Identification of CD146 as a novel molecular actor involved in systemic sclerosis

Abstract

Systemic sclerosis (SSc) is a fibrosing autoimmune disorder characterized by vascular damage, excessive fibrosis in the skin and internal organs, and immune dysfunction.1Gabrielli A. Avvedimento E.V. Krieg T. Scleroderma.N Engl J Med. 2009; 360: 1989-2003Crossref PubMed Scopus (1144) Google Scholar Pulmonary fibrosis and pulmonary arterial hypertension are the most serious complications and currently constitute the major causes of death. The identification of biomarkers for the presence and progression of clinical complications in SSc would help to clarify pathogenic mechanisms, assess disease activity, and offer new hopes for treatment. CD146, an adhesion molecule located essentially in the vascular system, is also present as a soluble form (sCD146) in the bloodstream.2Bardin N. Moal V. Anfosso F. Daniel L. Brunet P. Sampol J. et al.Soluble CD146, a novel endothelial marker, is increased in physiopathological settings linked to endothelial junctional alteration.Thromb Haemost. 2003; 90: 915-920Crossref PubMed Google Scholar Because of its role in endothelial integrity, angiogenesis, and inflammatory response,3Bardin N. Blot-Chabaud M. Despoix N. Kebir A. Harhouri K. Arsanto J.-P. et al.CD146 and its soluble form regulate monocyte transendothelial migration.Arterioscler Thromb Vasc Biol. 2009; 29: 746-753Crossref PubMed Scopus (97) Google Scholar processes modulated in SSc, we hypothesized that CD146/sCD146 could represent a new molecular actor involved in SSc pathogenesis. Sera from 50 patients with SSc admitted to internal medicine departments in Marseille were analyzed (see Table E1 in this article's Online Repository at www.jacionline.org). All patients fulfilled the 2013 ACR/EULAR Classification Criteria for Scleroderma4Van den Hoogen F. Khanna D. Fransen J. Johnson S.R. Baron M. Tyndall A. et al.2013 classification criteria for systemic sclerosis: an American College of Rheumatology/European League against Rheumatism collaborative initiative.Arthritis Rheum. 2013; 65: 2737-2747Crossref PubMed Scopus (1959) Google Scholar and were then subclassified according to LeRoy et al5LeRoy E.C. Medsger T.A. Criteria for the classification of early systemic sclerosis.J Rheumatol. 2001; 28: 1573-1576PubMed Google Scholar criteria. All samples came from a declared Biobank (DC 2012-1704) with respect to ethical directives. Soluble CD146 levels were significantly higher in patients with SSc than in healthy controls (867.1 ± 32.3 vs 497.7 ± 16.5; P < .001) (Fig 1, A). Regarding clinical association, we found a significant association between low sCD146 levels and digital gangrene (n = 5; P = .03) or pulmonary fibrosis (n = 20; P = .04), 2 severe manifestations of the disease, whereas no significant association was found between sCD146 levels and Raynaud phenomenon (n = 29), or digital scars/ulcers (n = 16) (Fig 1, B and C). Sera available 12 months later for 15 patients showed a significant association between the diminution of sCD146 levels and aggravation of the disease attested by digital necrosis, pulmonary arterial hypertension, and pericarditis (n = 15; P = .03) (Fig 1, D). To investigate whether sCD146 was active in the serum of patients, we examined its effect on endothelial cell proliferation. Results showed a significant correlation between seric CD146 levels and cell proliferation (P = .02) and a significant decrease in cell proliferation when sCD146 was immunodepleted (P < .0001) (see Fig E1 in this article's Online Repository at www.jacionline.org), demonstrating that sCD146 is biologically active. To further examine the impact of CD146, we used the bleomycin-induced skin fibrosis model in wild-type (WT) and CD146-deficient mice. At 1 μg of bleomycin per mouse, no modification was obtained in dermal thickness under bleomycin or control conditions in WT mice (P = .64; Fig 2, A, B, and G). In contrast in CD146 knockout (KO) mice, dermal thickness was significantly higher after bleomycin treatment than in control conditions (P = .0002; Fig 2, C, D, and G), showing that this concentration was sufficient to induce skin fibrosis in CD146KO mice but not in WT mice. In the same way, epidermal thickness was also increased after bleomycin treatment in comparison to control conditions in CD146KO mice (P = .04), whereas no effect of bleomycin treatment was observed