Resident synovial macrophages in synovial fluid: Implications for immunoregulation

Abstract

Resident synovial macrophages (RSMs) are anti-inflammatory, self-renewing macrophages that provide physical immune sequestration of the joint space from the peripheral immune system. Increased permeability of this structure is associated with peripheral immune cells in the synovial fluid (SF). Direct measures of synovial barrier integrity are possible with tissue histology, but after barrier breakdown, if these cells perpetuate or initiate chronic inflammation in SF remains unknown. We sought to identify RSM in human SF as an indirect measure of synovial barrier integrity. To validate findings, we created a novel ex vivo explant model using human synovium. scRNA-seq revealed these SF RSMs upregulated pro-fibrotic and pro-osteoclastic pathways in inflammatory arthritis, but not septic arthritis. Increased frequencies of RSMs in SF was associated with increased sRANKL regardless of underlying pathology. These findings suggest the frequency of RSMs in SF may correlate with synovial barrier damage and in turn, potential damage to joint structures.

Keywords: Arthritis; Immune regulation; Joint space; Macrophage; Synovitis; Synovium.

Figure 1:

Resident Synovial Macrophage-like cells in SF. Live, CD56⁻CD3⁻CD20⁻RANKL⁻CD11c⁻TREM2⁺CD68⁺CD11b⁺HLA-DR⁺CX3CR1⁺ cells were identified with an enrichment of CD14^dim cells in the SF compared to PBMCs (A, n=6 patients for preliminary evaluation, 2 = septic arthritis, 2 = inflammatory arthritis, 2 = control/no joint disease patients). These CD14^dim cells were increased in synovial fluid compared to PBMC in matched samples (B, Wilcoxon Rank Sum). The same cells were found in higher concentrations in patients with inflammatory arthritis based on SF analysis (C, one way ANOVA). Control patients were retrospectively found to have normal SF after diagnostics.

Figure 2:

Transwell synovial cell migration assay using ex vivo tissue in a proof-of-concept experiment. Schematic of the experimental set up (A). Gating scheme for migratory cells in the bottom well (B). Gating of the CD11b^+/+CD14^dim population was based on PBMC control (C). Sequential gating of the Live/Dead populations per treatment (D). Two well replicates per treatment condition. Statistics not performed. Alterations of the synovial lining with increasing doses of LPS (E). Macrophage (CD68) presence at the intimal lining verified using IF (F). * = autofluorescence of erythrocytes. Arrows = joint lining cells are lost from the intimal border leaving holes.

Figure 3:

ELISA results of SF supernatant protein concentration. Each data point = 1 patient. Multiple Mann-Whitney tests with FDR rate <0.05.

Figure 4:

Linear regressions comparing %CD14^dim (of total non-neutrophil, non-lymphocyte sample) M2 macrophage cells to ELISA where data was available for the same patient. Inflammatory and septic arthritis are analyzed as single “pathologic” group.

Figure 5:

Unsupervised Cluster Analysis showing composite of 6 patients (A). To broadly evaluate monocyte/macrophage clusters, we explored inflammatory and regulatory markers(B) in addition to functional markers (C) using combined module scores. Annotated clusters based on top 10 gene expression and identifying markers (D). Exploration of RSM markers using multi-gene feature plot (E), as well as markers of differentiation and proliferation (F). TAM = Tumor Associated Macrophage; cDC2 = Type 2 conventional Dendritic Cell; moDC = monocyte derived Dendritic Cell; ISG = Interferon Stimulated Gene; IFIT = interferon-induced proteins with tetratricopeptide repeats; cDC1 = Type 1 conventional Dendritic Cell.

Figure 6:

The osteodestructive potential of myeloid cells in infectious and inflammatory arthritis. Inflammatory arthritis patients displayed elevated osteoclastic and fibrotic potential compared to septic arthritis patients, specifically by cells within the M2 macrophage and TAM-like clusters.

Figure 7:

Differentially Expressed Genes (DEGs) (A). Conserved genes where Log2FC > 1.5 and the change in percent expression in both septic arthritis and inflammatory arthritis was > 0.7 in each cluster compared to all other clusters. Gene names in •bold represent genes that were also in the top 10 expressed genes.

Figure 8:

ReactomeGSA[18] of cell clusters by complement, fibroblast-related pathways, phagocytosis-related pathways, and extracellular matrix/joint fluid related pathway