The selective retinoic acid receptor-α agonist AM580 fails to control autoimmune neuroinflammation

Environmental cues provided by products from the diet, gut microbiota metabolism, and exposure to sunlight are sensed by specific receptors that drive a specific genomic profile in immune cells and influence the outcome of multiple sclerosis (MS), an autoimmune disease of the central nervous system (CNS),¹ which affects approximately 2.3 million people worldwide.² Among these receptors, the ligand-activated transcription factor retinoic acid receptor α (RAR-α) senses retinoic acid and other ligands to promote the transcription of genes that have retinoic acid receptor elements in their promoter region.³ Supplementation with all-trans retinoic acid (ATRA), a metabolite of vitamin A, ameliorates inflammation in MS and the corresponding animal model, experimental autoimmune encephalomyelitis (EAE), by shifting the balance between Th17/regulatory T cells (Tregs) and inducing tolerogenic dendritic cells.^4,5 To overcome the weaknesses associated with ATRA agonists, e.g., instability, poor bioavailability and nonselective binding to a broad range of retinoid receptors, a variety of synthetic agonists specifically targeting RAR have been developed.⁶ One of them, AM80, inhibits Th17 cells and suppress acute neuroinflammation; however, continuous AM80 treatment is ineffective, most likely because AM80 also inhibits Tregs and IL-10 production.⁶ AM580, a stable benzoic derivative of retinoic acid and a selective RAR-α agonist, has been recently shown to significantly reduce the production of Th1 cytokines but promote Th2 cytokines in human PBMCs⁷ and to inhibit microglial activation, thus acting beneficially in Alzheimer’s disease treatment.⁸ In addition, AM580 protects retinal cells against diabetes-induced apoptosis by inducing neurotrophic factors.⁹ However, whether it has a protective effect on neuroinflammation is unknown.

Given the dual effects of AM580, i.e., anti-inflammatory^7,8 and neuroprotective effects,⁹ we hypothesized that AM580 could be highly effective in suppressing EAE development. To test this hypothesis, EAE was induced in C57BL/6 mice, and disease severity was scored daily in a blinded manner by two investigators following a 0–5 scale previously described.⁵ The mice were then treated by intraperitoneal (i.p.) injections of either vehicle [80% dimethyl sulfoxide (DMSO) in distilled water] or 1 mg/kg or 2 mg/kg AM580 (Sigma-Aldrich, St. Louis, MO, USA) diluted in DMSO, following a protocol effective for the treatment of Alzheimer’s disease.⁸ Treatment was started at disease onset, i.e., on day 11 post immunization (p.i.), and administered every other day until day 21 p.i. 5. Our results showed that untreated EAE mice first showed signs of disease at day 12 p.i., and disease progressed to a peak at day 19 p.i. AM580 was ineffective in suppressing disease at either the 1 or 2 mg/kg dose (Fig. 1a). It should be noted that the mice treated with 2 mg/kg AM580 had delayed disease severity, but the severity reached the same level as that of the untreated mice at day 20 p.i. (Fig. 1a). These results suggest that AM580, even at a high dose, is not effective in fully suppressing EAE development. We then evaluated whether AM580 could influence demyelination and cellular infiltration into the CNS. Thus, the spinal cord was collected to analyze demyelination by Luxol fast blue staining and leukocyte infiltration by H&E staining. We found that demyelination and leukocyte infiltration was similar between the AM580-treated mice and vehicle-treated EAE mice (Fig. 1b, c). The number of infiltrating CD4⁺ T cells in the CNS of EAE mice remained unchanged (Fig. 1d).

Fig. 1.

AM580 fails to suppress EAE. C57BL/6 mice (8–10 weeks old) were immunized with MOG35-55 to induce EAE and treated i.p. with AM580 or vehicle every other day, starting on day 11 p.i. a Clinical disease was analyzed and scored daily by two blinded researchers. At day 21 p.i., the lumbar spinal cord was harvested for Luxol fast blue (b) and H&E staining (c). The total area of demyelination and cellular infiltration were calculated with ImageJ software. d CD4⁺ T cells in the spinal cord were determined by immunohistochemistry. Positive staining per area was calculated with ImageJ software. e Mononuclear cells from the spleen and CNS were analyzed by flow cytometry. Representative dot plots show the gated CD4⁺ T cells and a summary of the percentages of IL-17⁺ and IFN-γ⁺ cells among the total CD4⁺ T cells. Analyses between groups were performed with two-way ANOVA and the Bonferroni post hoc test. **p < 0.01. Data represent the mean ± SD (n = 5 each group). One representative of three independent experiments with similar outcomes is shown

IFN-γ-producing Th1 and IL-17-producing Th17 cells are major drivers of CNS inflammation and promote damage to myelin sheaths.¹⁰ We thus investigated whether the ineffectiveness of AM580 in suppressing EAE development was due to the failure of AM580 to modulate either or both of these encephalitogenic T cell populations. Mononuclear cells were harvested from the CNS and spleen of EAE mice at day 21 p.i., and IFN-γ and IL-17 production within the gated CD4⁺ T cell population was analyzed. We found that AM580 did not produce significant changes in the cells that produced these cytokines (Fig. 1e). No significant difference was observed for Foxp3 expression or IL-10 production in CD4⁺ T cells between the AM580- and vehicle-treated mice (data not shown). Collectively, our results show that the RAR-α agonist AM580 is ineffective in suppressing CNS inflammation and EAE development.

Overall, we provide evidence that the RAR-α-specific agonist AM580 fails to suppress the clinical development of EAE, even though it seemed to be a strong candidate for the treatment of MS/EAE based on its previously reported dual effects, i.e., the anti-inflammatory effect in vitro⁸ and neuroprotective effect in vivo.⁹ Attempting to find more effective and safer RAR-specific agonist(s) as an MS/EAE therapy is thus warranted.

Competing interests

The authors declare no competing interests.