Advertisement

Decreased CCN3 in Systemic Sclerosis Endothelial Cells Contributes to Impaired Angiogenesis

Open ArchivePublished:January 16, 2020DOI:https://doi.org/10.1016/j.jid.2019.11.026
      Systemic sclerosis (SSc) is a rare and severe connective tissue disease combining autoimmune and vasculopathy features, ultimately leading to organ fibrosis. Impaired angiogenesis is an often silent and life-threatening complication of the disease. We hypothesize that CCN3, a member of the CCN family of extracellular matrix proteins, which is an antagonist of the profibrotic protein CCN2 as well as a proangiogenic factor, is implicated in SSc pathophysiology. We performed skin biopsies on 26 patients with SSc, both in fibrotic and nonfibrotic areas for 17 patients, and collected 18 healthy control skin specimens for immunohistochemistry and cell culture. Histological analysis of nonfibrotic and fibrotic SSc skin shows a systemic decrease of papillary dermis surface as well as disappearance of capillaries. CCN3 expression is systematically decreased in the dermis of patients with SSc compared with healthy controls, particularly in dermal blood vessels. Moreover, CCN3 is decreased in vitro in endothelial cells from patients with SSc. We show that CCN3 is essential for endothelial cell migration and angiogenesis in vitro. In conclusion, CCN3 may represent a promising therapeutic target for patients with SSc presenting with vascular involvement.

      Graphical abstract

      Abbreviations:

      Fib (fibrotic), HC (healthy control), HDMEC (human microvascular dermal endothelial cell), NFib (nonfibrotic), rCCN3 (recombinant CCN3), SSc (systemic sclerosis)

      Introduction

      Systemic sclerosis (SSc) is a rare and severe connective tissue disease combining autoimmunity and widespread vasculopathy, ultimately leading to organ fibrosis (
      • Allanore Y.
      • Simms R.
      • Distler O.
      • Trojanowska M.
      • Pope J.
      • Denton C.P.
      • et al.
      Systemic sclerosis.
      ,
      • Gabrielli A.
      • Avvedimento E.V.
      • Krieg T.
      Scleroderma.
      ,
      • Varga J.
      • Abraham D.
      Systemic sclerosis: a prototypic multisystem fibrotic disorder.
      ). Vasculopathy is a major issue in SSc (
      • Allanore Y.
      • Distler O.
      • Matucci-Cerinic M.
      • Denton C.P.
      Review: Defining a unified vascular phenotype in systemic sclerosis.
      ), and Raynaud’s phenomenon is usually the first manifestation of the disease. Digital ulcers are frequent in patients with SSc, responsible for serious morbidity such as gangrene and amputation, and associated with higher mortality (
      • Meunier P.
      • Dequidt L.
      • Barnetche T.
      • Lazaro E.
      • Duffau P.
      • Richez C.
      • et al.
      Increased risk of mortality in systemic sclerosis-associated digital ulcers: a systematic review and meta-analysis.
      ). Other vasculopathic manifestations such as pulmonary hypertension or renal crisis are silent at an early stage but are potentially lethal. A current challenge lies in discovering new therapeutic targets to treat these manifestations.
      CTGF, which is also known as CCN2, is a well-known profibrotic factor in SSc (
      • Leask A.
      • Denton C.P.
      • Abraham D.J.
      Insights into the molecular mechanism of chronic fibrosis: the role of connective tissue growth factor in scleroderma.
      ,
      • Serratì S.
      • Chillà A.
      • Laurenzana A.
      • Margheri F.
      • Giannoni E.
      • Magnelli L.
      • et al.
      Systemic sclerosis endothelial cells recruit and activate dermal fibroblasts by induction of a connective tissue growth factor (CCN2)/transforming growth factor β-dependent mesenchymal-to-mesenchymal transition.
      ). CCN proteins, a family of six matricellular proteins sharing a common multimodular structure, are involved in major cellular pathways of interest in SSc (
      • Henrot P.
      • Truchetet M.E.
      • Fisher G.
      • Taïeb A.
      • Cario M.
      CCN proteins as potential actionable targets in scleroderma.
      ). CCN3 is thought to be an antagonist of CCN2 when induced by transforming growth factor-β (
      • Lemaire R.
      • Farina G.
      • Bayle J.
      • Dimarzio M.
      • Pendergrass S.A.
      • Milano A.
      • et al.
      Antagonistic effect of the matricellular signaling protein CCN3 on TGF-beta- and Wnt-mediated fibrillinogenesis in systemic sclerosis and Marfan syndrome.
      ,
      • Madne T.H.
      • Dockrell M.E.C.
      CCN3, a key matricellular protein, distinctly inhibits TGFβ1-mediated Smad1/5/8 signalling in human podocyte culture.
      ,
      • Riser B.L.
      • Najmabadi F.
      • Perbal B.
      • Peterson D.R.
      • Rambow J.A.
      • Riser M.L.
      • et al.
      CCN3 (NOV) is a negative regulator of CCN2 (CTGF) and a novel endogenous inhibitor of the fibrotic pathway in an in vitro model of renal disease.
      ). However, CCN3 is also proangiogenic, probably by binding to cell surface integrins such as integrin α5β1, α6β1, and αvβ5 (
      • Lin C.G.
      • Chen C.C.
      • Leu S.J.
      • Grzeszkiewicz T.M.
      • Lau L.F.
      Integrin-dependent functions of the angiogenic inducer NOV (CCN3): implication in wound healing.
      ). CCN3 has been shown to be decreased in the placenta of pregnant women with pre-eclampsia, a disease whose pathophysiology mimics SSc renal crisis (
      • Gellhaus A.
      • Schmidt M.
      • Dunk C.
      • Lye S.J.
      • Kimmig R.
      • Winterhager E.
      Decreased expression of the angiogenic regulators CYR61 (CCN1) and NOV (CCN3) in human placenta is associated with pre-eclampsia.
      ).
      Thus, our hypothesis was that CCN3 could be a key player in SSc, particularly in vascular function. The aim of our work was to study the alteration of CCN3 in SSc skin and its role in SSc angiogenic dysfunction.

