Neurotrophic Factors Cause Activation of Intracellular Signaling Pathways in Müller Cells and Other Cells of the Inner Retina, but Not Photoreceptors

Abstract

purpose. Intravitreal injection of brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF), or basic fibroblast growth factor (FGF2) promotes survival of photoreceptors exposed to various types of insults, but it is not known if these survival-promoting effects occur by direct action of the factors on photoreceptors or indirectly through the activation of other cells. In this study, the authors have sought to address this issue by determining which cells in the retina show evidence of activated intracellular signaling pathways acutely and at longer time points after intravitreal injection of these agents.

methods. Retinas were removed from C57BL/6J mice at 1, 6, or 24 hours after intravitreal injection of 1 μg of human BDNF, rat CNTF, human FGF2, or human transforming growth factor-α (TGFα), and immunohistochemically stained for phosphorylated extracellular signal–regulated kinase (pERK), phosphorylated cAMP responsive element binding protein (pCREB), or c-fos. Retinal organ cultures were incubated with 10 ng/ml of BDNF, CNTF, FGF2, or TGFα for 10 or 30 minutes or 1, 3, or 6 hours and then immunohistochemically stained for pERK, pCREB, or c-fos.

results. Intravitreal injection of BDNF, CNTF, or FGF2 resulted in a rapid increase in pERK immunoreactivity in Müller cells and a rapid increase in c-fos immunoreactivity in Müller, amacrine, and ganglion cells. Immunoreactivity for pERK and c-fos returned to baseline in all retinal cells at 6 or 24 hours after injection, but there was increased staining for glial fibrillary acidic protein (GFAP) in Müller cells at these time points. At no time after injection was there any staining for pERK or c-fos in photoreceptors. Similarly, retinal explants treated with FGF2, BDNF, or CNTF showed increased staining for pCREB, pERK, and c-fos in cells of the inner retina, but not photoreceptors.

conclusions. These data support the hypothesis that BDNF, CNTF, and FGF2 exert their effects on photoreceptors by acting indirectly through activation of Müller cells and perhaps other nonphotoreceptor cells.

Essentially all cells in the body are tenuously balanced between life and death. An insult may kill a cell outright or may induce sublethal damage that renders it useless and/or a threat to surrounding tissue. The latter situation often results in activation of a cell death program that eliminates damaged cells. The threshold for how much damage is tolerated and repair undertaken is modulated in part by survival factors (also called trophic factors). In addition, complete withdrawal of a critical trophic factor can result in cell death in the absence of damage, a mechanism used by the body to deal with overproduction of cells during development. 

Intraocular administration of neurotrophic factors, including brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF), or basic fibroblast growth factor (FGF2), significantly delays cell death in a number of models of photoreceptor degeneration. ^{1 2 3 4 5} These observations have raised the hope that neurotrophic factors may offer a therapeutic approach to currently untreatable retinal diseases, such as macular degeneration and retinitis pigmentosa. Extrapolation of these exciting experimental results to the treatment of patients, however, is hindered by our incomplete understanding of the mechanisms through which photoreceptors are rescued by neurotrophic factors and our lack of knowledge regarding side effects. It is not known whether neurotrophic factors act directly on photoreceptors or indirectly through activation of other cell types. Some neurotrophic factors may recruit inflammatory cells and/or stimulate the proliferation of nonneuronal cells, a source of potential undesirable side effects in the retina. ² It would clearly be helpful to determine which cells neurotrophic factors activate and how they exert their survival-promoting effects. 

One way to determine the site of action of neurotrophic factors is to take advantage of what is known about the molecular mechanisms through which they trigger cellular responses; they bind to specific cell surface receptors and stimulate receptor autophosphorylation and phosphorylation of downstream signal transduction molecules such as mitogen activated protein kinase (MAPK) and cAMP responsive element binding protein (CREB). ^{6 7 8} Activated cells also show rapid upregulation of the immediate early genes c-fos and c-jun. ⁹ The availability of antibodies directed against immediate early gene products and the phosphorylated forms of several signaling molecules makes it possible to detect immunohistochemically the cells that are immediately activated after exposure to a factor; these responses occur in minutes to hours, as opposed to survival-promoting effects that can only be detected several days, or even weeks after the onset of treatment. 

