Female and male zebrafish (Danio rerio) were used in this study and maintained at 28°C under normal light conditions (14 h light:10 h dark) in the Freimann Life Science Center at the University of Notre Dame as described previously (Vihtelic and Hyde, 2000). Fish were 6–12 months old and 4–5 cm in length. The fish lines utilized in this study include albino, and albino;Tg(gfap:EGFP) ^nt11 . All animal care protocols were approved by the University of Notre Dame Animal Care and Use Committee and in compliance with the Association of Research in Vision and Ophthalmology for the use of animal in vision research.

Adult albino and albino;Tg(gfap:EGFP) ^nt11 zebrafish were dark-adapted for 2 weeks and exposed to constant intense light for up to 96 h as previously described (Vihtelic and Hyde, 2000; Kassen et al., 2006; Lahne et al., 2021). The temperature of the tanks was maintained at 32°C, which is optimal to induce maximal photoreceptor cell death. Fish were euthanized in a 1:500 dilution of 2-phenoxyethanol (2-PE; 77699; Sigma-Aldrich, St. Louis, MO) in system water after each collection timepoint.

Morpholino oligos were synthesized by GeneTools LLC (Philomath, OR) and contained a positively charged lissamine tag. Dark-adapted albino; Tg(gfap:EGFP) ^nt11 zebrafish were intravitreally injected with 0.5 μL of either 1 mM Standard Control, il-1β or il-10 translation-blocking morpholinos. Electroporation were then performed as previously described (Thummel et al., 2011; Campbell et al., 2021) and immediately placed in constant light treatment following electroporation. The following morpholinos were used: Standard Control: 5′-CCT​CTT​ACC​TCA​GTT​ACA​ATT​TAT​A-3’ (Nasevicius and Ekker, 2000), il-1β: 5′- CCC​ACA​AAC​TGC​AAA​ATA​TCA​GCT​T -3' (López-Muñoz et al., 2011; Frame et al., 2020; Wattrus et al., 2022), il-10: 5′- AAT​CAG​TGG​CAC​TTA​CGT​TTA​TGT​T- 3′, tnfa: 5′- AGC​TTC​ATA​ATT​GCT​GTA​TGT​CTT​A- 3’ (Nelson et al., 2013).

Immunohistochemistry was performed as previously described (Lahne et al., 2021; Boyd et al., 2023). Eyes were collected and fixed in a 9:1 ethanolic formaldehyde solution overnight at 4°C. Eyes were rehydrated in a series of ethanol washes and incubated in 30% sucrose overnight at 4°C. Eyes were then cryoprotected in a mixture of 2:1 Tissue Freezing Media:30% sucrose overnight at 4°C. Retinas were then frozen in Tissue Freezing Media, sectioned at 14 μm thickness, and stored at −80°C until immunostaining.

Slides were rehydrated in PBS for 30 min at room temperature and blocked in 2% DMSO, 2% normal goat serum, 1% Tween-20, and 0.4% Triton X in PBS for 1 h at room temperature. Primary antibodies diluted in blocking solution were applied on the slide and incubated overnight at room temperature. Primary antibodies used in this study included: chicken anti-GFP (1:1000; AB13970; Abcam), mouse anti-PCNA (1:1000; P8825; Sigma-Aldrich) and rabbit anti-Lcp1 (1:400; GTX134687; GeneTex, Irvine, CA). Slides were washed with PBS/0.05% Tween-20 (PBS-T) and incubated in secondary antibodies diluted in blocking solution for 1 h at room temperature. Fluorescent-tagged secondary antibodies (Life Technologies, Carlsbad, CA) used in this study included: Alexa Fluor goat anti-chicken 488 (1:1000; A11039), Alexa Fluor goat anti-mouse 594 (1:1000; A11032), and Alexa Fluor goat anti-rabbit 647 (1:1000; A21245). 4,6-diamidino-2phenylindol (DAPI; 1:1000; D1306, Thermo Fisher Scientific; Waltham, MA) was also applied for nuclear localization. Slides were washed in PBS-T and mounted in Prolong Gold Antifade Reagent (P36930; Life Technologies).

