Effect of vasodilator and immunosuppressive therapy on the endothelial dysfunction in patients with systemic sclerosis

A comparative analysis of flow-mediated vasodilation (FMD), vasoactive angiogenic, and fibrogenic mediators between treatment-naive and treated systemic sclerosis (SSc) patients is an unmet need. (1)To assess the FMD and different pathogenic mediators in SSc patients about endothelial dysfunction. (2) To assess the proportion of circulating endothelial cells (CECs) in treatment-naïve patients. SSc patients were grouped into treatment-naïve (Group-I, n = 24) on vasodilator (Group-II, n = 10), on vasodilator + immunosuppressive (Group-III, n = 22)]. Age-sex matched healthy controls (n = 20) were included. Endothelial dysfunction (ED) was measured radiologically using FMD. Serum levels of NO, ET1, NO/ET1, sVCAM, sICAM, TGF, IL-6, and VEGF, as well as gene expressions of eNOS, iNOS, ET-1, and TGF, were measured to assess the status of ED in various study groups. CEC was measured in Group-I and HC. CEC was used as a marker to identify a key regulator of ED in SSc. FMD was significantly decreased in all SSc patients through receiving treatment. Upregulation of serum NO and ET concentrations was noted post-treatment with an unaltered NO/ET1 ratio. NO was positively correlated with FMD (r = 0.6) and negatively with TGFβ (r =  − 0.5). ET-1 showed a negative correlation with TGFβ (r =  − 0.5) but no significant correlation with FMD. Circulating endothelial cell (CEC) was significantly higher in Group-I (3.2%) than HC (0.8%) (p = 0.002), and it showed a good correlation with NO (r =  − 0.7, p = 0.0001) and NO/ET1 (r =  − 0.6, p = 0.007). Persistent ED was observed in all SSc patients irrespective of treatment. Dysbalance in NO/ET1 ratio might be the considering factor for the underlying progression of ED. Based on our findings, it may be hypothesized that reduced NO may be a contributing factor in the pathogenesis of endothelial dysfunction in SSc.


Introduction
The pathological hallmarks of systemic sclerosis (SSc) constitute an inter-related triad of vasculopathy, autoimmunity, and tissue remodeling. Vasculopathy in SSc is associated with tissue ischemia which is responsible for various clinical manifestations, Raynaud's phenomenon being the most common [1,2]. Vascular tonicity in healthy individuals is chiefly maintained by the vasoactive substances: vasodilator nitric oxide (NO) and vasoconstrictor endothelin-1(ET1) which are secreted by the endothelial layer. An imbalance in the production of these vasoactive substances leads to endothelial dysfunction (ED) [3]. Both NO and ET1 interact or have a concomitant regulative expression balance in maintaining vascular tone [4]. Systemic vasculopathy is the sine-quanon feature of pathogenesis that is initiated by an imbalance between NO and ET-1, associated with some other vasoactive molecules. From the pathogenic viewpoint, endothelial injury is believed to be the initial event in SSc which subsequently leads to endothelial dysfunction (ED) [5,6].
Flow-mediated dilatation (FMD) of the brachial artery by ultrasonography (USG) is a non-invasive method for the assessment of endothelial function. Impaired FMD has been reported in various rheumatic diseases and it is considered a predecessor of cardiovascular diseases. It reflects the effects of several pathophysiological events such as vessel tone regulation, cell proliferation, and inflammatory responses. Impairment of FMD in SSc patients has also been documented in the literature [7][8][9].
Downregulation of the endothelial NO synthase (eNOS) enzyme and NO bioavailability have been linked with ED. Three major isoforms of NOS catalyze NO production: neuronal (nNOS), inducible (iNOS), and endothelial (eNOS). eNOS is the major NOS isoform expressed in endothelial cells. It has been reported that with the advancement of the skin disease in SSc there is downregulation of eNOS expression, whereas iNOS is upregulated [10]. In addition to the NO and ET1 imbalance, fibrogenic cytokines, angiogenic mediators, and cell adhesion molecules have a putative role in SSc vasculopathy. Transforming growth factor-beta (TGFβ) and vascular endothelial growth factor (VEGF) can cause defective angiogenesis in SSc [11,12]. Overexpression of the adhesion molecules like soluble vascular cell adhesion molecule (sVCAM), and soluble intercellular adhesion molecule (sICAM) has been documented in SSc, and this is seen during the early part of microvascular damage of SSc [13,14]. Currently, circulating endothelial cells (CECs) have emerged as a reliable marker of ED [15]. As a consequence of endothelial injury, there is a lack of endothelial integrity, and the detached endothelial cells in the circulation can be detected as CECs. A higher proportion of CEC is seen in patients with SSc [16].
With this background, this study was conducted to measure the extent of endothelial dysfunction and the expressions of molecular markers between the treated and untreated SSc patients in comparison with healthy individuals and to evaluate the association of ED with such molecular markers. Additionally, this study also evaluated the proportion of CECs in treatment-naïve SSc patients and their association with ED. All participants were ≥ 18 years and ≤ 55 years of age and didn't have any kind of infection or did not receive any kind of antibiotics or antiviral drugs for the last one month. Inclusion criteria for patients: a) Patients (Group-I) didn't receive any vasodilator or immunosuppressives for at least a couple of months. b) Co-existence of other rheumatological diseases, current or past history of tobacco usage, and history of diabetes, hypertension, renal impairment, and pregnancy were excluded from the study. c) Disease duration ≤ 5 years.

