Acylcarnitine enrichment as a characteristic of rheumatoid arthritis fibroblast-like synoviocyte metabolic fingerprint

Abstract

Objective: In rheumatoid arthritis (RA), fibroblast-like synoviocytes (FLS) alter their metabolism to support their activation. We aimed to analyse the full spectrum of metabolic alterations associated with RA by performing untargeted metabolomics in RA FLS vs. non-inflamed (NI) FLS.

Methods: Untargeted annotated metabolomics was performed using mass spectrometry on ten primary RA and seven NI FLS culture extracts and 220 serum samples from participants with early RA from the randomised controlled NORD-STAR trial. Carnitine-related proteins were measured with Western blot. FLS bioenergetic profile was assessed with a Seahorse flux analyser.

Results: Metabolomics analysis based on 138 annotated metabolites revealed a distinct metabolic fingerprint between RA and NI FLS. Of the 12 metabolites enriched in RA FLS, 11 were acylcarnitines. Pro-inflammatory stimulation of NI FLS also led to acylcarnitine accumulation. RA FLS exhibited lower levels of CD36, a fatty acid transporter, but similar levels of L-carnitine transporter, and carnitine palmitoyltransferase 1 A and 2 compared to NI FLS. Seahorse analyses showed no difference in fatty acid oxidation between RA and NI FLS; however, RA FLS displayed mitochondrial dysfunction and energetic impairment. Serum acylcarnitine content decreased after 24 weeks of treatment with methotrexate combined with abatacept or tocilizumab in patients with early RA achieving remission.

Conclusion: Acylcarnitine accumulation is a characteristic of RA FLS metabolic fingerprint and could be linked to mitochondrial dysfunction. In patients with early RA, acylcarnitine content in serum decreases after successful anti-rheumatic treatment. These results indicate a dysregulation in acylcarnitine metabolism in RA at the joint level and systemically.

Keywords: Acylcarnitines; Fibroblast-like synoviocytes; Metabolomics; Rheumatoid arthritis.

Fig. 1

FLS metabolomics set-up overview. Created in https://BioRender.com.

Fig. 2

FLS metabolomics analysis. A. Clustering result shown as a heatmap. Distance measure using Euclidean distance. Samples are shown on the x axis, while metabolites are shown on the y axis. B. Enriched metabolites by volcano plot defined as fold change (x) ≥2 and Mann-Whitney test p-value (y) < 0.05. The red dots represent acylcarnitines. C. Partial least squares - discriminant analysis showing the 2-D scores plot between principal component 1 and 2. Each dot represents a sample. The cycles show the 95 % confidence interval. D. Dot plots of the peak intensities of three selected acylcarnitines and L-carnitine in NI and RA FLS. Mann-Whitney test was performed to determine the significance. E. Receiver operating characteristic curves of three selected acylcarnitines and L-carnitine. Sensitivity is on the y-axis, and the specificity on the x-axis. Results are from 10 RA FLS and 7 NI FLS. Abbreviations: RA, rheumatoid arthritis; NI, non-inflamed; FLS, fibroblast-like synoviocytes; AUC, area-under-the-curve.

Fig. 3

Acylcarnitine species, L-carnitine, L-glutamine, and adenosine in FLS. A. Average peak intensities of all identified acylcarnitine species stacked together and L-carnitine, B. significantly changed acylcarnitines, C. L-glutamine, and adenosine in NI-FLS unstimulated or stimulated with 10 ng/ml TNF or 10 ng/ml IL-1β for 16h prior to sample collection, as well as in RA FLS. Differences between groups were calculated using autoscaled data with 2-way ANOVA in A and Wilcoxon test in B and C to compare unstimulated and stimulated NI FLS, or Mixed effects model in A and Mann-Whitney test in B and C to compare between NI and RA FLS. Results are from 10 RA FLS and 7 NI FLS. Abbreviations: RA, rheumatoid arthritis; NI, non-inflamed; FLS, fibroblast-like synoviocytes; TNF, tumour necrosis factor; IL-1β, interleukin 1β.

Fig. 4

Expression of acylcarnitine-related transporters and enzymes. Expression of A. CD36 (cell membrane transporter of fatty acids), B. OCTN2 (cell membrane transporter of L-carnitine), C. ACSL5, D. CPT1A, and E. CPT2 shown as bar graphs and representative images. Data were normalised to vinculin and expression levels are shown as percentages of the same control sample loaded to different gels. Significance was determined with Mann-Whitney test. Results are from 9 RA FLS and 7 NI FLS. Abbreviations: OCTN2, solute carrier family 22 member 5; ACSL5, long-chain-fatty-acid—CoA ligase 5; CPT, carnitine palmitoyl transferase.

Fig. 5

Seahorse analysis. A. Quantification of total FAO at baseline and under FCCP-induced stress (maximal) with the Fatty Acid Oxidation test. B. OCR kinetics, C. quantification of respiratory parameters, and D. energetic maps obtained by plotting ECAR and OCR values at baseline and under FCCP-induced stress of RA vs NI FLS using glucose/glutamine as substrate with the Mito Stress test. E. GlucoATP (ATP production associated with conversion of glucose to lactate) and MitoATP (ATP production associated with OXPHOS in the mitochondria) production rates (pmol ATP/min/μg of proteins), F. percentages of total ATP and G. metabolic index calculated as the ratio of mitoATP/glycoATP production rate with the use of ATP Rate Assay test. In all tests, OCR was measured under basal conditions and after addition of oligomycin (1 μM) to inhibit ATP synthase, FCCP (1 μM) to uncouple OXPHOS, and antimycin A (2 μM) and rotenone (2 μM) to calculate non-mitochondrial respiration. Data were obtained from 4 – 8 replicates for each independent sample (3–4 samples per group). Data were normalised to total protein content and shown as mean ± SEM and p values obtained from unpaired t-test.

Fig. 6

Circulating acylcarnitines detected by metabolomics in participants with early RA in NORD-STAR study. A. Study description and metabolomics overview. Created in https://BioRender.com. B. Changes of inflammatory markers ESR, CRP, and CDAI before and after treatment. C. Total acylcarnitine content defined as the sum of all the detected acylcarnitines' peak intensities and L-carnitine at baseline and 24-week follow-up. D. Total acylcarnitine content after stratifying for remission at 24 weeks (CDAI≤2.8). E-H. Total acylcarnitine content at baseline and 24-week follow-up in each treatment arm and after stratifying for remission. Significance was determined with paired t-test for comparisons before and after treatment or unpaired t-test for comparisons between remission and non-remission. Abbreviations: MTX, methotrexate; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; CDAI, Clinical Disease Activity Index.