Dorsal retinas from adult albino zebrafish were isolated at specific time points of constant light and RNA isolation was performed with Trizol (15596-026; Thermo Fisher) as previously described (Campbell et al., 2021; Lahne et al., 2021). For each time point, 6-7 dorsal retinas were collected and RNA concentrations were measured. The cDNA samples were synthesized using LunaScript RT SuperMix kit (E3101; New England Biolabs, Ipswich, MA) and stored at −80°C. Sample reactions were assembled using Luna universal qPCR Master Mix (M3003, New England Biolabs) following the manufacturer’s instructions with TaqMan primers listed in Table 1 (Applied Biosystem). StepOnePlus Real-Time PCR system (4276600; Thermo Fisher) was used to perform qRT-PCR reactions with the following conditions: 1 min at 95°C, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. qRT-PCR reactions were analyzed using the comparative ΔΔCT method as previously described (Gorsuch et al., 2017; Lahne et al., 2021) using 18 s rRNA as the reference gene. The CT values for 18 s rRNA are consistent across all rounds of qRT-PCR experiments. Light-treated retinas used 0 h LT to generate a log₂-fold change in gene expression levels (Kassen et al., 2006) and undamaged retinas used PBS (vehicle) control to generate a log₂-fold change in gene expression levels. Three biological replicates were conducted with each sample.

RNAscope Multiplex Fluorescent v2 Assay (Advance Cell Diagnostics; Newark, CA) was utilized for RNA in situ hybridization. Sample fixation and immunohistochemistry with in situ hybridization were performed according to a protocol described previously (Campbell et al., 2021; Boyd et al., 2023). Briefly, frozen tissue sections were washed in PBS for 5 min before baking at 60°C for 1 h. Sections were then post-fixed in 4% paraformaldehyde at room temperature for 1 h. Sections were dehydrated in sequential 50%, 70%, 100% ethanol washes for 5 min each and twice in 100% before baking at 60°C for 1 h. Slides were given a hydrogen peroxide treatment provided by the manufacturer (Advance Cell Diagnostic) at room temperature for 1 h. The slides were washed in distilled water and immediately submerged into a mildly boiling Target Retrieval Reagent solution (Advance Cell Diagnostic) for 15 min. Sections were placed in distilled water at room temperature and dehydrated in 100% ethanol. A hydrophobic barrier (ImmEdge Hydrophobic pen; H-4000; Vector Laboratories; Burlingame, CA) was applied on the slides and baked at 60°C for 1 h before drying overnight at room temperature.

Sections were given a Protease III treatment at 40°C for 30 min in the HybEZ™ Oven (Advance Cell Diagnostic). Slides were washed with distilled water followed by probe hybridization at 40°C for 2 h. The probe used in this study was Dr-il1b-C1 (432971; Advance Cell Diagnostic), which was taken from NCBI accession of D. rerio il-1β: NM_212,844.2. A 3-plex negative control probe, provided by the manufacturer, was applied simultaneously with experiment slides. Following probe hybridization, sections were incubated with a probe amplification AMP1 solution at 40°C for 30 min followed by a buffer wash. Signal development with HRP was based on the manufacturer’s protocol with Opal 570 (1:1000; FP1488001KT; Akoya Biosciences; Menlo Park, CA). Once signal development was complete, slides were washed in PBS-T for 5 min at room temperature before proceeding with immunohistochemistry. Primary antibodies used were chicken anti-GFP (1:1000; AB13970; Abcam) and rabbit anti-Lcp1 (1:400, GTX134697; Genetex, Irvine, CA). Fluorescent-tagged secondary antibodies used were Alexa Fluor goat anti-chicken 488 (1:1000; A11039; Thermo Fisher Scientific) and Alexa Fluor goat anti-rabbit 647 (1:1000; A21245; Thermo Fisher Scientific) and DAPI (1:1000; D1306; Thermo Fisher Scientific) was applied to stain nuclei.

A Nikon A1 confocal microscope and Leica Stellaris eight confocal microscope with a 40x oil-immersion objective were used to acquire approximately 10 µm z-stack with 1.0 µm step size images of the central dorsal retina. Cell counts were manually performed using FIJI/ImageJ software and normalized to a 300 µm length of the retina. Scale bars for images are either 20, 10, or 5 μm and are indicated in the figures and figure legends.

