Colocalization of TrkB and Brain-Derived Neurotrophic Factor Proteins in Green-Red–Sensitive Cone Outer Segments

Abstract

purpose. To examine the distribution of neurotrophins (NTs) and their catalytic receptors in adult rat photoreceptors.

methods. Immunocytochemistry and Western blot analyses were performed using primary antibodies raised against NTs (nerve growth factor [NGF], brain-derived neurotrophic factor [BDNF], NT-3, and NT-4/5) and NT receptors (TrkA, TrkB, TrkC, and p75NTR). Double-labeling of retinal sections with opsin-specific antibodies was performed to identify each photoreceptor type. Competitive experiments using excess recombinant NT or Trk receptors confirmed the binding specificity of each antibody.

results. TrkB and BDNF immunoreactivity was colocalized in cone outer segments. TrkB and BDNF were detected in all green-red–sensitive cones, but not in blue-UV cones or rods, and other NTs and NT receptors were not detected in any of the photoreceptor types.

conclusions. The findings suggest a specific role for BDNF through its signaling receptor TrkB in the function and maintenance of green-red cones, the predominant cone type in the rat retina.

The neurotrophins [NTs; nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), NT-3, and NT-4/5] play important roles in the development, maintenance, and plasticity of neurons in the vertebrate visual system. ^{1 2 3} The biologic effects of NTs are mediated by two classes of cell surface receptors: One is the trk family of receptor tyrosine kinases comprising TrkA, the receptor for NGF; TrkB, the receptor for BDNF and NT-4/5; and TrkC, the receptor for NT-3. ^{4 5 6 7} The other is the receptor p75 (p75NTR), which binds all the NTs. ⁸ The trkB and trkC genes give rise to multiple transcripts encoding full-length catalytic receptors and truncated proteins that do not have the cytoplasmic tyrosine kinase domains. ^{6 9 10 11}  

Among NTs, BDNF has been identified as an important factor for the survival of injured neurons in the adult rodent retina. Photoreceptors can be protected from the damaging effect of constant light ¹² or from inherited retinal degeneration ¹³ by a single intravitreal injection of BDNF. Similarly, intraocular administration of BDNF recombinant protein ^{14 15 16} or BDNF gene transfer using viral vectors ¹⁷ confers protection on retinal ganglion cells that otherwise die soon after optic nerve transection. BDNF has also been shown to protect cells in the inner retina after ischemic damage in vivo. ¹⁸  

The cellular localization of NTs and their receptors in retinal neurons provides information relevant to the potential function and mechanism of action of these molecules. In the inner retina, there is substantial evidence for specific expression of TrkB and BDNF, both at the mRNA and protein levels, in retinal ganglion cells and amacrine cells of many species. ^{19 20 21 22 23 24 25 26 27 28 29} The localization of these proteins in photoreceptors has been more elusive. Two reports have identified TrkB-like immunoreactivity in primate ³⁰ and rat ²⁹ photoreceptors. However, at present, the specific cell-type distribution of these proteins in the photoreceptor layer has not been determined. 

In this study, we examined the distribution of all NTs and their receptors in the outer retina of adult rats using opsin-specific antibodies to identify each photoreceptor type. We demonstrated selective colocalization of TrkB and BDNF proteins in green-red cones, in contrast with the absence of immunostaining for all NTs and NT receptors in blue-UV cones and rods. 

Materials and Methods

Antibodies

Neurotrophins.

Antibodies (crude antisera) against amino-acids 168–177 of human BDNF (BDNF_168–177; diluted 1:500 for immunocytochemistry and 1:2000 for Western blot analysis), amino acids 123–131 of human NT-4/5 (NT-4/5_123–131; diluted 1:500), and amino acids 178–186 of human NT-3 (NT-3_178–186; diluted 1:500) ³¹ were provided by David Kaplan (McGill University, Montreal, Quebec, Canada). Two other antibodies against BDNF were used: a rabbit antiserum raised against a portion of mouse BDNF corresponding to amino acids 111–123 ³² (BDNF_111–123; diluted 1:200) provided by Yves Barde (Max-Planck Institute, Martinsried, Germany) and an affinity-purified rabbit polyclonal antibody raised against amino acids 128–147 of human BDNF (N-20; 1–5 μg/ml; Santa Cruz Biotechnology, Santa Cruz, CA). An affinity-purified antibody against an epitope at the amino terminus of human NGF (H-20; 2 μg/ml; Santa Cruz Biotechnology) was used. 

NT Receptors.

