2022 Volume 45 Issue 1 Pages 77-85
Cyclooxygenase (COX)-1-selective inhibitors have side effects such as itching and dryness of the skin. In this study, the degree of skin dryness and the onset mechanism of this condition were investigated by comparing the effects of three non-steroidal anti-inflammatory drugs (NSAIDs) in mice. Mice were orally administered either indomethacin, loxoprofen sodium, or celecoxib (n = 5 per group) once daily for four consecutive days, and blood samples as well as skin and jejunal tissues were isolated on day 5. In the mice treated with indomethacin, transepidermal water loss was significantly increased, and dry skin was observed. In addition, the expression of matrix metalloproteinase (MMP)-I, mast cells, CD163, CD23, CD21, histamine, and peroxisome proliferation-activated receptor (PPAR)γ in the skin and jejunum was increased, and the blood levels of interleukin-10 and immunoglobulin E were also increased. In contrast, the expression of collagen type I in the skin was decreased. These results show that indomethacin activates PPARγ in the skin and jejunum, changes the polarity of macrophages, increases the secretion of MMP-1 from mast cells, and decomposes collagen type I, leading to dry skin.
The number of elderly people in Japan is increasing, and the prevalence of chronic pain is also increasing. Non-steroidal anti-inflammatory drugs (NSAIDs) are widely used to treat pain and many inflammatory diseases.1) NSAIDs decrease the production of prostaglandins by inhibiting cyclooxygenase (COX). COX has two main isozymes, COX-1 and COX-2. COX-1 is constitutively expressed in most tissues. Specifically, it has a protective effect on the mucosa of the stomach and intestine.2) COX-2 is primarily localized to inflammatory cells and tissues. It induces inflammation, fever, and pain.3) Many NSAIDs inhibit COX-2 to suppress inflammation and the progression of inflammatory reactions. However, it has been reported that some NSAIDs, which are not selective toward COX, cause adverse effects in digestive organs upon long-term use.4)
Gut, including small intestine, influences skin conditions through various mechanisms. The intestine often causes systemic inflammation by release of inflammatory materials. This ability has important implications in skin disease. In addition, this ability is affected by intestinal conditions, such as increase of intestinal permeability.
We previously reported that dry skin may be associated with indomethacin-induced small intestine injury in mice.5) In that study, we found increased expression of mast cells in the small intestine. Mast cells are known to release tryptase and histamine.6) Therefore, it has been suggested that indomethacin-induced small intestinal mast cells may be involved in skin dryness. However, the detailed induction mechanism of skin dryness by indomethacin is unknown.
Furthermore, indomethacin acts as an agonist of peroxisome proliferation-activated receptor (PPAR)γ, a subtype of PPAR.7) PPARγ is a nuclear receptor and a nuclear transposable element belonging to the steroid hormone receptor superfamily. It is also involved in the differentiation and proliferation of adipose tissue cells, inflammation, and immune response.8,9) However, the effect of PPARγ on mast cell degranulation, which plays an important role in inflammation and dry skin, is unknown.
Therefore, in this study, we administered indomethacin, loxoprofen sodium, and celecoxib to mice to examine whether the degree of intestinal damage is correlated with the extent of skin disruption. In addition, the effects of indomethacin and mast cells on skin dryness were investigated.
Indomethacin was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Loxoprofen sodium was purchased from Wako Pure Chemical Corporation (Osaka, Japan). Celecoxib was purchased from TOCRIS Bioscience (Avonmouth, Bristol, U.K.). All other reagents were of analytical grade. Indomethacin selectively inhibits COX-1, whereas loxoprofen sodium inhibits both COX-1 and COX-2 to the same degree. Celecoxib selectively inhibits COX-2 expression.
AnimalsHairless mice (HOS: HR-1) were purchased from SLC (Hamamatsu, Shizuoka, Japan). The mice were housed under a 12-h light cycle at a constant temperature of 23 ± 2 °C and relative humidity of 55 ± 10%. The mice were fed laboratory chow (CE-2; Oriental Yeast Co., Ltd., Tokyo, Japan) and water ad libitum.
