All- Trans-Retinoic Acid Augments the Histopathological Outcome of Neuroinflammation and Neurodegeneration in Lupus-Prone MRL/lpr Mice

Abstract

Recently, we demonstrated that treatment with all- trans-retinoic acid (tRA) induced a paradoxical effect on immune activation during the development of autoimmune lupus. Here, we further describe its negative effects on mediating neuroinflammation and neurodegeneration. Female MRL/lpr mice were orally administered tRA or VARA (retinol mixed with 10% tRA) from 6 to 14 weeks of age. Both treatments had a significant effect on brain weight, which correlated with histopathological evidence of focal astrogliosis, meningitis, and ventriculitis. Infiltration of CD138- and Iba1-positve immune cells was observed in the third ventricle and meninges of treated mice that co-labeled with ICAM-1, indicating their inflammatory nature. Increased numbers of circulating plasma cells, autoantibodies, and total IgG were also apparent. IgG and C3 complement deposition in these brain regions were also prominent as was focal astrogliosis surrounding the ventricular lining and meninges. Using Fluoro-Jade staining, we further demonstrate that neuroinflammation was accompanied by neurodegeneration in the cortex of treated mice compared with vehicle controls. These findings indicate that vitamin A exposure exacerbates the immunogenic environment of the brain during the onset of systemic autoimmune disease. Vitamin A may therefore compromise the immuno-privileged nature of the central nervous system under a predisposed immunogenic environment.

Keywords: inflammation; lupus; retinoic acid.

Figure 1.

Brain weight in 14-week-old MRL/lpr mice treated with tRA or VARA. MRL mice were given, orally and daily, tRA (6 mg/kg BW) or VARA (11.2 mg retinyl palmitate/kg BW, equivalent to 6 mg retinol/kg BW, mixed with 0.6 mg tRA/kg BW that was 10% of the amount of retinol) from 6 to 14 weeks of age when brain and body weights were collected. (A) Brain weight in grams (g) graphically represented for MRL and MRL/lpr vehicle (canola oil)-treated and MRL/lpr tRA- or VARA-treated mice. (B) Brain weight as a ratio of total BW in the four groups. (C) Number of inflamed loci with leukocyte infiltration on each H&E slide that contained coronal brains sections of the following structures: olfactory bulb, head of caudate nucleus, rostral level of hippocampus, caudal level of hippocampus, mid-level of cerebellum with underlying medulla oblongata, and caudal level of cerebellum with underlying medulla oblongata. *p<0.05, **p<0.01, ***p<0.001, one-way ANOVA. Data are shown as mean ± standard error of the mean (SEM), n=6 mice in each group. (D) Representative images of inflammatory infiltrates are shown. Bar = 100 µm. Abbreviations: tRA, all-trans-retinoic acid; BW, body weight; H&E, hematoxylin and eosin; ANOVA, analysis of variance.

Figure 2.

Astrogliosis in the brain of MRL/lpr mice treated with tRA or VARA. Immunohistochemistry of brain sections from MRL/lpr mice treated with vehicle (A and B), tRA (C and D), or VARA (E and F), using antibodies against GFAP. GFAP staining was substantially increased around the DV3 and meningeal lining between the two cortical hemispheres (white arrows) in tRA- and VARA-treated compared with vehicle-treated MRL/lpr mice. Scale bar = 400 µm. Abbreviations: tRA, all-trans-retinoic acid; GFAP, glial-fibrillary acidic protein; DV3, dorsal third ventricle; DAPI, 4′,6-diamidino-2-phenylindole; Ctx, cortex.

Figure 3.

Identification of cellular infiltrates in the meninges of MRL/lpr mice treated with tRA. Brain sections of MRL/lpr tRA-treated mice were used to identify the cellular infiltrate in the meninges between the dorsal aspect of the cerebral hemispheres and dorsal to the CC at the level of the habenular nuclei (A). Antibodies against plasma cell-specific protein CD138 (red) showed accumulation of plasma cells in the DV3 and meninges (white arrows) (B) as well as monocyte/macrophages as seen by anti-Iba-1 staining (C). Sections were further characterized for inflammatory adhesion molecules ICAM-1 and E-selectin (D–F). High expression of ICAM (white arrows; D) was seen indicating their inflammatory nature whereas E-selectin was absent (E) in the infiltrates. Non-specific staining was also observed in several areas indicated by yellow arrow heads. Scale bar = 200 µm. Abbreviations: tRA, all-trans-retinoic acid; DV3, dorsal third ventricle; CC, corpus callosum; Ctx, cortex.

Figure 4.

Complement C3 and IgG deposition in the DV3 of MRL/lpr mice treated with tRA and VARA. Double immunofluorescence labeling of complement protein C3 (red) and IgG deposition (green) in the brain of vehicle (A–C), tRA (D–F), and VARA-treated (G–I) MRL/lpr mice at 14 weeks. Scale bar = 1000 µm. Expression of C3 protein and IgG accumulation was substantially increased following tRA and VARA administration compared with vehicle control (D1, G1, and A1 insets, respectively). Scale bar = 400 µm. Abbreviations: tRA, all-trans-retinoic acid; CC, corpus callosum; DV3, dorsal third ventricle.

Figure 5.

Frequency of CD44⁺Ly6c⁺ cells and levels of antibodies in the blood of MRL/lpr mice treated with tRA and VARA. (A) Percentage of CD44⁺Ly6c⁺ cells in the blood of tRA- and VARA-treated MRL/lpr mice was quantified using flow cytometry. A significant increase was observed in tRA and VARA compared with vehicle-treated MRL/lpr mice. (B) Anti-dsDNA IgG in the plasma of 14-week-old MRL/lpr mice as detected by ELISA. (C) Total IgG concentrations (mg/ml) in the plasma were also detected by ELISA. Data show significant increases in both anti-dsDNA and total IgG levels in tRA- and VARA-treated mice. *p<0.05, **p<0.01, one-way ANOVA. Data are shown as mean ± SEM, n=6 mice in each group. Abbreviations: dsDNA, anti-double-stranded DNA; ANOVA, analysis of variance; SEM, standard error of the mean.

Figure 6.

Neurodegeneration and astrogliosis in the cortex of tRA- and VARA-treated MRL/lpr mice. FJC staining (green) was used to visualize neurodegeneration in the cortex of (A) vehicle, (B) tRA-, and (C) VARA-treated MRL/lpr mice at 14 weeks. Compared with vehicle-treated mice, tRA and VARA display a significant increase in Fluoro-Jade positive neurons (yellow arrows), indicating ongoing neurodegeneration in the cortex following administration of these retinoic acid derivatives. Scale bar = 500 µm. (D–F) Representative images of anti-GFAP staining (red) in vehicle (D), tRA- (E), and VARA-treated MRL/lpr mice (F). (G–I) Representative H&E of the same brain region of the three groups of mice, respectively. (J) Quantified data of the estimated number of FJC-positive cells in the cortex. ***p<0.001. Data shown as mean ± SEM; n=5–6 mice per group. Abbreviations: tRA, all-trans-retinoic acid; GFAP, glial-fibrillary acidic protein; H&E, hematoxylin and eosin; FJC, Fluoro-Jade C; SEM, standard error of the mean.