Advertisement

Transcriptional Upregulation of Nrf2-Dependent Phase II Detoxification Genes in the Involved Epidermis of Vitiligo Vulgaris

      Oxidative stress is widely believed to be a contributing factor in vitiligo pathogenesis. To explore mechanisms by which epidermis responds to mounting oxidative stress, we investigated the involvement of phase II detoxification genes in vitiligo. Phase II detoxification pathways have recently been identified as being important in the regulation of epidermal skin homeostasis. In this study we show that the key transcription factor nuclear factor E2–related factor 2 (Nrf2) and the downstream genes NAD(P)H:quinone oxidase-1 (NQO-1), γ-glutamyl cystine ligase catalytic subunit (GCLC), and γ-glutamyl cystine ligase modifying subunit (GCLM) are upregulated in the lesional epidermal skin of subjects with vitiligo vulgaris. The differences between lesional and nonlesional skin were further investigated by studying the induced expression of Nrf2-dependent transcripts in skin punch biopsies using curcumin and santalol. Surprisingly, nonlesional skin showed induction of all transcripts while a similar effect was not observed for the skin punches from the lesional skin. The use of curcumin and santalol on epidermal cells showed that keratinocytes were more susceptible to apoptosis, whereas melanocytes induced phase II genes under the same concentrations with negligible apoptosis. Our studies provide new insights into the role of phase II detoxification pathway in maintaining skin homeostasis and sustaining redox balance in vitiligo patients.

      Abbreviations

      GCLC
      γ-glutamyl cystine ligase catalytic subunit
      GCLM
      γ-glutamyl cystine ligase modulatory subunit
      HO-1
      heme oxygenase-1
      Nrf2
      nuclear factor E2–related factor 2
      NQO-1
      NAD(P)H:quinone oxidase-1

