Treatment with maresin 1, a docosahexaenoic acid-derived pro-resolution lipid, protects skin from inflammation and oxidative stress caused by UVB irradiation

Acute exposure to UVB irradiation causes skin inflammation and oxidative stress, and long-term exposure to UVB irradiation may lead to carcinogenesis. Our organism has endogenous mechanisms to actively limit inflammation. Maresin 1 (MaR1; 7R,14S-dihydroxy-docosa-4Z,8E,10E,12Z,16Z,19Z-hexaenoic acid) is a pro-resolution lipid mediator derived from the docosahexaenoic acid, which presents anti-inflammatory and pro-resolution effects. However, it remains to be determined if treatment with MaR1 can inhibit inflammatory and oxidative alterations in the skin triggered by UVB. The treatment with MaR1 (0.1–10 ng/mice at −10 min relative to the UVB irradiation protocol) reduced UVB-induced skin edema, neutrophil recruitment (MPO; myeloperoxidase activity, and migration of LysM-eGFP+ cells), cytokine production, matrix metalloproteinase-9 activity, keratinocyte apoptosis, epidermal thickening, mast cells counts and degradation of skin collagen in hairless mice. UVB irradiation caused a decrease of GSH (reduced glutathione) levels, activity of the enzyme catalase, ferric reducing ability (FRAP), and ABTS radical scavenging capacity as well as induced lipid hydroperoxide, superoxide anion production, and gp91phox mRNA expression. These parameters that indicate oxidative stress were inhibited by MaR1 treatment. Therefore, these data suggest MaR1 as a promising pharmacological tool in controlling the deleterious effects related to UVB irradiation.


Results
MaR1 inhibits the development of skin edema, MPO activity and the recruitment of LysM-eGFP + cells triggered by UVB irradiation. Exposure to UVB induces an acute response in the skin that includes formation of edema and leukocyte recruitment 8,26 . Based on that, we first investigated the effect of MaR1 (0.1-10 ng) on two inflammation parameters: skin edema, a classical sign of inflammation and neutrophil recruitment, an important cellular event during inflammation. UVB irradiation triggered both skin edema (Fig. 1A) and MPO activity (Fig. 1B) compared to negative control group (non-irradiated group), which were amenable by the dose of 10 ng of MaR1. To determine whether the reduction of MPO activity was related to lowering the counts of recruited immune cells, we performed fluorescence analysis using LysM-eGFP mice. Corroborating the MPO data, we observed an increase of LysM-eGFP + cells in the skin tissue after UVB irradiation and treatment with the 10 ng-dose of MaR1 inhibited this cellular recruitment (Fig. 1C-F). Method differences might explain the discrepancies on the intensity of effect of MaR1 on skin inflammation. The MPO activity is an enzymatic assay and indicates indirectly the presence of neutrophils, but also reflects the activity of MPO in these cells 11,27 . The fluorescence data used a reporter mouse strain (LysM-eGFP) to determine cellular presence and not enzyme activity. Altogether, these data demonstrate that MaR1 reduces UVB irradiation-induced skin inflammation.

MaR1 inhibits UVB irradiation-induced apoptosis of keratinocytes.
After UVB radiation, keratinocytes undergo apoptosis (sunburn cells) 28 . Histologically, sunburn cells present altered morphology as observed by chromatin condensation and eosinophilic cytoplasm 28 . UVB irradiation induced an increase in the number of sunburn cells compared to negative control group ( Fig. 2A and B). On the other hand, MaR1 reduced the number of sunburn cells in a dose-dependent manner with significant effects with 1 and 10 ng (Fig. 2C-F). MaR1 at 10 ng presented a statistically different effect compared to the 0.1 ng dose (Fig. 2F). These data indicate a protective effect of MaR1 upon UVB irradiation-induced keratinocytes apoptosis.
MaR1 inhibits UVB irradiation-induced epidermal thickening. UVB irradiation also causes epidermal thickening due to edema, cellular infiltration and cellular proliferation 29 . UVB irradiation induced significant epidermal thickening in comparison to the negative control group (Fig. 3A and B), which was inhibited by all doses of MaR1 tested (0.1, 1 and 10 ng; Fig. 3C-F). The doses of 1 and 10 ng abolished the epidermal thickening www.nature.com/scientificreports www.nature.com/scientificreports/ caused by UVB (Fig. 3F). This result demonstrates that MaR1 reduces the deleterious impact imposed by UVB radiation to the skin.

