Microglia-Derived IL-6 Promotes Müller Glia Reprogramming and Proliferation in Zebrafish Retina Regeneration

Abstract

Purpose: Inflammation activates the Jak1-Stat3 signaling pathway in zebrafish Müller glia (MG), leading to their status transition and proliferation following retinal injury. However, the source of Stat3-activating molecules remains unclear. This study aims to explore the expression and function of a Stat3-activating cytokine IL-6 in zebrafish retina regeneration.

Methods: Mechanical retinal injury was induced in adult zebrafish by a needle-poke lesion. Single-cell RNA sequencing (scRNAseq) and PCR were used to determine gene expression. Microglia ablation was performed by using the mpeg1:nsfb-mcherry transgenic zebrafish. Morpholino oligonucleotides, a recombinant zebrafish IL-6 protein and drugs, were used to manipulate IL-6 or Stat3 signaling in the retina. The 5-Ethynyl-2′-deoxyuridine (EdU) labeling was used to evaluate MG proliferation and the formation of MG-derived progenitor cells (MGPCs). Neuronal regeneration in the retina was analyzed by lineage tracing and immunostaining.

Results: The scRNAseq reveals that IL-6 is mainly expressed by a subset of pro-inflammatory microglia in the injured retina. Loss- and gain-of-function experiments demonstrate that IL-6 signaling promotes MG proliferation and the formation of MGPCs following retinal injury. Additionally, IL-6 facilitates MG status transition by modulating Jak1-Stat3 signaling and the expression of regeneration-associated genes. Interestingly, IL-6 may also regulate MGPC formation via phase-dependent pro-inflammatory and anti-inflammatory mechanisms. Finally, IL-6 promotes the early differentiation of MGPCs and contributes to the regeneration of retinal neurons in the injured retina.

Conclusions: Our study unveils the critical role of microglia-derived IL-6 in zebrafish retina regeneration, with potential implications for mammalian MG reprogramming.

Unlike mammals, lower vertebrates, such as teleost fish, possess a remarkable capacity for regenerating their damaged retinas.^1,2 This regenerative process in the fish retina hinges on a type of macroglia cells called Müller glia (MG). Following retinal injury, MG in the zebrafish retina express genes associated with regeneration and undergo a reprogramming process, transitioning from a state of quiescence to activation and proliferation.^3,4 Activated MG undergo asymmetric cell division to generate multipotent MG-derived progenitor cells (MGPCs), which further proliferate and migrate to different retinal layers, eventually differentiating into all types of retinal neurons, thereby facilitating the repair of the damaged retina.^5–7 

MG reprogramming and proliferation are regulated by many intrinsic and extrinsic factors.^8,9 In recent years, it has been established that inflammation and immune cells, such as microglia/macrophages and regulatory T cells, play crucial roles in MG proliferation and retina regeneration in lower vertebrates.^10–17 We have previously shown that microglia-mediated inflammation is both necessary and sufficient for activating the mTOR signaling in MG in the injured zebrafish retina.¹⁴ In a recent study, we further demonstrated that inflammation upregulates the expression of Jak1-Stat3 signaling components in MG, facilitating their status transition from quiescence to activation/proliferation in the injured zebrafish retina.⁴ Our findings, along with those of other research groups, underscore the pivotal role of inflammation and immune cells in retina regeneration. Nevertheless, the specific signaling molecules derived from immune cells that regulate MG reprogramming and proliferation remain elusive. 

Our recent single-cell RNA sequencing (scRNAseq) study revealed that immune cells in the injured zebrafish retina express many Stat3-activating cytokines, including interleukin-4 (il4), il6, il11b, il13, il21, and m17 (LIF), whereas MG expresses many of their receptors.⁴ Among the Stat3-activating cytokines expressed by microglia, interleukin-6 (IL-6) exhibited the highest expression level, indicating its potential significance as a regulator of Stat3 signaling and MG reprogramming. In classical IL-6 signaling, IL-6 binds to the membrane-bond IL-6 receptor (mIL-6r) and the signal-transducing subunit glycoprotein 130 (gp130/Il6st).¹⁸ Alternatively, in IL-6 trans-signaling, IL-6 can bind to the soluble form of IL-6r (sIL-6r) and transmit signals through gp130 on cells lacking IL-6r expression.^18,19 In this study, we investigate the role of IL-6 signaling in MG reprogramming/proliferation, MGPC formation, regulation of the inflammatory response, and regeneration of retinal neurons in the adult zebrafish retina. 