Experimental DesignAt eight weeks of age, hairless mice were randomly assigned (n = 5 per group) to oral administration of indomethacin (40 mg/kg, suspended in saline), loxoprofen sodium (30 mg/kg, suspended in distilled water), or celecoxib (10 mg/kg, dissolved in 0.25% dimethyl sulfoxide) once a day for four days. The doses of indomethacin, loxoprofen and celecoxib (40, 30, and 10 mg/kg) used in the prelim examination were based on the method of Kuzumoto et al.,10) Shimada et al.,11) and Liang et al.12) Each value is the maximum concentration that does not affect the health changes (weight loss, behavioral abnormalities) of mice after long-term administration. Control mice were administered the corresponding delivery vehicles according to the same schedule. This study strictly followed the recommendations and guidelines for the care and use of laboratory animals at the Suzuka University of Medical Science (Approval No. 34). All surgical procedures were performed under anesthesia induced by pentobarbital, and every effort was made to minimize animal suffering.
PPARγ Antagonist TreatmentIn the PPARγ antagonist-treated group, mice were intraperitoneally injected with a PPARγ antagonist (GW9662: 1 mg/mL; Wako Pure Chemical Corporation) in saline (0.1 mL) daily for four days. Animals in the control group were administered saline only.13)
Measurement of Transepidermal Water Loss (TEWL) and CapacitanceOn day 5, the TEWL and capacitance in the dorsal skin of each mouse were measured. TEWL and capacitance were used as a marker of skin permeability, reflecting the barrier function of the skin; decreased TEWL and increased capacitance indicates dry skin.14–16) TEWL measurements were carried out using a Tewameter TM300 (Courage + Khazaka Electronic GmbH, Cologne, Germany).17) Values were recorded once the reading had stabilized, typically 10 s after the probe was placed on the skin. The data are presented as the average of three independent measurements. The level of hydration in the outermost layer of the skin, the stratum corneum, was measured on day 5 using a Corneometer CM825 (Courage + Khazaka Electronic GmbH).18) The Corneometer probe was applied to the dorsal skin surface of each mouse, and the degree of skin hydration was determined by the electrical capacitance of the surface of the skin, expressed in arbitrary units. The reported data are the average of three independent measurements per test area.
Measurement of Histamine, Interleukin (IL)-10, and Immunoglobulin (Ig)E LevelsSkin, jejunum, and plasma samples were collected on the last experimental day. We rinsed 100 mg of skin and jejunum tissues in phosphate-buffered saline (PBS) to remove excess blood before homogenization. The tissues were finely minced and homogenized in 1 mL of PBS with a glass homogenizer on ice. The cells were lysed by ultrasonication at room temperature two to three times. The samples were centrifuged at 1500 × g for 15 min. The supernatant was collected for assay. The levels of histamine in the skin and jejunum, and the plasma levels of IL-10 and IgE were determined using commercial enzyme-linked immunosorbent assay kits (histamine: Bertin Pharmacol, Montigny-le-Bretonnux, France; IL-10: Proteintech, Rosemont, IL, U.S.A.; IgE, Yamasa Shoyu Co., Ltd., Chiba, Japan) according to the manufacturer’s instructions. Optical density was measured using a microplate reader (Molecular Devices, Sunnyvale, CA, U.S.A.).