MaR1 inhibits UVB irradiation-induced increase of mast cell counts.
After UVB irradiation, mast cells secrete mediators that trigger inflammation and recruit other leukocytes such as neutrophils 30 . UVB irradiation induced significant increase of mast cells compared to basal conditions of negative control group ( Fig. 4A and B). All doses of MaR1 reduced the mast cell counts in the skin (Fig. 4C-F). These data indicate Results are presented as tissue weight in milligrams for skin edema and as neutrophils x10 4 per milligram of tissue for MPO activity. Original magnification 20x (images C-E); 50 μm. *p < 0.05 compared to non-irradiated group and # p < 0.05 compared to irradiated vehicle-treated group. 100 μm. *p < 0.05 compared to non-irradiated group; # p < 0.05 compared to irradiated vehicle-treated group; and **p < 0.05 compared to irradiated vehicle-and 0.1 ng of MaR1-treated groups.

MaR1 inhibits UVB irradiation-triggered production of cytokines and gp91 phox mRNA expression.
The cytokines IL-1β, TNFα and IL-10 are produced after excessive exposure to UVB radiation 11,12,28,31 . In the light of the pro-inflammatory contribution of IL-1β and TNFα and the role of IL-10 in limiting inflammation, the putative effect of MaR1 on cytokine production was addressed. We observed that the increase on cytokine levels after exposure to UVB was counteracted by MaR1 at 10 ng ( Fig. 7A-C). As a consequence of the inflammation caused by UVB, recruited cells such as neutrophils and macrophages produce large amounts of reactive oxygen species (ROS) such as superoxide anion 8,29 . Superoxide anion formation depends on gp91 phox (also known as NOX2), a subunit of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 11 . Thus, it was next addressed whether MaR1 reduces gp91 phox mRNA expression. We observed that MaR1 inhibited gp91 phox mRNA expression at the dose of 10 ng (Fig. 7D). These results demonstrate that MaR1 targets cytokines and oxidative metabolism to reduce UVB-irradiation induced inflammatory injury.

MaR1 inhibits UVB irradiation-induced oxidative stress.
Given UVB irradiation decreases antioxidant defenses as well as increases other oxidative stress markers in the skin 11,12,32,33 , we next evaluated the effect of MaR1 on UVB-induced oxidative stress. UVB irradiation depleted the antioxidant capacity of the skin as observed by reduced activity/levels in the FRAP assay (Fig. 8A), ABTS assay (Fig. 8B), GSH quantitation (Fig. 8C) and catalase activity assay (Fig. 8D). In addition to reducing antioxidant defenses ( Fig. 8A-D), UVB irradiation increased superoxide anion production ( Fig. 9A) and lipid peroxidation (Fig. 9B), which are markers of oxidative stress. MaR1 at 10 ng restored the skin antioxidant capacity as observed in the FRAP, ABTS, GSH and catalase assays (Fig. 8). Furthermore, all three doses of MaR1 inhibited superoxide anion production (the dose of 10 ng www.nature.com/scientificreports www.nature.com/scientificreports/ also present significant effect compared to the lower dose of 0.1 ng; Fig. 9A), while only MaR1 at 1 and 10 ng inhibited lipid peroxidation (Fig. 9B). Thus, the antioxidant effects of MaR1 contributed to the attenuation of the skin damages caused by UVB irradiation.