      Introduction

      Vitiligo is a depigmenting disorder of human skin that is characterized by localized loss of melanin from the lesional epidermis (
      • Lei T.C.
      • Vieira W.D.
      • Hearing V.J.
      In vitro migration of melanoblasts requires matrix metalloproteinase-2: implications to vitiligo therapy by photochemotherapy.
      ;
      • Spritz R.A.
      The genetics of generalized vitiligo.
      ;
      • Schallreuter K.U.
      • Bahadoran P.
      • Picardo M.
      • et al.
      Vitiligo pathogenesis: autoimmune disease, genetic defect, excessive reactive oxygen species, calcium imbalance, or what else?.
      ). It is clear that melanocytes—the melanin-synthesizing cells of the skin—are substantially decreased in vitiligo lesions (
      • Tobin D.J.
      • Swanson N.N.
      • Pittelkow M.R.
      • et al.
      Melanocytes are not absent in lesional skin of long duration vitiligo.
      ). Several theories about the etiology of this complex disorder have been proposed. These include genetic susceptibility (
      • Spritz R.A.
      The genetics of generalized vitiligo.
      ), autoimmunity (
      • Le Poole I.C.
      • Luiten R.M.
      Autoimmune etiology of generalized vitiligo.
      ), impaired melanocyte migration and/or proliferation (
      • Gauthier Y.
      • Cario Andre M.
      • Taieb A.
      A critical appraisal of vitiligo etiologic theories. Is melanocyte loss a melanocytorrhagy?.
      ), and oxidative stress (
      • Namazi M.R.
      Neurogenic dysregulation, oxidative stress, autoimmunity, and melanocytorrhagy in vitiligo: can they be interconnected?.
      ;
      • Schallreuter K.U.
      • Kruger C.
      • Wurfel B.A.
      • et al.
      From basic research to the bedside: efficacy of topical treatment with pseudocatalase PC-KUS in 71 children with vitiligo.
      ;
      • Kovacs D.
      • Raffa S.
      • Flori E.
      • et al.
      Keratinocyte growth factor down-regulates intracellular ROS production induced by UVB.
      ). Although the exact mechanism is far from clear, attempts have recently been made to develop a cumulative theory to explain the origin of the disorder (
      • Dell’anna M.L.
      • Picardo M.
      A review and a new hypothesis for non-immunological pathogenetic mechanisms in vitiligo.
      ;
      • Westerhof W.
      • d’Ischia M.
      Vitiligo puzzle: the pieces fall in place.
      ). In this context, identification of deregulated pathways in vitiligo would enable a better understanding of the underlying pathophysiology.
      Skin is the major target of insult for a broad spectrum of chemical pollutants and UV rays (
      • Brenner M.
      • Hearing V.J.
      The protective role of melanin against UV damage in human skin.
      ). These toxic factors induce, directly or indirectly, the production of reactive oxygen species. Reactive oxygen species include singlet oxygen, superoxide anions, hydroxyl radicals, and H2O2, all of which are short-lived entities continuously generated at low levels during the course of normal aerobic metabolism (
      • Zou C.G.
      • Banerjee R.
      Homocysteine and redox signaling.
      ). The inability of skin to regulate the levels of these species results in oxidative stress. The breakdown of this homeostasis defense mechanism in the skin can result in modification of proteins, lipids, and DNA, which in turn can bring about several dermatological diseases. To combat this cellular damage, mammals have evolved elaborate detoxification machinery associated with the phase I, II, and III classes of enzymes. The phase II detoxification enzymes are crucial, as they are directly involved in detoxification reactions of oxidative stress response (
      • Tirumalai R.
      • Rajesh Kumar T.
      • Mai K.H.
      • et al.
      Acrolein causes transcriptional induction of phase II genes by activation of Nrf2 in human lung type II epithelial (A549) cells.
      ). The key transcription factor that induces phase II gene expression is nuclear factor E2–related factor 2 (Nrf2), and it functions by binding to the antioxidant response element present in the promoter of many phase II genes (
      • Motohashi H.
      • Yamamoto M.
      Nrf2-Keap1 defines a physiologically important stress response mechanism.
      ;
      • Kensler T.W.
      • Wakabayashi N.
      • Biswal S.
      Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway.
      ). One of the important in vivo regulatory mechanisms involves formation of the Nrf2-Keap1 complex, which is known to regulate cellular defense mechanisms such as oxidative stress (
      • Wakabayashi N.
      • Itoh K.
      • Wakabayashi J.
      • et al.
      Keap1-null mutation leads to postnatal lethality due to constitutive Nrf2 activation.
      ;
      • Nguyen T.
      • Nioi P.
      • Pickett C.B.
      The Nrf2-antioxidant response element signaling pathway and its activation by oxidative stress.
      ).
      Recent studies have identified an important role for the Nrf2 pathway in maintaining skin homeostasis in epidermis (
      • Marrot L.
      • Jones C.
      • Perez P.
      • et al.
      The significance of Nrf2 pathway in (photo)-oxidative stress response in melanocytes and keratinocytes of the human epidermis.
      ). This pathway operates in both melanocytes and keratinocytes, enabling skin to adapt to various environmental stresses. Genetically, Nrf2 promoter polymorphisms have been associated with an increased risk of developing vitiligo (
      • Zhou M.N.
      • Xu A.E.
      • Lu L.J.
      • et al.
      [Association of single nucleotide polymorphisms of Nrf2 promoter region with susceptibility to vitiligo].
      ;
      • Guan C.P.
      • Zhou M.N.
      • Xu A.E.
      • et al.
      The susceptibility to vitiligo is associated with NF-E2-related factor2 (Nrf2) gene polymorphisms: a study on Chinese Han population.
      ). Altered cellular localization of Nrf2 in lesional vitiligo skin has also been reported (
      • Guan C.P.
      • Wei X.D.
      • Chen H.Y.
      • et al.
      [Abnormal nuclear translocation of nuclear factor-E2 related factor 2 in the lesion of vitiligo].
      ). Interestingly, vitiligo patients seem to exhibit systemic as well as localized oxidative stress (
      • Passi S.
      • Grandinetti M.
      • Maggio F.
      • et al.
      Epidermal oxidative stress in vitiligo.
      ;
      • Rokos H.
      • Beazley W.D.
      • Schallreuter K.U.
      Oxidative stress in vitiligo: photo-oxidation of pterins produces H(2)O(2) and pterin-6-carboxylic acid.
      ), and millimolar levels of H2O2 are reported to be present in their epidermis (
      • Schallreuter K.U.
      Successful treatment of oxidative stress in vitiligo.
      ;
      • Shalbaf M.
      • Gibbons N.C.
      • Wood J.M.
      • et al.
      Presence of epidermal allantoin further supports oxidative stress in vitiligo.
      ). Several antioxidant proteins have been shown to be altered in patient groups (
      • Ines D.
      • Sonia B.
      • Riadh B.M.
      • et al.
      A comparative study of oxidant-antioxidant status in stable and active vitiligo patients.
      ;
      • Arican O.
      • Kurutas E.B.
      Oxidative stress in the blood of patients with active localized vitiligo.
      ), and metabolic changes in the skin induced by H2O2 have also been reported (
      • Schallreuter K.U.
      • Gibbons N.C.
      • Zothner C.
      • et al.
      Butyrylcholinesterase is present in the human epidermis and is regulated by H2O2: more evidence for oxidative stress in vitiligo.
      ;
      • Spencer J.D.
      • Gibbons N.C.
      • Rokos H.
      • et al.
      Oxidative stress via hydrogen peroxide affects proopiomelanocortin peptides directly in the epidermis of patients with vitiligo.
      ). A clinical study suggested that narrow-band UVB-activated pseudocatalase controls H2O2 levels and thus might be effective for treatment of childhood vitiligo (
      • Schallreuter K.U.
      • Kruger C.
      • Wurfel B.A.
      • et al.
      From basic research to the bedside: efficacy of topical treatment with pseudocatalase PC-KUS in 71 children with vitiligo.
      ). However, it is currently not clear whether Nrf2-regulated phase II detoxification machinery has an important role in vitiligo etiopathogenesis.
      In this study, we examined the integrity of the phase II detoxification pathway in the epidermis of vitiligo patients. We found that the transcript levels of Nrf2 as well as the downstream detoxification genes NAD(P)H:quinone oxidase (NQO-1), γ-glutamyl cystine ligase catalytic subunit (GCLC), and γ-glutamyl cystine ligase modulatory subunit (GCLM) are upregulated in the lesional epidermis compared with the matched nonlesional skin. Furthermore, we evaluated the regulation of phase II genes in the punch biopsies from vitiligo patients as well as in cultured melanocytes and keratinocytes using the electrophilic compounds curcumin and santalol. Our study revealed a critical role of the phase II detoxification pathway in vitiligo and suggests that therapeutic intervention of this pathway may be useful in management of the disorder.