Recruitment of study participants and sample collection
All recruited patients had interstitial lung disease (ILD). SSc patients were subcategorized based on the ongoing treatment modalities into the following groups: treatmentnaive (Group-I, n = 24), patients receiving vasodilator only (Group-II, n = 10), and patients receiving a combination of vasodilator and immunosuppressive (Group-III, n = 22). This study was approved by the institutional ethical committee of IPGME&R (No: IPGME&R/IEC/2018/533). Written, informed consent was taken from each participant. This study was done following the Declaration of Helsinki, 1964, and its later amendments.
A total of 8 ml of peripheral blood was collected from all the study participants in ethylene diamine tetra-acetic acid (EDTA) vial as well as within clot vial to isolate serum, peripheral blood mononuclear cell (PBMC), and plasma to perform the following experimental proceedings. We measured FMD to identify the comparative status of ED in different study groups. Serum levels of NO, ET1, NO/ET1, sVCAM, sICAM, TGF, IL-6, and VEGF, as well as gene expressions of eNOS, iNOS, ET-1, and master regulator of fibrosis TGF, were measured to assess the status of ED and the effect of treatment modalities in controlling the progression of ED. CEC which is a well-known marker of ED was measured in naïve (Group-I) and HC. The potent regulator of ED in SSc was identified using CEC.

Measurement of FMD of the brachial artery by ultrasonography
FMD of the brachial artery was measured by an experienced radiologist who was blinded to the participant's data. Before the FMD analysis, the subject rested for 30 min in a temperature-controlled room. Measurements were taken in the supine position. The right brachial artery above the antecubital fossa was imaged by a linear array transducer (7-12 MHz) using My Lab. 25 gold (eSaote, Italy) USG platform. The center of the artery was identified, and the clearest pictures of anterior and posterior intima layers were obtained. Intima-to-intima measurement was obtained at baseline (D1), and then, a sphygmomanometer cuff was wrapped around the forearm distal to the scanned region and inflated to 200 mm Hg for 5 min. The luminal diameter was measured again 60 secs after cuff deflation (D2). FMD was calculated based on the following formula; (D2 − D1) * 100 D2 . An FMD value lower than 4.5% was considered a cutoff for ED [17][18][19].