      Results

      SSc skin is characterized by a major decrease in papillary dermis

      Skin biopsies from 26 patients with SSc were compared with 18 healthy control (HC) samples. Clinical characteristics of patients and controls are shown in Table 1. The thickness of the superficial (papillary) dermis was significantly decreased both in nonfibrotic (NFib; clinically uninvolved) and fibrotic (Fib) areas of SSc skin compared with HC skin (NFib: 0.0527 ± 0.006 mm, Fib: 0.0595 ± 0.006 mm; HC: 0.122 ± 0.008 mm; P < 0.0001 in both cases; Figure 1a and b), even at the early stages of the disease (Figure 1a, recent SSc: <1 year of evolution after diagnosis). A significant decrease in the interdigitation index, representing a flattening of the basal membrane as compared with HC, was also noted both in nonfibrotic and fibrotic areas (HC: 1.76 ± 0.442; SSc NFib: 1.31 ± 0.264; SSc Fib: 1.35 ± 0.240; P = 0.0007 and 0.002, respectively; Supplementary Figure S1).
      Table 1Characteristics of Patients with SSc and Controls
      CharacteristicsPatients with SSc (n = 26)Healthy controls (n = 18)P-value
      Demographic characteristics
       Age, years, mean (SD)58.88 (9.45)47.29 (9.85)0.0005
       Female, n (%)15 (53.8)16 (88.9)0.1037
       DcSSc, n (%)10 (38)
       Disease duration, years, mean (SD)5.69 (5.46)
      Cutaneous involvement
       mRSS, mean (SD)16.5 (12)
       Pigmentary disorders, n (%)14 (54)
      Vascular involvement
       Digital ulcers, n (%)16 (62)
       PAH, n (%)4 (15)
       SSc renal crisis, n (%)1 (0.03)
      Pulmonary involvement
       ILD, n (%)11 (42)
       Pulmonary fibrosis, n (%)3 (12)
      Gastrointestinal involvement
       Esophageal involvement, n (%)19 (73)
       Intestinal involvement, n (%)3 (12)
      Musculoskeletal involvement
       Synovitis, n (%)8 (30)
       Joint contractures, n (%)4 (15)
      Abbreviations: DcSSc, diffuse cutaneous systemic sclerosis; ILD, interstitial lung disease; mRSS, modified Rodnan skin score; PAH, pulmonary arterial hypertension; SD, standard deviation; SSc, systemic sclerosis.
      Figure thumbnail gr1
      Figure 1Morphological changes in SSc dermis compared with HCs. Apparently nonfibrotic SSc skin (clinically uninvolved skin) displays histological changes similar to fibrotic (involved) skin. Data represent mean ± SEM and are normalized by dermis surface. Each dot represents a patient or an HC sample (mean of three experiments for each dot). HC: n = 18, SSc: n = 26 (both zones for 17 patients). (a) Masson’s Trichrome staining visualizing collagen fibers (blue). Top, HC arm; bottom, SSc arm (disease <1 year), NFib zone and Fib zone of the same patient (magnification ×20). White dots represent the junction between papillary (superficial) and reticular (deep) dermis. Bar = 50 μm. (b) Major decrease in papillary dermis thickness in patients with SSc, both in NFib and Fib zones (P < 0.0001). (c) Top, HC arm; bottom, SSc arm, NFib zone and Fib zone of the same patient (CD31/DAPI staining). White dots mark limit between epidermis and dermis. Magnification ×40. Bar = 50 μm. (d) Significant decrease in number of dermal vessels in patients with SSc (counted by CD31 staining). HC versus SSc Fib: P < 0.0001; HC versus SSc NFib: P < 0.0001. (e) Loss of small capillaries (papillary dermal vessels) in SSc skin. HC versus SSc Fib: P = 0.0008; HC versus SSc NFib: P = 0.0009. Fib, fibrotic; HC, healthy control; NFib, nonfibrotic; SEM, standard error of the mean; SSc, systemic sclerosis.
      CD31 immunostaining (Figure 1c) showed a significant decrease in the number of dermal blood vessels in SSc skin, both in nonfibrotic and fibrotic areas, compared with HC (NFib: 6.14 ± 0.927 dermal vessels/field, Fib: 6.66 ± 0.891 dermal vessels/field; HC: 23.5 ± 2.41 dermal vessels/field; P < 0.0001 in both cases; Figure 1d). This was at least partly because of the disappearance of papillary dermal vessels (mainly capillaries), as evidenced by the significant decrease in the proportion of capillaries in nonfibrotic and fibrotic SSc skin versus HC (NFib: 62.7 ± 0.0597%, Fib: 65 ± 0.0582%; HC: 87.8 ± 0.0203%; P = 0.0009 and P = 0.0008, respectively; Figure 1e).

      SSc skin is characterized by a decrease in CCN3 expression in dermal vessels

      Dermal vessels were then stained for CCN3 and CD34 (Figure 2a), the latter to evidence both endothelial cells (inner layer, white arrow) and perivascular CD34+ cells (outer layer, yellow arrows) surrounding the dermal vessels. In HC samples, several layers of CCN3+ cells were observed; the CD34+/CCN3+ inner layer was identified as the tunica intima, the middle layer as the tunica media thanks to α-smooth muscle actin staining (Supplementary Figure S2), and the outer layer as CD34+/CCN3+ perivascular cells (Figure 2a). Other perivascular CCN3+ cells were identified as lymphocytes (CD3+) and fibroblasts (CD90+) (Supplementary Figure S2).
      Figure thumbnail gr2
      Figure 2CCN3 expression in situ in the dermis of patients with SSc (n = 26) and HC (n = 18). CCN3 is decreased in SSc patients’ dermal vessels. Data represent mean ± SEM of a minimum of three independent experiments. Representative pictures are shown. (a) CCN3/CD34 immunofluorescent staining of HC and SSc skin. CD34+ cells are endothelial cells (inner layer, white arrows, red dots) and surrounding perivascular CD34+ cells (outer layer, yellow arrow, yellow dots). Two layers of CD34+/CCN3+ cells can be seen in HC skin (top), whereas a single layer of CD34+/CCN3+ cells can be seen in SSc skin (bottom). Bar = 20 μm. (b) Significant decrease in number of CCN3+ layers around dermal reticular vessels in patients with SSc (P < 0.0001, Mann-Whitney). (c) CCN3/CD31 immunofluorescent staining of HC and SSc skin. CD31+ cells are endothelial cells. CCN3 intensity is decreased in SSc CD31+ cells (bottom) compared with HC cells (top). Bar = 20 μm. (d) Major decrease in CCN3 expression in endothelial cells of SSc dermal vessels compared with HCs (P < 0.0001). HC versus SSc NFib or Fib: unpaired t-tests with Welch’s correction or Mann-Whitney when indicated; SSc NFib vs Fib: paired t-tests. Fib, fibrotic; HC, healthy control; MFI, mean fluorescence intensity; NFib, nonfibrotic; SEM, standard error of the mean; SSc, systemic sclerosis.
      We observed a marked decrease in the number of CD34/CCN3+ layers in SSc dermal vessels compared with HC (P < 0.0001, Figure 2b), especially in reticular dermal vessels, which was mainly attributed to the loss of CD34+ perivascular cells. The number of CCN3+ layers did not differ regardless of the SSc skin area, fibrotic or not (Supplementary Figure S3a).
      We observed a marked decrease in CCN3 intensity in SSc endothelial cells (identified by CD31 staining) (1.58 ± 0.117 arbitrary units) compared with HC endothelial cells (2.71 ± 0.175 arbitrary units; P < 0.0001, Figure 2c and d). Again, we observed no difference between nonfibrotic and fibrotic areas of SSc skin (Supplementary Figure S3b). This result, as a mean CCN3 fluorescence intensity of all CD31+ cells, was neither influenced by the number nor by the size of dermal vessels. CCN3 expression in SSc endothelial cells did not differ in subgroup analysis regarding vascular involvement of patients with SSc (Supplementary Figure S3c). The mean CCN3 expression in endothelial cells was significantly correlated with the number of vessels (Spearman r = 0.6227, P < 0.0001, Supplementary Figure S3d).
      We also observed a significant decrease in dermal cell density in SSc skin compared with HC skin (data not shown), which partially accounts for the lack of other CCN3+ cells in the dermis as shown in Figure 2c. Epidermal expression of CCN3 was not found to be different between HC and SSc, although great heterogeneity was observed among SSc samples (data not shown).

      CCN3 level is decreased in vitro and CCN3 is differentially localized in SSc endothelial cells compared with HC cells

      Cultured human microvascular dermal endothelial cells (HDMECs) from SSc biopsies and HC skin were checked for phenotype by CD31 labeling and flow cytometry (Figure 3a, Supplementary Figure S4a). All cell populations (HC [n = 7] and SSc [n = 6]) were more than 95% CD31+. Western blotting revealed a significant decrease in intracellular CCN3 in SSc HDMECs compared with HC ones (P = 0.04, Figure 3b and c). Quantitative PCR did not reveal any difference in CCN3 expression at the transcriptional level (Supplementary Figure S4b).
      Figure thumbnail gr3
      Figure 3CCN3 expression is decreased in vitro in cultured endothelial cells of patients with SSc compared with HCs. (a) Purity of HDMECs grown from SSc skin. Flow cytometry after staining for CD31 (allophycocyanin) shows 98.2% of selected CD31+ population. (b, c) Significant decrease in CCN3 expression in SSc endothelial cells (HDMECs) (n = 4) in vitro compared with HCs (n = 7) (western blot) (P = 0.0445, unpaired t-test with Welch’s correction). β-actin is shown as loading control. Results are mean of three experiments. Data represent mean ± SD. One representative experiment is shown. (d) Immunocytochemistry of CCN3 (green) in HC and SSc HDMECs. CCN3 can be found in cytoplasmic granules in both types of cells, but intranuclear staining is mostly seen in HC HDMECs and absent from SSc HDMECs. Representative pictures of two experiments involving six HCs and six patients with SSc are shown. Bar = 100 μm. Close-up: bar = 20 μm. (e) Decrease in secreted CCN3 of SSc endothelial cells (HDMECs) (n = 5) in vitro compared with HCs (n = 5) (assessed by ELISA) (P = 0.0159, Mann-Whitney). Data represent mean ± SD. Experiment was performed in triplicate. HC, healthy control; HDMEC, human microvascular dermal endothelial cell; SD, standard deviation; SN, supernatant; SSc, systemic sclerosis.
      CCN3 staining in HC HDMECs revealed both a cytoplasmic and nuclear expression, with a granulated perinuclear staining, suggesting an accumulation in the Golgi apparatus, as described elsewhere (
      • Perbal B.
      New insight into CCN3 interactions - Nuclear CCN3 : fact or fantasy?.
      ) (Figure 3d). Intranuclear CCN3 staining was almost absent in SSc HDMECs, suggesting an intracytoplasmic retention of the protein. Moreover, the level of secreted CCN3 measured by ELISA in cell supernatants was significantly decreased in SSc HDMECs compared with HC HDMECs (P = 0.01, Figure 3e). Secreted CCN3 was undetectable in four out of five SSc samples.
      Altogether, these findings demonstrate that CCN3 is downregulated in SSc endothelial cells at both the intracellular and secreted level.
      It is interesting to note that, although CCN3 basal secretion is very low in SSc HDMECs, it seems to be amenable to modulation. We have stimulated HC and SSc HDMECs with the proinflammatory cytokine IL-1β and found that CCN3 secretion tended to increase in SSc HDMECs only (Supplementary Figure S5).