Methods

Animals and Retinal Injury

Adult wild type or Tg(gfap:GFP)²⁰ and Tg(mpeg1:nsfb-mcherry)¹⁴ transgenic zebrafish at the age of 6 to 10 months were used in the study. Zebrafish were treated in accordance with the Guidelines for Animal Use and Care at Nantong University. Fish were maintained in an automatic breeding system at 28°C on a 14:10 hour light/dark cycle. Methods for retinal needle-poke injury have been described in our previous study.¹⁴ Briefly, the fish were anesthetized in 0.02% Tricaine (Sigma-Aldrich) in system water. The right eye was gently pulled from the socket and stabbed 4 times (once in each quadrant) through the sclera with a sterile 30 G needle. During the injury, the needle was inserted up to the length of the bevel (approximately 0.5 mm) to assure the consistency. The uninjured left eye served as a negative control. 

scRNAseq Analysis of the Stab-Injured Zebrafish Retina

Detailed methods for the scRNAseq (10× Genomics) of stab-injured zebrafish retina were described in our recent study.⁴ Briefly, adult zebrafish retinal samples from the uninjured eyes, or those at 12 hours post injury (hpi), 1 day post injury (dpi), and 2 dpi were washed and digested with 1000 u/mL papain solution (Solarbio Life Science, Beijing, China) for 10 minutes at room temperature. The cells were further washed, filtered, and resuspended in cold 2% FBS (Thermo Fisher Scientific, Waltham, MA, USA). Single-cell solutions were examined for cell vitality, adjusted to approximately 1000 cells/µL, and added to the 10× Chromium Chip (10× Genomics) for further processing. The cDNA amplification and library construction followed standard protocols and libraries were sequenced by LC-Bio Technology (Hangzhou, China) on an Illumina NovaSeq 6000 system. 

The sequencing results were converted to FASTQ format and aligned to zebrafish GRCz11_98 genome using the CellRanger software. Low quality cells were filtered and cells were projected into 2D space using t-Distributed Stochastic Neighbor Embedding (t-SNE). Retinal cell types were identified based on known markers.⁴ Pseudotime analysis was performed using the Monocle2 software (http://cole-trapnell-lab.github.io/monocle-release/). 

Microglia Labeling and Ablation in Zebrafish Retina

1 µg of Dylight 594-labeled isolectin B4 (IB4, Vector Laboratories, Inc., Burlingame, CA, USA) was intravitreally injected 24 hours before euthanization to label retinal microglia. Retinal microglia was ablated using a previously described method.¹⁴ Briefly, adult Tg(mpeg1:nfsB-mCherry) transgenic zebrafish were first housed in fish water with or without 5 mM metronidazole (MTZ; M3764; Sigma-Aldrich, St. Louis, MO, USA) for 3 days. During cell ablation, fresh MTZ solution or fish water was changed twice a day. The retina was then injured at day 4 and the MTZ treatment continued until the end of the experiment. 

Intravitreous IL-6 Injection and Drug Treatment

Zebrafish IL-6 protein (Kingsfisher Biotech, St. Paul, MN, USA) was dissolved as 500 ng/µL stocks in 0.1% BSA solution. 1 µL of IL-6 solution of indicated concentrations was injected intravitreally through the front of the eye once daily for 2 to 4 days. JSI-124 (MedChemExpress, Monmouth Junction, NJ, USA) was dissolved as 1 mM stocks in DMSO solution. 1 µL of 5 µM JSI-124 was injected intravitreally through the front of the eye once daily for 4 days. Di-O-demethylcurcumin (DDC; MedChemExpress) was dissolved as 20 mM stocks in DMSO solution. 1 µL of DDC was injected intravitreally through the front of the eye once daily for 4 days. 

The method for immune suppression has been described previously.¹⁴ Briefly, to suppress retinal immune response, the fish were immersed in 15 mg/L dexamethasone (Dex; Sigma-Aldrich, St. Louis, MO, USA) in system water for 7 days before injury. Dex treatment continued through the entire period of the experiment. To enhance retinal immune response, the fish received intravitreous injection of zymosan A (Zym, 1 µL of 10 mg/mL; Sigma-Aldrich). 

il6r Morpholino Treatment

1 µL of a lissamine-tagged standard control morpholino (MO; 1 mM) that targets a human beta-globin intron mutation or a zebrafish il6r MO of indicated doses was intravitreally injected at the time of the stab injury. The protocol for electroporation of MOs into zebrafish retina has been previously described.⁷ The sequence of the standard control MO is 5′-CCTCTTACCTCAGTTACAATTTATA-3′ and the antisense il6r MO is 5′-GTGTGCAAAAGTCCTTACCCCCTAC-3′. The efficacy of the il6r MO to knock down endogenous zebrafish il6r has been validated in a previous study.²¹ 

Immunofluorescence Staining and MGPC Lineage Tracing

The methods for tissue preparation, cryosection, and immunofluorescence staining were performed using established protocols.⁷ The following primary antibodies were used: rabbit anti-GNAT2 (1:200; MBL, Tokyo, Japan, Cat # PM075); mouse anti-Zpr1 (1:200; Abcam, Cambridge, MA, USA, Cat # ab174435); and mouse anti-Hu-antigen C/D (HuC/D; 1:500; Thermo Fisher Scientific, Cat # A-21271). To label proliferating cells, 5-Ethynyl-2′-deoxyuridine (EdU) staining was performed using a BeyoClickTM EdU-488 Cell Proliferation Detection Kit (Beyotime, Shanghai, China). 

