A thyroid hormone network exists in synovial fibroblasts of rheumatoid arthritis and osteoarthritis patients

Abstract

While patients with rheumatoid arthritis (RA) sometimes demonstrate thyroidal illness, the role of thyroid hormones in inflamed synovial tissue is unknown. This is relevant because thyroid hormones stimulate immunity, and local cells can regulate thyroid hormone levels by deiodinases (DIO). The study followed the hypothesis that elements of a thyroid hormone network exist in synovial tissue. In 12 patients with RA and 32 with osteoarthritis (OA), we used serum, synovial fluid, synovial tissue, and synovial fibroblasts (SF) in order to characterize the local thyroid hormone network using ELISAs, immunohistochemistry, imaging methods, tissue superfusion studies, cell-based ELISAs, flow cytometry, and whole genome expression profiling. Serum/synovial fluid thyroid hormone levels were similar in RA and OA (inclusion criteria: no thyroidal illness). The degradation product termed reverse triiodothyronine (reverse T3) was much lower in serum compared to synovial fluid indicating biodegradation of thyroid hormones in the synovial environment. Superfusion experiments with synovial tissue also demonstrated biodegradation, particularly in RA. Cellular membrane transporters of thyroid hormones, DIOs, and thyroid hormone receptors were present in tissue and SF. Density of cells positive for degrading DIOs were higher in RA than OA. TNF increased protein expression of degrading DIOs in RASF and OASF. Gene expression studies of RASF revealed insignificant gene regulation by bioactive T3. RA and OA synovial tissue/SF show a local thyroid hormone network. Thyroid hormones undergo strong biodegradation in synovium. While bioactive T3 does not influence SF gene expression, SF seem to have a relay function for thyroid hormones.

Figure 1

Schematic drawing of the thyroid hormone network. Thyrotropin (TSH) from the pituitary gland stimulates the thyroid to produce T3 and T4. In serum, these hormones are bound to transport proteins. Thyroid hormones enter cells via thyroid hormone transporters (MCT8, solute carrier family 16 member 2). Intracellularly, the biologically inactive T4 is converted to the biologically active T3 and to the degradation product reverse T3 (rT3). T3 and rT3 can be further converted to T2, which has no classical thyroid hormone activities. Conversion happens with deiodinases 1–3 (DIO). T3 exerts its action via the retinoic acid receptor (RXR) and the thyroid hormone receptor alpha and beta (TRα, TRβ). Abbreviations: T3, triiodothyronine; T4, thyroxine; TRH, thyrotropin releasing hormone.

Figure 2

Serum levels of anti-thyroid peroxidase antibodies and anti-thyroglobulin antibodies in Patients with rheumatoid arthritis (RA) and osteoarthritis (OA) under study. Every point represents the mean value of two replicates of one patient. The dashed line represents the upper normal limit. The data show that patients did not demonstrate suspicious autoantibody levels indicative of autoimmune thyroid disease. The autoantibodies were measured by standard ELISAs (Cat. No. RE75511 and RE70951, IBL International, Hamburg). No statistical comparisons were carried out.

Figure 3

Levels of reverse T3 and ratio of reverse T3/free T4 in serum and synovial fluid. Every line represents the pair of values of one patient. The p-value for the comparison with Wilcoxon paired rank test is given. Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; synov. fluid, synovial fluid; T3, triiodothyronine; T4, thyroxine.

Figure 4

Interrelation between serum thyrotropin (TSH) and synovial fluid TSH. Every symbol in the graphs represent one individual patient with osteoarthritis (OA) and rheumatoid arthritis (RA). The linear regression line, the Spearman Rank correlation coefficient (R_Rank), and the respective p-value are given.

Figure 5

Synovial density of deiodinase (DIO)+ cells and levels of reverse T3 in synovial tissue superfusate of osteoarthritis and rheumatoid arthritis patients. (A) Immunohistochemistry of DIO2 and DIO3 including negative controls (neg. Co.). Scale bar = 60 µm. (B) Synovial density of DIO2+ and DIO3+ cells per mm². Every black symbol in B represents the mean value of 17 investigated synovial tissue high power fields of one patient. (C) Levels of reverse T3 in synovial tissue superfusate. Every black symbol in C represents the value of one patient. Box plots in (B,C) demonstrate the 10th (whisker), 25th, 50th (median), 75th, and 90th (whisker) percentile. For statistical comparisons in panel B and C, the Mann-Whitney-test was used. Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; T3, triiodothyronine.

Figure 6

Synovial cells positive for deiodinase 1 (DIO1), the thyroid hormone transporter MCT8 (solute carrier family 16 member 2), thyroid hormone receptor alpha (TRα), and thyroid hormone receptor beta (TRβ). Scale bar = 60 µm. Abbreviations: Co., control staining; OA, osteoarthritis; RA, rheumatoid arthritis.

Figure 7

Immunohistochemistry of trace amine associated receptor 1 (TA1) in synovial tissue of a patient with osteoarthritis (OA) and rheumatoid arthritis (RA). The scale bar represents 60 µm. A control with unspecific immunoglobulin as the primary antibody is demonstrated.

Figure 8

Immunocytochemistry of several markers of the thyroid hormone network in cultured synovial fibroblasts. The example of an RA patient is given. Abbreviations: TRα, thyroid hormone receptor type alpha; TRβ, thyroid hormone receptor type beta.

Figure 9

Modulation of protein expression of deiodinase 2 (DIO2), DIO3, thyroid hormone receptor alpha (TRα) and TRβ by TNF in synovial fibroblasts of patients with osteoarthritis (OA) and rheumatoid arthritis (RA). Box plots demonstrate the 10th (whisker), 25th, 50th (median), 75th, and 90th (whisker) percentile. Every black symbol represents the mean of duplicate values of one patient. The One-sample-Wilcoxon test was used for comparisons with the 100% control. *p < 0.05, #p < 0.008.

Figure 10

Median toll-like receptor 4 (TLR4) cellular surface expression on RA synovial fibroblasts. RA synovial fibroblasts were stimulated with 10⁻⁷ mol/l T3 over 24 and 48 hours. Surface expression of TLR4 was determined by flow cytometry analysis. Box plots demonstrate the 10th (whisker), 25th, 50th (median), 75th, and 90th (whisker) percentile. Every black symbol represents the mean of duplicate values of one patient (total: n = 4 patients). For comparisons of medians at 24 and 48 hours with the zero hour control (0 h), the Mann-Whitney-test was used.