      CCN3 is involved in HDMEC in vitro angiogenesis

      In vitro angiogenesis assays showed that incubation of HDMECs with anti-CCN3 antibody significantly impaired tube formation compared with control antibody (Figure 4a and b). We observed a significant decrease in the number of master junctions (P = 0.008), meshes (P = 0.01), and master segments (P = 0.01) (Figure 4c and d; Supplementary Figure S6b). A significant decrease in the total length of master segments (P = 0.01) and in the total segment length (P = 0.02) (Figure 4e; Supplementary Figure S6c) was also observed. The total length of isolated branches tended to increase in the presence of anti-CCN3 antibody compared with control antibody, without reaching statistical significance (Supplementary Figure S6d). Definition of each parameter is shown in Supplementary Figure S6a. No significant difference was observed between nontreated cells and HDMECs treated with control antibody (data not shown).
      Figure thumbnail gr4
      Figure 4CCN3 is involved in HDMEC in vitro angiogenesis. Each experiment was performed in quintuplicate. Data represent mean ± SEM of three independent experiments involving three different donors (for both HC and SSc cells). Representative pictures of one experiment are shown. Bar = 200 μm. Panels (a–e) represent CCN3 blockade in HC HDMECs; panels (f–j) represent CCN3 addition to SSc HDMECs. (a) Angiogenesis assay on matrigel of HC HDMECs treated with control antibody. Pictures taken 24 hours after beginning of experiment (left panel). Analysis of tube number and quality shows master junctions as blue dots circled in red, master segments in yellow (right panel; software: angiogenesis analyzer for ImageJ). (b) Angiogenesis assay on matrigel of HC HDMECs treated with anti-CCN3 antibody. Pictures taken 24 hours after beginning of experiment (left panel). (c) Significant decrease in number of master junctions in HC HDMECs treated with anti-CCN3 compared with control antibody (P = 0.0080). (d) Significant decrease in number of meshes in HC HDMECs treated with anti-CCN3 compared with control antibody (P = 0.0146). (e) Significant decrease in total length of master segments in HC HDMECs treated with anti-CCN3 compared with control antibody (P = 0.0104). (f) Angiogenesis assay on matrigel of untreated SSc HDMECs. Pictures taken 6 hours after beginning of experiment (left panel). Analysis of tube number and quality shows master junctions as blue dots circled in red, master segments in yellow (right panel; software: angiogenesis analyzer for ImageJ). (g) Angiogenesis assay on matrigel of SSc HDMECs treated with rCCN3 (10 μg/ml). Pictures taken 6 hours after beginning of experiment (left panel). (h) Significant increase in number of master junctions in SSc HDMECs treated with rCCN3 compared with untreated cells (P = 0.0066). (i) Significant increase in number of meshes in SSc HDMECs treated with rCCN3 compared with untreated cells (P = 0.0097). (j) Tendency to increase in total length of master segments in SSc HDMECs treated with rCCN3 compared with untreated cells (P = 0.0569). Two-way ANOVA with Sidak’s multiple comparisons test was performed for all analysis. Ab, antibody; ANOVA, analysis of variance; HC, healthy control; HDMEC, human microvascular dermal endothelial cell; NT, nontreated cells; rCCN3, recombinant CCN3; SSc, systemic sclerosis.
      We sought to determine the mechanism underlying this impaired angiogenesis. The anti-CCN3 antibody had no significant effect on cell proliferation and cell adhesion compared with control antibody and with nontreated cells (Supplementary Figure S7a and b). However, a significantly decreased migration was observed in HDMECs incubated with the anti-CCN3 antibody compared with the control condition (Supplementary Figure S7c), as assessed by the level of wound closure 8 hours after the beginning of the experiment (P = 0.02; Supplementary Figure S7d). No significant difference was observed between the control isotypic antibody and nontreated cells (data not shown). Altogether, these findings show that CCN3 blockade in HC HDMECs impairs in vitro angiogenesis at least partly by inhibiting migration.
      Next, we incubated SSc HDMECs with various doses of recombinant CCN3 (rCCN3: 0.1, 1, and 10 μg/ml). rCCN3 0.1 and 1 μg/ml showed a tendency to improve in vitro angiogenesis, although not statistically significant (data not shown). rCCN3 10 μg/ml significantly improved in vitro angiogenesis, as shown by the significant increase in the number of master junctions (P = 0.0066; Figure 4h) and the number of meshes (P = 0.0097; Figure 4i). A significant increase in the number of master segments (P = 0.0005, Supplementary Figure S6e) and the total segment length (P = 0.0066, Supplementary Figure S6f) was also observed. Finally, the total length of isolated branches tended to decrease in presence of rCCN3 (Supplementary Figure S6g). Addition of rCCN3 (10 μg/ml) to SSc HDMECs in a migration assay did not significantly change the level of wound closure (Supplementary Figure S7 e and f).
      Altogether, these data show that rCCN3 partly restores the angiogenesis defect observed in SSc HDMECs.