We have used an immunohistochemical approach to investigate the site of action of human BDNF, rat CNTF, and human FGF2, three neurotrophic factors known to promote the survival of photoreceptors. ^{2 3 10} These factors were tested in vivo and in organ cultures. Analysis with antibodies directed against c-fos, pERK or pCREB have shown that BDNF, CNTF, FGF2 and TGFα activate Müller cells, and in some cases inner retinal neurons, but photoreceptor activation was not observed with any of the factors. These results are consistent with the hypothesis that protection of photoreceptors by neurotrophic factors occurs indirectly through activation of other cells. 

Methods

Mice

All experimental procedures were designed to conform with the ARVO Statement for the Use of Animals in Ophthalmic and Visual Research. C57BL/6J^+/+ mice were obtained from the Jackson Laboratory (Bar Harbor, ME). Breeders were maintained in a 14-hour light/10-hour dark cycle. One- to 3-month-old C57 BL/6J mice were used for organ cultures and intravitreal injections. 

Intravitreal Injections

Mouse anesthesia was induced and maintained with inhaled methoxyflurane and topical 0.5% proparacaine HCl (Alcon, Humacao, Puerto Rico). Intravitreal injections were performed with needles pulled from borosilicate glass capillary tubes (cat. no. TW100-4; World Precision Instruments, Sarasota, FL), beveled to an approximate bore size of 10 to 20 μm (25° bevel angle) with a model BV-10 K. T. Brown Type micropipette beveler (Sutter Instrument Co., Novato, CA), with a fine diamond abrasive plate. Micropipettes were siliconized with hexamethyl-disilazane (Sigma, St. Louis, MO) for 7 days in a vacuum-tight desiccator and rinsed several times with sterile water. Under stereomicroscopic visualization, siliconized glass micropipettes were inserted just posterior to the superior limbus, and 1 or 2 μl of vehicle with or without factor were injected into the vitreous cavity using a nitrogen-pressurized picoinjector (Harvard Apparatus, South Natick, MA). To visualize the injection site, blue 0.8-μm diameter latex beads (Sigma) were added to the injected solution in some initial experiments. Injected factors (all from R&D systems, Minneapolis, MN) were recombinant human BDNF, recombinant rat CNTF (CNTF), recombinant human FGF2, and recombinant human transforming growth factor-α (TGFα). Factors were diluted to 1 μg/μl in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA). Xylazine HCl (Bayer, Shawnee Mission, KS) was also tested after intramuscular injection at a concentration of 20 mg/kg and intravitreally at a concentration of 20 μg/μl (Sigma), based on previous reports that it activates ERK. ¹¹ For each animal, the eye contralateral to the one receiving neurotrophic factors was injected with an equal volume of PBS containing 0.1% BSA. A total of 30 eyes were injected (4 BDNF, 4 FGF2, 3 CNTF, 2 TGFα, 2 xylazine, and 15 PBS) and analyzed 60 minutes after injection. Twenty-four additional eyes injected with 20 μg xylazine, 1 μg CNTF, or PBS were examined after 6 or 24 hours. 

Retinal Organ Cultures

Mice were killed by CO₂ inhalation. Eyes were enucleated and placed in Hanks’ balanced salt solution (HBSS), and retinas were gently dissected away from the retinal pigment epithelium (RPE) and sclera with forceps under a dissecting stereomicroscope. Quadrants of retinas were placed, photoreceptor side down, atop Millicell-CM filter membranes (Millipore, Bedford, MA) that were supported by Teflon rings such that filters were in direct contact with, but not covered by, the medium. The quadrants were grown as organ cultures in 0.5 ml of chemically defined medium (Dulbecco’s modified Eagle’s medium, pH 7.3, supplemented with 1.28 mg/l cytidine 5′-diphosphoethanolamine, 2.56 mg/l cytidine 5′-diphosphocholine, 16.6× 10⁻⁷ M insulin, 4× 10⁻⁸ M progesterone, 2× 10⁻⁴ M putrescine, 6× 10⁻⁸ M selenium, and 12.5× 10⁻⁸ M transferrin). ¹² Cultures were maintained at 37°C in a humidified chamber with 5% CO₂. After 24 hours, the organ cultures were treated with 10 ng/ml BDNF, CNTF, FGF2, or TGFα for 10 or 30 minutes or 1, 3, or 6 hours. Forskolin (100 μM; Calbiochem, La Jolla, CA) was used as a positive control for CREB phosphorylation at these same time points. 

