Introduction

Retinal degenerative diseases are a heterogeneous group of progressive neurologic disorders, including age-related macular degeneration (AMD) and retinitis pigmentosa (RP), which are worldwide leading causes of irreversible blindness. These diseases are characterized by progressive degeneration of the retinal pigment epithelium (RPE) and/or overlying photoreceptors [1, 2]. RPE is injured in response to oxidative stress and there is lack of effective anti-oxidative agents for RPE protection [3]. In clinical trials, the effect of anti-oxidative diet therapy is limited in most cases, and the trials on replacing dysfunctional/damaged RPE are still in progress [4,5,6].

IL-4, a pleiotropic cytokine, plays an essential role in the regulation of a variety of immune responses, including Th2 cell-mediated immunity, IgE class switching in B cells, and macrophage activation towards M2 phenotype [7,8,9,10]. Targeting IL-4 in Th2 cell-mediated immune response has led to the clinical application of IL-4 antagonist, dupilumab, in treatment of atopic dermatitis [11]. Meanwhile, the action of IL-4-induced M2 macrophage transition makes IL-4 a potential therapeutic target for macrophage-mediated diseases [12,13,14]. For instance, IL-4 improves myocardial repair and enhances function of damaged hearts through induction of M2-reparative macrophages [15]; Clinical trials targeting IL-4 has also been demonstrated safe for cancer treatment [16]. The IL-4 signaling is mediated through the IL-4R α-chain (IL-4Rα), which dimerizes either with the common γ-chain or with the IL-13Rα, constituting type I or II IL-4 receptors, respectively [17]. The IL-4Rα expression is not limited to the hematopoietic T cells, B cells and macrophages, but also in endothelial cells, fibroblasts, and RPE cells [18,19,20]. It would be valuable to know whether IL-4/IL-4Rα signaling is also involved in functions and related diseases of these cell types, thus to extend the repertoire of IL-4 clinical applications.

In this study, we investigated the role of IL-4 on RPE cells in two retinal neurodegeneration mouse models, including the sodium iodate (NaIO3)-induced mouse and the inherited rd10 mouse [21]. We observed excessive inflammation responses in the NaIO3-induced retinae with downregulation of IL-4 expression. The choroid complex of AMD patients and rd10 mice also displayed decreased IL-4 expression. Supplementing IL-4 was able to rescue RPE cells and protect against the degenerative retina in both NaIO3 induced retinae and rd10 mice. The salient protective property of IL-4 points to a novel therapeutic approach for treating retinal degenerative diseases. We further found the IL-4/IL-4Rα-Nrf2 axis being crucial for protecting RPE cells against inflammation and oxidative stress.

Results

Aberrant inflammation and reduction of IL-4 expression in the degenerative retinae

Inflammation is a hallmark of retinal degenerative diseases [22]. To obtain the gene expression landscape of inflammation in retinal degeneration, thereby identify potential mediators, we aimed to take an unbiased experimental approach of deep sequencing of the RPE/choroid/sclera complex, the main pathological lesions in the NaIO3-induced retinal degeneration model. NaIO3, a stable oxidizing agent, can trigger the accumulation of reactive oxygen species (ROS) in RPE cells by converting glycine to potentially toxic glyoxylate, resulting in rapid ablation of RPE and consequent photoreceptor loss, resembling the degeneration progression in the dry AMD [23, 24]. The Gene Ontology analysis of the entire set of upregulated genes in NaIO3 group revealed the significant enrichment for terms related to regulation of cytokine production (Fig. 1A), immune receptor activity, and cytokine activity (Fig. 1B) among the top ones.

Fig. 1: NaIO3-induced degenerative retinae underwent inflammatory responses and displayed decreased IL-4 expression.
figure 1

A, B GO enrichment of the upregulated DEGseq in NaIO3-induced degenerative retinae showed the enriched differentially expressed mRNAs in biological processes (A) and molecular function (B). C qPCR analysis showed decreased mRNA levels of IL-4 in NaIO3-induced degenerative retina, while the pro-inflammatory cytokines and chemokines/chemokine receptors, including Ifn-γ, Il-6, Tnf-α, Cxcl1, Cxcl2, Cxcl10, Cxcr1 and Cxcr2 were significantly increased. The anti-inflammatory Il-10 had no significant difference between the NaIO3-induced retina and normal control, n = 3. D Western blotting analysis showed marked decreased IL-4 expression in the NaIO3-retinae in comparison to controls, n = 3. E Lack of IL-4 expression (red) and broken ZO-1-positive RPE barrier (green) were observed in the NaIO3-induced RPE flat-mount sheet by immunohistochemistry analysis. Scale bar, 50 µm. n = 3. Data are shown as mean ± SEM, **P < 0.01, ***P < 0.001, n.s: not significant, unpaired Student’s t-test in C and D.