in WT mice (P = .63; Fig 2, H). Altogether these results show a higher susceptibility of CD146KO mice to develop skin fibrosis after bleomycin injection. Interestingly, injection of sCD146 in bleomycin-treated CD146KO mice showed a significant reduction in dermal and epidermal thickness as compared with bleomycin condition (P = .003 and P < .0001, respectively; Fig 2, E-H), suggesting that sCD146 protects bleomycin-treated CD146KO mice from fibrosis development. Because the Wnt pathway is involved in SSc fibrogenesis6Shi J. Chi S. Xue J. Yang J. Li F. Liu X. Emerging role and therapeutic implication of Wnt signaling pathways in autoimmune diseases.J Immunol Res. 2016; 2016: 1-18Crossref Scopus (6) Google Scholar and because interaction between CD146 and Wnt5a7Ye 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: 2803Crossref PubMed Scopus (53) Google Scholar was recently described, we investigated the effects of CD146. Immunochemistry experiments showed that Wnt5a expression was downregulated in KO compared with WT mice (see Fig E3, A, in this article's Online Repository at www.jacionline.org). Results were confirmed on mouse embryonic fibroblasts (MEFs) from WT and KO mice (Fig E3, B). In addition, Wnt5a induced Jnk phosphorylation in WT MEFs but not in KO MEFs (Fig E3, C), suggesting a downregulation of the noncanonical Wnt signaling in KO mice. In contrast, activation of canonical Wnt pathway with Wnt3a was evidenced in KO MEFs but not in WT MEFs (Fig E3, D) (see Table E1, Methods section, and Figs E1, E2, and E3 in this article's Online Repository at www.jacionline.org). Up to date, SSc-specific autoantibodies and cutaneous subclassification are the most useful biomarkers for diagnosis and predicting clinical features. If several biomarkers such as ICAM-1 or E-selectin were proposed with a potential prognostic interest,8Van Bon L. Affandi A.J. Broen J. Christmann R.B. Marijnissen R.J. Stawski L. et al.Proteome-wide analysis and CXCL4 as a biomarker in systemic sclerosis.N Engl J Med. 2014; 370: 433-443Crossref PubMed Scopus (285) Google Scholar none of them is used in clinical practice. Our results highlighted the role of CD146/sCD146 in SSc because (1) sCD146 levels were increased in patients with SSc compared with controls, (2) severe manifestations of the disease, digital gangrene and pulmonary fibrosis, were associated with low sCD146 levels in patients with SSc, (3) mice lacking CD146 were highly susceptible to develop skin fibrosis, (4) in this KO model, skin fibrosis development can be prevented by sCD146 injection, and (5) CD146/sCD146 is involved in the Wnt pathway. Of interest, patients producing high and sustained levels of sCD146 may be protected against the disease complications. Up to now, sCD146 has been mainly described in human serum as an endothelial biomarker modulated in inflammatory diseases. In inflammatory bowel diseases, we found an association between sCD146 levels and disease activity.9Reumaux D. Bardin N. Colombel J.-F. Dignat-George F. Duthilleul P. Vermeire S. Restoration of soluble CD146 in patients with Crohn's disease treated with the TNF-alpha antagonist infliximab.Inflamm Bowel Dis. 2007; 13: 1315-1317Crossref PubMed Scopus (8) Google Scholar Indeed, low level of sCD146 was associated with the active form of the disease and an upregulation of sCD146 occurred after anti-TNF treatment and a favorable disease evolution.9Reumaux D. Bardin N. Colombel J.-F. Dignat-George F. Duthilleul P. Vermeire S. Restoration of soluble CD146 in patients with Crohn's disease treated with the TNF-alpha antagonist infliximab.Inflamm Bowel Dis. 2007; 13: 1315-1317Crossref PubMed Scopus (8) Google Scholar In the present study, unfavorable evolution of SSc was associated with a decrease in sCD146 levels. In addition, vascular ischemic complication (digital gangrene) and pulmonary fibrosis were associated with low sCD146 levels. Although a prospective study is necessary to confirm our conclusion, high sCD146 levels could constitute a good prognostic marker of the disease. Along this line, we showed that CD146KO mice were strongly sensitive to bleomycin-induced skin fibrosis as compared with WT mice. The proof that this effect is CD146-dependent is given by the absence of fibrosis development after sCD146 injection in this animal model. Finally, we showed that CD146/CD146s modulates Wnt signaling. CD146 deficiency is associated