Measurement of nitrite by Griess assay
Supernatant from deproteinized plasma samples was then used for the determination of NO. Reduction of nitrate to nitrite was done by adding 100 μL vanadium(III) chloride (8 mg/ml) to each tube, and the Griess reagents [50 μl sulfanilamide (2%) and 50 μL0.1%N-(1-Naphthyl) ethylenediamine dihydrochloride (NEDD)] were added and incubated for 30 min at 37 °C. Absorbances were taken at 540 nm using a spectrophotometer (BIO-RAD, US). The concentration of NO in serum samples was determined from a linear standard curve established by 0-20 μmol/l sodium nitrite. [19]

Measurement of serum cytokines and cell adhesion molecules by ELISA
Blood samples were collected within a clot vial spun at 3000 Xg for 7 min to separate serum. The serum concentration of ET1, TGFβ, IL-6, sVCAM, sICAM, and VEGF was measured by using ELISA kits (KINESIS Dx, Los Angeles), according to the manufacturer's protocol.

Extraction of total RNA
One milliliter of collected blood was incubated at room temperature with RBC lysis buffer at a 1:10 ratio. Cells were pelleted down and washed twice with 1X × PBS. Cells were incubated with 1 ml of TRIzol reagent (Thermo Fisher Scientific, MA, USA) and chloroform (Himedia, Mumbai) (TRIzol: chloroform = 1: 0.25) in ice. After centrifugal phase separation, total RNA was salted out with isopropanol, washed with 70% ethanol (Himedia, Mumbai), and air-dried. Then, the air-dried RNA was dissolved in RNase-free water and stored at − 40 °C for downstream experiments.

Quantitative real-time PCR (qRT-PCR) to analyze gene expression
About 2.5 μg of total RNA was treated with 1unit of DNase-I (Roche Diagnostics, Mannheim, Germany), and the firststrand cDNA was synthesized using RevertAid Reverse Transcriptase at 42 °C following the manufacturer's protocol (Thermo Fisher Scientific, MA, USA). This cDNA was diluted (1:20) and used to quantify the expression of genes (eNOS, iNOS, ET1, TGFβ) using SyBr green (Roche Diagnostics, Mannheim, Germany) and respective primer sets in Quant studio 7 (Thermo Fisher Scientific). 18 s ribosomal RNA was used as an internal control for relative quantification of the gene expression; each reaction was run in triplicate for statistical accuracy.

Immunophenotyping of CECs by flow cytometry analysis
Peripheral blood (100 µl) was incubated at room temperature with 1 ml. RBC lysis buffer (BD Biosciences, BD Biosciences, USA.). After lysis, the cells were pelleted down and washed twice with 1 × PBS. Cells were incubated with specific antibodies: CD45 (cat no: 560777, V500; BD Pharmingen, USA), CD105 (cat no: 560819, BD Pharmingen, USA), CD31 (cat no: 560983, PE; BD Pharmingen, USA) to identify circulating endothelial cells (CECs). The percentage of circulating endothelial cell was measured by FACS. Whole blood cells were morphologically gated within characteristic linear forward scatter (FSC-A) and side scatter (SSc-A). The gated population was plotted against SSc-A and CD45. CD45-negative populations were plotted against CD105 and CD31. Cells that were positive for CD105/CD31 were gated and % were calculated as CECs.

Statistical analysis
All statistical analyses were performed using GraphPad prism v 5.0, (GraphPad Software INC, CA, and the USA) and Medcalf v 11.6, (Belgium). All normally distributed data were tested by applying the Shapiro-Wilk test. Normally distributed variables were presented as mean values with standard deviation (SD). The difference between groups was estimated by the independent t test and Mann-Whitney U test for normally distributed and skewed data, respectively.

3
Spearman (r) value was used for the correlation coefficient. All p values < 0.05 were considered statistically significant.

FMD among patients and HCs
Despite receiving the recommended standard of treatment, FMD proportion was significantly reduced in all groups of SSc patients compared to the HCs; [Group-I versus HC (p < 0.0001), Group-II versus HCs (p = 0.001), Group-III versus HCs (p < 0.0001). However, no significant differences in FMD were noted among the patient groups [p = 0.2 (Group-I vs Group-II), p = 0.7 (Group-I vs Group-III), p = 0.4 (Group-II vs Group-III)] (Fig. 1A).

