The microRNA miR-18a Links Proliferation and Inammation During Photoreceptor Regeneration in the Injured Zebrash Retina

In mammals, photoreceptor loss causes permanent blindness, but in zebrash (Danio rerio), Müller glia function as intrinsic stem cells, producing progenitor cells that regenerate photoreceptors and restore vision. MicroRNAs (miRNAs) critically regulate neurogenesis in the brain and retina, but the roles of miRNAs in injury-induced neuronal regeneration are largely unknown. The miRNA miR-18a regulates photoreceptor differentiation in the embryonic retina. The purpose of the current study was to determine the function of miR-18a during injury-induced photoreceptor regeneration. RT-qPCR, in-situ hybridization (ISH) and immunohistochemistry (IHC) showed that miR-18a expression increases throughout the retina by 1-day post-injury (dpi) and continues to increase through 5 dpi. Bromodeoxyuridine (BrdU) labeling showed that at 7 and 10 dpi, when regenerated photoreceptors are normally differentiating, there are more proliferating Müller glia-derived progenitors in homozygous miR-18a mutant (miR-18a mi5012 ) retinas compared with wild type (WT), indicating that miR-18a negatively regulates injury-induced proliferation. At 7 and 10 dpi, miR-18a mi5012 retinas have fewer mature photoreceptors than WT, but there is no difference at 14 dpi, revealing that photoreceptor regeneration is delayed. BrdU labeling showed that the excess progenitors in miR-18a mi5012 retinas migrate to other retinal layers besides the photoreceptor layer. Inammation is critical for photoreceptor regeneration and RT-qPCR showed that, in the absence of miR-18a, inammation is prolonged. Suppressing inammation with dexamethasone rescues the miR-18a mi5012 phenotype. Together, these data show that during injury-induced photoreceptor regeneration, miR-18a regulates proliferation and photoreceptor regeneration by regulating key aspects of the inammatory response during photoreceptor regeneration in zebrash. mature rods does not differ (miR-18a mi5012 98.5 ± 38.6 SD, WT 127.1 ± 72.1 SD cells/0.3 mm, p = 0.51). Hoechst labeling of cone nuclei at 7 dpi also showed that miR-18a mi5012 retinas have fewer cone nuclei than WT (miR-18a mi5012 86.8 ± 9.2 SD, WT 101.3 ± 5.2 SD cells/0.3 mm, p = 0.033). Together, these data indicate that both cone photoreceptor maturation and regeneration are delayed. At 10 dpi, miR-18a retinas have fewer mature cones and rods than WT (cones miR-18a mi5012 76.4 ± 2.5 SD, WT 107.6 ± 6.5 SD cells/0.3 mm, p = 0.001; rods miR-18a mi5012 146.4 ± 46.7 SD, WT 275.6 ± 32.5 SD cells/0.3 mm, p = 0.017) but, at 14 dpi, the number of mature photoreceptors does not differ (cones miR-18a mi5012 125.7 ± 11.4 SD, WT 120.3 ± 14.7 SD cells/0.3 mm, p = 0.462; rods miR-18a mi5012 237.7 ± 28.8 SD, WT 261.3 ± 32.1 SD cells/0.3 mm, p = 0.241) (Fig. 3c, d). Taken together, these data show that in miR-18a mi5012 retinas compared with WT, the same overall numbers of photoreceptors are regenerated by 14 dpi, but photoreceptor regeneneration and maturation are delayed.


Introduction
Photoreceptor loss in the human retina causes permanent blindness, but in the injured retina of zebra sh, Danio rerio, Müller glia (MG) are reprogrammed and divide, producing neuronal progenitors that fully regenerate the lost photoreceptors [1][2][3][4]. This regenerative capacity has established the zebra sh retina as an outstanding model for investigating photoreceptor degeneration and regeneration [5]. Mechanisms identi ed in the zebra sh retina that govern the reprogramming of Müller glia and neuronal regeneration have been used to develop methods to stimulate a limited MG-derived regeneration response in the mouse [6][7][8]. Understanding the mechanisms that govern photoreceptor regeneration in zebra sh could, therefore, be critical for developing regenerative therapies to treat human blindness.
