Advertisement

Vitiligo Skin: Exploring the Dermal Compartment

  • Author Footnotes
    4 These authors contributed equally to this work.
    Daniela Kovacs
    Footnotes
    4 These authors contributed equally to this work.
    Affiliations
    Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Author Footnotes
    4 These authors contributed equally to this work.
    Emanuela Bastonini
    Footnotes
    4 These authors contributed equally to this work.
    Affiliations
    Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Monica Ottaviani
    Affiliations
    Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Carlo Cota
    Affiliations
    Dermatopathological Laboratory, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Emilia Migliano
    Affiliations
    Department of Plastic and Reconstructive Surgery, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Maria Lucia Dell’Anna
    Affiliations
    Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Mauro Picardo
    Correspondence
    Correspondence: Mauro Picardo, Laboratory of Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute (IRCCS), Via Elio Chianesi 53, Rome 00144, Italy.
    Affiliations
    Cutaneous Physiopathology and Integrated Center of Metabolomics Research, San Gallicano Dermatologic Institute, IRCCS, Rome, Italy
    Search for articles by this author
  • Author Footnotes
    4 These authors contributed equally to this work.
Open ArchivePublished:October 09, 2017DOI:https://doi.org/10.1016/j.jid.2017.06.033
      There is an increasing interest in the apparently normal skin in vitiligo. Altered expression of the adhesion molecule E-cadherin and persistent deregulated intracellular redox status that promotes the acquisition of a stress-induced senescent phenotype in melanocytes of normally pigmented skin from patients with vitiligo have been described. Growing evidence has shown that such cellular and functional alterations are not necessarily restricted to melanocytes but may be extended to other cutaneous cell populations in both lesional and nonlesional areas. However, whether dermal fibroblasts exhibit related alterations that may contribute to the defects associated with melanocytes in vitiligo is not known. Here we reveal within the dermal compartment cells a myofibroblast phenotype and a predisposition to premature senescence, indicating the existence of altered cross-talk between dermal and epidermal components that may affect melanocyte functionality even in the apparently normal skin of patients with vitiligo.

      Abbreviations:

      α-SMA (alpha-smooth muscle actin), DKK1 (Dickkopf1), ECM (extracellular matrix), ET-1 (endothelin-1), HGF (hepatocyte growth factor), NHF (normal human fibroblast), qRT-PCR (quantitative real time reverse transcriptase-PCR), ROS (reactive oxygen species), SCF (stem cell factor)

      Introduction

      Several mechanisms have been considered to be responsible for melanocyte loss in vitiligo, including genetic, inflammatory, autoimmune, oxidative, and metabolic alterations, but the individual contribution of each of these alterations is still unclear. A deregulated redox state associated with increased cellular vulnerability to oxidative insults due to intrinsic metabolic abnormalities is a major factor able to trigger immune responses leading to melanocyte degeneration and disappearance (
      • Picardo M.
      • Dell'Anna M.L.
      • Ezzedine K.
      • Hamzavi I.
      • Harris J.E.
      • Parsad D.
      • et al.
      Vitiligo.
      ). Alterations in the distribution of the cell-cell adhesion molecule E-cadherin, as well as a deregulated intracellular redox status that, when persistent, promotes the acquisition of a stress-induced premature senescent-like phenotype, have been highlighted in melanocytes of nonlesional skin (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ,
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ). Increasing evidence suggests that the presence of cellular and functional abnormalities extends to other cutaneous cell populations in both lesional and nonlesional skin. Immunohistochemical analysis on biopsies showed the overexpression of senescence-associated markers distributed to the entire nonlesional epidermis (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ), and redox imbalance and modifications in the expression of proliferation and senescence markers have been highlighted in cultured lesional keratinocytes (
      • Bondanza S.
      • Maurelli R.
      • Paterna P.
      • Migliore E.
      • Giacomo F.D.
      • Primavera G.
      • et al.
      Keratinocyte cultures from involved skin in vitiligo patients show an impaired in vitro behaviour.
      ,
      • Kostyuk V.A.
      • Potapovich A.I.
      • Cesareo E.
      • Brescia S.
      • Guerra L.
      • Valacchi G.
      • et al.
      Dysfunction of glutathione S-transferase leads to excess 4-hydroxy-2-nonenal and H(2)O(2) and impaired cytokine pattern in cultured keratinocytes and blood of vitiligo patients.
      ). Dermal components, via extracellular matrix (ECM) proteins and fibroblasts, exert an important role in the regulation of melanocyte homeostasis. ECM proteins contribute to melanocyte adhesion to the basement membrane. Fibroblasts release growth factors and messengers that are part of a paracrine signaling network that controls melanocyte function. Most of these mediators such as hepatocyte growth factor (HGF), stem cell factor (SCF), keratinocyte growth factor, and neuregulin-1 act as promelanogenic factors favoring melanocyte growth, differentiation, migration, and survival. Others, such as Dickkopf1 (DKK1), function as negative regulators of pigmentation and melanocyte growth (
      • Bastonini E.
      • Kovacs D.
      • Picardo M.
      Skin pigmentation and pigmentary disorders: focus on epidermal/dermal cross-talk.
      ). An altered expression of growth factors controlling melanocyte homeostasis has been observed in several pigmentary disorders including vitiligo (
      • Kitamura R.
      • Tsukamoto K.
      • Harada K.
      • Shimizu A.
      • Shimada S.
      • Kobayashi T.
      • et al.
      Mechanisms underlying the dysfunction of melanocytes in vitiligo epidermis: role of SCF/KIT protein interactions and the downstream effector, MITF M.
      ,
      • Kovacs D.
      • Cardinali G.
      • Aspite N.
      • Cota C.
      • Luzi F.
      • Bellei B.
      • et al.
      Role of fibroblast-derived growth factors in regulating hyperpigmentation of solar lentigo.
      ,
      • Lee A.Y.
      • Kim N.H.
      • Choi W.I.
      • Youm Y.H.
      Less keratinocyte-derived factors related to more keratinocyte apoptosis in depigmented than normally pigmented suction-blistered epidermis may cause passive melanocyte death in vitiligo.
      ,
      • Moretti S.
      • Spallanzani A.
      • Amato L.
      • Hautmann G.
      • Gallerani I.
      • Fabiani M.
      • et al.
      New insights into the pathogenesis of vitiligo: imbalance of epidermal cytokines at sites of lesions.
      ). However, only few reports point to a specific involvement of the dermis in vitiligo, and most of the data are focused on lesional skin. Upregulation of the antiadhesive ECM protein tenascin and DKK1 has been demonstrated in lesional dermis (
      • Le Poole I.C.
      • van den Wijngaard R.M.
      • Westerhof W.
      • Das P.K.
      Tenascin is overexpressed in vitiligo lesional skin and inhibits melanocyte adhesion.
      ,
      • Oh S.H.
      • Kim J.Y.
      • Kim M.R.
      • Do J.E.
      • Shin J.Y.
      • Hann S.K.
      DKK1 is highly expressed in the dermis of vitiligo lesion: is there association between DKK1 and vitiligo?.
      ). By contrast, the expression and release of keratinocyte growth factor by lesional fibroblasts is reduced (
      • Purpura V.
      • Persechino F.
      • Belleudi F.
      • Scrofani C.
      • Raffa S.
      • Persechino S.
      • et al.
      Decreased expression of KGF/FGF7 and its receptor in pathological hypopigmentation.
      ). Because of the decisive role of the dermal compartment in regulating melanocyte activity and survival, we aimed to investigate the involvement of the dermis in vitiligo, focusing on the analysis of the features of fibroblasts to: (i) investigate whether the degeneration/senescence prone phenotype detected in vitiligo epidermis also involves the dermis of nonlesional skin and (ii) explore possible altered biological activities of fibroblasts, which may contribute to melanocyte alterations.