Immunohistochemistry

Mice were euthanatized by methoxyflurane overdose. Eyes were promptly enucleated and anterior chambers were removed with microsurgical scissors before fixation. Eye cups and retinal explants were fixed for 1.5 hours at 4°C in 4% paraformaldehyde in 0.1 M phosphate buffer containing 5% sucrose. This was followed by immersion in increasing concentrations of sucrose (5–20% in phosphate buffer) and overnight incubation in 20% sucrose solution at 4°C. The tissue was frozen on dry ice with isopentane in a 2:1 ratio of OCT and 20% sucrose and stored at −80°C. Retinal sections were cut to a thickness of 7 μm, thaw mounted onto Superfrost Plus glass slides (Fisher, Pittsburgh, PA), and stored at −20°C until needed. Endogenous peroxidases were inhibited with 0.75% H₂O₂ in PBS, and sections were permeabilized with 0.5% NP-40, blocked in 3% BSA/0.1% Triton X-100 and incubated overnight in primary antibody. 

Antibodies used at a dilution of 1:250 were anti-pCREB ¹³ (provided by David Ginty, Baltimore, MD), a rabbit polyclonal anti-pERK1/2 ¹¹ (New England Biolaboratories, Beverly, MA); anti–c-fos Ab-5 (Oncogene, Cambridge, MA) was used at a dilution of 1:2500. Control sections were processed omitting the primary antibody and, in the case of c-fos, after preadsorption of the antibody at 4°C with 0.1 mg/ml of c-fos peptide (Oncogene). Antibody binding was generally detected with the ABC method (Vector, Burlingame, CA), using diaminobenzidine tetrahydrochloride (DAB·4HCl; Polysciences, Inc., Warrington, PA) as a chromogen. 

Immunofluorescent double-labeling was performed to identify factor-activated cells, using anti-cellular retinaldehyde binding protein (anti-CRALBP, 1:50; provided by Jack Saari, Seattle, WA) to identify Müller glia and RPE, ^{14 15} anti–protein kinase C (anti-PKC, 1:50; Amersham, Piscataway, NJ) to identify rod-specific bipolar cells and cone inner and outer segments, ^{16 17} and anti–calbindin D (1:100; Sigma) to identify horizontal cells and certain subtypes of amacrine and ganglion cells. ^{18 19 20} Additionally, an antibody directed against glial fibrillary acidic protein (anti-GFAP, 1:300; Sigma) was used to identify reactive Müller glia after intravitreal injection of BDNF, FGF2, or CNTF. Primary antibodies were localized with goat anti-mouse IgG labeled with fluorescein isothiocyanate used at a dilution of 1:33 (Cappel, Durham, NC). c-fos, pERK, or pCREB were detected with a 1:100 dilution of biotinylated goat anti-rabbit secondary IgG (Vector) followed by incubation with a 1:100 dilution of rhodamine avidin D (Vector). Double exposures were recorded on film (Ektachrome 400, developed at 800 ASA; Eastmann Kodak, Rochester, NY) using a microscope equipped with epifluorescence (Fluophot; Nikon). 

Results

Intravitreal Injection of BDNF, CNTF, or FGF2: Rapid Increases in pERK Immunoreactivity in Müller Cells, but Not Photoreceptors

Control retinas from uninjected (not shown) or vehicle-injected eyes (Fig. 1A ) showed little or no background immunostaining with anti-pERK, but considerable immunoreactivity could be seen throughout the inner retina 60 minutes after intravitreal injection of BDNF, FGF2, or CNTF. It must be noted that CNTF has been shown to activate not only the JAK–STAT pathway, but also other signaling molecules including ERK-1 and ERK-2 ²¹ ; therefore, phosphorylation of ERK within minutes of CNTF presentation provides a good indicator of which cells are responding to this factor, overcoming limitations derived from difficulties in analyzing STAT activation by immunocytochemistry. Staining predominated in radial processes extending toward the inner and outer limiting membranes from cell bodies located in the middle of the inner nuclear layer. Stained processes adjacent to the vitreous cavity were typical of Müller cell endfeet (Figs. 1B 1C 1D) . These findings, suggesting that many of the immunoreactive cells are Müller glial cells, were confirmed by immunofluorescent double-labeling that showed colocalization of pERK and CRALBP (not shown). No immunoreactivity was seen in other retinal elements. As previously reported, intravitreal or intramuscular injection of theα ₂-adrenergic receptor agonist, xylazine, stimulated ERK phosphorylation in cells of the inner retina (not shown). ¹¹  