We used single-cell RNA-Seq (scRNA-Seq) data from whole-light-damaged retinas that was previously published (Hoang et al., 2020). scRNA-seq analysis was performed using Seurat (Hao et al., 2021). Violin plots were generated using cell clusters identified and maintained from Hoang et al. (2020). Differential expression analyses were performed with microglia and Müller glia.

The data in this study were analyzed with at least three independent trials. The sample size and the mean ± SEM are stated in the text and each figure legend. Statistical analysis was performed using GraphPad Prism 10 (San Diego, CA). Statistical significance was determined using either a Student’s t-test, a one-way ANOVA followed by Bonferroni’s pos-hoc test, or a two-way ANOVA followed by Bonferroni’s post hoc test. A p-value less than 0.05 indicated statistical significance, with graphs displaying * for p ≤ 0.05, ** for p ≤ 0.01, *** for p ≤ 0.001, **** for p ≤ 0.0001, and n. s. for no significance.

It was previously demonstrated that microglia are required for Müller glia proliferation following retinal damage (White et al., 2017; Conedera et al., 2019; Silva et al., 2020; Iribarne and Hyde, 2022). To confirm that microglia are required for retinal regeneration in light-damaged adult zebrafish retinas, we utilized two pharmacological treatments to reduce the number of activated microglia. Pexidartinib, or PLX3397 (PLX), is a small molecule that targets the colony-stimulating factor 1 receptor (CSF1R), which blocks CSF1 binding, a regulator of microglia and macrophage survival, production, and differentiation (Elmore et al., 2014; Conedera et al., 2019; Han et al., 2022). Dexamethasone (Dex) is an anti-inflammatory glucocorticoid that has immunosuppressive effects on microglia (Park et al., 2019; Hui et al., 2020; Iribarne and Hyde, 2022). Together, these drugs should reduce the number of microglia and the inflammatory response of any remaining microglia. Dark-adapted albino;Tg(gfap:EGFP) ^nt11 zebrafish were intravitreally injected with a combination of PLX and Dex every 24 h until 72 h of constant light treatment (LT). Retinas were collected at 0, 24, 36, and 72 h LT and immunolabeled for GFP and PCNA to label proliferating Müller glia and Lcp1, a marker commonly used for microglia, macrophages, and other leukocyte cells (Figures 1A–H). Because macrophages and other leukocyte cells are typically uncommon in the retina, unless the blood-brain barrier is compromised, Lcp1⁺ retinal cells are predominantly microglia.

The PLX + Dex coinjected retinas possessed significantly fewer Lcp1⁺ cells compared to the DMSO (vehicle) control group at all time points of constant light (Figure 1I; DMSO: 24 h LT: 9.78 ± 0.99, 36 h LT: 7.40 ± 1.08, 72 h LT: 6.10 ± 0.62; PLX + Dex: 24 h LT: 2.11 ± 0.20, p < 0.0001, 36 h LT: 2.10 ± 0.23, p < 0.0001, 72 h LT: 1.56 ± 0.24, p < 0.0001). The reduced number of microglia during LT also correlated with significantly fewer PCNA⁺ Müller glia in the INL at 36 h LT (Figure 1J; DMSO: 17.6 ± 3.50; PLX + Dex: 6.50 ± 1.41, p = 0.0005). At 72 h LT, when Müller glia-derived neuronal progenitor cells (NPCs) are present in both the INL and ONL, there were significantly fewer proliferating NPCs in the drug-treated retinas relative to the controls (Figure 1K; DMSO INL: 35.70 ± 3.29, ONL: 55.00 ± 3.98; PLX + Dex INL: 15.33 ± 1.36, p < 0.0001, ONL: 14.56 ± 1.42, p < 0.0001). These findings suggest that the combined treatment of PLX and Dex effectively reduced the number of microglia and proliferating Müller glia in light-damaged retinas.

