Advertisement

CD146 at the Interface between Oxidative Stress and the Wnt Signaling Pathway in Systemic Sclerosis

Open AccessPublished:June 08, 2022DOI:https://doi.org/10.1016/j.jid.2022.03.038
      CD146 involvement was recently described in skin fibrosis of systemic sclerosis through its regulation of the Wnt pathway. Because the interaction between Wnt and ROS signaling plays a major role in fibrosis, we hypothesized that in systemic sclerosis, CD146 may regulate Wnt/ROS crosstalk. Using a transcriptomic and western blot analysis performed on CD146 wild-type or knockout mouse embryonic fibroblasts, we showed a procanonical Wnt hallmark in the absence of CD146 that is reversed when CD146 expression is restored. We found an elevated ROS content in knockout cells and an increase in DNA oxidative damage in the skin sections of knockout mice compared with those of wild-type mice. We also showed that ROS increased CD146 and its noncanonical Wnt ligand, WNT5A, only in wild-type cells. In humans, fibroblasts from patients with systemic sclerosis presented higher ROS content and expressed CD146, whereas control fibroblasts did not. Moreover, CD146 and its ligand were upregulated by ROS in both human fibroblasts. The increase in bleomycin-induced WNT5A expression was abrogated when CD146 was silenced. We showed an interplay between Wnt and ROS signaling in systemic sclerosis, regulated by CD146, which promotes the noncanonical Wnt pathway and prevents ROS signaling, opening the way for innovative therapeutic strategies.

      Abbreviations:

      HDFC (control-derived human donor fibroblast), HDFS (systemic sclerosis‒derived human dermal fibroblast), KO (knockout), MEF (mouse embryonic fibroblast), SSc (systemic sclerosis), WT (wild type)

      Introduction

      Systemic sclerosis (SSc) is a rare and severe autoimmune disease characterized by fibrosis of the skin but also of multiple internal organs. The pathophysiological process involves the induction of an oxidative burst leading to oxidative stress (
      • Servettaz A.
      • Goulvestre C.
      • Kavian N.
      • Nicco C.
      • Guilpain P.
      • Chéreau C.
      • et al.
      Selective oxidation of DNA topoisomerase 1 induces systemic sclerosis in the mouse.
      ), ROS accumulation, and subsequent endothelial cell activation (
      • Mostmans Y.
      • Cutolo M.
      • Giddelo C.
      • Decuman S.
      • Melsens K.
      • Declercq H.
      • et al.
      The role of endothelial cells in the vasculopathy of systemic sclerosis: a systematic review.
      ). Hence, exacerbated ROS production induces chronic inflammation, autoimmunity, and fibroblast activation that contributes to extracellular collagen deposition and, subsequently, fibrosis of the skin and organs (
      • Doridot L.
      • Jeljeli M.
      • Chêne C.
      • Batteux F.
      Implication of oxidative stress in the pathogenesis of systemic sclerosis via inflammation, autoimmunity and fibrosis.
      ;
      • Sambo P.
      • Baroni S.S.
      • Luchetti M.
      • Paroncini P.
      • Dusi S.
      • Orlandini G.
      • et al.
      Oxidative stress in scleroderma: maintenance of scleroderma fibroblast phenotype by the constitutive up-regulation of reactive oxygen species generation through the NADPH oxidase complex pathway.
      ). Indeed, data from animal models have shown that chronic oxidative stress in the skin, as induced by bleomycin, a pro-oxidative agent, is sufficient to induce fibrosis, perturbation of the vascular system, and autoimmunity (
      • Servettaz A.
      • Goulvestre C.
      • Kavian N.
      • Nicco C.
      • Guilpain P.
      • Chéreau C.
      • et al.
      Selective oxidation of DNA topoisomerase 1 induces systemic sclerosis in the mouse.
      ). Therefore, a better understanding of the pathogenesis of SSc is needed to identify new molecular targets and develop alternative therapeutic approaches.
      CD146 was recently identified as a new molecular actor significantly involved in SSc (
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ). CD146 is a cell adhesion molecule of the Ig superfamily that exists as both a membrane-anchored form and a soluble form (soluble CD146) generated by membrane cleavage (
      • Boneberg E.M.
      • Illges H.
      • Legler D.F.
      • Fürstenberger G.
      Soluble CD146 is generated by ectodomain shedding of membrane CD146 in a calcium-induced, matrix metalloprotease-dependent process.
      ). Membrane CD146 is preferentially expressed on vascular endothelial cells (
      • Bardin N.
      • Anfosso F.
      • Massé J.M.
      • Cramer E.
      • Sabatier F.
      • Le Bivic A.
      • et al.
      Identification of CD146 as a component of the endothelial junction involved in the control of cell-cell cohesion.
      ) where it plays an important role in regulating vascular permeability. Moreover, CD146 is also expressed in various other cell types, such as T helper 17 immune cells (
      • Dagur P.K.
      • Biancotto A.
      • Wei L.
      • Sen H.N.
      • Yao M.
      • Strober W.
      • et al.
      MCAM-expressing CD4(+) T cells in peripheral blood secrete IL-17A and are significantly elevated in inflammatory autoimmune diseases.
      ). Physiologically, CD146 and its soluble form are mainly involved in regulating angiogenesis and inflammation (
      • Harhouri K.
      • Kebir A.
      • Guillet B.
      • Foucault-Bertaud A.
      • Voytenko S.
      • Piercecchi-Marti M.D.
      • et al.
      Soluble CD146 displays angiogenic properties and promotes neovascularization in experimental hind-limb ischemia.
      ). CD146 is required to activate the angiogenic function of VEGF-R2 (
      • Jiang T.
      • Zhuang J.
      • Duan H.
      • Luo Y.
      • Zeng Q.
      • Fan K.
      • et al.
      CD146 is a coreceptor for VEGFR-2 in tumor angiogenesis.
      ), and CD146 is also known as one of the receptors of WNT5A glycoprotein (
      • Ye Z.
      • Zhang C.
      • Tu T.
      • Sun M.
      • Liu D.
      • Lu D.
      • et al.
      Wnt5a uses CD146 as a receptor to regulate cell motility and convergent extension.
      ). In patients with SSc, we have shown that the decrease in serum concentration of soluble CD146 is associated with a poor prognosis of the disease in two independent cohorts of patients with SSc (
      • Gabsi A.
      • Heim X.
      • Dlala A.
      • Gati A.
      • Sakhri H.
      • Abidi A.
      • et al.
      TH17 cells expressing CD146 are significantly increased in patients with Systemic sclerosis.
      ;
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ), further confirming what was established by
      • Ito T.
      • Tamura N.
      • Okuda S.
      • Tada K.
      • Matsushita M.
      • Yamaji K.
      • et al.
      Elevated serum levels of soluble CD146 in patients with systemic sclerosis.
      . Next, in a mouse model of SSc induced by bleomycin, we showed that CD146 knockout (KO) mice were more prone to develop skin fibrosis than wild-type (WT) mice, with an upregulation of the canonical β-catenin‒dependent pathway (
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ). The Wnt pathway is generally divided into three subpathways: the canonical pathway, the noncanonical planar cell polarity pathway, and the noncanonical Wnt/calcium pathway. The canonical pathway activated inter alia by WNT3a initiates intracellular signaling through β-catenin nuclear translocation. The other two are qualified as β-catenin independent and are associated with differentiation, cell polarity, migration, and canonical pathway inhibition, especially with the noncanonical ligand WNT5A. In addition, in SSc, it has been observed that Wnt antagonists such as Dickkopf, SFRP, and WIF1 are decreased (
      • Dees C.
      • Schlottmann I.
      • Funke R.
      • Distler A.
      • Palumbo-Zerr K.
      • Zerr P.
      • et al.
      The Wnt antagonists DKK1 and SFRP1 are downregulated by promoter hypermethylation in systemic sclerosis.
      ;
      • Svegliati S.
      • Marrone G.
      • Pezone A.
      • Spadoni T.
      • Grieco A.
      • Moroncini G.
      • et al.
      Oxidative DNA damage induces the ATM-mediated transcriptional suppression of the Wnt inhibitor WIF-1 in systemic sclerosis and fibrosis.
      ).
      Recent publications have shed light on the link between Wnt and ROS signaling crosstalk (
      • Funato Y.
      • Michiue T.
      • Asashima M.
      • Miki H.
      The thioredoxin-related redox-regulating protein nucleoredoxin inhibits Wnt-beta-catenin signalling through dishevelled.
      ), particularly those involved in the fibrosis process (
      • Andersson-Sjöland A.
      • Karlsson J.C.
      • Rydell-Törmänen K.
      ROS-induced endothelial stress contributes to pulmonary fibrosis through pericytes and Wnt signaling.
      ;
      • Yu Y.
      • Guan X.
      • Nie L.
      • Liu Y.
      • He T.
      • Xiong J.
      • et al.
      DNA hypermethylation of sFRP5 contributes to indoxyl sulfate-induced renal fibrosis.
      ). However, the involvement of this signaling mechanism in SSc is not well-defined and requires identifying the molecular actors of this interplay. We hypothesized that CD146 could represent a mediator of the Wnt/ROS signaling crosstalk participating in the development of fibrosis in patients with SSc.