Recent research has improved our understanding of the molecular pathways that regulate neuronal regeneration in the zebra sh retina [reviewed in 4,3,9,10]. The primary focus of this research has been on identifying the transcriptional mechanisms involved in neuronal regeneration, but recent studies show that post-transcriptional regulation by non-coding RNAs, and speci cally microRNAs (miRNAs), also play critical roles in retinal neuronal regeneration [11][12][13][14][15]. Of the more than 2600 mature miRNAs coded by the vertebrate genome [16], the functional roles of a very small number have been established in retinal regeneration. MicroRNAs are likely to play key roles in regulating photoreceptor regeneration, and additional studies that determine the roles of miRNAs in photoreceptor regeneration are critically needed.
In ammation occurs in response to tissue injury, and several recent studies show that activation of neuroin ammatory pathways is both necessary and su cient to initiate neuronal (including photoreceptor) regeneration in the zebra sh retina [17][18][19][20][21][22] [reviewed in 23,24]. In ammatory signals seem to have the opposite effect in the injured mammalian retina, in which removal of microglia reduces in ammatory gene expression and increases the number of regenerated neurons [25]. Understanding mechanisms that control in ammation in the injured retina may, therefore, be critical to unlocking the regenerative potential in the mammalian retina. Importantly, several miRNAs have been identi ed as key regulators of in ammatory pathways [26] and as biomarkers of in ammatory disease [27], and identifying the roles of miRNAs in the injured retina could be critical for understanding the link between the retinal in ammatory response and neuronal regeneration.
MicroRNAs are small, 18-25 nucleotides long, non-coding RNAs that generally function by binding to the 3' untranslated region of mRNA and inhibit translation and/or promote mRNA degradation [28,29]. The miRNA, miR-18a, was recently found to regulate photoreceptor differentiation in larval zebra sh by suppressing levels of the transcription factor, NeuroD [30]. The role of miR-18a during photoreceptor regeneration in zebra sh is currently unknown. Bioinformatics tools predict that miR-18a can interact directly with mRNAs that encode more than 25 molecules that function in in ammatory pathways (http://www.targetscan.org/ sh_62/), suggesting that miR-18a might be an important regulator of in ammation. However, miR-18a has not been investigated in the context of in ammatory regulation.
The objective of this research was to determine if miR-18a governs photoreceptor regeneration by regulating in ammation. qPCR showed that miR-18a expression increases between 3 and 5 days post injury (dpi) and then decreases by 7 dpi. In situ hybridization for miR-18a, combined with green uorescent protein (GFP) immunolabeling in transgenic tg(gfap:egfp mi2002 ) sh [31] with GFP-labeled Müller glia showed that the miR-18a expression increases throughout the retina including in Müller glia (MG) by 1 dpi and in MG-derived progenitors by 3 dpi. 5-Bromo-20-Deoxyuridine (BrdU) labeling revealed that relative to wild-type (WT) animals, at 7 and 10 dpi, homozygous miR-18a mutants (miR-18a mi5012 ) have signi cantly more proliferating MG-derived progenitors. The regeneration of rods and cones is initially delayed in miR-18a mi5012 sh, but numerically matches wild type animals by 14dpi. There was no overproduction of regenerated photoreceptors. BrdU labeling showed that in miR-18a mi5012 retinas, more Müller glia-derived progenitors migrate to layers of the retina in addition to the outer nuclear layer (ONL), suggesting that the excess progenitors differentiate into a variety of other retinal cell types. RT-qPCR and in situ hybridization showed that at 5 and 7 dpi, when in ammation is normally resolving, in miR-18a mi5012 retinas, the expression of genes encoding pro-in ammatory cytokines and the cytokine regulator, nfkb1, are signi cantly elevated. Finally, suppressing in ammation with dexamethasone in miR-18a mi5012 sh fully rescues both the excess proliferation and the delay in the regeneration of cone photoreceptors. Together, these data show that following photoreceptor injury in zebra sh, miR-18a regulates the proliferation of MG-derived progenitors and photoreceptor regeneration by regulating key aspects of in ammation.