In pathological conditions, activated microglia play a critical role in regulating inflammatory signaling, which attracts other cells to the site of injury and modulates the surrounding microenvironment (Streit et al., 2004; Bosak et al., 2018; Mitchell et al., 2018; Iribarne and Hyde, 2022). To examine the expression of a subset of inflammatory cytokines during LT, we used quantitative real-time PCR (qRT-PCR). Dark-adapted albino zebrafish were light-damaged, and retinas were collected and RNA isolated at 0, 6, 12, 24, 36, and 72 h LT, as well as 1-day (120 h) and 7-days (264 h) after terminating the constant light treatment at 96 h LT. These time points were selected based on established events observed during light-induced damage and regeneration of the adult zebrafish retina, including: 1) photoreceptor cell death between 12 and 24 h, 2) Müller glia proliferation 31–36 h, and 3) amplification and migration of NPCs to the ONL between 68 and 96 h (Vihtelic and Hyde, 2000; Gorsuch and Hyde, 2014). Two additional time points were included: 1-day (120 h) and 7-days (264 h) after terminating the constant light treatment at 96 h LT, which represent time points that correspond to the regeneration of retinal neurons (Lyu et al., 2023). We investigated the temporal expression of three cytokine genes commonly associated with pro-inflammatory effects: interleukin-1β (il-1β), interleukin-6 (il-6), and interferon gamma one (ifng1) and three cytokine genes usually associated with anti-inflammatory effects: interleukin-4 (il-4), interleukin-10 (il-10), and interleukin-13 (il-13). The expression of all three cytokine genes in the first group rapidly increased during the first 24 h after initiating the LT and then decreased over the next several days (Figure 2A). In contrast, the three cytokine genes in the latter group peaked in their expressions between 72 and 120 h after the start of constant light. (Figure 2A). Thus, these six cytokines exhibit temporally distinct expression profiles following light damage.

We previously demonstrated that tumor necrosis factor-alpha (tnfa) expression increases in damaged/dying retinal neurons and is required and sufficient to induce Müller glia proliferation (Nelson et al., 2013). We hypothesized that TNFα expression from the dying photoreceptors might induce il-1β in the microglia. To test this, we intravitreally injected and electroporated either the Standard Control morpholino (S.C. MO), which is not complementary to any known sequence in the zebrafish genome, or the tnfa MO (Nelson et al., 2013) into dark-adapted albino;Tg(gfap:EGFP) ^nt11 zebrafish. Fish were then placed in constant bright light for either 0 h or 24 h and then retinas were collected, mRNA was purified from dorsal retinas, and subjected to qRT-PCR using primers for the il-1β mRNA. As expected, there was a significant increase in il-1β expression at 24 h LT relative to 0 h LT (Figure 2B). In addition, the tnfa morphant expressed significantly lower amounts of il-1β relative to the S.C. morphant (Figure 2B). This is consistent with TNFα being required to induce the increased expression of il-1β in the light-damaged retina.

We analyzed il-1β further because it had the highest level of expression of the three early onset expressing cytokines that we tested. To identify what cells expressed il-1β during constant light treatment, we analyzed il-1β expression in our previously published scRNA-Seq dataset (Hoang et al., 2020). The time points that were analyzed in this dataset were 0, 4, 10, 20, and 36 h LT. Expression of il-1β increased rapidly, peaking at 20 h LT and gradually declining (Figure 2C). Within this dataset of microglia from 0 to 36 h LT, the average il-1β expression peaked at 4 h LT (2.20), as did the percentage of microglia expressing il-1β (Figure 2C, 73%). We then examined if the Müller glia expressed il-1β and found that, while il-1β expression was significantly lower in the Müller glia relative to the microglia, the average il-1β expression peaked at 36 h LT (0.044) and the percentage Müller glia expressing il-1β peaked at 20 h LT (Figure 2D, 3.3%). Thus, il-1β expression is rapidly expressed in microglia after the start of constant light damage and then followed by expression in the Müller glia.

To confirm the spatial pattern of il-1β expression during LT, we utilized RNAscope in situ hybridization with immunohistochemistry (Figures 2E–J’’’). Low levels of il-1β probe signal were present at 0 h and increased through 24 h LT (Figures 2E’–H’), before decreasing in expression at 36 and 72 h LT (Figures 2I’, J’), consistent with the qRT-PCR results. Over the time course, the il-1β probe colocalized with both the Lcp1⁺ microglia (Figures 2E’’–J’’) and GFAP⁺ Müller glia cell populations (Figures 2E’’’–J’’’). Taken together, these data reveal that inflammatory cytokines are expressed in a dynamic temporal manner during LT, and the pro-inflammatory cytokine gene il-1β is expressed sequentially in both microglia and Müller glia.

