C for 15 min followed by enzyme inactivation at 85 °C for 5 s.

Real-time quantitative PCR

Real-time quantitative PCR was performed using each specific primer and SYBR Premix EX Taq II (Takara Bio, Japan). IL-8 (forward primer sequence of 5′–CCACACTGCGCCAACA–3; reverse primer sequence of 5′–GCATCTTCACTGATTCTTGGAT–3) and ribosomal protein S18 (RPS18) for internal control (forward primer sequences of 5′–TTTGCGAGTACTCAACACCAACATC–3′; reverse primer sequence of 5′–GAGCATATCTTCGGCCCACAC–3′) were purchased from Hokkaido System Science, Japan. The stability of RPS18 in the cells showed no variation under any of conditions used in this study. Real-time fluorescence detection was performed using a Thermal Cycler Dice Real Time System (Single PCR) (Takara Bio, Japan). The PCR cycling conditions were as follows: 95 °C for 30 s followed by 40 cycles at 95 °C for 5 s and 54–60 °C for 30 s.

Enzyme-linked immunosorbent assay (ELISA)

After the treatment, the cell supernatant was collected and stored at − 80 °C until measurement. Secreted IL-8 levels were estimated by ELISA assays using ELISA kits (R&D Systems, USA) according to the manufacturer’s instructions.

Statistical analysis

The results were analyzed using SPSS 22.0 (IBM, USA). The data were collected from at least three independent experiments and expressed as means ± SD. For all results, the data were analyzed by analysis of variance (ANOVA). To perform multiple comparisons, Dunnett’s test was used post hoc after ANOVA. A p value of < 0.05 was considered statistically significant. The levels of statistical significance were indicated as: *, p < 0.05; **, p < 0.01.

Results

RRFE suppresses IL-8 gene expression in poly(I:C)-treated NHEKs

We investigated whether the RRFE affected IL-8 gene expression in poly(I:C)-treated NHEKs. Poly(I:C) induced the expression of IL-8 in NHEKs significantly, and the RRFE suppressed IL-8 gene expression in a concentration-dependent manner (Fig. 3a). GK2 and allantoin, two anti-inflammatory components widely used in topically applied skin agents, showed no inhibitory effects on IL-8 gene expression (Fig. 3b and c).

Fig. 3
figure 3

Effects of RRFE, GK2, and allantoin on IL-8 gene expression in poly(I:C)-treated NHEKs. NHEKs were pretreated with a RRFE (0.37–10 µg/ml), b GK2 (7.41–200 µM), or c allantoin (7.41–200 µM) for 1 h and stimulated with poly(I:C) (1 µg/ml) for 4 h. The IL-8 mRNA expression was analyzed by quantitative real-time PCR. The relative expression level of IL-8 in the poly(I:C)-treated NHEKs was normalized by the housekeeping gene, RPS18. Data are expressed as means ± standard deviation (n = 4). **p < 0.01 versus poly(I:C) alone, Dunnett’s test

Full size image

Isolation and identification of the active compounds in RRFE that inhibit IL-8 gene expression

Upon observing the inhibitory activity of the RRFE on IL-8 gene expression (Fig. 3a), we decided to fractionate the extract to isolate the active compounds. Sephadex LH-20 gel column chromatography of the R. roxburghii fruit methanol extract yielded 8 fractions (Fig. 1). IL-8 gene expression was inhibited in the NHEKs treated with Fr. 5–8 (Fig. 4). Fr. 8, a fraction with high activity and low impurities, was further fractionated, and two hydrolyzed tannins, namely, strictinin and casuarictin, were isolated.

Fig. 4
figure 4

IL-8 inhibitory activity of RRFE fractions obtained by Sephadex LH-20 gel column chromatography. NHEKs were pretreated with a methanol extract (crude) or fraction (1.11–10 µg/ml) for 1 h and then stimulated with poly(I:C) (1 µg/ml) for 4 h. The IL-8 mRNA expression level was analyzed by quantitative real-time PCR. One value is shown for each fraction concentration (n = 1)

Full size image

Strictinin and casuarictin suppress IL-8 gene expression and IL-8 release in poly(I:C)-treated NHEKs

We investigated the effects of strictinin and casuarictin on IL-8 gene expression and IL-8 release in NHEKs treated with poly(I:C). Each tannin suppressed IL-8 gene expression in the poly(I:C)-treated NHEKs when applied at concentration above 100 ng/mL (Fig. 5a, c). A similar tendency was observed in the suppression of IL-8 release (Fig. 5b, d).