Methods
Fish husbandry, photolytic lesions and tissue preparation All sh were maintained at 28.5°C on a 14/10-h light/dark cycle under standard husbandry conditions [32]. AB wild-type (WT) strain zebra sh, purchased from the Zebra sh International Research Center (ZIRC; University of Oregon, Portland, OR, USA), were used for control experiments. The miR-18a mi5012 has a 25 bp insertion in the sequence coding for the primary transcript pri-miR-18a, and homozygous mutant sh, used for all experiments here, do not produce mature miR-18a [30]. The transgenic line, Tg(gfap:egfp) mi2002 [31], expresses Enhanced Green Fluorescent Protein (EGFP) in Müller glia and MGderived progenitors, and was used in conjunction with in situ hybridization to determine the cellular expression of miR-18a.
As previously described, photolytic lesions were used to selectively kill photoreceptors [33]. Brie y, sh were exposed to ultra-high intensity light (> 120,000 lux) in a 100 ml beaker for 30 minutes, using a SOLA SE II 365 White Light Engine (Lumencor, Beaverton, OR, USA). Following photolytic lesions, sh were maintained in normal system water and exposed to the standard 14/10-h light cycle.
Prior to collecting tissues, sh were euthanized in 0.1% Tricaine Methanesulfonate (MS-222) and then decapitated. For histology, eyes were removed, xed (overnight at 4°C) in 4% paraformaldehyde, in ltrated with 20% sucrose (in PBS) and then in a 2:1 mixture of 20% sucrose and optimal cutting temperature (OCT) compound, and then nally embedded and frozen in OCT. Sections, 10µm in thickness, were collected through the center of the eye and thaw mounted onto glass slides. For qPCR, retinas were isolated from two sh per biological replicate (4 retinas), and total RNA, including small RNAs, were puri ed using the miRvana miRNA puri cation kit (Invitrogen, Carlsbad, CA, USA).
Systemic labeling with BrdU, dexamethasone treatment, immunohistochemistry and in situ hybridization Cells in the S-phase of the cell cycle were labeled with BrdU by swimming sh in 5 mM BrdU for 24 hours. Fish were then either sacri ced immediately or at variable timepoints thereafter. For BrdU immunolabeling, sections were incubated in 100°C sodium citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0) for 30 minutes to denature DNA and cooled at room temperature for 20 minutes.
Sections then were subjected to standard immunolabeling as described below.
Fish treated with dexamethasone to inhibit in ammation were exposed in system water to 15 mg/L dexamethasone (Sigma-Aldrich, Corp, D1756) diluted in 0.1% MetOH [21]. Control sh were treated with 0.1% MetOH only. Fish were treated between 2-6 dpi. All solutions were changed daily, and sh were fed brine shrimp every other day.
Standard immunolabeling was performed using previously published protocols [34]. The primary and secondary antibodies and dilutions used here were: mouse anti- In situ hybridizations were performed using previously published protocols [35,36]. Riboprobes were generated from PCR products using the following primers and by adding a T3 polymerase sequence on the reverse primer (lowercase letters) [37]. Probes were generated for rods rhodopsin (F-GAGGGACCGGCATTCTACGTG, R-aattaaccctcactaaagggCTTCGAAGGGGTTCTTGCCGC) and cones arr3a (F-GAAGACCAGTGGAAATGGCCAG, R-aattaaccctcactaaagggTCAGAGGCAGCTCTACTGTCAC). In situ hybridization for mature miR-18a was performed using a miRCURY LNA detection probe (Exiqon/Qiagen, Germantown, MD), labeled with DIG at the 5′ and 3′ends. Standard in situ hybridization methods were used for miR-18a, as described above, but using a 0.25 µM probe working concentration at a hybridization temperature of 58°C. For comparisons of relative expression across post-injury time points or between WT and miR-18a mi5012 sh, all tissue sections were placed on the same slides and/or developed for identical periods of time.
Reverse transcriptase quantitative real-time PCR (RT-qPCR) As described above, for RT-qPCR, retinas were removed and total RNA puri ed from the eyes of 2 sh ( Student's t-tests were used to determine statistical signi cance (p values less than 0.05 were considered statistically sign cant).

