S100B Up-Regulates Macrophage Production of IL1β and CCL22 and Influences Severity of Retinal Inflammation

Abstract

S100B is a Ca2+ binding protein and is typically associated with brain and CNS disorders. However, the role of S100B in an inflammatory situation is not clear. The aim of the study was to determine whether S100B is likely to influence inflammation through its effect on macrophages. A murine macrophage cell line (RAW 264.7) and primary bone marrow derived macrophages were used for in vitro studies and a model of retinal inflammatory disease in which pathogenesis is highly dependent on macrophage infiltration, Experimental Autoimmune Uveoretinitis, for in vitro study. Experimental Autoimmune Uveoretinitis is a model for the human disease posterior endogenous uveoretinitis, a potentially blinding condition, with an autoimmune aetiology, that mainly affects the working age group. To date the involvement of S100B in autoimmune uveoretinitis has not been investigated. Real-time PCR array analysis on RAW 246.7 cells indicated up-regulation of gene expression for various cytokines/chemokines in response to S100B, IL-1β and CCL22 in particular and this was confirmed by real-time PCR. In addition flow cytometry and ELISA confirmed up-regulation of protein production in response to S100B for pro-IL-1β and CCL22 respectively. This was the case for both RAW 264.7 cells and bone marrow derived macrophages. Induction of EAU with retinal antigen in mice in which S100B had been deleted resulted in a significantly reduced level of disease compared to wild-type mice, as determined by topical endoscopic fundus imaging and histology grading. Macrophage infiltration was also significantly reduced in S100B deleted mice. Real-time PCR analysis indicated that this was associated with reduction in CCL22 and IL-1β in retinas from S100B knock-out mice. In conclusion S100B augments the inflammatory response in uveoretinitis and this is likely to be, at least in part, via a direct effect on macrophages.

Fig 1. RAGE gene expression of RAW 264.7 macrophages and BMDM.

A) Agarose gel electrophoresis of RT-PCR products indicates RAGE band at expected band size of 350 bp and β-actin as reference gene with band size 500 bp. Mouse lung sample used as positive control (+ve), with negative control (-ve) containing water in place of cDNA. B) Western blot for RAGE showing predicted 50 KDa band for RAW264.7 macrophages and BMDM and for positive control (+ve) mouse lung.

Fig 2. PCR array for inflammatory cytokines, chemokines and receptors on RAW 264.7 macrophages treated with or without S100B (2 μM for 6 h).

Graph shows relative fold change in response to S100B for those genes with >2 fold change after S100B treatment (2^-ΔΔCT). Fold regulation normalised to GAPDH. Mean of two individual experiments each with 3 pooled samples.

Fig 3. Confirmation of increased CCL22 and IL1β expression in RAW 264.7 macrophages in response to increasing dose of S100B.

Relative fold increase in mRNA expression for CCL22 (A) and IL-1β (B) determined by real-time PCR analysis. Error bars indicate SEM. n = 6. * P<0.05, **P<0.005, ***P≤0.001 (ANOVA).

Fig 4. CCL22 production by macrophages in response to S100B treatment for 24 h as determined by ELISA on supernatants.

CCL22 production by RAW 264.7 cells (A) and BMDM cells (B). There was a significant increase in CCL22 with 1 μM S100B and 2 μM S100B in RAW 264.7 macrophage cells and with 2 μM in BMDM cells compared to untreated control cultures. Representative of 3 individual experiments. *P<0.05, **P<0.001 (ANOVA). Error bars indicate SEM n = 3.

Fig 5. Treatment of macrophages with S100B does not reduce cell viability as indicated by cell morphology and acid phosphatase assay.

Images to show BMDM untreated (A) and treated with 2 μM S100B for 24 h (B) and RAW 264.7 untreated (C) and treated with 2 μM S100B for 24 h (D). Scale bars = 50 μM. Acid phosphatase cell viability assay on RAW 264.7 untreated or treated with 2μM or 5 μM S100B for 24 h (E) and on BMDM (F). No significant difference in viability following treatment with 2μM S100B for 24 h. Triton (0.1%) and DMSO (10%) as positive controls for the assay. Error bars indicate SEM n = 6.