To investigate how pro-inflammatory cytokine IL-1β may affect Müller glia proliferation following light damage, we inhibited IL-1β secretion by using caspase-1 inhibitor, Ac-YVAD-cmk (YVAD), which is a tetrapeptide that blocks caspase-1 cleavage of the pro-IL-1β precursor and obstructs the secretion of the mature IL-1β from the cell (Gemma et al., 2007; Vojtech et al., 2012; Lin et al., 2018; Liang et al., 2019). Dark-adapted albino;Tg(gfap:EGFP) ^nt11 zebrafish were intravitreally injected with YVAD 24 h prior to the start of LT and every 24 h afterward. Retinas were collected at 36 h LT and immunolabelled for GFP and PCNA (Figures 3A–D). YVAD-treated retinas showed a significant decrease in the number of PCNA⁺ Müller glia relative to the DMSO (vehicle) control group (Figure 3E; DMSO: 23.3 ± 2.29; YVAD: 8.83 ± 1.30, p < 0.0001), suggesting that reducing the amount of mature IL-1β resulted in fewer proliferating Müller glia during LT.

To independently confirm this result, we also used morpholinos to knockdown IL-1β expression. Dark-adapted albino;Tg(gfap:EGFP) ^nt11 zebrafish were electroporated with either a Standard Control morpholino (S.C. MO) or il-1β MO (Figures 3F–I). Similar to the YVAD-treated zebrafish, the il-1β morphants possessed significantly fewer number of PCNA⁺ Müller glia relative to the S.C. morphants (Figure 3J; DMSO: 27.2 ± 2.48; il-1β: 7.34 ± 2.00, p < 0.0001). Because the il-1β morphant did not possess significantly fewer pyknotic nuclei at 36 h LT relative to the S.C. morphant (Supplementary Figures S1A, B, D, E), the reduced number of proliferating Müller glia in the il-1β morphant relative to the S.C. morphant was not due to decreased cell death. These data confirm that the pro-inflammatory cytokine IL-1β is required for Müller glia proliferation in light-damaged retinas.

We next examined if the morpholino specifically knocked down IL-1β expression by testing if Müller glia proliferation in light-damaged albino;Tg(gfap:EGFP) ^nt11 il-1β morphant retinas was rescued by intravitreal injection of zebrafish IL-1β protein. Standard Control and il-1β morphant albino;Tg(gfap:EGFP) ^nt11 fish were intravitreally injected with recombinant zebrafish IL-1β protein every 24 h beginning at the start of constant light treatment (Figures 4A–F). As expected, fish that were electroporated with either il-1β MO alone or il-1β MO with PBS (vehicle) possessed significantly fewer PCNA⁺ Müller glia at 36 h LT relative to S.C. morphants (Figures 4A, B, D, E, G; S.C. MO alone: 28.70 ± 3.54, il-1β MO alone: 11.50 ± 1.88, p = 0.001; S.C. MO with PBS: 26.50 ± 2.62, il-1β MO with PBS: 10.1 ± 1.62, p = 0.002). In contrast, there was no significant difference in the number of PCNA⁺ Müller glia between the il-1β morphants injected with IL-1β protein and S.C. morphants injected with IL-1β protein (Figures 4C, F, G; S.C. MO with IL-1β protein: 22.70 ± 3.28, il-1β MO with IL-1β protein: 27.80 ± 3.61, p > 0.99). This data suggests that IL-1β protein rescued the Müller glia proliferation in il-1β morphants following light damage.

