Impaired Activation of the Nrf2-ARE Signaling Pathway Undermines

ORIGINAL ARTICLE

Impaired Activation of the Nrf2-ARE Signaling Pathway Undermines H2O2-Induced Oxidative Stress Response: A Possible Mechanism for Melanocyte Degeneration in Vitiligo Zhe Jian1,2, Kai Li1,2, Pu Song1,2, Guannan Zhu1, Longfei Zhu1, Tingting Cui1, Bangmin Liu1, Lingzhen Tang1, Xiaowen Wang1, Gang Wang1, Tianwen Gao1 and Chunying Li1 Vitiligo melanocytes possess higher susceptibility to oxidative insults. Consistent with this, impairment of the antioxidant defense system has been reported to be involved in the onset and progression of vitiligo. Our previous study showed that the nuclear factor E2-related factor 2–antioxidant response element (Nrf2-ARE) pathway and its downstream antioxidant enzyme heme oxygenase-1 (HO-1) are crucial for melanocytes to cope with H2O2-induced oxidative damage. Here, we sought to determine whether the diminished Nrf2-ARE activity that contributes to reduced downstream antioxidant enzymes and increased oxidative stress could be the reason why melanocytes are more vulnerable to vitiligo. We found that vitiligo melanocytes exhibited hypersensitivity to H2O2-induced oxidative injury because of reduced Nrf2 nuclear translocation and transcriptional activity, which led to decreased HO-1 expression and aberrant redox balance. Moreover, we also found that the level of serum HO-1 was significantly decreased and that of IL-2 was markedly increased in 113 vitiligo patients when compared with healthy controls. These data demonstrate that impaired activation of Nrf2 under oxidative stress could result in decreased expression of antioxidant enzymes and increased death of vitiligo melanocytes. Journal of Investigative Dermatology advance online publication, 17 April 2014; doi:10.1038/jid.2014.152

INTRODUCTION Vitiligo is a common skin depigmenting disorder characterized by the death of melanocytes from lesional epidermis (Schallreuter et al., 2008; Spritz, 2008). Despite much research, the etiology of vitiligo and the causes of melanocyte death are not yet fully understood. Recent experimental and clinical evidence suggests that oxidative stress and 1

Department of Dermatology of Xijing Hospital, Fourth Military Medical University, Xi’an, China

2

These authors contributed equally to this work.

Correspondence: Tianwen Gao or Chunying Li, Department of Dermatology, Xijing Hospital, Fourth Military Medical University, Xi’an 710032, China. E-mail: [email protected] or [email protected] Abbreviations: ARE, antioxidant response element; CAT, catalase; DMEM, Dulbecco’s Modified Eagle’s Medium; DTNB, 5, 5-dithio-bis (2-nitrobenzoic) acid; ER, endoplasmic reticulum; FRDA, Friedreich ataxia; GAPDH, glyceraldehyde 3 phosphate dehydrogenase; GCLC, g-glutamyl cystine ligase catalytic subunit; GCLM, g-glutamyl cystine ligase modulatory subunit; GPx, glutathione peroxidase; GSH, reduced glutathione; GSHt, total glutathione; GSSG, oxidized disulfide; GST, glutathione-S-transfereases; H2DCF-DA, dihydrodichlorofluorescein diacetate; HO-1, heme oxygenase-1; Keap1, Kelch-like ECH-associated protein; MDA, malondialdehyde; NQO1, NAD(P)H: quinone reductase; Nrf2, nuclear factor E2-related factor 2; ROS, reactive oxygen species; SOD, superoxide dismutase; TBA, thiobarbituric acid; TNB, 5-thio-2-nitrobenzoic acid Received 20 November 2013; revised 7 February 2014; accepted 25 February 2014; accepted article preview online 24 March 2014

& 2014 The Society for Investigative Dermatology

accumulation of free radicals in the epidermis of affected skin leads to melanocyte degeneration (Taieb, 2000; Gauthier et al., 2003). However, the exact disappearance mechanisms of melanocytes lack a clear explanation. Studies have demonstrated that melanocytes involved in vitiligo may have inherent aberrations that make them more vulnerable to extracellular insults such as chemical oxidants or UVB, both of which are able to induce reactive oxygen species (ROS) generation (Maresca et al., 1997; Jimbow et al., 2001; Kroll et al., 2005; Lee et al., 2005). Among the great variety of ROS, hydrogen peroxide (H2O2) has a pivotal role in the onset and progression of vitiligo (Schallreuter et al., 2007). This point was supported by the increased levels of H2O2 detected in the epidermis of vitiligo patients (Schallreuter et al., 1999; Shalbaf et al., 2008). Vitiligo melanocytes display cellular and biological impairments when compared with normal melanocytes, namely: (a) low proliferation rate and inefficient subculture (Puri et al., 1987), (b) dilative endoplasmic reticulum (ER) and aberrant melanosome clustering (Boissy et al., 1991; Im et al., 1994), (c) grossly enlarged and retracted dendrites (Iyengar and Misra, 1988), (d) defective arrangement of the membrane lipids and altered lipid components (Dell’Anna et al., 2007), and (e) lower levels of decay accelerating factor (CD 55) and membrane cofactor protein (CD 46) that could render them www.jidonline.org

