Systemic sclerosis-associated pulmonary arterial hypertension and pulmonary fibrosis: exploring biomarker discriminators with advanced omics in a Caucasian cohort

Introduction: Systemic sclerosis (Scleroderma; SSc) is associated with high morbidity and mortality, particularly in patients with pulmonary arterial hypertension (SSc-PAH) and pulmonary fibrosis (SSc-PF). Effective risk stratification and treatment of SSc remains a significant challenge. This proof-of-concept study aimed to identify potential biomarkers capable of distinguishing between three SSc patient groups, defined by no pulmonary involvement (SSc-NLD; n=30), SSc-PAH (n=30), SSc-PF (n=30) compared to healthy controls (HC; n=30).

Methods: The study employed Olink-based proteomics using the Cardiovascular II and Immuno-oncology panels, and untargeted metabolomic profiling using Ultra-high Performance Liquid Chromatography-Tandem Mass Spectroscopy (UPLC-MS/MS), to discover distinct molecular signatures.

Results: Proteomics analysis revealed significantly elevated levels of MCP-1, MCP-3, and MCP-4 in SSc-PF compared to all other groups. However, no robust discriminatory cytokines were identified for SSc-PAH or SSc-NLD. Validation of systemic MCP-1 and IL-6 by ELISA supported the proteomics findings. IL-33 levels were found to be reduced in the SSc-PAH group. Increased levels of pro-inflammatory sIL-6R were also identified in SSc-PAH and SSc-PF, indicating shared inflammatory pathways. Protein-protein interaction analyses demonstrated greater network complexity in SSc-PF, with pathway analysis suggesting overlapping biological mechanisms across pulmonary groups. Metabolomics analysis uncovered a unique panel of metabolites altered exclusively in SSc-PAH, including quinolinate, dimethylarginines, hydroxyasparagine and orotidine. In contrast, no metabolites were uniquely discriminatory for SSc-PF or SSc-NLD. Metabolite-metabolite interaction networks revealed nicotinate and nicotinamide metabolism as the more significantly enriched metabolic pathways in SSc-PAH. Correlation analyses identified distinct protein-metabolite profiles across groups. Of note is the loss of IL-33-related metabolic associations specific to SSc-PAH.

Discussion: This study identified a candidate biomarker panel comprising three cytokines and ten metabolites capable of differentiating between SSc-PAH, SSc-PF, SSc-NLD, and HC. Biomarkers of SSc-PAH were linked to nicotinate and nicotinamide, as well as tryptophan metabolism, whereas those of SSc-PF reflected immune cell infiltration and fibrosis. These findings highlight the potential biomarker panels for diagnosis and targeted therapeutic development.

1 Introduction

Systemic sclerosis (Scleroderma; SSc) is a chronic, autoimmune disease characterized by vascular damage and activation of the immune system which triggers the excessive production and deposition of extracellular matrix (ECM) proteins. This process is driven by heightened inflammation, autoimmunity, and small vessel vasculopathy promoting abnormal connective tissue remodeling and scarring. These changes lead to progressive fibrosis affecting the skin and internal organs including the heart, kidneys, and lungs. Currently, the highest mortality rates among SSc patients are those with cardiopulmonary manifestations, accounting for ~50% of SSc-related deaths (1–3). Major pulmonary complications are systemic sclerosis-associated pulmonary artery hypertension (SSc-PAH), and systemic sclerosis-associated pulmonary fibrosis (SSc-PF) (4–6). SSc-PAH and SSc-PF are distinct disease endotypes, involving common and unique pathologic mechanisms that are not yet fully understood, leading to distinct clinical associations, predictive factors, and treatment regimens (7, 8).

SSc-PAH results from the extensive remodeling of the small pulmonary arteries, resulting in a narrowed lumen and increased pulmonary vascular resistance. The pathogenesis has been attributed to an imbalance of vasoactive mediators, pro-inflammatory cytokines, dysregulated thrombosis and disruption of several pathways, including cytokines involved in the Transforming Growth Factor-beta (TGF-β) signaling pathway. This results in altered endothelial and smooth muscle cell proliferation within the pulmonary arterioles (9, 10). There is also an indication that SSc-PAH patients exhibit distinctive, unfavorable metabolic profiles, which are attributable to oxidative stress and mitochondrial dysfunction (11, 12).

SSc-PF is an interstitial lung disease characterized by inflammation, fibrosis, and deposition of ECM within the lung parenchyma leading to diminished gas exchange and decreased vital lung capacity (5). Endothelial cell injury with subsequent vascular damage and alveolar epithelial cell injury are key initial insults that precede fibrosis. Upon injury, pro-inflammatory and pro-fibrotic mediators are released, that activate interstitial fibroblasts (13). Over time fibroblasts acquire features of smooth muscle cells and become myofibroblasts, resulting in dysregulated accumulation of collagen and other ECM components and ultimately leading to fibrosis (14). Cytokines and growth factors, such as TGF-β, interleukin-6 (IL-6) and connective tissue growth factor (CTGF), have all been implicated in the pathogenesis of the disease (15–17). Targeted anti-inflammatory and anti-fibrotic therapies such as tocilizumab, mycophenolate mofetil and nintedanib have been shown to improve patient prognosis, however these lack disease specificity (18, 19).

In the early stages of disease, the pathogenesis of SSc-PAH and SSc-PF appear to involve common mechanisms including vascular dysfunction and inflammatory processes, creating a disparity between mediators and autoimmunity (20–22). These are likely to reflect some shared factors whose overlapping roles are not completely understood, yet lead to distinct features as the diseases progress. SSc-PF can also contribute to pulmonary hypertension in SSc, due to chronic hypoxia and vessel damage from fibrosis in advanced interstitial lung disease (23). This complex overlapping of disease pathogenesis makes risk stratification, early intervention and management challenging. Thus, robust methods to predict the onset and nature of pulmonary complications associated with SSc would help in guiding early diagnosis and intervention. Moreover, specific discriminatory markers between SSc-PAH and SSc-PF could allow for the identification of new disease mechanisms and therapeutics targeting the underlying causes.

In recent years, omics-approaches including transcriptomics, proteomics and metabolomics have made significant advances in understanding disease pathogenesis and in detecting new biomarkers and therapeutic targets (24–26). A transcriptomic study of SSc-PF patients detected abnormalities in levels of IL-6, insulin growth factor binding protein-2 (IGFBP-2), insulin growth factor like-2 (IGFL-2) and toll-like receptor-8 (TLR-8) (27). In SSc-PAH, proteomics has identified changes in collagen IV, endostatin, IGFBP-2, IGFBP-7, matrix metallopeptidase-2 (MMP-2), neuropilin-1, N-terminal pro-brain natriuretic peptide and RAGE (28). More recently, using bioactive metabolomics analysis, metabolites of fatty acid oxidation, eicosanoid metabolism, and sex hormones were reported to distinguish between SSc-PAH and IPAH (29). However, to date, there have been no studies using a multi-omics approach to identify markers of SSc-PAH and SSc-PF within SSc cohorts. Therefore, the aim of this study was to use proteomics and metabolomics, to uncover systemic proteome factors and metabolites with the potential to differentiate between SSc-PAH and SSc-PF patients.

2 Materials and methods

2.1 Ethical approval and study design

The study was approved by UK nationally registered ethical committee (REC ref. 6398) and the experiments were conducted in accordance with national research guidelines.