      Results

      CD146 mediates a balanced ratio between canonical and noncanonical Wnt signaling

      In a previous study, we described an upregulation of the canonical Wnt pathway in the absence of CD146 (
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ). To further investigate the molecular mechanisms underlying this initial observation, we performed a transcriptomic analysis focusing on 84 genes implicated in the Wnt signaling pathway as effector or regulatory molecules of the canonical and noncanonical pathways using WT or Cd146 KO mouse embryonic fibroblasts (MEFs) (Figure 1a). Transcriptomic analysis confirmed our previous in vitro data, which revealed overexpression of 18 genes implicated in the Wnt canonical pathway in MEFs from KO mice. Thus, an accumulating body of evidence points toward a procanonical Wnt hallmark in the absence of CD146. Among the overexpressed genes, the protein expression levels of WNT3a, a typical activator of the canonical pathway, and WIF1 were further analyzed, owing to their well-defined role in the pathophysiology of SSc (
      • Svegliati S.
      • Marrone G.
      • Pezone A.
      • Spadoni T.
      • Grieco A.
      • Moroncini G.
      • et al.
      Oxidative DNA damage induces the ATM-mediated transcriptional suppression of the Wnt inhibitor WIF-1 in systemic sclerosis and fibrosis.
      ;
      • Wei J.
      • Fang F.
      • Lam A.P.
      • Sargent J.L.
      • Hamburg E.
      • Hinchcliff M.E.
      • et al.
      Wnt/β-catenin signaling is hyperactivated in systemic sclerosis and induces Smad-dependent fibrotic responses in mesenchymal cells.
      ). Next, we performed western blots on different batches of WT or KO MEFs. These results confirmed those obtained by transcriptomic analysis by showing a robust increase in both WNT3a and WIF1 in KO MEFs (Figure 1b‒d) compared with that in WT MEFs. We also analyzed the protein expression of β-catenin by western blot because it is a major mediator in the Wnt canonical pathway even if its expression was not modified at the transcriptional level. Notably, we found significantly higher protein expression of β-catenin in the MEFs from KO mice than in those from WT mice in two different batches of cells (Figure 1b and e). To further confirm these results, KO MEFs were transiently transfected with a CD146-coding plasmid before the analysis of protein expression. Replenishing CD146 expression in KO MEFs was associated with the loss of WNTa, WIF1, and β-catenin expression and showed that the restoration of CD146 expression inhibited the procanonical signature, confirming the results obtained with WT MEFs (Figure 1b‒e). Altogether, our results showed a procanonical Wnt hallmark in the absence of CD146, suggesting that CD146 helps to maintain a balanced ratio between the Wnt canonical and noncanonical signaling pathways.
      Figure thumbnail gr1
      Figure 1CD146 deficiency is associated with increased Wnt canonical signaling. (a) Transcriptional profiling of Wnt signaling components showed 18 upregulated but 3 downregulated genes in MEFs from KO mice compared with those from WT mice. (b‒e) Representative western blots for the selected proteins (b) WNT3A, WIF1, and β-catenin and their corresponding relative quantification (n ≥ 3) for (c) WNT3A, (d) WIF1, and (e) β-catenin are shown. KO MEFs were also transfected with a plasmid coding for CD146. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM. KO, knockout; MEF, mouse embryonic fibroblast; WT, wild type.

      Absence of CD146 enhances ROS production and sensitivity to ROS effects

      To elucidate whether CD146 participates in regulating ROS production, we assessed the generation of ROS in KO and WT MEFs. Basal conditions for up to 6 hours resulted in higher ROS generation in KO cells than in WT cells from the second hour until the sixth hour (Figure 2a). Next, we used bleomycin as an exogenous source of oxidative stress and compared the ROS production of WT MEFs with that of KO MEFs. Bleomycin treatment at a high dose (10 μg/ml) significantly increased ROS generation in both WT and KO MEFs in a similar fashion to that induced by indoxyl sulfate (200 μM), which was used as a positive control (
      • Dou L.
      • Jourde-Chiche N.
      • Faure V.
      • Cerini C.
      • Berland Y.
      • Dignat-George F.
      • et al.
      The uremic solute indoxyl sulfate induces oxidative stress in endothelial cells.
      ) (Figure 2b). Of interest, this effect was abolished when N-acetylcysteine, an inhibitor of oxidative stress, was added to the treatment (Figure 2c). As shown in Figure 2d, bleomycin at a high concentration, 10 μg/ml, efficiently induced ROS generation in both WT and KO MEFs; however, at doses as low as 1 μg/ml, bleomycin was able to induce ROS only in KO MEFs (Figure 2d). We next examined the level of DNA base oxidative damage in skin sections from WT and KO mice. As shown in Figure 3a (left panel) and b, there was a significant increase in the staining intensity of 8-oxo-2'-deoxyguanosine, a marker of oxidized DNA bases, in the skin sections of KO mice compared with that in WT mice at steady state, confirming that KO mice are more sensitive to ROS than WT mice. After the induction of oxidative stress by bleomycin treatment, skin sections from WT mice showed increased DNA oxidative, suggesting that the protective effect of CD146 from ROS is lost under high oxidative stress conditions (Figure 3a [left panel] and b).
      Figure thumbnail gr2
      Figure 2CD146 inhibits ROS production and decreases cell sensitivity to ROS. (a) Time course of ROS generation by WT (black) or KO (red) MEFs in basal culture media as assessed by cytofluorimetry. (b) ROS generation by MEFs 6 hours after induction by bleomycin (10 μg/ml) or IS (200 μM) relative to nonstimulated cells. (c) ROS generation by MEFs 6 hours after induction by bleomycin (10 μg/ml). NAC (at a final concentration of 50 mM) was added to the cultured MEFs before treatment with bleomycin. (d) KO and WT MEFs were treated with either a low or a high bleomycin concentration, and the effect on ROS generation was detected by cytofluorimetry. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM. IS, indoxyl sulfate; KO, knockout; MEF, mouse embryonic fibroblast; NAC, N-acetyl cysteine; n.s., not significant; WT, wild type.
      Figure thumbnail gr3
      Figure 3CD146 inhibits ROS-induced skin fibrosis. (a) Representative images of anti‒8-OH-dG immunostaining (left panel), Masson’s trichrome staining (middle panel), and anti‒α-SMA immunostaining (right panel) on skin biopsies from WT and KO mice injected with either saline or bleomycin. Examples of cells positive for 8-OH-dG and α-SMA immunostaining are indicated with red and blue arrowheads, respectively. (b) Quantification of ROS-induced DNA damage in KO and WT mice treated with or without bleomycin. (c) Quantification of dermal thickness in KO and WT mice treated with or without bleomycin. (d) Positive correlation between dermal thickness and anti‒α-SMA immunostaining in skin sections from WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM. Bars = 50 μm. 8-OH-dG, 8-oxo-2'-deoxyguanosine; α-SMA, α-smooth muscle actin; KO, knockout; n.s., not significant; WT, wild type.