1

Z Jian et al. Faulty Nrf2-HO-1 Signaling in Vitiligo Melanocytes

more vulnerable to autologous complement attack (van den Wijngaard et al., 2002). In addition, recent studies have confirmed an imbalance in the intracellular redox status in vitiligo patients in peripheral blood, lesional epidermis, and cultured melanocytes from the perilesional skin of vitiligo patients (Ines et al., 2006; Arican and Kurutas, 2008; Khan et al., 2009). Actually, several antioxidant proteins have also been found to be altered in patients with vitiligo, such as catalase (CAT), superoxide dismutase (SOD), glutathione peroxidase (GPx), and vitamins C and E (Ines et al., 2006; Arican and Kurutas, 2008). However, thus far, the reason why vitiligo melanocytes are hypersensitive to oxidative stress has not been elucidated. Previous studies indicate that vitiligo keratinocytes are vulnerable to apoptosis by TNF-a treatment because of reduced activation of NF-kB via impaired PI3K/Akt activation (Kim et al., 2007). On the basis of these findings, identification of signaling pathways that account for redox system imbalance and melanocyte disappearance may provide a better understanding of the pathomechanisms underlying vitiligo and possibly improve the therapeutic approach. We recently revealed that the nuclear factor E2-related factor 2–antioxidant response element (Nrf2-ARE) antioxidant pathway and its downstream antioxidant enzyme heme oxygenase-1 (HO-1) have a crucial role in the ability of melanocytes to cope with H2O2-induced oxidative stress (Jian et al., 2011). It is known that the key transcription factor Nrf2 regulates the expression of phase II detoxifying and antioxidant genes by binding to the ARE sequence (Kensler et al., 2007). Under unstimulated conditions, Nrf2 is sequestered in the cytosol, where it is bound to Kelch-like ECH-associated protein 1 (Keap1) (Hayes and McMahon, 2001). When cells are under oxidative stress, Nrf2 is rapidly released from Keap1; it then translocates to the nucleus where it binds to ARE and induces the phase II antioxidant genes (Dinkova-Kostova et al., 2002). These genes encode heme oxygenase-1 (HO-1), CAT, SOD, glutathione-Stransferase (GST), GPx, NADH quinone oxidoreductase 1 (NQO1), glutamate-cysteine ligase catalytic subunit (GCLC), and glutamyl cystine ligase modulatory subunit (GCLM) (Zhu et al., 2005), which are important antioxidants in melanocytes. Given that the inability of melanocytes to deal with oxidative stress has been implicated in vitiligo, and the importance of the Nrf2-ARE pathway in the defense against oxidative stress, we hypothesized that dysfunction in the Nrf2ARE signaling pathway may result in increased sensitivity to H2O2-induced oxidative insult of vitiligo melanocytes. RESULTS Decreased expression of HO-1 and increased sensitivity of vitiligo melanocytes to H2O2-induced oxidative stress

To investigate whether vitiligo melanocytes are more susceptible to oxidative stress, we used H2O2 as a means of administering oxidative stress and performed an MTS assay to measure cell viability. No remarkable difference in cell viability was found between the epidermal primary normal human melanocytes taken from healthy controls (control) and 2

Journal of Investigative Dermatology

the perilesional skin samples taken from patients with vitiligo (vitiligo) when H2O2 was absent, whereas the presence of 1.0 mM H2O2 caused a statistically significant (Po0.01) reduction in vitiligo melanocyte cell viability compared with control melanocytes (Figure 1a). These data demonstrate that vitiligo melanocytes are more vulnerable to H2O2-induced oxidative stress. Our previous study confirmed that HO-1 is the main target antioxidant gene of the Nrf2-ARE pathway in protecting human melanocytes from H2O2-induced oxidative stress (Jian et al., 2011). Therefore, we measured the expression of HO-1 mRNA and protein levels in the control and vitiligo melanocytes with or without exposure to H2O2 treatment. In control melanocytes, H2O2 led to a greater than fivefold increase in HO-1 mRNA, whereas in vitiligo melanocytes, the increases in HO-1 mRNA were far smaller, with only a less than threefold increase compared with that in the H2O2untreated group (Figure 1b). Similarly, a significant difference was also observed in HO-1 protein levels (Figure 1c and d). However, the basal expression of HO-1 in normal control and vitiligo melanocytes showed no significant difference (Figure 1b–d). Abnormal cellular sublocalization of the Nrf2 transcription factor and impaired Nrf2 transcriptional activity in vitiligo melanocytes

Transcription of inducible antioxidant genes is chiefly controlled by the Nrf2 transcription factor (Zhu et al., 2005). Therefore, we examined the localization of Nrf2 in control and vitiligo melanocytes by laser confocal scanning microscopy. As shown in Figure 2a, nuclear Nrf2 localization was mainly observed in the control melanocytes, whereas vitiligo melanocytes primarily showed cytoplasmic staining. Next, we performed western blot analyses of nuclei-enriched and cytoplasmic fractions from control and vitiligo melanocytes. Under normal conditions, the expression of nuclear Nrf2 in the control melanocytes was higher than that in the vitiligo melanocytes, whereas the expression of cytoplasmic Nrf2 in the vitiligo melanocytes was higher than that in the control melanocytes. Moreover, in the nuclei-enriched fractions of control and vitiligo melanocytes, Nrf2 content increased in response to H2O2 (Figure 2b and c). Far smaller Nrf2 increases were seen in the nuclei-enriched fractions of vitiligo melanocytes (Figure 2b and c). This suggests that nuclear translocation of Nrf2 is relatively reduced in vitiligo melanocytes in response to H2O2-induced oxidative stress. To compare the Nrf2-ARE transcriptional activity of control with vitiligo melanocytes, we used a dual-luciferase reporter assay to measure ARE luciferase activity in the presence and absence of H2O2. PIG1 (a normal human epidermal melanocyte cell line) and PIG3V (a vitiligo melanocyte cell line) cells were transiently transfected with an ARE-luciferase reporter plasmid (pGL3-ARE) and a Renilla luciferase reporter plasmid (pRLtk), and treated with 1.0 mM H2O2 for 6, 12, and 24 h before the luciferase activity was measured. As shown in Figure 2d, treatment with H2O2 increased the transcriptional activity of Nrf2 in terms of ARE luciferase activity in a timedependent manner in both groups, whereas ARE luciferase


activity of PIG3V was much lower than that of PIG1 in the presence of H2O2. These data clearly indicate that Nrf2-driven transcriptional activation of ARE is aberrant in vitiligo melanocytes.

a

Control

Cell viability (% of control cells)

100

Vitiligo

75 * 50

25

0

b

0.0 1.0 H2O2 concentration (mM)

7

Vitiligo

HO-1 mRNA level (fold)

In order to evaluate the status of the oxidative stress and antioxidant defense system, we detected ROS levels, malondialdehyde (MDA) levels, and the activities or contents of major antioxidant enzymes, including SOD, GPx, CAT, and glutathione (GSH), in primary normal human melanocytes taken from healthy controls (control) and primary melanocytes from perilesional skin samples of vitiligo patients (vitiligo). As shown in Figure 4a and b, vitiligo melanocytes exhibit a significantly higher intracellular ROS level (54.6±7.5 in vitiligo and 14.2±4.9 in control melanocytes, Po0.001). Moreover, a higher MDA level was detected in vitiligo melanocytes (20.3±2.1 in vitiligo and 6.1±1.3 in control melanocytes, Po0.01) (Figure 4c), further supporting the hypothesis that the oxidative stress level is higher in vitiligo melanocytes.