Results
Following photoreceptor injury, miR-18a is expressed in both the inner and outer nuclear layers, including in Müller glia and Müller glia-derived progenitors RT-qPCR and in situ hybridization were performed to determine the temporal and spatial expression of miR-18a following photoreceptor injury. Taqman qPCR showed that, compared with control uninjured retinas, miR-18a expression is signi cantly higher at 3 days (2.15 ± 0.44 SD fold higher, p = 0.003), 5 days (2.06 ± 0.69 SD fold higher, p = 0.0.024) and 7 days (1.54 ± 0.30 SD fold higher, p = 0.045) post-retinal injury (dpi) (Fig. 1a). In situ hybridization using an LNA ribroprobe for mature miR-18a, combined with immunolabeling for green uorescent protein (GFP) in Tg(gfap:egfp) mi2002 sh, showed that at in control uninjured retinas, there is only very slight expression of miR-18a ( Fig. 1b) but by 24 hpi, around the time that Müller glia divide once to produce a neuronal progenitor [see 39], miR-18a is expressed in many cells throughout the inner nuclear layer (INL), including Müller glia (Fig. 1c). Then at 3 dpi, during the peak of progenitor proliferation, miR-18a remains strongly expressed in the INL and ONL, including in Müller glia and MG-derived progenitors (Fig. 1d). Finally, by 7 dpi, when MG-derived progenitors are normally exiting the cell cycle, miR-18a expression is again similar to the expression pattern seen between 0 and 24 hpi (Fig. 1e, f). Together, these results show that miR-18a expression is upregulated in the retina during the time periods when Müller glia and then MG-derived progenitors divide.
During photoreceptor regeneration, miR-18a regulates proliferation among MG-derived progenitors To determine if miR-18a regulates cell proliferation among MG-derived progenitors, BrdU labeling and immunostaining were used to quantitatively compare the number of proliferating cells in WT and miR-18a mi5012 retinas. In the injured WT retina, progenitor proliferation peaks around 3 dpi, some MG-derived progenitors normally stop dividing between 4 and 5 dpi, and the rst regenerated photoreceptors can be detected between 5 and 6 dpi [see 40]. By 7 dpi, many new photoreceptors have normally regenerated and fewer progenitors continue to proliferate and, by 10 dpi, very few progenitors normally continue to proliferate. Compared with WT retinas, the number of proliferating progenitors in miR-18a mi5012 retinas did not differ at 3 dpi (miR-18a mi5012 81.7 ± 7.2 SD, WT 87.6 ± 8.1 SD cells/0.3 mm, p = 0.60) (Fig. 2a, b), indicating that the initial proliferative response is unaltered in miR-18a mi5012 retinas. At 7 dpi, however, there were signi cantly more BrdU-positive cells in the miR-18a mi5012 retinas than in WT (miR-18a mi5012 64.3 ± 9.8 SD, WT 30.6 ± 7.2 SD cells/0.3 mm, p = 0.001) (Fig. 2c, d). In miR-18a mi5012 retinas, there were also signi cantly more proliferating cells at 10 dpi (miR-18a mi5012 45.7 ± 3.2 SD, WT 14.0 ± 3.6 SD cells/0.3 mm, p < 0.001) (Fig. 2e, f). These data show that following photoreceptor injury, in the absence of miR-18a, the timing of the initial proliferative response is unchanged, but that MG-derived progenitors continue to proliferate longer than in WT retinas.
During photoreceptor regeneration, miR-18a regulates the timing, but not the extent, of photoreceptor regeneration In miR-18a mi5012 retinas, the prolonged period of proliferation among MG-derived progenitors suggests that photoreceptor regeneration might be delayed and/or that more photoreceptors might be produced.
