Metabolic signatures of post-viral syndromes: insights from long COVID, experimental models, and comparative studies

Thesis

Post-viral syndromes, such as long COVID, represent a significant and growing public health challenge, yet their underlying biological mechanisms remain incompletely understood. This thesis explores the molecular basis of the host response to viral illness and its long-term sequelae through an integrated, multi-modal analysis spanning behaviour, inflammation, and metabolism. In a foundational murine study, distinct sex-dependent behavioural and metabolic responses to TLR-4 and TLR-7 activation were identified, revealing divergent immune-metabolic pathways and highlighting N-acetyl aspartate as a potential marker of resilience. Building on this, a longitudinal human study applied 1H NMR metabolomics to serum and urine from individuals with long COVID and fully recovered controls. Multivariate analyses uncovered robust metabolic signatures distinguishing long COVID, with subtypes linked to systemic inflammation or altered neuroendocrine stress responses. Amino acid availability emerged as a key determinant of symptom persistence. Expanding this work, a multi-omics analysis of fibromyalgia (FM) revealed shared metabolic disruptions with the non-recovered long COVID patients in amino acid pathways (notably glutamine and histidine), alongside elevated central CRP, supporting a role for neuroinflammation and metabolic dysfunction in pain sensitisation. Quantitative sensory testing further identified biologically distinct FM subtypes based on pain thresholds. Parallel work in osteoarthritis patients used serum and CSF metabolomics to identify glutamine and histidine as consistent markers of pain and disability, strengthening the case for a shared metabolic substrate across chronic pain states. Together, these studies demonstrate that chronic post-viral syndromes are underpinned by measurable metabolic and inflammatory alterations, affecting both central and peripheral systems, and that as these post viral syndromes evolve and for long COVID they seem to move toward a metabolic signature that is similar to FM. By combining metabolomics, proteomics, and sensory phenotyping, this work identifies candidate biomarkers and biological subtypes, laying a foundation for more accurate diagnostics and personalised therapeutic approaches to complex, chronic conditions emerging in the aftermath of viral illness.

Authors: Dunstan, IK

• DOI: 10.5287/ora-o8qnwp6am
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: long COVID; nuclear magnetic resonance; metabolomics; fibromyalgia (FM); chronic pain; osteoarthritis (OA); post-viral syndromes
• Subjects: Neuroimmunology; Multi-omics; Sickness behaviour; Metabolism