with down- and upregulation of the noncanonical and the canonical Wnt pathway, respectively, leading to a profibrotic state. In conclusion, our data led us to propose CD146/sCD146 as a novel biomarker in SSc for the assessment of the disease activity. Further studies will be conducted to confirm the role of CD146/sCD146 in the fibrotic process, to get more insights into the Wnt pathway, and to investigate the potential interest of the molecule for therapeutic applications. We thank Joëlle Fiteni, Nathalie Boitano, and Corinne Derrien for their help and continuous support. Sera from 50 patients with SSc admitted to internal medicine departments in Marseille (South of France) were analyzed. All patients fulfilled the 2013 ACR/EULAR Classification Criteria for SclerodermaE1LeRoy E.C. Medsger T.A. Criteria for the classification of early systemic sclerosis.J Rheumatol. 2001; 28: 1573-1576PubMed Google Scholar and were then subclassified according to LeRoy et alE2Shi J. Chi S. Xue J. Yang J. Li F. Liu X. Emerging role and therapeutic implication of Wnt signaling pathways in autoimmune diseases.J Immunol Res. 2016; 2016: 1-18Crossref Scopus (74) Google Scholar criteria. For 15 patients, sera were available 12 months later. As controls, we studied the sera of 50 age- and sex-matched blood donors. All samples came from a declared Biobank (DC 2012-1704) with respect to ethical directives and were stored to −80°C before use. A capture ELISA (CY-QUANT ELISA) was used to determine the levels of sCD146 (Biocytex, Marseille, France), according to the manufacturer's protocol. Coefficients of variation of repeatability, reproducibility lot to lot, and reproducibility within batch were respectively 8.2%, 2.9%, and 3.5%. Bioactivity of seric sCD146 in patients with SSc was determined using cell proliferation experiments performed on endothelial progenitor cells (EPCs) previously incubated with sera of patients with SSc with high (>734 ng/mL; n = 24) or low (<734 ng/mL; n = 17) sCD146 levels, or with sCD146-immunodepleted samples (n = 21). In 3 independent experiments, sCD146 was removed from the plasma (n = 21). First, sera containing high levels of sCD146 (>734 ng/mL) were incubated with S-Endo-1 antibody, and tubes were loaded on a test tube rotator and agitated overnight at +4°C. Then, the immune complex was covalently immobilized on protein-A Sepharose beads. The unabsorbed fraction was recovered by centrifugation at 1500 rpm for 10 minutes, and was used in the following tests. Human umbilical cord blood samples from donors were collected, in compliance with French legislation, in a heparinized tube. Mononuclear cells (MNCs) were isolated by density gradient centrifugation as previously described.E3Bompais H. Chagraoui J. Canron X. Crisan M. Liu X.H. Anjo A. et al.Human endothelial cells derived from circulating progenitors display specific functional properties compared with mature vessel wall endothelial cells.Blood. 2004; 103: 2577-2584Crossref PubMed Scopus (230) Google Scholar EPCs were seeded on 96-well plates (7.103/well) and starved in RPMI-10% FCS overnight. Cells were then incubated in EBM2 medium with 1% of serum for 24 hours. Cell proliferation was assayed by 5-bromo-2’-deoxy-uridine (BrdU) incorporation into cellular DNA using the BrdU Labeling and Detection Kit III (Roche Diagnostics, Mannheim, Germany). In brief, cells were incubated for 15 hours with BrdU labeling solution in EBM-2 medium. Cellular DNA was partially digested by nuclease treatment and incorporated BrdU was detected with a peroxidase-conjugated anti-BrdU primary antibody. The absorbance was measured at 450 nm using an Opsys MR microplate reader (Dynex Technologies, Chantilly, Va). Results were expressed as arbitrary units. CD146 KO miceE4Jouve 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.Int J Cancer. 2015; 137: 50-60Crossref PubMed Scopus (40) Google Scholar aged between 11 and 18 weeks were treated with subcutaneous bleomycin and were compared with littermates of the same age and sex WT mice. Skin fibrosis was induced, according to protocol modified from Ould-Ali et alE5Ould-Ali D. Hautier A. Andrac-Meyer L. Bardin N. Magalon G. Granel B. Bleomycin-induced scleroderma in nude mice can be reversed by injection of adipose tissue: evidence for a novel therapeutic intervention in systemic sclerosis.J Clin Exp Dermatol Res. 2013; (Available at: http://www.omicsonline.org/2155-9554/2155-9554-3-164.digital/2155-9554-3-164.html. doi: http://dx.doi.org/10.4172/2155-9554.1000164)Google Scholar by local subcutaneous injections of 100 μL of bleomycin dissolved in 0.9% NaCl, at a concentration of 10 μg/mL,E6Yamamoto T. Nishioka K. Cellular and molecular mechanisms of bleomycin-induced murine scleroderma: current update and future perspective.Exp Dermatol. 