Intravitreal Injection of BDNF, CNTF, or FGF2: Rapid Increases in c-fos Immunoreactivity in Müller, Amacrine, and Ganglion Cells, but Not Photoreceptors

Expression of the immediate early gene c-fos was also used to monitor cellular responses to neurotrophic factors. As shown in Figure 2A , untreated and vehicle-injected eyes showed some diffuse, weak staining in ganglion cells, some cells of the inner nuclear layer adjacent to the inner plexiform layer, the processes in the plexiform layers, and in photoreceptor inner segments. No immunoreactivity was observed in nuclei or in specimens incubated with anti–c-fos preadsorbed with c-fos peptide (not shown). Sixty minutes after intraocular injection of BDNF, CNTF, or FGF2, there was a marked increase in staining for c-fos in several cell types located in the inner retina (Figs. 2B 2C 2D 2E) . Immunoreactivity was localized in cell nuclei, as is typically observed in factor-activated cells. ⁷ With all three factors, some of the cells that showed increased immunoreactivity were tentatively identified as Müller cells on the basis of their position and morphology. In eyes injected with BDNF or FGF2, there also were some cells with increased staining for c-fos in the inner nuclear layer adjacent to the inner plexiform layer and in the ganglion cell layer (Figs. 2B 2C 2D) . This pattern was similar to that observed in retinas from eyes injected with TGFα (Fig. 2F) , a factor previously reported to induce c-fos gene expression in Müller cells. ²² Colocalization of increased c-fos immunoreactivity with CRALBP (Fig. 3C ) or calbindin D (Fig. 3A) suggested activation of Müller cells or amacrine and ganglion cells, respectively. As was the case for pERK, photoreceptors failed to show c-fos immunoreactivity in vehicle- or factor-injected eyes. In light of our consistent finding of a lack of photoreceptor immunoreactivity with both anti-pERK and anti–c-fos antibodies, it is worth noting that photoreceptors have been reported to express c-fos under other circumstances, such as during the period of cell death in the rd mouse, ²³ and we have successfully reproduced this observation (data not shown). 

Six to 24 Hours after CNTF Injection: Decreased pERK and c-fos Immunoreactivity in the Retina, and Increased GFAP Immunoreactivity in Müller Cells

Immunocytochemical signals observed with antibodies directed against c-fos or pERK 60 minutes after intravitreal injection with CNTF, xylazine, or vehicle, were no longer detectable 6 or 24 hours after injection, and no additional immunoreactivity was seen anywhere else in the retina. However, Müller cell activation at these late time points was suggested by their increased GFAP immunoreactivity. In untreated and vehicle-injected eyes, GFAP immunoreactivity was only observed in astrocytes located in the nerve fiber layer or inner retina, possibly along blood vessels (Fig. 4A ), whereas Müller cells were completely negative in the central retina and showed only minimal immunoreactivity near the periphery. On the other hand, eyes injected with CNTF showed prominent GFAP immunoreactivity in Müller cells at 6 or 24 hours after injection that was particularly obvious in the cell processes extending into the inner retina (Figs. 4B 4C) . Similar results were observed 24 hours after intravitreal injection of xylazine. 

Retinal Explants Treated with FGF2, BDNF, or CNTF: Increased Staining for pCREB, pERK, and c-fos in Cells of the Inner Retina, but Not Photoreceptors

A possible explanation for the lack of photoreceptor activation after intravitreal injection of trophic factors is that the injected factors may not reach the photoreceptors. To investigate this possibility, retinas were grown as organ cultures on porous membranes and pulsed with trophic factors in the medium bathing the photoreceptors. Retinal explants treated with FGF2, BDNF, or CNTF showed increased staining for pCREB, pERK, and c-fos in the inner nuclear and ganglion cell layers, which were not detectable in control explants. Similar results were obtained with explants cultured with the ganglion cell layer or the photoreceptor layer closest to the medium. Increased pCREB immunoreactivity, detectable in some cases after 10 minutes (not shown), was very strong after 60 minutes (Fig. 5) , and was much weaker or even undetectable after 3 hours (not shown). No immunoreactivity was observed after 24 hours in any cell layer. Photoreceptor cell activation was never observed. 