      Discussion

      Our results highlight a systemic architectural modification in SSc skin, as shown by the loss of papillary dermis and capillaries. These changes are associated with a global downregulation of CCN3 in dermal vessels and endothelial cells. CCN3 was found to play a prominent role in HDMEC migration and angiogenesis in vitro.
      One striking histological finding was the disappearance of papillary dermis, seen whatever the area of biopsy in SSc skin, as mirrored by a dramatic decrease in the number of capillaries. Comparison of the same clinical sites that we analyzed (dorsal forearm and upper inner arm) also found similar histological features between the two zones (
      • Van Praet J.T.
      • Smith V.
      • Haspeslagh M.
      • Degryse N.
      • Elewaut D.
      • De Keyser F.
      Histopathological cutaneous alterations in systemic sclerosis: a clinicopathological study.
      ). Studies analyzing fibroblasts taken from both nonfibrotic and fibrotic zones in the same patients have either shown a difference in the molecular patterns (
      • Corallo C.
      • Santucci A.
      • Bernardini G.
      • Figura N.
      • Leoncini R.
      • Riolo G.
      • et al.
      Proteomic investigation of dermal fibroblasts isolated from affected and unaffected skin samples from patients with limited cutaneous systemic sclerosis: 2. distinct entities?.
      ) or no difference (
      • 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.
      ). This decrease in papillary dermal surface seems to be different than papillary changes observed in other dermatological conditions such as lichen sclerosus et atrophicus, where the papillary dermis is often edematous (
      • Brănişteanu D.E.
      • Brănişteanu D.C.
      • Stoleriu G.
      • Ferariu D.
      • Voicu C.M.
      • Stoica L.E.
      • et al.
      Histopathological and clinical traps in lichen sclerosus: a case report.
      ,
      • Uitto J.
      • Santa Cruz D.J.
      • Bauer E.A.
      • Eisen A.Z.
      Morphea and lichen sclerosus et atrophicus. Clinical and histopathologic studies in patients with combined features.
      ). In any case, there is a need to decipher the pathophysiological mechanism behind the selective destruction of papillary dermis. It could be due to the enhanced secretion of matrix metalloproteinases by inflammatory cells or to tissular destruction by local hypoxia in the context of SSc-impaired angiogenesis.
      Inflammation and hypoxia are factors that modulate the level of CCN proteins, in particular CCN3 (
      • Jun J.I.
      • Lau L.F.
      Taking aim at the extracellular matrix: CCN proteins as emerging therapeutic targets.
      ,
      • Wolf N.
      • Yang W.
      • Dunk C.E.
      • Gashaw I.
      • Lye S.J.
      • Ring T.
      • et al.
      Regulation of the matricellular proteins CYR61 (CCN1) and NOV (CCN3) by hypoxia-inducible factor-1{alpha} and transforming-growth factor-{beta}3 in the human trophoblast.
      ). We observed a decreased dermal protein level of CCN3 in situ in the skin of patients with SSc. This result is not in contradiction with the increase in CCN3 found by
      • Lemaire R.
      • Farina G.
      • Bayle J.
      • Dimarzio M.
      • Pendergrass S.A.
      • Milano A.
      • et al.
      Antagonistic effect of the matricellular signaling protein CCN3 on TGF-beta- and Wnt-mediated fibrillinogenesis in systemic sclerosis and Marfan syndrome.
      , because they analyzed whole skin samples at the transcriptional level. The high rate of post-translational modifications of CCN3 (
      • Kyurkchiev S.
      • Yeger H.
      • Bleau A.M.
      • Perbal B.
      Potential cellular conformations of the CCN3(NOV) protein.
      ,
      • Yang W.
      • Wagener J.
      • Wolf N.
      • Schmidt M.
      • Kimmig R.
      • Winterhager E.
      • et al.
      Impact of CCN3 (NOV) glycosylation on migration/invasion properties and cell growth of the choriocarcinoma cell line Jeg3.
      ) demonstrates the need for assessment at the protein level, as shown by the discrepancy between quantitative PCR and Western blot results in our work. Here, degradation of CCN3 protein in SSc HDMECs could be enhanced, or CCN3 stabilization (through dimerization, for example) could be decreased. Moreover, we found CD34+/CCN3+ perivascular cells in HC skin that were almost absent in SSc skin. Because of the expression of CD34, the perivascular localization, and peculiar morphological characteristics (small cell body and long cytoplasmic prolongations), these cells could either be telocytes, which have been reported as decreased in SSc skin (
      • Manetti M.
      • Guiducci S.
      • Ruffo M.
      • Rosa I.
      • Faussone-Pellegrini M.S.
      • Matucci-Cerinic M.
      • et al.
      Evidence for progressive reduction and loss of telocytes in the dermal cellular network of systemic sclerosis.
      ), or mesenchymal stem cells (
      • Barron A.M.S.
      • Mantero J.C.
      • Ho J.D.
      • Nazari B.
      • Horback K.L.
      • Bhawan J.
      • et al.
      Perivascular adventitial fibroblast specialization accompanies T cell retention in the inflamed human dermis.
      ). The origin of these CD34+ perivascular cells remains unclear, yet their loss contributes to the decrease of CCN3 expression in dermal vessels.
      We show that CCN3 blockade impairs HC HDMEC in vitro angiogenesis and that adding rCCN3 to SSc HDEMCs improves in vitro angiogenesis. These results are consistent with the impaired in vitro angiogenesis already observed in SSc endothelial cells compared with HC cells (
      • Tsou P.S.
      • Wren J.D.
      • Amin M.A.
      • Schiopu E.
      • Fox D.A.
      • Khanna D.
      • et al.
      Histone deacetylase 5 is overexpressed in scleroderma endothelial cells and impairs angiogenesis via repressing pro-angiogenic factors.
      ). Thus, CCN3 seems to play a prominent role in HDMEC homeostasis, which is consistent with other reports (
      • Lin C.G.
      • Leu S.J.
      • Chen N.
      • Tebeau C.M.
      • Lin S.X.
      • Yeung C.Y.
      • et al.
      CCN3 (NOV) is a novel angiogenic regulator of the CCN protein family.
      ,
      • van Roeyen C.R.C.
      • Boor P.
      • Borkham-Kamphorst E.
      • Rong S.
      • Kunter U.
      • Martin I.V.
      • et al.
      A novel, dual role of CCN3 in experimental glomerulonephritis: pro-angiogenic and antimesangioproliferative effects.
      ). In SSc HDMECs, low CCN3 expression seems to be amenable to modulation by stimulation with IL-1β (whereas no change in CCN3 extracellular secretion was seen in HC cells), although these data need to be confirmed with a larger data set. IL-1β stimulation activates many pathways in endothelial cells, triggering a response program including potent proangiogenic factors, notably via NF-кB pathway (
      • Pober J.S.
      • Sessa W.C.
      Evolving functions of endothelial cells in inflammation.
      ). These data could lead to deciphering the IL-1β/CCN3 axis in endothelial cells on a therapeutic prospect.
      Impaired tube formation can be due to decreased migration or dysfunction in cell-cell attachment (
      • Irvin M.W.
      • Zijlstra A.
      • Wikswo J.P.
      • Pozzi A.
      Techniques and assays for the study of angiogenesis.
      ). This is consistent with our results, showing a more important decrease in tube formation than in migration. This cannot be attributed to a difference in cell proliferation, given that it was not altered by anti-CCN3 treatment. Thus, in our model, CCN3 blockade probably impairs mostly cell-cell attachment through defect in integrin binding, directly or indirectly. A direct interaction between CCN3 and endothelial cell surface integrins (αVβ3, α6β1, and α5β1) has indeed already been shown (
      • Lin C.G.
      • Chen C.C.
      • Leu S.J.
      • Grzeszkiewicz T.M.
      • Lau L.F.
      Integrin-dependent functions of the angiogenic inducer NOV (CCN3): implication in wound healing.
      ,
      • Lin C.G.
      • Leu S.J.
      • Chen N.
      • Tebeau C.M.
      • Lin S.X.
      • Yeung C.Y.
      • et al.
      CCN3 (NOV) is a novel angiogenic regulator of the CCN protein family.
      ). CCN3 blockade could also impair CCN3’s binding to extracellular matrix constituents (
      • Liu S.
      • Liu Z.
      • Bi D.
      • Yuan X.
      • Liu X.
      • Ding S.
      • et al.
      CCN3 (NOV) regulates proliferation, adhesion, migration and invasion in clear cell renal cell carcinoma.
      ) and thus inhibiting tubulogenesis by preventing extracellular matrix degradation. Adding rCCN3 to SSc HDMECs partly restores in vitro angiogenesis, which is particularly interesting from a therapeutic point of view.
      However, the full role of CCN3 in endothelial cells remains unknown. Intracellular CCN3, and particularly the nuclear form, may act as a cotranscriptional factor to regulate angiogenesis as well as other cellular functions. We have shown that SSc HDMECs do not exhibit intranuclear CCN3 staining, unlike healthy HDMECs. Further studies are needed to fully understand if intranuclear CCN3 regulates HDMEC homeostasis.
      Altogether, our data suggest the involvement of CCN3 in SSc skin pathophysiology, especially vasculopathy. On another note, CCN3 has been reported as antagonizing the profibrotic role of CCN2 in several models (
      • Borkham-Kamphorst E.
      • Huss S.
      • Van de Leur E.
      • Haas U.
      • Weiskirchen R.
      Adenoviral CCN3/NOV gene transfer fails to mitigate liver fibrosis in an experimental bile duct ligation model because of hepatocyte apoptosis.
      ,
      • Lafont J.
      • Jacques C.
      • Le Dreau G.
      • Calhabeu F.
      • Thibout H.
      • Dubois C.
      • et al.
      New target genes for NOV/CCN3 in chondrocytes: TGF-beta2 and type X collagen.
      ,
      • Lemaire R.
      • Farina G.
      • Bayle J.
      • Dimarzio M.
      • Pendergrass S.A.
      • Milano A.
      • et al.
      Antagonistic effect of the matricellular signaling protein CCN3 on TGF-beta- and Wnt-mediated fibrillinogenesis in systemic sclerosis and Marfan syndrome.
      ,
      • Riser B.L.
      • Najmabadi F.
      • Perbal B.
      • Peterson D.R.
      • Rambow J.A.
      • Riser M.L.
      • et al.
      CCN3 (NOV) is a negative regulator of CCN2 (CTGF) and a novel endogenous inhibitor of the fibrotic pathway in an in vitro model of renal disease.
      ) but not in the skin. Here, we show that CCN3 plays a proangiogenic role, CCN2 also being reported as a proangiogenic factor (
      • Brigstock D.R.
      Regulation of angiogenesis and endothelial cell function by connective tissue growth factor (CTGF) and cysteine-rich 61 (CYR61).
      ,
      • Chintala H.
      • Liu H.
      • Parmar R.
      • Kamalska M.
      • Kim Y.J.
      • Lovett D.
      • et al.
      Connective tissue growth factor regulates retinal neovascularization through p53 protein-dependent transactivation of the matrix metalloproteinase (MMP)-2 gene.
      ,
      • Ponticos M.
      Connective tissue growth factor (CCN2) in blood vessels.
      ). It is likely that CCN3’s role is organ- and function-dependent, and its antifibrotic role in the skin remains to be demonstrated.
      In conclusion, our results indicate that decrease of CCN3 in SSc endothelial cells is linked to SSc-impaired angiogenesis. A future challenge is to develop therapeutic approaches to restore CCN3 homeostasis in the skin.