Comparative serum concentration of vasoactive substances associated with maintenance of vascular tone
The NO and ET1 are the major regulative factor in regulating the vascular tone. As there is no difference in FMD (%) among different patient groups, we measured the concentration of NO and ET1. Serum NO concentration (pg/ml) was significantly  were present between the treated and untreated groups ( Fig. 2A). Serum concentration ET1 (pg/ml) showed no significant difference between the HCs and Group-I (p = 0.1). Interestingly, patients of Group-II and Group-III showed significantly higher ET1 expression than HCs (p = 0.03, and p < 0.0001, respectively) as well as those from Group-I (p = 0.001, and p < 0.0001, respectively). Inter-group analysis showed higher expression of ET1 in Group-III compared to the Group-II (p = 0.04) patients (Fig. 2B).
The ratio of serum concentration between NO and ET1 (NO/ET1) was also analyzed among the study groups, as this ratio is an important factor for maintaining vascular tone. NO/ET1 was significantly decreased in all patient groups compared to HCs (Group-I: p < 0.0001, Group-II: p = 0.04, Group-III: p = 0.001). Intergroup analysis showed no difference in NO/ET1 among different patient subgroups (Fig. 2C).  (Fig. 2D).

Comparative mRNA expression of the mediators associated with maintenance of vascular tone
Compared to the HCs, significant downregulation of iNOS was noted in patient groups [Group-I (p = 0.0002), Group-II (p = 0.0007), Group-III (p = 0.0002)]. Intergroup analysis among SSc patients did not show any significant differences in iNOS mRNA expression (Fig. 2E).
No significant differences were observed in the ET1 mRNA expression among the study groups (Fig. 2F).

Comparative account on molecular parameters associated with fibrosis, angiogenesis, and cell adhesion
The serum concentration of TGFβ was significantly higher in all patient sub-groups compared to HCs (p < 0.0001, p = 0.004, and p = 0.002for Group-I, Group-II, and Group-III, respectively). However, treated SSc patients had a significant decrease in the serum concentration of TGFβ. Group-I had a significantly higher concentration among the patient groups (p = 0.004 and p < 0.0001for Group-II and Group-III, respectively). Group-III had a significantly lower concentration of TGFβ compared to Group-II (p = 0.04) (Fig. 3A).
Concerning the serum concentration of TGF-β, its mRNA expression also showed a similar trend. All patient  (Fig. 3B).
We also assessed the serum concentration of cell adhesion molecules (VCAM and ICAM) in response to the conventional therapeutic approaches. Significant upregulation of sVCAM was observed in all patient groups compared to HC (Group-I: p = 0.01, Group-II: p = 0.03, Group-III: p = 0.01). No differences were present in the sVCAM concentration between the treated groups (Fig. 3C). sICAM showed no statistically significant differences between the study groups (Fig. 3D).
Regarding serum VEGF expression, treated groups (Group-II and Group-III) had a significantly higher level than HCs (p = 0.03, p = 0.006, respectively) and Group-I (p = 0.04, p = 0.007, respectively). No difference was found regarding serum concentration of VEGF between Group-II and Group-III (Fig. 3E).
Within the patient groups, Group-I has the highest serum concentration of IL-6 compared to HCs (p = 0.001), whereas no significant differences were found between the treated groups and HCs. Intergroup analysis of serum IL-6 concentration in SSc patients showed significant downregulation in Group-III compared to Group-I. No differences were found between Group-I and Group-II, Group-II and Group-III (Fig. 3F).

Correlation analysis between the molecules associated with endothelial dysfunction and FMD
Correlation statistics were done between NO, ET1, NO/ ET1, and the mediators of fibrosis, angiogenesis, and FMD to study the comparative impact of NO and ET-1 on ED. A strong positive correlation of FMD was noted with NO (r = 0.6, p < 0.0001) (Fig. 4A), whereas no correlation was found between FMD and ET1 (r = − 0.04, p = 0.8) (Fig. 4B). NO/ET1 had a positive correlation with FMD (r = 0.3, p = 0.02) (Fig. 4C).