To identify those cell types responding to the neurotrophic factors tested, immunofluorescent double labeling was performed with anti-pCREB and antibodies for cell type–specific markers (Fig. 6) . After administration of FGF2, pCREB was increased in Müller cells, amacrine cells, and ganglion cells. Administration of BDNF resulted in a slightly more general response, including not only Müller, amacrine, and ganglion cells, but also bipolar cells. Injection of CNTF also resulted in Müller, amacrine, ganglion, and bipolar cell activation. None of the neurotrophic factors stimulated any signs of intracellular signaling in either horizontal cells or photoreceptors. These results support our in vivo evidence that indicates that under the conditions tested and with the reagents used, photoreceptors fail to respond directly to BDNF, CNTF, or FGF2 (Fig. 3) . 

Discussion

The results reported in this article can be summarized as follows: (1) although considerable CREB phosphorylation was observed in untreated and vehicle-injected retinas, very little or no ERK phosphorylation or c-fos nuclear immunoreactivity was observed under those conditions; (2) intraocular injection of BDNF, CNTF, FGF2, or TGFα results in ERK phosphorylation and increased c-fos in Müller glial cells; BDNF and FGF2 also resulted in increased pERK in some ganglion and amacrine cells; (3) these responses were transient and were not detectable 6 or 24 hours after factor administration; (4) no evidence of photoreceptor activation was observed with any of the trophic factors tested in these experiments; and (5) similar results (e.g., activation of Müller cells and some inner retinal neurons without signs of photoreceptor activation) were observed with explants of adult mouse retina exposed to the same neurotrophic factors, regardless of whether the photoreceptors or the ganglion cells were closest to the factor-containing medium. 

It has been well established in recent years that intraocular administration of neurotrophic factors can promote the survival of photoreceptors in retinal degenerations of genetic or environmental origin. Although most studies have been done in rats, La Vail et al. ³ reported rescue of photoreceptors in rd/rd, Q344ter, and in nr/nr mice with CNTF or its analog Axokine, and in the nr/nr mice with BDNF. These authors indicated that the actual mechanism of such protection remains unknown and suggested the possible involvement of intermediate cell type(s) in the protective responses, because CNTFα receptors have been found in the inner retina but not in photoreceptors. ^{24 25 26} Our results are consistent with this hypothesis, because we observed no sign of short-term activation in photoreceptor cells in retinas exposed to CNTF, BDNF, or FGF2, while observing considerable activation of Müller cells and some inner retinal neurons. 

The reasons why we were unable to detect activation in photoreceptors are not clear. A trivial explanation for the lack of photoreceptor responsiveness would be that factors injected intravitreally may fail to reach photoreceptors, at least in suitable concentrations; however, the results of experiments with retinas grown in organ culture showed similar lack of photoreceptor activation. The lack of responsiveness of photoreceptor cells to BDNF and CNTF could possibly relate to the reported absence of corresponding receptors from these cells in rodent retinas ^{27 28 29 30} ; TrkB has been immunocytochemically detected in primate cones but not in rods, ³¹ the predominant photoreceptor type in the mouse. The absence of photoreceptor activation by FGF is more difficult to explain, because there are reports that FGF2 and its receptors are present in photoreceptors, ³² that photoreceptor-specific expression of dominant-negative FGF receptors leads to photoreceptor degeneration ³³ and that purified dissociated rat photoreceptors respond to FGF2 with increased protein phosphorylation and increased cell survival. ³⁴ The reason for this apparent discrepancy remains unclear, but possible explanations would be that FGF signal transduction in photoreceptors is mediated by factors other than FGF2 and/or that constant stimulation may result in receptor downregulation, which could make it difficult to detect effects of exogenous FGF2. Alternatively, photoreceptor activation by FGF, BDNF, or CNTF may be mediated by signaling pathways that are not detected by the assay systems we used; consistent with this possibility, it has recently been reported that different cell types may respond to the same growth factor through different second-messenger pathways. ³⁵  