The rabbit polyclonal antibodies TrkA_in, against amino acids 462–481 of human TrkA (diluted 1:200), TrkB_in, against amino acids 482–501 of rat TrkB (diluted 1:500 for immunocytochemistry and 1:5000 for Western blot analysis), and TrkC_in, against amino acids 637–653 of rat TrkC (diluted 1:200) were generated against the intracellular, catalytic domains specific for full-length TrkA, TrkB, or TrkC, respectively. ^{33 34} In addition, the rabbit polyclonal antibody pan-Trk203 (diluted 1:500) was generated against the C-terminal 15 amino acids common to all full-length Trk receptors. ^{5 35} All anti-Trk antibodies used in this study were crude sera provided by David Kaplan. The monoclonal antibody 192-IgG (hybridoma supernatant, diluted 1:1000) against an extracellular epitope of p75NTR ³⁶ was provided by Phil Barker (McGill University, Montreal, Quebec, Canada). 

Visual Pigments.

Monoclonal antibodies against chick cone opsins were provided byÁ goston Szél (Semmelweiss University, Budapest, Hungary): COS-1 recognizes middle-wave or green-red–sensitive cone opsins (diluted 1:100), and OS-2 recognizes short-wave or blue-UV–sensitive cone photopigments in mammals (diluted 1:100). ^{37 38} The specificity of these antibodies in the retina has been demonstrated by competitive inhibition using synthetic peptides corresponding to each of the visual pigments. ³⁹ The monoclonal antibody rho4D2 against bovine rhodopsin (diluted 1:50) ⁴⁰ was provided by Robert Molday (Jules Stein Eye Institute, University of California, Los Angeles). All antibodies against visual pigments used in this study were in the form of hybridoma supernatant or ascites fluid. 

Secondary Antibodies.

Affinity-purified fluorophore-conjugated goat anti-mouse IgG (red, 4μ g/ml, Alexa 594; Molecular Probes, Eugene, OR); fluorophore-conjugated goat anti-rabbit IgG (green, 3 μg/ml; Alexa 488, Molecular Probes), biotinylated anti-rabbit Fab fragment (5μ g/ml; Jackson ImmunoResearch, West Grove, PA), and horseradish peroxidase–conjugated anti-rabbit IgG (2 μg/ml, Amersham Pharmacia Biotech, Piscataway, NJ). 

Tissue Source and Processing

Animal procedures were performed in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the McGill University Animal Care Committee guidelines for the use of experimental animals. Under general anesthesia, adult Sprague–Dawley rats (Charles River Breeders, St-Constant, Quebec, Canada) were perfused intracardially with 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.4), and the eyes were immediately enucleated. The anterior part of the eye and the lens were removed, and the remaining eyecup was immersed in the same fixative for 2 hours at 4°C. Eyecups were equilibrated in graded sucrose solutions (10–30% in PB) for several hours at 4°C, embedded in optimal cutting temperature compound (Tissue-Tek; Miles, Elkhart, IN), and frozen in a 2-methylbutane-liquid nitrogen bath. Radial cryosections (6–12 μm) were collected onto gelatin-coated slides and processed for immunocytochemistry. In some cases, rats were deeply anesthetized, and after removal of the eyes, the retinas were rapidly dissected and processed for Western blot analysis. 

Immunocytochemistry

Retinal cryosections were incubated in 10% normal goat serum (NGS) and 0.2% Triton X-100 (Sigma, St. Louis, MO) in phosphate-buffered saline (PBS) for 30 minutes at room temperature to block nonspecific binding. Primary antibodies were added in 2% NGS and 0.2% Triton X-100 and incubated overnight at 4°C. Sections were then processed with fluorophore-conjugated secondary antibodies and mounted with an anti-fade reagent (SlowFade; Molecular Probes). After incubation with the appropriate primary antibody, some sections were processed with biotinylated anti-rabbit Fab fragment, avidin-biotin-peroxidase reagent (ABC Elite; Vector, Burlingame, CA) and reacted in a solution containing 0.05% diaminobenzidine tetrahydrochloride (DAB) and 0.06% hydrogen peroxide in PB (pH 7.4) for 5 minutes. Control sections were treated in the same way but with omission of primary antibodies. Sections were visualized with light or fluorescence microscopy (Polyvar; Reichert–Jung, Vienna, Austria) or by confocal microscopy with a laser scanning microscope (model 410; Carl Zeiss, Oberkochen, Germany). 

Antibody Adsorption Experiments

Preadsorption experiments were performed using TrkA, TrkB, and TrkC proteins expressed in Sf9 insect cells. Sf9 cells (Invitrogen, Carlsbad, CA) were cultured in Grace’s medium (Gibco, Burlington, Ontario, Canada) supplemented with 10% fetal bovine serum in a nonhumidified incubator at 27°C. Cells were infected with baculovirus, to express each of the full-length Trk receptors, at multiplicities of infection ranging from 1 to 10. After 48 hours, cells were collected by centrifugation (10 minutes at 1000g, 4°C) and resuspended in PBS. The TrkA- and TrkC-expressing cells were combined. Next, the TrkB-expressing and the combined TrkA- and TrkC-expressing cells were washed three times in PBS. To expose the intracellular domain of the expressed Trk receptors, cells were lysed with four cycles of freezing and thawing using dry ice-methanol and a 37°C water bath. Cell lysis was confirmed by light microscopy. For antibody adsorption, each suspension was incubated with 5 μl of TrkB_in overnight at 4°C. Cellular debris were then removed by centrifugation, and the supernatant was used for immunostaining or Western blot analysis. 