Fig 6. S100B treatment of macrophages results in increased pro-IL1β expression.

Representative dot plots showing flow cytometric analysis of RAW 264.7 cells using anti-CD11b (PerCP-Cy 5.5) to gate macrophages and anti-pro-IL1β (PE) to detect intracellular staining in untreated cells (A) and in cells treated with 2μM S100B (B). Significant increase in the percentage of CD11b positive cells expressing pro-IL-1β in response to S100B is shown graphically in both RAW 264.7 macrophage cells (*P<0.05) (C) and BMDM (***P<0.001) (D). Error bars indicate SEM n = 4.

Fig 7. Increase in S100B was detected in retinal sections from WT mice with EAU compared to untreated mice using immunohistochemistry.

(A) Positive staining was observed in the naïve retina, specifically in the outer plexiform layer (a), inner plexiform layer (b) and retinal ganglion layer (c). Increased staining was observed in EAU diseased sections (B) in the retinal ganglion layer (d) and in the rod outer segments where positive staining was located around infiltrating cells (e). No positive staining was observed in control sections, where S100B antibody was replaced with PBS in naïve retina (C) or in EAU diseased retina (D). Serial sections from naïve and diseased mice taken at spaced intervals throughout the eye.

Fig 8. EAU TEFI grading in C57BL/6 and S100B KO mice at day 15 and day 21 pi.

A, vascular cuffing (a) indicating inflammation in C57BL/6 day 15 pi. Severe inflammation in C57BL/6 at day 21 pi, with swollen, barely visible optic disk (b), cell infiltrates (c), and progression of vascular cuffing. In comparison, S100B KO images showing reduced disease but evidence of inflammation. Optic disk appears swollen day 21 pi (d) with vascular cuffing occurring on both sides of vessels (e). Approximate diameter of eye 3.3 mm. B, TEFI clinical grading scores for EAU at day 15 pi and C, day 21 pi indicating significantly reduced disease in S100B KO mice (*P<0.05). TEFI scores for individual eyes from the same mouse were averaged and each point represents one animal. Results from two independent experiments were combined.(C57BL/6 n = 18 mice S100B KO n = 22 mice). Error bars indicate SEM.

Fig 9. Histological grading of EAU in C57BL/6 WT and S100B KO retinal sections at day 24 pi.

Eyes were snap frozen and serially sectioned at 6 μm. Sections were graded for EAU. A, section showing EAU disease in C57BL/6 mouse (infiltrative grade 3) with folding of retinal layers (a) and cell infiltrates in the vitreous (b) and retinal layers. B, section showing reduced EAU disease (infiltrative grade 2) in S100B KO mice where layers have remained intact, showing cells gathering around vessel wall (c) and infiltrating cells in the vitreous (d). C, significantly decreased levels of retinal infiltrating cells were seen in sections from S100B KO mice (*P<0.05). Grading scores for individual eyes from the same mouse were averaged and each point represents one animal. C57BL/6 n = 8 mice, S100B KO n = 10 mice. Error bars indicate SEM.

Fig 10. Monocyte/Macrophage infiltration in sections from C57BL/6 and S100B KO mice.

A, C57BL/6 WT retinal section from mouse with EAU showing cells staining positively for MOMA-2 (red). B, positive MOMA-2 staining in S100B KO with EAU was also observed in the rod outer segments and vitreous. No positively stained cells were observed in negative control staining in C57BL/6 mice (C) or S100B KO mice with EAU (D), where primary antibody was replaced with TBS. No positive staining was observed in naïve C57BL/6 WT (E) or S100B KO mice (F). Cells positive for MOMA-2 were counted in a total area of 1.5 mm². An average cell number was calculated from 3 sections taken at spaced intervals through the eye. A significant reduction in positively stained cells was observed in S100B KO mice compared to C57BL/6 WT mice both in the retinal layers overall (G) and specifically within the rod outer segments (ROS) (H). Error bars indicate SEM; n = 8 randomly selected eyes per genotype. Scale bars indicate 100 μm.