      Results

      Nonlesional vitiligo fibroblasts display increased basal ROS levels associated with the upregulation of the stress-induced marker p53

      As most vitiligo fibroblasts appeared flattened and enlarged, features resembling a senescent phenotype (
      • Kovacs D.
      • Cardinali G.
      • Aspite N.
      • Cota C.
      • Luzi F.
      • Bellei B.
      • et al.
      Role of fibroblast-derived growth factors in regulating hyperpigmentation of solar lentigo.
      ), we measured cell areas and actin cytoskeleton organization, demonstrating a significant enlargement in the size with respect to control (Figure 1a and b). Vitiligo-associated fibroblasts also exhibited an increase in actin stress fibers detected using phalloidin (Figure 1a). A significant increase in reactive oxygen species (ROS) content was also observed in vitiligo fibroblasts (Figure 1c), similar to that reported in melanocytes (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ). Consequently, we analyzed the expression of the stress-induced cell cycle regulator p53, which is upregulated in vitiligo melanocytes and keratinocytes (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ,
      • Salem M.M.
      • Shalbaf M.
      • Gibbons N.C.
      • Chavan B.
      • Thornton J.M.
      • Schallreuter K.U.
      Enhanced DNA binding capacity on up-regulated epidermal wild-type p53 in vitiligo by H2O2-mediated oxidation: a possible repair mechanism for DNA damage.
      ). Vitiligo fibroblasts showed a marked increase in the levels of the p53 protein in most cultures by western blot analysis (Figure 1d). Immunofluorescence analysis confirmed a higher number of cells displaying p53-positive nuclei in vitiligo with respect to control, as assessed by the nuclear counterstaining with DAPI and the double staining with the nuclear membrane marker lamin B1 (Figure 1e and f). As a positive control of oxidative stress, control fibroblasts were exposed to H2O2. An increased number of p53-stained cells were detected (Figure 1g), which appeared comparable to that observed in vitiligo fibroblasts (Figure 1h). The p53-responsive stress gene GADD45 evaluated by quantitative real time reverse transcriptase-PCR (qRT-PCR) was also significantly induced in vitiligo (Figure 1i).
      Figure 1
      Figure 1Vitiligo fibroblasts display increased ROS levels associated with upmodulation of stress-induced markers. (a) Phase-contrast and TRITC-phalloidin staining. (b) Cell area measurement (mean value ± SD, μm2). (c) ROS detection by FACS analysis. (d) Western blot of p53 expression and corresponding densitometric analysis. (e) Immunofluorescence of p53 expression. (f) Magnified image of p53 staining in vitiligo fibroblasts and double immunofluorescence with anti-p53 and anti-lamin B1 antibodies. (g) Immunofluorescence of p53 in H2O2-treated control fibroblasts. (h) Percentage ± SD of p53-positive cells on control, vitiligo, and control fibroblasts treated with H2O2. (i) mRNA transcripts of GADD45. Nuclei are stained with DAPI. The arrows point at positive cells. The cell shape is outlined. Scale bars: a, e, and g: 50 μm; f: 20 μm. DCFH-DA, 2’,7’-dichlorofluorescein diacetate; NHF, normal human fibroblast; ROS, reactive oxygen species; SD, standard deviation; TRITC, tetramethylrhodamine; VHF, vitiligo fibroblast.

      Nonlesional vitiligo fibroblasts exhibit a myofibroblast-like phenotype, which is correlated with the increased intrinsic oxidative stress