Discussion
A current concept is that pro-resolution lipid mediators actively reduce inflammation. Imbalance on pro-resolution lipid mediator profile can be found in skin diseases such as psoriasis. Interestingly, despite the presence of pro-resolution lipid mediators and their precursors in the skin of patients with psoriasis, it is also possible to observe a high increase on omega-6 fatty acid-oxidized derivatives, which shift the response to a pro-inflammatory profile 34,35 . Therefore, this imbalance can favor the development of skin inflammation and indicates that treatment with pro-resolution lipid mediators may work as a therapeutic approach to treat skin diseases. In fact, we have recently demonstrated that treatment with Lipoxin A 4 and Resolvin D1 (RvD1), which are pro-resolution lipid molecules derived from arachidonic acid and DHA, respectively, reduce UVB irradiation-induced skin damages 11,36 . However, to our knowledge, the present study shows for the first time that treatment with MaR1 efficiently reduces the deleterious impact of UVB-induced skin injury through the inhibition of inflammation and oxidative stress.
MaR1 reduced clinical signs of inflammation and essential pathophysiological inflammatory parameters in the UVB-induced skin injury model. MaR1 (10 ng per mouse) inhibited skin edema, neutrophils/macrophages recruitment (MPO and LysM-eGFP data), number of sunburn cells, epidermal thickening and mast cell counts triggered by UVB exposure in a dose-dependent manner. These results are in line with the inhibitory effect of MaR1 on neutrophil chemotaxis and stimulation of efferocytosis by macrophages 25,37 . Furthermore, human neutrophil interacts with platelet to increase endogenous production of MaR1, which might contribute to limit neutrophil recruitment toward tissue 17 . UVB also induces the apoptosis of keratinocytes, which can be observed as sunburn cells 28 . In a murine model of cancer, RvD1, RvD2, and RvE1 reduce apoptotic tumor cell www.nature.com/scientificreports www.nature.com/scientificreports/ debris-stimulation of tumor growth, indicating that enhancing the clearance of apoptotic cells and debris can reduce inflammation 38 . Herein we show that MaR1 inhibited keratinocyte apoptosis, thus, reducing the number of sunburn cells. Therefore, concurrent data reveal the inhibition of UVB-induced skin inflammation by MaR1.
Keratinocytes contribute to the initiation of the inflammatory response upon UVB irradiation by producing cytokines and chemokines 28 . These mediators induce vascular permeability, edema and recruitment of inflammatory cells (e.g. neutrophils) 31,39 . Tissue resident cells, such as mast cells also release histamine, TNFα, and IL-1β to stimulate the expression of metalloproteinase (especially MMP-9) in skin keratinocytes and fibroblasts. MMP-9 degrades collagen causing the skin photoaging [40][41][42] . In fact, considering the vulnerability of the skin to UV irradiation, MMP-9 is strongly linked to pathophysiology of tumors 43 . Interestingly, the treatment with MaR1 reduced mast cells count as well as inhibited the MMP-9 activity and the production of TNFα and IL-1β, indicating that MaR1 targets these physiopathological mechanisms to reduce UVB-induced skin inflammation and injury. TNFα also contributes to the generation of sunburn cells 28,44 . Thus, the inhibition of TNFα by MaR1 is in accordance with the reduced number of sunburn cells detected in our model. Additionally, both IL-1β and TNFα are edematogenic 45 and induce neutrophil recruitment 46 . Therefore, it is likely that MaR1 inhibition of TNF-α and IL-1β might have accounted to reduce edema, keratinocytes apoptosis, mast cell count, neutrophil recruitment, MMP-9 activity and collagen degradation induced by UVB irradiation. IL-10 is an anti-inflammatory cytokine that is co-released with pro-inflammatory cytokines to limit inflammation 47 . The present results demonstrated that MaR1 inhibited UVB-induced IL-10 production, which indicates that enhancing IL-10 levels is not a therapeutic mechanism of MaR1 in UVB irradiation-triggered inflammation. It is also possible that as MaR1 reduced the release of pro-inflammatory cytokines, the co-release of IL-10 was also reduced 48 . Collectively, these data corroborate other studies using different models and demonstrate that inhibiting cytokine production is a consistent mechanism of action of MaR1 22,23 . Interestingly, in a model using human bronchial epithelial cells line (BEAS-2B) and mouse lung slices stimulated with hog confinement facility-derived organic dust extract (HDE), MaR1 inhibited the release of pro-inflammatory cytokines and chemokines (TNF-α, IL-6, and CXCL1). In terms of mechanism, the HDE extract induced DNA biding activities of NFκB, AP-1 (activator protein 1), SP-1 (trans-acting transcription factor 1), and SRE (serum response element) as determined using luciferase reporter assays. However, MaR1 only inhibited the SRE DNA binding activity 21 , which demonstrated that MaR1 mechanisms may depend on inflammatory stimulus context and supports the importance of investigating MaR1 activities and mechanisms in every condition it may be useful.