2.2 Study recruitment

Participants were recruited from the Royal Free Hospital Scleroderma Centre, a national referral center with a database of over 2000 SSc patients and around 1000 control subjects. The SSc cohort consisted of 120 Caucasian participants categorized into four groups of equal numbers (n=30 per group): SSc-PAH, SSc-PF, SSc-NLD and HC.

2.3 Inclusion criteria for participant selection

All SSc patients recruited were diagnosed according to the EULAR 2013 classification (30). Stringent clinical criteria were used to ensure the four study groups were clearly defined. All patients selected were European Caucasian to limit population differences. Patients were also selected depending on their autoantibody profiles, as 95% of SSc patients have detectable autoantibodies which, in most cases, are specific for the disease (31) (Table 1).

Table 1

Table 1. Cohort demographics and clinical characteristics.

SSc-PAH participants were selected based on 2015 ESC/ERS Guidelines for PAH diagnosis and management with a mean pulmonary artery pressure (mPAP) >25 millimeters of mercury (mmHg) (32). A pulmonary artery wedge pressure (PAWP) of ≤15mmHg and a pulmonary vascular resistance (PVR) of >3.0 wood units (UW), measured using a right heart catheter were also used to confirm the presence of SSc-PAH. Only subjects positive for anti-centromere antibodies (ACA) and absence of significant pulmonary fibrosis were recruited.

SSc-PF patients were classified as those with >30% fibrosis of the lung when measured using high resolution computed tomography (HRCT) or 10-30% fibrosis of the lung with a forced vital capacity (FVC) <70% (33, 34). For this group, anti-topoisomerase (ATA) positive patients with an absence of significant pulmonary hypertension (measured on an echocardiogram using the tricuspid regurgitation velocity (TR vel) were selected (with cut-off value of peak TR vel<2.8 m/s) (35).

Patients in the SSc-NLD group had no evidence of pulmonary fibrosis or pulmonary arterial hypertension, as defined above, and had a disease duration of at least five years, reflecting the typical timeframe for the development of pulmonary complications (33). SSc-NLD patients could be either ATA or ACA positive. HC participants had no diagnosis of SSc or any other known disease and were not blood relatives of individuals with SSc. For the SSc-PAH and SSc-PF groups, the earliest available sample following diagnosis of pulmonary complications was analyzed.

2.4 Blood sampling and processing

Plasma and serum were extracted from participant’s whole blood samples. Blood (25ml) was taken from consenting participants using a 20-gauge needle, and serum and plasma (EDTA) were separated by centrifugation at 3000 rpm (17000 x g) for 10min at room temperature and stored at -20°C until analysis. Plasma was used for proteomic and metabolomic analyses, while serum samples were used for ELISA validation and investigation of IL-6 trans signaling and CTGF levels.

2.5 Olink proteomics analysis

Plasma protein levels were measured using the Olink Proximity Extension platform (Olink proteomics AB, Uppsala Sweden). The Cardiovascular II and Immuno-oncology 96 plex immunoassay panels were used, as they included the best selection of cytokines that are implicated in cardiovascular, inflammatory, metabolic diseases and oncology. Each panel consisted of 92 proteins (n=184). Duplicated panel analytes (n=17) were removed from the Immuno-oncology panel prior to analysis, leaving a total of 167 analytes. Protein levels were reported as Olink normalized protein expression (NPX). NPX values were calculated from cycle threshold (Ct) values and data was pre-processed to minimize variations between and within assays. A high NPX value is correlated with a high protein concentration (36).

2.6 Enzyme-linked immunosorbent assay

For targeted cytokine analysis, serum concentrations in 120 samples (n=30 per group) of IL-6, IL-33 MCP-1, CTGF, sIL-6R and sgp130 were determined using commercial ELISA (R&D systems, Oxfordshire, UK). The assay procedure was carried out according to the recommended protocols. Assay dilutions were IL-6/1:5; IL-33/no dilution; MCP-1/1:2; sIL-6R, sgp130 and CTGF/1:100.

2.7 Protein-protein interactions

Network analysis to identify protein-protein interactions between cytokines altered in each group was performed using the STRING database (https://string-db.org/). The STRING platform assesses direct physical interactions and indirect functional associations from five sources: primary curated databases, text mining, experimental interactions, co-expression interaction analyses or interactions seen in organisms with great homology (37). The network was assessed by node degree (the average number of interactions a protein has in the network), and clustering coefficients (a measure of how connected the nodes in the network are. A high clustering coefficient is representative of a highly connected network). Biologics, either FDA-approved or undergoing clinical trials, targeting these cytokines were included in the network. Pathway analysis was conducted using the gene ontology (GO) database in the Enrichr tool to identify biological and molecular pathways for cytokines altered per group (https://maayanlab.cloud/enrichr-kg).

2.8 Metabolomics analysis

A total of 120 plasma samples (n=30 per group) were analyzed by untargeted metabolomics. Sample preparation, processing and QC generation were completed at the Anti-Doping Lab Qatar (ADLQ, Doha, Qatar) while final QC review and curation were completed by Metabolon (Metabolon Inc., Durham, NC, USA). Metabolomics profiling was performed via Ultra-high Performance Liquid Chromatography-Tandem Mass Spectroscopy (UPLC-MS/MS) using established protocols. Raw data were extracted, peak-identified and quality control processed using Metabolon’s hardware and software (26). Compounds were compared to a library of >3300 purified standard compounds. Unknown compounds were also recorded and may be identified in future by acquisition of a matching purified standard or by classical structural analysis (38).

2.9 Statistics and bioinformatics

Using R Programming software, a one-way ANOVA with Tukey’s multiple comparisons test and Pearson’s correlations were used to analyze Olink proteomics (R v.4.1.3, R studio build 485). ELISAs utilized Independent-Samples Kruskal-Wallis tests with Dunn’s multiple comparisons test, and Spearman’s Rho correlations (SPSS 7.0). All significantly different proteins per group, as well as cytokines investigated by ELISA, underwent network and pathway analysis. Network analyses were performed using the STRING database. Pathway analysis was conducted using the Enrichr gene ontology (GO) database. Statistical significance was determined using Fischer exact test. Benjamini-Hochberg corrections were applied and a p-value <0.05 was considered significant throughout.

Prior to analysis, unnamed metabolites were excluded from the dataset. Principal Component Analysis (PCA) was performed. Unpaired t-tests were subsequently conducted to compare each group against the remaining three groups. To account for multiple comparisons, the Benjamini-Hochberg correction was applied, with an adjusted p-value <0.01 considered significant. Metabolites were retained only if they exhibited significant dysregulation in the target group compared to all other groups, without showing significant directional dysregulation (up or down) in any other group. Finally, the retained metabolites were validated by training a Random Forest model to confirm their predictive capability. All metabolomics analyses were conducted using Python (v3.10.19) and the scikit-learn library.

Metabolite network analysis and pathway enrichment was performed using the MetaboAnalyst 6.0 software (39). Protein-metabolite correlations utilized Pearson’s correlation test. A p-value <0.05 was used to determine the pathway analysis and correlation significance. All graphs were created either using R software or GraphPad Prism (9.4.0).

3 Results

Key cohort demographics and clinical characteristics are shown in Table 1.