      Results

      Expression analyses of phase II detoxification pathway genes

      To study the detoxification pathway in vitiligo, nonlesional and lesional skin biopsies were collected from vitiligo vulgaris patients after written informed consent and institutional ethical committee clearance had been obtained. The levels of Nrf2 and phase II detoxification gene transcripts were compared across the lesional and nonlesional epidermis of vitiligo patients, and this was considered a measure of cellular response to the prevailing oxidative stress. Total RNA was isolated from the lesional and nonlesional epidermis. Quantitative real-time PCR analysis was carried out to examine the expression of heme oxygenase-1 (HO-1), NQO-1, GCLC, and GCLM transcripts. TaqMan probes used for the experiment were designed at the exon–intron boundary and have been validated to be mRNA specific (
      • Marrot L.
      • Jones C.
      • Perez P.
      • et al.
      The significance of Nrf2 pathway in (photo)-oxidative stress response in melanocytes and keratinocytes of the human epidermis.
      ). Our preliminary analysis showed that β-actin mRNA was differentially expressed in vitiligo as compared with nonlesional skin, and therefore 18S ribosomal RNA (rRNA) was used as the normalization control in duplex TaqMan assays. Quantitative PCR analysis was carried out for each sample in duplicate. Replicates that differed by an SD of ≥0.5 were eliminated from the analysis. The ΔCt value was defined as the difference between the Ct value for the gene of interest and that for the endogenous control 18S rRNA. The ΔCt values of nonlesional and lesional skin for each of the patient samples were plotted as a column scatter plot (Figure 1a). Median ΔCt values were significantly lower for the expression of GCLC, GCLM, and NQO-1 in vitiliginous skin as compared with nonlesional skin (P-value <0.001, Mann–Whitney test). Interestingly, there was no significant difference in the HO-1 mRNA levels between nonlesional and lesional skin in this cohort of vitiligo vulgaris patients. Paired Student's t-test for the matched samples of nonlesional and lesional skin also showed significant trends (P-value <0.001), confirming upregulation of detoxification genes in the lesional skin.
      Figure thumbnail gr1
      Figure 1Levels of phase II detoxification gene transcripts are altered in the lesional epidermis as compared with the nonlesional epidermis. (a) Real-time PCR analyses of GCLC, GCLM, NQO-1, and HO-1 carried out using duplex assays were normalized to 18S ribosomal RNA (rRNA) and represented as a column scatter plot of ΔCt values for both nonlesional and lesional skin. The horizontal line shows the median (n=15). **Significant P-value (<0.001), as calculated using the Mann–Whitney test. ΔCt values for phase II genes in (b) foreskin and (c) normal epidermis. Fold change with respect to corresponding nonlesional skin for each vitiligo patient is represented as a bar graph. (d) GCLC, (e) GCLM, (f) NQO-1, and (g) HO-1. Fold change was calculated using 2−ΔΔCt, where ΔΔCt represents ΔCt of lesional–ΔCt of nonlesional skin. GCLC, γ-glutamyl cystine ligase catalytic subunit; GCLM, γ-glutamyl cystine ligase modulatory subunit; HO-1, heme oxygenase-1; NQO-1, NAD(P)H:quinone oxidase-1.
      To understand the significance of this upregulation, we also investigated the expression levels in normal and foreskin epidermis (Figure 1b and c). The ΔCt values between lesional and nonlesional samples can be readily compared as these were obtained from the same subjects. However, the same criterion cannot be applied to the foreskin and normal skin samples. With our limited data, it is tempting to speculate that the foreskin has a higher expression of phase II genes than normal skin does.
      We calculated the fold change in expression of each matched pair of nonlesional and lesional skin. The lesional skin of most of the individuals had elevated expression of GCLC, GCLM, and NQO-1 genes (Figure 1d–f, Table 1), whereas only 3 of 15 individuals showed remarkable elevated expression of the HO-1 gene (Figure 1g). Although HO-1 is an important element of the antioxidant response, its regulation has been shown to be more complex than that of other phase II genes (
      • Baglole C.J.
      • Sime P.J.
      • Phipps R.P.
      Cigarette smoke-induced expression of heme oxygenase-1 in human lung fibroblasts is regulated by intracellular glutathione.
      ;
      • Numata I.
      • Okuyama R.
      • Memezawa A.
      • et al.
      Functional expression of heme oxygenase-1 in human differentiated epidermis and its regulation by cytokines.
      ). The GCLC and GCLM proteins interact to form the functional γ-glutamyl cystine ligase enzyme, and, accordingly, the elevated GCLC and GCLM levels in the lesional skin assume functional significance. The data suggest a definite involvement of NQO-1, as it was upregulated in the lesional skin in almost all subjects with a fold change of >2.0.
      Table 1Details of vitiligo vulgaris patients recruited for the study along with the heat map showing regulation of phase II detoxification genes in lesional as compared with nonlesional epidermis
      Table thumbnail gr6

      Expression of Nrf2 in the lesional and nonlesional epidermis

      Given that Nrf2 is the upstream transcription factor regulating the induction of phase II genes upon oxidative stress, we examined the levels of Nrf2 mRNA in skin punch samples from vitiligo patients. The ΔCt values showed a distinctly higher expression of Nrf2 in the lesional skin as compared with the nonlesional skin (Figure 2a). The relative Nrf2 levels vary between two- and eightfold in different samples (Figure 2b). The transcript of Nrf2 was also studied in the normal skin and foreskin samples, as shown in Figure 2c. We further examined the protein expression of Nrf2 in the epidermis using immunohistochemistry on the skin sections derived from the foreskin and the nonlesional and lesional skin (Figure 2d). Interestingly, although the distribution of Nrf2 in the nonlesional skin is uniform, the lesional skin showed Nrf2 distribution in patches.
      Figure thumbnail gr2
      Figure 2Analysis of nuclear factor E2–related factor 2 (Nrf2) mRNA levels in nonlesional and lesional epidermis. (a) ΔCt values for Nrf2 determined by real-time PCR in the nonlesional and the lesional epidermis of vitiligo vulgaris patients. **Significant P-value (<0.005), as calculated using the Mann–Whitney test. (b) Changes in the expression levels of Nrf2 between the lesional skin and the corresponding nonlesional skin represented as fold change for each vitiligo patient. (c) ΔCt values for Nrf2 determined by real-time PCR in the foreskin and normal epidermis. (d) Immunohistochemical analysis of Nrf2 levels in nonlesional and lesional skin sections and comparison with foreskin. Nrf2 was stained brown using rabbit Nrf2 antibody and developed. Scale bar=10μm.