Tissue Section StainingJejunal and dorsal skin tissue samples were isolated and fixed in PBS containing 4% paraformaldehyde (Wako Pure Chemical Corporation). Fixed tissue specimens were embedded in a frozen Tissue–Tek OCT Compound (Sakura Finetek, Tokyo, Japan) and sliced into 5-µm-thick sections. The skin and jejunal specimens were stained with hematoxylin-eosin (H&E) in accordance with established procedures to enable histological analysis. We microscopically evaluated H&E stained skin tissues. To determine overall skin thickness, 10 regions in which the skin appeared flat in acquired images were randomly selected. The length from the outer layer of the epidermis to the border of the subcutis was measured, and the average value was calculated. The cells were stained with toluidine blue to evaluate the number of mast cells via fluorescence microscopy using ImageJ software. Mast cells were quantified by counting cell numbers per mm2 field in 10 randomly selected regions. In addition, the skin specimens were stained using antibodies for immunological analysis according to a previously published method.5) Briefly, the specimens were incubated with either mouse monoclonal anti-CD23 (B cell marker, 1 : 100; Proteintech) or rabbit polyclonal anti-CD21 (1 : 100; Abcam, Cambridge, U.K.) primary antibodies. The samples were then incubated with fluorescein isothiocyanate-conjugated anti-mouse and tetramethylrhodamine isothiocyanate-conjugated anti-rabbit (1 : 30; Dako Cytomation, Glostrup, Denmark) secondary antibodies. The expression of CD23 and CD21 was evaluated immunohistochemically using fluorescence microscopy.
Western Blot AnalysisThe dorsal skin samples were homogenized in a lysis buffer (Kurabo, Osaka, Japan) and then centrifuged at 8000 × g for 10 min. The supernatant from each sample was collected and stored at −80 °C until analysis. After thawing, equal amounts of protein (5 µg/lane) were loaded onto a 4–12% BIS-TRIS Bolt gel (Life Technologies, Carlsbad, CA, U.S.A.) and the gel was subjected to electrophoresis at 200 V for 30 min. Following separation, proteins were transferred onto a nitrocellulose membrane using an iBlot Western blotting system (Life Technologies). The membranes were blocked with 5% skim milk at 4 °C overnight. After blocking, the membranes were incubated at 25 °C for 1 h with primary antibodies against type I collagen (1 : 1000; EMD Chemicals Inc., Gibbstown, NJ, U.S.A.), matrix metalloproteinase (MMP)-1 (1 : 1000; Santa Cruz Biotechnology Inc., Santa Cruz, CA, U.S.A.), mast cell tryptase (mast cell marker, 1 : 500; Santa Cruz Biotechnology Inc.), chemokine receptor 7 (M1 macrophage marker, 1 : 1000; Santa Cruz Biotechnology Inc.), CD163 (M2 macrophage marker, 1 : 1000; Abcam), PPARγ (1 : 1,000; Santa Cruz Biotechnology Inc.), p-PPARγ (1 : 1000; Bioss Inc, Woburn, MA, U.S.A.), CD23 (1 : 1000; Proteintech), and β-actin (1 : 5000; Sigma-Aldrich). The membranes were incubated with a horseradish peroxidase-labeled secondary antibody (Life Technologies), and the signal was detected with ImmunoStar Zeta (Wako Pure Chemical Corporation) using a lumino-image analyzer (LAS-4000; FUJIFILM, Tokyo, Japan). Protein bands were subjected to densitometric analysis, and the values were normalized to those of β-actin.
Statistical AnalysisAll data are presented as the mean ± standard deviation. Data were analyzed using Microsoft Excel 2010 and one-way analysis of variance followed by Tukey’s post-hoc test with SPSS software version 20 (IBM, Armonk, NY, U.S.A.). Heatmap analysis and hierarchical clustering were performed using R software. Differences were considered significant at p < 0.05.
The delivery vehicle of each NSAID was also administered to the control mice. However, no differences were observed following treatment (data not shown).
Effect of NSAIDs on Skin Dryness in MiceFirst, we examined various indicators of dry skin. The TEWL and capacitance in the dorsal skin showed that the mice in the indomethacin and loxoprofen sodium treatment groups had a higher TEWL than that in the control mice. In particular, the TEWL increased significantly in the indomethacin-treated mice compared to that in the control mice (Fig. 1a). The capacitance was only reduced in the indomethacin-treated mice (Fig. 1b). The skin of indomethacin-treated mice was thicker than that of control mice; however, the skin of loxoprofen or celecoxib treated mice were not different from control mice (Fig. 1c). In addition, indomethacin treatment markedly reduced the expression of collagen type 1 in the skin (Fig. 1d) and markedly increased the expression of MMP-1 (Fig. 1e).