Full size image

Next, we performed the qPCR to validate the expressions of common cytokines. Interestingly, among the eight cytokines or chemokines tested, IL-4 was the only one significantly reduced in the NaIO3-induced RPE/choroid/sclera complex while the others were either drastically elevated, such as Ifn-γ, Il-6, Tnf-α, Cxcl1, Cxcl2, and Cxcl10, or unchanged (Il-10) (Fig. 1C). The expressions of chemokines receptors Cxcr1 and Cxcr2 were also increased (Fig. 1C). The lower expression of IL-4 in the NaIO3-treated group of the RPE/choroid/sclera was further confirmed at protein level by Western blotting (Fig. 1D). Immunohistochemistry analysis of flat-mount RPE sheet further validated very low level of IL-4 expression in the NaIO3-induced RPE monolayer (Fig. 1E).

In addition, we analyzed a published single-cell transcriptomics data of human RPE and choroid from the health and macular degeneration patients [25, 26] (GSE135922 from NCBI GEO database). The cells were clustered into RPE, endothelial cells, pericytes, fibroblasts, melanocytes, schwann cells, and immune cells (Fig. S1A, B). The violin plot data showed that IL-4 was mainly enriched in immune cells and was particular decreased in the sample from macular degeneration patients (Fig. S1C, D). We further interrogated those immune cells into different types, including microglia/macrophage, NK/T cells, B cells, mast cells, and others with respective makers according to previous studies [25, 27, 28]. Interestingly, IL-4 was mainly expressed in microglia/macrophage (Fig. S1E), which are consistent with the branchy morphology of the microglia in Fig. 1E. The RPE-sclera complex from rd10 mice also displayed decreased IL-4 expression in comparison to the same-age wild type mice (Fig. S3A). These data collectively suggested that the reduction of IL-4 expression involved in the process of retinal degeneration.

Exogenous IL-4 preserved neuroretinal structure and function in the degenerative retina

Given IL-4 as the critical decreased cytokine in the degenerative retina in addition to its anti-inflammation function, we wondered whether IL-4 supplement might be protective for the retinal degeneration. To test this speculation, IL-4 was injected intravitreally to the NaIO3-induced mice and rd10 mice. As shown in Fig. 2A, pale-colored retinae with structural damages such as white dot-like or linear lesions were observed by funduscopy in NaIO3 injected retinae contrasting to that of PBS-injected reddish-orange controls; whilst IL-4 treatment significantly alleviated these pathological changes. When performed OCT scan, RPE band in the NaIO3-treated eye exhibited aberrant hyperreflective discrete foci (Fig. 2B, red arrows), indicating the RPE lesions. In contrast, the IL-4-treated RPE exhibited as a smooth, continuous, and uniform reflectiveness RPE band, more close to that of the PBS-treated controls (Fig. 2B). In addition, the reduced thickness of ONL in NaIO3-induced retina was recovered after IL-4 treatment (Fig. 2B). Consistently, many crumby-structure patches (yellow arrows), possibly the pathologic aggregation of melanin in the RPE cells, were observed in the NaIO3-retina in H&E staining (Fig. 2C). Besides, the discontinue degenerated RPE layer was usually associated with retinal folds (yellow stars) in the NaIO3-retina, which were severer in the mid-periphery location than those in the center or periphery (Fig. 2C). Notably, IL-4 treatment largely preserved the well-organized RPE layer and inhibited the associated photoreceptor degeneration, with less melanin rich aggregation and ONL folds (Fig. 2C). On flat-mount RPE sheet, NaIO3 disrupted the hexagonal-packaging of RPE cells and barrier as indicated by a tight junction marker, ZO-1 staining, particular in the center and mid-periphery. IL-4 treatment again remarkably rescued these pathogenic phenotypes (Fig. 2D). Finally, ERG analysis was used to determine neuroretinal function after IL-4 treatment. Remarkably, the typically reduced amplitudes of a-wave and b-wave in NaIO3-induced retina were restored by IL-4 treatment (Fig. 2E). A spontaneous retinal degeneration rd10 mouse model was also treated by IL-4 intravitreal injection. Similarly, IL-4 treatment improved retinal neurofunction of rd10 mice with increased amplitudes in ERG (Figure S2A) and restored vision in optomotor task (Fig. S2B). In H&E staining, the IL-4 treatment group displayed increased ONL thickness (Fig. S2C). Taken together, the IL-4 treatment exhibits a neuroprotective effect in the retinal degenerative process.

