Machine Learning Reveals Synovial Fibroblast Genes Associated with Pain Affect Sensory Nerve Growth in Rheumatoid...

Area of Science:

• Immunology
• Neuroscience
• Genetics

Background:

• Rheumatoid arthritis (RA) joint pain is traditionally linked to synovial inflammation, but patient pain levels don't always correlate with observed inflammation.
• This disconnect suggests alternative mechanisms driving pain in RA patients.

Approach:

• A novel machine learning method, Graph-based Gene expression Module Identification (GbGMI), was employed to identify gene modules associated with pain.
• The approach utilized gene expression data from RA patient samples, including those with longstanding disease and early untreated RA, for robust validation.

Key Points:

• GbGMI identified an 815-gene module linked to pain in RA.
• Single-cell RNA-seq revealed these genes are highly expressed in lining layer fibroblasts.
• These fibroblasts show predicted crosstalk with sensory neurons (CGRP+).
• Netrin-4, produced by fibroblasts, significantly enhanced nerve fiber outgrowth in vitro.

Conclusions:

• GbGMI effectively identifies gene modules associated with clinical features like pain.
• Synovial fibroblasts produce Netrin-4, which can promote the growth of pain-sensing nerve fibers, independent of inflammation.
• This finding offers a potential new avenue for understanding and treating RA-associated pain.