*

5 4

*# 3 2 1 0 H2O2:

–

+

–

+

Vitiligo

Control

c HO-1

32 kDa

β-Actin

43 kDa

H2O2:

d

HO-1 expression (/β-actin expression)

1.00

–

+

Control Vitiligo

–

+

*

0.75

*#

0.50

0.25

0.00 H2O2:

–

Our previous study has shown that transfection with pCMV6XL5-Nrf2 increases the expression of Nrf2 and HO-1 in melanocytes. To investigate whether the upregulation of Nrf2 expression can restore the cell viability of vitiligo melanocytes under oxidative stress, PIG3V cells were transfected with the Nrf2 over-expression plasmid pCMV6-XL5Nrf2. Next, we used the MTS assay to investigate the cell viability in vitiligo melanocytes in response to exogenous oxidative stress induced by H2O2. In Figure 3, H2O2 caused an B40% decrease in cell viability when PIG3V cells were treated at a concentration of 1.0 mM H2O2 for 24 hours. Compared with the normal group, upregulation of Nrf2 expression by cells transfected with pCMV6-XL5-Nrf2 led to reduced cell death (o15%) under oxidative stress, whereas such changes were not observed in the mock transfection group (transfected with pCMV6-XL5) (36%). These results suggested that upregulation of Nrf2 enhanced the ability of vitiligo melanocytes to cope with H2O2-induced cytotoxicity. High oxidative stress levels and low antioxidant enzyme activities in vitiligo melanocytes

Control

6

Upregulation of Nrf2 restores the antioxidative ability of vitiligo melanocytes

+

–

+

Figure 1. Detection of cell viability and heme oxygenase-1 (HO-1) expression after treatment with or without H2O2 in control and vitiligo melanocytes. (a) Cell viability was determined with an MTS assay 24 hours after exposure to 1.0 mM H2O2. Viability was calculated as the percentage of living cells in treated compared with untreated control cells. *Po0.01 when compared with H2O2-treated control melanocytes. (b) Modulation of HO-1 mRNA expression levels was detected by real-time PCR in control and vitiligo melanocytes with or without 24-h H2O2 treatment. Data are shown as ratios of gene expression in treated cells to that in untreated controls after normalization on the basis of the expression of the GAPDH housekeeping gene. *Po0.01 compared with the control group (H2O2 untreated). *#Po0.05 compared with the control group (H2O2 treated). (c) HO-1 protein levels were measured by western blot. Representative gel blots depicting HO-1 protein by using a HO-1-specific antibody. b-Actin was used as an internal control. (d) The intensity of each band was quantified by densitometry analysis. All protein expression was normalized to that of b-actin. *Po0.05 compared with the control group (H2O2 untreated). *#Po0.01 compared with the control group (H2O2 treated). Error bars represent means±SD across three independent cultures (n ¼ 3).

www.jidonline.org

3


a

Anti-Nrf2

DAPI

100

Merge

Untreated Normal

pCMV6-XL5 pCMV6-XL5-Nrf2

1

2

Cell viability (% of control cells)

Control

*#

3

Vitiligo

50 μm

4

5

6

c Control

Vitiligo

Nuclear Nrf2

57 kDa

Cytosolic Nrf2

57 kDa

Lamin B1

68 kDa

β-Actin

43 kDa

H2O2

–

+

–

Protein expression (/lamin B1 or β-actin expression)

b

3

N-Nrf2 C-Nrf2

#

1

Luciferase activity (fold)

15.0 12.5

Δ ∇

*

–

+

–

+

Control

d

Control Vitiligo

Vitiligo

#

10.0

#

* Untreated

6h

12 h 1.0 mM H2O2

24 h

Figure 2. Nrf2 location, amount, and dual-luciferase reporter assays of antioxidant response element (ARE) activity in control and vitiligo patient melanocytes under normal and oxidative stress conditions. (a) Nrf2 localization in control and vitiligo melanocytes under normal conditions was observed by laser confocal scanning microscopy after 24 hours of culture. Nuclear Nrf2 translocation occurred in control melanocytes, but not in vitiligo melanocytes. Scale bar ¼ 50 mm. (b) Western blots of nuclear and cytoplasmic fractions from control and vitiligo melanocytes under normal conditions (lanes 1, 3) or after H2O2 treatment (lanes 2, 4). Representative gel blots depicting nuclear and cytosolic Nrf2 proteins by using specific antibodies. Fraction purity was assessed by labeling with b-actin (cytoplasm) and lamin B1 antibody (nuclei). (c) The intensity of each band was quantified by densitometry analysis. Nuclear protein expression was normalized to that of lamin B1 and cytoplasmic protein expression was normalized to that of b-actin. N-Nrf2: nuclear Nrf2, C-Nrf2: cytosolic Nrf2. *Po0.01 compared with N-Nrf2 in control group (H2O2 untreated). #Po0.05 compared with C-Nrf2 in control group (H2O2 untreated). DPo0.001 compared with N-Nrf2 in control group (H2O2 treated). ,Po0.01 compared with C-Nrf2 in control group (H2O2 treated). Error bars represent means±SD across three independent cultures (n ¼ 3). (d) PIG1 and PIG3V cells were co-transfected with an ARE-luciferase reporter plasmid (pGL3-ARE) and a Renilla luciferase reporter plasmid (pRLtk) for 24 hours and treated with 1.0 mM H2O2 for 6, 12, and 24 hours before luciferase activity was measured. Firefly luciferase activity in relative light units per second (RLU s 1) was normalized to Renilla luciferase activity and was expressed as x-fold multiples of the control, a ratio of the experimental to the control (control cells without treatment of H2O2). *Po0.05 compared with control cells (H2O2 treated for 6-, 12-, and 24-hour groups, respectively). # Po0.01 compared with control cells (H2O2 untreated).