2005; 14: 81-95Crossref PubMed Scopus (71) Google Scholar every other day for 28 days. sCD146, dissolved in PBS, was also injected at a concentration of 100 μg/mL, every other day, jointly with bleomycin. Bleomycin with or without sCD146 was injected in defined areas on the right upper back of CD146 KO (10 and 3 mice, respectively) and WT (11 and 3 mice, respectively) mice. Vehicle (NaCl ± PBS) was injected in defined areas on the left upper back of CD146 KO and WT mice and was used as controls for the bleomycin ± sCD146 experiment, respectively. After approval of the protocol by the institution's Animal Care and Use Committee, the experiments were performed in an authorized laboratory and under the supervision of an authorized researcher (GUILLET 13-328). Murine skin sections were fixed in 4% formalin and embedded into paraffin. Several 7-μm sections were stained with hematoxylin-eosin stain. Dermal and epidermal thickness was quantitatively evaluated using a stereological method, using a Merz gridE7Massacrier A. Nègre-Aminou P. Couraud F. Cau P. Quantitative analysis of fetal rat brain neurons developing in primary cultures, I: stereological study of the neuronal differentiation.Brain Res. 1988; 468: 161-170Crossref PubMed Scopus (6) Google Scholar, E8Weibel E.R. Stereological methods. Vol 1: Practical methods for biological morphometry. Academic, London1979: 415Google Scholar (Fig E2). The grid is composed of a wavy array of lines (108 tests points, 40-mm spacing, total length of test lines, 6785.8 mm).E7Massacrier A. Nègre-Aminou P. Couraud F. Cau P. Quantitative analysis of fetal rat brain neurons developing in primary cultures, I: stereological study of the neuronal differentiation.Brain Res. 1988; 468: 161-170Crossref PubMed Scopus (6) Google Scholar Pictures of murine skin sections were taken, using Leica microscope (magnification ×10) and the grid was surimposed on each picture. The volume density of derm or epiderm was estimated by counting the points of the Merz grid located on the derm or epiderm. The surface area was estimated by counting the number of intersections of the sinous line with the border of each tissue. The dermal and epidermal thickness was thus quantitatively calculated using the volume density/surface area ratio. Impact of bleomycin (1 μg/mouse) and/or sCD146 injection (10 μg/mouse) on dermal and epidermal thickness was evaluated after normalization with corresponding control condition (vehicle). Immunohistochemistry was performed with avidin–biotin-peroxidase complex, using the Wnt5a antibody (anti-Wnt5a from Santa-Cruz (Dallas, Tex) [sc365370] diluted 1/200) on skin biopsy from WT and KO mice after vehicle (NaCl ± PBS) injection. A semiquantitative evaluation of Wnt5a expression was performed using a 4-point scale: − indicates absence; +, low expression; ++, intermediate expression; and +++, high expression. A blind reading of the staining was performed. MEFs were isolated from day-13.5 embryos from CD146 KO mice or WT mice. MEF isolation has been previously described.E9J Xu Preparation, culture, and immortalization of mouse embryonic fibroblasts.Curr Protoc Mol Biol. 2005; (Chapter 28:Unit 28.1)Google Scholar Briefly, the placental membranes, 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 18-G needle and a 100-mm strainer to remove large fragments, and placed in a 25-cm2 flask containing Dulbecco modified Eagle medium (GIBCO, ThermoFisher Scientific, Illkirch, France), 10% (v/v) FBS (GIBCO), 50 units/mL penicillin, and 50 mg/mL streptomycin. At this and subsequent stages of culture, cells were maintained in 5% oxygen. At confluency, the cells were transferred to a 75-cm2 flask. Western blot analysis was performed as previously described.E10Harhouri 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: 3843-3851Crossref PubMed Scopus (65) Google Scholar Membranes were probed with specific primary antibodies (anti-Wnt5a [sc-365370] diluted 1/200 [Santa-Cruz], anti-SAPK/JNK [#9252] diluted 1/1000, and anti-phospho-SAPK/JNK [#9251] diluted 1/1000 [Cell Signalling Technology, Leiden, The Netherlands]) followed by secondary antibodies coupled to peroxidase. Blots were revealed with the ECL substrate (Pierce, ThermoFisher Scientific, Illkirch, France). Canonical Wnt signaling was monitored by analysis of β-catenin/TCF transcription activity, as determined by luciferase reporter assay. Briefly, MEFs from WT and CD146 KO mice were seeded in 96-well tissue culture dishes at an initial density of 104 cells/well. After 18 to 24 hours, cells were transiently transfected with the Super8XTOPFlash reporter and pBARLS lentiviral reporter expression system containing 12 tandem repeats of TCF-binding sites,E11Biechele T.L. Moon R.T. Assaying beta-catenin/TCF transcription with beta-catenin/TCF transcription-based reporter constructs.Methods Mol Biol. 2008; 468: 99-110Crossref PubMed Scopus (85) Google Scholar using Fugene 6 (cat. no. E2691, Promega, Madison, Wis) and according to the manufacturing protocol. Forty-eight hours after transfection, cells were switched to serum-free medium for 6 hours, and then treated using Wnt3a-conditioned medium or control medium (control condition). Cell lysates were prepared with reporter lysis buffer from the Luciferase Assay System kit (cat. no. E1910, Promega) according to the protocol. Luciferase activity was normalized for transfection efficiency using Renilla luciferase values. The assay was done in quadruplicate. Results were then normalized in comparison to control condition (control medium). Statistical analysis was performed with the Prism software (GraphPad Software, Inc, San Diego, Calif). Significant differences were determined using parametric t test or nonparametric Mann Whitney test. A P value of less than .05 was considered significant.Fig E2Merz grid. The grid consists of a square that delimits a test area containing a system of 108 points marked on a sinuous line formed by the succession of 9 chained semicircles. The thickness was thus calculated using the volume density/surface area ratio.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3CD146/sCD146 modulates Wnt signaling. A, Semiquantification of Wnt5a expression was performed on 16 skin biopsies from WT or KO mice, showing a significant downregulation of Wnt5a expression in KO mice in comparison to WT mice (P < .05). B, Reduced Wnt5a expression on MEFs from KO mice compared with WT (Western blot experiments). C, Wnt5a induces Jnk phosphorylation in WT but not in KO mice (Western blot experiment). D, Activation of canonical Wnt signaling (β-catenin/TCF transcription activity; Luciferase assay) was observed in MEFs from KO mice after stimulation with Wnt3a-conditioned medium, but not in MEFs from WT mice. Results were normalized in comparison to control condition (control medium) (P < .05).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table E1Demographic data and disease characteristics of patients with SScPatients with SSc (n = 50)(25%-75% percentile)Median age (y)62 (56-71)Median disease duration (y)13 (10-20)Male/female06/44Clinical featuresn (%)Diffuse cutaneous form∗According to the classification of LeRoy et al.E123 (46)Ulcers, scars, gangrene21 (42)Pulmonary fibrosis†Pulmonary fibrosis was defined on HRCT imaging (when available) and pulmonary function tests.20 (40)Pulmonary arterial hypertension‡Pulmonary arterial hypertension was detected by echocardiography and confirmed by right heart catheterism if necessary.7 (14)Digestive involvement39 (78)Kidney involvement5 (10)Immunologic resultsn (%)Positive for antinuclear antibodies50 (100)Centromeric staining (ACA)18 (36)Homogeneous nucleoplasmic and nucleolar staining15 (30)Nucleolar staining7 (14)Homogeneous and speckled nucleoplasmic staining8 (16)Speckled nucleoplasmic staining2 (4)Positive for anti–topoisomerase I (anti–topo I)25 (50)ACA, Anticentromeric autoantibodies; HRCT, high-resolution computed tomography.∗ According to the classification of LeRoy et al.E1LeRoy E.C. Medsger T.A. Criteria for the classification of early systemic sclerosis.J Rheumatol. 2001; 28: 1573-1576PubMed Google Scholar† Pulmonary fibrosis was defined on HRCT imaging (when available) and pulmonary function tests.‡ Pulmonary arterial hypertension was detected by echocardiography and confirmed by right heart catheterism if necessary. Open table in a new tab ACA, Anticentromeric autoantibodies; HRCT, high-resolution computed tomography.