      Absence of CD146 enhances skin fibrosis under oxidative stress conditions

      To reinforce the link between CD146 and fibrosis mediated by ROS, dermal thickness was analyzed using Masson’s trichrome staining, which stains collagen fibers blue (
      • Yang L.
      • Serada S.
      • Fujimoto M.
      • Terao M.
      • Kotobuki Y.
      • Kitaba S.
      • et al.
      Periostin facilitates skin sclerosis via PI3K/Akt dependent mechanism in a mouse model of scleroderma.
      ), and the presence of myofibroblasts in dermal tissue was investigated. As shown in Figure 3a (middle panel) and c, dermal thickness was significantly increased in KO mice after bleomycin treatment compared with that after saline solution injection, whereas no difference was observed in WT mice nor between WT mice and KO mice after saline solution injection. These results are consistent with our previously published data (
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ), which described an increase in dermal thickness in KO mice after bleomycin injection compared with that after the control condition (saline solution), whereas this concentration of bleomycin (10 μg/ml) was not sufficient to induce a dermal thickness increase in WT mice. Finally, because myofibroblasts produce collagen in the extracellular matrix, we investigated the presence of this type of differentiated fibroblast in our model using α-smooth muscle actin marker (
      • Hinz B.
      Formation and function of the myofibroblast during tissue repair.
      ). No difference in the number of positive cells between WT and KO mice was observed (data not shown). The dermal thickness was positively correlated with anti‒α-smooth muscle actin immunostaining, indicating that dermal collagen content is associated with the presence of myofibroblasts (Figure 3d [right panel] and e). Altogether, our data led us to confirm that KO mice developed more skin fibrosis than WT mice under oxidative stress.

      ROS induction increases CD146 and WNT5A expression

      Because our results showed a higher basal ROS content in the absence of CD146, we wondered whether ROS could conversely modulate CD146 expression. To assess this hypothesis, WT MEFs were treated with low (1 μg/ml) or high (10 μg/ml) doses of bleomycin for 6 hours, and then the mRNA and protein expression of CD146 was analyzed. Remarkably, not only the high dose of bleomycin induced a steep increase in CD146 protein expression in WT MEFs but also the lower dose (Figure 4a). These results were also confirmed by RT-qPCR, which showed a significant increase in Cd146 transcripts after 10 μg/ml bleomycin treatment (Figure 4b). Next, we investigated whether the CD146 ligand WNT5A was also altered by bleomycin treatment. As shown in Figure 4c, bleomycin treatment significantly increased WNT5A expression in the same way as CD146 in WT MEFs, whereas there was no variation in WNT5A protein expression in KO cells. The addition of N-acetylcysteine, a ROS inhibitor, to bleomycin treatment abrogated the upregulation of CD146 and WNT5A on WT MEFs (Figure 4d), reinforcing the fact that oxidative stress increases the expression of CD146 and WNT5A. Furthermore, in both WT and KO MEFs, the mRNA levels of Wnt5a were not modified by bleomycin treatment (Figure 4e). All our results showed that ROS induction increases the synthesis of CD146 and the expression of both CD146 and WNT5A. To further elucidate the mechanism, we first investigated whether CD146 inhibits the degradation of WNT5A after a cycloheximide multiple-chase experiment and analyzed WNT5A expression (Figure 4f). We confirmed an upregulation of WNT5A expression by bleomycin only in WT MEFs and not in KO MEFs regardless of the time of bleomycin treatment. In WT MEFs, no significant difference between bleomycin and bleomycin/cycloheximide treatment was found at 6, 12, or 24 hours, showing a translation-independent increase in WNT5A expression. This suggests that in the presence of CD146, WNT5A could be protected from degradation. Then, we tested whether WNT5A protein levels were modified after proteasome blockade (Supplementary Figure S1). A significant reduction of WNT5A was observed in KO MEFs after proteasome blockade, suggesting, as described in the literature (
      • Fels D.R.
      • Ye J.
      • Segan A.T.
      • Kridel S.J.
      • Spiotto M.
      • Olson M.
      • et al.
      Preferential cytotoxicity of bortezomib toward hypoxic tumor cells via overactivation of endoplasmic reticulum stress pathways.
      ;
      • Harhouri K.
      • Navarro C.
      • Depetris D.
      • Mattei M.G.
      • Nissan X.
      • Cau P.
      • et al.
      MG132-induced progerin clearance is mediated by autophagy activation and splicing regulation.
      ), that another degradation pathway is stimulated. In contrast, such a decrease was not observed in WT MEFs, suggesting that degradation pathways are less activated in the presence of CD146. We investigated the potential activation of autophagy by analyzing the expression of LC3. The LC3BII/LC3BI ratio was increased after MG132 treatment in both WT and KO MEFs. Interestingly, the activation of autophagy was significantly higher in KO MEFs than in WT MEFs (P = 0.03) (Supplementary Figure S1).
      Figure thumbnail gr4
      Figure 4CD146 prevents WNT5A degradation under oxidative stress conditions. (a, b) WT and KO MEFs were treated with increasing concentrations of BLM, and the effect on (a) CD146 protein expression with a representative western blot (left) and the relative quantification (right) and mRNA was studied. (c) Representative western blot (left) and the relative quantification (right) of WNT5A protein expression after BLM treatment are shown. (d) Representative western blot analysis for the proteins CD146 and WNT5A after BLM treatment with or without NAC and the relative quantification are shown. (e) mRNA levels of Wnt5a after BLM treatment are shown. (f) WT and KO MEFs were treated with CHX and/or BLM for 6, 12, and 24 h, and the effect on WNT5A protein expression was studied. Relative quantification (right) and representative western blot at 12 h (left) are shown. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM. All experiments were performed in triplicate. BLM, bleomycin; CHX, cycloheximide; h, hour; KO, knockout; MEF, mouse embryonic fibroblast; NAC, N-acetyl cysteine; WT, wild type.

      ROS impact CD146/WNT5A regulation in human cells

      To assess whether these data were reproducible in human cells, we isolated human dermal fibroblasts from five patients with SSc (i.e., SSc-derived human dermal fibroblasts [HDFSs]) and five healthy controls (i.e., control-derived human dermal fibroblasts [HDFCs]). Patients with SSc presented different clinical presentations and autoimmune profiles to consider the disease heterogeneity (Supplementary Table S1). First, we observed significantly higher ROS production in HDFSs than in HDFCs, using two distinct methods (Figure 5a and Supplementary Figure S2). Second, we evaluated the expression of WNT5A and CD146 on both HDFSs and HDFCs at basal conditions or after bleomycin treatment. Of interest, WNT5A was initially significantly more expressed in HDFSs than in HDFCs (Figure 5b), whereas bleomycin treatment increased significantly WNT5A in HDFCs (Figure 5c). In accordance with previous data (
      • Blasi A.
      • Martino C.
      • Balducci L.
      • Saldarelli M.
      • Soleti A.
      • Navone S.E.
      • et al.
      Dermal fibroblasts display similar phenotypic and differentiation capacity to fat-derived mesenchymal stem cells, but differ in anti-inflammatory and angiogenic potential.
      ), we did not detect CD146 expression on HDFCs (Figure 5d); however, HDFSs solely expressed this marker under basal conditions (Figure 5e). This expression of CD146 was confirmed on fibroblasts from different patients with SSc regardless of their clinical presentation or antibody profile (Supplementary Figure S3). Bleomycin treatment induced, to a similar extent to that of hydrogen peroxide, an upregulation in the surface expression of CD146 on HDFCs (Figure 5f), which suggests, at least in part, that the induction of CD146 is ROS dependent in healthy subjects. To confirm these results, the downregulation of CD146 by small interfering RNA was performed in HDFSs before submitting the cells to bleomycin treatment. As shown in Figure 5g and h, the increase in bleomycin-induced WNT5A expression was abrogated when CD146 was silenced on HDFSs, showing that CD146 regulates the augmentation of WNT5A expression after oxidative stress.
      Figure thumbnail gr5
      Figure 5ROS in human fibroblasts enhance WNT5A and CD146 expression. (a) HDFCs and HDFSs were serum starved for 6 h, and the generated ROS were evaluated by cytofluorimetry (left) or flow cytometry (right). (b, c) Flow cytometry analysis of WNT5A surface expression and the corresponding quantification of the MFI on both HDFCs and HDFSs at (b) basal conditions or (c) after 48 h of bleomycin treatment (10 μg/ml). (d, e) CD146 expression levels on both (d) HDFCs and (e) HDFSs were assessed by flow cytometry and western blot. (f) Flow cytometry analysis of CD146 surface expression on HDFCs after bleomycin (10 μg/ml) or H2O2 (200 μM) treatment and the corresponding quantification of the MFI. (g, h) Flow cytometry analysis of (g) WNT5A and (h) CD146 expression after transfection with siCTRL or siCD146 (left) and the corresponding quantification of the MFI (right) on HDFSs 48 h after bleomycin (10 μg/ml). ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM for a minimum of three independent experiments on two batches of skin fibroblasts. CTRL, control; h, hour; H2O2, hydrogen peroxide; HDF, human donor fibroblast; HDFC, control-derived human donor fibroblast; HDFS, systemic sclerosis‒derived human dermal fibroblast; HUVEC, human vein endothelial cell; Max, maximum; MFI, median fluorescence intensity; siCD146, CD146-targeted small interfering RNA; siCTRL, control-targeted small interfering RNA; siRNA, small interfering RNA.