      As high levels of intracellular ROS favor the conversion of fibroblasts into transdifferentiated myofibroblasts (
      • Cat B.
      • Stuhlmann D.
      • Steinbrenner H.
      • Alili L.
      • Holtkötter O.
      • Sies H.
      • et al.
      Enhancement of tumor invasion depends on transdifferentiation of skin fibroblasts mediated by reactive oxygen species.
      ), we analyzed the expression of the alpha-smooth muscle actin (α-SMA) marker in our cell cultures. A significantly higher α-SMA mRNA level was detected in vitiligo fibroblasts compared with control as demonstrated by qRT-PCR (Figure 2a), which was reflected in a clear increase at the protein level detected by western blot (Figure 2b). Immunofluorescence analysis revealed the staining pattern for α-SMA as positive stress fibers distributed throughout the cytoplasm, whereas only a weak cytosolic reactivity was evident in a few normal fibroblasts (Figure 2c and d). The treatment of control fibroblasts with H2O2 was able to induce an increase in the number of α-SMA-positive cells comparable to that measured in vitiligo (Figure 2d and e). The mRNA transcript of the splice variant of extra domain A-fibronectin, typically expressed by myofibroblasts (
      • Serini G.
      • Bochaton-Piallat M.L.
      • Ropraz P.
      • Geinoz A.
      • Borsi L.
      • Zardi L.
      • et al.
      The fibronectin domain ED-A is crucial for myofibroblastic phenotype induction by transforming growth factor-β1.
      ), also showed a clear increase in vitiligo (Figure 2f). α-SMA has been reported to be a direct transcriptional target of p53 (
      • Comer K.A.
      • Dennis P.A.
      • Armstrong L.
      • Catino J.J.
      • Kastan M.B.
      • Kumar C.C.
      Human smooth muscle alpha-actin gene is a transcriptional target of the p53 tumor suppressor protein.
      ). Consistent with this, we found a positive correlation between the percentage of stained α-SMA and those positive for p53 (r = +0.88).
      Figure 2
      Figure 2Nonlesional vitiligo fibroblasts show a myofibroblast phenotype. (a) α-SMA mRNA level evaluated by qRT-PCR. (b) Western blot analysis of α-SMA expression and corresponding densitometric analysis on cell lysates from normal and vitiligo fibroblasts. (c, d) Immunofluorescence analysis of α-SMA expression in control, vitiligo, and H2O2-treated control fibroblasts. (e) Quantitative analysis of the percentage ± standard deviation of α-SMA-positive cells on control, vitiligo, and control fibroblasts treated with H2O2.. (f) mRNA transcript levels of EDA-fibronectin evaluated qRT-PCR. Nuclei are stained with DAPI. Scale bars: c and d: 50 μm. α-SMA, alpha-smooth muscle actin; EDA-fibronectin, extra domain A-fibronectin; NHF, normal human fibroblast; qRT-PCR, quantitative real time reverse transcriptase-PCR; VHF, vitiligo fibroblast.
      On the basis of these results, we deepened the analysis of the features of fibroblasts by evaluating the expression of growth factors and cytokines, as well as collagen isoforms and ECM molecules associated with the myofibroblast phenotype (
      • Desmoulière A.
      • Guyot C.
      • Gabbiani G.
      The stroma reaction myofibroblast: a key player in the control of tumor cell behavior.
      ,
      • Powell D.W.
      • Mifflin R.C.
      • Valentich J.D.
      • Crowe S.E.
      • Saada J.I.
      • West A.B.
      Myofibroblasts. I: Paracrine cells important in health and disease.
      ). We analyzed the inflammatory cytokines IL-1β, IL-6, and the growth factors HGF and SCF, for their role in controlling melanocyte homeostasis and for their known upmodulation in stress-induced senescent cells (
      • Kovacs D.
      • Cardinali G.
      • Aspite N.
      • Cota C.
      • Luzi F.
      • Bellei B.
      • et al.
      Role of fibroblast-derived growth factors in regulating hyperpigmentation of solar lentigo.
      ,
      • Miyazaki M.
      • Gohda E.
      • Kaji K.
      • Namba M.
      Increased hepatocyte growth factor production by aging human fibroblasts mainly due to autocrine stimulation by interleukin-1.
      ,
      • Waldera Lupa D.M.
      • Kalfalah F.
      • Safferling K.
      • Boukamp P.
      • Poschmann G.
      • Volpi E.
      • et al.
      Characterization of skin aging-associated secreted proteins (SAASP) produced by dermal fibroblasts isolated from intrinsically aged human skin.
      ). A significant induction of expression of the IL-1β, IL-6, and HGF genes was observed in vitiligo fibroblasts with respect to normal control cells as demonstrated by qRT-PCR, whereas SCF transcripts showed no difference (Figure 3a). An enhanced release of IL-6 and HGF was also observed at the protein level by ELISA on culture supernatants (Figure 3b and c), whereas the secretion of SCF showed a tendency to decrease, although not significant (Figure 3d). Moreover, IL-6 and HGF levels correlated with the expression of α-SMA (r = +0.85 and +0.82 for IL-6 and HGF, respectively). The ECM glycoprotein fibronectin and the intermediate filament-associated protein vimentin, both associated with myofibroblast differentiation and known to be induced by ROS and in senescent cells, were also upmodulated in vitiligo (Figure 3e–g), as well as the fibroblast-derived factor DKK1 (Figure 3h). Among the different collagens produced by myofibroblasts (
      • Powell D.W.
      • Mifflin R.C.
      • Valentich J.D.
      • Crowe S.E.
      • Saada J.I.
      • West A.B.
      Myofibroblasts. I: Paracrine cells important in health and disease.
      ), the expression of collagen IV, which represents the major molecule of the basement membrane and contributes to the maintenance of the correct location of melanocytes, was significantly upregulated in vitiligo, as assessed by qRT-PCR (Figure 3i). Because dermal fibroblasts subjected to stress-induced premature senescence display an increased amount of cholesterol and its oxidative products oxysterols (
      • Briganti S.
      • Flori E.
      • Bellei B.
      • Picardo M.
      Modulation of PPARγ provides new insights in a stress induced premature senescence model.
      ), we evaluated the levels of cholesterol and of the cholesterol oxidation derivatives 7-beta-hydroxycholesterol and 7-ketocholesterol. We found a higher amount of cholesterol (128.36 ± 36; P < 0.05) and of the two oxysterols in vitiligo compared with normal fibroblasts (230 ± 97% and 148 ± 33%, respectively; P < 0.05). Finally, we analyzed the expression of the transforming growth factor-β and of endothelin-1 (ET-1), as they represent important mediators in promoting myofibroblast induction (
      • Desmoulière A.
      • Geinoz A.
      • Gabbiani F.
      • Gabbiani G.
      Transforming growth factor-beta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts.
      ,
      • Shi-Wen X.
      • Chen Y.
      • Denton C.P.
      • Eastwood M.
      • Renzoni E.A.
      • Bou-Gharios G.
      • et al.
      Endothelin-1 promotes myofibroblast induction through the ETA receptor via a rac/phosphoinositide 3-kinase/Akt-dependent pathway and is essential for the enhanced contractile phenotype of fibrotic fibroblasts.
      ). The results showed no significant differences either at the mRNA (Supplementary Figure S1a online) or protein levels (Supplementary Figure S1b) for transforming growth factor-β between vitiligo and control fibroblasts. The overall amount of ET-1 released in vitiligo culture supernatant tended to be higher in comparison to control (Supplementary Figure S1c). However, the difference was not statistically significant.
      Figure 3
      Figure 3Vitiligo fibroblasts show increased expression of markers related to the myofibroblast phenotype. (a) mRNA transcripts of IL-1β, IL-6, HGF, and SCF evaluated by qRT-PCR in control and vitiligo fibroblasts. (b) IL-6, (c) HGF, (d) SCF quantitation by ELISA. (e) mRNA transcript levels of fibronectin evaluated by qRT-PCR in normal and vitiligo fibroblasts. Western blot analysis and corresponding densitometric analysis of (f) fibronectin and (g) vimentin expression in normal and vitiligo fibroblasts. mRNA transcript levels of (h) DKK1 and (i) collagen IV evaluated by qRT-PCR in normal and vitiligo fibroblasts. DKK1, Dickkopf1; HGF, hepatocyte growth factor; NHF, normal human fibroblast; qRT-PCR, quantitative real time reverse transcriptase-PCR; SCF, stem cell factor; VHF, vitiligo fibroblast.
      To evaluate whether the myofibroblast phenotype could be ascribed to the augmented intracellular ROS content, the expression of α-SMA was analyzed after the treatment with the ROS scavenger N-acetyl-l-cysteine. Western blot and immunofluorescence analyses demonstrated that α-SMA protein levels and the number of positive cells were significantly reduced in the presence of N-acetyl-l-cysteine in a dose-dependent manner (Supplementary Figure S2a–c online). The reduction is preceded and accompanied by the decrease of the intracellular ROS production (P < 0.05) (Supplementary Figure S2d).

      Conditioned medium from vitiligo fibroblasts downregulates E-cadherin expression on melanocytes

      Dermal end epidermal constituents release growth factors and cytokines, which regulate melanocyte functionality both in an autocrine and a paracrine manner (
      • Bastonini E.
      • Kovacs D.
      • Picardo M.
      Skin pigmentation and pigmentary disorders: focus on epidermal/dermal cross-talk.
      ). We therefore investigated the possible influences of vitiligo fibroblasts on primary melanocytes by treating them with conditioned medium obtained from normal and vitiligo fibroblasts. Emerging data point on an altered expression of the adhesion molecule E-cadherin in vitiligo melanocytes, which occurs even earlier than the appearance of clinical lesions (
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ). We first evaluated the basal expression level of the protein on a panel of normal and vitiligo melanocytes collected from nonlesional areas. Most vitiligo melanocytes demonstrated a significant reduction of E-cadherin expression with respect to control by western blot and immunofluorescence analyses (Figure 4a and b), confirming in vitro the altered expression pattern demonstrated ex vivo (
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ). We then treated melanocytes with conditioned medium collected from control and vitiligo fibroblasts with the aim of simulating in vitro the dermal-epidermal cross-talk existing in the cutaneous microenvironment. The treatment with normal fibroblast-conditioned medium did not significantly modify the expression of E-cadherin in normal melanocytes, whereas it induced a decrease in the E-cadherin level in vitiligo melanocytes, as assessed by western blot and immunofluorescence (Figure 4c and d). Vitiligo fibroblast-conditioned medium induced a reduction in E-cadherin expression in both vitiligo and normal melanocytes that was more pronounced in vitiligo cells, as assessed by densitometric analysis (Figure 4c). Although most untreated melanocytes displayed a regular and constant membranous distribution of E-cadherin, cells treated with conditioned medium from vitiligo fibroblasts showed a reduction associated with a discontinuous expression pattern of the adhesion molecule (Supplementary Figure S3 online). We next evaluated whether the reduction of E-cadherin expression observed after the treatment with vitiligo fibroblast-conditioned medium could be ascribed to the higher production of HGF observed in vitiligo. To this aim, we first treated normal melanocytes with recombinant HGF and, as expected (
      • Li G.
      • Schaider H.
      • Satyamoorthy K.
      • Hanakawa Y.
      • Hashimoto K.
      • Herlyn M.
      Downregulation of E-cadherin and Desmoglein 1 by autocrine hepatocyte growth factor during melanoma development.
      ,
      • Soong J.
      • Chen Y.
      • Shustef M.E.
      • Scott G.
      Sema4D, the ligand for plexin B1, suppresses c-Met activation and migration and promotes melanocyte survival and growth.
      ), the results showed that the growth factor induced a reduction in E-cadherin expression (Supplementary Figure S4a online). The specific involvement of HGF in modulating this adhesion molecule was ascertained by neutralizing the growth factor activity with an anti-HGF antibody, which led to the inhibition of E-cadherin downregulation (Supplementary Figure S4b). The pretreatment of vitiligo fibroblast-conditioned medium with the neutralizing antibody to HGF resulted in a partial reduction of E-cadherin downmodulation in response to vitiligo culture medium, confirming the relevant role of HGF in mediating such effect (Supplementary Figure S4b). To analyze whether the reduction of E-cadherin in response to the treatment with vitiligo fibroblast-conditioned medium could be associated with altered expression of other cadherin adhesion molecules, we analyzed the expression of N-cadherin. Immunofluoresence analysis showed that normal melanocytes displayed basal low levels of the adhesion molecule, which were not modified by the treatment with conditioned medium either from control or vitiligo fibroblasts (Supplementary Figure S4c and d).
      Figure 4
      Figure 4Influences of nonlesional vitiligo fibroblasts on E-cadherin expression in melanocytes. (a) Western blot analysis and corresponding densitometric analysis of E-cadherin expression in normal and nonlesional vitiligo melanocytes. (b) Immunofluorescence confocal microscopy of E-cadherin expression in normal and vitiligo melanocytes. (c) Western blot analysis and corresponding densitometric analysis of E-cadherin expression in normal and vitiligo melanocytes treated for 48 hours with normal and vitiligo fibroblasts-conditioned medium. (d) Immunofluorescence confocal microscopy of E-cadherin expression in normal and vitiligo melanocytes treated for 48 hours with conditioned medium collected from normal and vitiligo fibroblasts. Scale bars: b and d: 20 μm. NHF, normal human fibroblast; NHM, normal human melanocyte; VHF, vitiligo fibroblast; VHM, vitiligo melanocyte.