Fig. 5
figure 5

Effects of strictinin and casuarictin on IL-8 gene expression and IL-8 release in poly(I:C)-treated NHEKs. NHEKs were treated with strictinin or casuarictin for 1 h and incubated a, c for 4 h or b, d for 24 h in the presence of 1 µg/ml poly(I:C). The IL-8 mRNA expression was analyzed by quantitative real-time PCR. The IL-8 protein in the supernatant was quantified by the IL-8 ELISA method. Data are expressed as means ± standard deviation (n = 4). *p < 0.05, **p < 0.01 versus poly(I:C) alone, Dunnett’s test

Full size image

On the other hand, the amounts of strictinin and casuarictin contained in the RRFE and Fr. 8 that significantly suppressed IL-8 gene expression in the poly(I:C)-treated NHEKs were lower than the amounts of strictinin and casuarictin that achieved comparable effects when administered alone (Tables 2, 3). We speculated that the coexistence of strictinin and casuarictin in the RRFE or Fr. 8 may have strengthened the suppression of IL-8 gene expression through a synergistic effect. To investigate, we conducted an experiment to confirm whether the combined use of strictinin and casuarictin affected the IL-8 gene expression and IL-8 production in poly(I:C)-treated NHEKs.

Table 2 The amounts of strictinin and casuarictin contained in the RRFE
Full size table
Table 3 The amounts of strictinin and casuarictin contained in Fr. 8
Full size table

When the strictinin-to-casuarictin ratio was set to 1:1 in the RRFE, IL-8 gene expression and IL-8 release were both significantly inhibited by the compounds applied in combination at 100 ng/ml concentrations (Fig. 6a, b). While we also expected to see significant effects from the compounds in combination at the lower concentrations of 30 ng/ml, IL-8 gene expression was not inhibited under that condition (Fig. 6a, b). When the strictinin-to-casuarictin ratio was set to 1:5 in Fr. 8, the combined compounds applied at respective concentrations of 10 ng/ml and 50 ng/ml showed additive and significant inhibitory effects on IL-8 gene expression and IL-8 release (Fig. 6c, d).

Fig. 6
figure 6

Effects of the combined treatment of strictinin and casuarictin on IL-8 gene expression and IL-8 release in poly(I:C)-treated NHEKs. The NHEKs were treated with strictinin and casuarictin in combination for 1 h at concentration ratios of a, b 1:1 and c, d 1:5 and incubated a, c for 4 h and b, d for 24 h in the presence of 1 µg/ml poly(I:C). The IL-8 mRNA expression was analyzed by quantitative real-time PCR. The IL-8 protein in the supernatant was quantified by the IL-8 ELISA method. Data are expressed as means ± standard deviation (n = 4). *p < 0.05, **p < 0.01 versus poly(I:C) alone, Dunnett’s test

Full size image

Discussion

The RRFE significantly suppressed IL-8 gene expression in poly(I:C)-stimulated NHEKs in our experiments (Fig. 3). Chemokines are generally thought to be important mediators of UV-induced inflammatory responses [26]. The production of IL-8 in UVB-irradiated epidermal cells has been reported both in vitro [8, 27] and in vivo [28]. IL-8 effectively activates neutrophils, migrates to the site of inflammation, and exacerbates the inflammation by generating a cytokine storm. Our results indicate that the topical application of RRFE to skin before sun exposure can block the cytokine cascade by suppressing IL-8 production in the UV-damaged epidermis, and thereby reduce skin inflammation.

Casuarictin has been reported to suppress IL-8 production by inhibiting the NF-κB pathway in AGS stimulated with TNF-α [20]. Both the casuarictin and strictinin contained in the RRFE in our present experiments suppressed IL-8 gene expression in NHEKs stimulated by poly(I:C), a TLR-3 ligand (Fig. 5). While it was reported that strictinin inhibited the NO production and iNOS expression in the macrophage-like cell line, RAW 264.7, and that orally administered strictinin contained in Pimenta racemosa leave extract exhibited anti-inflammatory activity using rat footpad edema induced by carrageenan, our present study is the first report to demonstrate the anti-dermatitis effect at least in poly(I:C) treated NHEKs [29, 30].