4

50

25

0 1.0 mM H2O2 (24 h) Figure 3. Determination of cell viability in vitiligo melanocytes pretreated with pCMV6-XL5-Nrf2 at 48 hours after exposure to 1.0 mM H2O2. The immortalized vitiligo melanocytes PIG3V were pretreated with pCMV6-XL5Nrf2 or pCMV6-XL5. Cell viability was determined with an MTS assay 24 hours after exposure to 1.0 mM H2O2. Cell viability was calculated as the percentage of living cells in the treated compared with the untreated cells. Data are expressed as means±SD of three independent experiments. *Po0.01 when compared with the untreated group. *#Po0.05 compared with normal (mock transfection group treated with H2O2).

* #

2.5 0.0

*

*

7.5 5.0

*

2

0 H2O2

+

75


The SOD and GPx activities in vitiligo melanocytes were comparatively lower than in control melanocytes (Figure 4d). Interestingly, there was no significant difference in CAT activity between the two groups. In the meanwhile, low total GSH and GSH content was detected in the vitiligo group (Figure 4e). Many studies suggested that the GSH/GSSG ratio may act as an indicator of oxidative stress. As presented in Figure 4f, the GSH/GSSG ratio in the normal group was higher than in the vitiligo group. These results suggest that the antioxidant enzyme system is impaired in vitiligo melanocytes. Decreased serum HO-1 levels are correlated with increased serum IL-2 levels in patients with vitiligo

Besides abnormalities in redox balance, several biochemical abnormalities are commonly found in patients with vitiligo. We tested the serum levels of HO-1 and IL-2 in 113 patients with non-segmental vitiligo and 113 healthy control subjects by using ELISA. Serum HO-1 levels were significantly lower in patients with vitiligo compared with those in healthy controls (Figure 5a). However, the overall level of serum IL-2 in the patients with vitiligo was obviously higher than that of healthy controls (Figure 5b). In both progressive and stable stages of vitiligo, the levels of IL-2 and HO-1 were significantly different from the levels in healthy controls. As HO-1 might suppress IL-2 production by regulating T-cell activation (Pae et al., 2004), we examined the relationship between serum HO-1 levels and IL-2 levels in vitiligo patients. We found that the level of serum HO-1 was inversely correlated with that of serum IL-2 in vitiligo patients (r ¼ 0.4482, P ¼ 0.0001, n ¼ 113; Figure 5c).


a 100

Negative control

100 101

102

103 104 Vitiligo

100

Control

100

101 102 103

104 100 101 102

Activity (unit per mg protein)

* 20 15 10 5

30 20

*

10

60 50

Vitiligo

Control Vitiligo

NC

Control Vitiligo

40 30 20

*

10

* 0

SOD

GPx

CAT

f 60

* *

50

Control Vitiligo

40 30 20 10 0 Total GSH

GSH

GSSG

GSH/GSSG content (μmol per mg protein)

MDA level (μmol per mg protein)

d 25

e Content (μmol per mg protein)

40

0

c

Control

50

103 104

FL1-H

0

#

60

DCF-H fluorescence

Counts

b

17.5 15.0

*

12.5 10.0 7.5 5.0 2.5 0.0 Control

Vitiligo

Figure 4. Antioxidant enzyme activities or content, reactive oxygen species (ROS) and malondialdehyde (MDA) levels in control and vitiligo melanocytes. (a) Representative results for ROS production. (b) Bar graphic representations of the fluorescence intensity in control and vitiligo melanocytes. Bar graphic representations of (c) MDA, (d) SOD, GPx, and CAT, (e) GSH and GSSG, (f) GSH/GSSG in control and vitiligo melanocytes. Error bars represent means±SD across three independent cultures (n ¼ 3). *Po0.05 compared with control melanocytes. #Po0.01 when compared with NC (negative control).

DISCUSSION In this study, we provided a rational explanation for the hypersensitivity of vitiligo melanocytes to H2O2-induced oxidative damage. We found that vitiligo melanocytes under H2O2-induced oxidative stress conditions showed lower HO1 levels, resulting in the decline of antioxidative ability to cope with abnormally high levels of H2O2. These data support the views that melanocytes from vitiligo subjects show increased susceptibility to high levels of epidermal ROS production (Maresca et al., 1997; Jimbow et al., 2001; Boissy and Manga, 2004; Kroll et al., 2005).

Recent evidence, including our findings, indicated that nuclear transcription factor Nrf2 mainly regulates HO-1 expression, and is involved in protecting human melanocytes against H2O2-induced oxidative stress (Surh, 2003; Jian et al., 2011). Thus, we focused on the expression, localization, and transcriptional activity of Nrf2. Interestingly, although we did not observe any difference in Nrf2 expression between the control and vitiligo melanocytes (data not shown), Nrf2 nuclear localization and transcriptional activity are decreased in vitiligo melanocytes. Despite severe oxidative stress that should have induced activation of Nrf2 and www.jidonline.org

5


***

a

***

70

Total PV SV Control

***

HO-1 (pg ml–1)

60 50 40 30 20 10 0

Total

SV

PV

Control

***

b

*** 150

**

Total PV SV Control

IL-2 (pg ml–1)

125 100 75 50 25 0

Serum IL-2 (pg ml–1)

c

Total

PV

SV

Control

200

r = –0.4482 p = 0.0001 n = 113

150

100

50

0 0

10

20

30 40 50 Serum HO-1 (pg ml–1)

60

70

Figure 5. Serum HO-1 and interleukin (IL)-2 levels in healthy controls (n ¼ 113) and patients with vitiligo (n ¼ 113) and correlation between serum HO-1 and IL-2 in patients with vitiligo. (a) Decreased serum HO-1 levels and (b) increased serum IL-2 levels in patients with vitiligo. **Po0.01 and ***Po0.001 as determined by the unpaired Mann–Whitney U-test. Control, healthy controls; PV, progressive vitiligo; SV, stable vitiligo; Total, total patients. (c) Association between serum HO-1 and IL-2 levels. Serum HO-1 levels were inversely correlated with serum IL-2 levels in vitiligo patients (r ¼ 0.4482, P ¼ 0.0001, n ¼ 113, Spearman’s rank correlation test). The solid line represents the regression line.