To determine if miR-18a regulates the timing and/or extent of photoreceptor regeneration, in situ hybridization was used to differentially label mature cones and rods in the retina at 7 dpi, when large numbers of newly differentiated photoreceptors can rst be detected, at 10 dpi, when most new photoreceptors have been normally regenerated, and 14 dpi, when photoreceptor regeneration is normally complete. In situ hybridization for mature cones (arr3a) and rods (rho) (Fig. 3a, b) and quanti cation of these cells (Fig. 3c, (Fig. 3c, d). Taken together, these data show that in miR-18a mi5012 retinas compared with WT, the same overall numbers of photoreceptors are regenerated by 14 dpi, but photoreceptor regeneneration and maturation are delayed.
Following photoreceptor injury, miR-18a regulates the number of neuronal progenitors that are produced In miR-18a mi5012 retinas compared with WT, MG-derived progenitors proliferate for a longer period of time, suggesting that excess neuronal progenitors are produced. However, since extra photoreceptors are not generated, these excess progenitors might either die or migrate to other retinal layers, possibly differentiating into other cell types. To investigate this, TUNEL labeling was used to label dying cells at 10 dpi, when the largest differences in photoreceptor numbers were identi ed between WT and miR-18a mi5012 retinas. This experiment showed that, at 10 dpi, there were no differences in the number of TUNEL-positive cells between WT and miR-18a mi5012 retinas (data not shown), indicating that excess progenitors in miR-18a mi5012 retinas are not eliminated by cell death. To establish the potential fates of the excess MG-derived progenitors in miR-18a mi5012 retinas, sh were exposed to 5 mM BrdU from 3 to 4 dpi during the peak of cell proliferation, and then sh were sacri ced at 14 dpi when photoreceptor regeneration is complete. Immunolabeling for BrdU at 14 dpi shows that compared with WT, miR-18a mi5012 retinas have more BrdU + cells in both the INL (miR-18a mi5012 22.7 ± 4.3 SD, WT 11.0 ± 3.2 SD cells/200 mm, p = 0.019) and ganglion cell layer (GCL) (miR-18a mi5012 15.1 ± 2.2 SD, WT 6.2 ± 1.9 SD cells/200 mm, p = 0.006) (Fig. 4), indicating that the excess progenitors produced at 3-4 dpi either remain in the INL or migrate to the GCL, suggesting that they may differentiate into a variety of other cell types.
Following photoreceptor injury, miR-18a regulates the extent and duration of the in ammatory response The effect of miR-18a on the cell cycle during photoreceptor regeneration is strikingly different from embryonic development, in which miR-18a regulates photoreceptor differentiation but not cell proliferation [30]. This indicates that in the injured retina, miR-18a regulates pathways that are speci c to the post-injury response. Silva et al. [21] showed that in injured mmp9 mutant retinas, there was increased in ammation resulting in excess proliferation among MG-derived progenitors. This nding helped lead to the rationale that increased or prolonged in ammation might cause the excess proliferation observed in miR-18a mi5012 retinas. Indeed, the miRNA target database TargetscanFish (http://www.targetscan.org/ sh_62) predicts that miR-18a interacts with mRNA for many molecules involved in in ammatory pathways. To determine if miR-18a regulates in ammation, RT-qPCR was used to compare the mRNA expression levels of key in ammatory molecules at 1, 3, 5 and 7 dpi in WT and miR-18a mi5012 retinas. These data show that at 1 dpi, around the time that Müller glia normally begin to divide, expression of tnfα is higher in miR-18a mi5012 retinas compared with WT (miR-18a mi5012 4.51 ± 1.14 SD, WT 2.31 ± 0.99 SD, p = 0.033) (Fig. 5a). Then at 3 dpi, around the normal peak of MG-derived progenitor proliferation, il1β expression is higher in miR-18a mi5012 retinas compared with WT (miR-18a mi5012 1.61 ± 0.21 SD, WT 0.97 ± 0.06 SD, p = 0.007) (Fig. 5b). At 5 dpi, when many photoreceptor progenitors normally stop proliferating and begin to differentiate, expression of il6, il1β, and the cytokine regulator Nuclear Factor Kappa B 1 (nfkb1) are higher in miR-18a mi5012 retinas compared with WT (il6 miR-18a mi5012 1.32 ± 0.26 SD, WT 0.85 ± 0.25 SD, p = 0.042; il1b miR-18a mi5012 1.71 ± 0.47 SD, WT 0.97 ± 0.08 SD, p = 0.028; nfkb1 miR-18a mi5012 1.68 ± 0.23 SD, WT 1.19 ± 0.17 SD, p = 0.021) (Fig. 5c). Finally, at 7 dpi, when many photoreceptor progenitors have normally stopped proliferating and have fully differentiated, nfkb1 expression is still higher in miR-18a mi5012 retinas (miR-18a mi5012 1.33 ± 0.22 SD, WT 1.03 ± 0.07 SD, p = 0.044) (Fig. 5d). Taken together, these results indicate that, compared with WT, there is both a higher level of in ammatory pathway activity and a prolonged in ammatory response in miR-18a mi5012 retinas following photoreceptor injury and death.
