Hydroxychloroquine Alleviates EAU by Inhibiting Uveitogenic T Cells and Ameliorating Retinal Vascular Endothelial Cells Dysfunction

Abstract

Purpose: Inflammation triggers the activation of CD4⁺T cells and the breakdown of blood-retinal barrier, thus contributing to the pathology of experimental autoimmune uveitis (EAU). We explored the anti-inflammatory effect of hydroxychloroquine (HCQ) on EAU and the potential mechanisms active in T cells and retinal vascular endothelial cells (RVECs).

Methods: C57BL/6J mice were immunized with interphotoreceptor retinoid binding protein 1-20 (IRBP_1-20) to induce EAU and then treated with the vehicle or HCQ (100 mg/kg/day). On day 7, 14, 21, 30 and 60 after immunization, clinical scores were evaluated. On day 14, histopathological scores were assessed, and retinas, spleens, and lymph nodes were collected for quantitative polymerase chain reaction or flow cytometry analysis. RVEC dysfunction was induced by tumor necrosis factor α (TNF-α) stimulation. The expression of cytokines, chemokines, adhesion molecules, and lectin-like oxidized LDL receptor-1 (LOX-1)/nuclear factor κB (NF-κB) was measured in RVECs with or without HCQ.

Results: HCQ treatment protected mice from uveitis, evidenced by reduced expression of inflammatory factors, chemokines, and adhesion molecules in the retina. In systemic immune response, HCQ inhibited the activation of naïve CD4⁺T cells and frequencies of T effector cells, and promoted T regulatory cells. HCQ decreased IRBP_1-20-specific T cell responses and proliferation of CD4⁺T cells in vitro. Further studies established that TNF-α induced RVECs to express inflammatory cytokines, chemokines, and adhesion molecules, whereas HCQ alleviated the alterations via the LOX-1/NF-κB pathways.

Conclusions: HCQ alleviates EAU by regulating the Teff/Treg balance and ameliorating RVECs dysfunction via the LOX-1/NF-κB axis. HCQ may be a promising therapeutic candidate for uveitis.

Keywords: Hydroxychloroquine; T cells; experimental autoimmune uveitis; lectin-like oxidized LDL receptor-1 (LOX-1); nuclear factor κB (NF-κB); retinal vascular endothelial cells.

Figure 1

Hydroxychloroquine (HCQ) alleviates retinal inflammation in experimental autoimmune uveitis (EAU). C57BL/6J mice immunized with hIRBP1-20 were treated with HCQ (100 mg/kg) or vehicle daily. (A, B) Representative photos of fundus showed attenuation of retinal inflammation by HCQ, characterized by fewer linear lesions and minimal vasculitis of the retina. HCQ had persistent anti-inflammatory effects on EAU till day 60 after immunization. The clinical scores of EAU in the two groups were evaluated and compared. (N=5). (C, D) The eyes were enucleated and prepared for staining. Representative histopathological images showing that HCQ considerably reduced the structural damages. The histopathological scores were statistically different (N=5). (E–K) RT-qPCR results showed that the expression of IL-17A, IFN-γ, IL-1β, CXCL2, CXCL10, CXCL11, and ICAM mRNAs in retina was upregulated in uveitic retinas, whereas HCQ treatment considerably decreased the mRNA expression of these genes (N=4). The data are presented as the mean ± SD (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure 2

Hydroxychloroquine (HCQ) maintains systemic Teff/Treg balance to ameliorate experimental autoimmune uveitis (EAU). The lymphocytes were isolated from the spleens and draining lymph nodes (dLNs) of C57BL/6J mice on day 14 after immunization (N=5). Flow cytometry experiments were gated on CD4⁺ cells. (A, B) HCQ inhibited the naïve CD4⁺T cell (CD62L+CD44-) differentiation into memory CD4⁺T cells (CD62L-CD44+) compared with EAU treated with the vehicle, in both spleens and dLNs. (C, D) The proportion of Treg cells in spleens and dLNs was promoted by HCQ administration in EAU mice. (E‐H) HCQ decreased the frequencies of Th1 and Th17 cells in spleens and dLNs. The data are presented as the mean ± SD. (*P < 0.05, **P < 0.01, ****P < 0.0001).