      Materials and Methods

      Patients and controls

      Within the biomedical research cohort Vasculopathy and Inflammation in Systemic Sclerosis approved by the institutional ethical committee (CPP, 2012 A00081-42, Aquitaine), 26 patients with SSc were recruited over a two-year period (from September 2015 to September 2017). All patients met the classification criteria proposed by the American College of Rheumatology and the European League Against Rheumatism 2013 (
      • van 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.
      ). All patients provided written, informed consent before entering the study. Skin biopsies were performed on the forearm, on a fibrotic zone whenever possible, as spindle-shaped biopsies of 1 cm of wide axis. If the forearm was fibrotic (17 patients), another similar biopsy was performed on the proximal part of the same arm, in a nonfibrotic zone. For HC skin, 18 biopsy specimens were isolated from arm skin that had been discarded during plastic surgery (brachioplasty) thanks to a research convention with the plastic surgery department. For patients with SSc, vascular involvement was defined as current or past history of digital ulcers, pulmonary hypertension, and/or renal crisis.

      Histological assessment, immunohistochemistry, and immunofluorescence

      Part of the skin biopsy was formalin-fixed and paraffin-embedded before undergoing 4-μm-thick microtome sections, another part being devoted for cell culture. Histological analyses were performed using Masson’s Trichrome Aniline Blue staining. Immunohistochemistry was performed as previously described (
      • Cario-Andre M.
      Analysis of CCN expression by immunofluorescence on skin cells, skin, and reconstructed epidermis.
      ,
      • Marie J.
      • Kovacs D.
      • Pain C.
      • Jouary T.
      • Cota C.
      • Vergier B.
      • et al.
      Inflammasome activation and vitiligo/nonsegmental vitiligo progression.
      ), using antibodies against CCN3 (ab137677; Abcam, Cambridge, United Kingdom), CD31 (clone JC70A; Dako, Carpinteria, CA), and CD34 (QBEnd-10; Thermo Fisher Scientific, Waltham, MA). The same dilutions were used for all immunohistochemical stains. Images were acquired using an epifluorescence microscope (Leica). For analysis, see Supplementary Materials and Methods.

      Endothelial cell primary cell culture

      HDMECs were obtained from five patients with SSc and seven HC skin samples, based on a previous protocol (
      • Normand J.
      • Karasek M.A.
      A method for the isolation and serial propagation of keratinocytes, endothelial cells, and fibroblasts from a single punch biopsy of human skin.
      ), improved with purification using a CD31 microbead kit (Miltenyi Biotec, Bergisch Gladbach, Germany). The cells were then cultured in MV2 EC medium (PromoCell, Heidelberg, Germany) and used between passages 2 and 5.

      Flow cytometry

      One passage after purification, HDMECs were labeled with an anti-CD31 allophycocyanin-conjugated antibody (clone AC128, Miltenyi Biotec). Cells were analyzed using a Canto I flow cytometer with FACSDiva software (BD Biosciences, San Jose, CA), and data analysis was performed with FlowJo 10.1 software.

      Immunocytochemistry

      Cultured cells were fixed in 4% formaldehyde, washed with phosphate buffered saline, permeabilized with phosphate buffered saline–TritonX100 0.2%, saturated with fetal calf serum 2.5%, and incubated overnight (4 °C) with anti-CCN3 antibody (ab137677; Abcam). Images were acquired using an epifluorescence microscope (Nikon Eclipse).

      Western blotting

      Cells were scraped with RIPA buffer and lysates were centrifuged for 20 minutes at 12,500 r.p.m and 4 °C. After electrophoresis on a 10% acrylamide gel, proteins were transferred to a polyvinylidene fluoride membrane. Membranes were saturated in Tris-buffered saline-5% milk, then incubated with antibodies against CCN3 (ab191425; Abcam), β-actin (Sigma-Aldrich, St. Louis, MO), and corresponding secondary antibodies (Vector Laboratories, Burlingame, CA) and revealed with ECL+ (Amersham). Chemiluminescence acquisitions were performed using LAS-3000 Reader (Fujifilm Life Science). Quantification analyses were performed using ImageLab 6.0 software (Bio-Rad, Hercules, CA).

      Determination of secreted CCN3 level

      CCN3 levels in serum and cell culture supernatants were quantified using an ELISA kit (R&D Systems, Minneapolis, MN), according to the manufacturer’s instructions. The cut-off for CCN3 positivity was defined as values equal or superior to 8 pg/ml, which corresponds to the detection limit of the ELISA.

      Angiogenesis assay

      HDMECs from three different healthy and SSc donors were seeded on 6-well plates (200,000 cells/well) and, after allowing them to adhere, were incubated overnight with either 0.5 μg/ml of anti-CCN3 (ab137677; Abcam) or control antibody (ab37415; Abcam) and either rCCN3 (0.1, 1, or 10 μg/mL) (120-26; PeproTech, Rocky Hill, NJ). The day after, cells were seeded in complete growth medium MV2 on Ibidi microslides (Ibidi, Gräfelfing, Germany; cat. No. 81506) coated with Matrigel Matrix Growth Factor Reduced, Phenol Red-Free (BD Biosciences, cat. No. 356231) and allowed to form tubes at 37 °C. 10,000 cells were used per well, five wells were used per condition in a single experiment, and experiments were replicated three times. Pictures were taken 6 and 24 hours after the beginning of the experiment with a phase contrast microscope (Leica). For analysis, see Supplementary Materials and Methods.

      Data availability statement

      No datasets were generated or analyzed during this study.

      Conflict of Interest

      TS: Abbvie: advice, consultancy, clinical trials; Lilly: advice, consultancy, clinical trials; Pfizer: advice, consultancy, research grant; Novartis: advice, clinical trials, Abivax: clinical trials; Biogen: clinical trials. The other authors state no conflict of interest.

      Acknowledgments

      This work was funded by Fédération de Recherche Transbiomed , Association des Sclérodermiques de France , and Société Française de Dermatologie . The authors wish to thank Sébastien Marais, from the Bordeaux Imaging Center, for his precious help concerning ImageJ analysis; Catherine Pain, for technical assistance; and all the clinicians (Claire Carcaud, Gaël Galli) and patients who took part in the VISS cohort.

      Author Contributions

      Conceptualization: PH, MET, FM, MC; Formal Analysis: PH; Funding Acquisition: PH, MET, MC; Investigation: PH, PL, PM, VJ, VL; Resources: FM, PKC, FP, CB, TS, JS, EL; Supervision: MET, FM, HRR, AT, MC; Visualization: PH, MET; Writing - Original Draft Preparation: PH, MET, FM, MC; Writing- Review and Editing: PH, FM, MET, MC.