The comparative proportion of CECs in HCs and Group-I patients and its correlation with FMD, NO, ET1, and NO/ET1
Percentages of CECs were significantly higher in Group-I compared to HCs (p = 0.002) (Fig. 5B). FMD showed a strong negative correlation with the proportion of CEC (r = − 0.6, p = 0.003) (Fig. 5D). A strong negative correlation was found between NO and CEC (r = − 0.7, p < 0.0001) and NO/ET-1 with CEC (r = − 0.6, p = 0.007) (Fig. 5E, G).

Discussion
SSc is a clinically heterogeneous autoimmune disorder typically characterized by vasculopathy, autoimmunity, inflammation, and fibrosis [20]. Vasculopathy is considered a leading event in the SSc pathogenesis, which subsequently acts as a connection between autoimmunity and dysregulated fibrosis. Systemic autoimmunity and environmental triggers are the possible pioneers of vasculopathy, which is characterized by both abnormal activation and damaged endothelial cells [21]. Brachial arterial FMD measurement is a surrogate marker of endothelial function, and a significant reduction in the FMD in SSc patients has been documented in the literature [9,[22][23][24]. There is a paucity of data regarding the effect of treatment on ED of SSc and the correlation among the markers of ED, vasoactive, angiogenic, and fibrogenic mediators. We stratified our SSc patients based on the therapeutic regimen and found a significant reduction in the FMD across all patient groups. It indicates that current therapeutic modalities may not be adequate to restore vasculopathy in SSc patients. This finding is also supported by the fact that currently there is no approved disease-modifying drug for SSc [25]. An imbalance between NO and ET1is already established in SSc vasculopathy [26,27]. However, the role of ET-1 is yet to be elucidated properly. The current EULAR guidelines recommend using phosphodiesterase-5 inhibitor or endothelial receptor antagonist for the management of different symptoms related to vascular insufficiency, aimed at counteracting the NO and ET-1 imbalance [28]. In reality, the symptoms of vascular insufficiency show a fluctuating course in the majority of SSc patients despite ongoing treatment strategies, and irreversible damage is seen with almost 1/5 of patients requiring digital amputation due to gangrene [29]. The impaired balance between NO and ET1 has an impact on the progression of vascular dysfunction. We found a significant downregulation of NO level in the treatment-naïve SSc patients than HCs. In the treated groups, increased levels were observed. This is probably because all patients recruited in this study were receiving phosphodiesterase-5 inhibitors as vasodilator therapy.
Interestingly, a similar increasing trend was found in serum ET-1 concentration in treated SSc patients. Vasodilator treatment to counteract the downregulated mediator of vasodilator might have some impact on the balancing mechanism within the biological system to increase the ET1 level. We did not find any significant alteration in NO/ ET1 ratio between treatment-naïve and treated groups. This result indicates that the ratio between NO and ET1 could be more important in SSc vasculopathy, rather than the absolute value of either one. This finding also explains the persistence of FMD abnormality across all SSc patients irrespective of the treatment. From the clinical viewpoint, it indicates that the current treatment modalities may not be sufficient enough to achieve the optimum benefit in SSc vasculopathy. We also showed a fair correlation of FMD with the NO and NO/ET1 ratio. In concurrence with this finding, one recent study documented the superior efficacy of combined phosphodiesterase-5 inhibitors with ET1 receptor antagonists in the management of pulmonary arterial hypertension of connective tissue disease than monotherapy [30]. eNOS is the major isoform of NOS secreted by the endothelial cells, and iNOS is mainly induced as an effect of oxidative stress. Reduced eNOS expression in SSc has been reported in the literature [31]. The gene expression of eNOS and iNOS did not vary significantly between treatment-naïve and treated patients. ( Fig: 2A and 2D). This might be linked to the underlying progression of ED despite treatment. Prospective studies may be conducted to validate the hypothesis that the current standard of treatment does not have any impact on the gene expression of eNOS and iNOS.
On the other hand, expression of the eNOS gene showed an increased pattern of expression as follows Group-III > Group-II > naïve, which might have reflected in the similar increase in serum NO concentration. Thus, this increase in NO had no effect on FMD which is dependent more on the iNOS mediated NO release.
