TLR7 Signaling Drives the Development of Sjögren's Syndrome

Abstract

Sjögren's syndrome (SS) is a chronic systemic autoimmune disease that affects predominately salivary and lacrimal glands. SS can occur alone or in combination with another autoimmune disease like systemic lupus erythematosus (SLE). Here we report that TLR7 signaling drives the development of SS since TLR8-deficient (TLR8ko) mice that develop lupus due to increased TLR7 signaling by dendritic cells, also develop an age-dependent secondary pathology similar to associated SS. The SS phenotype in TLR8ko mice is manifested by sialadenitis, increased anti-SSA and anti-SSB autoantibody production, immune complex deposition and increased cytokine production in salivary glands, as well as lung inflammation. Moreover, ectopic lymphoid structures characterized by B/T aggregates, formation of high endothelial venules and the presence of dendritic cells are formed in the salivary glands of TLR8ko mice. Interestingly, all these phenotypes are abrogated in double TLR7/8-deficient mice, suggesting that the SS phenotype in TLR8-deficient mice is TLR7-dependent. In addition, evaluation of TLR7 and inflammatory markers in the salivary glands of primary SS patients revealed significantly increased TLR7 expression levels compared to healthy individuals, that were positively correlated to TNF, LT-α, CXCL13 and CXCR5 expression. These findings establish an important role of TLR7 signaling for local and systemic SS disease manifestations, and inhibition of such will likely have therapeutic value.

Keywords: Sjögren’s syndrome (SS); TLR7; TLR8-deficient mice; autoimmunity; dendritic cells; innate immunity; transgenic mice.

Figure 1

TLR8ko mice develop TLR7-dependent SS. Spleens and salivary gland were harvested from 8- and 14-months-old TLR8ko, TLR7/8ko and WT female mice. (A) Spleen weight of TLR8ko, TLR7/8ko and WT mice. (B) Salivary gland tissues were sectioned and representative areas of H&E stained sections are shown. Panels in upper row magnification x1, panels in lower row magnification x4 of the black frames indicated in the upper row panels. Black arrows denote lymphocytic infiltration. (C) Pathological scores of SG were evaluated and results are expressed as mean +/- SD of total foci area versus SG area x 100. Scale bars: 0.5 mm. (D) Levels of anti-SSA, anti-SSB and anti-RNA antibodies in sera of 8- and 14-months-old TLR8ko, TLR7/8ko and WT female mice were evaluated by ELISA. (E) Representative microphotographs of immunofluorescence detection of IgG and IgM in SG sections from TLR8ko, TLR7/8ko and WT female mice. Scale bars: 100 μm. Statistical analysis was done by Kruskal-Wallis followed by Wilcoxon rank sum tests and correction for multiple comparisons using the Benjamini-Hochberg method. In (A, B, D) each time point represents the value of one mouse and horizontal bars denote mean. In (A–D) pooled data from 4 independent experiments, and in (E) data are representative of 2 independent experiments with 3-5 mice per genotype. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2

Lung inflammation in TLR8ko mice. Lung inflammation scoring at (A) 8-months and (B) 14 months old WT, TLR8ko and TLR7/8ko mice. (C) Representative micrographs of lungs from 8- and 14-months old WT, TLR8ko and TLR7/8ko mice. H&E-stained lung sections show inflammatory foci (arrows). Scale bars: 0.5 mm. Statistical analysis was done by Kruskal-Wallis followed by Wilcoxon rank sum tests and correction for multiple comparisons using the Benjamini-Hochberg method. Data are representative of 2 independent experiments combined. *P < 0.05, ***P < 0.001.

Figure 3

Identification of TLS in the infiltrates of TLR8ko SMG. Immunofluorescent microscopic analysis of lymphocyte infiltrates in SMG of 8-months old TLR8ko mice. Frozen sections were stained with fluorescent labelled (A) CD3 for T cells (red), B220 for B cells (violet) and CD11c for DCs (green) and (B) Meca79 for HEV (green), CD11c for DCs (red) and CD3 for T cells (violet). Scale bars: 100 μm; magnification x20. (C) Expression of the HEV markers Glycam1 and Gcnt1 was evaluated in the SMG of WT, TLR8ko and TLR7/8ko mice by Q-PCR. Plots represent mean ± SD of triplicates, with n=5-6 mice combined per genotype. Statistical analysis was done by one-way ANOVA followed by Tukey’s multiple-comparison test. (A–C) Results are representative of two independent experiments with 5-6 mice per genotype and per experiment. **P < 0.01, ****P < 0.0001.

Figure 4

Increased levels of cytokines and B cell markers in TLR8ko SMG. Expression of (A) Il-6, Tnf and Lt-α and (B) Cxcl13, Cxcr5 and Baff mRNA levels were evaluated in the SMG of 8- and 14-months old female WT, TLR8ko and TLR7/8ko mice by Q-PCR. Plots represent mean ± SD of triplicates, with n=5-6 mice combined per genotype. Statistical analysis was done by one-way ANOVA followed by Tukey’s multiple-comparison test. Data are representative of at least two independent experiments. *P <0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5

Increased TLR7 expression in the SG of the SS patients is positively correlated with TNF, LT-α, CXCL13 and CXCR5 expression levels. RNA was extracted from the salivary glands of female patients with primary SS and age matched sicca donors and the expression levels of (A) TLR7 and (B) TNF, LT-α, CXCL13 and CXCR5 were evaluated by Q-PCR. Each point represents one individual and horizontal lines correspond to mean. Number of individuals (n) per group are denoted in parenthesis. Statistical analysis was done by non-parametric Mann-Whitney test. (C) Correlation analysis was performed to investigate the expression relationship between TLR7 and TNF, LT-α, CXCL13 and CXCR5 expression levels, and were expressed as Pearson correlation coefficient (r). Pearson’s correlation and P values are listed at the top of each mini-plot. Each point represents one individual. For SS patients (n=19-40) and for sicca controls (n=7-11). *P < 0.05, **P < 0.01, ****P < 0.0001, ns, non statistical.