      Effect of curcumin and santalol on human melanocytes and keratinocytes

      Enhanced levels of phase II genes indicate a disrupted redox homeostasis in the vitiligo epidermis. We decided to explore this further by investigating the effect of induced oxidative stress response in cultured melanocytes and keratinocytes. Two common skin-care compounds—curcumin and santalol—were chosen for this study. Curcumin possesses two phenolic rings connected by two α,β-unsaturated carbonyl groups (Figure 3a). Santalol consists of a bicyclic ring system covalently linked to an aliphatic unsaturated chain containing a hydroxyl functional group (Figure 3b). Both compounds are electrophilic and similar to other known phase II gene inducers, including butylated hydroxyanisol, tert-butylhydroquinone, green tea polyphenol, (-)-epicatechin-3-gallate, and isothiocyanates such as sulforaphane (
      • Rushmore T.H.
      • King R.G.
      • Paulson K.E.
      • et al.
      Regulation of glutathione S-transferase Ya subunit gene expression: identification of a unique xenobiotic-responsive element controlling inducible expression by planar aromatic compounds.
      ). Curcumin has previously been shown to activate Nrf2 in several cell types (
      • Jeong S.O.
      • Oh G.S.
      • Ha H.Y.
      • et al.
      Dimethoxycurcumin, a synthetic curcumin analogue, induces heme oxygenase-1 expression through Nrf2 activation in RAW264.7 macrophages.
      ;
      • Yang C.
      • Zhang X.
      • Fan H.
      • et al.
      Curcumin upregulates transcription factor Nrf2, HO-1 expression and protects rat brains against focal ischemia.
      ).
      Figure thumbnail gr3
      Figure 3Cell viability of melanocytes and keratinocytes upon curcumin and santalol treatment. Chemical structures of (a) curcumin and (b) santalol. Bright-field images of cultured (c) keratinocytes and (d) melanocytes before treatment; scale bar=100μm. Cell viability measured colorimetrically using the methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay at various concentrations of curcumin (e) in keratinocytes and (f) in melanocytes and of santalol (g) in keratinocytes and (h) in melanocytes. Cells were treated with indicated concentrations of the compounds for 48hours followed by the addition of MTT for the last 3hours, and the purple formazan was dissolved in DMSO.
      Keratinocyte and melanocyte cultures were established from foreskins derived from healthy donors (Figure 3c and d). MTT (methylthiazolyldiphenyl-tetrazolium bromide) assays were performed to determine the optimal concentration of the compounds, and two concentrations were chosen on the basis of differences in the viability of these two cell types (Figure 3e–h). There are indications in the literature to suggest that cells in the early apoptotic phase are viable and thus are also positive in the MTT assay (
      • Gomez-Lechon M.J.
      • O’Connor E.
      • Castell J.V.
      • et al.
      Sensitive markers used to identify compounds that trigger apoptosis in cultured hepatocytes.
      ). Apoptosis assays were performed on melanocytes and keratinocytes with 12.5 and 25μM curcumin and 25 and 100μM santalol incubated for 48hours. To assess the extent of cell death induced by these two compounds, cell viability was determined via propidium iodide (PI) and annexin V staining followed by flow cytometry. The percentage of cells undergoing apoptosis (PI and annexin V double positive) was observed to be higher in keratinocytes than in melanocytes (Figure 4a). Even at low concentrations of curcumin (12.5μM) and santalol (25μM), approximately 50% of the keratinocytes were alive, whereas at the same concentrations in melanocytes the extent of cell death was negligible (Figure 4b). Although both these cell types were epidermally derived, the observed responses to chemicals were strikingly different.
      Figure thumbnail gr4
      Figure 4Curcumin and santalol alter nuclear factor E2–related factor 2 (Nrf2) and phase II detoxification transcripts in cultured human melanocytes and keratinocytes and their effect on apoptosis. Representative flow cytometric density plots of curcumin- and santalol-treated (a) keratinocytes (top) and melanocytes (bottom) are given. In each panel the lower left quadrant represents unstained live cells, and the lower right and upper right quadrants show early (annexin V-positive cells) and late (annexin V–propidium iodide (PI)-positive) apoptotic cells, respectively. Upper left quadrant represents PI-positive necrotic cells. The inset in the upper left quadrant shows the percentages of cells falling in each quadrant. (b) Percentage of total apoptotic cells detected (mean of sum of early and late apoptotic cells of four samples each) for indicated concentrations of curcumin and santalol. (c) Nrf2 levels measured by real-time PCR for each concentration of curcumin and santalol in keratinocytes and melanocytes. Effect of curcumin (12.5 and 25μM) and santalol (25 and 100μM) on the levels of phase II detoxification transcripts in (d) keratinocytes and (e) melanocytes. Error bars represent mean±SEM across five independent cultures (n=5).
      To assess the expression levels of Nrf2 and phase II detoxification genes, cells at 60% confluence were treated with the chosen concentrations of curcumin and santalol for 48hours. The Nrf2 mRNA levels showed induced expression in melanocytes, and no significant change could be observed in keratinocytes (Figure 4c). For the detoxification genes, HO-1 showed increased levels in keratinocytes on treatment with curcumin and santalol (Figure 4d). In contrast, melanocytes showed an increase in all four transcript levels (Figure 4e). It is tempting to speculate that keratinocytes succumb to mounting levels of oxidative stress because of their inability to induce Nrf2. Melanocytes, by virtue of elevated levels of Nrf2 and also the phase II transcripts, resisted apoptosis induced by curcumin and santalol.