      Discussion

      To date, the interaction between ROS and the Wnt signaling pathway has been established and defined in the literature; however, how adhesion molecules such as CD146 intervene and regulate the interplay between ROS and Wnt signaling still must be elucidated. In this study, we provided insight into the relationship between oxidative stress and Wnt signaling by showing that CD146 is an essential factor mediating the exchange between ROS and Wnt signaling, which to our knowledge, is previously unreported. We also show that CD146 intercedes under oxidative stress in SSc to enhance WNT5A expression and to inhibit the profibrotic Wnt canonical pathway (Figure 6).
      Figure thumbnail gr6
      Figure 6Illustrative summary of events occurring in an oxidative stress environment. In response to oxidative stress, ROS are released from the mitochondria and subsequently activate CD146 gene transcription, which enhances CD146 surface expression. The interaction between CD146 and its ligand, WNT5A, inhibits ROS secretion (1), protects the WNT5a intracellular pool from degradation (2), and favors the noncanonical Wnt pathway while inhibiting the canonical pathway through selective induction of β-catenin degradation (3). This homeostatic balance between the canonical and noncanonical Wnt pathways, as mediated by CD146 (4), attenuates collagen synthesis and deposition in the extracellular matrix, thereby impeding SSc disease progression. The decrease in β-catenin availability and therefore canonical Wnt signaling has been associated with decreased chronic inflammation, autoimmune diseases, and fibrotic events. SSc, systemic sclerosis.
      A large body of evidence attributes the fibrotic events in SSc to exacerbated canonical Wnt signaling (
      • Bergmann C.
      • Distler J.H.W.
      Canonical Wnt signaling in systemic sclerosis.
      ). In this study, we confirmed our previous findings by showing that KO mice developed more skin fibrosis than WT mice under oxidative stress in SSc. Although the increase in dermal thickness is associated with fibroblast differentiation into myofibroblasts, which secrete collagen in the extracellular matrix (
      • Hinz B.
      Formation and function of the myofibroblast during tissue repair.
      ), the dermal thickness increase can also be attributed to other factors, such as ROS, which activate profibrotic gene transcription and collagen synthesis. ROS also inhibit extracellular matrix degradation, participating in the dermal thickness increase and fibrosis (
      • Thuan D.T.B.
      • Zayed H.
      • Eid A.H.
      • Abou-Saleh H.
      • Nasrallah G.K.
      • Mangoni A.A.
      • et al.
      A potential link between oxidative stress and endothelial-to-mesenchymal transition in systemic sclerosis.
      ). In our model, we observed an increase in dermal thickness and collagen content, partly due to activation of myofibroblasts (as indicated by the correlation between α-smooth muscle actin and trichrome staining) and ROS production.
      In addition, we showed a significant increase in WNT3a, WIF1, and β-catenin expression in KO but not in WT MEFs. The restoration of CD146 expression in KO MEFs transfected with a plasmid encoding human CD146 inhibited the expression of these procanonical proteins.
      We showed in this study that CD146 negatively regulates canonical Wnt signaling preferentially by inhibiting β-catenin expression. Otherwise, because β-catenin is known to be a central mediator of profibrotic Wnt signaling in SSc, our results highlight the important role of CD146 in the regulation of fibrosis (
      • Beyer C.
      • Schramm A.
      • Akhmetshina A.
      • Dees C.
      • Kireva T.
      • Gelse K.
      • et al.
      β-catenin is a central mediator of pro-fibrotic Wnt signaling in systemic sclerosis.
      ). The inhibitory function of CD146 on β-catenin was previously reported by
      • Ye Z.
      • Zhang C.
      • Tu T.
      • Sun M.
      • Liu D.
      • Lu D.
      • et al.
      Wnt5a uses CD146 as a receptor to regulate cell motility and convergent extension.
      . In addition, CD146 is a receptor for WNT5A, a major activator of noncanonical Wnt signaling. These results led us to assume that CD146 is a central regulator of the interrelated canonical and noncanonical Wnt signaling pathways and subsequently fibrosis.
      The first finding of our study is that CD146 acts as a mediator linking ROS and the Wnt signaling pathway through WNT5A ligand. Using bleomycin as a ROS-inducing agent, we showed that both CD146 and the main Wnt noncanonical ligand, WNT5A, were highly upregulated only in the presence of CD146. However, because Wnt5a gene expression did not change at the transcriptional level after bleomycin treatment, we speculated that CD146 could protect WNT5A from degradation. In agreement with our hypothesis, we showed that cycloheximide multiple-chase experiments did not affect the upregulation of WNT5A expression in WT CD146 cells after bleomycin treatment. Moreover, after proteasome blockade, a significant diminution of WNT5A was observed in KO MEFs but not in WT MEFs, in which CD146 may protect WNT5A from the activation of other degradation pathways (
      • Ding W.-X.
      • Ni H.-M.
      • Gao W.
      • Yoshimori T.
      • Stolz D.B.
      • Ron D.
      • et al.
      Linking of autophagy to ubiquitin-proteasome system is important for the regulation of endoplasmic reticulum stress and cell viability.
      ;
      • Harhouri K.
      • Navarro C.
      • Depetris D.
      • Mattei M.G.
      • Nissan X.
      • Cau P.
      • et al.
      MG132-induced progerin clearance is mediated by autophagy activation and splicing regulation.
      ;
      • Kato M.
      • Ospelt C.
      • Gay R.E.
      • Gay S.
      • Klein K.
      Dual role of autophagy in stress-induced cell death in rheumatoid arthritis synovial fibroblasts.
      ). Accordingly, an analysis of autophagy showed preferential activation in the absence of CD146, suggesting that CD146 could protect WNT5A from autophagic degradation.
      The regulation of Wnt signaling by ROS was first proposed by
      • Funato Y.
      • Michiue T.
      • Asashima M.
      • Miki H.
      The thioredoxin-related redox-regulating protein nucleoredoxin inhibits Wnt-beta-catenin signalling through dishevelled.
      , who showed that DVL activates the Wnt/β-catenin pathway under oxidative conditions. Then, the crosstalk between ROS and Wnt/β-catenin signaling was further investigated in cancer diseases (
      • Aceto G.M.
      • Catalano T.
      • Curia M.C.
      Molecular aspects of colorectal adenomas: the interplay among microenvironment, oxidative stress, and predisposition.
      ). CD146 acts as a proinflammatory and proangiogenic factor in several types of cancers. SSc disease and cancers share a bundle of common pathological phenomena, as attested by epidemiological studies (
      • Mellemkjaer L.
      • Pfeiffer R.M.
      • Engels E.A.
      • Gridley G.
      • Wheeler W.
      • Hemminki K.
      • et al.
      Autoimmune disease in individuals and close family members and susceptibility to non-Hodgkin’s lymphoma.
      ;
      • Rosenthal A.K.
      • McLaughlin J.K.
      • Gridley G.
      • Nyrén O.
      Incidence of cancer among patients with systemic sclerosis.
      ;
      • Zeineddine N.
      • Khoury L.E.
      • Mosak J.
      Systemic sclerosis and malignancy: a review of current data.
      ). Similarly, genetic instabilities; environmental factors; and some signaling pathways, such as Wnt, are behind the initiation of many cancers as well as autoimmune diseases, including SSc. Thus, our work reinforces the intriguing relationship between cancer and SSc by identifying CD146 as a principal regulator of Wnt/ROS crosstalk.
      Although the susceptibility to autoimmune diseases is multifactorial, ROS are increased in many pathologies, especially in SSc (
      • Doridot L.
      • Jeljeli M.
      • Chêne C.
      • Batteux F.
      Implication of oxidative stress in the pathogenesis of systemic sclerosis via inflammation, autoimmunity and fibrosis.
      ;
      • Smallwood M.J.
      • Nissim A.
      • Knight A.R.
      • Whiteman M.
      • Haigh R.
      • Winyard P.G.
      Oxidative stress in autoimmune rheumatic diseases.
      ). In this study, we found higher levels of ROS in patients with HDFS than in those with HDFC. These results are in agreement with previous data showing increased endogenous oxidative stress in patients with SSc (
      • Sambo P.
      • Baroni S.S.
      • Luchetti M.
      • Paroncini P.
      • Dusi S.
      • Orlandini G.
      • et al.
      Oxidative stress in scleroderma: maintenance of scleroderma fibroblast phenotype by the constitutive up-regulation of reactive oxygen species generation through the NADPH oxidase complex pathway.
      ). The authors evidenced oxidative stress as an early marker in disease pathogenesis, with higher ROS levels detected in fibroblasts extracted from the skin of patients with SSc than from the skin of control individuals, suggesting that ROS are the main drivers of disease progression.
      Furthermore, we validated by western blot and flow cytometry that CD146 is uniquely expressed on HDFSs and not on HDFCs. The expression of CD146 was confirmed on fibroblasts of all patients with SSc independent of their clinical symptoms and autoantibodies profile. In addition, after bleomycin treatment of HDFCs, we induced CD146 expression. We then hypothesized that ROS-induced CD146 expression acts as a feedback regulator on the functional consequences of ROS signaling.
      In conclusion, our study evidences the existence of an interplay between ROS and Wnt signaling orchestrated by CD146 in the pathological processes of SSc and opens up a new field of study in autoimmune diseases. These findings add to our understanding of pathological mechanisms in the field of autoimmune diseases and provide potential therapeutic targets and strategies.