As documented in the literature, TGFβ is considered the master regulator of fibrosis and it may have a role in the angiogenic defect seen in SSc [5,11]. TGFβ was significantly upregulated in all SSc patients. SSc patients treated with both the vasodilator and immunosuppressive agents had a lower TGFβ expression compared to the other two groups, but it was still significantly higher than HCs. Serum TGFβ was responsive to vasodilators and more to the combination of vasodilators and immunosuppressives. Simultaneously, TGFβ gene expression was significantly downregulated in treated groups than in non-treated patients. However, these favorable findings were not translated into the significant increment of the FMD in the treated patients, though FMD was numerically higher in the treated groups than in naïve patients. Treatment modalities showed effectiveness in reducing the serum IL-6 concentration which was most prominent in patients receiving immunosuppressive therapy compared to treatment-naïve patients. Interestingly IL6 was effectively downregulated in Group-II. This raises the possibility of IL6 being reactionary to ischemia rather than primary inflammation driving SSc. We also noted significant upregulation of sVCAM level but not of sICAM level in our SSc patients, and one recent study also reported a similar result [32]. According to the previous studies, TGFβinduced endothelial injury is mediated by the upregulation of sICAM, whereas it has no relationship with sVCAM [33,34]. In this regard, our data indicate the efficacy of the treatment modalities in controlling TGFβ. Significant differences were not found in the serum concentration of sVCAM and sICAM; this might help us to conclude that treatment modalities do not affect these two adhesion molecules.
CEC has gained much attention in recent years and it is considered a marker of endothelial damage [15,19]. One recently published study showed that CEC can be a direct indicator of the systemic vascular damage of SSc [35]. Considering both FMD and CEC as indicators of ED, we tried to find the CEC level in treatment-naïve SSc patients, as this group had the lowest FMD. We found a significantly higher proportion of CEC in the treatment-naïve group, and the CEC proportion showed a strong negative correlation with both FMD and NO but no correlation with ET1. These findings indicate that NO downregulation has an immense impact on the ED in SSc patients compared to ET1 upregulation. Moreover, the altered ratio of NO/ET-1 ratio is also regulated by the downregulation of NO, and persistent alteration in this ratio leads to the progression of ED in SSc.
The relatively small size of Group-II and cross-sectional design are the major limitations of our study. A prospective study with SSc patients aiming at FMD evaluation before initiation of any treatment and follow-up at different time frames after initiation of vasodilator and/or immunosuppressive treatment would be a more pragmatic approach, though Comparative % of CEC and its correlation with vasoactive molecules (NO, ET1) and NO/ET1: A and B Representative quadrant plot of immunophenotyping of CECs (%) from whole blood of healthy controls and SSc patients by flow-cytometry. Whole blood cells were morphologically gated within characteristic linear forward scatter (FSC-A) and side scatter (SSc-A). The gated population was plotted against SSc-A and CD45.CD45 − populations were plotted against CD105 and CD31. Cells that were positive for CD105/CD31 were gated and % was calculated as CECs. C Box and whisker plot showing the comparative frequency(%) of CECs in naïve SSc patients (Group-I) and healthy control. D-G A comparative account of the correlation between CECs with FMD, NO, ET1, and NO/ET1 our study will also provide valuable information regarding the additional benefit of immunosuppressive agents in addition to the vasodilators in counteracting SSc vasculopathy if any. Despite these limitations, our study has some unique aspects. A comparative assessment of the FMD in treatmentnaïve SSc and different treatment groups has been done in our study to get information regarding treatment effects on ED. Simultaneously, we also assessed different vasoactive and inflammatory mediators and their gene expression in SSc patients. Most importantly, we evaluated the CEC proportion in treatment-naïve SSc patients and showed its correlation with FMD. To the best of our knowledge, this would be the first instance when such findings are being reported.