For MGPC lineage tracing, the fish received an intraperitoneal EdU injection (20 µL of 20 mM) at 4 dpi. The fish were then euthanized at 7, 21, or 30 dpi to identify the progeny of MGPCs in the retina. 

Reverse Transcription PCR and Quantitative PCR

Fish retinas were collected and total RNA was isolated using the TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA). RT-PCR was performed as described previously.¹⁴ The qPCR was performed in triplicate, as described previously.⁷ The ΔΔCt method²² was used to determine relative mRNA levels which were then normalized to that of rpl13. Primers used in the study are listed in the Table. 

Table.

 

Primers Used for PCR and qPCR in the Study

 

Imaging and Cell Count

Retinal cryosections were imaged using a Zeiss Imager M2 fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany) equipped with an Axiocam 506 mono camera. Retinal cryosection images for cell counting were captured with a 10× or 20× objective. Cells in each of the four injured retinal region were counted using the Cell Counter plugin of the ImageJ software. 

Statistical Analysis

All experiments were performed in triplicate. The Student's t-test was used for single comparisons, and 1-way analysis of variance (ANOVA) followed by Tukey's post hoc test was used for multiple comparisons. For dose-dependent effects, a correlation analysis was performed and the correlation coefficient (r value) was indicated in the graphs. Error bars represent standard error and statistical significance was defined as P < 0.05. 

Results

Expression of IL-6 and IL-6r in Stab-Injured Zebrafish Retina

We first examined the expression of IL-6 signaling components in the intact and stab-injured zebrafish retina using our recent scRNAseq data.⁴ Gene expression analysis showed that il6 and il6r were mainly expressed by microglia and neutrophil in the retina (Fig. 1A, upper panels). Although MG do not express il6r, increased expression of il6st and stat3 was observed in MG from the injured retinas (see Fig. 1A, lower panels), suggesting that IL-6 released from immune cells might activate the Stat3 signaling pathway in MG via trans-signaling. In our recent study, retinal microglia were clustered into three subtypes: proinflammatory microglia-1, anti-inflammatory microglia-2, and proliferative microglia (Fig. 1B, upper panels).⁴ Pseudotime analysis demonstrated their dynamic status transition following retinal injury (see Fig. 1B, dashed arrows). The expression of il6 and a typical inflammatory cytokine il1b in microglia gradually increased during their transition toward the proinflammatory microglia-1 direction (see Fig. 1B, lower panels). The expression of il6 in microglia peaked at 12 hpi, coinciding with the expression pattern of il6r in neutrophil, and il6st and stat3 in MG (Fig. 1C), suggesting a connection between microglia-derived IL-6 and the Stat3 signaling in MG. This result was further supported by RT-PCR analysis of the expression of these genes in the injured retina (Fig. 1D). Importantly, microglia ablation using the Tg(mpeg1:nsfB-mcherry) transgenic zebrafish, which resulted in the elimination of approximately 70% of microglia in the injured retina,¹⁴ significantly reduced retinal il6 expression at 12 hpi and 1 dpi (Fig. 1E), supporting the notion that microglia are the main source of il6 in the injured zebrafish retina. 

Figure 1.

 

Expression of the IL-6 signaling components in control and stab-injured zebrafish retinas. (A) Single-cell RNAseq analysis showing the expression of il6, il6r, il6st, and stat3 in adult zebrafish retina. (B) Pseudotime analysis showing the expression of il6 and il1b during microglia status transition. The dashed arrows show the direction of the status transition. (C) Dotplot showing the expression of IL-6 signaling component genes in MG, microglia, and neutrophil. (D) PCR analysis showing the expression of il6, il6r, il6st, stat3, and house-keeping gene rpl13 at the indicated time points in zebrafish retina. (E) Microglia ablation using the Tg(mpeg1:nsfB-mCherry) transgenic zebrafish combined with metronidazole (MTZ) treatment abolished injury-dependent retinal il6 expression at 12 hpi and 1 dpi, as shown by qPCR. *, P < 0.05. AC, amacrine cells; BC, bipolar cells; HC, horizontal cells; MG, Müller glia; MTZ, metronidazole; RGC, retinal ganglion cells; RPE, retinal pigment epithelium; tSNE, t-Distributed Stochastic Neighbor Embedding.

Zebrafish IL-6 Injection Upregulates the Expression of its Own Transcript, il11a and Jak-Stat3 Signaling Reporters in the Injured Retina

To validate the efficacy of IL-6 in activating the Stat3 signaling pathway in the retina, zebrafish retinas were stab-injured and received daily intravitreal injection of a recombinant zebrafish IL-6 protein (1 µL of 100 ng/µL) or PBS control for 2 consecutive days. The qPCR was then performed to assess the expression of Jak-Stat3 signaling components in the retina at 2 dpi. Intact retinas were used as the uninjured control. The qPCR demonstrated that IL-6 injection significantly increased the expression of Jak-Stat3 signaling reporters socs3a and socs3b (Fig. 2A), indicating that IL-6 could activate this signaling pathway in the injured zebrafish retina. IL-6 injection also increased the retinal expression of stat3 and il6st, although the increase was not statistically significant (see Fig. 2A). 