upregulation of its downstream gene products, vitiligo melanocytes showed progressive reduction of nuclear Nrf2 content and a markedly diminished Nrf2 activation. This suggests that in vitiligo melanocytes, some process may block the activity of Nrf2 nuclear translocation, potentially 6


contributing to melanocyte dysfunction and/or loss. Similar results were found by Paupe et al. (2009) in Friedreich ataxia (FRDA) cultured fibroblasts. Moreover, lower ARE luciferase activity of PIG3V cells was detected by a dual-luciferase reporter assay with or without H2O2 stimulus. These data strongly suggested that the Nrf2-ARE pathway is dysfunctional in vitiligo melanocytes and provides an explanation for the impaired induction of antioxidants seen in these cells in response to H2O2-induced oxidative stress. Another interesting finding from our study was that upregulation of Nrf2 could restore the antioxidant capability of PIG3V cells, which allowed them to cope with H2O2-induced cytotoxicity. This was accompanied by significant upregulation of the Nrf2 target antioxidant genes, including HO-1, GST, GCLC, GCLM, and NQO1. This effect has been confirmed in human melanocytes by research from our group (Jian et al., 2011) and that by Marrot et al. (2008). In fact, many studies have found that activation of the Nrf2-ARE pathway with different means has an important protective role in several types of cells and diseases. For instance, studies in neurodegenerative disease have found that activation of Nrf2 protects against H2O2-induced PC12 cell death and apoptosis (Tanaka et al., 2010; Tusi et al., 2010). In addition, Beyer et al. (2007) demonstrated that upregulation of Nrf2 had an antiinflammatory role in skin wounds, and that activation of this pathway protects skin from carcinogenesis induced by chemicals. An increasing number of studies have demonstrated that melanocyte impairment in vitiligo may be related to increased oxidative stress. Excess production of ROS or insufficient antioxidant enzymes may cause oxidative damage in melanocytes. Recent studies using lesional epidermis and blood (serum) of vitiligo patients have found that an imbalance in redox status is a possible pathogenic clue in vitiligo (Taieb, 2000; Ines et al., 2006; Arican and Kurutas, 2008; Khan et al., 2009). SOD, CAT, GSH, and GPx are a group of antioxidant enzyme systems that scavenge ROS to prevent cell damage. In this study, we confirmed that the levels of ROS and MDA are increased and that these antioxidant systems are insufficient in vitiligo melanocytes. Previous studies have shown that the serum, plasma, and vitiliginous tissue samples from different types of vitiligo patients have higher MDA levels, which is in concordance with our findings (Yildirim et al., 2003; Koca et al., 2004; Yildirim et al., 2004; Arican and Kurutas, 2008). However, our study identified higher levels of MDA in vitiligo melanocytes. There are many conflicting studies on SOD activity in patients with vitiligo. Some studies show higher levels of SOD (Dell’Anna et al., 2001; Yildirim et al., 2003; Agrawal et al., 2004; Yildirim et al., 2004; Hazneci et al., 2005; Shajil and Begum, 2006; Arican and Kurutas, 2008), whereas others show lower levels of SOD in vitiligo patients (Arican and Kurutas, 2008; Khan et al., 2009). GPx activity in patients with vitiligo has been found to be normal (Picardo et al., 1994; Passi et al., 1998) in some studies and lower in others (Beazley et al., 1999; Shajil and Begum, 2006; Khan et al., 2009). In fact, we failed to find any significant difference in CAT activity between normal and vitiligo melanocytes in the present study.


a

b Normal Nrf2-ARE/HO-1 pathway

Aberrant Nrf2-ARE/HO-1 pathway H 2O 2

ROS

ROS Cytoplasm Epidermis

Epidermis Melanocytes

Melanocytes Dissociation Keap1

Lymphocytes IL-2 levels

Dissociation

Keap1

Nrf2

Nrf2

Increased nuclear translocation Transcriptional upregulation

HO-1

Increase

P

Keap1

Keap1

Decreased nuclear translocation P

P

Nrf2

Nrf2

ARE

ARE

P

Lymphocytes IL-2 levels

Transcriptional downregulation HO-1

Decrease

Nuclear Melanocyte

Normal (control)

Melanocytes degeneration

Vitiligo

Figure 6. Schematic proposal of the Nrf2-ARE/HO-1 pathway in normal (control) human melanocytes and vitiligo melanocytes under H2O2-induced oxidative stress. (a) Under H2O2-induced oxidative stress conditions in normal (control) human melanocytes, Nrf2 escapes from Kelch-like ECH-associated protein (Keap1) and phosphorylated Nrf2 translocates to the nucleus where it binds to antioxidant response element (ARE) and transcriptionally upregulates HO-1 gene expression. HO-1 can further protect human melanocytes from H2O2-induced oxidative damage and may decrease IL-2 levels by inhibiting T-cell activation that maintains skin homeostasis. (b) In vitiligo melanocytes, impaired activation of the Nrf2-ARE/HO-1 pathway (reduced Nrf2 nuclear translocation, decreased transcriptional activity, and aberrant antioxidant defense system) leads to excessive ROS generation and decreased HO-1 induction, which fails to inhibit H2O2-induced oxidative stress and IL-2 production. This may eventually result in melanocyte degeneration and cause vitiligo.

This is in accordance with a previous report that detected no difference in CAT levels in the erythrocytes isolated from normal and vitiligo patients by Ines et al. (2006), whereas other studies revealed lower CAT activity in vitiligo melanocytes (Schallreuter et al., 1991; Dell’Anna et al., 2001; Arican and Kurutas, 2008). Our results showed lower activities of SOD and GPx, as well as diminished levels of GSH in vitiligo melanocytes when compared with normal melanocytes. We believe that these varying results may be due to the differences in the innate levels of these enzymes in the different samples that were studied (the serum, erythrocytes, melanocytes, and epidermis). Besides, they can be influenced by the duration and activity of disease as well as the laboratory techniques. Notably, the limited sample size in our study should be taken into consideration as well, and a larger sample size is still needed to confirm our findings. It has been reported that T cells in vitiligo patients are activated (Le et al., 1993), and that the increased levels of IL-2 present in these patients (Khan et al., 2012) is a sign of T-cell activation (Rubin and Nelson, 1990). Moreover, HO-1 may suppress IL-2 production by inhibiting T-cell activation (Pae et al., 2004). Consistently, in this study, we found that the IL-2 levels in patients with vitiligo were significantly higher than those in the healthy control group, and the decreased serum HO-1 level was inversely correlated with serum IL-2

