TH17 cells expressing CD146 are significantly increased in patients with Systemic sclerosis

Systemic sclerosis (SSc) is an autoimmune disorder characterized by vascular damage, excessive fibrosis and abnormal T cells immune-regulation. CD146 is an adhesion molecule essentially expressed in the vascular system, but also on TH17 lymphocytes. In view of the recently described role of CD146 in SSc, we hypothesized an involvement of CD146 positive TH17 cells in this disease. Compared to healthy controls, we showed that both soluble form of CD146 (sCD146), and IL17A levels were increased in patients with SSc with a positive correlation between both factors. A significant increase in TH17 cells attested by an increase of RORγT, IL17A mRNA and CD4+ IL17A+ cell was observed in patients with SSc. Interestingly, the percentage of TH17 cells expressing CD146 was higher in patients with SSc and inversely correlated with pulmonary fibrosis. In vitro experiments showed an augmentation of the percentage of TH17 cells expressing CD146 after cell treatment with sCD146, suggesting that, in patients the increase of this sub-population could be the consequence of the sCD146 increase in serum. In conclusion, TH17 cells expressing CD146 could represent a new component of the adaptive immune response, opening the way for the generation of new tools for the management of SSc.

differentiation and for production of cytokines which participates in protective immunity at boundary tissues 15 . IL17A primarily binds to IL-17RA 16 and initiates a pro-inflammatory signal, which is strongly involved in the progression of many autoimmune diseases in humans 17,18 .
In SSc, the mRNA and protein expression of IL17A was found to be positively correlated to an outcome index, the modified Rodnan skin score 17,19 .
In view of the role played by CD146 in SSc, we hypothesized a preponderant role of CD146 positive TH17 cells in SSc. To investigate this hypothesis, we explored in the present study, the concentrations of sCD146 and IL17A in serum of patients with SSc as compared to healthy controls. In parallel, we evaluated the proportion of TH17 cells expressing CD146 (CD146+ TH17) over total TH17 cells and analyzed the potential correlations with clinical events. At last, effect of sCD146 was tested on the proportion of CD146+ TH17 among total TH17cells of patients with SSc.

Materials and Methods
Ethics statement. All Table 1. Fifty sera of age and sex matched blood controls (median age 61 (57-70) and sex ratio 10 male/40 female).

Separation of blood cells on a FICOLL density gradient. Peripheral blood mononuclear cells
(PBMCs) were prepared by centrifugation over a Ficoll-Hypaque Gradient (Histopaque; Sigma Diagnostics, St Louis, MO) and cryopreseved in 10% dimethyl sulphoxide (DMSO) in fetal bovine serum (FBS). Sera were also collected and stocked in −20° until use. Soluble CD146 detection. We measured serum sCD146 concentration using a capture enzyme-linked immunosorbent assay kit (CYQUANT ELISA, Biocytex Marseille, France) according to the manufacturer's protocol. The ELISA assay was validated previously 4 . A plastic support coated with specific mouse monoclonal anti human CD146 F(ab') 2 fragments binds to the sCD146 to be measured.
Next the mouse monoclonal antibody anti-CD146 coupled with peroxidase binds to a remaining free antigenic determinant of the CD146. The bound enzyme peroxidase is then revealed by its activity in a predetermined time on the TMB substrate. After stopping the reaction, the intensity of the signal is directly related to the concentration of sCD146 initially contained in the sample.
Quantification of IL17A. IL17A was quantified in serum of healthy controls and patients with SSc, using commercially available ELISA kit (Quantikinie ELISA human IL-17A immunoassay R&D Systems France). All the amount of cytokine present in diluted serum will be captured. After several washes to remove the excess, a polyclonal antibody bound to an enzyme specific for human IL17A is added. After washes, the reaction is revealed with a substrate (TMB), giving a color whose intensity is proportional to the amount of IL17A bound and the plates are read at 450 nm and optical density were converted to concentration using a four parametric logistic regression. mRNA quantification. RNA isolation and cDNA synthesis. Total cellular RNA was extracted from cryopreseved PBMCs using the acid guanidine-phenol-chloroform technique with Trizol (GIBCO, BRL, Bethesda, MD), according to the manufacturer's instruction. After incubation for 10 min at 65 °C, RNA samples were stored Flow cytometry analysis. PBMCs were prepared by centrifugation over a Ficoll technique and cryopserved in 10% SVF-DMSO. Thawed cells were routinely tested for viability by trypan blue exclusion. For cytokine secretion assays, PBMCs were stimulated with complete cell culture medium (Eurobio France) with 1/1500-inonomycine. -1/2000 PMA -1/3000 Golgi Stop overnight. Then cells were washed twice with 5% fetal bovine serum (FBS)-phosphate-buffered saline (PBS) prior to staining. First of all, cells were incubated for ten minutes with F c blocking (Human TrueStain FcX TM Ozyme FRANCE) as an isotype control. Then, PBMCs were stained with up to five antibodies for 30 min at room temperature. After staining, the cells were fixed with 2% formaldehyde-PBS prior to analysis on a BD canto flow cytometry (BD Bioscience, San Diego, CA, USA). Lymphocytes were gated based on forward-scatter and side-scatter properties; at least 10,000-gated events were analyzed using the Flow Software (Flow LLC Ashland, Oregon). We calculated absolute cell percentages based on flow cytometry data and complete blood counts. Six colors were used to analyze the surface phenotype and intracellular cytokine production of PBMCs. The antibodies used were CD3 PerCP Cy5/CD4 APC-Cy7/CD25 PE-Cy7/CD146 Alexa fluor 488 for cell surface staining and IL17A PE and FOXP3 APC for intracellular staining and cytokine staining. Flow cytometry analysis was performed on the canto 2 flow cytometry (BD Bioscience, San Diego, CA, USA). Lymphocytes subsets were gated by first forward versus side light scatter to identify lymphocytes, then CD4+ T cells were gated using a dot plot showing CD3 versus CD4. First of all, TH17 cells were identified using a dot plot showing CD4 and IL17A and then TH17 cells expressing CD146 were gated using a dot plot showing IL17A and CD146. Frequencies of cells were determined using statistical option in FLOWJO V10.5.3.
Flow cytometry analysis was performed on the canto 2 flow cytometry (BD Bioscience, San Diego, CA, USA). Lymphocytes subsets were gated as described above in flow cytometry analysis section.
Finally, using these cells stimulated or not with sCD146, total RNA was extracted and cDNA was synthesized as described above (2.6) and Q-RT PCR was performed using CD146 primers (Table 2). Statistical analysis. Results presented as means ± standard deviation. Statistical analyses were performed using Prism (Graph Pad Software, San Diego, CA). A p value < 0.05 was considered statistically significant. Correlation and linear regression are the most commonly used techniques for quantifying the association between two parameters. The analyses are based on linear relationship between the two variables. A scatter plot is essential before embarking on any correlation-regression analysis to show that this is indeed the case. In our study, correlation and linear regression let us study correlation between sCD146 concentrations, IL17A concentration and TH17 expressing CD146 cells percentages.