      Effect of curcumin and santalol on lesional and nonlesional epidermis

      Studies on cultured cells in the presence of curcumin and santalol have shown that both Nrf2 and phase II detoxification genes could be induced differentially in the two epidermis cell types. We therefore decided to investigate the overall oxidative homeostasis in epidermis using these two electrophilic reagents. For this experiment we used organ culture of intact skin biopsies (
      • Lateef H.
      • Stevens M.J.
      • Varani J.
      All-trans-retinoic acid suppresses matrix metalloproteinase activity and increases collagen synthesis in diabetic human skin in organ culture.
      ). The biopsies were disinfected with 70% ethanol, rinsed with antibiotics, and placed in M254 medium containing curcumin or santalol for 48hours. These experiments were performed with 12.5μM curcumin and 25μM santalol. Under these conditions the expression of several housekeeping genes was minimally perturbed.
      In the nonlesional skin, both curcumin and santalol caused an increase in the levels of all four phase II transcripts (Figure 5a), although the change in GCLC seems to be marginal. Interestingly, in the lesional skin these compounds did not show an analogous increase in the levels of transcripts (Figure 5b). It seems that the elevated basal levels of phase II transcripts are not further altered by these compounds. The foreskin also did not show corresponding changes in the transcripts as observed for the nonlesional skin (Figure 5c). Corroborating these observations, corresponding changes in the Nrf2 levels were also observed in the nonlesional skin (Figure 5d). Immunohistochemical staining of melanocytes in the nonlesional and lesional skin shows the differences in melanocyte numbers for the two matched skins (Figure 5e and f). Our results thus strongly suggest that the cellular physiology of lesional epidermis is indeed perturbed in comparison with nonlesional skin.
      Figure thumbnail gr5
      Figure 5Effect of curcumin and santalol on whole epidermis. (a) Nonlesional and (b) lesional vitiligo skin biopsies and (c) foreskin were cultured in the presence of 12.5μM curcumin and 25μM santalol for 48hours, and the levels of phase II detoxification transcripts were quantitated by real-time PCR in the epidermis. Error bars represent mean±SEM across four individuals. Dotted line represents a twofold upregulation threshold, which is considered significant. (d) Fold changes in the nuclear factor E2–related factor 2 (Nrf2) levels upon curcumin and santalol treatment in nonlesional and lesional skin of a vitiligo patient and in foreskin. Immunohistochemistry using S-100 antibody for detecting melanocytes in the (e) nonlesional and (f) lesional skin of the vitiligo patient. Dark red–stained cells are S100-positive melanocytes. Scale bar=10μm.

      Discussion

      Oxidative stress has a significant role in the induction of several inflammatory skin diseases (
      • Bickers D.R.
      • Athar M.
      Oxidative stress in the pathogenesis of skin disease.
      ). Increased oxidative stress has also been implicated in the etiology of depigmented patches in vitiligo (
      • Passi S.
      • Grandinetti M.
      • Maggio F.
      • et al.
      Epidermal oxidative stress in vitiligo.
      ;
      • Schallreuter K.U.
      Successful treatment of oxidative stress in vitiligo.
      ). The transcription factor Nrf2 is known to regulate the expression of a network of cytoprotective enzymes, resulting in protection against toxicity after exposure to oxidative stress or various electrophilic and oxidative chemicals. In this study we described the transcriptional activation of Nrf2 as well as downstream phase II detoxification genes from the lesional regions. Comparative analysis between lesional and nonlesional skin from the same individuals across several subjects showed upregulation of NQO-1, GCLC, and GCLM genes in lesional skin. Interestingly, the expression of HO-1, another phase II detoxification response gene, is not altered in the majority of individuals. Indeed, several studies have reported that HO-1 regulation may be more complex (
      • Zhang J.
      • Ohta T.
      • Maruyama A.
      • et al.
      BRG1 interacts with Nrf2 to selectively mediate HO-1 induction in response to oxidative stress.
      ;
      • Baglole C.J.
      • Sime P.J.
      • Phipps R.P.
      Cigarette smoke-induced expression of heme oxygenase-1 in human lung fibroblasts is regulated by intracellular glutathione.
      ). Across all 15 subjects, the lesional level of Nrf2 was also consistently higher than in the matched nonlesions. Interestingly, under basal conditions Nrf2 is bound to Keap1 in the cytoplasm, and the dissociation of this complex is critical to the activation of Nrf2. In certain instances, mRNA regulation of Nrf2 has also been demonstrated (
      • Gong P.
      • Cederbaum A.I.
      Nrf2 is increased by CYP2E1 in rodent liver and HepG2 cells and protects against oxidative stress caused by CYP2E1.
      ), and Nrf2 is proposed to autoregulate its own expression through an antioxidant response element–like element located in the proximal region of its promoter, leading to persistent nuclear accumulation of Nrf2 and protracted induction of phase II genes (
      • Miao W.
      • Hu L.
      • Scrivens P.J.
      • et al.
      Transcriptional regulation of NF-E2 p45-related factor (NRF2) expression by the aryl hydrocarbon receptor-xenobiotic response element signaling pathway: direct cross-talk between phase I and II drug-metabolizing enzymes.
      ). In this context, consistent upregulation of Nrf2 mRNA and phase II detoxification genes in lesional skin gains relevance.
      We also independently observed that the levels of Nrf2 mRNA and of the detoxification genes seem to be higher in foreskin samples than in normal skin, based on data from about five samples. Although the normal skin and foreskin samples were from unmatched individuals, the consistent pattern of high Nrf2 expression and lack of induction of phase II genes suggests that the foreskin may be under high oxidative stress. Studies using keratinocytes from foreskin have shown an augmented innate immune response owing to the presence of microflora (
      • Chung W.O.
      • Dale B.A.
      Innate immune response of oral and foreskin keratinocytes: utilization of different signaling pathways by various bacterial species.
      ). This could be a reason for the increased Nrf2 and phase II genes observed in this skin type. For GCLC and GCLM, the median levels in nonlesional skin were seemingly higher than in normal epidermis from individuals not suffering from vitiligo. Although the number of normal samples analyzed was small, this pattern suggests that the nonlesional skin was also under oxidative stress, although at a lower level than in the corresponding lesional skin.
      Our results showed that the nonlesional skin was able to actively respond to the electrophilic compounds curcumin and santalol and induce a classic Nrf2-dependent antioxidant response. The lesional punch samples, however, did not respond in an analogous manner. We suggest that this was due to the elevated levels of phase II genes in the lesional skin. The effects of curcumin and santalol on keratinocytes and melanocytes in culture highlight the differential responses between the two cell types. Whereas keratinocytes do not respond to the compounds by elevating Nrf2 and phase II genes and eventually succumb to apoptosis, melanocytes upregulate the Nrf2 pathway and are well protected. This suggests an integrated role of Nrf2-dependent phase II genes in oxidative stress–induced apoptosis. Although the cells provide a suitable model for studying skin homeostasis, the clear-cut differences with the skin punch suggest a complex behavior of this tissue. Thus, our studies using skin punch biopsies provided an alternative organ model and revealed oxidative stress regulation in vitiligo.
      The roles of Nrf2-dependent pathways have begun to emerge from various model systems, ranging from pulmonary disorders to cancer. The beneficial effect of enhanced Nrf2 effectors seems to be the prime protective function conferred by this pathway on tissue homeostasis. Apart from the protective response, it is noteworthy that the consequences of this increased response of phase II genes could be in sustaining the depigmented lesions in vitiligo patches. Although oxidative stress in lesional and nonlesional skin has been reported previously, our study demonstrates the involvement of Nrf2 and phase II genes during skin homeostasis in vitiligo. We also showed that nonlesional pigmented skin from vitiligo patients does not undergo the same level of oxidative stress. Thus, therapeutic intervention in this altered oxidative stress pathway may have significant relevance in the management and treatment of vitiligo.