Supressing in ammation in miR-18a mutants rescues both the excess proliferation and delayed photoreceptor regeneration The increased and prolonged expression of in ammatory genes in the injured miR-18a mi5012 retina led to the hypothesis that, in the absence of miR-18a, prolonged and excessive in ammation is causally related to the excess proliferation and delayed photoreceptor maturation observed in miR-18a mutants. To test this hypothesis, dexamethasone was used to suppress in ammation in WT and miR-18a mi5012 sh from 2 to 6 dpi, the time range during which miR-18a expression is upregulated in injured retinas (see Fig. 1a, b) and when in ammatory molecules are expressed higher than WT in miR-18a mi5012 retinas (see Fig. 5). Fish were then exposed to 5 mM BrdU from 6 to 7 dpi, to label proliferating cells. BrdU immunolabeling was used to label BrdU + cells in S-phase of the cell cycle, and in situ hybridization was used to label mature rod or cone photoreceptors. Compared with controls, the dexamethasone treatment fully rescued the excess proliferation in miR-18a mi5012 retinas, reducing the number of BrdU + cells to that observed in  (Fig. 6c, d). Together, these data indicate that in the injured retina, miR-18a regulates MGderived progenitor proliferation and photoreceptor regeneration by regulating in ammation.

Discussion
Recent studies have dramatically improved our understanding of the mechanisms that govern neuronal regeneration in the injured zebra sh retina from dedifferentiated Müller glia [reviewed in 4,41,10]. Some of this information has led to studies showing that mammalian Müller glia possess latent regenerative potential that can be augmented by reprogramming these cells, and that MG-derived progenitors can regenerate some neurons (including photoreceptors) [7,25,8,[42][43][44], but this neuronal regeneration is very ine cient [9,41]. In the injured zebra sh retina, in ammation is necessary and su cient for neuronal regeneration to begin [45,17,18,46,20,19,23,21,22], but in the injured mammalian retina, in ammation prevents Müller glia from initiating a regenerative response [25]. When in ammation is misregulated in the injured zebra sh retina, this leads to aberrant proliferation of MG-derived progenitors and alters photoreceptor regeneration [21,20], showing the importance of precise in ammatory regulation in controlling these events.
Several miRNAs have been identi ed as important regulators of in ammation and could be key neuroin ammatory regulators in the injured retina [26,27], but studies linking miRNAs with retinal in ammation are lacking. The miRNA miR-18a was recently identi ed as an important regulator of photoceptor differentiation in the developing embryonic retina [30], and miR-18a is predicted to interact with mRNAs of more than 25 in ammation-related molecules (http://www.targetscan.org/ sh_62), suggesting that miR-18a could regulate neuroin ammation in the retina, but the roles of miR-18a in injury-induced in ammation and photoreceptor regeneration had not been previously investigated.