Because IL-1β is required for Müller glia proliferation, we examined if IL-1β is also sufficient to induce Müller glia proliferation in undamaged retinas. We intravitreally injected recombinant zebrafish IL-1β protein into undamaged albino;Tg(gfap:EGFP) ^nt11 zebrafish every 24 h and collected at 2- and 3-days following the initial injection (dpi) (Figures 4H–K). At 2dpi, there were no significant differences in the number of proliferating Müller glia between retinas injected with PBS and IL-1β (Figure 4L; PBS: 0.55 ± 0.16, IL-1β: 8.64 ± 2.06, p = 0.105). By 3dpi, retinas treated with IL-1β protein had significantly greater number of PCNA⁺ Müller glia and NPCs in the INL compared to the PBS (vehicle) control (Figure 4L; PBS: 0.40 ± 0.27, IL-1β: 66.50 ± 4.86, p < 0.0001) and in the ONL (Figure 4M; PBS: 1.7 ± 0.76, IL-1β: 64.33 ± 7.66, p < 0.0001). Furthermore, the number of Lcp1⁺ cells also significantly increased in retinas injected with IL-1β protein in both the INL and ONL at 2dpi (Figures 4N, O; PBS INL: 4.91 ± 0.68, IL-1β INL: 7.73 ± 0.72, p = 0.0054; PBS ONL: 2.73 ± 0.62, IL-1β ONL: 9.91 ± 1.32, p < 0.0001) and 3dpi (Figure N–O; PBS INL: 6.60 ± 0.58, IL-1β INL: 10.08 ± 0.50, p = 0.0006; PBS ONL: 4.50 ± 0.69, IL-1β ONL: 16.50 ± 1.25, p < 0.0001). However, we did not observe any signs of cell loss (TUNEL+ cells, Supplementary Figure S2) or pyknotic nuclei in the IL-1β-injected retinas (Figures 4J, K). This demonstrates that the pro-inflammatory cytokine IL-1β protein is sufficient to induce Müller glia proliferation even without retinal damage.

We next examined if injections of IL-1β affected the gene expression profiles of other inflammatory cytokines in the undamaged retinas. We performed qRT-PCR on albino zebrafish that were intravitreally injected with IL-1β protein for 3 days (Figure 4P). Only one of the five cytokines examined showed a significant increase in gene expression, il-10 (Figure 4P; il-10: 2.90 ± 1.1, p = 0.0420) relative to the PBS control. Thus, not only is IL-1β necessary and sufficient to induce Müller glia proliferation following light damage, but it also likely activates the expression of il-10 and possibly other cytokines.

Cytokine IL-10 often possesses anti-inflammatory effects that can regulate the expression of pro-inflammatory cytokines released by immune cells, such as microglia (O’Garra and Vieira, 2007; Lobo-Silva et al., 2017; Shemer et al., 2020). To determine if IL-10 is required for Müller glia proliferation following light damage, Standard Control and il-10 morpholinos were intravitreally injected alone or with PBS every 24 h until 36 h LT (Figures 5A–F). Similar to the il-1β morphants, the il-10 morphants had significantly fewer proliferating Müller glia in the INL compared to the S.C. morphants either MO alone or with PBS groups (Figures 5A, B, D, E, G; S.C. MO alone: 28 ± 2.56, il-10 MO alone: 12.07 ± 2.24, p = 0.0002; S.C. MO with PBS: 25.40 ± 3.00, il-10 MO with PBS: 12.44 ± 1.29, p = 0.007). Because the il-10 morphant did not possess significantly fewer pyknotic nuclei at 36 h LT relative to the S.C. morphant (Supplementary Figures S1A, B, D, E), the reduced number of proliferating Müller glia in the il-10 morphant relative to the S.C. morphant was not due to decreased cell death. Additionally, Müller glia proliferation was rescued when IL-10 protein was intravitreally injected every 24 h into the il-10 morphants (Figures 5C, F, G; S.C. MO with IL-10: 21.56 ± 2.45. il-10 MO with IL-10: 27.22 ± 2.91, p > 0.99). This demonstrated that the il-10 morpholino specifically knocked down IL-10 expression, which is required for Müller glia proliferation following light damage.

To determine if IL-10 is sufficient to induce Müller glia proliferation, IL-10 protein was intravitreally injected into undamaged retinas every 24 h and collected at 2dpi and 3dpi (Figures 5H–K). There was no significant difference in the number of PCNA⁺ Müller glia between the PBS and IL-10 protein injected groups in either the INL or ONL at 2dpi (Figures 5L, M; PBS INL: 0.75 ± 0.25, IL-10 INL: 0.63 ± 0.38, p > 0.99; PBS ONL: 5.88 ± 0.97, IL-10 ONL: 5.38 ± 0.63, p > 0.99) and 3dpi (Figures 5L, M; PBS INL: 0.25 ± 0.16, IL-10 INL: 0.50 ± 0.17, p > 0.99; PBS ONL: 5.00 ± 1.10, IL-10 ONL: 7.50 ± 0.92, p > 0.99). There were also no significant difference in the number of Lcp1⁺ cells relative to the PBS controls in either the INL or ONL at 2dpi (Figures 5N, O; PBS INL: 5.00 ± 0.78, IL-10 INL: 4.88 ± 0.58, p > 0.99; PBS ONL: 2.88 ± 0.77, IL-10 ONL: 4.88 ± 1.04, p > 0.99) and at 3dpi (Figures 5N, O; PBS INL: 5.88 ± 0.58, IL-10 INL: 5.70 ± 0.82, p > 0.99; PBS ONL: 5.50 ± 0.68, IL-10 ONL: 5.20 ± 0.53, p > 0.99). Thus, IL-10 is necessary, but not sufficient to induce Müller glia proliferation.