      Supplementary Materials and Methods

      Histological and immunohistological analysis

      Each staining was performed three times to ensure reproducibility of the experiment. For each analysis, three to five sections per sample were randomly chosen and a prespecified mask was applied to each section. Annexes were excluded from the analysis. Papillary dermis surface, interdigitation index, and epidermal surface were measured by two blinded observers (PH and VJ). For dermal cell counting, the number of cells was automatically counted using the particle analyzer tool (DAPI staining) and the results were normalized in relation to the surface of dermis analyzed. Dermal vessels were counted manually by a blinded observer (PH) and the results were normalized in relation to the surface of dermis analyzed. Results per sample are the mean of three randomly chosen sections. Papillary dermal vessels were detected manually according to size and localization. The mean number of CCN3+ layers on the vessels of each section was counted manually by a blinded observer (PH) and results per sample are the mean of all vessels of three randomly chosen sections. To assess CCN3 intensity in endothelial cells, dermal endothelial cells were identified by CD31 staining and mean CCN3 fluorescence of all dermal endothelial cells was quantified, thus avoiding a possible bias because of the decrease in the absolute number of vessels. This fluorescence intensity was then divided by the fluorescence of the whole dermis to obtain a ratio allowing comparisons between multiple samples. The mean fluorescence intensity was influenced neither by vessel size, the number of cells in the vessel, nor by the number of vessels. It therefore reliably reflects the level of CCN3 expression in dermal endothelial cells. All analyses were performed using ImageJ software (1.51s).

      CCN3 antagonization assays

      Human microvascular dermal endothelial cells from three different healthy donors were incubated with various concentrations of an anti-CCN3 antibody (ab137677; Abcam, Cambridge, United Kingdom), ranging from 0.05 μg/ml to 1 μg/ml (Figure 4a). The lowest concentration, 0.05 μg/ml, did not show any effect on proliferation. The middle concentrations, 0.1 and 0.5 μg/ml, showed a tendency to an increased proliferation, but without statistical significance. The highest concentration, 1 μg/ml, showed an inhibition of proliferation, but was not statistically significant. The concentration of 0.5 μg/ml was chosen for the subsequent experiments, as it had proven not to be toxic for cell growth. An isotypic control antibody (ab37415; Abcam) was tested and used at the same concentration.

      Angiogenesis assay analysis

      Three wells per condition per experiment (out of five wells) were chosen blindly for the analysis. Quality of tube formation was measured at 24 hours after the beginning of the experiment with the plug-in Angiogenesis Analyzer of Image J software (Gilles Carpentier). The following parameters were analyzed: number of master junctions, number of master segments, total length of master segments, number of meshes, length of total segments, and length of isolated branches. Definition of each parameter is shown in Supplementary Figure 4a.

      Proliferation assay

      Endothelial cells from three different healthy donors were seeded on a 96-well culture plate (2,500 cells/well, each condition in triplicate) and TetraZ assay was performed according to the manufacturer’s instructions (TetraZ Cell Counting Kit, BioLegend, San Diego, CA).

      Adhesion assay

      Overnight, 96-well culture plates were coated with fibronectin (10 μg/ml diluted in phosphate buffered saline [PBS]-ABC [PBS + 1 mM MgCl2 + 1 mM CaCl2]), then saturated the day after with denaturated PBS-BSA and washed with complete cell growth medium. The coated wells were seeded with 20,000 cells/well (5 wells per condition per experiment) and incubated at 37 °C for 30 minutes. The wells were then washed several times with PBS-ABC and incubated for 5 minutes at room temperature with Cristal Violet 0.2% in ethanol 10%. The wells were washed three times with PBS-ABC and flushed with solubilization buffer (50% NaH2PO4 0.1M/50% absolute ethanol, pH 4.5). All chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Optical density was immediately read at 570 nm.

      Migration assay

      Human microvascular dermal endothelial cells from three different healthy donors or one patient with systemic sclerosis were seeded on a 96-well culture plate (30,000 cells/well to reach immediate confluence; five wells per condition per experiment), and after allowing them to adhere overnight, a scratch was performed manually at the center of the well with a 200-μl tip. The wells were rinsed with Hank's Balanced Salt Solution (Gibco, Thermo Fisher Scientific, Waltham, MA) and complete EC growth medium (MV2, PromoCell, Heidelberg, Germany). Anti-CCN3 or control isotypic antibody was added (0.5 μg/ml), or recombinant CCN3 (10 μg/ml). Images were obtained using an optical inverted microscope (Eclipse Ti-U, Nikon) 0 and 8 hours after the beginning of the experiment.

      RNA isolation and quantitative real-time reverse transcriptase–PCR

      Total RNA was isolated from trypsinized endothelial cells and keratinocytes with the NucleoSpin RNA Plus extraction system (Macherey-Nagel, Düren, Germany), and cDNA was synthesized from 0.35 μg of total RNA (endothelial cells) and 1 μg of total RNA (keratinocytes) using random hexamers and Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Gene expression was quantified by SYBR green on a MX3000P instrument (Stratagene, San Diego, CA). The specific primer pairs for CCN3 are (Forward) CGG-CTG-CTC-ATG-CTG-TCT-GG and (Reverse) TTA-TCT-CCC-TCT-ACC-GCC-GTG-C. Each reaction was performed in triplicate. Stable housekeeping genes GAPDH and HUPO were selected for normalization. The oligonucleotides were obtained from Eurogentec (Liège, Belgium). The differences were calculated using the threshold cycle (Ct) and the comparative Ct method for relative quantification.