Measurement of transepidermal water loss (TEWL) (a), capacitance (b), histological damage (c), and the expression of Collagen type I (d) and MMP-1 (e) in the dorsal skin of mice treated with indomethacin, loxoprofen sodium, or celecoxib. Values represent the mean ± standard deviation. * p < 0.05. Capacitance is represented in arbitrary units (a.u.). Scale bar = 100 µm.
Toluidine blue staining of the skin and jejunum sections indicated that the number of mast cells increased in each treatment group compared with that in the control group. In particular, the increase was the greatest in the indomethacin-treated group, followed by the loxoprofen sodium and celecoxib treatment groups (Figs. 2a, b). The results of measuring mast cells using Western blot showed the same tendency as the results of toluidine blue staining (Figs. 2c, d). In contrast, histamine levels were greatly increased by indomethacin treatment. The histamine levels with the other NSAID treatments were unchanged from those in the control group (Figs. 2c, d).
The number of mast cells in the control and NSAID-treated mice was counted following toluidine blue staining of the skin specimens. Values represent the mean ± standard deviation. * p < 0.05. Scale bar = 100 µm.
Next, we examined the expression of PPARγ, for which indomethacin acts as an agonist. The PPARγ expression in the skin was reduced by indomethacin treatment, and PPARγ expression in the jejunum was the same as that in the control group after all NSAID treatments (Figs. 3c, d). Conversely, activated PPARγ and p-PPARγ showed high values in both the skin and jejunum (Figs. 3a, b).
Values represent the mean ± standard deviation. * p < 0.05.
We investigated the localization of two types of macrophages. The expression of CCR7, an indicator of M1 macrophages, was unchanged in all groups (Figs. 4a, b). However, the expression of CD163, an indicator of M2 macrophages, was very high in the skin and jejunum of the indomethacin-treated group (Figs. 4c, d).
CCR7 is a marker of M1 macrophages and CD163 is a marker of M2 macrophages. Values represent the mean ± standard deviation. * p < 0.05.
We measured the plasma levels of IL-10 and IgE, as well as the expression of CD23 (mature B cells), in the skin and jejunum, which are closely associated with mast cells. The plasma levels of IL-10 and IgE were greatly increased by indomethacin treatment (Figs. 5a, b). However, in the other treatment groups, there was no difference from the levels in the control group. Similarly, the expression of CD23 in the skin and jejunum was also greatly increased by indomethacin treatment (Figs. 5c, d). In addition, we observed an increase in CD23 expression as well as co-localization of CD23 with CD21 in the mice treated with indomethacin (Fig. 5e).
Values represent the mean ± standard deviation. * p < 0.05. Scale bar = 100 µm.
We administered an antagonist of PPARγ to confirm the role of indomethacin in PPARγ. The TEWL, an indicator of skin dryness, increased due to the administration of indomethacin but was improved by administration of a PPARγ antagonist (Fig. 6a). In contrast, expression of p-PPARγ, which is the active form of PPARγ, was greatly increased by treatment with indomethacin, but decreased by administration of an antagonist, and was the same as that of the control group (Figs. 6b, c).
Values represent the mean ± standard deviation. * p < 0.05.
In the present study, we examined the interaction between dry skin and small intestinal injury induced by NSAIDs. We observed that treatment with NSAIDs (indomethacin, loxoprofen sodium, or celecoxib) increased the TEWL and decreased the capacitance in the skin of mice. However, these effects were most prominent in the indomethacin-treated group.