Figure 2.

 

Intravitreous injection of zebrafish IL-6 activates Jak-Stat3 signaling and enhances the inflammatory response in injured zebrafish retinas. (A, B) The qPCR analysis showing the mRNA expression of Jak-Stat3 signaling component genes (stat3, il6st, il6r, socs3a, and socs3b) and inflammatory cytokines (il1b, il6, il11a, il11b, tnfa, and tnfb), in the uninjured control, or stab-injured retinas that received daily intravitreal IL-6 or PBS injection at 1 or 2 dpi. (C, D, E) IL-6 injection significantly increased the number and spatial distribution of IB4+ microglia at the injury site at 2 dpi. White *, site of the stab injury. (F) EdU immunofluorescence showing the cell proliferation in PBS control or IL-6-treated retinas at 4 days post injection. (G) Quantification of the total number of INL EdU+ cells in the retina at 4 days post injection. *, P < 0.05; **, P < 0.01, ***, P < 0.001 compared with PBS control. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer.

Because IL-6 could exhibit both pro-inflammatory and anti-inflammatory properties,²³ we further examined its impact on the inflammatory response in the retina after it was administered at the time of injury. The qPCR analysis revealed that IL-6 injection did not elevate the expression of typical inflammatory cytokines, IL-1β (il1b) and TNF-α (tnfa and tnfb), in the injured retina (Fig. 2B). However, it significantly upregulated the expression of its own transcript (il6) and il11a, both of which are known activators of the Stat3 signaling pathway (see Fig. 2B). Fluorescence microscopy showed that IL-6 injection led to a significant increase in both the number and spatial distribution of IB4+ microglia in the retina at the injury site at 2 dpi (Figs. 2C–E). These findings suggest that zebrafish IL-6 may function as an atypical proinflammatory cytokine in the retina. 

To investigate if IL-6 alone is sufficient to trigger a regenerative response in the intact zebrafish retina, various doses of zebrafish IL-6 (100, 200, and 400 ng) were intravitreally injected once daily for 4 days, and the retinas were harvested on the fourth day to examine MG proliferation. EdU staining in the inner nuclear layer (INL) showed that the number of INL proliferating cells in all IL-6 groups was comparable to that of the PBS control (Figs. 2F, 2G), suggesting that zebrafish IL-6 alone is insufficient to promote MG proliferation without additional injury signals in the intact retina. 

IL-6 Regulates the Formation of MGPCs in the Injured Zebrafish Retina

To investigate the role of IL-6 signaling in retina regeneration, we examined the impact of IL-6 loss-of-function on the formation of MGPCs in the retina at 4 dpi. For this purpose, a curcumin analog Di-O-demethylcurcumin (DDC) which inhibits IL-6 production,²⁴ was used to suppress IL-6 activity in the retina (Fig. 3A). The qPCR showed that DDC treatment significantly reduced retinal il6 expression in a dose-dependent manner (Fig. 3B), confirming its effectiveness in IL-6 inhibition. EdU immunofluorescence demonstrated that DDC treatment dose-dependently reduced the number of INL EdU+ cells in the injured region at 4 dpi (see Figs. 3A, 3C), which have been previously identified as MGPCs.²⁵ Furthermore, knocking down il6r by electroporation of a validated il6r MO²¹ (Supplementary Fig. S1) also significantly reduced the number of MGPCs in the injured retina in a dose-dependent manner, as shown by the EdU staining at 4 dpi (Figs. 3D, 3E). These findings suggest that IL-6 signaling is necessary for MGPC formation in the injured zebrafish retina. 

Figure 3.

 

IL-6 promotes MGPC formation in the injured zebrafish retina. (A) EdU immunofluorescence showing the cell proliferation in PBS control or Di-O-demethylcurcumin (DDC)-treated retinas at 4 dpi. (B) The qPCR analysis of il6 mRNA levels in the PBS- or DDC-treated retinas at 12 hpi. Uninjured retina (0 day) served as a negative control. (C) Quantification of the number of INL EdU+ cells per injury at 4 dpi of A. (D) EdU immunofluorescence showing the cell proliferation in retinas electroporated with lissamine-tagged control (ctrl) MO or il6r MO at 4 dpi. (E) Quantification of the number of INL EdU+ cells per injury at 4 dpi of D. (F) EdU immunofluorescence showing the cell proliferation in retinas treated with PBS control or indicated doses of zebrafish IL-6 at 4 dpi. (G) Quantification of the number of INL EdU+ cells per injury in PBS or IL-6-treated retinas at 4 dpi. White *, site of the stab injury. *, P < 0.05; **, P < 0.01; ***, P <0.001. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; r, correlation coefficient.

We next investigated the effect of IL-6 overexpression on MGPC formation in the injured retina. Various doses of zebrafish IL-6 (12.5, 25, 50, 100, and 200 ng) were intravitreally injected once daily for 4 days, and the retinas were collected at 4 dpi. EdU immunofluorescence showed that IL-6 dose-dependently increased the number of MGPCs in the injured retina (Figs. 3F, 3G). At a dosage of 100 ng, IL-6 achieved a statistically significant difference of P < 0.001 in MGPC formation (see Fig. 3G). Therefore, we chose this dosage for subsequent experiments. Together, these findings demonstrate that IL-6 signaling regulates MGPC formation in the stab-injured zebrafish retina. 