      Modifications in the dermal compartment are detectable ex vivo

      To validate the in vitro results, we performed the immunohistochemical evaluation of some markers detected in vitro on skin biopsies collected from the same patients from whom the cells were derived. Vitiligo specimens were first analyzed for the number and distribution of melanocytes. Positive cells for the melanocyte markers MITF and MART1 were detected in nonlesional vitiligo skin with no significant differences in the overall number of cells in comparison to control. However, some vitiligo melanocytes appeared both to extend into the dermis and to be localized suprabasally. The melanocytes detected in the dermis were higher in number but not significantly different with respect to control, whereas the melanocytes found in the suprabasal layers were significantly more abundant than those in control skin (P < 0.05) (Figure 5a and b). To evaluate whether the melanocytes located suprabasally could show alterations in the E-cadherin expression, double immunofluorescence with anti-c-kit and anti-E-cadherin antibodies was performed. Immunostaining revealed the presence of suprabasal c-kit positive melanocytes characterized by a discontinued and low expression of E-cadherin (Supplementary Figure S5a online). Parallel immunohistochemical analysis of the expression of N-cadherin did not show melanocyte positive reactivity in any vitiligo sample (Supplementary Figure S5b). Immunohistochemical analysis of p53 revealed the presence of positive cells in the dermis from nonlesional areas and, although a variable degree of reactivity was observed among the samples, the number of labeled cells was higher in vitiligo with respect to control (Figure 5c). The evaluation of α-SMA expression revealed, as expected, positive staining of smooth muscle cells in vessel walls in both normal and vitiligo skin. Some positive cells were detected in the dermis of vitiligo, whereas no immunoreactivity was observed in control dermis (Figure 5d). Immunohistochemical analysis for the expression of IL-6 showed an increased immunoreactivity for the cytokine in vitiligo in comparison to control skin (Figure 5e). Also, immunofluorescence analysis of fibronectin expression showed a higher reactivity in vitiligo dermis compared with control (Figure 5f).
      Figure 5
      Figure 5Ex vivo assessment of modifications in the dermal compartment of nonlesional vitiligo skin. (a) Immunohistochemical expression of MART1 and MITF in control and nonlesional vitiligo skin. (b) Mean value ± standard deviation of the number of positive cells/mm of basal membrane counted in control and vitiligo samples. Immunohistochemical staining for (c) p53, (d) α-SMA, and (e) IL-6 in control skin and nonlesional vitiligo skin. The arrows point at positive cells. (f) Immunofluorescence staining for fibronectin. Nuclei are stained with DAPI. The basal membrane is outlined as a white dashed line. The boxed areas represent the enlarged view of the selected fields. Scale bars: a, b, c, and f: 50 μm; (d): 100 μm; enlarged view of the boxed area: 20 μm. α-SMA, alpha-smooth muscle actin.