Generally, TEWL increases and capacitance decreases in dry skin.19) Collagen, along with other proteins in the skin, act as natural moisturizers, and mast cell-induced inflammation decreases their levels to dry the skin.20,21) Mast cells cause skin dryness induced by lesions in the digestive tract.5,22) Histamine is a well-known bioactive amine that is released from mast cells when IgE binds to the Fcε receptor on mast cells.23,24) Additionally, tryptase released from mast cells activates MMPs. MMP-1 decomposes type I collagen, and MMP-9 decomposes type IV collagen, which then causes skin disruption.20,21) Therefore, we confirmed the number of mast cells in the skin of mice treated with NSAIDs (Fig. 2). The number of mast cells in the skin increased in each NSAID treatment group, and this increase was the greatest in the indomethacin-treated mice.
The extent of skin dryness caused by NSAIDs depends on the COX selectivity of each NSAID. COX-1-selective inhibitors have a greater potential to cause gastrointestinal tract disturbances.25) Indomethacin selectively inhibits COX-1. Accordingly, in our study, indomethacin induced gastrointestinal tract disturbance by increasing the number of neutrophils to the greatest extent among all the tested NSAIDs. As celecoxib is a COX-2-selective inhibitor, it is not expected to induce gastrointestinal tract disturbance.26) Consistently, small intestinal injury was observed in the celecoxib-treated mice. Moreover, skin dryness in the celecoxib-treated mice was mild compared to that in the mice of the other groups.
The second reason why the indomethacin-treated mice experienced the driest skin in this study is considered to be the relationship between indomethacin and PPARγ. It is generally known that indomethacin activates PPARγ without suppressing the expression of the transcription factors activator protein 1 (AP-1) and nuclear factor-kappaB (NFκB).27) The role of PPARγ in inflammation has been advocated. An increase in bone marrow mononuclear cells has been observed when examining cell proliferation of mouse bone marrow cells with the PPARγ agonist troglitazone, indicating that PPARγ may be involved in an increase in mast cells.28) Indomethacin may have caused an increase in mast cells in the skin and jejunum in this study via PPARγ. Furthermore, activation of PPARγ changes the polarity of M1/M2 macrophages and is important for inducing differentiation into M2 macrophages.29,30) M2 macrophages activate the secretion of IL-10, an anti-inflammatory cytokine that promotes B-cell maturation. CD23 was examined as a B-cell marker in this study. It is known that CD23 is mainly expressed on B cells, binds to an antigen-IgE immune complex, and is involved in the regulation of IgE production.31) In particular, CD23 interacts with CD21 in B cells and preferentially promotes IgE production.32) Mast cells release various mediators (histamine, tumor necrosis factor (TNF)-α, and matrix metalloproteinase) through IgE-mediated stimulation. These mediators have been suggested to cause dry skin. To prove this data, a PPARγ antagonist was administered instead of indomethacin and drying of the skin was induced, suggesting that this pathway may be involved in skin dryness (Fig. 6).
In this study, administration of indomethacin increased histamine. Histamine H1 and H2 receptors expressed on keratinocytes are activated to delay the recovery of skin barrier function.33) It has also been reported that TNF-α has many effects on the skin barrier function.34,35) However, this study does not examine histamine receptors or TNF-α. Therefore, further studies are needed on histamine receptors and TNF-α. In addition, topical application of PPARγ normalized skin barrier homeostasis.36) Therefore, it is necessary to apply indomethacin locally and study it.
From the above results, it is evident that small intestinal injury caused by indomethacin induces skin dryness. In addition, it is suggested that indomethacin changes the M2 type of macrophages by activating PPARγ in the skin and jejunum, and moves the downstream signal for IL-10/CD23 (B cell)/IgE/mast cells, leading to dry skin. Additionally, we confirmed that COX-2-selective NSAIDs may be preferred over other NSAIDs for their anti-inflammatory and analgesic effects, as they do not cause significant skin dryness (Fig. 7). This information may be critical when selecting NSAIDs for patients predisposed to skin disorders.
This study was supported by JSPS KAKENHI (Grant No. 18K06082).
The authors declare no conflict of interest.