Casuarictin has been found to suppress IL-8 expression by strongly inhibiting the NF-κB pathway in TNF-α stimulated AGS, while leaving the activator protein 1 (AP-1) pathway unchanged [20]. In NHEKs stimulated by poly(I:C), in contrast, the AP-1 pathway appears to contribute more strongly to the poly(I:C)-induced IL-8 gene expression than the NF-κB pathway, as the IL-8 expression peaks within 4 h after the poly(I:C) exposure [8]. The early control of AP-1 is thought to be crucial in preventing the exacerbation of inflammation after UVR irradiation, as AP-1 is produced within 30 min after UVR exposure [31], and IL-8 gene expression is first detected within 1 h [8]. A future question to explore is whether strictinin and casuarictin can regulate the expression of AP-1.

Strictinin and casuarictin, the two ellagitannins isolated and identified from the RRFE, inhibited IL-8 gene expression and IL-8 protein release in poly(I:C)-stimulated NHEKs when administered alone at high concentrations (Fig. 5) or in combination at low concentrations (Fig. 6c, d). Both ellagitannins can thus be considered active ingredients of RRFE. No sufficient effects were observed, meanwhile, in the combination test using the amounts of strictinin and casuarictin contained in RRFE naturally—amounts that were expected to show significant IL-8 gene expression inhibitory activity (Table 2), (Fig. 6a, b). This dose discrepancy is thought to be partly explained by factors in RRFE that enhance the efficacy of strictinin and casuarictin. The search for such factors will be a subject for future study.

GK2 has been reported to directly bind to high-mobility group box-1 protein (HMGB1), a non-histone DNA binding protein and a type of DAMP released by cells during inflammation [32]. Blocking the binding to HMGB1 receptors (TLR-2, 4) [33] and receptors for advanced glycation end products (RAGE) [34] inhibits the NF-κB signal and suppresses the inflammatory response. Allantoin has been reported to suppress carrageenan-induced footpad edema [35], and thus may inhibit the inflammatory response induced by carrageenan via TLR-4 [36, 37]. The absence of any suppressive activity of GK2 or allantoin against the TLR-3-mediated IL-8 expression using poly(I:C) (Fig. 3) is thought to be explained by the binding selectivity of GK2 and allantoin with respect to the pattern recognition receptors.

Dexamethasone, a steroid used for the treatment of sunburn, suppresses IL-8 expression in several ways. Strong IL-8 suppression can take place via the TLR-3-toll/IL-1 receptor domain containing the adapter-inducing interferon-β (TRIF) signaling pathway when epidermal cells are stimulated with poly(I:C) [38], or via other IL-8 production pathways such as the suppression of secondary IL-8 production by IL-1β released from epidermal cells by TLR-3-mediated Caspase-4 activation [39, 40]. While strictinin and casuarictin suppress IL-8 expression via the TLR-3-TRIF signaling pathway, they entirely lack the ability to suppress IL-8 secondarily produced via other pathways. As such, we believe that the IL-8 mRNA expression at 4 h after poly(I:C) stimulation (Figs. 5a, c and 6a, c) was suppressed more strongly than the IL-8 release at 24 h after stimulation (Figs. 5b, d and 6b, d).

While it remains unclear whether strictinin and casuarictin inhibit poly(I:C), suppress endocytosis, or inhibit the signal transduction system of TLR-3, both compounds suppress inflammation originating from self-RNAs released from UVR-damaged cells that resist control by the plant-derived anti-inflammatory agents used previously. It was reported that topical treatment on human reconstituted skin with pomegranate extract, which contains ellagitannnins, punicalin, pedunculagin and punicalagin as its major constituent, resulted in inhibition of UVB-induced formation of cyclobutane pyrimidine dimers (CPD) and 8–dihydro–2′–deoxyguanosine (8–OHdG) [41]. In our clinical trials, it has been shown that pre-application of lotion-containing RRFE-suppressed erythema caused by UV irradiation (data not shown). From these, it is expected that Strictinin, Casuarictin and RRFE have some degree of cutaneous absorption. However, we do not know the exact percutaneous absorption of these compounds, and their measurement is for further study.

We found that the RRFE and its strictinin and casuarictin components suppressed IL-8 expression, a biomarker of the cytotoxicity of UV irradiation induced by TLR-3 ligand. RRFE and its strictinin and casuarictin components are potentially beneficial to the skin, and thus provide a safe and natural strategy for the development of new topical anti-inflammatory agents.