Our observations also add to the growing recognition of the importance of Müller cells in the maintenance and support of other retinal cell types; it has been known for quite some time that Müller cells participate in a series of metabolic activities, including homeostatic regulation of the ionic milieu of the retina, and they have been observed to react to mechanical injuries (including retinal detachment, laser photo-coagulation, or subretinal injection) and in photoreceptor degeneration of both genetic and environmental origin. ^{36 37 38} Müller cells in culture and in situ have been shown to be responsive to treatment with different growth factors, ^{22 39} and to show second messenger activation in response to intramuscular administration of anα ₂-adrenergic agonist, xylazine, recently reported to have survival promoting activity in the retina. ¹¹ In this context, it is relevant that they show responsiveness in situ not only to factors known to prevent photoreceptor degeneration, such as BDNF, CNTF, and FGF2, but also to factors that do not do so, such as TGFα; it will be of interest to investigate possible differences in the metabolic responses of Müller cells to these different factors. It must be noted, however, that although the possibility that Müller cells mediate the photoreceptor rescuing effects is attractive, the evidence supporting this possibility remains largely circumstantial, and the issue awaits direct experimental demonstration. Although Müller cells were the only ones detected by all the factors under investigation, possible contributions of inner retinal neurons cannot be dismissed, particularly taking into consideration that neurons are not only capable of responding to neurotrophic factor support, but can also be sources of trophic factors active on other cells. ⁴⁰ It also appears important to determine whether the patterns of cell activation by neurotrophic factors observed in wild-type animals in this study are also present in retinas affected by photoreceptor degenerations; this issue is currently under investigation using several mouse and rat retinal degeneration models. 

 Supported by Grants EY04859, EY05951, EY10017, EY09769, a core grant P30EY1765 from the National Eye Institute; by a center grant from the Foundation Fighting Blindness; by a Lew R. Wasserman Merit Award (PAC), a career development award (DJZ), a Senior Investigator Award (RA) and an unrestricted departmental grant from Research to Prevent Blindness; by a Wilmer Institute AMD Research Grant; and a grant from Mrs. Harry J. Duffey. RA is the Arnall Patz Distinguished Professor of Ophthalmology. PAC is the George S. and Dolores D. Eccles Professor of Ophthalmology.

 Submitted for publication August 2, 1999; revised October 13, 1999; accepted October 26, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

Immunocytochemical analysis of ERK phosphorylation in retina 60 minutes after intravitreal injection of PBS (A), FGF2 (B), BDNF (C), or CNTF (D). PBS-injected eyes showed light, diffuse immunoreactivity, whereas FGF2, BDNF, or CNTF injections resulted in prominent phospho-ERK signal in cells located in the central part of the inner nuclear layer, which show processes extending radially toward the inner and outer limiting membranes. Strongly immunoreactive structures near the vitreal surface of the retina apparently correspond to endfeet of these processes. As discussed in the text, these cells are likely to be Müller cells. Note that photoreceptor cells are negative and that immunoreactivity in the ONL corresponds to processes from putative Müller cells. Scale bar, 30 μm. GCL, ganglion cell layer; IPL, inner plexiform layer; OPL, outer plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer.

Figure 2.

 

Immunolocalization of c-fos expression in retinas of mice fixed 60 minutes after intravitreal injection of PBS (A), FGF2 (B, C), BDNF (D), or CNTF (E); a retina treated for 3 hours with TGFα is shown in (F). PBS-injected eyes showed only diffuse, low level expression (A). In contrast, retinas injected with FGF2 (B, C), BDNF (D), or TGFα (F) showed strong c-fos immunoreactivity in nuclei of cells whose laminar position suggests that they are likely to be Müller cells, amacrine cells, or ganglion cells. CNTF-injected retinas (E) only showed nuclear immunoreactivity in Müller-like cells. Note that the ONL was consistently negative with all the factors studied. Scale bar, 50 μm. GCL, ganglion cell layer; IPL, inner plexiform layer; OPL, outer plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer; RPE, retinal pigment epithelium.

Figure 3.