Fig. 2: IL-4 protected neuroretinal structure and function in the NaIO3-induced retinal degenerative mice.
figure 2

A The funduscopy of NaIO3-induced degenerative retina was characterized by pale appearance and some dot-like or linear lesions showing structural damage (blue arrows), which were alleviated after IL-4 treatment. n = 6. B Representative images of the OCT scanning. The NaIO3 model demonstrated obvious discontinuity of RPE layer with high-reflective deposits (red arrows), while IL-4-treated RPE exhibited as a smooth, continuous, and uniform reflectiveness layer. The ONL thickness (distance between yellow lines) was calculated and compared. n = 6. C H&E staining showed crumby-structure patches and melanin rich aggregation in the RPE layer of the NaIO3-retina, particularly in the mid-peripheral locations. After IL-4 treatment, the RPE layer was preserved and well-organized with less melanin rich aggregation and the retinal folds were relieved. Scale bar: 50 μm and 25 μm. D Representative images of RPE flat-mount in the center, mid-periphery and periphery areas sheet showing RPE integrity by ZO-1 staining. NaIO3 disrupted the well-organized hexagonal-shaped RPE cells, whereas IL-4 treatment could preserve it with more organized barrier structures specifically in the mid-periphery and periphery areas. Statistical analysis showed the amount of ZO-1+ cells. n = 6, Scale bar: 20 μm. E ERG examinations showed a- and b-waves were almost extinct in NaIO3-retina. However, IL-4 treatment protected the neuroretinal functions with increased amplitudes of both a- and b-waves. n = 6. Data are shown as mean ± SEM, ***P < 0.001, n.s: not significant, one-way ANOVA test.

Full size image

Transcriptome analysis revealed the inhibiting effects of IL-4 on inflammatory and oxidative stress pathways

Next, we examined the transcription profile of IL-4-treated degenerative retina in order to find out the underlying mechanism. The RNA was extracted from the NaIO3-induced RPE/choroid/sclera complex with or without IL-4 treatment followed by RNA-seq assay. In total, 7663 differentially expressed genes were identified when the NaIO3 + IL-4 treatment and NaIO3 control group were compared. Among these genes, 3567 (46.5%) genes were upregulated whereas 4096 (53.45%) were down-regulated upon the IL-4 treatment. In particular, a list of pro-inflammatory genes such as Cxcr2, Cxcl1, Cxcl10, Stat1, Ccl8 were down-regulated, while the anti-inflammatory genes, including Il1f5, Klf14, Drd5, Ifnk were upregulated after IL-4 treatment (Fig. 3A). In addition, the GSEA analysis revealed that the genes involved in positive regulation of axon extension and neurogenesis including Lrp1, Ntrk3, Ntrk2, Ntn1, Wnt2, Serpine2, and Notch1 were enriched in the IL-4-treated degenerative retina (Fig. 3B), indicating the possible neuroprotective effects of IL-4. Particularly, in comparison to NaIO3 control group, IL-4 treatment suppressed the expressions of pro-inflammatory or pro-oxidative stress enriched gene sets, including antigen processing and presentation, immune response, chemotaxis, chemokine activity, IL-23-mediated signaling, cellular response to IFN-γ or IFN-β, NAD ADP-Ribosyltransferase activity and so on (Fig. 3B). These results further suggested that IL-4 supplements could suppress the inflammation or oxidative stress in the degenerative retina and orchestrate an immune niche for neuroprotection.

Fig. 3: The transcriptional changes of NaIO3-induced RPE sheet.
figure 3

A Heatmap showing differentially expressed inflammation-related genes in NaIO3 mice with and without IL-4 treatment. The pro-inflammatory genes such as Cxcr2, Cxcl1, Cxcl10, Stat1, Ccl8, etc were down-regulated in the IL-4-treated group, while the anti-inflammatory genes, including Il1f5, Klf14, Drd5, Ifnk, etc were upregulated. B GSEA analysis revealed significantly enriched gene sets such as axon extension and neurogenesis in the IL-4 group, and the enriched gene sets, including the inflammation and oxidative stress-related gene sets in the NaIO3-RPE without IL-4. n = 3, FDR value <0.25. C The enrichment analysis of IL-4Rα and anti-oxidative genes such as Nrf2, Hmox1, and Arg1 in different groups were presented by circos plot. The thickness and color of the ribbons correlated to the genes expression.