IL-6 Signaling Promotes MG Reprogramming and Proliferation via the Jak-Stat3 Pathway

To investigate the role of IL-6 signaling in MG proliferation, zebrafish IL-6 or PBS were intravitreally injected into the retina of the Tg(gfap:GFP) transgenic zebrafish, where GFP is specifically expressed by MG.²⁰ EdU immunofluorescence staining showed that IL-6 injection significantly increased the number of proliferating MG at 2 dpi (EdU/GFP double positive cells; Figs. 4A, 4B). Similarly, knocking down il6r expression by electroporation of the il6r MO into the retina significantly reduced the number of INL EdU+ cells at 2 dpi (Figs. 4C, 4D), which are known to be proliferating MG at this timepoint.^14,26 These results indicate that IL-6 signaling promotes MG proliferation in the injured zebrafish retina. 

Figure 4.

 

IL-6 regulates MG reprogramming and proliferation via Stat3 signaling. (A) EdU immunofluorescence showing the cell proliferation in retinas of the Tg(gfap:GFP) fish at 2 dpi. (B) Quantification of the number of INL GFP+ EdU+ cells per injury of A. (C) EdU immunofluorescence showing MG proliferation in the INL at 2 dpi. Retinas were electroporated with 1 mM of control or il6r MO at the time of the stab injury. (D) Quantification of the number of INL EdU+ cells per injury of C. (E, F, G, H) The qPCR analysis of the expression levels of MG subtype markers and Jak1-Stat3 signaling components in retinas treated with PBS or indicated doses of DDC. Day 0 served as the uninjured control. The statistical analyses were between DDC groups and PBS control. (I) JSI-124 treatment abolished the promoting effect of IL-6 on MG proliferation at 4 dpi. White *, site of the stab injury. *, P < 0.05; **, P < 0.01; ***, P <0.001; DDC, Di-O-demethylcurcumin; GCL, ganglion cell layer; INL, inner nuclear layer; ns, non-significant; ONL, outer nuclear layer.

In our recent scRNAseq study, we demonstrated that in the stab-injured zebrafish retina, resting MG transitioned to immunoregulatory MG and activated MG, with the latter expressing known regeneration-associated genes and cell-cycle markers.⁴ To investigate the role of IL-6 signaling in MG reprogramming, we examined the expression of MG subtype-specific markers by qPCR in PBS control or DDC-treated retinas. The qPCR analysis showed that compared with the PBS control group, DDC treatment significantly alleviated the injury-induced reduction of resting MG marker expression but decreased the expression of immunoregulatory MG markers in a largely dose-dependent manner (Figs. 4E, 4F). Inhibition of IL-6 production by DDC also significantly reduced injury-dependent induction of activated MG markers, including clcf1, lepb, crlf1a, and lin28a (Fig. 4G). These findings indicate that IL-6 signaling regulates MG reprogramming by facilitating their state transition toward the immunoregulatory and activated MG status. To determine if IL-6 signaling regulates MG reprogramming and proliferation via the Jak-Stat3 pathway, we examined the mRNA expression of Jak1-Stat3 signaling components in PBS control or DDC-treated retinas. The qPCR showed that DDC treatment significantly decreased the expression of jak1, stat3, and il6st, as well as the Stat3 signaling reporter genes socs3a and socs3b in the injured retina at 1 dpi (Fig. 4H). Importantly, the promoting effect of IL-6 on MG proliferation was abolished in the presence of a Stat3 inhibitor JSI-124 (Fig. 4I, Supplementary Fig. S2A). TUNEL staining showed that the reduced MG proliferation in JSI-124-treated retinas was independent of increased cell death (Supplementary Figs. S2B, S2C). Together, these results suggest that IL-6 signaling regulates MG reprogramming and proliferation via the Jak-Stat3 signaling pathway in the injured zebrafish retina. 

IL-6 May Also Regulate MGPC Formation by Exerting Phase-Dependent Pro-Inflammatory and Anti-Inflammatory Properties

To further understand the role of IL-6 signaling in the regulation of retina regeneration by inflammation and microglia, we investigated whether IL-6 alone could rescue the MGPC formation defect in immune-suppressed retinas. For this purpose, the zebrafish were immersed in the glucocorticoid Dex solution (15 mg/L) in system water for 7 days prior to injury, and then received daily intravitreal injections of PBS or IL-6 for 4 days following retinal injury (Fig. 5A, experimental timeline). Our results showed that intravitreal injection of 100 ng of zebrafish IL-6 was able to rescue the MGPC defect in the Dex-treated retina (see Fig. 5A). Specifically, the number of INL EdU+ cells in the retinas treated with Dex and IL-6 was significantly higher than those treated with Dex and PBS, and this number was comparable to that of the control group (see Figs. 5A, 5B). IL-6 injection also restored the number of IB4+ microglia in the injury region in Dex-treated retinas at 2 dpi (Fig. 5C). This suggests that, in addition to its role in MG reprogramming and proliferation, IL-6 may also regulate MGPC formation by exerting pro-inflammatory properties when administered from the early phase of retinal injury onward. 