3.1 SSc blood proteome profiling

Analysis of plasma samples revealed that the 167 Olink protein analytes were detected in samples examined. Using one-way ANOVA, 12 cytokines were found to be significantly altered in the four groups. These were IL-10, IL-12, IL-16, MCP-1, MCP-3, MCP-4, IL-1RL2, IL-33, IL-6, IL-8, IL-1RA and TNFRSF4. Cytokines found to be significantly elevated in more than one group were excluded from further analysis, as they would be non-specific, leaving 8 cytokines (Figure 1). IL-10 was significantly elevated in SSc-NLD compared to HC but not when compared to SSc-PAH or SSc-PF. IL-12 was significantly elevated in SSc-PAH compared to SSc-NLD and HC but not SSc-PF. IL-16 was significantly elevated in SSc-PAH when compared to HC but not SSc-NLD or SSc-PF. MCP-1, MCP-3 and MCP-4 were all significantly and specifically elevated in SSc-PF compared to all other groups, however IL1RL2 and IL-33 were significantly elevated in SSc-PF only when compared to HC.

Figure 1

Figure 1. Distinct cytokines are differentially expressed in SSc disease cohorts. Relative cytokine levels identified by proteomic analysis are shown for healthy controls (HC), systemic sclerosis with no lung disease (SSc-NLD), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH), and systemic sclerosis with pulmonary fibrosis (SSc-PF). Data are illustrated as box plots showing the median and interquartile ranges. Cytokines predominantly altered in each subgroup are shown as follows: (A) SSc NLD (1. IL-10); (B) SSc PAH (1. IL-12, 2. IL-16); and (C) SSc PF (1. MCP-1, 2. MCP-3, 3. MCP-4, 4. IL1RL2, 5. IL-33). Statistical significance is indicated by *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. NPX denotes Olink protein normalization units (AU).

3.2 Validation of selected biomarkers and their inter-relatedness in the SSc sub-sets

Serum levels of IL-6, sIL-6R, sgp130, MCP-1, IL-33 and CTGF were measured by ELISA in the four groups (Figure 2). MCP-1 was significantly elevated in SSc-PF compared to all other groups. IL-33 was significantly reduced in SSc-PAH compared to HC. IL-6 was significantly upregulated in all SSc groups compared to HC, and in SSc-PF when compared to SSc-NLD, but no significant difference was found in IL-6 concentrations between SSc-PAH and SSc-PF. With IL-6 trans signaling molecules, sIL-6R was significantly elevated in both SSc-PAH and SSc-PF compared to HC but not when compared to SSc-NLD. There was no significant difference in the levels of the sgp130 and CTGF between groups (data not shown). Interactions between these cytokines were analyzed using Spearman’s correlation analysis (Figure 3). When all participants were pooled together, there was a significant positive correlation between IL-6 and sIL-6R, and between IL-6 and MCP-1. In the SSc-NLD group, sgp130 was found to be significantly positively correlated with IL-6. In the SSc-PF group, MCP-1 was significantly positively correlated with CTGF.

Figure 2

Figure 2. Targeted cytokine profiles in SSc disease cohorts. Cytokine concentrations (ng/ml) identified by ELISA are shown for healthy controls (HC), systemic sclerosis with no lung disease (SSc-NLD), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH), and systemic sclerosis with pulmonary fibrosis (SSc-PF). Data are illustrated as box plots showing the median and interquartile ranges. Cytokines measured were (A) IL-6, (B) sIL-6R, (C) MCP-1 and (D) IL-33. Statistical significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 3

Figure 3. Correlation analysis of targeted cytokines in SSc disease groups. Linear regression analyses of cytokine concentrations (ng/ml) are shown for all groups, systemic sclerosis with no lung disease (SSc-NLD) and systemic sclerosis with pulmonary fibrosis (SSc-PF). Graphs denote significant correlations between IL-6 and sIL-6R (A1) and between IL-6 and MCP-1 (A2) across the entire cohort. Subgroup analyses demonstrate significant correlations between IL-6 and sgp130 in SSc-NLD (B), and between CTGF and MCP-1 in SSc-PF (C). A p value <0.05 was considered statistically significant. The correlation coefficient (r) and p-value (p) and are shown for each analysis.

3.3 Protein-protein interactions and protein pathway analysis in SSc

Network analysis identified SSc-NLD as the group with the lowest number of interactions (node degree = 3.67) but a high degree of connectivity between proteins (clustering coefficient = 0.839) (Figure 4A.1). The SSc-PAH group had a node degree of 4.0 and had the lowest clustering coefficient (0.713) (Figure 4B.1). The SSc-PF group had the highest node degree (6.67) and a clustering coefficient of 0.794 (Figure 4C.1). Analysis of current FDA-approved biologics showed that only tocilizumab is routinely used therapeutically in SSc. Several other biologics, such as siltuximab, spesolimab, and anakinra, may potentially be repurposed for treatment of SSc, both in the presence and absence of pulmonary complications (Figures 4A.1-C.1).

Figure 4

Figure 4. Protein-protein interactions and biological and molecular pathway analysis. Network analysis identified using the STRING database for systemic sclerosis with no lung disease (SSc-NLD; A1), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH; B1), and systemic sclerosis with pulmonary fibrosis (SSc-PF; C1). The network also includes FDA approved biologics and mAbs undergoing clinical trials (*) for cytokines selected for each group. Colored nodes indicate different cytokines, edge colors represent type of evidence: Experimental (purple), Text-mining (yellow), Database (blue) and Co-expression (black). The image illustrates tables of significantly enriched biological pathways for SSc-NLD (A2), SSc-PAH (B2) and SSc-PF (C2), with IL-6 related pathways such as T-cell proliferation and Regulation of Janus kinase (JAK) activity being the most significantly enriched. The image also denotes tables of the most significantly enriched molecular pathways for SSc-NLD (A3), SSc-PAH (B3) and SSc-PF (C3), with Tumor Necrosis Factor Receptor (TNFR) activity and IL-6 receptor binding being the most significantly enriched. A p value <0.05 was considered statistically significant.

Pathway analysis identified T cell proliferation and tumor necrosis factor receptor (TNFR) activity as the most significantly enriched biological and molecular pathways in SSc-NLD. For the SSc-PAH and SSc-PF groups, the most significantly enriched pathways were activation of Janus kinase (JAK) activity, and IL-6R binding (Figures 4A.2.3-C.2.3).

3.4 Untargeted metabolomics

A total of 1189 metabolites were detected, comprising 1001 compounds of known identity (named metabolites) and 188 compounds of unknown structural identity (unnamed metabolites). Unknown metabolites were not included in any further analysis. Additionally, 36 known metabolites were removed due to not fulfilling assay QC criteria. The final analysis performed included 965 known metabolites.

3.5 Identification of metabolites and metabolite profiles in SSc

PCA revealed a modest separation of the SSc-PAH group from the other cohorts, whereas the SSc-PF, SSc-NLD, and HC groups exhibited significant overlap (Figure 5). Initial univariate testing identified 122, 37, and 27 differentially expressed metabolites in the SSc-PAH, SSc-PF, and SSc-NLD groups, respectively. However, following the application of strict inclusion criteria to isolate group-specific signatures, only 10 metabolites were retained for the SSc-PAH group, while no metabolites met the criteria for the SSC-PF or SSC-NLD groups (Figure 6). The SSc-PAH group exhibited a modest spatial separation from the other cohorts, whereas the SSc-PF, SSc-NLD, and HC groups showed considerable overlap.

Figure 5

Figure 5. Principal Component Analysis (PCA) 3D score plot of metabolite profiles. The plot illustrates the spatial distribution of samples across the first three principal components (PC1, PC2, PC3), explaining 7.4%, 6.9%, and 5.7% of the variance, respectively. Colored points represent individual samples from the study groups: healthy controls (HC; blue), systemic sclerosis with no lung disease (SSc-NLD; orange), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH; green), and systemic sclerosis with pulmonary fibrosis (SSc-PF; purple). Shaded ellipses indicate 95% confidence regions.