Full size image

IL-4 drove the expression of IL-4Rα and Nrf2

Further, the circos plot analysis of the RNA-seq data revealed a positive correlation between the expression of IL-4Rα and antioxidant genes such as Nrf2, Hmox1, and Arg1 (Fig. 3C). IL-4 usually binds to the high-affinity subunit of IL-4Rα to induce immune response [19]. Here we found the IL-4Rα was highly expressed in the IL-4-treated RPE complex both in mRNA and protein level (Fig. 4A, B and Fig. S3A). IL-4-flat-mount RPE sheet revealed that the IL-4Rα distributed alongside the ZO-1, indicating the possible involvement of IL-4/IL-4Rα axis in retinal neuroprotection (Fig. 4C and Fig. S3B). In addition, oxidative stress contributes to a range of neurodegenerative diseases [29,30,31]. Nrf2, a transcription factor, which is essential for protection against oxidative stresses, has been known to attenuate degeneration and inflammation [32]. Here we found that whereas the total and phosphorylated expression levels of Nrf2 were reduced in the NaIO3-induced RPE/choroid/sclera complex when compared to the ones in the PBS control, the addition of IL-4 increased their levels closed to those in the control (Fig. 4D). In the rd10 mice, phosphorylated Nrf2 was decreased in the PBS control group, whereas increased Nrf2 phosphorylation level was observed after IL-4 treatment (Fig. S3A, B). Consistently, in the in vitro experiments, H2O2, a usually used oxidative stress inducer, decreased the levels of p-Nrf2 and IL-4Rα in the cultured RPE cells, which were restored after IL-4 treatment (Fig. 4E). These results indicate that Nrf2 is probably implicated in the reparative and protective role of IL-4.

Fig. 4: IL-4R and Nrf2 activation in response to IL-4 in vivo and in vitro.
figure 4

A, B qPCR (A) and western blotting analysis (B) showed elevated IL-4Ra expression in RPE complex from the NaIO3-induced degenerative mice after IL-4 treatment. n = 3. C Representative images of RPE flat-mount sheet revealed that increased IL-4Ra distributed alongside the RPE tight junction marker ZO-1. Scale bar: 20 μm. The hexagonal ZO-1+ cells per field and the percentage of IL-4Rα + ZO-1+ cells and total ZO-1+ cells were calculated for statistics. D Western blotting analysis showed reduction of total and phosphorylated Nrf2 expression in NaIO3 model, while IL-4 treatment induced their elevations. n = 3. E Representative images of cultured RPE cells under oxidative stress insults showing lack of IL-4Rα and phosphorylated Nrf2 expression, which were induced after IL-4 stimulation. Data are shown as mean ± SEM, *P < 0.05, **P < 0.01, ***P < 0.001, n.s: no significance, one-way ANOVA test.

Full size image

Nrf2 deficiency abrogated the neuroprotective effect of IL-4 in NaIO3-induced retinal degeneration mice

To further validate the involvement of Nrf2, we performed the NaIO3-induced retinal degeneration in the Nrf2 knockout (Nrf2-/-) mice. The 8 to 12-week-old Nrf2-/--mice did not show signs of degeneration as previously reported [33], and after NaIO3 induction, they displayed similar hyperreflective foci (red arrows) in the RPE layer as the WT-NaIO3, indicating the Nrf2 depletion alone at the young age would not worsen the pathological process of NaIO3-induced retinal degeneration examined by OCT (Fig. 5A). On the other hand, while treatment of IL-4 resulted in rescue of integrity of RPE layer and elevated thickness of ONL in IL-4-WT-NaIO3-mice, no recovery of hyperreflective foci and ONL thickness in IL-4-Nrf2-/--NaIO3-mice were observed, indicating Nrf2 deficiency abolished the protective effects of IL-4 (Fig. 5A). Although OCT has provided exquisite methods to identify and evaluate multiple layers of ocular structures, the OCT signals could be easily affected by the image quality and structural reflection, leading to biased results. Therefore, we further performed histological analysis by H&E staining, the images revealed clearly that the melanin rich aggregation in the RPE in both WT and Nrf2-/--NaIO3-mice with slightly more severe pathology in the Nrf2-/- mice (Fig. 5B). Consistent with the results from the OCT examination, while IL-4 treatment alleviated the RPE injuries in WT- NaIO3-mice, it failed to do so in the IL-4-Nrf2-/--NaIO3-mice (Fig. 5B). To investigate the functional impact, ERG was also utilized in all individual groups. The WT-NaIO3 and Nrf2-/--NaIO3- mice presented nearly similar extinct a- and b- waves. Once again, while IL-4-treated WT-NaIO3- mice had active waves with elevated amplitudes of a-wave and b-wave, the Nrf2-/--NaIO3- mice treated with IL-4 displayed the least amplitudes of a-wave and b-wave (Fig. 5C), indicating IL-4 could not rescue the dysfunction of RPE and photoreceptors without Nrf2. Thus, these data suggested that IL-4 exerted neuroprotective role in a Nrf2-dependent manner.