      Materials and Methods

      Detailed materials and methods are described in Supplementary Materials and Methods.

      MEFs culture

      MEFs were isolated from day 13.5 embryos from KO mice or WT mice as previously described (
      • Xu J.
      Preparation, Culture, and Immortalization of Mouse Embryonic Fibroblasts.
      ).

      Human dermal fibroblasts culture

      The culture of fibroblasts was established from five patients with SSc (i.e., HDFSs) and from controls (i.e., HDFCs) using explant cultures from skin biopsies obtained during surgery. Written informed consent was obtained from patients with SSc in the context of Scleradec II clinical trial: ClinicalTrials.gov NCT02558543 authorized by the Comité de protection des personnes Sud méditerranée V (2014-003023-22). Dermal fibroblasts from healthy donors were obtained from surgical residues of skin for cosmetic purposes. All healthy donors provided written informed nonobjection for the scientific use of surgical residues.

      Measurement of ROS

      ROS production in MEFs and human dermal fibroblasts was measured by cytofluorimetry or flow cytometry with carboxy-2',7'-dichlorodihydrofluorescein diacetate (Thermo Fisher Scientific, Waltham, MA).

      Histological analysis and immunostaining of murine skin biopsy

      We performed the immunostaining and immunochemistry in the skin section using 8-oxo-2'-deoxyguanosine with avidin-biotin-peroxidase complex (Life Technologies, Carlsbad, CA) for 8-oxo-2'-deoxyguanosine and on Ventana Benchmark system for α-smooth muscle actin and Masson’s trichrome.

      Western blotting

      Western blotting of whole-cell lysate extracts was performed to evaluate protein expression. Lysis buffer and primary antibodies are listed in the Supplementary Materials and Methods.

      RT-qPCR

      Total RNA was isolated and subjected to real-time RT-qPCR using the probes obtained from Thermo Fisher Scientific.

      MEFs transfection

      Briefly, cells were transiently transfected using jetPRIME transfection agent (Polyplus transfection, Illkirch-Graffenstaden, France) according to the manufacturer’s instruction.

      Cycloheximide chase analysis

      Protein biosynthesis was inhibited by adding cycloheximide (Sigma-Aldrich, Vienna, Austria) to the culture media simultaneously with or without bleomycin.

      Statistical analysis

      Statistical analysis was performed with Prism software (GraphPad Software, San Diego, CA). Significant differences were determined using the nonparametric Mann–Whitney U test. When comparing more than two groups, a nonparametric one-way ANOVA followed by Dunn’s multiple comparison test was used. A P < 0.05 was considered to be significant.

      Data availability statement

      No datasets were generated or analyzed during this study. All the data that support the findings of our study are available from the corresponding author on request.

      Conflict of Interest

      The authors state no conflict of interest.

      Acknowledgments

      We thank BioCytex (Marseille, France) for providing recombinant soluble CD146 and S-Endo-1 antibodies and Stephanie Simoncini, Karim Harhouri, Lucile Tuchtan-Torrents, and Marie-Dominique Piercecchi-Marti for their help and support. The work was done in Marseille, France. Aix-Marseille University, INSERM, INRAE, and GFRS (Groupe Francophone de Recherche sur la Sclérodermie) were funders of the study.

      Author Contributions

      Conceptualization: XH, JB, RB, AFB, ASL, MBC, NB; Data Curation: XH, JB, AJ, EK, RB, MN, MV, LD, ASL, ABe, AD, AFB, ASL, BG; Formal Analysis: XH, JB, ABr, AJ, EK, NB; Funding Acquisition: FDG, MBC, NB; Investigation: XH, JB, AJ, EK, RB, MN, MV, LD, ABr, ASL, ABe, AD, AFB, ASL, BG; Methodology: XH, JB, AJ, ABr, ASL, AFB, LD, AD, NB, MBC; Project Administration: XH, JB, NB, MBC; Resources: RB, MV, ASL, AD, FS, BG, MBC, NB; Supervision: MBC, NB; Validation: XH, NB; Visualization: XH, JB, AJ, EK, ABr, NB; Writing - Original Draft Preparation: XH, AJ, RB, MBC, NB; Writing - Review and Editing: XH, JB, AJ, EK, RB, MN, MV, LD, ABr, AFB, ASL, ABe, AD, BG, FS, FDG, MBC, NB

      Supplementary Materials and Methods

      Human dermal fibroblasts culture

      The culture of fibroblasts was established from five patients with systemic sclerosis (i.e., systemic sclerosis‒derived human dermal fibroblast) and from controls (i.e., control-derived human donor fibroblast). The skin biopsies were obtained as surgical waste during a surgical operation, and nonopposition with informed consent was obtained from the patients according to the Helsinki rules. Tissues were rinsed with DMEM (Gibco, Paisley, United Kingdom) containing 400 IU/ml penicillin, 400 μg/ml streptomycin, and 4 μg/ml amphotericin B and cut into small pieces. Explants were deposited in a 25 cm two-culture flask and incubated in a humidified atmosphere of 5% carbon dioxide in DMEM supplemented with 20% fetal calf serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 μg/ml amphotericin B. In these conditions, fibroblasts leave the explant in 3–5 days and settle at the entire flask surface in 3 weeks. Fibroblasts were grown in DMEM supplemented with 10% fetal calf serum (Gibco), 100 μg/ml streptomycin, and 100 IU/ml penicillin.