Excessive absorption of UV irradiation can lead to the formation of molecular oxygen (O 2 ) increasing ROS production that trigger NFκB activation and cytokine production 6,33,49 . UVB also induces H 2 O 2 directly and indirectly though O 2 •− production 4 , and increases cyclooxygenase 2 (COX-2) activity resulting in pro-inflammatory prostaglandin production 50 . On the other hand, the activity of GSH and catalase are endogenous skin mechanisms www.nature.com/scientificreports www.nature.com/scientificreports/ to maintain its integrity. The first one chelates metal ions forming inert complexes preventing the oxidation process, while catalase converts H 2 O 2 into H 2 O and O 2 that prevents hydroxyl radicals ( • OH) formation. However, UVB irradiation-induced ROS production overwhelms the endogenous antioxidant systems resulting in tissue lesion 33 . Pro-resolution lipid mediators reduce oxidative stress in vitro as evidenced by two recent studies showing RvD1 51 and RvE1 52 reduce ROS production in macrophages. Interestingly, despite reducing ROS production, pro-resolution lipid mediators are not immunosuppressive. For instance, RvD1, RvD5, and PD1 lower the requirement of antibiotics during E. coli infection since these pro-resolution lipids enhance bacterial killing by neutrophils and macrophages 16 . This effect could be particularly interesting since disruption of skin barrier can increase susceptibility to infections 53 . In the present study, MaR1 restored the levels of GSH and catalase activity as well as reduced O 2 •− production and gp91 phox (NOX2) mRNA expression. These data are consistent with the effect of MaR1 on attenuating TNFα-induced ROS production, expression of NOX2, adhesion molecules and NFκB activation in vascular smooth muscle and endothelial cells 24 . The reduction of gp91 phox mRNA expression observed upon treatment with MaR1 could be related to the lower amount of recruited immune cells in the UVB irradiation model of skin sterile inflammation and does not directly indicates immunosuppression, which was not www.nature.com/scientificreports www.nature.com/scientificreports/ addressed in the present work. We also observed that MaR1 restored the ability of the skin to reduce the ion Fe +3 to Fe +2 and the ability to scavenge the ABTS + radical. Notably, oxidative stress is also involved in sunburn cells formation through mechanisms that lead to cytochrome C release after mitochondrial injury and modulation of p53 activity 54,55 . Consistent with these concepts, aldose reductase (an enzyme transcriptionally regulated during cellular response against oxidative insults and toxic aldehydes) diminishes human keratinocytes (HaCaT cells) apoptosis after UVB exposure. These effects were attributed to be dependent on p53 regulation, possibly through a suppression of ROS generation after UVB irradiation 54 . Therefore, besides inhibition of TNFα, counteracting oxidative damage is also a contributing mechanism of MaR1 to reduce the formation of sunburn cells. www.nature.com/scientificreports www.nature.com/scientificreports/ Summing up, intraperitoneal administration of the pro-resolution lipid mediator MaR1 protected the skin against inflammatory and oxidative alterations triggered by irradiation with UVB light. MaR1 diminished the following UVB skin physiopathological alterations: edema, neutrophil recruitment, the number of sunburn cells, epidermal thickening, mast cell counts, collagen degradation and MMP-9 activity. The inhibition of cytokine and ROS production explained the therapeutic actions of MaR1. Importantly, MaR1 presented a dose-dependent effect indicating a rational dosing that favors clinical development. Housing was under pathogen-free conditions in cages with individual ventilation in a rack system designed for mouse with regular shaving bedding and free access to water and food, with light/dark cycle of 12/12 h, exhausted air and controlled temperature (22 ± 2 °C). The Ethics Committee for Animal Use (CEUA/UEL, process number 1447.2015.10) approved all procedures of this study. All methods were performed in accordance with the relevant guidelines and regulations. All efforts were made in order to minimize the number of animals used and their suffering. Euthanasia at the end of experiments involved the sequential procedures of anesthesia with isoflurane 5% (Abbott Park, IL, USA), terminal killing by cervical dislocation and decapitation. Euthanasia was always performed during the light cycle. The mice were continuously monitored regarding welfare-related assessment before, during and after the experiments.