levels. A previous study found that HO-1/CO induced suppressive effects on T-cell proliferation and IL-2 secretion, possibly via its inhibition of the ERK MAP kinase pathway (Pae et al., 2004). Thus, the increased levels of IL-2 observed in the serum of patients with vitiligo may result from the decreased levels of HO-1. This phenomenon may be the result of the dysfunction of the Nrf2-ARE redox signaling pathway. Until now, no study has examined the reason why vitiligo melanocytes are more vulnerable to oxidative stress. Our findings identified impairment in the antioxidant system in vitiligo melanocytes, and showed that ROS-mediated damage in vitiligo leads to melanocyte degeneration. Nrf2 regulates the expression of various antioxidant enzymes, including CAT, GPx, and SOD, by using GSH as their substrate (Kim and Vaziri, 2010). Therefore, decreased antioxidant enzyme activity and increased ROS may be associated with impairment of the Nrf2-ARE pathway in vitiligo melanocytes. As the Nrf2-ARE pathway has a vital role in the protection against oxidative stress–induced cellular injury (Harvey et al., 2009), dysfunction of the pathway can lead to excessive ROS generation and low levels of antioxidant enzymes, which is evident in the death of vitiligo melanocytes. Our current understanding is summarized in the scheme in Figure 6. www.jidonline.org

7


In summary, we found that vitiligo melanocytes exhibit a paradoxical reduction in Nrf2 activation leading to decreased activation of downstream antioxidant molecules. We have provided a rational explanation for the hypersensitivity of vitiligo melanocytes to H2O2-induced oxidative stress. Accordingly, the impairment of the Nrf2ARE signaling pathway in vitiligo melanocytes, shown here, contributes to the severity of oxidative stress and the failure of antioxidant induction in response to oxidative stress. It also renders the intracellular environment prone to ROSmediated toxicity and aberrant redox homeostasis. Future studies are needed to determine how Nrf2 localization is regulated and why it is aberrant in vitiligo melanocytes. By further understanding this crucial pathway and its alteration in vitiligo progression, more clues about how melanocytes lose their endogenous protection may be presented, leading to potential strategies to restore this function before melanocyte loss. MATERIALS AND METHODS A detailed overview of the experimental procedure is given in the Supplementary Information online.

growth medium. Melanocytes in their third or fourth in vitro passage were used for the experiments. An immortalized normal human epidermal melanocyte cell line (PIG1) and a vitiligo melanocyte cell line (PIG3V; both gifts from Dr Caroline Le Poole, Loyola University Chicago, Maywood, IL) were similarly cultured in Medium 254 supplemented with human melanocyte growth supplement, 5% fetal bovine serum, and a penicillin-streptomycin antibiotic mix at 37 1C in the presence of 5% CO2. Normal human epidermal melanocytes were isolated from neonatal foreskin of a Caucasian infant, and vitiligo melanocytes were isolated from nonlesional skin of a 34-yearold female Caucasian vitiligo patient. These cell lines were immortalized by retroviral introduction with the HPV16 E6 and E7 genes (Le et al., 1997, 2000). Cell treatment was induced by adding H2O2 (analytical pure grade; Xi’an Chemical Reagent Factory, Xi’an, China) at 1.0 mM in serum-free, antibiotic-free DMEM (Invitrogen) for 24 hours.

Cell viability testing (MTS assay) The MTS assay was performed as described (Jian et al., 2011) using Cell Titer 96AQUeous One Solution Cell Proliferation Assay (Promega, Madison, WI).

Real-time PCR Specimens of patients and controls After obtaining informed written consent, full-thickness skin biopsies were taken from the perilesional skin (for melanocyte culture) of six patients (three men and three women; age range, 21–57 years; mean age, 37.5 years) affected by active (on the basis of the progression or appearance of lesions in the last 3 months) non-segmental symmetrical vitiligo. Samples were also obtained from six age- and sexmatched normal controls. Perilesional skin is defined as skin along the edge of the white patch. All biopsies were taken from sununexposed areas. To measure HO-1 activity and IL-2 levels, blood was obtained from 113 vitiligo patients with non-segmental disease (64 men and 49 women; 29 in active and 84 in stable; age range, 18–57 years; mean age, 33 years) and from 113 age- and sex-matched control subjects. None of the patients underwent any treatment (either systemic or topical) 3 months before the biopsies and blood samples were obtained. Autoimmune diseases or familiarity were not noted in any vitiligo patient. The local ethics committee of Xijing Hospital approved the study, which was performed in strict compliance with the principles of the Declaration of Helsinki.

Real-time PCR was performed as described (Jian et al., 2011) with SYBR Premix Ex Taq II (TaKaRa, Ohtsu, Japan) on a Chromo4 continuous fluorescence.

Western blot analysis Western blots were performed as described (Jian et al., 2011) using anti-Nrf2, anti-HO-1, anti-Lamin B1, and anti-b-Actin antibodies (Santa Cruz, CA).

Transient transfection of Nrf2 Transient transfection of Nrf2 was performed as described (Jian et al., 2011) using the mammalian expression plasmid pCMV6-XL5 (OriGene, MD).

Transient transfection and dual-luciferase reporter assay The dual-luciferase reporter assay were performed as described (Shunsuke et al., 2007) using Lipofectamine 2000 (Invitrogen) with a pGL3-ARE reporter plasmid containing three copies of ARE (Kindly donated by Dr Guodong Yang, Forth Military Medical University, Xi’an, China) and pRLtk plasmid (Promega).

Cell culture and treatment Skin specimens obtained from the perilesional skin of vitiligo patients and normal controls were used for cell culture. The epidermis was separated from the dermis after treatment with 2.4 U ml 1 dispase (Roche, Mannheim, Germany) at 4 1C overnight. The epidermal sheets were treated with 0.05% trypsin for 10 minutes to prepare a suspension of individual epidermal cells. The cells were resuspended in Medium 254 conditioned medium (Cascade Biologics/Invitrogen, Portland, OR) supplemented with a human melanocyte growth supplement (Cascade Biologics/Invitrogen), 5% fetal bovine serum (Invitrogen, San Diego, CA), and penicillin–streptomycin antibiotic mix (Invitrogen) at 37 1C in the presence of 5% CO2, and analyzed during the subconfluent phase at passages 0–2. Three days after confluence, the primary cultures were trypsinized, and the cell suspension (4 104 cells per cm2) was cultured in the melanocyte 8


Laser scanning confocal immunofluorescence microscopy Laser scanning confocal microscopy was performed as described (Shunsuke et al., 2007).