Figure 6

Figure 6. Metabolites selectively increased in SSc-PAH. Relative metabolite levels identified by metabolomics are shown for healthy controls (HC), systemic sclerosis with no lung disease (SSc-NLD), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH), and systemic sclerosis with pulmonary fibrosis (SSc-PF). Data are illustrated as box plots showing the median and interquartile ranges of metabolites significantly elevated in the SSc-PAH group compared with SSc-PF, SSc-NLD and HC. Metabolites include dimethylarginines (SDMA+ADMA; A), quinolinate (QUIN; B), N6-carbamoylthreonyladenosine (T6A; C), vanillactate (VLA; D), 2,3-dihydroxy-5-methylthio-4-pentenoate (DMTPA; E) N-acetylalanine (AYA; F), N-acetylserine (G), N-acetyl-isoputreanine (H), hydroxyasparagine (I), orotidine (J). Statistical significance is indicated as **p<0.01, ***p<0.001, ****p<0.0001.

These 10 retained metabolites were subsequently used to train a Random Forest classifier to evaluate their predictive capability (Figure 7). The model demonstrated robust performance in identifying the SSc-PAH group, achieving a precision of 0.70, a recall of 0.78, and an F1-score of 0.74. In contrast, the model exhibited poor predictive power for the remaining groups, with F1-scores ranging from 0.00 to 0.27 (Table 2).

Figure 7

Figure 7. Confusion matrix for the Random Forest classifier. Matrix showing healthy controls, systemic sclerosis with no lung disease (SSc-NLD), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH), and systemic sclerosis with pulmonary fibrosis (SSc-PF). The heatmap displays the classification performance of the Random Forest model trained on the 10 selected metabolites. The y-axis represents the true clinical diagnosis, while the x-axis represents the predicted class. Darker blue cells along the diagonal indicate higher numbers of correct predictions. The cohort was split into training and testing groups - with 21 of each group acting as training group and the remaining (n=9) as the testing cohort.

Table 2

Table 2. Performance metrics of the Random Forest model for group classification.

3.6 Metabolite pathway analysis in SSc

Pathway enrichment analysis was performed using all metabolites altered in SSc-PAH to provide greater insight into dysregulated metabolic pathways within this group. An enrichment ratio was calculated by computing hits/expected, where hits = observed hits; expected = expected hits (39). The most enriched pathways in SSc-PAH were those regulating the metabolism of nicotinate and nicotinamide and tryptophan (Figure 8). The metabolites involved in these pathways were quinolinate (QUIN), 1-methylnicotinamide, nudifloramide, xanthurenic acid and kynurenine (KYN).

Figure 8

Figure 8. Enriched pathway interaction networks for SSc associated metabolites. Pathway enrichment analysis identified using MetaboAnalyst 6.0 for the systemic sclerosis with pulmonary arterial hypertension (SSc-PAH) group. Enriched pathway network for all metabolites significantly altered in the SSc-PAH group (A). Nodes indicate different metabolic pathways, size of node refers to the extent of enrichment, while the color indicates significance, with lowest p-values (high significance) in red, and highest p-values in yellow (low significance). The nicotinate and nicotinamide metabolism and tryptophan metabolism pathways exhibited the most metabolite hits. A p-value <0.05 was considered significant (B).

3.7 Cytokine-metabolite interactions

Association matrix patterns revealed distinct protein–metabolite profiles that were broadly comparable across the four groups (Figure 9). However, a marked disruption in IL-33-related associations was observed in the SSc-PAH group. Of the 10 metabolites analyzed, 9 showed negative correlations with IL-33 in HC and SSc-NLD. These associations were reduced in SSc-PF (4 negatively correlating metabolites) and largely lost in SSc-PAH, with IL-33 exhibiting more positive correlations. Moreover, a negative correlation between IL-33 and IL-12 emerged in the pulmonary groups, which was absent in both SSc-NLD and HC.

Figure 9

Figure 9. Correlation matrix of cytokines and metabolites. Correlation matrices illustrating relationships between cytokines and metabolites across study groups. Pearson correlation matrices are shown for healthy controls (HC), systemic sclerosis with no lung disease (SSc-NLD), systemic sclerosis with pulmonary arterial hypertension (SSc-PAH), and systemic sclerosis with pulmonary fibrosis (SSc-PF). Red indicates positive correlations and blue indicates negative correlations, with color intensity proportional to correlation strength. A p value <0.05 was considered statistically significant.

4 Discussion

Pulmonary complications associated with SSc (SSc-PAH and SSc-PF) are the leading cause of disease-related death. Both conditions have complex etiologies which are not fully understood. Available therapeutics aim to alleviate symptoms, but do not resolve the underlying inflammatory and metabolic dysfunction. Identification of discriminatory biomarkers and novel therapeutic targets would improve patient prognosis and outcomes. This is the first study to produce a combined cytokine-metabolite panel which may potentially distinguish between SSc-PAH and SSc-PF.

MCP-1, MCP-3, and MCP-4 were identified as discriminatory markers for SSc-PF, as their levels were significantly elevated in this group compared to others. These chemokines are members of the monocyte chemoattractant protein family, and are strong chemotactic factors, recruiting immune cells to sites of infection, inflammation, and stress. MCP-1 has been associated with initiation of inflammation in SSc. High levels of its receptor, CCR2, has been detected on CD4⁺ T cells within the fibrotic lung of bleomycin-treated mice (40, 41). Elevated serum MCP-1 has also been shown to correlate with the development of pulmonary fibrosis in SSc patients (42). Similarly, serum MCP-3 levels in SSc patients levels have been associated with greater skin fibrosis and reduced FVC, suggesting a central role in fibrogenesis (43). MCP-4, while not as extensively studied as other members of the MCP family, has been reported to be expressed in cells of the bronchoalveolar lavage and may also contribute to fibrotic disease processes. However, members of the MCP family have also been reported to be elevated in other inflammatory pulmonary conditions, such as idiopathic pulmonary fibrosis (IPF). To date, proteomic studies in IPF have not consistently identified MCPs as established biomarkers. A direct comparison between IPF and SSc associated pulmonary fibrosis would therefore be required to determine whether MCPs represent a robust and disease-specific biomarker (44, 45).

The discriminatory value of MCP-1 between SSc-PAH and SSc-PF was validated in this study by ELISA. IL-33, IL-6, sIL-6R, sgp130 and CTGF, while not found to distinguish between the pulmonary groups, were also investigated. IL-33 levels were found to be reduced in SSc-PAH compared to HC. In trans-signaling, IL-6 bound to its soluble receptor, sIL-6R, and the signal transducer gp130 to promote inflammation, while soluble gp130 blocks this interaction and dampens the response. In this study, IL-6 and sIL-6R were found to be elevated in both SSc-PAH and SSc-PF. However, this may occur through distinct pathogenic pathways in the disease subsets, as suggested by the efficacy of anti-IL-6R therapy with tocilizumab in SSc-PF, but not in SSc-PAH (46, 47). Interestingly, in the SSc-NLD group, there was a significant positive correlation between IL-6 and sgp130. This is potentially a protective mechanism, halting progression and development of pulmonary complications. CTGF has also been implicated in the pathogenesis of SSc. Representational difference analysis (RDA) demonstrated increased CTGF gene transcription in fibroblasts isolated from SSc patients (48). In addition, selective deletion of CTGF in mouse models of PAH and PF significantly reduced pulmonary interstitial scarring and vascular remodeling (49). While no significant differences in CTGF levels were observed across patient groups in this study, there was a strong positive correlation between CTGF and MCP-1 in the SSc-PF group, suggesting that CTGF signaling may contribute to this disease pathway.