      Discussion

      Despite the important role exerted by fibroblasts and ECM proteins in regulating melanocyte functionality, few studies examine the involvement of the dermal constituents in vitiligo. Our data show that changes linked to redox imbalance previously detected in vitiligo melanocytes (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ) also extend to the dermal fibroblasts of normally pigmented areas. The generation of oxidative stress detected in vitiligo fibroblasts was correlated with morphological and functional modifications. We found a higher number of cells with biological characteristics resembling a myofibroblast phenotype. Myofibroblasts may arise from different precursors. Besides their origin from adjacent local tissue fibroblasts, they may derive from other resident cells including smooth muscle, endothelial and epithelial cells, tissue-derived mesenchymal stem/stromal cells, as well as bone-marrow-derived cells such as fibrocytes (
      • Kendall R.T.
      • Feghali-Bostwick C.A.
      Fibroblasts in fibrosis: novel roles and mediators.
      ,
      • Leask A.
      Potential therapeutic targets for cardiac fibrosis: TGFbeta, angiotensin, endothelin, CCN2, and PDGF, partners in fibroblast activation.
      ). As the origin of myofibroblasts may be heterogeneous, even the factors controlling their induction and persistence are under the control of a complex interplay of promoting and suppressing mediators (
      • Hinz B.
      Myofibroblasts.
      ). Although the specific contribution of the different cell sources in the induction of myofibroblasts in vitiligo remains to be characterized, the presence of myofibroblasts may be the consequence of a deregulation in the interplay between mediators and messengers released in vivo in the entire skin. The elevated basal ROS levels detected in vitiligo may represent a pivotal player because it has been shown that ROS represent positive mediators in the induction of myofibroblasts. Mitochondria have been proposed as a possible site of the increased production of ROS in vitiligo, and an altered expression and activity of complex I has been demonstrated both in peripheral blood mononuclear vitiligo cells and in melanocytes from nonlesional skin (
      • Dell'Anna M.L.
      • Urbanelli S.
      • Mastrofrancesco A.
      • Camera E.
      • Iacovelli P.
      • Leone G.
      • et al.
      Alterations of mitochondria in peripheral blood mononuclear cells of vitiligo patients.
      ,
      • Dell'Anna M.L.
      • Ottaviani M.
      • Albanesi V.
      • Vidolin A.P.
      • Leone G.
      • Ferraro C.
      • et al.
      Membrane lipid alterations as a possible basis for melanocyte degeneration in vitiligo.
      ). Impairment of mitochondrial function leading to high basal intracellular ROS production may also be present as an intrinsic defect in dermal fibroblasts and encourage, in turn, myofibroblast differentiation. In fact, it has been observed that fibroblasts carrying dysfunctional mitochondrial complex I produce high levels of ROS that correlate with elevated α-SMA expression and consequently with their transdifferentiation into myofibroblasts (
      • Taddei M.L.
      • Giannoni E.
      • Raugei G.
      • Scacco S.
      • Sardanelli A.M.
      • Papa S.
      • et al.
      Mitochondrial oxidative stress due to complex I dysfunction promotes fibroblast activation and melanoma cell invasiveness.
      ). Dysfunction in the mitochondrial complex I associated with increased ROS levels also drives irradiation-induced myofibroblast differentiation of lung fibroblasts (
      • Yang X.
      • Liu T.
      • Chen B.
      • Wang F.
      • Yang Q.
      • Chen X.
      Grape seed proanthocyanidins prevent irradiation-induced differentiation of human lung fibroblasts by ameliorating mitochondrial dysfunction.
      ). Accordingly, strong production of ROS has been found in nonlesional vitiligo skin biopsies in comparison to control (
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ). At the same time, we cannot exclude a combined effect guided by local inflammation, as the presence of positive CD3 in the dermis has been detected in clinically normal pigmented skin areas (
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ).
      The upregulation of p53 observed in our cell cultures was also previously demonstrated in vitiligo melanocytes and keratinocytes (
      • Bellei B.
      • Pitisci A.
      • Ottaviani M.
      • Ludovici M.
      • Cota C.
      • Luzi F.
      • et al.
      Vitiligo: a possible model of degenerative diseases.
      ,
      • Salem M.M.
      • Shalbaf M.
      • Gibbons N.C.
      • Chavan B.
      • Thornton J.M.
      • Schallreuter K.U.
      Enhanced DNA binding capacity on up-regulated epidermal wild-type p53 in vitiligo by H2O2-mediated oxidation: a possible repair mechanism for DNA damage.
      ), and could be an additional element involved in the induction of the myofibroblast phenotype, because α-SMA is a direct transcriptional target of p53 (
      • Comer K.A.
      • Dennis P.A.
      • Armstrong L.
      • Catino J.J.
      • Kastan M.B.
      • Kumar C.C.
      Human smooth muscle alpha-actin gene is a transcriptional target of the p53 tumor suppressor protein.
      ). Accordingly, the treatment of control fibroblasts with H2O2 resulted in an increased expression of both p53 and α-SMA proteins, supporting the involvement of oxidative stress in mediating such an effect.
      Consistent with the idea that the acquisition of stress-related modifications could be extended to the entire skin, we observed that vitiligo fibroblasts display an increased content of cholesterol and oxysterols, which are associated with the occurrence of a senescence-like phenotype in a stress-induced premature senescence model caused by the exposure of fibroblasts to 8-meyhoxypsoralen plus UVA irradiation (
      • Briganti S.
      • Flori E.
      • Bellei B.
      • Picardo M.
      Modulation of PPARγ provides new insights in a stress induced premature senescence model.
      ). Oxysterols may be involved in the induction of the myofibroblast phenotype by contributing to the unbalanced redox status in the dermis and favoring over time the induction of premature senescence. Interestingly, an increase in the number of α-SMA expressing cells has been observed in senescent fibroblasts with respect to young cells (
      • Cáceres M.
      • Oyarzun A.
      • Smith P.C.
      Defective wound-healing in aging gingival tissue.
      ,
      • Yanai H.
      • Shteinberg A.
      • Porat Z.
      • Budovsky A.
      • Braiman A.
      • Ziesche R.
      • et al.
      Cellular senescence-like features of lung fibroblasts derived from idiopathic pulmonary fibrosis patients.
      ). Vitiligo cells also display an enlarged shape associated with a higher expression of fibronectin, vimentin, and stress fibers, all features exhibited by senescent fibroblasts (
      • Kim H.J.
      • Kim K.S.
      • Kim S.H.
      • Baek S.H.
      • Kim H.Y.
      • Lee C.
      • et al.
      Induction of cellular senescence by secretory phospholipase A2 in human dermal fibroblasts through an ROS-mediated p53 pathway.
      ,
      • Nishio K.
      • Inoue A.
      • Qiao S.
      • Kondo H.
      • Mimura A.
      Senescence and cytoskeleton: overproduction of vimentin induces senescent-like morphology in human fibroblasts.
      ). As previously demonstrated in melanocytes, vitiligo fibroblasts showed an increased synthesis and production of IL-6, a contributing factor to aging-associated secreted proteins released by prolonged damaged and senescent cells in skin (
      • Waldera Lupa D.M.
      • Kalfalah F.
      • Safferling K.
      • Boukamp P.
      • Poschmann G.
      • Volpi E.
      • et al.
      Characterization of skin aging-associated secreted proteins (SAASP) produced by dermal fibroblasts isolated from intrinsically aged human skin.
      ), further supporting the presence of a stress-mediated senescence phenotype extended to the entire skin. Interestingly, some features observed in vitiligo fibroblasts resemble the phenotype described in fibroblasts of scleroderma/systemic sclerosis (
      • Hinz B.
      • Phan S.H.
      • Thannickal V.J.
      • Prunotto M.
      • Desmouliere A.
      • Varga J.
      • et al.
      Recent developments in myofibroblast biology: paradigms for connective tissue remodeling.
      ). Skin fibroblasts from patients with systemic sclerosis generate excessive ROS and show increased expression of α-SMA (
      • Spadoni T.
      • Svegliati Baroni S.
      • Amico D.
      • Albani L.
      • Moroncini G.
      • Avvedimento E.V.
      • et al.
      A reactive oxygen species-mediated loop maintains increased expression of NADPH oxidases 2 and 4 in skin fibroblasts from patients with systemic sclerosis.
      ). Transforming growth factor-β represents a pivotal factor in the accumulation of myofibroblasts in scleroderma (