Results
Epidemiologic study. We included 50 patients with SSc from the internal medicine department of university hospital RABTA Tunis, having a score of 9 or higher according to the ACR/EULAR criteria. Our cohort was composed of 43 women and 7 men and a mean age of the onset of illness equal to 60 ± 16 years [28-68 years]. Rodnan's score ranged from one to 41 in our cohort and averaged 17. Then, clinical data highlighted several inaugural events such as Raynaud's syndrome, skin involvement, arthralgia, cutaneous sclerosis, gastro-esophageal reflux. Pulmonary Arterial Hypertension (PAH), is one of the most serious disorders of systemic sclerosis, and is diagnosed by cardiac Doppler ultrasound after exercise dyspnea. Another serious manifestation is pulmonary fibrosis demonstrated by Chest CT examination. Our cohort was constituted by 20 patients with PAH and 19 patients with pulmonary fibrosis (Table 1).

Soluble CD146 and IL17A concentrations are increased in sera of patients with SSc. Serum
concentrations of sCD146 and IL17A were measured in patients with SSc (n = 50) and compared to age -and sexmatched controls (n = 50). As shown in Fig. 1A,B, sCD146 and IL17A concentrations were significantly higher in patients with SSc as compared to healthy controls (respectively, 329.0 ± 24.90 ng/ml versus 168.4 ± 19.39 ng/ ml, p < 0, 0001 and 20.05 ± 1.97 ng/ml versus 3.32 ± 0.54 ng/ml, p < 0, 0001). Regarding clinical association, we found a significant association between low sCD146 concentrations and pulmonary fibrosis (p = 0.04) as shown in Fig. 1C, while there was no correlation with other clinical disorders described in Table 1. For IL17A, no correlation was observed with clinical disorders. Interestingly, we found a positive correlation between sCD146 and IL17A concentrations (p = <0.0001), as illustrated in Fig. 1D. The mRNA expression of RORγT and IL17A was first quantified by qRT-PCR in peripheral blood mononuclear cells (PBMCs) of patients with SSc and healthy controls. RORγT and IL17A mRNA expression were significantly higher in patients with SSc as compared to healthy controls (p = 0, 0013 and p = <0.0001) ( Fig. 2A,B). For RORγT and IL17A mRNA expression, no significant association was observed with clinical features.
In a second step, we studied the proportion of TH17 cells over the total CD4+ cells population using a flow cytometry analysis (Fig. 3A). We found a higher percentage of TH17 cells in patients with SSc (Mean 16.95 ± 0.8482), as compared to healthy controls (Mean 9.135 ± 0.2522, p < 0.05) as shown in Fig. 3B.
Then, we analyzed CD146 expression on TH17 cells, we found a higher percentage of TH17 cells expressing CD146 (CD146+ IL17A+ subset) in patients with SSc as compared to healthy controls (Fig. 4A,B). Interestingly the percentage of TH17 cells expressing CD146 was inversely correlated to the pulmonary fibrosis (4 C) but no association with other clinical features was detected.
To better understand the link between TH17 cells and sCD146, the ratio between CD146+ TH17 cells and total TH17 cells was calculated (percentage of TH17 cells expressing CD146/TH17 cells). We found a significant correlation between sCD146 concentration and this ratio (p = 0, 0135), as shown in Fig. 4D. sCD146 affects T lymphocytes in PBMCs from patients with SSc. We evaluated sCD146 effects on TH17 cells after sCD146 treatment (100 ng/ml for 24 hours) of PBMCs of patients with SSc. Interestingly, an up regulation of the percentage of TH17 cells expressing CD146 (CD146+ IL17A+ subset) was found after sCD146 treatment (Fig. 5A). The number of cells was determined at T0, and after 24 h of sCD146 treatment with 492 million +/− 6088 at T0 and 491 million +/− 6441 at T24, no significant difference was observed (p = 0.8992) attesting for the absence of cell proliferation. In addition expression of CD146 was up regulated after sCD146 treatment as demonstrated by fluorescence intensity analysis (Fig. 5B). Besides, after treatment with sCD146, mRNA expression of CD146 was increased in PBMCs treated with sCD146 as compared to untreated cells (Fig. 5C).