Network analysis identified SSc-PF as the group with the highest node degree, indicating that cytokines within this subset exhibited a greater number of interactions with one another. This finding aligns with previous studies suggesting that SSc-PF represents a more inflammation-driven phenotype. Collectively, the cytokines involved and their interconnectedness point to a state of inflammatory dysregulation within this group. Of the seven FDA-approved drugs targeting the cytokines within these networks, four (tocilizumab, sarilumab, siltuximab and satralizumab) target the IL-6 signaling pathway and have been trialed in SSc with varying degrees of success. Anakinra is an anti-IL-1 inhibitor. While increasing evidence suggests that IL-1-mediated inflammation may play a role in SSc-associated cardiac failure, to date, there are no reports of its use in SSc-related disease (50). Ustekinumab targets the p40 subunit, shared by IL-12 and IL-23, and has demonstrated efficacy in immune mediated conditions such as psoriasis. However, there is currently no evidence supporting its use in SSc (51). Spesolimab is a monoclonal antibody targeting the IL-36 receptor and has shown clinical benefit in the treatment of generalized pustular psoriasis; however, its role in SSc has not been investigated (52). Repurposing these drugs may potentially offer a wider range of therapeutic options for patients.

A 10-metabolite signature was uncovered for SSc-PAH. Dimethylarginines (SDMA + ADMA together), Hydroxyasparagine, QUIN, AYA, T6A, VLA, DMTPA, N-acetylserine, N-isoputreanine and Orotidine were significantly elevated in this group compared to all others. These metabolites are important for energy metabolism, mitochondrial oxidation, amino acid catabolism, and inflammation. Dimethylarginine (ADMA) is known to regulate vascular tone, response to injury, and is a known inhibitor of nitrous oxide (NO) synthase. Decreased levels of NO is a key driver of chronic inflammation and cellular disruption, as it leads to endothelial dysfunction. ADMA levels have been shown to be elevated in the plasma of IPAH and chronic thromboembolic pulmonary hypertension (CETPH) patients compared to healthy controls, and this correlated with low endothelial NOS expression (53, 54). Clinically, serum ADMA has been shown to positively correlate with mPAP and pulmonary vascular resistance, and so could serve as a biomarker for monitoring the effectiveness of treatment (55). However, ADMA was not measured separately in this study.

Hydroxyasparagine is important in post-translational modification of calcium-binding epidermal growth factor (cbEGF)-like domains. These domains are present in a wide variety of extracellular proteins such as fibrillin. Fibrillin has been shown to induce the trans differentiation of resident fibroblasts into myofibroblasts, which produce ECM proteins and pro-inflammatory cytokines that exacerbate inflammation (56). This is a key inflammatory pathway involved in regulating chemotaxis of inflammatory cells, which can lead to inflammatory cascades within the vasculature (57).

High levels of QUIN, a metabolite formed through the kynurenine pathway (KP) of tryptophan metabolism, has been associated with lower survival rates in treatment naïve PAH patients. Interestingly, in human lung primary cells, treatment with IL-6/sIL-6R complex lead to similar KP metabolic profiles as in PAH, suggesting that the metabolic dysfunction within the KP is driven by IL-6 trans signaling (58).

N-acetyl-isoputreanine is a by-product in the catabolism and metabolism of spermidine. Spermidine levels have been shown to be elevated in IPAH patients, and may promote proliferation and migration of PASMCs, encouraging vascular remodeling in rat pulmonary hypertensive models (59). Vanillactate (VLA), N-acetylalanine, T6A, DMTPA, N-acetyl-serine and orotidine were also shown to be elevated in the plasma of SSc-PAH patients compared to the other three groups. However, none of these have been well-characterized in PAH, and so further research into the mechanisms of these metabolites is necessary. However, these metabolites may still act as biomarkers if the results are reproduced. The robustness of the identified metabolites were evaluated using Random Forest. The analysis demonstrated that the identification of metabolite markers in the SSc-PAH group was significantly robust compared to other groups.

Nicotinate and nicotinamide metabolism was the most significantly affected metabolic pathway in SSc-PAH. Nicotinate and nicotinamide are precursors of the co-enzyme’s nicotinamide-adenine dinucleotide (NAD+) and NADP +. CD38, which hydrolyses NAD and nicotinamide, is overexpressed in skin of patients with diffuse SSc compared to healthy individuals. Depletion of NAD+ by CD38 impairs the ability for PARP to repair DNA lesions which encourage pro-inflammatory and pro-fibrotic phenotypes (60). Enrichment of the nicotinate and nicotinamide pathways may therefore imply an imbalance of the resolution of inflammation and fibrosis.

In this study, most of the protein-metabolite, protein-protein and metabolite-metabolite correlations were found to be comparable across groups. However, several notable differences were observed. In HC and SSc-NLD groups, IL-33 showed strong negative correlations with most metabolites, consistent with its role in maintaining metabolic homeostasis through ST2-mediated regulation of key metabolic checkpoints (61). These associations weakened in SSc-PF, indicating diminished IL-33–mediated control of immune–metabolic balance, and were completely lost in SSc-PAH, suggesting disruption of this pathway. Reduced systemic IL-33 levels have been reported in idiopathic PAH, suggesting a loss of its protective function. IL-33 has been implicated in cardiopulmonary protection by reducing inflammation and limiting vascular remodeling, mediated primarily via the ST2L receptor. Notably, IL-33 exhibits context dependent activity and can either activate, promoting inflammation, or reduce resolving the inflammatory response. In this context, our findings of reduced IL-33 levels in PAH may reflect a loss of protective IL-33 signaling (62–64). Moreover, IL-33 and IL-12 were not correlated in HC, but positively associated in SSc-NLD, likely reflecting early immune activation where IL-33 release enhances IL-12 production (65). In contrast, the pulmonary groups exhibited a negative correlation, suggesting immune polarization, where chronic IL-33 signaling suppresses IL-12 activity, driving a Th2-dominant, fibrotic environment (66).

In conclusion, this proof-of-concept study proposes a combined cytokine–metabolite signature which may potentially distinguish between SSc-PAH and SSc-PF. While the biomarker panel identified in this study largely aligns with previous findings, it may also hold novel diagnostic, prognostic, and therapeutic potential. However, its clinical applicability requires validation in a larger, independent cohort.