      • Lafyatis R.
      Transforming growth factor beta—at the centre of systemic sclerosis.
      ). In vitiligo skin, we did not find a significant increase in the expression of transforming growth factor-β, suggesting that myofibroblast transdifferentiation is not directly mediated by this growth factor, at least not via autocrine production. However, we cannot exclude the action of other factors, for example, IL-6 and ET-1, whose production is also increased in scleroderma fibroblasts (
      • Feghali C.A.
      • Bost K.L.
      • Boulware D.W.
      • Levy L.S.
      Mechanisms of pathogenesis in scleroderma. I: Overproduction of interleukin 6 by fibroblasts cultured from affected skin sites of patients with scleroderma.
      ,
      • Kawaguchi Y.
      • Suzuki K.
      • Hara M.
      • Hidaka T.
      • Ishizuka T.
      • Kawagoe M.
      • et al.
      Increased endothelin-1 production in fibroblasts derived from patients with systemic sclerosis.
      ,
      • Shi-Wen X.
      • Chen Y.
      • Denton C.P.
      • Eastwood M.
      • Renzoni E.A.
      • Bou-Gharios G.
      • et al.
      Endothelin-1 promotes myofibroblast induction through the ETA receptor via a rac/phosphoinositide 3-kinase/Akt-dependent pathway and is essential for the enhanced contractile phenotype of fibrotic fibroblasts.
      ). In addition, some reports have shown a similar gene expression profile in both affected and unaffected fibroblasts (
      • Fuzii H.T.
      • Yoshikawa G.T.
      • Junta C.M.
      • Sandrin-Garcia P.
      • Fachin A.L.
      • Sakamoto-Hojo E.T.
      • et al.
      Affected and non-affected skin fibroblasts from systemic sclerosis patients share a gene expression profile deviated from the one observed in healthy individuals.
      ,
      • Whitfield M.L.
      • Finlay D.R.
      • Murray J.I.
      • Troyanskaya O.G.
      • Chi J.T.
      • Pergamenschikov A.
      • et al.
      Systemic and cell type-specific gene expression patterns in scleroderma skin.
      ) and higher levels of ROS in both fibrotic and nonfibrotic skin of patients with scleroderma (
      • Bourji K.
      • Meyer A.
      • Chatelus E.
      • Pincemail J.
      • Pigatto E.
      • Defraigne J.O.
      • et al.
      High reactive oxygen species in fibrotic and nonfibrotic skin of patients with diffuse cutaneous systemic sclerosis.
      ). Despite the occurrence of these alterations, the absence of clinical manifestations related to fibrosis in uninvolved scleroderma skin suggests that other additional factors are necessary to develop a macroscopically detectable fibrotic phenotype. A similar phenomenon may occur in vitiligo skin, in which the myofibroblast phenotype does not appear to be adequate to develop a clinically manifested fibrosis. In fact, although only few data are available in the literature on the histological examination of vitiligo normal appearing skin, fibrosis does not appear to have a disclosed characteristic.
      Most of our vitiligo melanocytes showed a reduction of the E-cadherin level with respect to control cells, confirming in vitro the altered expression of E-cadherin demonstrated ex vivo (
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ). We observed E-cadherin downmodulation after treatment with conditioned-medium derived from fibroblasts, which appeared significant only for vitiligo melanocytes when treated either with normal or nonlesional vitiligo fibroblast-conditioned media. By contrast, in control melanocytes, E-cadherin expression was diminished only when treated with conditioned medium derived from vitiligo cells. In the latter condition, the distribution of E-cadherin appeared discontinuous and interrupted, features resembling the expression pattern described by
      • Wagner R.Y.
      • Luciani F.
      • Cario-André M.
      • Rubod A.
      • Petit V.
      • Benzekri L.
      • et al.
      Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
      ex vivo. Although these results indicate a higher intrinsic susceptibility of vitiligo melanocytes to external influences, it also appears that vitiligo fibroblasts release an elevated amount of biologically active messengers able to affect E-cadherin expression. Among them, both HGF and ET-1 downmodulate E-cadherin in melanocytes (
      • Haass N.K.
      • Herlyn M.
      Normal human melanocyte homeostasis as a paradigm for understanding melanoma.
      ). The increased secretion of HGF by vitiligo cells may therefore affect basal E-cadherin expression as detected in normal appearing skin. The partial abrogation of such a reduction after the treatment of vitiligo fibroblast-conditioned medium with an HGF neutralizing antibody confirms the involvement of HGF in modulating E-cadherin expression in vitiligo skin. However, additional and often synergistic effects can be caused by other factors released in the surrounding microenvironment, for example ET-1 itself and/or IL-6, which are both able to affect E-cadherin expression. Among the bioactive messengers involved in the intricate dermal-epidermal cross-talk, also the fibroblast-derived factor DKK1 reduces E-cadherin in HEK293 cells (
      • Kuang H.B.
      • Miao C.L.
      • Guo W.X.
      • Peng S.
      • Cao Y.J.
      • Duan E.K.
      Dickkopf-1 enhances migration of HEK293 cell by beta-catenin/E-cadherin degradation.
      ). Previous work showed a higher DKK1 level in lesional skin compared with nonlesional vitiligo biopsies (
      • Oh S.H.
      • Kim J.Y.
      • Kim M.R.
      • Do J.E.
      • Shin J.Y.
      • Hann S.K.
      DKK1 is highly expressed in the dermis of vitiligo lesion: is there association between DKK1 and vitiligo?.
      ). Our data demonstrate increased expression of DKK1 in fibroblasts cultures from nonlesional areas, indicating a deregulation of this mediator in the entire skin. DKK1 may act dually on melanocytes both by negatively regulating their proliferation and melanogenesis (
      • Yamaguchi Y.
      • Itami S.
      • Watabe H.
      • Yasumoto K.
      • Abdel-Malek Z.A.
      • Kubo T.
      • et al.
      Mesenchymal-epithelial interactions in the skin: increased expression of dickkopf1 by palmoplantar fibroblasts inhibits melanocyte growth and differentiation.
      ) and by reducing E-cadherin expression. As a consequence, melanocytes may be more likely to dissociate themselves from the neighboring keratinocytes, relocate suprabasally, and progressively disappear. In some conditions, such as melanoma progression, the downregulation of E-cadherin expression is often associated with the parallel induction of N-cadherin (
      • Hsu M.
      • Andl T.
      • Li G.
      • Meinkoth J.L.
      • Herlyn M.
      Cadherin repertoire determines partner-specific gap junctional communication during melanoma progression.
      ). The absence of positive N-cadherin immunoreactivity in vitiligo suggests that the dermal protrusion of melanocytes does not appear to be mediated by cadherin switching. Rather, since the presence of focal gaps in the basal membrane of vitiligo skin has been described (
      • Bose S.K.
      • Ortonne J.P.
      Focal gaps in the basement membrane of involved and uninvolved skin of vitiligo. Are they normal?.
      ,
      • Panuncio A.L.
      • Vignale R.
      Ultrastructural studies in stable vitiligo.
      ), melanocytes may drop into the dermis through these basement membrane abnormalities.
      The higher expression of HGF, as well as of IL-1β, detected in vitiligo fibroblasts further supports that these cells, similar to melanocytes, display a predisposition to a premature senescence phenotype. In fact, the production of HGF in dermal fibroblasts increases with aging, mainly due to autocrine stimulation by IL-1 (
      • Miyazaki M.
      • Gohda E.
      • Kaji K.
      • Namba M.
      Increased hepatocyte growth factor production by aging human fibroblasts mainly due to autocrine stimulation by interleukin-1.
      ). Moreover, although in aged fibroblasts fibronectin is increased, it results in defective cell adhesion (
      • Chandrasekhar S.
      • Millis A.J.
      Fibronectin from aged fibroblasts is defective in promoting cellular adhesion.
      ). We can speculate that despite its higher production, in vitiligo fibronectin fails to properly exert adhesion activity because of the senescence-prone phenotype of fibroblasts.
      Collectively, our results point to the involvement of the entire skin in vitiligo, even in normal appearing skin, showing the presence in the dermal compartment of cells with a myofibroblast and a premature senescence phenotype. These cells, producing skin aging-associated secreted proteins, can in turn affect melanocyte functionality favoring their loss. Focusing on normal-appearing skin may allow us to recognize the occurrence of cellular phenomena before the clinically manifested onset of the disease and possibly restrict the spread of the lesions.