      Materials and Methods

      Skin punches used for the experiments were taken from patients after obtaining informed written consent; the procedure is elaborated in the Supplementary Methods online. The procedures followed were approved by the institutional human ethics committee at the National Institute of Immunology and Dr Ram Manohar at Lohia Hospital, and were in agreement with the Declaration of Helsinki Principles.

      Real-time PCR for phase II detoxification gene transcripts

      Gene expression analysis was carried out using the ABI Prism 7000 real-time PCR system (Applied Biosystems, Foster City, CA). Reactions (10μl) were carried out in replicates using TaqMan universal PCR master mix (Applied Biosystems). FAM- and TAMRA-labeled TaqMan probes used in the study were Hs00157965_m1 (HO-1), Hs00168547_m1 (NQO-1), Hs00155249_m1 (GCLC), and Hs00157649_m1 (GCLM); normalization was carried out using 18S rRNA probes labeled with VIC and MGB 4319413E-0710034. For Nrf2, the primer sequences used were 5′-GCTCATACTCTTTCCGTCGC-3′ and 5′-ATCATGATGGACTTGGAGC-3′ and for 18S rRNA, 5′-CGAAAGCATTTGCCAAGAAT-3′ and 5′-AGTCGGCATCGTTTATGGTC-3′ were used. Real time PCR was carried out using the SYBR Green (Fermentas Intl, Ontario, Canada) method. All quantitative PCR reactions were initiated at 50°C for 2minutes, 95°C for 10minutes, followed by a cycling condition of 95°C for 0.15minutes and 60°C for 1minute for 40 cycles. The data were analyzed using SDS 1.2 sequence-detection software (Applied Biosystems), and Ct values were obtained.

      Data analysis

      For comparison, the Ct method (ΔCt value) was used. Ct values were normalized to their corresponding 18S rRNA levels to obtain ΔCt values. The ΔCt values were calculated by subtracting the value of 18S RNA from that of the gene of interest. The reactions were performed in replicates, and ΔCt values that differed by an SD of ≥0.5 between the replicates were eliminated from further analysis. Comparison of the median ΔCt values was carried out using the Mann–Whitney test. Fold change was calculated using 2−ΔΔCt, where ΔΔCt represents ΔCt of vitiligo–ΔCt of nonlesional skin. Changes in expression level were calculated as fold change in the level of transcript between nonlesional and lesional skin and normalized to 18S rRNA levels. Paired Student's t-test was performed to study the significance of the ΔCt value pairs. All statistical analyses performed in this study were carried out using Graph-Pad Prism software (San Diego, CA).

      Apoptosis assay

      At 48hours after the aforesaid treatment, both floating and adherent cells were harvested and stained with annexin V–FITC and PI as described by the manufacturer (Annexin V–FITC detection kit, Promokine, Heidelberg, Germany). In brief, 5μl of annexin V–FITC and 5μl of PI were added to cells after washing and resuspension in 500μl of binding buffer. The mixture was incubated in the dark for 5minutes at room temperature. FITC and PI fluorescence was measured using a BD LSR Flow cytometer (BD Biosciences, Franklin Lake, NJ). Approximately 30,000 cells were analyzed for each sample. All data analysis was performed using WinMDI version 2.9 software (Joseph Totter, Scripps Research Institute, San Diego, CA).