Fig. 5: Nrf2 deficiency diminished the protective function of IL-4.
figure 5

A In OCT examination, NaIO3 model in Nrf2-/-mice presented discontinuity of RPE layer with highly reflective granules (red arrows). Importantly, IL-4 treatment in these mice showed no significant improvement. In addition, the calculated ONL thickness (distance between yellow lines) was reduced in NaIO3 model of both WT and Nrf2-/- mice in comparison to those in PBS controls, respectively. IL-4 treatment could partially restore the ONL thickness in NaIO3 model of WT mice but not Nrf2-/- mice. n = 6. B H&E staining revealed melanin rich aggregation (yellow arrows) in the NaIO3-RPE from both WT and Nrf2-/- mice. IL-4 treatment rescued RPE injuries in WT-NaIO3-mice with much less melanin rich aggregation. However, the IL-4-treated Nrf2-/-NaIO3 mice still presented with crumby structure with melanin rich aggregation. Scale bar: 20 μm. n = 3. C ERG showed increased amplitudes of a- and b- waves after IL-4 treatment in WT-NaIO3 retina but not Nrf2-/- NaIO3 mice. The Nrf2-/- NaIO3 mice with or without IL-4 treatment presented almost extinct waves. n = 6. Data are shown as mean ± SEM. **P < 0.01, n.s: no significance, one-way ANOVA test.

Full size image

IL-4 Induced Anti-oxidative Stress and Anti-inflammatory Genes Expression in Degenerative RPE

IL-4 has been reported to induce the expression of Arg1, which has also been shown to mediate anti-oxidative stress or anti-inflammatory effects [34]. The Nos2 and Arg1 catalysis usually compete against each other to activate a common metabolic pathway and exert the divergent effects [35]. Thus, we examined the expressions of both Nos2 and Arg1 in vivo upon the IL-4 treatment. As shown in the retinal cryosection images of Fig. 6A, whereas NaIO3 promoted the expression of Nos2 and down-regulated the Arg1 expression in RPE layer in comparison to PBS control, IL-4 treatment reversed the expression patterns of both genes. The impacts of IL-4 on the expression of Nos2 and Arg1 was further verified via the immunostaining of the RPE flat-mount sheets showing that IL-4 treatment suppressed the expression of Nos2, as well as promoted the Arg1 staining in RPE (Fig. 6B). Consistently, the qPCR analysis also revealed the increased Arg1 and decreased Nos2 at the mRNA level (Fig. 6C). Western blot analysis showed that IL-4 treatment also increased Arg1 protein expression while decreased Nos2 expression in the degenerative rd10 retina (Fig. S4A, B). In addition, the expression of a panel of anti-oxidative stress-related genes, including Hmox1, Nqo1, Foxm1, and Gpx1 were highly increased after IL-4 treatment (Fig. 6C). The luminex assay found the protein expressions of inflammatory genes such as G-Csf, Cxcl1, and Ccl2 were reduced in the IL-4 group (Fig. 6D). These data suggested that IL-4 may regulate the inflammatory and oxidative signaling.

Fig. 6: IL-4 induced antioxidant and anti-inflammatory genes expression in vivo.
figure 6

A Immunoflorence of cryosection showed high expression of Nos2 in response to NaIO3 in the ZO-1+ RPE monolayer, whereas IL-4 treatment decreased Nos2 and increased Arg1 expression. Scale bar: 20 μm. B In the RPE flat-mounts, NaIO3 disrupted the well-organized hexagonal-shaped RPE cell and induced Nos2 expression elevation, whereas IL-4 treatment reversed it with more organized barrier structures of RPE cells. IL-4 could also suppress the Nos2 expression and promote Arg1 expression within RPE cells. Scale bar: 20 μm. C qPCR data showed the reduction of Nos2 and increase of Arg1 in the NaIO3 + IL-4 group. In addition, IL-4 treatment induced the expressions of anti-oxidative stress-related genes, including Hmox1, Nqo1, Foxm1, and Gpx1 in the NaIO3-retina-RPE complex. n = 3. D Luminex results showed elevated expression of pro-inflammatory cytokines in the RPE complex of NaIO3-induced model, including G-Csf, Cxcl1, and Ccl2, which decreased after IL-4 treatment. n = 3. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n.s: no significance, one-way ANOVA test.

Full size image

Discussion

Neurodegenerative diseases are characterized by chronic degenerative pathological changes due to oxidative stress and inflammation [3, 22]. These diseases have very limited therapeutic options, and it is highly needed to explore more effective treatment [36]. In this study, IL-4, a pleiotropic cytokine in immune response, was reported to preserve the structure and neuroretinal function in the NaIO3-induced retinal degeneration mouse model and the rd10 mouse, underlining the function and therapeutic potential of IL-4 beyond its classical regulation of Th2-mediated immune responses. RNA-sequencing analysis showed that IL-4 treatment modulated the oxidative stress-related pathways and attenuated the inflammatory injuries. This notion was further confirmed by revealing the requirement of Nrf2 in IL-4-mediated protection using knockout mice. These observations suggested that IL-4 may serve as a therapeutic candidate to treat retinal degeneration via activating Nrf2 signaling.

