Translational targeting of inflammation and fibrosis in frozen shoulder: Molecular dissection of the T cell/IL-17A axis

Abstract

Frozen shoulder is a common fibroproliferative disease characterized by the insidious onset of pain and restricted range of shoulder movement with a significant socioeconomic impact. The pathophysiological mechanisms responsible for chronic inflammation and matrix remodeling in this prevalent fibrotic disorder remain unclear; however, increasing evidence implicates dysregulated immunobiology. IL-17A is a key cytokine associated with inflammation and tissue remodeling in numerous musculoskeletal diseases, and thus, we sought to determine the role of IL-17A in the immunopathogenesis of frozen shoulder. We demonstrate an immune cell landscape that switches from a predominantly macrophage population in nondiseased tissue to a T cell-rich environment in disease. Furthermore, we observed a subpopulation of IL-17A-producing T cells capable of inducing profibrotic and inflammatory responses in diseased fibroblasts through enhanced expression of the signaling receptor IL-17RA, rendering diseased cells more sensitive to IL-17A. We further established that the effects of IL-17A on diseased fibroblasts was TRAF-6/NF-κB dependent and could be inhibited by treatment with an IKKβ inhibitor or anti-IL-17A antibody. Accordingly, targeting of the IL-17A pathway may provide future therapeutic approaches to the management of this common, debilitating disease.

Keywords: IL-17A; T cell; adhesive capsulitis; frozen shoulder; inflammation.

Fig. 1.

T cells produce IL-17A in frozen shoulder. (A) Percentage and subsets of immune cells phenotyped from disaggregated shoulder capsule tissue. Results are mean ± SD, n = 4 control capsule and n = 5 frozen shoulder. Statistical analysis using unpaired Student’s t test. * indicates significant difference from control capsule tissue, *P < 0.05. (B) Uniform Manifold Approximation and Projection (UMAP) embedding of single-cell RNA sequencing and distribution of immune cells from all shoulder capsule tissue (n = 7, k = 3,347) and split into shoulder capsule of control (n = 3, k = 555 and frozen shoulder tissue (n = 4, k = 2,792). (C) UMAP embedding and distribution of eight delineated T cell populations (CD161 CCR6 T cells, inactive T cells, CD4 T cells, CD8 T cells, natural killer [NK] cells, NEAT1 T Cells, γδ T cells) in all shoulder capsule tissue (n = 7, k = 1,562) and split into shoulder capsule of control (n = 3, k = 120) and frozen shoulder tissue (n = 4, k = 1,442). (D) IL-17A gene and protein expression in shoulder capsule tissue. IL-17A expression in control and frozen shoulder capsule, 2^-ΔCT relative to GAPDH, mean ± SD, n = 5 control capsule and n = 10 frozen shoulder. Statistical analysis using unpaired Student’s t test, * indicates significant difference from control capsule tissue, *P < 0.05. Shoulder capsule tissue stained for IL-17A, isotype IgG in bottom left corner, using rabbit polyclonal IL-17A antibody at 10× and 40× magnification. Graph illustrates percentage of cells stained positive for IL-17A, mean ± SEM, n = 10 for control, n = 10 for frozen shoulder tissue. Statistical analysis using unpaired Student’s t test, ****P < 0.0001. (E) IL-17A is produced by T cells from disaggregated frozen shoulder tissue. Representative flow cytometric plots of unstimulated and stimulated CD3⁺ T cells following intracellular staining with antibodies against IFN-γ and IL-17A. Graph displays the proportion of CD4⁺ and CD8⁺ T cells positive for IFN-γ and IL-17A (n = 4).

Fig. 2.