Measurement of intracellular ROS levels The ROS levels were determined as described (Lluis et al., 2007) using the ROS assay kit (Beyotime, Jiangsu, China).

Biochemical assays The contents of malondialdehyde (MDA), total glutathione (GSHt), reduced glutathione (GSH), and oxidized disulfide (GSSG), and the activity of SOD, GPx, and catalase were measured using a commercially available assay kit (Beyotime) according to the manufacturer’s instructions.


Analysis of serum HO-1 activity and IL-2 levels Serum activity of HO-1 and levels of IL-2 were measured by human HO-1 or IL-2ELISA kits (Xi Tang Biotechnology, Shanghai, China) according to the manufacturer’s instructions.

Statistical analysis The data represent the means±SD for at least three independent experiments. Statistical significance was assessed with the unpaired two-tailed Student’s t-test. The differences of serum HO-1 and IL-2 levels between patients with vitiligo and healthy controls were analyzed using the Mann–Whitney U-test. All statistical analyses were performed by GraphPad Prism (GraphPad Software 3.0; San Diego, CA) and P values o0.05 were considered statistically significant. CONFLICT OF INTEREST The authors state no conflict of interest.

ACKNOWLEDGMENTS The work described in this article was supported by the National Natural Science Foundation of China (no. 81130032 and no. 81373844). We thank Dr Caroline Le Poole for providing the immortalized human epidermal melanocyte cell line PIG1 and vitiligo melanocyte cell line PIG3V and thank Dr Hua Wang and Yu Ye for assistance with the flow cytometer analyses. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid

REFERENCES Agrawal D, Shajil EM, Marfatia YS et al. (2004) Study on the antioxidant status of vitiligo patients of different age groups in Baroda. Pigment Cell Res 17:289–94

Hayes JD, McMahon M (2001) Molecular basis for the contribution of the antioxidant responsive element to cancer chemoprevention. Cancer Lett 174:103–13 Hazneci E, Karabulut AB, Ozturk C et al. (2005) A comparative study of superoxide dismutase, catalase, and glutathione peroxidase activities and nitrate levels in vitiligo patients. Int J Dermatol 44:636–40 Im S, Hann SK, Kim HI et al. (1994) Biologic characteristics of cultured human vitiligo melanocytes. Int J Dermatol 33:556–62 Ines D, Sonia B, Riadh BM et al. (2006) A comparative study of oxidantantioxidant status in stable and active vitiligo patients. Arch Dermatol Res 298:147–52 Iyengar B, Misra RS (1988) Neural differentiation of melanocytes in vitiliginous skin. Acta Anat (Basel) 133:62–5 Jian Z, Li K, Liu L et al. (2011) Heme oxygenase-1 protects human melanocytes from H(2)O(2)-induced oxidative stress via the Nrf2-ARE pathway. J Invest Dermatol 131:1420–7 Jimbow K, Chen H, Park JS et al. (2001) Increased sensitivity of melanocytes to oxidative stress and abnormal expression of tyrosinase-related protein in vitiligo. Br J Dermatol 144:55–65 Kensler TW, Wakabayashi N, Biswal S (2007) Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway. Annu Rev Pharmacol Toxicol 47:89–116 Khan R, Gupta S, Sharma A (2012) Circulatory levels of T-cell cytokines (interleukin [IL]-2, IL-4, IL-17, and transforming growth factor-beta) in patients with vitiligo. J Am Acad Dermatol 66:510–1 Khan R, Satyam A, Gupta S et al. (2009) Circulatory levels of antioxidants and lipid peroxidation in Indian patients with generalized and localized vitiligo. Arch Dermatol Res 301:731–7 Kim HJ, Vaziri ND (2010) Contribution of impaired Nrf2-Keap1 pathway to oxidative stress and inflammation in chronic renal failure. Am J Physiol Renal Physiol 298:F662–71 Kim NH, Jeon S, Lee HJ et al. (2007) Impaired PI3K/Akt activation-mediated NF-kappaB inactivation under elevated TNF-alpha is more vulnerable to apoptosis in vitiliginous keratinocytes. J Invest Dermatol 127:2612–7 Koca R, Armutcu F, Altinyazar HC et al. (2004) Oxidant-antioxidant enzymes and lipid peroxidation in generalized vitiligo. Clin Exp Dermatol 29:406–9

Arican O, Kurutas EB (2008) Oxidative stress in the blood of patients with active localized vitiligo. Acta Dermatovenerol Alp Panonica Adriat 17:12–6

Kroll TM, Bommiasamy H, Boissy RE et al. (2005) 4-Tertiary butyl phenol exposure sensitizes human melanocytes to dendritic cell-mediated killing: relevance to vitiligo. J Invest Dermatol 124:798–806

Beazley WD, Gaze D, Panske A et al. (1999) Serum selenium levels and blood glutathione peroxidase activities in vitiligo. Br J Dermatol 141:301–3

Le PI, Boissy RE, Sarangarajan R et al. (2000) PIG3V, an immortalized human vitiligo melanocyte cell line, expresses dilated endoplasmic reticulum. In Vitro Cell Dev Biol Anim 36:309–19

Beyer TA, Auf DKU, Braun S et al. (2007) Roles and mechanisms of action of the Nrf2 transcription factor in skin morphogenesis, wound repair and skin cancer. Cell Death Differ 14:1250–4 Boissy RE, Liu YY, Medrano EE et al. (1991) Structural aberration of the rough endoplasmic reticulum and melanosome compartmentalization in longterm cultures of melanocytes from vitiligo patients. J Invest Dermatol 97:395–404 Boissy RE, Manga P (2004) On the etiology of contact/occupational vitiligo. Pigment Cell Res 17:208–14 Dell’Anna ML, Maresca V, Briganti S et al. (2001) Mitochondrial impairment in peripheral blood mononuclear cells during the active phase of vitiligo. J Invest Dermatol 117:908–13 Dell’Anna ML, Ottaviani M, Albanesi V et al. (2007) Membrane lipid alterations as a possible basis for melanocyte degeneration in vitiligo. J Invest Dermatol 127:1226–33 Dinkova-Kostova AT, Holtzclaw WD, Cole RN et al. (2002) Direct evidence that sulfhydryl groups of Keap1 are the sensors regulating induction of phase 2 enzymes that protect against carcinogens and oxidants. Proc Natl Acad Sci USA 99:11908–13 Gauthier Y, Cario AM, Taieb A (2003) A critical appraisal of vitiligo etiologic theories. Is melanocyte loss a melanocytorrhagy? Pigment Cell Res 16:322–32 Harvey CJ, Thimmulappa RK, Singh A et al. (2009) Nrf2-regulated glutathione recycling independent of biosynthesis is critical for cell survival during oxidative stress. Free Radic Biol Med 46:443–53

