National and Kapodistrian University of Athens
Pharmacometabolomic study of osteoarthritis and rheumatoid arthritis using mass spectrometry and nuclear magnetic resonance spectroscopy
Abstract
dc:descriptionIntroduction: Rheumatoid arthritis (RA) is a chronic, systemic autoimmune disease marked by synovial inflammation, pain, and progressive joint damage with substantial extra-articular morbidity. Despite therapeutic advances, durable remission remains elusive for many patients, underscoring the need for robust biomarkers for early diagnosis, monitoring disease activity and pain, and evaluation of therapeutic efficacy. Metabolomics and lipidomics enable quantitative phenotyping of perturbed amino-acid, energy, lipid and lipoprotein pathways in RA. This dissertation established a multi-platform framework combining proton nuclear magnetic resonance (1H-NMR) metabolomics and NMR-based lipoprotein and glycoprotein profiling, targeted mass-spectrometry (MS) lipidomics, urinary 1H-NMR metabolomics, and targeted oxylipin analysis to characterize RA biofluids, define clinically relevant metabolic signatures, and evaluate cross-platform analytical congruence. Aims: The principal aims of this dissertation were: (1) to characterize NMR-derived small molecules, lipoproteins and glycoprotein signals in RA plasma and explore the relation to inflammation, disease activity, and pain; (2) to define therapy-specific plasma metabolomic and lipidomic responses; (3) to profile the urinary metabolome as a non-invasive reflection of disease status; (4) to quantify lipid mediators related to inflammatory resolution, and (5) to validate an integrated single-aliquot NMR to MS workflow for translational application. Methods: RA patients and healthy volunteers from the ArthritisHeal project (ArthritisHeal) were recruited at the rheumatology outpatient clinic of General Hospital of Athens “Laikon”, and fasted plasma and preprandial urine were collected. RA patients were stratified by DAS28-CRP (Remission/Low/Moderate/High), therapy stage and type, and systemic inflammation (Low/High). 1H NMR metabolomics was employed to profile small molecules in plasma and urine. Glycoprotein signals (GlycA/B) were extracted from 1H-NMR spectra. Lipoprotein subclasses were quantified via the Bruker IVDr B.I.-LISA platform. Targeted plasma lipidomics (DMS-Shotgun Lipidomics Assistant; FIA-MRM) quantified over 700 isoform-resolved lipid species across multiple lipid classes. Targeted lipid mediator analysis, which allows for the quantification of structural lipids, fatty acids and pro-resolving mediators, was performed by reversed-phase liquid chromatography-tandem mass spectrometry (RPLC-MS/MS). A single-aliquot workflow enabled sequential IVDr B.I.-LISA and DMS-SLA analysis, facilitating quantitative cross-platform evaluation and integrative lipid-lipoprotein correlation mapping. Results: Plasma metabolites, lipoproteins and glycoproteins in RA (1H-NMR): RA patients exhibited a distinct plasma metabolic, lipoprotein and glycoprotein fingerprint, with reduced levels of lactic acid, formic acid, sarcosine and valine, elevated phenylalanine, tyrosine, triglycerides, cholesterol, apolipoproteins (A1, A2, B100), together with marked remodeling of LDL and HDL subclasses, and increased GlycA/B signals. Glycoproteins were significantly higher in DMARD-naive RA (AUC ≈ 0.9), validating these NMR-derived signals as robust inflammation markers. Remission was characterized by higher small dense HDL-4 and lower LDL-triglycerides than high activity RA. H4A1 alone distinguishes RA remission from active disease (AUC ≈ 0.8). DAS28 correlated positively with small dense LDL-triglycerides and glycoproteins, and inversely with large LDL and small HDL lipids, while VAS pain score was associated with LDL-triglycerides. Lipoprotein and lipid fingerprints after 6-week monotherapy: Short-term biologic and targeted synthetic DMARD treatment was accompanied by therapy-specific metabolic and lipoprotein signatures. IL-6R blockade induced marked increases in total and free cholesterol, triglycerides and apoB-containing lipoproteins (VLDL/IDL) with parallel elevations in TAG(18:1/18:2) and CE(18:2), consistent with enhanced hepatic lipoprotein synthesis. JAK inhibition promoted a more favorable profile with enrichment of HDL-4 particles, higher apoA1/apoA2 and reductions IDL/LDL cholesterol, triglycerides and selected sphingolipids. TNF inhibition produced more modest lipoprotein and lipid remodeling, while methotrexate and glucocorticoids showed characteristic small-molecule changes compatible with altered energy metabolism and increased lipolysis. These therapy-related patterns were statistically supported in some comparisons but remained exploratory due to sample size and heterogeneity. Targeted plasma lipidomics (DMS-SLA): RA patients displayed pronounced reductions in lysophosphatidylcholines (LPCs; class p≈1×10⁻⁵), particularly LPC(16:0), LPC(18:0), and LPC(20:4), together with lower cholesteryl esters CE(18:2) and CE(18:0). The most significant perturbations occurred in moderate disease activity, characterized by further LPC depletion and increased TAGs-containing arachidonic acid (FA(20:4)). Clinical monitoring markers, DAS28 and CRP positively correlated