Upon closer examination, undamaged retinas injected with IL-10 possessed Müller glia that displayed a gliotic morphology after 3dpi, with hypertrophied soma and processes (Figures 5I’, K’, arrows and arrowhead, respectively). To confirm that these Müller glia were gliotic, we performed qRT-PCR on various gliotic markers including gfap, yap1, glutamine synthase (glula/b), mesencephalic astrocyte-derived neurotrophic factor (manf), vimentin (vim), nestin (nes), and aquaporin 4 (aqp4) (Figure 5P). The gfap, glulb, vim, and nes gliotic marker genes were significantly upregulated in expression in the IL-10 injected retinas relative to the PBS (vehicle) control group after 3dpi (Figure 5P; gfap: 0.85 ± 0.30, p = 0.021; yap1: 0.35 ± 0.15, p > 0.99; glula: 0.43 ± 0.15, p = 0.73; glulb: 1.10 ± 0.25, p = 0.0026; manf: 0.49 ± 0.34, p = 0.46; vim: 0.81 ± 0.33, p = 0.030; nes: 1.3 ± 0.55, p = 0.0002; aqp4: 0.58 ± 0.25, p = 0.22). Thus, IL-10 is necessary for Müller glia proliferation in damaged retinas, while injection of IL-10 in undamaged retinas stimulated the Müller glia to exhibit a gliotic phenotype.

We demonstrated that il-1β expression is induced before il-10 and both genes are required for Müller glia proliferation in light-damaged retinas. Additionally, injection of IL-1β into undamaged retinas induced il-10 expression and stimulated Müller glia proliferation. However, injection of IL-10 into the undamaged retina did not induce Müller glia proliferation. Thus, IL-1β may be required to induce il-10 expression in the undamaged retina in order to stimulate Müller glia and/or NPC proliferation. To test this hypothesis, we injected IL-1β protein into undamaged retinas with and without the il-10 morpholino. The albino;Tg(gfap:EGFP) ^nt11 zebrafish were electroporated with il-10 morpholino at 0dpi and then injected with recombinant zebrafish IL-1β protein every 24 h and collected at 3dpi (Figures 6B–D). These were compared to undamaged albino;Tg(gfap:EGFP) ^nt11 zebrafish that were injected with only recombinant zebrafish IL-1β protein every 24 h and collected at 3dpi (Figure 6A). The albino;Tg(gfap:EGFP) ^nt11 fish that were only injected with IL-1β possessed a large number of PCNA⁺ Müller glia and NPCs at 3dpi in both the INL and ONL (Figures 6A, E, F; INL: 56.44 ± 7.52, ONL: 58 ± 9.65). The il-10 morphants alone and the il-10 morphants injected with PBS (vehicle) control possessed very low numbers of PCNA⁺ Müller glia and NPCs in the INL (Figures 6B, C, E; il-10 MO:1.78 ± 0.68; il-10 MO with PBS: 5.40 ± 1.07) and ONL (Figures 6B, C, F; il-10 MO: 11.56 ± 2.19, il-10 MO with PBS: 11.40 ± 1.27). However, il-10 morphants injected with IL-1β protein possessed similar numbers of PCNA⁺ Müller glia and NPCs to zebrafish injected with only IL-1β protein in the INL (Figures 6A, D, E; IL-1β: 56.44 ± 7.52, il-10 MO with IL-1β: 64.73 ± 10.50, p > 0.99), but not in the ONL (Figures 6A, D, F; IL-1β: 58.33 ± 9.65, il-10 MO with IL-1β: 27.36 ± 5.92, p = 0.0027). This demonstrates that IL-1β induces Müller glia proliferation independent of IL-10 expression in the undamaged retina.