Abstract: BACKGROUND: Rheumatoid arthritis is a complex autoimmune disease, which involves the interaction of many factors including heredity, environment and immunity. Traditional research focuses on the relationship between individual protein and rheumatoid arthritis, but lacks systematic evaluation of the causal relationship between large-scale plasma proteins and rheumatoid arthritis. Mendelian randomization, as an innovative causal inference method, uses genetic variation as a tool variable, which avoids the interference of confounding factors in traditional epidemiological research and provides a new way to reveal the causal relationship between plasma protein and rheumatoid arthritis.
OBJECTIVE: To study the potential relationship between human plasma protein and rheumatoid arthritis, and to explore the possible biological pathways involved in the pathogenesis of rheumatoid arthritis.
METHODS: The public data of plasma protein (exposure factor) and rheumatoid arthritis (outcome factor) were collected. Among them, the data of 4 907 types of plasma proteins came from genome-wide association studies published in the journal Nature Genetics in 2022, which is now included in the Icelandic deCODE Genetics database. The study covered 35 559 Icelanders and found 18 084 protein quantitative trait loci. Genome-wide association study data of rheumatoid arthritis came from FinnGen database (R11 version), including 14 818 cases of rheumatoid arthritis and 287 796 cases in the control group. All participants were from the European population. The potential causal relationship between plasma protein and rheumatoid arthritis was evaluated by two-way Mendelian randomized analysis. Mendelian randomized analysis included five analysis methods: inverse variance weighted method, MR-Egger regression method, weighted median method, weighted model method and simple model method. Heterogeneity test, horizontal pleiotropy test and leave-one-out analysis were used to test whether the results of Mendelian randomized analysis are reliable. Reverse Mendelian randomization analysis took the plasma protein that met the screening conditions in the positive Mendelian randomization analysis as the outcome for effect analysis and sensitivity analysis.
RESULTS AND CONCLUSION: (1) The results of inverse variance weighting method showed that plasma proteins CRAT and LYG1 were directly and positively correlated with rheumatoid arthritis. Plasma proteins FAM177A1, JPH4 and SPAG11B had a direct negative causal relationship with rheumatoid arthritis. (2) Reverse Mendelian randomization results showed that rheumatoid arthritis had a direct positive causal relationship with LYG1. (3) No heterogeneity or horizontal pleiotropy was found in two-way sensitivity analysis. (4) With plasma proteins as the exposure and rheumatoid arthritis as the outcome, two plasma proteins CRAT and LYG1 were positively correlated with rheumatoid arthritis, and three plasma proteins FAM177A1, JPH4 and SPAG11B were negatively correlated with rheumatoid arthritis. With rheumatoid arthritis as the exposure and plasma proteins CRAT, LYG1, FAM177A1, JPH4, SPAG11B as the outcome, rheumatoid arthritis had a positive causal relationship with the plasma protein LYG1. The above results suggest that there may be causal relationship between these five plasma proteins and rheumatoid arthritis, which will help the development of potential therapeutic and monitoring targets for rheumatoid arthritis in the future. Although the protein targets found in European databases need to be verified in the China population, the preliminary evidence of such causal associations can guide domestic research to give priority to related protein pathways and accelerate targeted drug research and development. At the same time, it calls for strengthening the construction of a local multi-center biological sample bank and a proteomics-genome-wide association studies integrated platform to overcome the limitations of European databases and promote precision medical research based on Asian populations.

Key words: plasma protein, rheumatoid arthritis, Mendelian randomization, genome-wide association studies, sensitivity analysis, single nucleotide polymorphism, genetics, heterogeneity, horizontal pleiotropy