Figure 5.

 

IL-6 may regulate MGPC formation through phase-dependent pro-inflammatory and anti-inflammatory mechanisms. (A) The experimental timeline, and EdU immunofluorescence showing the formation of MGPCs (INL EdU+ cells) in control or Dex-treated retinas at 4 dpi that received daily PBS- or IL-6 injection. (B) Quantification of the number of INL EdU+ MGPCs per injury in A. (C) Quantification of the number of IB4+ microglia at the injury site in retinas that received the same treatment of A at 2 dpi. (D) The experimental timeline, and EdU immunofluorescence showing the formation of MGPCs at 4 dpi in the retinas that received intravitreous injection of PBS, Dex, or Zym. (E) Quantification of the number of IB4+ microglia at the injury site in the retina in D at 4 dpi. (F) Quantification of the MGPC number per injury at 4 dpi in D. (G) The experimental timeline, and immunofluorescence showing the INL EdU+ MGPCs and IB4+ microglia at 4 dpi in retinas that received intravitreous injection of PBS or zebrafish IL-6 protein (100 ng). (H, I) Quantification of the number of MGPCs and IB4+ cells per injury in G at 4 dpi. (J) The experimental timeline, and qPCR analysis of the mRNA expression of indicated cytokines at 3 dpi in retinas that received intravitreous injection of PBS or zebrafish IL-6 protein (100 ng). White *, site of the stab injury. *, P < 0.05; **, P < 0.01; ***, P <0.001; Dex, dexamethason; GCL, ganglion cell layer; INL, inner nuclear layer; ns, non-significant; ONL, outer nuclear layer; Zym, zymosan A.

We previously showed that inflammation exhibits context-dependent effects on MG/MGPC proliferation and retinal neuron regeneration.¹¹ It was also shown that immune suppression before or after retinal injury had different impacts on MG proliferation in larval zebrafish.¹⁵ To investigate whether inflammation has phase-dependent effects on MGPC formation, the immune response was manipulated within a later time window, from 2 to 4 dpi, in the stab-injured retina (Fig. 5D, 5E). Fluorescence microscopy showed that Dex treatment significantly reduced the number of IB4+ microglia in the injured retina (see Fig. 5E). We did not observe an increase in the number of IB4+ cells at the injury site in retinas treated with Zymosan A (Zym; see Fig. 5E), suggesting that Zym may enhance retinal inflammation through mechanisms independent of microglia infiltration/proliferation during this phase. Surprisingly, Dex treatment led to an increase in the number of EdU+ MGPCs, whereas the Zym injection had an opposite effect (see Figs. 5D, 5F), in sharp contrast to the results of immune manipulation before or immediately following retinal injury (see Fig. 5A).^11,14 This indicates that inflammation during this late phase (2–4 dpi) inhibits MGPCs formation, and a rapid resolution of inflammation after the early phase is required for proper retina regeneration. 

We next investigated the effects of IL-6 injection on MGPC formation when administered between 2 and 4 dpi (Fig. 5G, experimental timeline). We hypothesized that if IL-6 exhibits pro-inflammatory properties during this phase, it might mimic the adverse effect of Zym injection on MGPC formation (see Figs. 5D, 5F). Interestingly, EdU immunofluorescence showed that IL-6 injection during this phase significantly increased the number of MGPCs at 4 dpi (see Figs. 5G, 5H), similar to the effect of Dex treatment administered from 2 to 4 dpi. This suggests that IL-6 may exert anti-inflammatory properties during the late phase of retinal injury. Supporting this, IL-6 injection starting at 2 dpi significantly reduced the number of IB4+ microglia at the injury site at 4 dpi (see Figs. 5G, 5I), and decreased the expression of a typical proinflammatory cytokine IL-1β in the retina (Fig. 5J). These findings align with an anti-inflammatory role of IL-6 in the late phase. Collectively, these results demonstrate that, beyond its role in MG reprogramming and proliferation, IL-6 may also regulate MGPC formation via phase-dependent pro-inflammatory and anti-inflammatory mechanisms. 

IL-6 Signaling Facilitates the Regeneration of Major Retinal Neurons in Injured Zebrafish Retina

To investigate the role of IL-6 signaling in the regeneration of retinal neurons, we first knocked down il6r expression at the time of the stab injury, labeled the MGPCs with EdU at 4 dpi, and determined their early differentiation by examining the colocalization of EdU and HuC/D signals at 7 dpi.¹⁴ EdU immunofluorescence showed that il6r knock down led to a significant reduction in the number of MGPCs at 7 dpi (Figs. 6A, 6B). Additionally, il6r knock down significantly reduced the number of EdU and HuC/D double-positive cells (see Figs. 6A, 6C), indicating a reduced number of early differentiating MGPCs. Interestingly, il6r knock down also decreased the proportion of EdU+ cells expressing HuC/D at this time point (see Figs. 6A, 6D), suggesting that in addition to its effect on the MGPC population, IL-6 signaling may also be required for the early differentiation of MGPCs in the injured retina. 