IL-4, a signature cytokine of Th2-immune response, plays a critical role in the regulation of immune responses, as well as a nonimmune function [37]. In the context of the retina, which consists of multiple well-arranged layers with neurons, glial cells, and RPE cells, the functions of IL-4 are quite complex. IL-4 has been shown to enhance the recruitment of CCR2+ bone marrow cells into retinochoroid lesion and promote the choroid neovascularization [38], whereas others reported that IL-4 drove macrophage sFlt-1 production, and the sFlt-1+Arg1+ macrophages attenuated laser-induced choroidal neovascularization [39]. In this study, we found IL-4 induced IL-4Rα activation and protected the RPE from either inherited or chemically induced degeneration through Nrf2 signaling, expanding the functions of IL-4. Thus, at the context of retinal degeneration, IL-4 exhibited a new neuroprotective feature.

It is noteworthy of the distribution pattern of IL-4 in normal retina indicating that IL-4 is likely derived from microglia based on the shape of IL-4-positive cells with filopodia-like protrusions, closely resembling those of branchy M2 phenotypes [27, 28]. This is also consistent with the human scRNA-seq data (Fig. S1) and previous studies reporting that lymphocytes and myeloid cell were the main cellular sources of IL-4 [40,41,42]. Interestingly, the expression of IL-4 was abolished by the NaIO3 stimulation, suggesting that NaIO3 targets on not only RPE cells as previously considered, but also other types as well. In this study, even though we focused on the effects of IL-4 on RPE in vivo and in vitro, the impacts of other types of cells, particularly the microglia, which also expressed IL-4R, would probably respond to this immune cytokine as well [20, 43].

Oxidative stress and inflammation form a vicious circle to promote the neurodegeneration pathogenesis [44]. RPE cells, under oxidative stress initiated by various factor such as aging, play a critical role in the chronic inflammation network leading to the degenerative retina [45]. Our RNA-seq data revealed that IL-4 treatment could not only suppress the expressions of some pro-inflammatory enriched gene sets, but also inhibit the genes of oxidative stress pathways. Specifically, RPE cells were shown to highly express Nos2, a gene to promote reactive nitrogen specie production and mediate inflammation in tissues. In contrast, Arg1, a known competing enzyme, which induces expressions of anti-oxidative and anti-inflammatory genes [46, 47], displayed increased expression upon the IL-4 treatment. The strong inhibitory effects of IL-4 on inflammation and oxidative stress in NaIO3-induced RPE might be the underlying mechanism responsible for the potent protection. Consistently, we found Nrf2, a master regulator of cellular defense against oxidative stress [48], was required for IL-4-induced anti-inflammation and anti-oxidative effects. Although Nrf2 has been linked to neuroprotection for long time [49], it still lacks of available pharmacological modulators due to the elusive mechanism of current compounds for inhibiting Nrf2 [50, 51]. Our findings suggest IL-4 might be a new modulator for Nrf2 to treat neurodegenerative diseases.

In conclusion, our study showed that IL-4 was decreased in retinal degeneration, and the application of exogenous IL-4 protected the RPE cells and improved neurofunction through anti-oxidative and anti-inflammatory pathways. The protection of IL-4 was mediated via the IL-4/IL-4Rα-Nrf2 axis. These findings make IL-4 a new potential therapeutic approach for neurodegenerative diseases.

Materials and methods

Animals

All the experiments were carried out in accordance with the Association for Research in Vision and Ophthalmology Statement for Use of Animals in Ophthalmic and Vision Research and with the approved guidelines of Animal Care and Use Committee of Zhongshan Ophthalmic Center (Ethics ID: 2018-105). Nrf2-/- mice were bought from the Jackson Laboratory (Strain #:017009), and the mice with rd8 mutation in the Crb1 gene detected by PCR were excluded for this study. Eight- to 12-week-old C57BL/6J and Nrf2-/- mice were used in this study. All mice were age and sex-matched and then randomized into the different groups. Retinal fundus was examined before NaIO3 model establishment. To establish the retinal degeneration model, the mice were injected intravenously with 20 mg/kg of NaIO3 (Sigma-Aldrich Corp., St. Louis, MO, USA), which was dissolved in phosphate-buffered saline (PBS). The mice receiving the same volume of PBS worked as controls. When receiving treatment, the NaIO3 mice were injected intravitreally either with IL-4 (2 μg/μl in 1 μl) or 0.2% BSA as control. Seven days later, the mice were sacrificed by carbon dioxide euthanasia and the eyes were collected for further investigation. The rd10 mouse model, homozygous for the Pde6brd10 on a C57BL/6J background (Strain #:004297) were bought from the Jackson Laboratory, and IL-4 (2 μg/μl in 1 μl) were injected intravitreally every two days for 3 times beginning at P16 for rd10 mice. The same-age C57BL/6J mice received PBS treatment as comparison and they are sacrificed at P25 with eyes collected for further investigations.