      Statistical analysis

      Statistical analyses were performed using GraphPad Prism 6. A paired Student t-test was used to compare nonfibrotic and fibrotic areas of the same patients with systemic sclerosis. To compare healthy controls and patients with systemic sclerosis, an unpaired t-test with Welch’s correction (in the event of significantly different variances across populations) was used. When values did not follow a Gaussian distribution, a Mann-Whitney test was used. For angiogenesis assays, analysis of variance was used. A P-value < 0.05 was considered statistically significant.
      Figure thumbnail fx2
      Supplementary Figure S1Histological changes in SSc dermis. Significant decrease in the interdigitation index in SSc skin, whereas in fibrotic or nonfibrotic area (NFib vs. HC: P = 0.0007, Fib vs. HC: P = 0.002, NFib vs. Fib: P = 0.8073). The interdigitation index is computed by dividing the length of the dermo-epidermal junction by the length of the stratum corneum and represents the degree of incurvation of the basal membrane. Fib, fibrotic; HC, healthy control; NFib, nonfibrotic; SSc, systemic sclerosis.
      Figure thumbnail fx3
      Supplementary Figure S2Identification of CCN3+ layers in HC and SSc dermal vessels. Double immunohistochemistry stainings for CCN3 and CD34, α- smooth muscle actin, CD3, CD90, and CD31 to respectively identify endothelial cells and perivascular CD34+ cells, pericytes and smooth muscle cells, T lymphocytes, fibroblasts, and endothelial cells were performed. * indicates vessel lumen. (a) In HC skin, all three dermal vessels layers are CCN3-positive (inner layer composed of endothelial cells; middle layer composed of smooth muscle cells; outer layer composed of fibroblasts, lymphocytes and perivascular CD34+ cells). (b) In SSc skin, two layers of CCN3+ cells can be seen, composed of endothelial cells and smooth muscle cells. No perivascular CD34+ cell is seen. CCN3+ perivascular fibroblasts can be observed. As for perivascular lymphocytes, none were seen for this section (patient with SSc with advanced disease not in inflammatory phase) but isolated CD3+ cells were CCN3-positive. HC, healthy control; SSc, systemic sclerosis. Bar = 20 μm.
      Figure thumbnail fx4
      Supplementary Figure S3CCN3 expression in the dermis. (a) Mean number of CCN3+ layers is not statistically different whether in the nonfibrotic or fibrotic zone of SSc skin (NFib vs. HC: P < 0.0001, Fib vs. HC: P = 0.0002, NFib vs. Fib: P = 0.4469). (b) CCN3 expression in dermal endothelial cells is not statistically different whether in the nonfibrotic or fibrotic zone of SSc skin (NFib vs. HC: P = 0.0008, Fib vs. HC: P < 0.0001, NFib vs. Fib: P = 0.8760). (c) CCN3 expression in dermal endothelial cells is not statistically different in a subgroup analysis when taking the presence of vascular involvement in patients with SSc into account (P = 0.9310). (d) Significant positive correlation between mean CCN3 expression in dermal vessels and number of dermal vessels (Spearman r = 0.6227, P < 0.0001). AU, arbitrary units; Fib, fibrotic; HC, healthy control; NFib, nonfibrotic; NVasc, nonvascular; SSc, systemic sclerosis; Vasc, vascular.
      Figure thumbnail fx5
      Supplementary Figure S4CCN3 expression in endothelial cells of patients with SSc and controls. (a) Pictures of HDMECs cultured from a healthy control, a patient with SSc from a nonlesional zone, and a patient with SSc from a lesional zone. Note morphological changes especially in fibrotic zone of patient with SSc. Bar = 200 μm. (b) No significant difference in CCN3 expression at the transcriptional level between patients with SSc and HC HDMECs (qPCR) (P = 0.7056, fold increase compared with one HC sample). HC, healthy control; HDMEC, human microvascular dermal endothelial cell; qPCR, quantitative PCR; SSc, systemic sclerosis.
      Figure thumbnail fx6
      Supplementary Figure S5CCN3 secretion after stimulation of HDMECs with IL-1β. Cells were treated with IL-1β at 50 ng/ml for 24 hours. Data are mean ± SEM of one experiment involving three different donors for both HC and SSc cells and performed in duplicate. Paired t-test was used for statistical analysis. (a) CCN3 secretion is unchanged in stimulated HC HDMECs (P = 0.7233). (b) CCN3 secretion shows a tendency to increase in SSc HDMECs stimulated with IL-1β, although not statistically significant (P = 0.2180). HC, healthy control; HDMEC, human microvascular dermal endothelial cell; NT, nontreated cells; SEM, standard error of the mean; SSc, systemic sclerosis.
      Figure thumbnail fx7
      Supplementary Figure S6Angiogenesis assay. (a) Adapted from angiogenesis assay analysis, plug-in Angiogenesis Analyzer for ImageJ (Gilles Carpentier). A segment (arrow 2 in cartoon b, for example) is delimited by two junctions (red nodes in cartoon a). A master segment (yellow branch in cartoon c) is delimited by two master junctions (blue node circled in red in cartoon c), which are joining at least three segments. Meshes are polygons formed by the reunion of several master segments (light blue polygon in cartoon d) and represent the section of one tube in vitro. Isolated branches are segments delimited by only one junction. (b) Significant decrease in number of master segments in HDMECs treated with anti-CCN3 compared with control antibody (P = 0.0118). (c) Significant decrease in total segment length in HDMECs treated with anti-CCN3 compared with control antibody (P = 0.0114). (d) No significant difference in total length of isolated branches in HDMECs treated with anti-CCN3 compared with control antibody, despite an increase (P = 0.2571). (e) Significant increase in number of master segments in SSc HDMECs treated with recombinant CCN3 (10 μg/ml) compared with NT cells (P = 0.0005). (f) Significant increase in total segment length in SSc HDMECs treated with recombinant CCN3 (10 μg/ml) compared with NT cells (P = 0.0066). (g) No significant difference in total length of isolated branches in SSc HDMECs treated with recombinant CCN3 (10 μg/ml) compared with NT cells, despite a decrease (P = 0.4329). Two-way ANOVA with Sidak’s multiple comparisons test was performed for all analyses. Ab, antibody; ANOVA, analysis of variance; HC, healthy control; HDMEC, human microvascular dermal endothelial cell; NT, nontreated cells; SSc, systemic sclerosis.
      Figure thumbnail fx8
      Supplementary Figure S7Mechanisms of CCN3’s action on in vitro angiogenesis. For HC HDMECs, data represent mean ± SEM of three independent experiments (performed with three different donors, in triplicate), and one-way ANOVA with Sidak’s multiple comparisons test was performed for all analyses. For SSc HDMECs, data represent mean ± SD of one experiment performed in triplicate (as such experiment requires a great number of cells, which is difficult to obtain for SSc HDMECs). (a) No significant difference in proliferation in HC HDMECs treated with anti-CCN3 compared with control antibody or NT cells. Anti-CCN3 Ab versus Control Ab: P = 0.9998; anti-CCN3 Ab versus NT cells: P = 0.9343; Control Ab versus NT cells: P = 0.9918. (b) No significant difference in adhesion in HC HDMECs treated with anti-CCN3 compared with control antibody (P = 0.7319). (c) Migration assay of HC HDMECs incubated with control or anti-CCN3 antibody. Pictures were taken 8 hours after initial wound. Bar = 200 μm. Slowdown of migration is observed in HDMECs incubated with anti-CCN3 antibody. (d) Significant difference in mean percentage of wound closure between control and anti-CCN3 antibody. Significantly impaired migration is observed with anti-CCN3 antibody compared with control cells (P = 0.0248). (e) Migration assay of SSc HDMECs untreated or treated with rCCN3 (10 μg/ml). Pictures were taken 8 hours after initial wound. Bar = 400 μm. (f) Tendency to an increase in the mean percentage of wound closure in SSc HDMECs incubated with rCCN3, although not statistically significant (P = 0.2279). Ab, antibody; ANOVA, analysis of variance; HC, healthy control; HDMEC, human microvascular dermal endothelial cell; NT, nontreated cells; rCCN3, recombinant CCN3; SD, standard deviation; SEM, standard error of the mean; SSc, systemic sclerosis.