The objective of the current study was to determine the role of miR-18a in photoreceptor regeneration and to determine if it regulates in ammation during this response. Following photoreceptor injury, miR-18a is expressed in the INL and ONL, including in dividing Müller glia as early as 1 dpi and in both Müller glia and proliferating MG-derived progenitors at 3-5 dpi, indicating that miR-18a functions during key times of cell division during photoreceptor regeneration. In miR-18a mi5012 retinas compared with WT, more cells continue to proliferate at 7 and 10 dpi, when most photoreceptor progenitors have normally exited the cell cycle, indicating that miR-18a regulates proliferation of MG-derived progenitors. This differs from ndings in the developing embryonic retina in which miR-18a regulates the timing of photoreceptor differentiation but does not regulate cell proliferation [30]. The function of miR-18a to limit progenitor proliferation also differs from the general function of miR-18a in cancer, in which it typically promotes the proliferation of tumor cells [47,48] including glioblastoma cells [49]. Also, in the developing mouse neocortex, the miR-17-92 cluster, which includes miR-18a, has been shown to promote proliferation of neuronal progenitors [50]. Interestingly, however, miR-18a has also been shown to suppress cell proliferation in other contexts such in pancreatic progenitor cells [47] and myoblast cells [51], and even has anti-proliferation/anti-tumor effects in colorectal and breast cancers [52,48,53]. The effects of miR-18a on the cell proliferation are, therefore, dependent on the cell type and injury/disease state of the tissue involved.
In miR-18a mi5012 retinas, the prolonged proliferation among neuronal progenitors results in a delay in photoreceptor regeneration and maturation, and causes excess neuronal progenitors to be produced. At least some of these excess progenitors remain in the INL or migrate to the GCL, suggesting that they may differentiate into other types of neurons besides photoreceptors. This is consistent with data showing that following damage to retinal neurons, Müller glia derived progenitors are multipotent and generate multiple cell types even though they preferentially generate the neurons that were lost [54][55][56]. The presence of additional MG-derived cells in the INL and GCL could also be due to aberrant migration of progenitor cells. Chemokines, produced in response to in ammatory cytokines, guide the migration of neural progenitors [57]. While it is unknown if miR-18a regulates chemokines directly, the altered cytokine expression observed in miR-18a mi5012 retinas could possibly lead to altered chemokine expression and aberrant progenitor migration, resulting in more progenitors failing to migrate to the ONL.
The excess proliferation of MG-derived progenitors and higher expression of key in ammatory molecules in miR-18a mi5012 retinas compared with WT is similar to what is observed in mmp9 mutant retinas, but there are also some important differences. First, the excess progenitors in miR-18a mi5012 retinas do not generate excess photoreceptors, and this differs from what is observed in mmp9 mutant retinas, in which excess progenitors do generate excess photoreceptors [21]. This difference indicates that miR-18a and Mmp9 may also differentially regulate molecules downstream of pathways that regulate in ammation and cell proliferation. In the subventricular zone of the CNS, Mmp9 has been found to regulate differentiation of neuronal progenitor cells [58], indicating that it has some capacity to regulate neuronal differentiation. As a microRNA, miR-18a likely regulates several molecules involved in different steps of the regeneration response, and members of the miR-17-92 cluster of miRNAs, which includes miR-18a, are key regulators of neurogenesis, involved in cell proliferation, progenitor fate determination and differentiation [reviewed in 59]. Previous work found that NeuroD governs the cell cycle and differentiation among photoreceptor progenitors during both embryonic development and regeneration [60,40] and that miR-18a regulates NeuroD protein levels in the embryonic retina [30]. Downstream of in ammatory pathways, miR-18a is also likely to regulate NeuroD in the injured retina, and this could affect differentiation of photoreceptors and other cells. The dominant phenotype in injured adult miR-18a mi5012 retinas, however, is the increased in ammation and cell proliferation, and pathways downstream of in ammation have not yet been investigated.
A second key difference between the miR-18a mi5012 and mmp9 mutant retinal phenotypes is that, although both mutations result in an elevated in ammatory response following photoreceptor injury, they result in higher expression of different in ammatory molecules and at different time points. In miR-18a mi5012 retinas, tnfα expression is increased only at 1 dpi, differing markedly from observations in mmp9 mutant retinas that have increased expression of tnfα (but not other in ammatory molecules) at all post-injury time points [21]. Also in contrast to mmp9 mutants, miR-18a mi5012 retinas have increased expression of other cytokines (il6 and il1b) and the cytokine regulator nfkb1 at 5 dpi, a time point when in ammation is typically subsiding and progenitors are beginning to differentiate. Further, nfkb1 expression continues to be higher in miR-18a mi5012 retinas at 7dpi, when widespread differentiation is typically occurring among photoreceptor progenitors. These data indicate that, like Mmp9, miR-18a negatively regulates in ammation, leading to a converging phenotype (excess neuronal progenitors).