Electroretinogram (ERG) recordings

After dark adaptation overnight, the mice were anesthetized and the pupils were dilated with 0.5% tropicamide. ERG was recorded with gold plated wire loop electrodes contacting the corneal surface as the active electrode. Stainless steel needle electrodes were inserted into the tail serving as ground leads. The mice were exposed to full-field scotopic flashes with intensity of 0.03, 0.1, 0.3,1.0, 3.0, and 10.0 log cd.s/m2 [52,53,54]. Amplitudes of a- and b-wave were measured (Celeris-Diagnosys system) as in previous reports [55].

Visual acuity

The visual acuity of mice was examined as previously described [56]. In brief, the test began when the mouse on top of the platform no longer actively moving around, with contrast set at 100% and a grating of low spatial frequency (0.042 cycles per degree), and the tracking behavior of the mice was observed and recorded. Then the same stimuli were rotated in the anticlockwise direction (thus effectively testing the right eye). The acuity threshold was determined by a staircase method. When the mouse ceased to respond to a particular spatial frequency, a lower frequency grating was presented; when the mouse responded to a frequency, the frequency was increased, until reaching the closest none-responding spatial frequency. The acuity threshold was set as the highest spatial frequency to which the animal responded.

Optical coherence tomography (OCT)

The images were obtained with SPECTRALIS-OCT (Heidelberg, Germany) with a mouse objective lens. After focusing on the retina with the optic cup placed in the scanning center, OCT scans were acquired in automatic real-time (ART) mode to improve the signal to noise ratio. The quality index of each scan was verified to be more than 28 to show the retinal structures clearly. Twelve scanning-images were recorded in 7.8 × 7.8 mm scanning frame for each eye at 30° field-of-view in a circular order. The images were acquired with 60–100 frames for each scan. To measure the thickness of outer nuclear layer (ONL), quantitative analysis was performed by the Image J software (US National Institutes of Health) as previously described [57]. In brief, the ONL was defined as the distance between the outer plexiform layer to the outer limiting membrane, and was assessed by the averaged thicknesses from 10 locations on each image, measured at 700 μm intervals by 5 locations on each side of the optic disc. The measurements of ONL thickness were performed by experienced investigators in a masked fashion [58,59,60].

Retinal fundus imaging and grading

After anesthesia and pupil dilation, the retinal fundus was captured with the Micron IV fundus camera with mouse objective and 50-degree field of view (1.8-mm diameter) centering on the optic cup (Phoenix Research Labs, Pleasanton, CA, USA). Retinal infiltrates and folds were counted in the resulting digital images. The grading system were based on the retinal structural damage scores from 0 to 4. Score 1 refers to 1–4 dot-like lesion or 1 linear lesion, score 2 refers to 5–10 dot-like lesions or 2–3 linear lesions, score 3 refers to >10 dot-like lesions or >3 linear lesions, score 4 refers to confluent linear lesions [23, 61].

Hematoxylin & Eosin staining and thickness analysis

Eyes were fixed with 4% formalin overnight, embedded in paraffin and cut into 3 μm vertical slices. Sections were washed and treated with hematoxylin buffer followed by eosin solution as previously described [62]. The retinal morphology was observed and evaluated under a microscope (Leica DM4000, Germany). The ONL thickness were measured and the number of melanin rich aggregation in RPE layer was counted in a masked fashion manually and analyzed in 6 fields/eye from 3 eyes. Images were processed and analyzed with Image J software.

RPE-sclera Isolation, RNA extraction, library construction, and sequencing

Briefly, the anterior segment of the eyeball was cutoff with the lens, vitreous, and the neuroretina carefully removed. The RPE-sclera were collected and put into the Trizol (Thermo Fisher Scientific) for RNA extraction. Total RNA was purified, and the RNA integrity was assessed using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA, USA). Libraries were constructed using the NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB, USA) following manufacturer’s recommendations and index codes were added to attribute sequences to each sample. After PCR products were purified (AMPure XP system) and library quality was assessed on the Agilent Bioanalyzer 2100 system, libraries were pooled and sequenced on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumia).

RNA-sequencing data processing

Raw data (raw reads) of fastq format were firstly processed through in-house perl scripts with clean data (clean reads) obtained and low-quality reads removed. Retained reads were then aligned to mus musculus genome using STAR (v 2.7.3a) in the pairend mode with default parameters. Reads number mapped to each gene was counted using Hisat2 v2.0.5. The output matrix from Hisat2 was input into the DESeq2 (v1.26.0) for identification of differentially expressed genes between diabetes group and control group. We determined differentially expressed genes with the criteria of absolute fold change >2 and false discovery rate (FDR) adjusted P-value < 0.05. In addition, log-fold change of all genes between two groups were used to conduct Gene Ontology (GO) enrichment analysis. GO terms with corrected P-value less than 0.05 were considered significantly enriched by differential expressed genes. GO and circos plots were generated by ggplot2 (v 3.2.1) and clusterProfiler. Gene-set enrichment analysis (GSEA) was conducted using the GSEA software (http://www.gsea-msigdb.org/gsea/index.jsp). The NaIO3 induced RPE was class A, and IL-4 treated NaIO3 RPE was class B. Gene-set permutations were performed 1000 times for each analysis. The cutoff values were predefined as FDR < 0.25. Our sequencing data are available at Gene Expression Omnibus under the accession number GSE193808.