      Immunohistochemistry for Nrf2 and S100

      Immunohistochemistry was performed using standard protocols. Briefly, skin biopsies were fixed with 10% formalin, dehydrated, and embedded in paraffin. Sections (4μm thick) were placed on polylysine-coated slides, deparaffinized, hydrated, antigen retrieved with 10mM citrate buffer pH 6.0, and blocked with 5% normal goat serum. Slides were stained with rabbit monoclonal antibodies to Nrf2 (dilution 1:100; Epitomics, Burlingame, CA) and S100 (dilution 1:100; Dako Cytomation, Glostrup, Denmark) for 45minutes at room temperature. Slides were washed and incubated with labeled polymer alkaline phosphatase (Dako Cytomation) for 30minutes at room temperature. Reaction was developed with fast red for Nrf2 staining and with fuchsin for S100 staining for 5minutes each, and both were counterstained with hematoxylin.

      ACKNOWLEDGMENTS

      RSG is the recipient of a Swarnajayanti Fellowship from the Department of Science and Technology and is also supported by a Tata Innovative Fellowship from the Department of Biotechnology, India. This work was supported by grants for the “Program Support for skin pigmentation and melanocyte-keratinocyte biology” from the Department of Biotechnology, India. AAK is a Junior Research Fellow of the Council of Scientific and Industrial Research, India. We thank Tarun Chopra for his initial help in culturing melanocytes from the epidermis and Chandrima Shaha for helping in IHC staining.

      SUPPLEMENTARY MATERIAL

      Supplementary material is linked to the online version of the paper at http://www.nature.com/jid