The prolonged expression of important in ammatory molecules in miR-18a mi5012 retinas during photoreceptor regeneration indicates that miR-18a functions during the resolution phase of the in ammatory response, potentially regulating a negative feedback loop that resolves in ammation and restores homeostasis. Resolving neuroin ammation promotes normal tissue repair and if acute in ammation remains unresolved, it can result in inadequate repair or further neuronal damage [61]. Following photoreceptor injury, expression levels of nfκb1 and certain cytokines are normally upregulated and peak between 24 and 48 hpi and then their expression levels decrease, returning to homeostasis sometime after 7 dpi [see 21]. The removal of NFκB activity and in ammatory cytokine signaling are critical steps in resolving in ammation [62], indicating that the time frame between 48 hpi and 7 dpi in the retina is key for regulating this process. In miR-18a mi5012 retinas, the higher expression of nfκb1 and in ammatory cytokines at 5 dpi compared with WT, and the continued higher expression of nfκb1 at 7 dpi, indicate that resolution of the in ammatory response is delayed and that miR-18a regulates this process. Treatment of miR-18a mi5012 sh with dexamethasone from 48 hpi to 6 dpi was a means to chemically resolve retinal in ammation during the normal time frame and, because this fully rescued the miR-18a mi5012 phenotype (excess proliferation and delayed photoreceptor regeneration), this indicates that miR-18a functions during the resolution phase to regulate key aspects of in ammation. This function is in line with functions of several other miRNAs, including miR-9, miR-21, miR-146 and miR-155, which also play important roles in resolving in ammation by regulating pathways involving macrophages [63]. Importantly, to our knowledge, this is the rst study to show that miR-18a regulates in ammation, and the rst to show that any miRNA regulates in ammation in the injured retina.
In ammation in the retina following photoreceptor death is generated by the release of cytokines from  [17]. An underlying assumption is that these cells are responsible for the elevated and prolonged in ammation in the miR-18a mi5012 retinas, however, given that miR-18a expression is elevated relatively ubiquitously in the injured retina, the increased and prolonged in ammation in the mutants could originate from other cellular types. Müller glia are known to secrete in ammatory cytokines in response to retinal injury [23,17,19] and cytokine signals from microglia and/or dying cells are necessary for Müller glia to divide [17,19].
The early expression of miR-18a in Müller glia and other retinal cells (by 1 dpi) might, therefore, be a mechanism to limit the level of retinal in ammation by regulating the secretion and responses to certain in ammatory cytokines. The highest expression of miR-18a occurs between 3 and 5 dpi, during the resolution phase of retinal in ammation [see 21]. Based on its 7-base seed sequence (CACCUUA), miR-18a is predicted to interact directly with mRNAs of more than 25 molecules that function in in ammatory pathways (http://www.targetscan.org/ sh_62), some of which are cytokines and other intercellular signaling molecules (e.g. 1l-16, cxcl12a, bmp6), but many of which are transmembrane receptors or molecules that function downstream of in ammatory cytokines (e.g. Il-7r, cxcr4a, tnfaip3). It is therefore probable that miR-18a not only reduces in ammatory signals from Müller glia and other cells (e.g. microglia, dying photoreceptors), but also reduces the effects of those in ammatory signals on the MGderived progenitors.
In conclusion, this study is the rst to show that a miRNA regulates in ammation in the injured retina and is the rst to show that miR-18a is an important in ammatory regulator. This work adds to the growing body of knowledge that miRNAs are key regulators of neurogenesis throughout the central nervous system [67] and neuronal regeneration in the retina [15]. Importantly, the differential responses of Müller glia to in ammation following retinal injury in zebra sh compared with mammals could be key to unlocking the potential for mammalian Müller glia to robustly regenerate neurons. Like miR-18a, other miRNAs could be potent in ammatory regulators in the injured retina and may be key to fully augmenting the regenerative potential of the mammalian retina.