Figure 6.

 

IL-6 signaling promotes the regeneration of retinal neurons in injured retinas. (A) EdU and HuC/D immunofluorescence showing the early MGPC differentiation at 7 dpi in retinas electroporated with 1 mM of control MO (ctrl MO) or il6r MO. (B, C, D) Quantification of the total number of EdU+ cells per injury B, the number of EdU and HuC/D double positive cells per injury C, and the proportion of EdU+ cells expressing HuC/D per injury D in A. (E) Immunofluorescence showing the regeneration of retinal neurons at 30 dpi in retinas electroporated with 1 mM of ctrl MO or il6r MO. (F) Quantification of the number of EdU+ cells in each layer per injury at 30 dpi in E. (G) Quantification of the number of regenerated photoreceptor (ONL Gnat+/EdU+), amacrine cells (INL HuC/D+/EdU+), and RGCs (GCL HuC/D+/EdU+) per injury at 30 dpi in E. (H) Immunofluorescence showing the regeneration of retinal neurons at 21 dpi in retinas treated with PBS control or IL-6. (I) Quantification of the number of EdU+ cells in each layer per injury at 21 dpi in H. (J) Quantification of the number of regenerated photoreceptor (ONL Zpr1+/EdU+), amacrine cells (INL HuC/D+/EdU+), and RGCs (GCL HuC/D+/EdU+) per injury at 21 dpi in H. *, P < 0.05; **, P < 0.01; ***, P <0.001; GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer.

We next investigated the role of IL-6 signaling in the regeneration of major retinal neurons at later time points. For this purpose, the fish retina was first electroporated with a standard control- or il6r-MO at the time of the stab injury. MGPCs were then labeled by a pulse of EdU at 4 dpi, and their progeny cells were examined for expression of retinal neuron markers at 30 dpi. Immunofluorescence staining showed that il6r knockdown led to a significant reduction of EdU+ cells across all 3 nuclear layers at 30 dpi (Figs. 6E, 6F). Importantly, loss of IL-6 signaling led to a significant reduction in the number of regenerated photoreceptors (ONL Gnat2⁺/EdU⁺ cells), amacrine cells (INL HuC/D⁺/EdU⁺ cells), as well as RGCs (GCL HuC/D⁺/EdU⁺ cells; see Figs. 6E, 6G). Conversely, intravitreal IL-6 injection following retinal injury produced the opposite effect on retinal neuron regeneration. Our results showed that IL-6 injection increased the number of MGPC progeny cells in the retina (Figs. 6H, 6I), and led to the regeneration of more photoreceptors (ONL Zpr1⁺/EdU⁺ cells), amacrine cells (INL HuC/D⁺/EdU⁺ cells), and RGCs (GCL HuC/D⁺/EdU⁺ cells) at 21 dpi (see Figs. 6H, 6J). Together, these findings demonstrate a crucial role of IL-6 signaling in the neuronal regeneration of injured zebrafish retinas. 

Discussion

Inflammation and microglia play essential roles in retina regeneration in fish and avian species. Our recent study demonstrated that inflammation upregulates the expression of Jak-Stat3 signaling components in MG, promoting their reprogramming and proliferation.⁴ However, the specific inflammatory molecules mediating inflammation's effect on the Jak-Stat3 signaling pathway have remained unclear. In this study, microglia-derived IL-6 was identified as a crucial factor required for retina regeneration in the stab-injured zebrafish retina. We show that IL-6 regulates MG reprogramming and status transition via Jak-Stat3 signaling. Furthermore, IL-6 signaling promotes the formation of MGPCs and their early differentiation, as well as the regeneration of major retinal neurons. Interestingly, IL-6 may also regulate MGPC formation via phase-dependent pro-inflammatory and anti-inflammatory mechanisms. These findings underscore the pivotal roles of IL-6 in zebrafish retina regeneration. 

A previous study reported that Leptin and IL-6-family cytokines, such as IL-11 and CNTF, were expressed in MGPCs shortly after retinal injury, synergizing to regulate MG reprogramming and proliferation in zebrafish.²⁷ However, PCR analysis in that study failed to detect il6 mRNA expression in stab-injured adult zebrafish retina. In our study, PCR analysis found rapid il6 induction in the stab-injured zebrafish retina (see Fig. 1),⁴ primarily expressed by microglia as shown by scRNAseq (see Fig. 1). The reason for this discrepancy is unclear, and we speculate it may stem from differences in zebrafish genetic backgrounds or the severity of retinal injury between the two studies. Additionally, the previous study showed that intravitreal injection of a recombinant human IL-6 protein was sufficient to induce MG proliferation in the uninjured zebrafish retina.²⁷ In contrast, we showed that whereas intravitreal injection of a recombinant zebrafish IL-6 protein activated the Stat3 signaling in the injured retina, various dosages of this protein failed to induce MG proliferation in the intact zebrafish retina (see Fig. 2). The use of IL-6 proteins from different species likely contributed to these divergent outcomes, highlighting the need for further research to explore species-specific effects of IL-6. Nevertheless, our results align with findings from our previous studies showing that zymosan-induced inflammation was sufficient to activate Jak-Stat3 signaling in MG, yet insufficient to induce MG proliferation in the uninjured zebrafish retina.^4,11 