Drugs and reagents. Reagents from
General experimental procedures. Hairless mice (n = 6 mice per group per experiment) or LysM-eGFP mice (n = 5 mice per group per experiment) were randomly assigned in three different groups: non-irradiated, vehicle-treated irradiated group, and MaR1-treated irradiated group. The stock solution contained 10 ng/μL of MaR1 in 100% ethanol and was kept in a −80 °C freezer until use. During the preparation of the doses for the www.nature.com/scientificreports www.nature.com/scientificreports/ treatments, caution was taken to avoid exposure of MaR1 to air. The doses of MaR1 were chosen based on previous reports 18,22,56 . Mice were treated with MaR1 (0.1, 1, or 10 ng/mice, diluted in sterile saline, intraperitoneally, 200 µL [1 μL of ethanol plus 199 μL of saline]) 10 min before the beginning of UVB irradiation. Based on the dose-dependent results of MPO activity, we used the dose of 10 ng/mice for experiments involving LysM-eGFP mice. Samples of skin were dissected 2 h, 4 h, or 12 h after the UVB exposure, depending on the assay. Each parameter was evaluated at a specific time point, which was previously determined as suitable to detect significant differences between negative and positive control groups, therefore, being valid for determination of possible treatment effect [10][11][12]57 . Experiments were conducted twice.
Irradiation protocol. UVB lamp (Philips TL/12 RS 40 W, Medical-Holand) emits light between 270 and 400 nm, peaking at 313 nm, and placed on the top of the irradiation chamber and positioned 20 cm above the mice. This distance results in an irradiation of 0.384 mW/cm 2 . The radiation dose for induction of inflammation and oxidative stress was 4.14 J/cm 1, [10][11][12]32 . Specific time points after the UVB exposure were used based on previous standardization protocols 10-12,36 . Skin edema. Dorsal skin biopsy was carefully removed from euthanized mice and weighed using a precision scale. All samples presented a constant diameter of 5 mm. Results are expressed in mg of skin tissue obtained from the weight of each sample 10,11 . MPO activity assay. MPO colorimetric assay was used to determine neutrophil migration to the skin 10,27 . Reading was performed at 450 nm. A standard curve of neutrophils was used to compare the results, which are presented as MPO activity (number of neutrophils x 10 4 per mg of skin).
Skin fluorescence. Shaved LysM-eGFP + mice were used to visualize leukocyte recruitment to the skin as described previously 11 . This report mouse strain expresses enhanced green fluorescent protein (eGFP) expression controlled by the lysozyme M promoter (LysM) present in neutrophils granules. Imaging was performed using a confocal microscope (Leica TCS SP8, Leica, Wetzlar, Germany) with a 20x long distance objective. Images were processed using Leica EL6000 software (Leica, Wetzlar, Germany). The intensity of fluorescence was quantified by an investigator blinded to the treatment in randomly selected fields (one field per sample, n = 5) of different groups as an indicative of neutrophils recruitment to the skin tissue. The results are expressed as eGFP fluorescence intensity (%) 11 . Histopathological analysis. Skin samples were fixed in buffered formaldehyde, embedded in paraffin, sectioned (5 mm), and stained with Masson's trichrome stain for collagen fiber analysis. Collagen fiber intensity bundles shown in blue were analyzed by ImageJ Program as described previously 11 . Tissue sections were also stained with hematoxylin and eosin (H&E), and images were analyzed for epidermal thickness and apoptotic cells 11 . Toluidine blue staining was also used to determine mast cells count 35 . Histopathological scores are presented together with the representative images quantifying the alterations detected between the groups. Cytokine measurement. The levels of skin IL-1β, TNFα and IL-10 were measured using commercial enzyme-linked immunosorbent assay (ELISA) kits according to manufacturer's instructions (eBioscience) 11 . For that, skin samples were dissected and homogenized into sterile saline (500 µL). The supernatant was used to determine cytokine levels. Reading was performed at 450 nm in a microplate spectrophotometer reader and the results are expressed as picograms (pg) of each cytokine/mg of skin tissue.
Real time and quantitative polymerase chain reaction (RT-qPCR). Skin samples were dissected into TRIzol reagent (Invitrogen) and total RNA was extracted as recommended by manufacturer. The ratio of the reading at 260 and 280 nm was used to determine RNA purity (between 1.8 and 2.0 for all preparations).
Reverse transcription of total RNA to cDNA and qPCR were performed using GoTaq ® 2-Step RT-qPCR System (Promega, Madison, WI, USA) on a StepOnePlus ™ Real-Time PCR System (Applied Biosystems ® , Thermo Fisher Scientific, Waltham, MA, USA). The relative gene expression was determined using the comparative 2 − (∆∆Ct) method and GAPDH mRNA expression was used a reference gene to normalize data. Primer sequences were described previously 11 . Matrix metalloproteinase (MMP)-9 activity measurement. SDS-PAGE (sodium dodecyl sulphate polyacrylamide gel electrophoresis) substrate-embedded enzymography was performed as described previously 11 . After electrophoresis, the gels were incubated with 2.5% Triton X-100 (1 h), with 0.05 M Tris-HCl (pH 7.4), 0.01 M CaCl 2 (overnight, 37 °C), and stained with Brilliant blue R. The zone of enzyme activity was analyzed by comparing the groups in the ImageJ Program, after detaining in 20% acetic acid (NIH, Bethesda, MD, USA). A total of 4 gels were analyzed and each one presented the results of pools of 6 mice per group. GSH assay. GSH levels were determined as described previously 11,36 . Reading was performed at 405 nm.
Results are presented as μM of GSH per mg of skin, compared to standard curve (GSH concentration ranging 5-150 μM).

Catalase activity assay.
To determine catalase activity, it was measured the decay on H 2 O 2 concentration and the oxygen generation 11,36 . Reading was performed at 240 nm (25 °C) and catalase activity was calculated based on the difference between the reading before and 30 sec after H 2 O 2 . The catalase values were expressed as unit of catalase/mg of skin/minute.