      Mouse embryonic fibroblasts culture

      Mouse embryonic fibroblasts (MEFs) were isolated from day 13.5 embryos from Cd146 knockout (KO) mice or wild-type (WT) mice as previously described (
      • Xu J.
      Preparation, Culture, and Immortalization of Mouse Embryonic Fibroblasts.
      ). Briefly, the placental membrane, amniotic sac, head, and primordial blood organs were removed, and the remaining carcass was rinsed with PBS and minced in 2 ml PBS. The tissue fragments were passed through a 100 mm strainer to remove large fragments and placed in a 25 cm2 flask containing DMEM (Gibco), 10% (v/v) fetal calf serum (Gibco), 50 units/ml penicillin, and 50 mg/ml streptomycin. At this and subsequent stages of culture, cells were maintained in 5% oxygen. At confluence, the cells were transferred to a 75 cm2 flask.

      Measurement of ROS

      For measurement of ROS production by cytofluorimetry, MEFs and human dermal fibroblasts were seeded in 96-well plates at 15,000 cells per well. The next day, fibroblasts were labeled for 45 minutes with 10 μM of di(acetoxymethylester)6-carboxy-2',7'-dichlorodihydrofluorescein diacetate. Cells were washed with PBS and treated with bleomycin at 0, 1, and 10 μg/ml in DMEM medium without phenol red. Fluorescence intensity was measured every 1 hour during a 6-hour period. In some experiments, MEFs were incubated with indoxyl sulfate as a positive control (
      • Dou L.
      • Jourde-Chiche N.
      • Faure V.
      • Cerini C.
      • Berland Y.
      • Dignat-George F.
      • et al.
      The uremic solute indoxyl sulfate induces oxidative stress in endothelial cells.
      ) or in the presence of antioxidant N-acetylcysteine (10 mM). The production of intracellular ROS was detected by measuring the fluorescence of di(acetoxymethylester)6-carboxy-2',7'-dichlorodihydrofluorescein diacetate on a Glomax plate reader (Promega, Madison, WI). The fluorescence intensity reflecting ROS production was normalized as expressed in arbitrary units.
      For measurement of ROS production by flow cytometry, human dermal fibroblasts were seeded on a 24-well plate. The next day, fibroblasts were labeled for 45 minutes with 2 μM of di(acetoxymethylester)6-carboxy-2',7'-dichlorodihydrofluorescein diacetate. Cells were washed with PBS and incubated for 6 hours in basal medium and harvested. Unlabeled cells were used as negative controls to adjust the positivity of the cells in each experiment. Cell fluorescence was analyzed with a flow cytometer (BD FACSCanto II, BD Biosciences, Franklin Lakes, NJ) in the channel FL1.

      Animal model of systemic sclerosis

      Mice were the same as previously described (
      • Kaspi E.
      • Heim X.
      • Granel B.
      • Guillet B.
      • Stalin J.
      • Nollet M.
      • et al.
      Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
      ). As a reminder, Cd146-KO mice (
      • Jouve N.
      • Bachelier R.
      • Despoix N.
      • Blin M.G.
      • Matinzadeh M.K.
      • Poitevin S.
      • et al.
      CD146 mediates VEGF-induced melanoma cell extravasation through FAK activation: Host CD146 mediates melanoma metastasis.
      ) aged between 11 and 18 weeks were treated with subcutaneous bleomycin and were compared with WT littermates of the same age and sex. Skin fibrosis was induced, according to a protocol modified by Ould-Ali et al. (2013), by local subcutaneous injections of 100 μl of bleomycin dissolved in 0.9% sodium chloride at a concentration of 10 μg/ml (
      • Yamamoto T.
      • Nishioka K.
      Cellular and molecular mechanisms of bleomycin-induced murine scleroderma: current update and future perspective.
      ) every other day for 28 days. Bleomycin was injected in defined areas on the right upper back of Cd146-KO and WT mice. Vehicle (sodium chloride) was injected in defined areas on the left upper back of Cd146-KO and WT mice and was used as controls for the bleomycin experiment. On day 28, mice were killed, and the skin was fixed in 4% formalin and then embedded in paraffin. The protocol received approval from the committee of Aix Marseille University for Animal Care and Use, and experiments were performed in an authorized laboratory and under the supervision of an authorized researcher (GUILLET 13-328).

      Histological analysis and immunostaining of murine skin biopsy

      Tissues were fixed in 4% formalin and embedded into paraffin. Several 5-μm sections were stained with Masson’s trichrome or were used for immunohistochemistry. Masson’s trichrome staining was performed on four WT mice versus five Cd146-KO mice. Dermal thickness was quantitatively evaluated using ImageJ software (National Institutes of Health, Bethesda, MD) (six measurements per section). Manual immunohistochemistry was performed with avidin-biotin-peroxidase complex (Life Technologies, Carlsbad, CA) using the anti‒8-oxo-2'-deoxyguanosine antibody (sc-66036) diluted by 1:50 (Santa Cruz Biotechnology, Dallas, TX) on skin biopsy on seven WT versus seven Cd146-KO mice after vehicle or bleomycin injection.
      The percentage of positive nuclei/cells was manually evaluated. Blind counts were performed by two distinct operators. Nuclei were counted on 100 cells on two different locations of each section. For each condition, seven sections were counted.
      Automatic immunohistochemistry was performed on Ventana Benchmark system using the anti‒α-smooth muscle actin antibody from Thermo Fisher Scientific (Waltham, MA) (MS-113-R7, ready to use) on skin biopsy from three WT and four KO mice after vehicle or bleomycin injection. Automatic quantification of immunostaining intensity was performed using ImageJ software. In detail, RGB (red, green, blue) images were converted to 8 bits 256 levels, a mask was applied on pictures to select only concerned tissue, and the mean intensity was measured after background subtraction. For each section, four areas were analyzed in the derm.

      Western blot experiments

      Experiments were performed as previously described (
      • Harhouri K.
      • Kebir A.
      • Guillet B.
      • Foucault-Bertaud A.
      • Voytenko S.
      • Piercecchi-Marti M.D.
      • et al.
      Soluble CD146 displays angiogenic properties and promotes neovascularization in experimental hind-limb ischemia.
      ). Cells were lysed in 150 μl of nondenaturing lysis buffer (containing 10 mM Tris, pH 8; 1 mM EDTA, pH8; 150 mM sodium chloride; and 1% NP40) to which we added protease and phosphatase inhibitors (numbers 88665 and 88667, Thermo Fisher Scientific), which we recovered and placed in tubes. After 10 minutes in the ice, samples were centrifuged for 10 minutes at 10,000g at a temperature of 4 °C. Finally, we recovered the supernatant containing the proteins of cell lysates. Protein concentration was evaluated with a Pierce BCA Protein Assay Kit (number 23225, Thermo Fisher Scientific). Membranes were probed with specific primary antibodies: anti-WNT3A (ab28472), anti-WIF1 (ab186845), and anti-CD146 (ab75769) diluted 1:1,000 (Abcam, Cambridge, United Kingdom); anti‒β-catenin (number 9562), anti‒beta-actin (number 8457), and anti-LC3B (number 2775) diluted at 1:1,000, 1:5,000, and 1:1,000, respectively (Cell Signaling Technology, Danvers, MA); anti-WNT5A (sc-365370) diluted at 1:200 (Santa Cruz Biotechnology); antimonoubiquitinylated and polyubiquitinylated diluted 1:1,000 (FK2 clone, Enzo Biochem, Farmingdale, NY); and anti-LC3A/LC3B (number PA5-109226) diluted 1:1,000 (Thermo Fisher Scientific), followed by hybridization (incubation) with secondary antibodies coupled to peroxidase. Blots were revealed with ECL substrate (Pierce). Quantifications of band intensities on immunoblots were performed with GeneTools GBoxImageQuant software (Syngene International, Bangalore, India) and normalized by actin intensity.
      Western blot analysis was performed as described earlier with a mouse antimonoubiquitinylated and polyubiquitinylated conjugates mAb (FK2, used at 1:1,000 for the western blot analyses, Enzo Biochem).