Le PI, Das PK, van den Wijngaard RM et al. (1993) Review of the etiopathomechanism of vitiligo: a convergence theory. Exp Dermatol 2:145–53 Le PI, van den Berg FM, van den Wijngaard RM et al. (1997) Generation of a human melanocyte cell line by introduction of HPV16 E6 and E7 genes. In Vitro Cell Dev Biol Anim 33:42–9 Lee AY, Kim NH, Choi WI et al. (2005) Less keratinocyte-derived factors related to more keratinocyte apoptosis in depigmented than normally pigmented suction-blistered epidermis may cause passive melanocyte death in vitiligo. J Invest Dermatol 124:976–83 Lluis JM, Buricchi F, Chiarugi P et al. (2007) Dual role of mitochondrial reactive oxygen species in hypoxia signaling: activation of nuclear factor{kappa}B via c-SRC and oxidant-dependent cell death. Cancer Res 67:7368–77 Maresca V, Roccella M, Roccella F et al. (1997) Increased sensitivity to peroxidative agents as a possible pathogenic factor of melanocyte damage in vitiligo. J Invest Dermatol 109:310–3 Marrot L, Jones C, Perez P et al. (2008) The significance of Nrf2 pathway in (photo)-oxidative stress response in melanocytes and keratinocytes of the human epidermis. Pigment Cell Melanoma Res 21:79–88 Pae HO, Oh GS, Choi BM et al. (2004) Carbon monoxide produced by heme oxygenase-1 suppresses T cell proliferation via inhibition of IL-2 production. J Immunol 172:4744–51 Passi S, Grandinetti M, Maggio F et al. (1998) Epidermal oxidative stress in vitiligo. Pigment Cell Res 11:81–5

www.jidonline.org

9


Paupe V, Dassa EP, Goncalves S et al. (2009) Impaired nuclear Nrf2 translocation undermines the oxidative stress response in Friedreich ataxia. PLoS One 4:e4253

Shunsuke T, Makoto F, Hou DX (2007) Action of Nrf2 and Keap1 in AREmediated NQO1 expression by quercetin. Free Radic Biol Med 42: 1690–703

Picardo M, Passi S, Morrone A et al. (1994) Antioxidant status in the blood of patients with active vitiligo. Pigment Cell Res 7:110–5

Spritz RA (2008) The genetics of generalized vitiligo. Curr Dir Autoimmun 10:244–57

Puri N, Mojamdar M, Ramaiah A (1987) In vitro growth characteristics of melanocytes obtained from adult normal and vitiligo subjects. J Invest Dermatol 88:434–8

Surh YJ (2003) Cancer chemoprevention with dietary phytochemicals. Nat Rev Cancer 3:768–80

Rubin LA, Nelson DL (1990) The soluble interleukin-2 receptor: biology, function, and clinical application. Ann Intern Med 113:619–27 Schallreuter KU, Bahadoran P, Picardo M et al. (2008) Vitiligo pathogenesis: autoimmune disease, genetic defect, excessive reactive oxygen species, calcium imbalance, or what else? Exp Dermatol 17:139–40, 141–60 Schallreuter KU, Gibbons NC, Zothner C et al. (2007) Hydrogen peroxidemediated oxidative stress disrupts calcium binding on calmodulin: more evidence for oxidative stress in vitiligo. Biochem Biophys Res Commun 360:70–5 Schallreuter KU, Moore J, Wood JM et al. (1999) In vivo and in vitro evidence for hydrogen peroxide (H2O2) accumulation in the epidermis of patients with vitiligo and its successful removal by a UVB-activated pseudocatalase. J Investig Dermatol Symp Proc 4:91–6 Schallreuter KU, Wood JM, Berger J (1991) Low catalase levels in the epidermis of patients with vitiligo. J Invest Dermatol 97:1081–5 Shajil EM, Begum R (2006) Antioxidant status of segmental and non-segmental vitiligo. Pigment Cell Res 19:179–80 Shalbaf M, Gibbons NC, Wood JM et al. (2008) Presence of epidermal allantoin further supports oxidative stress in vitiligo. Exp Dermatol 17:761–70

10


Taieb A (2000) Intrinsic and extrinsic pathomechanisms in vitiligo. Pigment Cell Res 13(Suppl 8):41–7 Tanaka A, Hamada N, Fujita Y et al. (2010) A novel kavalactone derivative protects against H2O2-induced PC12 cell death via Nrf2/ARE activation. Bioorg Med Chem 18:3133–9 Tusi SK, Ansari N, Amini M et al. (2010) Attenuation of NF-kappaB and activation of Nrf2 signaling by 1,2,4-triazine derivatives, protects neuronlike PC12 cells against apoptosis. Apoptosis 15:738–51 van den Wijngaard RM, Asghar SS, Pijnenborg AC et al. (2002) Aberrant expression of complement regulatory proteins, membrane cofactor protein and decay accelerating factor, in the involved epidermis of patients with vitiligo. Br J Dermatol 146:80–7 Yildirim M, Baysal V, Inaloz HS et al. (2004) The role of oxidants and antioxidants in generalized vitiligo at tissue level. J Eur Acad Dermatol Venereol 18:683–6 Yildirim M, Baysal V, Inaloz HS et al. (2003) The role of oxidants and antioxidants in generalized vitiligo. J Dermatol 30:104–8 Zhu H, Itoh K, Yamamoto M et al. (2005) Role of Nrf2 signaling in regulation of antioxidants and phase 2 enzymes in cardiac fibroblasts: protection against reactive oxygen and nitrogen species-induced cell injury. FEBS Lett 579:3029–36