      Material and Methods

      Skin biopsies and cell cultures

      Specimens were collected from nonlesional gluteal skin areas of eight vitiligo subjects (two men and six women, age range: 26–66 years) with active nonsegmental disease (on the basis of the progression or appearance of lesions in the last 6 months) observed in the San Gallicano Dermatologic Institute. At the time of patient enrollment, none of the subjects had received either local or systemic therapy for at least 5 months. Normal human skin samples matched for gender, age, and anatomic site were taken from eight healthy volunteers subjected to plastic surgery. The study was approved by the Medical Ethical Committee of the San Gallicano Dermatologic Institute and was conducted according to the Declaration of Helsinki principles. Participants gave their written informed consent. Primary cultures of dermal fibroblasts and melanocytes were isolated and grown as previously described (
      • Flori E.
      • Mastrofrancesco A.
      • Kovacs D.
      • Ramot Y.
      • Briganti S.
      • Bellei B.
      • et al.
      2,4,6-Octatrienoic acid is a novel promoter of melanogenesis and antioxidant defence in normal human melanocytes via PPAR-γ activation.
      ,
      • Kovacs D.
      • Cardinali G.
      • Aspite N.
      • Cota C.
      • Luzi F.
      • Bellei B.
      • et al.
      Role of fibroblast-derived growth factors in regulating hyperpigmentation of solar lentigo.
      ). All the experiments were performed employing cells from short-term cultures (2–10 cell culture passages).

      Statistical analysis

      Student’s t-test was used to assess statistical significance with thresholds of *P ≤ 0.05 and **P ≤ 0.01. The correlation analysis was determined by the coefficient of Pearson’s test (r).
      For immunofluorescence, western blot analysis, flow cytometry, immunohistochemical, real-time RT-PCR, sandwich ELISA, and gas chromatography-mass spectrometry analyses methods, see Supplementary Material online.

      Conflict of Interest

      The authors state no conflict of interest

      Acknowledgments

      We thank Marco Zaccarini for technical assistance. This work was partially supported by the grant RF-2013-02359621 from Ministero della Salute, Italy.