      References

        • Allanore Y.
        • Distler O.
        • Matucci-Cerinic M.
        • Denton C.P.
        Review: Defining a unified vascular phenotype in systemic sclerosis.
        Arthritis Rheumatol. 2018; 70: 162-170
        • Allanore Y.
        • Simms R.
        • Distler O.
        • Trojanowska M.
        • Pope J.
        • Denton C.P.
        • et al.
        Systemic sclerosis.
        Nat Rev Dis Primers. 2015; 1: 15002
        • Barron A.M.S.
        • Mantero J.C.
        • Ho J.D.
        • Nazari B.
        • Horback K.L.
        • Bhawan J.
        • et al.
        Perivascular adventitial fibroblast specialization accompanies T cell retention in the inflamed human dermis.
        J Immunol. 2019; 202: 56-68
        • Borkham-Kamphorst E.
        • Huss S.
        • Van de Leur E.
        • Haas U.
        • Weiskirchen R.
        Adenoviral CCN3/NOV gene transfer fails to mitigate liver fibrosis in an experimental bile duct ligation model because of hepatocyte apoptosis.
        Liver Int. 2012; 32: 1342-1353
        • Brănişteanu D.E.
        • Brănişteanu D.C.
        • Stoleriu G.
        • Ferariu D.
        • Voicu C.M.
        • Stoica L.E.
        • et al.
        Histopathological and clinical traps in lichen sclerosus: a case report.
        Rom J Morphol Embryol. 2016; 57: 817-823
        • Brigstock D.R.
        Regulation of angiogenesis and endothelial cell function by connective tissue growth factor (CTGF) and cysteine-rich 61 (CYR61).
        Angiogenesis. 2002; 5: 153-165
        • Cario-Andre M.
        Analysis of CCN expression by immunofluorescence on skin cells, skin, and reconstructed epidermis.
        Methods Mol Biol. 2017; 1489: 63-76
        • Chintala H.
        • Liu H.
        • Parmar R.
        • Kamalska M.
        • Kim Y.J.
        • Lovett D.
        • et al.
        Connective tissue growth factor regulates retinal neovascularization through p53 protein-dependent transactivation of the matrix metalloproteinase (MMP)-2 gene.
        J Biol Chem. 2012; 287: 40570-40585
        • Corallo C.
        • Santucci A.
        • Bernardini G.
        • Figura N.
        • Leoncini R.
        • Riolo G.
        • et al.
        Proteomic investigation of dermal fibroblasts isolated from affected and unaffected skin samples from patients with limited cutaneous systemic sclerosis: 2. distinct entities?.
        J. Rheumatol. 2017; 44: 40-48
        • 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
        • Gabrielli A.
        • Avvedimento E.V.
        • Krieg T.
        Scleroderma.
        N Engl J Med. 2009; 360: 1989-2003
        • Gellhaus A.
        • Schmidt M.
        • Dunk C.
        • Lye S.J.
        • Kimmig R.
        • Winterhager E.
        Decreased expression of the angiogenic regulators CYR61 (CCN1) and NOV (CCN3) in human placenta is associated with pre-eclampsia.
        Mol Hum Reprod. 2006; 12: 389-399
        • Henrot P.
        • Truchetet M.E.
        • Fisher G.
        • Taïeb A.
        • Cario M.
        CCN proteins as potential actionable targets in scleroderma.
        Exp Dermatol. 2019; 28: 11-18
        • Irvin M.W.
        • Zijlstra A.
        • Wikswo J.P.
        • Pozzi A.
        Techniques and assays for the study of angiogenesis.
        Exp Biol Med (Maywood). 2014; 239: 1476-1488
        • Jun J.I.
        • Lau L.F.
        Taking aim at the extracellular matrix: CCN proteins as emerging therapeutic targets.
        Nat Rev Drug Discov. 2011; 10: 945-963
        • Kyurkchiev S.
        • Yeger H.
        • Bleau A.M.
        • Perbal B.
        Potential cellular conformations of the CCN3(NOV) protein.
        Cell Commun Signal. 2004; 2: 9
        • Lafont J.
        • Jacques C.
        • Le Dreau G.
        • Calhabeu F.
        • Thibout H.
        • Dubois C.
        • et al.
        New target genes for NOV/CCN3 in chondrocytes: TGF-beta2 and type X collagen.
        J Bone Miner Res. 2005; 20: 2213-2223
        • Leask A.
        • Denton C.P.
        • Abraham D.J.
        Insights into the molecular mechanism of chronic fibrosis: the role of connective tissue growth factor in scleroderma.
        J Invest Dermatol. 2004; 122: 1-6
        • Lemaire R.
        • Farina G.
        • Bayle J.
        • Dimarzio M.
        • Pendergrass S.A.
        • Milano A.
        • et al.
        Antagonistic effect of the matricellular signaling protein CCN3 on TGF-beta- and Wnt-mediated fibrillinogenesis in systemic sclerosis and Marfan syndrome.
        J Invest Dermatol. 2010; 130: 1514-1523
        • Lin C.G.
        • Chen C.C.
        • Leu S.J.
        • Grzeszkiewicz T.M.
        • Lau L.F.
        Integrin-dependent functions of the angiogenic inducer NOV (CCN3): implication in wound healing.
        J Biol Chem. 2005; 280: 8229-8237
        • Lin C.G.
        • Leu S.J.
        • Chen N.
        • Tebeau C.M.
        • Lin S.X.
        • Yeung C.Y.
        • et al.
        CCN3 (NOV) is a novel angiogenic regulator of the CCN protein family.
        J Biol Chem. 2003; 278: 24200-24208
        • Liu S.
        • Liu Z.
        • Bi D.
        • Yuan X.
        • Liu X.
        • Ding S.
        • et al.
        CCN3 (NOV) regulates proliferation, adhesion, migration and invasion in clear cell renal cell carcinoma.
        Oncol Lett. 2012; 3: 1099-1104
        • Madne T.H.
        • Dockrell M.E.C.
        CCN3, a key matricellular protein, distinctly inhibits TGFβ1-mediated Smad1/5/8 signalling in human podocyte culture.
        Cell Mol Biol (Noisy-le-grand). 2018; 64: 5-10
        • Manetti M.
        • Guiducci S.
        • Ruffo M.
        • Rosa I.
        • Faussone-Pellegrini M.S.
        • Matucci-Cerinic M.
        • et al.
        Evidence for progressive reduction and loss of telocytes in the dermal cellular network of systemic sclerosis.
        J Cell Mol Med. 2013; 17: 482-496
        • Marie J.
        • Kovacs D.
        • Pain C.
        • Jouary T.
        • Cota C.
        • Vergier B.
        • et al.
        Inflammasome activation and vitiligo/nonsegmental vitiligo progression.
        Br J Dermatol. 2014; 170: 816-823
        • Meunier P.
        • Dequidt L.
        • Barnetche T.
        • Lazaro E.
        • Duffau P.
        • Richez C.
        • et al.
        Increased risk of mortality in systemic sclerosis-associated digital ulcers: a systematic review and meta-analysis.
        J Eur Acad Dermatol Venereol. 2019; 33: 405-409
        • Normand J.
        • Karasek M.A.
        A method for the isolation and serial propagation of keratinocytes, endothelial cells, and fibroblasts from a single punch biopsy of human skin.
        In Vitro Cell Dev Biol Anim. 1995; 31: 447-455
        • Perbal B.
        New insight into CCN3 interactions - Nuclear CCN3 : fact or fantasy?.
        Cell Commun Signal. 2006; 4: 6
        • Pober J.S.
        • Sessa W.C.
        Evolving functions of endothelial cells in inflammation.
        Nat Rev Immunol. 2007; 7: 803-815
        • Ponticos M.
        Connective tissue growth factor (CCN2) in blood vessels.
        Vascul Pharmacol. 2013; 58: 189-193
        • Riser B.L.
        • Najmabadi F.
        • Perbal B.
        • Peterson D.R.
        • Rambow J.A.
        • Riser M.L.
        • et al.
        CCN3 (NOV) is a negative regulator of CCN2 (CTGF) and a novel endogenous inhibitor of the fibrotic pathway in an in vitro model of renal disease.
        Am J Pathol. 2009; 174: 1725-1734
        • Serratì S.
        • Chillà A.
        • Laurenzana A.
        • Margheri F.
        • Giannoni E.
        • Magnelli L.
        • et al.
        Systemic sclerosis endothelial cells recruit and activate dermal fibroblasts by induction of a connective tissue growth factor (CCN2)/transforming growth factor β-dependent mesenchymal-to-mesenchymal transition.
        Arthritis Rheum. 2013; 65: 258-269
        • Tsou P.S.
        • Wren J.D.
        • Amin M.A.
        • Schiopu E.
        • Fox D.A.
        • Khanna D.
        • et al.
        Histone deacetylase 5 is overexpressed in scleroderma endothelial cells and impairs angiogenesis via repressing pro-angiogenic factors.
        Arthritis Rheumatol. 2016; 68: 2975-2985
        • Uitto J.
        • Santa Cruz D.J.
        • Bauer E.A.
        • Eisen A.Z.
        Morphea and lichen sclerosus et atrophicus. Clinical and histopathologic studies in patients with combined features.
        J Am Acad Dermatol. 1980; 3: 271-279
        • van 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: 2737-2747
        • Van Praet J.T.
        • Smith V.
        • Haspeslagh M.
        • Degryse N.
        • Elewaut D.
        • De Keyser F.
        Histopathological cutaneous alterations in systemic sclerosis: a clinicopathological study.
        Arthritis Res Ther. 2011; 13 (1:R35)
        • van Roeyen C.R.C.
        • Boor P.
        • Borkham-Kamphorst E.
        • Rong S.
        • Kunter U.
        • Martin I.V.
        • et al.
        A novel, dual role of CCN3 in experimental glomerulonephritis: pro-angiogenic and antimesangioproliferative effects.
        Am J Pathol. 2012; 180: 1979-1990
        • Varga J.
        • Abraham D.
        Systemic sclerosis: a prototypic multisystem fibrotic disorder.
        J Clin Invest. 2007; 117: 557-567
        • Wolf N.
        • Yang W.
        • Dunk C.E.
        • Gashaw I.
        • Lye S.J.
        • Ring T.
        • et al.
        Regulation of the matricellular proteins CYR61 (CCN1) and NOV (CCN3) by hypoxia-inducible factor-1{alpha} and transforming-growth factor-{beta}3 in the human trophoblast.
        Endocrinology. 2010; 151: 2835-2845
        • Yang W.
        • Wagener J.
        • Wolf N.
        • Schmidt M.
        • Kimmig R.
        • Winterhager E.
        • et al.
        Impact of CCN3 (NOV) glycosylation on migration/invasion properties and cell growth of the choriocarcinoma cell line Jeg3.
        Hum Reprod. 2011; 26: 2850-2860