IL-17A induces fibrosis and inflammation in vitro. (A) Effect of recombinant IL-17A on control and frozen shoulder fibroblasts viability, BCL2 gene expression, and mitochondrial and cytosolic cytochrome C content, mean ± SD, n = 4 control fibroblasts and n = 5 frozen shoulder fibroblasts, * indicates significant difference from untreated cells, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, # indicates significant difference from control fibroblasts, ^#P < 0.05, ^##P < 0.01. (B) Effect of IL-17A on COL3A1, FN1, MMP1, MMP3 gene expression and MMP3 protein secretion. mRNA gene expression expressed as fold change following normalization to housekeeping gene (GAPDH) and then to relevant untreated cells, n = 4 control fibroblasts and n = 5 frozen shoulder fibroblasts, * indicates significant difference from untreated cells *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, # indicates significant difference from control fibroblasts, ^#P < 0.05, ^##P < 0.01 ^###P < 0.001. (C) Effect of IL-17A on IL-6, IL-8, CCL20 production from control or frozen shoulder fibroblasts, n = 4 control fibroblasts and n = 5 frozen shoulder fibroblasts, * indicates significant difference from untreated cells *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, # indicates significant difference from control fibroblasts, ^#P < 0.05, ^##P < 0.01 ^###P < 0.001. All statistical analyses use two-way ANOVA with Dunnet’s correction or Sidak’s test for multiple comparisons.

Fig. 3.

Enhanced IL-17RA expression in frozen shoulder fibroblasts. IL-17RA and TRAF6 protein expression in cultured fibroblasts from control and frozen shoulder capsule. Image of Western blot of protein from control and frozen shoulder fibroblasts immunoblotted for GAPDH, TRAF6, and IL-17RA. Graph illustrates TRAF6 and IL-17RA protein quantification relative to housekeeping (GAPDH), mean ± SD, n = 4 control and frozen shoulder fibroblasts. Statistical analysis using Mann–Whitney U rank-sum test. * indicates significant difference from control capsule tissue, *P < 0.05.

Fig. 4.

IL-17A signaling is NF-κB dependent in frozen shoulder fibroblasts. (A) Schematic diagram of NF-κB pathway following IL-17A stimulation. Frozen shoulder fibroblasts were pretreated with IKKβ inhibitor before exposure to IL-17A (100 ng/mL) (B) Fibroblast’s viability, BCL2 gene expression, and mitochondrial and cytosolic cytochrome C content. (C) Gene expression of matrix and inflammatory regulators. mRNA gene expression expressed as fold change following normalization to housekeeping gene (GAPDH) and then to untreated cells (dashed line). (D) Effect on MMP3 production. (E) Effect on IL-6, IL-8, and CCL20 production. All results are mean ± SD, n = 4 control and frozen shoulder fibroblasts, * indicates significant difference from untreated cells, *P < 0.05, **P < 0.01, ***P < 0.001. ^#P < 0.05, ^##P < 0.01, ^###P < 0.001. All statistical analyses use ANOVA with Bonferroni or Dunn’s correction for multiple comparisons depending on normality.

Fig. 5.

Anti–IL-17A monoclonal targeting of IL-17A induced fibrotic and inflammatory effects in frozen shoulder. Frozen shoulder fibroblasts were pretreated with an anti–IL-17A monoclonal antibody or IgG control before exposure to IL-17A (100 ng/mL). (A) Fibroblast’s viability, BCL2 gene expression, and mitochondrial and cytosolic cytochrome C content. (B) Gene expression of matrix and inflammatory regulators. mRNA gene expression expressed as fold change following normalization to housekeeping gene (GAPDH) and then to untreated cells (dashed line). (C) Effect on MMP3 production. (D) Effect on IL-6, IL-8 and CCL20 production. All results are mean ± SD, n = 4 control and frozen shoulder fibroblasts, * indicates significant difference from untreated cells, *P < 0.05, **P < 0.01, ****P < 0.001, ^#P < 0.05, ^##P < 0.01, ^####P < 0.0001. All statistical analyses use ANOVA with Bonferroni or Dunn’s correction for multiple comparisons depending on normality.

Fig. 6.

Targeting IL-17A in frozen shoulder. Schematic diagram illustrating possible IL-17A–driven pathogenesis in frozen shoulder. The presence of IL-17A–secreting T cells results in the fibrosis, inflammation, and further recruitment of pathogenic T cells, resulting in disease chronicity, which can be attenuated by use of an anti–IL-17A monoclonal antibody.