      Supplementary Material

      References

        • Bastonini E.
        • Kovacs D.
        • Picardo M.
        Skin pigmentation and pigmentary disorders: focus on epidermal/dermal cross-talk.
        Ann Dermatol. 2016; 28: 279-289
        • Bellei B.
        • Pitisci A.
        • Ottaviani M.
        • Ludovici M.
        • Cota C.
        • Luzi F.
        • et al.
        Vitiligo: a possible model of degenerative diseases.
        PLoS One. 2013; 8: e59782
        • Bondanza S.
        • Maurelli R.
        • Paterna P.
        • Migliore E.
        • Giacomo F.D.
        • Primavera G.
        • et al.
        Keratinocyte cultures from involved skin in vitiligo patients show an impaired in vitro behaviour.
        Pigment Cell Res. 2007; 20: 288-300
        • Bose S.K.
        • Ortonne J.P.
        Focal gaps in the basement membrane of involved and uninvolved skin of vitiligo. Are they normal?.
        J Dermatol. 1994; 21: 152-159
        • Bourji K.
        • Meyer A.
        • Chatelus E.
        • Pincemail J.
        • Pigatto E.
        • Defraigne J.O.
        • et al.
        High reactive oxygen species in fibrotic and nonfibrotic skin of patients with diffuse cutaneous systemic sclerosis.
        Free Radic Biol Med. 2015; 87: 282-289
        • Briganti S.
        • Flori E.
        • Bellei B.
        • Picardo M.
        Modulation of PPARγ provides new insights in a stress induced premature senescence model.
        PLoS One. 2014; 9: e104045
        • Cáceres M.
        • Oyarzun A.
        • Smith P.C.
        Defective wound-healing in aging gingival tissue.
        J Dent Res. 2014; 93: 691-697
        • Cat B.
        • Stuhlmann D.
        • Steinbrenner H.
        • Alili L.
        • Holtkötter O.
        • Sies H.
        • et al.
        Enhancement of tumor invasion depends on transdifferentiation of skin fibroblasts mediated by reactive oxygen species.
        J Cell Sci. 2006; 119: 2727-2738
        • Chandrasekhar S.
        • Millis A.J.
        Fibronectin from aged fibroblasts is defective in promoting cellular adhesion.
        J Cell Physiol. 1980; 103: 47-54
        • Comer K.A.
        • Dennis P.A.
        • Armstrong L.
        • Catino J.J.
        • Kastan M.B.
        • Kumar C.C.
        Human smooth muscle alpha-actin gene is a transcriptional target of the p53 tumor suppressor protein.
        Oncogene. 1998; 16: 1299-1308
        • Dell'Anna M.L.
        • Ottaviani M.
        • Albanesi V.
        • Vidolin A.P.
        • Leone G.
        • Ferraro C.
        • et al.
        Membrane lipid alterations as a possible basis for melanocyte degeneration in vitiligo.
        J Invest Dermatol. 2007; 127: 1226-1233
        • Dell'Anna M.L.
        • Urbanelli S.
        • Mastrofrancesco A.
        • Camera E.
        • Iacovelli P.
        • Leone G.
        • et al.
        Alterations of mitochondria in peripheral blood mononuclear cells of vitiligo patients.
        Pigment Cell Res. 2003; 16: 553-559
        • Desmoulière A.
        • Geinoz A.
        • Gabbiani F.
        • Gabbiani G.
        Transforming growth factor-beta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts.
        J Cell Biol. 1993; 122: 103-111
        • Desmoulière A.
        • Guyot C.
        • Gabbiani G.
        The stroma reaction myofibroblast: a key player in the control of tumor cell behavior.
        Int J Dev Biol. 2004; 48: 509-517
        • Feghali C.A.
        • Bost K.L.
        • Boulware D.W.
        • Levy L.S.
        Mechanisms of pathogenesis in scleroderma. I: Overproduction of interleukin 6 by fibroblasts cultured from affected skin sites of patients with scleroderma.
        J Rheumatol. 1992; 19: 1207-1211
        • Flori E.
        • Mastrofrancesco A.
        • Kovacs D.
        • Ramot Y.
        • Briganti S.
        • Bellei B.
        • et al.
        2,4,6-Octatrienoic acid is a novel promoter of melanogenesis and antioxidant defence in normal human melanocytes via PPAR-γ activation.
        Pigment Cell Melanoma Res. 2011; 24: 618-630
        • Fuzii H.T.
        • Yoshikawa G.T.
        • Junta C.M.
        • Sandrin-Garcia P.
        • Fachin A.L.
        • Sakamoto-Hojo E.T.
        • et al.
        Affected and non-affected skin fibroblasts from systemic sclerosis patients share a gene expression profile deviated from the one observed in healthy individuals.
        Clin Exp Rheumatol. 2008; 26: 866-874
        • Haass N.K.
        • Herlyn M.
        Normal human melanocyte homeostasis as a paradigm for understanding melanoma.
        J Investig Dermatol Symp Proc. 2005; 10: 153-163
        • Hinz B.
        Myofibroblasts.
        Exp Eye Res. 2016; 142: 56-70
        • Hinz B.
        • Phan S.H.
        • Thannickal V.J.
        • Prunotto M.
        • Desmouliere A.
        • Varga J.
        • et al.
        Recent developments in myofibroblast biology: paradigms for connective tissue remodeling.
        Am J Pathol. 2012; 180: 1340-1355
        • Hsu M.
        • Andl T.
        • Li G.
        • Meinkoth J.L.
        • Herlyn M.
        Cadherin repertoire determines partner-specific gap junctional communication during melanoma progression.
        J Cell Sci. 2000; 113: 1535-1542
        • Kawaguchi Y.
        • Suzuki K.
        • Hara M.
        • Hidaka T.
        • Ishizuka T.
        • Kawagoe M.
        • et al.
        Increased endothelin-1 production in fibroblasts derived from patients with systemic sclerosis.
        Ann Rheum Dis. 1994; 53: 506-510
        • Kendall R.T.
        • Feghali-Bostwick C.A.
        Fibroblasts in fibrosis: novel roles and mediators.
        Front Pharmacol. 2014; 5: 123
        • Kim H.J.
        • Kim K.S.
        • Kim S.H.
        • Baek S.H.
        • Kim H.Y.
        • Lee C.
        • et al.
        Induction of cellular senescence by secretory phospholipase A2 in human dermal fibroblasts through an ROS-mediated p53 pathway.
        J Gerontol A Biol Sci Med Sci. 2009; 64: 351-362
        • Kitamura R.
        • Tsukamoto K.
        • Harada K.
        • Shimizu A.
        • Shimada S.
        • Kobayashi T.
        • et al.
        Mechanisms underlying the dysfunction of melanocytes in vitiligo epidermis: role of SCF/KIT protein interactions and the downstream effector, MITF M.
        J Pathol. 2004; 202: 463-475
        • Kostyuk V.A.
        • Potapovich A.I.
        • Cesareo E.
        • Brescia S.
        • Guerra L.
        • Valacchi G.
        • et al.
        Dysfunction of glutathione S-transferase leads to excess 4-hydroxy-2-nonenal and H(2)O(2) and impaired cytokine pattern in cultured keratinocytes and blood of vitiligo patients.
        Antioxid Redox Signal. 2010; 13: 607-620
        • Kovacs D.
        • Cardinali G.
        • Aspite N.
        • Cota C.
        • Luzi F.
        • Bellei B.
        • et al.
        Role of fibroblast-derived growth factors in regulating hyperpigmentation of solar lentigo.
        Br J Dermatol. 2010; 163: 1020-1027
        • Kuang H.B.
        • Miao C.L.
        • Guo W.X.
        • Peng S.
        • Cao Y.J.
        • Duan E.K.
        Dickkopf-1 enhances migration of HEK293 cell by beta-catenin/E-cadherin degradation.
        Front Biosci (Landmark Ed). 2009; 14: 2212-2220
        • Lafyatis R.
        Transforming growth factor beta—at the centre of systemic sclerosis.
        Nat Rev Rheumatol. 2014; 10: 706-719
        • Leask A.
        Potential therapeutic targets for cardiac fibrosis: TGFbeta, angiotensin, endothelin, CCN2, and PDGF, partners in fibroblast activation.
        Circ Res. 2010; 106: 1675-1680
        • Lee A.Y.
        • Kim N.H.
        • Choi W.I.
        • Youm Y.H.
        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. 2005; 124: 976-983
        • Le Poole I.C.
        • van den Wijngaard R.M.
        • Westerhof W.
        • Das P.K.
        Tenascin is overexpressed in vitiligo lesional skin and inhibits melanocyte adhesion.
        Br J Dermatol. 1997; 137: 171-178
        • Li G.
        • Schaider H.
        • Satyamoorthy K.
        • Hanakawa Y.
        • Hashimoto K.
        • Herlyn M.
        Downregulation of E-cadherin and Desmoglein 1 by autocrine hepatocyte growth factor during melanoma development.
        Oncogene. 2001; 20: 8125-8135
        • Miyazaki M.
        • Gohda E.
        • Kaji K.
        • Namba M.
        Increased hepatocyte growth factor production by aging human fibroblasts mainly due to autocrine stimulation by interleukin-1.
        Biochem Biophys Res Commun. 1998; 246: 255-260
        • Moretti S.
        • Spallanzani A.
        • Amato L.
        • Hautmann G.
        • Gallerani I.
        • Fabiani M.
        • et al.
        New insights into the pathogenesis of vitiligo: imbalance of epidermal cytokines at sites of lesions.
        Pigment Cell Res. 2002; 15: 87-92
        • Nishio K.
        • Inoue A.
        • Qiao S.
        • Kondo H.
        • Mimura A.
        Senescence and cytoskeleton: overproduction of vimentin induces senescent-like morphology in human fibroblasts.
        Histochem Cell Biol. 2001; 116: 321-327
        • Oh S.H.
        • Kim J.Y.
        • Kim M.R.
        • Do J.E.
        • Shin J.Y.
        • Hann S.K.
        DKK1 is highly expressed in the dermis of vitiligo lesion: is there association between DKK1 and vitiligo?.
        J Dermatol Sci. 2012; 66: 163-165
        • Panuncio A.L.
        • Vignale R.
        Ultrastructural studies in stable vitiligo.
        Am J Dermatopathol. 2003; 25: 16-20
        • Picardo M.
        • Dell'Anna M.L.
        • Ezzedine K.
        • Hamzavi I.
        • Harris J.E.
        • Parsad D.
        • et al.
        Vitiligo.
        Nat Rev Dis Primers. 2015; 1: 15011
        • Powell D.W.
        • Mifflin R.C.
        • Valentich J.D.
        • Crowe S.E.
        • Saada J.I.
        • West A.B.
        Myofibroblasts. I: Paracrine cells important in health and disease.
        Am J Physiol. 1999; 277: C1-C9
        • Purpura V.
        • Persechino F.
        • Belleudi F.
        • Scrofani C.
        • Raffa S.
        • Persechino S.
        • et al.
        Decreased expression of KGF/FGF7 and its receptor in pathological hypopigmentation.
        J Cell Mol Med. 2014; 18: 2553-2557
        • Salem M.M.
        • Shalbaf M.
        • Gibbons N.C.
        • Chavan B.
        • Thornton J.M.
        • Schallreuter K.U.
        Enhanced DNA binding capacity on up-regulated epidermal wild-type p53 in vitiligo by H2O2-mediated oxidation: a possible repair mechanism for DNA damage.
        FASEB J. 2009; 23: 3790-3807
        • Serini G.
        • Bochaton-Piallat M.L.
        • Ropraz P.
        • Geinoz A.
        • Borsi L.
        • Zardi L.
        • et al.
        The fibronectin domain ED-A is crucial for myofibroblastic phenotype induction by transforming growth factor-β1.
        J Cell Biol. 1998; 142: 873-881
        • Shi-Wen X.
        • Chen Y.
        • Denton C.P.
        • Eastwood M.
        • Renzoni E.A.
        • Bou-Gharios G.
        • et al.
        Endothelin-1 promotes myofibroblast induction through the ETA receptor via a rac/phosphoinositide 3-kinase/Akt-dependent pathway and is essential for the enhanced contractile phenotype of fibrotic fibroblasts.
        Mol Biol Cell. 2004; 15: 2707-2719
        • Soong J.
        • Chen Y.
        • Shustef M.E.
        • Scott G.
        Sema4D, the ligand for plexin B1, suppresses c-Met activation and migration and promotes melanocyte survival and growth.
        J Invest Dermatol. 2012; 132: 1230-1238
        • Spadoni T.
        • Svegliati Baroni S.
        • Amico D.
        • Albani L.
        • Moroncini G.
        • Avvedimento E.V.
        • et al.
        A reactive oxygen species-mediated loop maintains increased expression of NADPH oxidases 2 and 4 in skin fibroblasts from patients with systemic sclerosis.
        Arthritis Rheumatol. 2015; 67: 1611-1622
        • Taddei M.L.
        • Giannoni E.
        • Raugei G.
        • Scacco S.
        • Sardanelli A.M.
        • Papa S.
        • et al.
        Mitochondrial oxidative stress due to complex I dysfunction promotes fibroblast activation and melanoma cell invasiveness.
        J Signal Transduct. 2012; 2012: 684592
        • Wagner R.Y.
        • Luciani F.
        • Cario-André M.
        • Rubod A.
        • Petit V.
        • Benzekri L.
        • et al.
        Altered E-cadherin levels and distribution in melanocytes precede clinical manifestations of vitiligo.
        J Invest Dermatol. 2015; 135: 1810-1819
        • Waldera Lupa D.M.
        • Kalfalah F.
        • Safferling K.
        • Boukamp P.
        • Poschmann G.
        • Volpi E.
        • et al.
        Characterization of skin aging-associated secreted proteins (SAASP) produced by dermal fibroblasts isolated from intrinsically aged human skin.
        J Invest Dermatol. 2015; 135: 1954-1968
        • Whitfield M.L.
        • Finlay D.R.
        • Murray J.I.
        • Troyanskaya O.G.
        • Chi J.T.
        • Pergamenschikov A.
        • et al.
        Systemic and cell type-specific gene expression patterns in scleroderma skin.
        Proc Natl Acad Sci USA. 2003; 100: 12319-12324
        • Yamaguchi Y.
        • Itami S.
        • Watabe H.
        • Yasumoto K.
        • Abdel-Malek Z.A.
        • Kubo T.
        • et al.
        Mesenchymal-epithelial interactions in the skin: increased expression of dickkopf1 by palmoplantar fibroblasts inhibits melanocyte growth and differentiation.
        J Cell Biol. 2004; 165: 275-285
        • Yanai H.
        • Shteinberg A.
        • Porat Z.
        • Budovsky A.
        • Braiman A.
        • Ziesche R.
        • et al.
        Cellular senescence-like features of lung fibroblasts derived from idiopathic pulmonary fibrosis patients.
        Aging (Albany NY). 2015; 7: 664-672
        • Yang X.
        • Liu T.
        • Chen B.
        • Wang F.
        • Yang Q.
        • Chen X.
        Grape seed proanthocyanidins prevent irradiation-induced differentiation of human lung fibroblasts by ameliorating mitochondrial dysfunction.
        Sci Rep. 2017; 7: 62