Abstract
Chronic widespread pain (CWP) is a complex trait with poorly understood pathogenesis, affecting 10.6% of the world population and 14.2% of the UK population. Heritability estimates of CWP range between 48-54%, indicating a substantial genetic contribution. CWP is one of the major diagnostic criteria for fibromyalgia (FM), representing a subgroup at the more severe end of the CWP spectrum. The traditional candidate gene approach to study genetic factors contributing to CWP has predominantly examined genes expressed in the central and peripheral nervous system. However, results have been inconsistent; for example, catechol-O-methyltransferase (COMT) involvement in CWP could not be confirmed. Genetic factors are shared between chronic pain conditions, and an inconclusive role of COMT was also observed for temporomandibular disorders. Genome-wide association studies (GWAS) bring the advantage of agnostic methods, but to date, such studies of CWP are of limited size and scope. Therefore, I aimed to identify genetic variants/genes involved in CWP through GWAS and obtain additional insights into risk factors and disease pathogenesis using in-silico follow-up of GWAS summary data.Observational studies have repeatedly identified higher body mass index (BMI) as a risk factor for CWP. One potential mechanism underlying the association might include a steady but low level of inflammation (low-grade inflammation), as fat tissue actively secretes inflammatory markers. Some autoimmune diseases occurring with CWP (i.e., secondary FM) also support the role of inflammation in CWP. Mast cells in the skin producing low-grade inflammation have been associated with FM. These mast cells can be activated by immunoglobulin-G (IgG) to promote inflammatory process. Transferring IgG from FM patients to mice has shown increased pain sensitivity compared to the control IgG, and FM patients treated with IgG have a short-term reduction in pain and tenderness and improvement in strength. The role of IgG in CWP pathogenesis might be influenced by glycans attached to the nitrogen atom of asparagine residue of proteins (N-glycosylation), known as N-glycans, which plays a role in immune regulation. A role of N-glycosylation has been observed in several chronic diseases but not limited to rheumatoid arthritis (RA), hypertension, inflammatory bowel disease, metabolic conditions, several cancer types, Alzheimer’s disease, and mild cognitive impairments. Many of these diseases co-occur with CWP and are age-related as CWP is. It was suggested that with increasing age, Nglycans attached to IgG starts producing inflammatory loops in the body that affects the homeostasis of peripheral functions (e.g., skeletal homeostasis), causing age-related diseases. A recent study from TwinsUK found that several N-glycans are associated with low back pain, a pain-site commonly affected in CWP. Collectively, the role of IgG Nglycosylation in CWP can be suspected. Therefore, I investigated the relevance of N-glycan components of IgGs in CWP at the population level and integrated summary data (GWAS data for N-glycans and CWP) to infer causality and obtain additional biological insight.
To explore the genetic determinants of CWP, I performed a GWAS using UK Biobank (UKB) as discovery data. Independent replication of UKB findings was attempted using six European cohorts. Genome-wide association analyses were performed using a linear mixed-effects model assuming an additive genetic model for SNP effect on CWP, adjusted for age, sex, genotyping platform, and the first ten genetic principal components. In silicofollow-up (e.g., genetic correlation, partial genetic correlation, tissue-specificity, colocalisation) was performed using publicly available data. The cross-sectional association between IgG N-glycans and CWP was examined using TwinsUK data. Association between N-glycans and CWP was assessed using a generalised linear mixed-effects model adjusted for age, sex, and smoking status as fixed-effects and zygosity, family relatedness, and batch run-day as random-effects. Several sensitivity analyses were performed, such as female-specific analysis. I used summary data to assess the causal association between Nglycans and CWP and performed colocalisation analyses to identify shared causal variants. Correction for multiple testing was duly applied in all analyses performed in the thesis.
The GWAS study identified three genome-wide significant loci (rs1491985, rs10490825, rs165599) residing within the genes RNF123, ATP2C1, and COMT. The RNF123 locus was replicated (meta-analysis p = 0.0002), the ATP2C1 locus showed suggestive association (p = 0.0227), and the COMT locus was not replicated. In silico follow-up highlighted the relevance of skeletal muscle and calcium regulation in CWP. Using the Mendelian randomisation approach, I found that body mass index is causally associated with CWP in a bi-directional manner. The N-glycan study found that CWP is associated with several N-glycans promoting or blocking antibodydependent cell-mediated cytotoxicity. Sensitivity analyses supported the internal validity of the primary findings. Using GWAS summary statistics, I found a causal involvement of Nglycans containing bisecting N-acetylglucosamine (GlcNAc) and core fucose with CWP. I found no evidence of shared genetic variants between CWP and IgG N-glycans.
I have identified a novel association of RNF123 locus and a suggestive association of ATP2C1 with CWP. The association with COMT, one of the most studied genes in the chronic pain field, was found associated in the discovery sample but not in the replication analysis. Also, I have identified several N-glycans causally associated with CWP, suggesting a role of immune regulation in the pathogenesis of CWP. Both GWAS and Nglycan studies add to the aetiological basis of CWP.
| Date of Award | 1 Jul 2022 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Frances Williams (Supervisor) |