Local expression of tumor necrosis factor-receptor 1:immunoglobulin G can induce salivary gland dysfunction in a murine model of Sjögren's syndrome

Abstract

Introduction: Tumor necrosis factor is a pleiotropic cytokine with potent immune regulatory functions. Although tumor necrosis factor inhibitors have demonstrated great utility in treating other autoimmune diseases, such as rheumatoid arthritis, there are conflicting results in Sjögren's syndrome. The aim of this study was to assess the effect of a locally expressed tumor necrosis factor inhibitor on the salivary gland function and histopathology in an animal model of Sjögren's syndrome.

Methods: Using in vivo adeno associated viral gene transfer, we have stably expressed soluble tumor necrosis factor-receptor 1-Fc fusion protein locally in the salivary glands in the Non Obese Diabetic model of Sjögren's syndrome. Pilocarpine stimulated saliva flow was measured to address the salivary gland function and salivary glands were analyzed for focus score and cytokine profiles. Additionally, cytokines and autoantibody levels were measured in plasma.

Results: Local expression of tumor necrosis factor-receptor 1:immunoglobulin G fusion protein resulted in decreased saliva flow over time. While no change in lymphocytic infiltrates or autoantibody levels was detected, statistically significant increased levels of tumor growth factor-beta1 and decreased levels of interleukin-5, interleukin-12p70 and interleukin -17 were detected in the salivary glands. In contrast, plasma levels showed significantly decreased levels of tumor growth factor-beta1 and increased levels of interleukin-4, interferon-gamma, interleukin-10 and interleukin-12p70.

Conclusions: Our findings suggest that expression of tumor necrosis factor inhibitors in the salivary gland can have a negative effect on salivary gland function and that other cytokines should be explored as points for therapeutic intervention in Sjögren's syndrome.

Figure 1

In vitro biological activity assay for TNFR1:IgG. The biological activity of the TNFR1:IgG was determined by adding supernate from TNFR1:IgG expressing cells to WEHI fibrosarcoma cells and measuring its ability to block TNF induced apoptosis in these cells. Increasing concentrations of TNFR1:IgG resulted in inhibiting TNF (62.5 pg/ml) induced apoptosis of WEHI cells. Data shown are mean ± SD (N = 2 independent experiments).

Figure 2

Diabetes, saliva flow, and body weight in untreated and vector treated groups. A) Blood sugar was monitored starting at week 13 (five weeks post vector delivery) as described in the Materials and Methods section. Data shown are the mean percentage from two independent experiments (N = 8 for untreated, N = 20 for LacZ and N = 19 for TNFR1:IgG). B) Saliva was collected as described in the Materials and Methods section over a 20-minute period after stimulation with 0.5 mg/kg BW pilocarpine. Data shown are mean values +/- SEM (N = 8 for untreated, N = 20 for LacZ and N = 19 for TNFR1:IgG). P-values are indicated and were determined by non-parametric Wilcoxon's ranksum test. Delivery of TNFR1:IgG resulted in significant decreased saliva flow not seen in control groups. C) The weight of the mice (in grams) was measured at the indicate times for LacZ, TNFR1:IgG, and untreated mice. There were no significant changes in bodyweight between all three groups as determined by unpaired student's t-test. Data shown are the mean (+/- SD) of two independent experiments (N = 8 for untreated, N = 11 for LacZ and N = 12 for TNFR1:IgG per experiment).

Figure 3

Focus score and infiltrate analysis in untreated and vector treated groups. A) Salivary glands were removed for histologic analysis at 24 weeks, fixed and stained with Hematoxylin and eosin. Histopathologic assessment was performed and presented as a focus score (see Materials and Methods). Data shown are the mean (+/- SD). No significant change was detected by one-way ANOVA test followed by unpaired student's t-test. Untreated (N = 7), LacZ (N = 8) and TNFR1:IgG (N = 6) mice. B) Acetone fixed frozen sections from SGs were analyzed for CD4, CD8, CD19 and CD138 by immunohistochemistry followed by digital analysis as described in the Material and Methods section. Data shown are the mean positive cell counts per mm² (+/- SEM). Higher means were observed for CD4, CD8 and CD138 in the TNFR1:IgG treated mice, but no significant changes were detected by non-parametric Wilcoxon's ranksum test. LacZ (N = 4) and TNFR1:IgG (N = 8).

Figure 4

Correlation cytokines with salivary flow. Cytokine data were analyzed for correlation with salivary flow from mice at 20 weeks of age using the non-parametric Spearman's Rho test. The correlation coefficient (r) and P-values are indicated. Data on the x-axis for figure A are plotted in log-scale. Strong correlations were found for hTNFR1 (A) in SGs and IL-12p70 (B), IL-4 (C), IL-5 (D) in plasma.

Figure 5

Autoantibodies in plasma. Plasma samples were collected and autoantibody levels were measured as described in the Material and Methods section using samples collected at 24 weeks and compared with four different pools of pre-disease six-week-old mice. Antibody levels for SSA/Ro (A) SSB/La (B) and ANA (C) are shown (mean ± SD). Significant differences are indicated (*) and were determined by one-way ANOVA test followed by unpaired student's t-test. LacZ (N = 8) and TNFR1:IgG (N = 6).