Key limitations of this proof-of-concept study include the small sample size and the absence of detailed information on comorbidities and treatment regimens. Although the cohorts were ethnicity and age matched, there were sex differences between the groups, reflecting the higher prevalence of SSc in women. As a result, potential sex related differences may not have been fully captured in the analysis. Finally, differences in medication in the groups may play a role in the metabolite signatures of each of these groups and the possibility that these metabolite profiles may be, at least in part, a medication-associated phenomenon cannot be discounted.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by University College London (REC ref. 6398). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

NM-A: Writing – original draft, Writing – review & editing. VA: Writing – original draft, Writing – review & editing. MA-J: Writing – review & editing. RB: Writing – review & editing. IA-M: Writing – review & editing. KS: Writing – review & editing. AB: Writing – review & editing. DK: Writing – review & editing. CD: Writing – review & editing. VO: Writing – review & editing. MA-N: Writing – review & editing. DA: Writing – review & editing. MA-M: Writing – review & editing. MP: Writing – review & editing. VM-A: Writing – review & editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Pope JE, Denton CP, Johnson SR, Fernandez-Codina A, Hudson M, Nevskaya T, et al. State-of-the-art evidence in the treatment of systemic sclerosis. Nat Rev Rheumatol. (2023) 19:212–26. doi: 10.1038/s41584-023-00909-5

PubMed Abstract | Crossref Full Text | Google Scholar

2. Volkmann ER and Fischer A. Update on morbidity and mortality in systemic sclerosis-related interstitial lung disease. J Scleroderma Relat Disord. (2021) 6:11–20. doi: 10.1177/2397198320915042

PubMed Abstract | Crossref Full Text | Google Scholar

3. Komócsi A, Vorobcsuk A, Faludi R, Pintér T, Lenkey Z, Költő G, et al. The impact of cardiopulmonary manifestations on the mortality of systemic sclerosis: a systematic review and meta-analysis of observational studies. Rheumatol (Oxford). (2012) 51:1027–36. doi: 1093/rheumatology/ker357

PubMed Abstract | Crossref Full Text | Google Scholar

4. Bousseau S, Sobrano Fais R, Gu S, Frump A, and Lahm T. Pathophysiology and new advances in pulmonary hypertension. BMJ Med. (2023) 2:e000137. doi: 10.1136/bmjmed-2022-000137

PubMed Abstract | Crossref Full Text | Google Scholar

5. Liakouli V, Ciancio A, Del Galdo F, Giacomelli R, and Ciccia F. Systemic sclerosis interstitial lung disease: unmet needs and potential solutions. Nat Rev Rheumatol. (2024) 20:21–32. doi: 10.1038/s41584-023-01044-x

PubMed Abstract | Crossref Full Text | Google Scholar

6. Cueto-Robledo G, Tovar-Benitez D, Alfaro-Cruz A, and Gonzalez-Hermosillo LM. Systemic scleroderma: review and updated approach and case description to addressing pulmonary arterial hypertension and idiopathic pulmonary fibrosis: a dual challenge in treatment. Curr Probl Cardiol. (2024) 49:102404. doi: 10.1016/j.cpcardiol.2024.102404

PubMed Abstract | Crossref Full Text | Google Scholar

7. Denton CP, Wells AU, and Coghlan JG. Major lung complications of systemic sclerosis. Nat Rev Rheumatol. (2018) 14:511–27. doi: 10.1038/s41584-018-0062-0

PubMed Abstract | Crossref Full Text | Google Scholar

8. Denton CP and Khanna D. Systemic sclerosis. Lancet. (2017) 390:1685–99. doi: 10.1016/S0140-6736(17)30933-9

PubMed Abstract | Crossref Full Text | Google Scholar

9. Bahi M, Li C, Wang G, and Korman BD. Systemic sclerosis-associated pulmonary arterial hypertension: from bedside to bench and back again. Int J Mol Sci. (2024) 25:4728. doi: 10.3390/ijms25094728

PubMed Abstract | Crossref Full Text | Google Scholar

10. Alamri AK, Ma CL, and Ryan JJ. Novel drugs for the treatment of pulmonary arterial hypertension: where are we going? Drugs. (2023) 83:577–85. doi: 10.1007/s40265-023-01862-z

PubMed Abstract | Crossref Full Text | Google Scholar

11. Deidda M, Piras C, Cadeddu Dessalvi C, Locci E, Barberini L, Orofino S, et al. Distinctive metabolomic fingerprint in scleroderma patients with pulmonary arterial hypertension. Int J Cardiol. (2017) 241:401–6. doi: 10.1016/j.ijcard.2017.04.024

PubMed Abstract | Crossref Full Text | Google Scholar

12. Paulin R and Michelakis ED. The metabolic theory of pulmonary arterial hypertension. Circ Res. (2014) 115:148–64. doi: 10.1161/CIRCRESAHA.115.301130

PubMed Abstract | Crossref Full Text | Google Scholar

13. Mouawad JE and Feghali-Bostwick C. The molecular mechanisms of systemic sclerosis-associated lung fibrosis. Int J Mol Sci. (2023) 24:2963. doi: 10.3390/ijms24032963

PubMed Abstract | Crossref Full Text | Google Scholar

14. Schoenfeld SR and Castelino FV. Interstitial lung disease in scleroderma. Rheum Dis Clin North Am. (2015) 41:237–48. doi: 10.1016/j.rdc.2014.12.005

PubMed Abstract | Crossref Full Text | Google Scholar

15. Munger JS, Huang X, Kawakatsu H, Griffiths MJ, Dalton SL, Wu J, et al. A mechanism for regulating pulmonary inflammation and fibrosis: the integrin αvβ6 binds and activates latent TGF-β1. Cell. (1999) 96:319–28. doi: 10.1016/S0092-8674(00)80545-0

PubMed Abstract | Crossref Full Text | Google Scholar

16. Hasegawa M, Sato S, Ihn H, and Takehara K. Enhanced production of interleukin-6, oncostatin M and soluble IL-6 receptor by cultured peripheral blood mononuclear cells from patients with systemic sclerosis. Rheumatol (Oxford). (1999) 38:612–17. doi: 10.1093/rheumatology/38.7.612

PubMed Abstract | Crossref Full Text | Google Scholar

17. Sato S, Nagaoka T, Hasegawa M, Tamatani T, Nakanishi T, Takigawa M, et al. Serum levels of connective tissue growth factor are elevated in patients with systemic sclerosis: association with extent of skin sclerosis and severity of pulmonary fibrosis. J Rheumatol. (2000) 27:149–54.

PubMed Abstract | Google Scholar

18. Higuero Sevilla JP, Memon A, and Hinchcliff M. Learnings from clinical trials in patients with connective tissue disease-associated interstitial lung disease. Arthritis Res Ther. (2023) 25:118. doi: 10.1186/s13075-023-03090-y

PubMed Abstract | Crossref Full Text | Google Scholar

19. Daoussis D and Liossis SN. Treatment of systemic sclerosis-associated fibrotic manifestations: current options and future directions. Mediterr J Rheumatol. (2019) 30:33–7. doi: 10.31138/mjr.30.1.33

PubMed Abstract | Crossref Full Text | Google Scholar

20. Chaisson NF and Hassoun PM. Systemic sclerosis-associated pulmonary arterial hypertension. Chest. (2013) 144:1346–56. doi: 10.1378/chest.12-2396

PubMed Abstract | Crossref Full Text | Google Scholar

21. Le Pavec J, Humbert M, Mouthon L, and Hassoun PM. Systemic sclerosis-associated pulmonary arterial hypertension. Am J Respir Crit Care Med. (2010) 181:1285–93. doi: 10.1164/rccm.200909-1331PP

PubMed Abstract | Crossref Full Text | Google Scholar

22. Shah FA, Mahler N, Braford M, and Greene N. Limited scleroderma-induced pulmonary arterial hypertension resulting in impaired postoperative respiratory function. Cureus. (2021) 13:e13742. doi: 10.7759/cureus.13742

PubMed Abstract | Crossref Full Text | Google Scholar

23. Naranjo M and Hassoun PM. Systemic sclerosis-associated pulmonary hypertension: spectrum and impact. Diagnostics (Basel). (2021) 11:911. doi: 10.3390/diagnostics11050911