Figure 3

Hydroxychloroquine (HCQ) suppresses the differentiation and proliferation of T cells. (A–C) The lymphocytes from draining lymph nodes (dLNs) of experimental autoimmune uveitis (EAU) mice were isolated and cultured in vitro with hIRBP_1-20 for 3 days. The percentages of viable lymphocytes (zombie-) were similar at concentrations of 0, 20, 40, and 60 μM of HCQ, whereas 80 μM HCQ treatment decreased the percentage. HCQ (40 and 60 µM) suppressed the differentiation of Th17 and Th1 cells significantly with statistical differences, and 20 µM HCQ suppressed Th17 differentiation similarly (N= 5). (D, E) The labeled T cells were sorted and activated with anti-CD3/CD28 beads for 4 days. HCQ (40 and 60 µM) inhibited the proliferation of CD4⁺T cells obviously with statistical difference (N=5). The data are presented as the mean ± SD (^nsP > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure 4

Hydroxychloroquine (HCQ) alleviates retinal vascular endothelial cells (RVECs) dysfunction induced by inflammation. (A) Differentially expressed genes (DEGs) were explored and are reported in a heatmap of mRNA abundance. Rows denote expression of RNAs according to their enrichment in WT and spontaneous experimental autoimmune uveitis (EAU). (B) The heatmap of the gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) analyses showed that compared with the WT group, uveitic RVECs presented with enhanced inflammatory response, cell adhesion molecules CAMS, and antigen processing and presentation. (C) Chord chart equally illustrates that the antigen presentation mediated by MHC-II molecules was significantly enhanced in RVECs of EAU mice. The inflammatory dysfunction of RVECs was induced by TNF-α stimulation. (D) The toxic effects of HCQ on the cell viability of RVECs were measured by CCK8 assay. RVECs were incubated with 10, 20, 40, 80 and 100 μM HCQ for 48 h, and 100 μM HCQ induced cytotoxicity on RVECs (N=3). (E–J) RT-qPCR results show that in the model of RVEC dysfunction, the expression of ICAM, VCAM, E-selectin, MCP-1, MMP-9, and CD74 mRNAs in the retina was markedly upregulated, whereas HCQ (10, 20, 40, and 80 µM) considerably and dose-dependently decreased the expression of these mRNAs (N=5). (K, L) Flow cytometry analysis shows that the T cells were activated and produced more IL-17 and IFN-γ after coculture with RVECs, especially with TNF-α–stimulated RVECs. Pretreatment of TNF-α–stimulated RVECs with HCQ resulted in lower cytokine production in T cells (N=3). The data are presented as the mean ± SD (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure 5

Hydroxychloroquine (HCQ) regulates TNF-α–stimulated endothelial dysfunction via the LOX-1/NF-κB axis. (A) In the in vivo study, RT-qPCR results showed that HCQ suppressed the LOX-1 and NF-κB mRNA expression in retinal tissue of experimental autoimmune uveitis (EAU) mice (N = 4). (B) In the in vitro studies, RT-qPCR revealed that HCQ (10, 20, 40, and 80 µM) considerably and dose-dependently decreased LOX-1 and NF-κB mRNA expression in the retinal vascular endothelial cell (RVEC) dysfunction model. Furthermore, we verified these results at the protein level (N = 5). (C, D) Flow cytometry analysis showed that the mean fluorescence intensity (MFI) of LOX-1 in TNF-α–stimulated RVECs was higher than that in the vehicle group. HCQ reduced its MFI at a concentration of 40 µM with a significant difference(N = 4). (E, F) Western blot analysis showed upregulated expression of total p65 (t-p65) and phosphorylated p65 (p-p65) in TNF-α–stimulated RVECs compared with those in the vehicle group. HCQ (40 µM) treatment suppressed their expression and lowered the p-p65/t-p65 ratio (N = 7). (G) Immunofluorescent staining showed colocalization of p-p65 (green) and the nuclei (blue) of RVECs. Elevated expression of p-p65 was observed in RVECs after TNF-α stimulation, whereas HCQ treatment reduced the expression of p-p65 in RVECs. The data are presented as the mean ± SD (*P < 0.05, **P < 0.01, ****P < 0.0001).