In this study, we demonstrate that IL-6 regulates MG status transition via the Jak1-Stat3 signaling. The qPCR showed that inhibition of IL-6 production blocked MG status transition toward the immunoregulatory and activated state in the injured zebrafish retina (see Fig. 4). The impact of IL-6 inhibition on MG status transition mirrors that of immune suppression or Stat3 inhibition, as observed in our recent study,⁴ indicating a shared mechanism in their regulation of MG reprogramming. Previous studies have highlighted the necessity of Jak-Stat3 signaling in MG reprogramming and MGPC formation in fish and avian retinas.^27,28 It has been reported that Stat3 binds to the promoter of a key reprogramming factor ascl1a, promoting its transcription in MG following in injured zebrafish retina.²⁷ Our recent research further demonstrated that Stat3 signaling also regulates the expression of a pluripotent factor lin28a in MG following retinal injury.⁴ These findings underscore the role of Jak-Stat3 signaling in MG reprogramming and proliferation through the transcriptional regulation of critical reprogramming factors. Importantly, our study found that inhibition of IL-6 production or inflammation suppressed the expression of several Stat3-activating cytokines by MG, including lepb, clcf1, and crlf1a⁴ (see Fig. 4). This suggests that microglia-derived IL-6 may act as the initial injury signal triggering Stat3 activation in MG, which then transitions to an activated status expressing other Stat3-activating cytokines to maintain or amplify this signaling in the regenerative niche. 

IL-6 can signal through classical pathway by binding to IL-6r on the membrane of target cells, or through a trans-signaling mechanism by binding to soluble IL-6r (sIL-6r).²⁹ The classical IL-6 signaling is mainly anti-inflammatory and regenerative,^19,23 whereas its trans-signaling has been traditionally viewed as predominantly pro-inflammatory and implicated in disease pathogenesis over the past decade.^18,19,23,29 However, recent studies have also highlighted the involvement of IL-6 trans-signaling in tissue regeneration,^30,31 indicating a dual role for IL-6 in both disease progression and regeneration. Previous studies have reported the expression of membrane-bound IL-6r by MG in mammalian retinas.^32,33 However, our scRNAseq data revealed that il6r was not expressed in MG in either uninjured or injured zebrafish retinas at the tested time points (see Fig. 1). This suggests that microglia-derived IL-6 may signal via trans-signaling to activate the Jak-Stat3 pathway in MG. We demonstrated that knocking down IL-6R by an il6r MO in the injured retina suppressed MG proliferation and MGPC formation (see Figs. 3, 4). It is likely that the il6r MO inhibited IL-6R protein translation in retinal microglia or neutrophils that express il6r, resulting in the reduction or elimination of its soluble form (sIL-6r) in the extracellular environment. Notably, we discovered that IL-6 may also regulate MGPC formation via phase-dependent pro-inflammatory and anti-inflammatory properties (see Fig. 5). IL-6’s pro-inflammatory role in the early phase aligns with its proposed trans-signaling to MG, a known source of inflammatory factors and chemokines in the injured retina.^4,34,35 Conversely, IL-6’s anti-inflammatory function in the late phase (2–4 dpi) is also consistent with its potential classical signaling to microglia or neutrophil, which express IL-6 receptor (il6r), as shown in our scRNAseq data. However, further investigations that dissect the specific roles of IL-6 classic versus trans-signaling in MG reprogramming/proliferation and its dual regulation of the inflammatory response, are necessary to fully understand how IL-6 regulates retina regeneration. 

Acknowledgments

The authors thank the Institute of Neuroscience, Chinese Academy of Science for sharing the Tg(gfap:GFP) zebrafish. 

Supported by the National Natural Science Foundation of China (81970820, 81801331, and 31930068), National Key Research and Development Project of China (2017YFA0701304 and 2017YFA0104100), and Shanghai Yangzhi Rehabilitation Hospital (Shanghai Sunshine Rehabilitation Center) Talent introduction plan (KYPT202204). 

Author Contributions: H.X. and L.C. conceived the study; J.X. and Y.L. performed the experiments; J.X., Y.L., X.L., X.T., L.L., L.C., and H.X. analyzed the data; H.X. designed the experiments and wrote the paper. 

Ethics Statements: All fish used in this study were treated in accordance with the ARVO Animal Statement. 

Data Availability: The data supporting the results reported in the article are available upon request to the corresponding authors. 

Disclosure: J. Xu, None; Y. Li, None; X. Li, None; X. Tan, None; L. Liu, None; L. Cao, None; H. Xu, None