PubMed Abstract | Crossref Full Text | Google Scholar

24. Mehta BK, Espinoza ME, Hinchcliff M, and Whitfield ML. Molecular “omic” signatures in systemic sclerosis. Eur J Rheumatol. (2020) 7:S173–80. doi: 10.5152/eurjrheum.2020.19192

PubMed Abstract | Crossref Full Text | Google Scholar

25. Álvez MB, Edfors F, von Feilitzen K, Zwahlen M, Mardinoglu A, Edqvist PH, et al. Next generation pan-cancer blood proteome profiling using proximity extension assay. Nat Commun. (2023) 14:4308. doi: 10.1038/s41467-023-39765-y

PubMed Abstract | Crossref Full Text | Google Scholar

26. Ullah E, El-Menyar A, Kunji K, Elsousy R, Mokhtar HRB, Ahmad E, et al. Untargeted metabolomics profiling reveals perturbations in arginine–NO metabolism in Middle Eastern patients with coronary heart disease. Metabolites. (2022) 12:517. doi: 10.3390/metabo12060517

PubMed Abstract | Crossref Full Text | Google Scholar

27. Renaud L, da Silveira WA, Takamura N, Hardiman G, and Feghali-Bostwick C. Prominence of IL6, IGF, TLR, and bioenergetics pathway perturbation in lung tissues of scleroderma patients with pulmonary fibrosis. Front Immunol. (2020) 11:383. doi: 10.3389/fimmu.2020.00383

PubMed Abstract | Crossref Full Text | Google Scholar

28. Bauer Y, de Bernard S, Hickey P, Ballard K, Cruz J, Cornelisse P, et al. Identifying early pulmonary arterial hypertension biomarkers in systemic sclerosis: machine learning on proteomics from the DETECT cohort. Eur Respir J. (2020) 57(6):2002591. doi: 10.1183/13993003.02591-2020

PubMed Abstract | Crossref Full Text | Google Scholar

29. Alotaibi M, Shao J, Pauciulo MW, Nichols WC, Hemnes AR, Malhotra A, et al. Metabolomic profiles differentiate scleroderma-PAH from idiopathic PAH and correspond with worsened functional capacity. Chest. (2022) 163(1):204–15. doi: 10.1016/j.chest.2022.08.2230

PubMed Abstract | Crossref Full Text | Google Scholar

30. van den Hoogen F, Khanna D, Fransen J, Johnson SR, 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) 65:2737–47. doi: 10.1002/art.38098

PubMed Abstract | Crossref Full Text | Google Scholar

31. Ho KT and Reveille JD. The clinical relevance of autoantibodies in scleroderma. Arthritis Res Ther. (2003) 5:80–93. doi: 10.1186/ar628

PubMed Abstract | Crossref Full Text | Google Scholar

32. Galiè N, Humbert M, Vachiery JL, Gibbs S, Lang I, Torbicki A, et al. 2015 ESC/ERS guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Heart J. (2016) 37:67–119. doi: 10.1093/eurheartj/ehv317

PubMed Abstract | Crossref Full Text | Google Scholar

33. Solomon JJ, Olson AL, Fischer A, Bull T, Brown KK, Raghu G, et al. Scleroderma lung disease. Eur Respir Rev. (2013) 22:6. doi: 10.1183/09059180.00005512

PubMed Abstract | Crossref Full Text | Google Scholar

34. Goh NS, Desai SR, Veeraraghavan S, Hansell DM, Copley SJ, Maher TM, et al. Interstitial lung disease in systemic sclerosis: a simple staging system. Am J Respir Crit Care Med. (2008) 177:57–63. doi: 10.1164/rccm.200706-877OC

PubMed Abstract | Crossref Full Text | Google Scholar

35. Sumimoto K, Tanaka H, Mukai J, Yamashita K, Tanaka Y, Shono A, et al. Optimal cut-off of tricuspid regurgitation velocity according to the new definition of pulmonary hypertension—its use in predicting pulmonary hypertension. Circ Rep. (2020) 2:625–29. doi: 10.1253/circrep.CR-20-0094

PubMed Abstract | Crossref Full Text | Google Scholar

36. Olink. What is NPX? Available online at: https://www.olink.com/faq/what-is-npx/ (Accessed September 29, 2025).

Google Scholar

37. Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J, et al. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. (2015) 43:D447–52. doi: 10.1093/nar/gku1003

PubMed Abstract | Crossref Full Text | Google Scholar

38. Al-Nesf A, Mohamed-Ali N, Acquaah V, Al-Jaber M, Al-Nesf M, Yassin MA, et al. Untargeted metabolomics identifies a novel panel of markers for autologous blood transfusion. Metabolites. (2022) 12:517. doi: 10.3390/metabo12050425

PubMed Abstract | Crossref Full Text | Google Scholar

39. Pang Z, Lu Y, Zhou G, Hui F, Xu L, Viau C, et al. MetaboAnalyst 6.0: toward a unified platform for metabolomics data processing, analysis, and interpretation. Nucleic Acids Res. (2024) 52:W398–406. doi: 10.1093/nar/gkae316

PubMed Abstract | Crossref Full Text | Google Scholar

40. Distler O, Pap T, Kowal-Bielecka O, Meyringer R, Guiducci S, Landthaler M, et al. Overexpression of monocyte chemoattractant protein 1 in systemic sclerosis: role of platelet-derived growth factor and effects on monocyte chemotaxis and collagen synthesis. Arthritis Rheumatol. (2001) 44:2665–78. doi: 10.1002/1529-0131(200111)44:11<2665::AID-ART446>3.0.CO;2-S

PubMed Abstract | Crossref Full Text | Google Scholar

41. Osafo-Addo AD and Herzog EL. CCL2 and T cells in pulmonary fibrosis: an old player gets a new role. Thorax. (2017) 72:967–68. doi: 10.1136/thoraxjnl-2017-210517

PubMed Abstract | Crossref Full Text | Google Scholar

42. Hasegawa M, Sato S, and Takehara K. Augmented production of chemokines (monocyte chemotactic protein-1, macrophage inflammatory protein-1α, and MIP-1β) in patients with systemic sclerosis: MCP-1 and MIP-1α may be involved in the development of pulmonary fibrosis. Clin Exp Immunol. (1999) 117:159–65. doi: 10.1046/j.1365-2249.1999.00929.x

PubMed Abstract | Crossref Full Text | Google Scholar

43. Yanaba K, Komura K, Kodera M, Matsushita T, Hasegawa M, Takehara K, et al. Serum levels of monocyte chemotactic protein-3/CCL7 are raised in patients with systemic sclerosis: association with extent of skin sclerosis and severity of pulmonary fibrosis. Ann Rheum Dis. (2006) 65:124–26. doi: 10.1136/ard.2005.040782

PubMed Abstract | Crossref Full Text | Google Scholar

44. Iyonaga K, Takeya M, Saita N, Sakamoto O, Yoshimura T, Ando M, et al. Monocyte chemoattractant protein-1 in idiopathic pulmonary fibrosis and other interstitial lung diseases. Hum Pathol. (1994) 25:455–63. doi: 10.1016/0046-8177(94)90117-1

PubMed Abstract | Crossref Full Text | Google Scholar

45. Wang L, Zhu M, Li Y, Yan P, Li Z, Chen X, et al. Serum proteomics identifies biomarkers associated with the pathogenesis of idiopathic pulmonary fibrosis. Mol Cell Proteomics. (2023) 22:100524. doi: 10.1016/j.mcpro.2023.100524