with diacylglycerols and ceramides, and inversely with CEs, sphingomyelins, and LPCs. Pain intensity (VAS) showed positive correlations with AA-containing TAGs (r≈0.30, p=1×10⁻⁵). In DMARD-naïve RA, reduced LPC species and elevated phosphatidic acid PA(34:5) achieved fair discrimination versus controls (AUC>0.70), while refractory RA showed pronounced LPC(20:4) depletion (AUC=0.91) and increased TAG species. Urine 1H-NMR metabolomics: Untargeted ¹H-NMR of urine revealed a reproducible RA signature involving amino-acid, energy and gut-derived metabolites. Active RA exhibited lower branched-chain amino acids, 3-hydroxyisovalerate, glucose and citrate, consistent with increased catabolic demand and altered tricarboxylic acid cycle flux. DAS28, ESR and VAS were inversely associated with valine, citrate, creatinine and 3-hydroxybutyrate and positively associated with 3-indoxyl sulfate and phenylacetylglutamine, implicating BCAA, TCA and tryptophan/indole pathways in RA activity and symptom burden. DMARDnaïve and refractory RA showed distinct urinary profiles, and therapy-type comparisons indicated differential modulation of creatinine, citrate and TMAO. Cross-fluid coupling linked urinary citrate with small HDL-4 features, supporting urine as a non-invasive surrogate of systemic metabolic remodeling. Oxylipins and inflammation resolution: Targeted lipid mediator profiling detected 31 oxylipins, of which 13 individual mediators and 5 ratios differed significantly between RA and controls and across RA inflammation strata. HETE and HDHA isomers were prominently altered. 5-HETE and the AA:5-HETE ratio showed the strongest discrimination between high- and low-inflammation RA (AUC ≈ 0.74-0.77), indicating perturbed 5-lipoxygenase pathway flux and impaired resolution in highly inflamed disease. Multiple HETE and HDHA species correlated with DAS28 and even more strongly with VAS (r ≈ 0.50-0.68), and multivariate models demonstrated that composite oxylipin profiles best captured pain in high inflammation RA (R² ≈ 0.82). Methodological congruence and complementarity of IVDr B.I.-LISA and DMS-SLA: Sequential singlealiquot IVDr B.I.-LISA and DMS-SLA analyses showed excellent agreement at the lipid-class level and comparable diagnostic performance in distinguishing RA from controls. The validated workflow enabled robust lipoprotein–lipid correlation mapping, revealing biologically coherent lipid-lipoprotein relationships and underscoring the complementarity of NMR- and MS-based lipid measurements. Conclusions: This thesis establishes a clinically anchored, multi-modal framework linking RA phenotypes to coordinated alterations across plasma lipoproteins, glycoproteins, lipid species, urinary metabolites, and lipid mediators. Key conclusions are: 1. Lipoprotein profiles in RA correlate with disease activity, inflammation, and pain. Large HDL, intermediate LDL and glycoproteins serve for RA monitoring as well as potential molecular markers of pain. 2. Therapy-specific lipid remodeling reflects both drug action and disease activity in RA, with IL-6R blockade elevating VLDL/IDL lipids and JAK inhibition enhancing HDL subclasses. These findings highlight distinct metabolic signatures that may serve as biomarkers of therapeutic response. 3. RA is defined by consistent LPC depletion and selective remodeling of cholesteryl ester and TAG species, which closely reflect disease activity and pain intensity. These findings highlight specific lipid species as sensitive biomarkers of inflammatory burden and disease state in RA. 4. The urinary metabolome in RA reflects integrated metabolic disturbances encompassing amino acid catabolism, mitochondrial energy turnover, and gut-microbial co-metabolism. Reduced levels of BCAAs, citrate, and creatinine, together with elevated 3-indoxyl sulfate and TMAO, define a reproducible urinary signature associated with inflammation and pain. Urinary NMR metabolomics may thus complement blood-based profiling for non-invasive monitoring of RA disease activity and therapeutic response. 5. This study identifies HETE and HDHA derivatives, particularly 5-HETE and AA:5-HETE, as potential lipid biomarkers reflecting inflammatory burden and pain in RA. The findings suggest that dysregulation of arachidonic acid metabolism contributes to persistent inflammation and nociception, highlighting the therapeutic potential of targeting lipoxygenase-derived lipid mediators in RA. 6. A validated single-aliquot NMR/MS workflow demonstrates cross-platform congruence and enables joint lipid-lipoprotein correlation maps that are sensitive to disease state, supporting clinical translation and future mechanistic studies.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Tsezou, Konstantina-Ismini
- Τσέζου, Κωνσταντίνα-Ισμήνη
Subjects
dc:subject × 18- Μεταβολομική ανάλυση
- Λιπιδομική ανάλυση
- Βιοδείκτες
- Ρευματοειδής αρθρίτιδα
- Metabolomic analysis
- Lipidomics analysis
- Biomarkers
- Rheumatoid arthritis
- Άλλες Ιατρικές Επιστήμες
- Ιατρική και Επιστήμες Υγείας
- Ιατρική έρευνα και Πειραματική ιατρική
- Βασική Ιατρική
- Ανοσολογία
- Other Medical Sciences
- Medical and Health Sciences
- Research and Experimental medicine
- Basic Medicine
- Immunology
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/61217