Document resource
Introduction Fibrotic interstitial lung disease (ILD), with no definitive treatment that stabilizes or regresses the progressive fibrosis, is frequently lethal in both systemic sclerosis (SSc) and idiopathic pulmonary fibrosis (IPF). Understanding the metabolomic foundations of IPF and SSc could lead to new discoveries for diagnosis and treatment.Material and Methods Serum samples from 63 SSc patients — 17 limited cutaneous SSc (lcSSc) without ILD, 18 lcSSc ILD, 10 diffuse cutaneous SSc (dcSSc) without ILD, and 18 dcSSc ILD, 16 IPF patients, and 13 healthy controls — were analyzed via chemical isotope labeling liquid chromatography-mass spectrometry. Factor analysis, clustering, and pathway enrichment analysis, were used to identify significant metabolic differences.Results An untargeted metabolomics approach detected 3,183 unique peak pairs in the peripheral blood serum samples ( table 1), of which 318 peak pairs were identified as metabolites with high confidence based on accurate mass and retention time in the labeled CIL metabolite library of more than 1,500 entries. With an ANOVA/ANCOVA permutation test adjusting for the covariates of age, sex, and BMI with 999,999 permutations; and FDR < 0.1, metabolomic changes in six out of the 318 profiled metabolites were found (figure 1A). The abundances cis-4-hydroxy-D-proline, propionic acid, and homovanillic acid exhibited significant differences between lcSSc without ILD and dcSSc ILD patients. Six metabolomic signatures identified in IPF patients were distinct from those observed in dcSSc ILD patients. Factor analysis of mixed data showed how the metabolites distinguish groups (figure 1B). Dimension 1 separates IPF patients from all groups, whereas dimension 2 captures the spectrum of SSc subtypes and controls. Metabolite set enrichment analysis (figure 1C) via MetaboAnalyst (KEGG version 5.0) identified eight significant (FDR < 0.15) metabolic pathways based on the results of pairwise 15 comparisons in the six groups (permutation t test, p < 0.05; and the FC cutoff). Arginine and proline metabolism associated with cis-4-hydroxy-D-proline was upregulated in dcSSc ILD patients compared with healthy controls and IPF patients, and it was also increased in IPF patients compared with the controls. Taurine and hypotaurine metabolism are lower in IPF patients than in SSc ILD patients and lcSSc patients without ILD. Aminoacyl-tRNA biosynthesis metabolism was downregulated in IPF patients compared with controls.Conclusions Unique serum metabolic signatures have been identified in patients with SSc (particularly SSc ILD) and IPF. Several metabolic pathways associated with collagen, short-chain fatty acids, and dopamine were dysregulated.Abstract P.242 Figure 1Differential abundance across groups (ANOVA/ANCOVA permutation test adjusted for the covariates age, BMI, and sex). B. Mixed-data factor analysis of abundances of 6 metabolites adjusted for age, BMI, and sex. (The ellipses represent the 95% confidence ellipses.) C. Heatmap of 15 pairwise comparisons in 6 groups of differential abundances of metabolites (FDR < 0.05, FC > 1.2 or FC <1/1.2). Eight significant pathways of metabolism (FDR < 0.15) presented in pathways 1 and 2Abstract P.242 Table 1Characteristics of patients and controls