PubMed Abstract | Crossref Full Text | Google Scholar

46. Rose-John S. IL-6 trans-signaling via the soluble IL-6 receptor: importance for the proinflammatory activities of IL-6. Int J Biol Sci. (2012) 8:1237–47. doi: 10.7150/ijbs.4989

PubMed Abstract | Crossref Full Text | Google Scholar

47. Taga T, Hibi M, Hirata Y, Yamasaki K, Yasukawa K, Matsuda T, et al. Interleukin-6 triggers the association of its receptor with a possible signal transducer, gp130. Cell. (1989) 58:573–81. doi: 10.1016/0092-8674(89)90438-8

PubMed Abstract | Crossref Full Text | Google Scholar

48. Shi-wen X, Pennington D, Holmes A, Leask A, Bradham D, Beauchamp JR, et al. Autocrine overexpression of connective tissue growth factor maintains fibrosis: representational difference analysis of fibrosis genes in systemic sclerosis. Exp Cell Res. (2000) 259:213–24. doi: 10.1006/excr.2000.4972

PubMed Abstract | Crossref Full Text | Google Scholar

49. Tam AYY, Horwell AL, Trinder SL, Khan K, Xu S, Ong V, et al. Selective deletion of connective tissue growth factor attenuates experimentally induced pulmonary fibrosis and pulmonary arterial hypertension. Int J Biochem Cell Biol. (2021) 134:105957. doi: 10.1016/j.biocel.2021.105961

PubMed Abstract | Crossref Full Text | Google Scholar

50. Cavalli G and Dinarello CA. Anakinra therapy for non-cancer inflammatory diseases. Front Pharmacol. (2018) 9:1157. doi: 10.3389/fphar.2018.01157

PubMed Abstract | Crossref Full Text | Google Scholar

51. Gumkowska-Sroka O, Kotyla K, Mojs E, Palka K, and Kotyla P. Novel therapeutic strategies in the treatment of systemic sclerosis. Pharm (Basel). (2023) 16:1066. doi: 10.3390/ph16081066

PubMed Abstract | Crossref Full Text | Google Scholar

52. Kodumudi V, Bibb LA, Adalsteinsson JA, Shahriari N, and Skudalski L. Emerging therapeutics in the management of connective tissue disease. Part II: Dermatomyositis and scleroderma. J Am Acad Dermatol. (2022) 87:21–38. doi: 10.1016/j.jaad.2021.12.068

PubMed Abstract | Crossref Full Text | Google Scholar

53. Kielstein JT, Bode-Böger SM, Hesse G, Martens-Lobenhoffer J, Takacs A, Fliser D, et al. Asymmetrical dimethylarginine in idiopathic pulmonary arterial hypertension. Arterioscler Thromb Vasc Biol. (2005) 25:1414–18. doi: 10.1161/01.ATV.0000168414.06853.f0

PubMed Abstract | Crossref Full Text | Google Scholar

54. Skoro-Sajer N, Mittermayer F, Panzenboeck A, Bonderman D, Sadushi R, Hitsch R, et al. Asymmetric dimethylarginine is increased in chronic thromboembolic pulmonary hypertension. Am J Respir Crit Care Med. (2007) 176:1154–60. doi: 10.1164/rccm.200702-278OC

PubMed Abstract | Crossref Full Text | Google Scholar

55. Zhang S, Yang T, Xu X, Wang M, Zhong L, Yang Y, et al. Oxidative stress and nitric oxide signaling related biomarkers in patients with pulmonary hypertension: a case-control study. BMC Pulm Med. (2015) 15:50. doi: 10.1186/s12890-015-0045-8

PubMed Abstract | Crossref Full Text | Google Scholar

56. Yoshiba N, Yoshiba K, Ohkura N, Takei E, Edanami N, Oda Y, et al. Correlation between fibrillin-1 degradation and mRNA downregulation and myofibroblast differentiation in cultured human dental pulp tissue. J Histochem Cytochem. (2015) 63:438–48. doi: 10.1369/0022155415580622

PubMed Abstract | Crossref Full Text | Google Scholar

57. Smallridge RS, Whiteman P, Werner JM, Campbell ID, Handford PA, Downing AK, et al. Solution structure and dynamics of a calcium-binding epidermal growth factor-like domain pair from the neonatal region of human fibrillin-1. J Biol Chem. (2003) 278:12199–206. doi: 10.1074/jbc.M208266200

PubMed Abstract | Crossref Full Text | Google Scholar

58. Cai Z, Tian S, Klein T, Tu L, Geenen LW, Koudstaal T, et al. Kynurenine metabolites predict survival in pulmonary arterial hypertension: a role for IL-6/IL-6Rα. Sci Rep. (2022) 12:12326. doi: 10.1038/s41598-022-15039-3

PubMed Abstract | Crossref Full Text | Google Scholar

59. He YY, Yan Y, Jiang X, Zhao J, Wang Z, Wu T, et al. Spermine promotes pulmonary vascular remodelling and its synthase is a therapeutic target for pulmonary arterial hypertension. Eur Respir J. (2020) 56:1902004. doi: 10.1183/13993003.00522-2020

PubMed Abstract | Crossref Full Text | Google Scholar

60. Avvedimento EV and Gabrielli A. Linking NAD metabolism and DNA repair to inflammation in systemic sclerosis. Nat Rev Rheumatol. (2021) 17:381–82. doi: 10.1038/s41584-021-00629-8

PubMed Abstract | Crossref Full Text | Google Scholar

61. Lin J, Liu J, Ma R, Hao J, Liang Y, Zhao J, et al. Interleukin-33: Metabolic checkpoints, metabolic processes, and epigenetic regulation in immune cells. Front Immunol. (2022) 13:900826. doi: 10.3389/fimmu.2022.900826

PubMed Abstract | Crossref Full Text | Google Scholar

62. Smith DE. IL-33: a tissue derived cytokine pathway involved in allergic inflammation and asthma. Clin Exp Allergy. (2010) 40:200–8. doi: 10.1111/j.1365-2222.2009.03384

PubMed Abstract | Crossref Full Text | Google Scholar

63. Miller AM, Xu D, Asquith DL, Denby L, Li Y, Sattar N, et al. IL-33 reduces the development of atherosclerosis. J Exp Med. (2008) 205:339–46. doi: 10.1084/jem.20071868

PubMed Abstract | Crossref Full Text | Google Scholar

64. Monticelli LA, Osborne LC, Noti M, Tran SV, Zaiss DM, and Artis D. Il-33 promotes an innate immune pathway of intestinal tissue protection dependent on amphiregulin-egfr interactions. Proc Natl Acad Sci USA. (2015) 112:10762–7. doi: 10.1073/pnas.1509070112

PubMed Abstract | Crossref Full Text | Google Scholar

65. Yang Q, Li G, Zhu Y, Liu L, Chen E, Turnquist H, et al. IL-33 synergizes with TCR and IL-12 signaling to promote the effector function of CD8+ T cells. Eur J Immunol. (2011) 41:3351–60. doi: 10.1002/eji.201141629

PubMed Abstract | Crossref Full Text | Google Scholar

66. de Kleer IM, Kool M, de Bruijn MJ, Willart M, van Moorleghem J, Schuijs MJ, et al. Perinatal activation of the interleukin-33 pathway promotes type 2 immunity in the developing lung. Immunity. (2016) 45:1285–98. doi: 10.1016/j.immuni.2016.10.031

PubMed Abstract | Crossref Full Text | Google Scholar