      REFERENCES

        • Arican O.
        • Kurutas E.B.
        Oxidative stress in the blood of patients with active localized vitiligo.
        Acta Dermatovenerol Alp Panonica Adriat. 2008; 17: 12-16
        • Baglole C.J.
        • Sime P.J.
        • Phipps R.P.
        Cigarette smoke-induced expression of heme oxygenase-1 in human lung fibroblasts is regulated by intracellular glutathione.
        Am J Physiol Lung Cell Mol Physiol. 2008; 295: L624-L636
        • Bickers D.R.
        • Athar M.
        Oxidative stress in the pathogenesis of skin disease.
        J Invest Dermatol. 2006; 126: 2565-2575
        • Brenner M.
        • Hearing V.J.
        The protective role of melanin against UV damage in human skin.
        Photochem Photobiol. 2008; 84: 539-549
        • Chung W.O.
        • Dale B.A.
        Innate immune response of oral and foreskin keratinocytes: utilization of different signaling pathways by various bacterial species.
        Infect Immun. 2004; 72: 352-358
        • Dell’anna M.L.
        • Picardo M.
        A review and a new hypothesis for non-immunological pathogenetic mechanisms in vitiligo.
        Pigment Cell Res. 2006; 19: 406-411
        • Gauthier Y.
        • Cario Andre M.
        • Taieb A.
        A critical appraisal of vitiligo etiologic theories. Is melanocyte loss a melanocytorrhagy?.
        Pigment Cell Res. 2003; 16: 322-332
        • Gomez-Lechon M.J.
        • O’Connor E.
        • Castell J.V.
        • et al.
        Sensitive markers used to identify compounds that trigger apoptosis in cultured hepatocytes.
        Toxicol Sci. 2002; 65: 299-308
        • Gong P.
        • Cederbaum A.I.
        Nrf2 is increased by CYP2E1 in rodent liver and HepG2 cells and protects against oxidative stress caused by CYP2E1.
        Hepatology. 2006; 43: 144-153
        • Guan C.P.
        • Wei X.D.
        • Chen H.Y.
        • et al.
        [Abnormal nuclear translocation of nuclear factor-E2 related factor 2 in the lesion of vitiligo].
        Zhonghua Yi Xue Za Zhi. 2008; 88: 2403-2406
        • Guan C.P.
        • Zhou M.N.
        • Xu A.E.
        • et al.
        The susceptibility to vitiligo is associated with NF-E2-related factor2 (Nrf2) gene polymorphisms: a study on Chinese Han population.
        Exp Dermatol. 2008; 17: 1059-1062
        • Ines D.
        • Sonia B.
        • Riadh B.M.
        • et al.
        A comparative study of oxidant-antioxidant status in stable and active vitiligo patients.
        Arch Dermatol Res. 2006; 298: 147-152
        • Jeong S.O.
        • Oh G.S.
        • Ha H.Y.
        • et al.
        Dimethoxycurcumin, a synthetic curcumin analogue, induces heme oxygenase-1 expression through Nrf2 activation in RAW264.7 macrophages.
        J Clin Biochem Nutr. 2009; 44: 79-84
        • Kensler T.W.
        • Wakabayashi N.
        • Biswal S.
        Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway.
        Annu Rev Pharmacol Toxicol. 2007; 47: 89-116
        • Kovacs D.
        • Raffa S.
        • Flori E.
        • et al.
        Keratinocyte growth factor down-regulates intracellular ROS production induced by UVB.
        J Dermatol Sci. 2009; 54: 106-113
        • Lateef H.
        • Stevens M.J.
        • Varani J.
        All-trans-retinoic acid suppresses matrix metalloproteinase activity and increases collagen synthesis in diabetic human skin in organ culture.
        Am J Pathol. 2004; 165: 167-174
        • Le Poole I.C.
        • Luiten R.M.
        Autoimmune etiology of generalized vitiligo.
        Curr Dir Autoimmun. 2008; 10: 227-243
        • Lei T.C.
        • Vieira W.D.
        • Hearing V.J.
        In vitro migration of melanoblasts requires matrix metalloproteinase-2: implications to vitiligo therapy by photochemotherapy.
        Pigment Cell Res. 2002; 15: 426-432
        • Marrot L.
        • Jones C.
        • Perez P.
        • et al.
        The significance of Nrf2 pathway in (photo)-oxidative stress response in melanocytes and keratinocytes of the human epidermis.
        Pigment Cell Melanoma Res. 2008; 21: 79-88
        • Miao W.
        • Hu L.
        • Scrivens P.J.
        • et al.
        Transcriptional regulation of NF-E2 p45-related factor (NRF2) expression by the aryl hydrocarbon receptor-xenobiotic response element signaling pathway: direct cross-talk between phase I and II drug-metabolizing enzymes.
        J Biol Chem. 2005; 280: 20340-20348
        • Motohashi H.
        • Yamamoto M.
        Nrf2-Keap1 defines a physiologically important stress response mechanism.
        Trends Mol Med. 2004; 10: 549-557
        • Namazi M.R.
        Neurogenic dysregulation, oxidative stress, autoimmunity, and melanocytorrhagy in vitiligo: can they be interconnected?.
        Pigment Cell Res. 2007; 20: 360-363
        • Nguyen T.
        • Nioi P.
        • Pickett C.B.
        The Nrf2-antioxidant response element signaling pathway and its activation by oxidative stress.
        J Biol Chem. 2009; 284: 13291-13295
        • Numata I.
        • Okuyama R.
        • Memezawa A.
        • et al.
        Functional expression of heme oxygenase-1 in human differentiated epidermis and its regulation by cytokines.
        J Invest Dermatol. 2009; 129: 2594-2603
        • Passi S.
        • Grandinetti M.
        • Maggio F.
        • et al.
        Epidermal oxidative stress in vitiligo.
        Pigment Cell Res. 1998; 11: 81-85
        • Rokos H.
        • Beazley W.D.
        • Schallreuter K.U.
        Oxidative stress in vitiligo: photo-oxidation of pterins produces H(2)O(2) and pterin-6-carboxylic acid.
        Biochem Biophys Res Commun. 2002; 292: 805-811
        • Rushmore T.H.
        • King R.G.
        • Paulson K.E.
        • et al.
        Regulation of glutathione S-transferase Ya subunit gene expression: identification of a unique xenobiotic-responsive element controlling inducible expression by planar aromatic compounds.
        Proc Natl Acad Sci USA. 1990; 87: 3826-3830
        • Schallreuter K.U.
        Successful treatment of oxidative stress in vitiligo.
        Skin Pharmacol Appl Skin Physiol. 1999; 12: 132-138
        • Schallreuter K.U.
        • Bahadoran P.
        • Picardo M.
        • et al.
        Vitiligo pathogenesis: autoimmune disease, genetic defect, excessive reactive oxygen species, calcium imbalance, or what else?.
        Exp Dermatol. 2008; 17: 139-140
        • Schallreuter K.U.
        • Gibbons N.C.
        • Zothner C.
        • et al.
        Butyrylcholinesterase is present in the human epidermis and is regulated by H2O2: more evidence for oxidative stress in vitiligo.
        Biochem Biophys Res Commun. 2006; 349: 931-938
        • Schallreuter K.U.
        • Kruger C.
        • Wurfel B.A.
        • et al.
        From basic research to the bedside: efficacy of topical treatment with pseudocatalase PC-KUS in 71 children with vitiligo.
        Int J Dermatol. 2008; 47: 743-753
        • Shalbaf M.
        • Gibbons N.C.
        • Wood J.M.
        • et al.
        Presence of epidermal allantoin further supports oxidative stress in vitiligo.
        Exp Dermatol. 2008; 17: 761-770
        • Spencer J.D.
        • Gibbons N.C.
        • Rokos H.
        • et al.
        Oxidative stress via hydrogen peroxide affects proopiomelanocortin peptides directly in the epidermis of patients with vitiligo.
        J Invest Dermatol. 2007; 127: 411-420
        • Spritz R.A.
        The genetics of generalized vitiligo.
        Curr Dir Autoimmun. 2008; 10: 244-257
        • Tirumalai R.
        • Rajesh Kumar T.
        • Mai K.H.
        • et al.
        Acrolein causes transcriptional induction of phase II genes by activation of Nrf2 in human lung type II epithelial (A549) cells.
        Toxicol Lett. 2002; 132: 27-36
        • Tobin D.J.
        • Swanson N.N.
        • Pittelkow M.R.
        • et al.
        Melanocytes are not absent in lesional skin of long duration vitiligo.
        J Pathol. 2000; 191: 407-416
        • Wakabayashi N.
        • Itoh K.
        • Wakabayashi J.
        • et al.
        Keap1-null mutation leads to postnatal lethality due to constitutive Nrf2 activation.
        Nat Genet. 2003; 35: 238-245
        • Westerhof W.
        • d’Ischia M.
        Vitiligo puzzle: the pieces fall in place.
        Pigment Cell Res. 2007; 20: 345-359
        • Yang C.
        • Zhang X.
        • Fan H.
        • et al.
        Curcumin upregulates transcription factor Nrf2, HO-1 expression and protects rat brains against focal ischemia.
        Brain Res. 2009; 1282: 133-141
        • Zhang J.
        • Ohta T.
        • Maruyama A.
        • et al.
        BRG1 interacts with Nrf2 to selectively mediate HO-1 induction in response to oxidative stress.
        Mol Cell Biol. 2006; 26: 7942-7952
        • Zhou M.N.
        • Xu A.E.
        • Lu L.J.
        • et al.
        [Association of single nucleotide polymorphisms of Nrf2 promoter region with susceptibility to vitiligo].
        Zhonghua Yi Xue Za Zhi. 2008; 88: 969-972
        • Zou C.G.
        • Banerjee R.
        Homocysteine and redox signaling.
        Antioxid Redox Signal. 2005; 7: 547-559