Transcriptome analysis of rheumatoid arthritis uncovers genes linked to inflammation-induced pain

Abstract

Autoimmune diseases such as rheumatoid arthritis (RA) can promote states of chronic inflammation with accompanying tissue destruction and pain. RA can cause inflammatory synovitis in peripheral joints, particularly within the hands and feet, but can also sometimes trigger temporomandibular joint (TMJ) arthralgia. To better understand the effects of ongoing inflammation-induced pain signaling, dorsal root ganglia (DRGs) were acquired from individuals with RA for transcriptomic study. We conducted RNA sequencing from the L5 DRGs because it contains the soma of the sensory neurons that innervate the affected joints in the foot. DRGs from 5 RA patients were compared with 9 non-arthritic controls. RNA-seq of L5 DRGs identified 128 differentially expressed genes (DEGs) that were dysregulated in the RA subjects as compared to the non-arthritic controls. The DRG resides outside the blood brain barrier and, as such, our initial transcriptome analysis detected signs of an autoimmune disorder including the upregulated expression of immunoglobulins and other immunologically related genes within the DRGs of the RA donors. Additionally, we saw the upregulation in genes implicated in neurogenesis that could promote pain hypersensitivity. Overall, our DRG analysis suggests that there are upregulated inflammatory and pain signaling pathways that can contribute to chronic pain in RA.

Keywords: Inflammation; Pain; Rheumatoid arthritis; Transcriptomics.

Fig. 1

Histological examination of RA DRGs: Representative H&E images of (A) control and (B) RA L4 DRGs. No major pathological differences were identified between the controls and RA samples in terms of ganglionic cells nor inflammation (Supplemental Table 3). (C) No statistical differences were seen in DRG neuronal densities between the donors.

Fig. 2

RNA-seq analysis identified 128 DEGs between RA and non-arthritic controls: (A) As seen in the heatmap (Z-score), 67 of these genes were upregulated while 61 were downregulated in the RA donors compared to the non-arthritic controls. Bilateral L5 DRGs (a & b) from the RA subjects were analyzed. Only one of the L5 DRGs from the non-arthritic controls was sequenced. (B) Scatterplot depicts base mean versus fold difference. (C) Plot of log2 Fold Change for each significant gene versus -log10 of the p adjusted value (see Supplemental Tables S7). (D). Ligand-receptor interactome data that depicts possible changes in cellular communication, particularly since our analysis showed that cell signaling seems most affected with RA.

Fig. 3

DEGs from bulk RNA-seq was analyzed using the ARCHS4 dataset to characterize how the different DRG cell types are affected by RA: Heatmaps were generated of genes associated with (A) peripheral neurons and (B) glial cells. Of note, there is some overlap in genes within each of these two cell types. (C) Signature genes for immune cells were also detected within the DRG in our transcriptomic analysis.

Fig. 4

Characterization of dysregulated neuronal genes detected in the RNAseq analysis: (A) A harmonized cross-species DRG cell atlas was used to map out which neuronal subtypes could be most impacted with RA. RNAscope was performed on sections of (B) control and (C) RA DRGs to label NTRK1⁺ (White) and SLC17A7⁺ (Green) neurons (DAPI—Blue). NTRK1 was upregulated in the DRGs of the RA donor and is generally expressed in unmyelinated C-fibers. SLC17A7 was alternatively used as a marker for large, myelinated fibers. (D) Count plot for NTRK1 (male donors—blue; female donors—red). Expression analysis of RNA-seq data shows that NTRK1 was significantly upregulated in the RA donors. (E) While there was no significant difference in the proportion of either neuronal population, (F) semi-quantitative scoring shows that the percent of strongly expressing NTRK1 cells (> 15 dots/neuron) is higher in the RA donors than the controls, which correlates with the transcriptomic data.

Fig. 5

Indications of altered immune responses in the DRGs of the RA donors: CD163 immunostaining of (A) control and (B) RA DRGs. 3 out of 5 RA donors have > 50 CD163⁺ cells/high-power field (HPF; 40X) compared with just 2 out of 7 controls (Supplemental Table S3). (C) Network analysis on deregulated genes using Ingenuity Pathway Analysis (IPA) that identified immune response related networks are perturbed with RA, where IL-15 producing kinases such as MST1R and NTRK1 were upregulated. (D) Heatmap of DEGs involved in immune system processes (GO:0002376).