 

Immunofluorescent double labeling of retinal sections with an antibody for c-fos (detected with a rhodamine-labeled secondary antibody), and antibodies against protein kinase C (D through J), calbindin D (E through K), or CRALBP (F through L). As shown by single labeling in (A) through (C), these antibodies respectively recognize bipolar cells and cone outer and inner segments (PKC), horizontal cells, and, more lightly, amacrine and ganglion cells (calbindin), and processes and cell bodies of Müller glia as well as RPE (CRALBP). FGF2-injected eyes showed strong immunoreactivity for c-fos in Müller cells (I) and to a lesser extent in amacrine and ganglion cells (H). BDNF injections induced c-fos expression in Müller cells (F) in addition to less robust yet detectable responses in amacrine and ganglion cells (E). Likewise, CNTF injections resulted in strong c-fos expression in Müller cells (L) and in some cases within ganglion cells (K). No detectable c-fos immunoreactivity was observed in horizontal cells and photoreceptor cells. b, bipolar cells; h, horizontal cells; a, amacrine cells; g, ganglion cells; m, Müller cells; GCL, ganglion cell layer; IP, inner plexiform layer; INL, inner nuclear layer; OP, outer plexiform layer; ONL, outer nuclear layer; IS, photoreceptor inner segment; OS, photoreceptor outer segment. Scale bar, 30 μm.

Figure 4.

 

Immunolocalization of glial fibrillary acidic protein (GFAP) in mouse retinas after 6- or 24-hour exposure to intraocularly administered CNTF. Normal untreated control eyes or PBS-injected eyes (A) express GFAP in presumptive blood vessels of the nerve fiber layer, astrocytes, and Müller cell endfeet and no detectable signal in Müller cell bodies or processes. (B) Six hours after CNTF injections, retinas show increased signal localized to Müller cell processes and endfeet. (C) Twenty-four hours after CNTF injection, there was strong signal in Müller cell processes and endfeet. Positive processes radiating outward from the inner nuclear layer were most conspicuously seen in the inner retina. The nuclear layers were counterstained with DAPI. Scale bar, 30μ m. GCL, ganglion cell layer; IP, inner plexiform layer; INL, inner nuclear layer; OP, outer plexiform layer; ONL, outer nuclear layer; RPE, retinal pigment epithelium.

Figure 5.

 

Immunolocalization of phosphorylated CREB in mouse retinal organ cultures. Explants were grown in vitro for 24 hours before treatment with DMEM or growth factors. Few phospho-CREB–positive nuclei were present in explants receiving DMEM (A). A marked increase in the number of phospho-CREB immunoreactive nuclei of the inner nuclear and ganglion cell layers can be seen 60 minutes after the addition of FGF2 (B), BDNF (C), or CNTF (D). Scale bar, 30 μm.

Figure 6.

 

Colocalization of phospho-CREB and cell type specific markers in mouse retinal organ cultures treated for 60 minutes with FGF2, BDNF, or CNTF. Cryostat sections of factor-treated retinal explants were labeled with one of these cell-specific antibodies (green), plus anti–phospho-CREB (red). The cell types recognized by the various cell markers are illustrated in (A through C); anti-PKC identifies rod-specific bipolars, anti–calbindin D recognizes horizontal, amacrine, and ganglion cell types, and anti-CRALBP identifies Müller glial cells. Treatment with BDNF (D through F) resulted in increased levels of CREB phosphorylation in amacrine cells (E), ganglion cells (E), bipolar cells (D), and Müller glia (F), although it had no detectable effects in photoreceptors or horizontal cells. FGF2 treatment (G through I) resulted in detectable amounts of CREB phosphorylation in ganglion cells, amacrine cells, and Müller glia, but failed to show a response in photoreceptors, bipolars, or horizontals. CNTF treatment elicited phosphorylation in bipolars, amacrine, ganglion, and Müller cells; however, this treatment also failed to show a response in photoreceptors and horizontal cells. b, bipolar cells; h, horizontal cells; a, amacrine cells; g, ganglion cells; m, Müller cells; GCL, ganglion cell layer; IP, inner plexiform layer; INL, inner nuclear layer; OP, outer plexiform layer; ONL, outer nuclear layer. Scale bar, 30 μm.

The authors thank David Ginty and John Saari for generously providing us with antibodies against pCREB and CRALBP, respectively.