Discussion
In systemic sclerosis, we recently reported a major role of CD146, a molecule ubiquitously distributed on vascular cells, but also expressed on a subset of lymphocytes involved in the pathogenesis of the disease, TH17 cells. In this study we demonstrated for the time that TH17 lymphocytes expressing CD146 are increased in patients with SSc, suggesting their involvement in the pathology.
In these patients, we found a higher percentage of TH17 cells expressing CD146 and a correlation between sCD146 and IL17A serum concentrations. Interestingly the ratio of TH17 cells expressing CD146/Total TH17 cells in patients with SSc was inversely correlated to pulmonary fibrosis. Finally, in vitro treatment of TH17 cells with sCD146 led to an up regulation of CD146 expression. Afterward, PBMCs treatment with sCD146 induced an increase in the mRNA expression of CD146 as compared to untreated cells.
In a previous study, unfavorable evolution of SSc was associated with a decrease of sCD146 concentrations. We recently showed that sCD146 concentrations were increased in patients with SSc and that severe manifestations of the disease, as pulmonary fibrosis, were associated with low sCD146 concentrations 4 . Patients producing high and sustained levels of sCD146 may be protected against the disease complications, accounting for the reduction in the concentration of the soluble form in the case of pulmonary fibrosis 4 . In the present study, we confirm these data, highlighting sCD146 as a novel biomarker for the assessment of the disease. In addition, we evidence a www.nature.com/scientificreports www.nature.com/scientificreports/ significant correlation between sCD146 and IL17A concentrations. This positive correlation could explain the protective role of sCD146, interfering with the counterbalance of inflammation and fibrosis mediated by IL17A.
In patients with autoimmune disease, an increased proportion of TH17 cells were described in the peripheral circulation and at the sites of active inflammation 20 . Several studies have shown that in SSc, TH17 cells could affect the inflammatory response of the disease, and potentially the fibrosis process 21 . Our results confirm  www.nature.com/scientificreports www.nature.com/scientificreports/ a higher percentage of TH17 cells in patients with SSc and show for the first time that this increase is partially due to a higher percentage of TH17 cells expressing CD146. We propose that in patients, the up-regulation of this sub-population could be related to the sera sCD146 increase. In accordance with this hypothesis, we showed an in vitro increase of TH17 cells expressing CD146 after sCD146 treatment. This increase was attributed to an augmentation of CD146 expression since we showed an up-regulation of fluorescence intensity without TH17 proliferation. Such an increase of membrane CD146 expression has been already reported in endothelial cells 22 . We also showed that sCD146 concentrations correlate with the TH17 expressing CD146/TH17 total cells ratio in patients with SSc.
Altogether our results led us to hypothesize that sCD146 could be involved in TH17 cells differentiation and their homing to the site of the disease. Along this line, it has been described that CD146 positive TH17 cells was the sub-population of TH17 cells with an enhanced ability to bind to, and to cross through, vascular endothelium. In addition, CD146 was described as a marker of mouse NK cell maturation 23 . At last, TH17 expressing CD146 are able to migrate from the peripheral circulation to sites of active inflammation 24 . Further studies will be conducted to confirm the role of this sub-population CD146+ TH17 in the fibrotic process, to get more insights into mechanistic pathways and to investigate the potential interest for therapeutic applications.
In conclusion, TH17 cells expressing CD146 could represent a new component of the adaptive immune response and could be potential targets for therapy, opening the way for the generation of new tools for the management of SSc.