      Flow cytometry analysis

      Human fibroblasts were incubated for 48 hours in basal medium ± 10 μg/ml bleomycin or 200 μM hydrogen peroxide, harvested, fixed with 2% paraformaldehyde, and labeled in permeabilization buffer with either anti-CD146 antibody coupled with phycoerythrin or with anti-WNT5A antibody, followed by incubation with a secondary antibody coupled with Alexa 488 (Thermo Fisher Scientific). Cell fluorescence was analyzed with a flow cytometer (BD FACSCanto II, BD Biosciences).

      RT-qPCR

      Total RNA was extracted with the RNeasy Mini Kit (Qiagen, Hilden, Germany). A total of 1 μg of total RNA was used for reverse transcription to cDNA using a high-capacity cDNA synthesis kit (Archive, Applied Biosystems, Foster, CA). Real-time PCR amplification was performed using an RT2 Profiler PCR array specific for Wnt signaling pathway components (number 330231 PAMM-043ZA, Qiagen) or Taqman Fast Advanced Master Mix predesigned primers (CD146: mm00522397_m1, WNT5A: mm00437347_m1, RPL13: mm02526700-g1; Applied Biosystems) for bleomycin-induced WNT5A and CD146 expression on a StepOne Plus thermocycler Real-Time PCR System (Applied Biosystems). The cycling conditions were as follows: 2 minutes at 50 °C, 2 minutes at 95 °C, 40 cycles of 1 second at 95 °C, and 20 seconds at 60 °C. Relative gene expression levels were normalized to that of Rpl13 or Gapdh and quantified using the ΔΔCT method.

      MEFs transfection

      Briefly, cells were transiently transfected with a plasmid coding for CD146 long isoform (PND28), small interfering RNA inhibiting CD146 (HSS106379, Thermo Fisher Scientific), or control-targeted small interfering RNA (number 6568, Cell Signaling Technology) using jetPRIME transfection agent (Polyplus transfection, Illkirch-Graffenstaden, France) according to the manufacturer’s instructions.

      Cycloheximide chase analysis

      Protein biosynthesis was inhibited by adding 20 μg/ml cycloheximide (Sigma-Aldrich, Vienna, Austria) to the medium of the MEF for 6, 12, and 24 hours. Cells were treated simultaneously with bleomycin (0, 10 μg/ml) for the same time. Then cells were rinsed and lysed for western blot analysis.

      Inhibition of proteasome and activation of autophagy

      MEFs from WT and KO mice were cultured in a culture medium containing MG132 (number 474790, MiliporeSigma, Burlington, MA) at a final concentration of 5 μM or with 10 μg/ml bleomycin for 24 hours. Then, cells were rinsed and lysed for western blot analysis.
      Figure thumbnail fx1
      Supplementary Figure S1Proteasome blockade increases WNT5A degradation by autophagy. WT and KO MEFs were treated with the proteasome inhibitor MG132 (5 μM) and/or bleomycin (10 μg/ml) for 24 hours. (a) Protein ubiquitination was observed by western blot. (b, c) WNT5A and LC3B protein expression were studied by western blot. Representative western blot (left) and relative quantification (right) are shown. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data represent the mean ± SEM. All experiments were performed in triplicate. KO, knockout; MEF, mouse embryonic fibroblast; WT, wild type.
      Figure thumbnail fx2
      Supplementary Figure S2Fibroblasts from patients with SSc display more ROS than fibroblasts from healthy donors. HDFC and HDFS were serum starved for 6 hours, and the generated ROS were evaluated by flow cytometry. Each column corresponds to a different batch. DAM-C-H2DCF-DA, di(acetoxymethylester)6-carboxy-2',7'-dichlorodihydrofluorescein diacetate; HDFC, control-derived human donor fibroblast; HDFS, systemic sclerosis‒derived human dermal fibroblast; Max, maximum; SSc, systemic sclerosis.
      Figure thumbnail fx3
      Supplementary Figure S3Fibroblasts from patients with SSc express CD146, whereas fibroblasts from healthy donors do not express CD146. The expression level of CD146 on human dermal fibroblasts from healthy donors and donors with SSc was assessed by flow cytometry. Each column corresponds to a different batch. HDFC, control-derived human donor fibroblast; HDFS, systemic sclerosis‒derived human dermal fibroblast; Max, maximum; SSc, systemic sclerosis.
      Supplementary Table S1Clinical and Biological Data of Patients with SSc from Whom Fibroblasts Were Collected
      PatientSexAge at DiagnosisCutaneous FormUlcereRenal CrisisInterstitial Lung DiseasePulmonary HypertensionAntinuclear AutoantibodySpecificity
      1F53LimitedYesNoYesYes>1,280RNA polymerase III
      2F47DiffuseNoNoNoNo>1,280No specificity
      3F36DiffuseYesNoNoNo>1,280Topoisomerase I
      4F56LimitedNoNoNoNo>1,280Centromer
      5F53LimitedNoNoYesYes>1,280Topoisomerase I
      Abbreviations: F, female; SSc, systemic sclerosis.

