Laser Capture Microscopy RNA Sequencing for Topological Mapping of Synovial Pathology During Rheumatoid Arthritis

Abstract

Objective: Rheumatoid arthritis (RA) is an autoimmune disease in which the joint lining or synovium becomes highly inflamed and majorly contributes to disease progression. Understanding pathogenic processes in RA synovium is critical for identifying therapeutic targets. We performed laser capture microscopy (LCM) followed by RNA sequencing (LCM-RNAseq) to study regional transcriptomes throughout RA synovium.

Methods: Synovial lining, sublining, and vessel samples were captured by LCM from seven patients with RA and seven patients with osteoarthritis (OA). RNAseq was performed on RNA extracted from captured tissue. Principal component analysis was performed on the sample set by disease state. Differential expression analysis was performed between disease states based on log₂ fold change and q value parameters. Pathway analysis was performed using the Reactome Pathway Database on differentially expressed genes among disease states. Significantly enriched pathways in each synovial region were selected based on the false discovery rate.

Results: RA and OA transcriptomes were distinguishable by principal component analysis. Pairwise comparisons of synovial lining, sublining, and vessel samples between RA and OA revealed substantial differences in transcriptional patterns throughout the synovium. Hierarchical clustering of pathways based on significance revealed a pattern of association between biologic function and synovial topology. Analysis of pathways uniquely enriched in each region revealed distinct phenotypic abnormalities. As examples, RA lining samples were marked by anomalous immune cell signaling, RA sublining samples were marked by aberrant cell cycle, and RA vessel samples were marked by alterations in heme scavenging.

Conclusion: LCM-RNAseq confirms reported transcriptional differences between the RA synovium and the OA synovium and provides evidence supporting a relationship between synovial topology and molecular anomalies in RA.

Figure 1.

LCM followed by RNA sequencing of RA and OA synovial tissue. (A) Representative images of microdissection of RA synovial lining, sublining, and vessel samples. Images are at ×10 magnification and show substructures before microdissection (left), after microdissection (middle), and captured on the cap (right). (B) Principal component analysis by disease state was performed on synovial lining, sublining, and vessel samples from seven patients with RA and seven patients with OA. Proportion of variability explained is shown in parentheses. (C) Heatmap depicts sample Pearson correlations based on all genes satisfying preliminary screening criteria (see Materials and methods). Positive correlations are visualized in shades of red, and negative correlations are visualized in shades of blue. (D) Venn diagram of differentially expressed genes between RA samples and OA samples in the synovial lining, sublining, and vessels (q value <0.10 [Mann-Whitney U test] and ∣log₂ fold change∣ > 1.0). (E) Heatmap shows false discovery rate values of selected pathways found to be significantly altered (false discovery rate <1 ×10⁻³) in all three synovial regions. Unsupervised clustering of pathways based on false discovery rate values in each region resulted in pathway grouping by biologic function. LCM, laser capture microscopy; OA, osteoarthritis; PC, principal component; RA, rheumatoid arthritis.

Figure 2.

Unique pathways and related genes in RA synovial lining. (A) Reactome pathway analysis was performed using differentially expressed genes in the RA versus OA synovial linings. The graph shows the four pathways found to be significant (false discovery rate < 1 × 10⁻³) only in the lining. Pathways are color coded by relation to immune cell signaling (brown) or ECM organization (green). (B and C) Heatmaps show the Z score of expression levels of differentially expressed genes found in the pathways related to (B) immune cell signaling or (C) ECM organization that are uniquely significant (false discovery rate < 1 × 10⁻³) in the synovial lining. (D) Heatmap of expression of immune cell signaling pathway differentially expressed genes in synovial lining macrophages by GeoMx. (E) The dot plot shows the log₂ fold change of synovial lining expression between RA and OA of key marker genes previously identified in the study by Zhang et al in fibroblasts of the synovial lining (group SC-F4). ECM, extracellular matrix; MHC, major histocompatibility complex; OA, osteoarthritis; RA, rheumatoid arthritis; TCR, T cell receptor.

Figure 3.

Unique pathways and related genes in RA synovial sublining. (A) Reactome pathway analysis was performed using differentially expressed genes in the RA versus OA synovial sublining. The graph shows the four pathways found to be significant (false discovery rate < 1 × 10⁻³) only in the sublining. Pathways are color coded by relation to cell cycle (brown) or MHC class I (green). (B and C) Heatmaps show the Z score of expression levels of differentially expressed genes found in the pathways related to (B) cell cycle or (C) MHC class I that are uniquely significant (false discovery rate < 1 × 10⁻³) in the synovial sublining. (D) Heatmap of expression of MHC class I pathway differentially expressed genes in synovial sublining macrophages by GeoMx. (E) The dot plot shows the log₂ fold change of synovial sublining expression between RA and OA of key marker genes previously identified in the study by Zhang et al in fibroblasts of the synovial sublining (group SC-F2). MHC, major histocompatibility complex; OA, osteoarthritis; RA, rheumatoid arthritis.