Analysis of scRNA-seq data

The scRNA-seq datasets were retrieved from NCBI’s Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE135922, No. GSE135922), which included sequenced single cells from human RPE/choroid of both healthy controls and AMD patients (Donors 1–3) as previously described [25, 26]. After quantification and normalization for each sample, unsupervised clustering analysis was further performed to identify the highly variable genes. Then principal component analysis was performed on the highly variable genes. Non-linear dimensional reduction (tSNE) was used to visualize clustering results based on the top PCs. Graph-based clustering was performed to cluster cells by using the FindNeighbors and FindClusters function in Seurat (clustering resolution = 1.0). We screened the top-ranking marker gene lists for each cluster to perform cluster annotation to specific cell types. Transcriptomic differences of IL-4 between different cell types were statistically compared and analyzed using visualization tools of DotPlot. The expression level of IL-4 between normal and AMD patient was visualized as VlnPlot.

ARPE-19 cell line culture

The human ARPE-19 cell line was purchased from ScienCell, China. The RPE cells were treated with IL-4 (100 ng/ml, dissolved in PBS, 96-214-14-50, PeproTech) or PBS for 2 h and then stimulated with hydrogenperoxide (H2O2, 300 μM). Twenty-four hours later, the cells were harvested for further analysis.

Quantitative PCR analysis and western blotting

The qPCR was performed as previously described [63]. The total volume of 20 μl contained 2 μl of cDNA, 10 μl of 2×SYBR Premix Ex Taq, 7 μl ddH2O and 10 μmol/l of the primer pairs. The sequences of the used primers were listed in the supplement Table 1 and qPCR was performed using Roche LightCycler480 II. The western blotting was performed as previously described [64]. Primary antibodies included anti-IL-4 antibody (Santa Cruz, sc-53084), anti-IL-4Rα antibody (Santa Cruz, sc-28361), anti-Nrf2 (Abcam, ab137550), anti-p-Nrf2 antibody (Abcam, ab76026), anti-Nos2 (Santa Cruz, sc-7271), anti-Arg1 (Santa Cruz, sc-18355), and anti-β-actin (Cell Signaling Technology, 4970S).

Immunofluorescence

For cryosections, eyes were removed carefully and embedded in optimal cutting temperature compound (Sakura Fine Technical). The frozen samples were then sliced transversely (8 μm) with a cryostat (Leica CM1950) at –20 °C. Only the cross-sections throughout optic nerve were used for staining and analysis. For retinal-RPE flat-mount, the eyes were fixed in 4% PFA for 60 min at room temperature and the RPE-sclera complex were dissected out. The cryosection, RPE flat-mount sheet samples and ARPE cells were blocked with 0.5% Triton-X100/5% BSA and incubated with primary antibodies overnight at 4 °C. The primary antibodies included anti-ZO-1 antibody (Santa Cruz, sc-33725), anti-Arg1 antibody (Santa Cruz, sc-18355), anti-Nos2 antibody (Santa Cruz, sc-7271), anti-IL-4Ra antibody (Santa Cruz, sc-28361), anti-IL-4 antibody (Santa Cruz, sc-53084), and anti-p-Nrf2 antibody (Abcam, ab76026). Then they were incubated with secondary antibodies for 1 h and counterstained with DAPI (Invitrogen, D1306) for 5 min before mounted. Negative control was done by omission of the primary antibody. The images were obtained using a confocal microscope (LSM880, Carl Zeiss).

Multiplex immunoassay

Retinal-RPE-mixture protein were collected and the concentration were measured by BCA protein assay and diluted to the concentration of 1 mg/ml. Cytokine/chemokine levels were analyzed using MilliPlex map Mouse Angiogenesis/Growth Factor Magnetic Bead Panel (15 plex) from Millipore according to manufacturer’s instructions [65]. All data was collected on the Luminex 200 at Laizee Biotech CO., LTD (Beijing, China).

Statistics

All experiments including immunostaining, qPCR, and WB were repeated at least three times with representative results shown in the figures. Data were presented as mean ± standard error of mean (SEM) with the number of samples provided in the figure legends, and were analyzed statistically using an unpaired Student’s t-test (two groups) or one-way ANOVA (>2 groups) with a Turkey’s post hoc comparisons test. P-values < 0.05 were considered statistically significant.