      References

        • Aceto G.M.
        • Catalano T.
        • Curia M.C.
        Molecular aspects of colorectal adenomas: the interplay among microenvironment, oxidative stress, and predisposition.
        Biomed Res Int. 2020; 20201726309
        • Andersson-Sjöland A.
        • Karlsson J.C.
        • Rydell-Törmänen K.
        ROS-induced endothelial stress contributes to pulmonary fibrosis through pericytes and Wnt signaling.
        Lab Invest. 2016; 96: 206-217
        • Bardin N.
        • Anfosso F.
        • Massé J.M.
        • Cramer E.
        • Sabatier F.
        • Le Bivic A.
        • et al.
        Identification of CD146 as a component of the endothelial junction involved in the control of cell-cell cohesion.
        Blood. 2001; 98: 3677-3684
        • Bergmann C.
        • Distler J.H.W.
        Canonical Wnt signaling in systemic sclerosis.
        Lab Invest. 2016; 96: 151-155
        • Beyer C.
        • Schramm A.
        • Akhmetshina A.
        • Dees C.
        • Kireva T.
        • Gelse K.
        • et al.
        β-catenin is a central mediator of pro-fibrotic Wnt signaling in systemic sclerosis.
        Ann Rheum Dis. 2012; 71: 761-767
        • Blasi A.
        • Martino C.
        • Balducci L.
        • Saldarelli M.
        • Soleti A.
        • Navone S.E.
        • et al.
        Dermal fibroblasts display similar phenotypic and differentiation capacity to fat-derived mesenchymal stem cells, but differ in anti-inflammatory and angiogenic potential.
        Vasc Cell. 2011; 3: 5
        • Boneberg E.M.
        • Illges H.
        • Legler D.F.
        • Fürstenberger G.
        Soluble CD146 is generated by ectodomain shedding of membrane CD146 in a calcium-induced, matrix metalloprotease-dependent process.
        Microvasc Res. 2009; 78: 325-331
        • Dagur P.K.
        • Biancotto A.
        • Wei L.
        • Sen H.N.
        • Yao M.
        • Strober W.
        • et al.
        MCAM-expressing CD4(+) T cells in peripheral blood secrete IL-17A and are significantly elevated in inflammatory autoimmune diseases.
        J Autoimmun. 2011; 37: 319-327
        • Dees C.
        • Schlottmann I.
        • Funke R.
        • Distler A.
        • Palumbo-Zerr K.
        • Zerr P.
        • et al.
        The Wnt antagonists DKK1 and SFRP1 are downregulated by promoter hypermethylation in systemic sclerosis.
        Ann Rheum Dis. 2014; 73: 1232-1239
        • Ding W.-X.
        • Ni H.-M.
        • Gao W.
        • Yoshimori T.
        • Stolz D.B.
        • Ron D.
        • et al.
        Linking of autophagy to ubiquitin-proteasome system is important for the regulation of endoplasmic reticulum stress and cell viability.
        Am J Pathol. 2007; 171: 513-524
        • Doridot L.
        • Jeljeli M.
        • Chêne C.
        • Batteux F.
        Implication of oxidative stress in the pathogenesis of systemic sclerosis via inflammation, autoimmunity and fibrosis.
        Redox Biol. 2019; 25: 101122
        • Dou L.
        • Jourde-Chiche N.
        • Faure V.
        • Cerini C.
        • Berland Y.
        • Dignat-George F.
        • et al.
        The uremic solute indoxyl sulfate induces oxidative stress in endothelial cells.
        J Thromb Haemost. 2007; 5: 1302-1308
        • Fels D.R.
        • Ye J.
        • Segan A.T.
        • Kridel S.J.
        • Spiotto M.
        • Olson M.
        • et al.
        Preferential cytotoxicity of bortezomib toward hypoxic tumor cells via overactivation of endoplasmic reticulum stress pathways.
        Cancer Res. 2008; 68: 9323-9330
        • Funato Y.
        • Michiue T.
        • Asashima M.
        • Miki H.
        The thioredoxin-related redox-regulating protein nucleoredoxin inhibits Wnt-beta-catenin signalling through dishevelled.
        Nat Cell Biol. 2006; 8: 501-508
        • Gabsi A.
        • Heim X.
        • Dlala A.
        • Gati A.
        • Sakhri H.
        • Abidi A.
        • et al.
        TH17 cells expressing CD146 are significantly increased in patients with Systemic sclerosis.
        Sci Rep. 2019; 9: 17721
        • Harhouri K.
        • Kebir A.
        • Guillet B.
        • Foucault-Bertaud A.
        • Voytenko S.
        • Piercecchi-Marti M.D.
        • et al.
        Soluble CD146 displays angiogenic properties and promotes neovascularization in experimental hind-limb ischemia.
        Blood. 2010; 115: 3843-3851
        • Harhouri K.
        • Navarro C.
        • Depetris D.
        • Mattei M.G.
        • Nissan X.
        • Cau P.
        • et al.
        MG132-induced progerin clearance is mediated by autophagy activation and splicing regulation.
        EMBO Mol Med. 2017; 9: 1294-1313
        • Hinz B.
        Formation and function of the myofibroblast during tissue repair.
        J Invest Dermatol. 2007; 127: 526-537
        • Ito T.
        • Tamura N.
        • Okuda S.
        • Tada K.
        • Matsushita M.
        • Yamaji K.
        • et al.
        Elevated serum levels of soluble CD146 in patients with systemic sclerosis.
        Clin Rheumatol. 2017; 36: 119-124
        • Jiang T.
        • Zhuang J.
        • Duan H.
        • Luo Y.
        • Zeng Q.
        • Fan K.
        • et al.
        CD146 is a coreceptor for VEGFR-2 in tumor angiogenesis.
        Blood. 2012; 120: 2330-2339
        • Kaspi E.
        • Heim X.
        • Granel B.
        • Guillet B.
        • Stalin J.
        • Nollet M.
        • et al.
        Identification of CD146 as a novel molecular actor involved in systemic sclerosis.
        J Allergy Clin Immunol. 2017; 140: 1448-1451.e6
        • Kato M.
        • Ospelt C.
        • Gay R.E.
        • Gay S.
        • Klein K.
        Dual role of autophagy in stress-induced cell death in rheumatoid arthritis synovial fibroblasts.
        Arthritis Rheumatol. 2014; 66: 40-48
        • Mellemkjaer L.
        • Pfeiffer R.M.
        • Engels E.A.
        • Gridley G.
        • Wheeler W.
        • Hemminki K.
        • et al.
        Autoimmune disease in individuals and close family members and susceptibility to non-Hodgkin’s lymphoma.
        Arthritis Rheum. 2008; 58: 657-666
        • Mostmans Y.
        • Cutolo M.
        • Giddelo C.
        • Decuman S.
        • Melsens K.
        • Declercq H.
        • et al.
        The role of endothelial cells in the vasculopathy of systemic sclerosis: a systematic review.
        Autoimmun Rev. 2017; 16: 774-786
        • Rosenthal A.K.
        • McLaughlin J.K.
        • Gridley G.
        • Nyrén O.
        Incidence of cancer among patients with systemic sclerosis.
        Cancer. 1995; 76: 910-914
        • Sambo P.
        • Baroni S.S.
        • Luchetti M.
        • Paroncini P.
        • Dusi S.
        • Orlandini G.
        • et al.
        Oxidative stress in scleroderma: maintenance of scleroderma fibroblast phenotype by the constitutive up-regulation of reactive oxygen species generation through the NADPH oxidase complex pathway.
        Arthritis Rheum. 2001; 44: 2653-2664
        • Servettaz A.
        • Goulvestre C.
        • Kavian N.
        • Nicco C.
        • Guilpain P.
        • Chéreau C.
        • et al.
        Selective oxidation of DNA topoisomerase 1 induces systemic sclerosis in the mouse.
        J Immunol. 2009; 182: 5855-5864
        • Smallwood M.J.
        • Nissim A.
        • Knight A.R.
        • Whiteman M.
        • Haigh R.
        • Winyard P.G.
        Oxidative stress in autoimmune rheumatic diseases.
        Free Radic Biol Med. 2018; 125: 3-14
        • Svegliati S.
        • Marrone G.
        • Pezone A.
        • Spadoni T.
        • Grieco A.
        • Moroncini G.
        • et al.
        Oxidative DNA damage induces the ATM-mediated transcriptional suppression of the Wnt inhibitor WIF-1 in systemic sclerosis and fibrosis.
        Sci Signal. 2014; 7: ra84
        • Thuan D.T.B.
        • Zayed H.
        • Eid A.H.
        • Abou-Saleh H.
        • Nasrallah G.K.
        • Mangoni A.A.
        • et al.
        A potential link between oxidative stress and endothelial-to-mesenchymal transition in systemic sclerosis.
        Front Immunol. 2018; 9: 1985
        • Wei J.
        • Fang F.
        • Lam A.P.
        • Sargent J.L.
        • Hamburg E.
        • Hinchcliff M.E.
        • et al.
        Wnt/β-catenin signaling is hyperactivated in systemic sclerosis and induces Smad-dependent fibrotic responses in mesenchymal cells.
        Arthritis Rheum. 2012; 64: 2734-2745
        • Xu J.
        Preparation, Culture, and Immortalization of Mouse Embryonic Fibroblasts.
        in: Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Curr. Protoc. Mol. Biol. John Wiley & Sons, Inc., Hoboken, NJ, USA2005
        • Yang L.
        • Serada S.
        • Fujimoto M.
        • Terao M.
        • Kotobuki Y.
        • Kitaba S.
        • et al.
        Periostin facilitates skin sclerosis via PI3K/Akt dependent mechanism in a mouse model of scleroderma.
        PLoS One. 2012; 7e41994
        • Ye Z.
        • Zhang C.
        • Tu T.
        • Sun M.
        • Liu D.
        • Lu D.
        • et al.
        Wnt5a uses CD146 as a receptor to regulate cell motility and convergent extension.
        Nat Commun. 2013; 4: 2803
        • Yu Y.
        • Guan X.
        • Nie L.
        • Liu Y.
        • He T.
        • Xiong J.
        • et al.
        DNA hypermethylation of sFRP5 contributes to indoxyl sulfate-induced renal fibrosis.
        J Mol Med (Berl). 2017; 95: 601-613
        • Zeineddine N.
        • Khoury L.E.
        • Mosak J.
        Systemic sclerosis and malignancy: a review of current data.
        J Clin Med Res. 2016; 8: 625-632