Labeling specificity for the NTs was determined by preadsorbing each NT antibody with recombinant human BDNF, NT-4, NT-3, or NGF (Regeneron Pharmaceuticals, Tarrytown, NY) at 1 μg/μl overnight at 4°C, followed by staining of retinal sections or immunoblots. 

Western Blot Analyses

Fresh retinas were rapidly dissected and homogenized with an electric pestle (Kontes, Vineland, NJ) in lysis buffer: 20 mM Tris (pH 8.0), 135 mM NaCl, 1% NP-40, 0.1% sodium dodecyl sulfate (SDS) and 10% glycerol supplemented with protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 1 μg/ml leupeptin, and 0.5 mM sodium orthovanadate). After incubation for 30 minutes on ice, homogenates were centrifuged at 10,000 rpm for 10 minutes, and the supernatants were removed and resedimented for an additional 10 minutes to yield solubilized extracts. Protein content was determined with a protein assay kit (Bio-Rad, Hercules, CA). Retinal extracts (100–150 μg) were resolved on 8% (for TrkB) or 15% (for BDNF) SDS-polyacrylamide gels and transferred to nitrocellulose filters (Xymotech Biosystems, Montréal, Quebec, Canada). To block nonspecific binding, filters were placed in 10 mM Tris (pH 8.0), 150 mM NaCl, 0.2% Tween-20 and 5 g dry skim milk for 1 hour at room temperature. Blots were incubated for 16 to 18 hours at 4°C with primary antibodies followed by incubation in peroxidase-linked secondary antibodies. Blots were developed with a chemiluminescence reagent (ECL; Amersham Pharmacia) and exposed to imaging film (X-OMAT; Eastman Kodak, Rochester, NY). 

Results

Full-length TrkB immunostaining was localized in cones (Figs. 1 and 2) , whereas TrkA, TrkC, or p75NTR immunoreactivity was not detected in the photoreceptor layer (not shown). Immunostaining with anti-TrkB_in resulted in robust labeling of cone outer segments as well as a subpopulation of cells in the inner nuclear layer (INL) and throughout the ganglion cell layer (GCL; Fig. 1A ). Within the photoreceptor layer, TrkB immunoreactivity was restricted to the cone outer segments, whereas no staining was detected in photoreceptor nuclei or inner segments (Fig. 1B) . Omission of primary antibodies resulted in sections devoid of stain. A second antibody, pan-Trk203, that recognizes a different epitope within the intracellular domain of TrkB, produced a pattern of immunostaining in cone outer segments identical with that observed with TrkB_in (Fig. 1C) . Using these antibodies, we did not detect any immunostaining in the retinal pigment epithelium (RPE; Figs. 1B 1C ). 

To verify the specificity of TrkB immunostaining, we preadsorbed the TrkB_in antibody with Trk proteins expressed in Sf9 cells. Adsorption of TrkB_in with recombinant TrkB eliminated all staining (Fig. 2A) , whereas incubation with recombinant TrkA and TrkC did not alter the labeling pattern (Fig. 2B) . The finding that two antibodies generated against different parts of TrkB exhibit identical and restricted patterns of immunostaining, together with our preadsorption data, support the specificity of the signal observed using these antibodies. 

To examine the localization of NTs in photoreceptors, antisera generated against BDNF, NT-4/5, NT-3, or NGF were compared. Several antibodies, each raised against a different epitope of BDNF, were used: BDNF_168–177, BDNF_111–123, and N-20. BDNF_168–177 produced labeling of cone outer segments and cells throughout the GCL (Fig. 3a ). Preadsorption of BDNF_168–177 with recombinant BDNF protein abolished all labeling (Fig. 3B) , whereas preadsorption with recombinant NT-3, NT-4/5, or NGF did not alter the staining pattern (not shown). A second antibody, BDNF_111–123, produced staining of cone outer segments similar to that observed with BDNF_168–177 (Fig. 3C) . No labeling was observed in the adjacent RPE using either BDNF_168–177 or BDNF_111–123. We were not able to detect BDNF immunoreactivity in cones using the N-20 antibody (Santa Cruz Biotechnology). This is consistent with another study in which this antibody was found unsuitable for immunocytochemical detection of BDNF in the rodent brain; however, it was effective in recognizing denatured BDNF protein on Western blot analysis. ⁴¹ Antibodies against NT-3, NT-4/5, or NGF did not produce a detectable signal in photoreceptor segments or nuclei (not shown). The common pattern of staining provided by BDNF antibodies raised against two different epitopes and our demonstration that this labeling is blocked by recombinant BDNF protein support the specificity of the BDNF signal detected in cones. 

To further validate the antibodies used in the current study, we tested their ability to recognize TrkB and BDNF proteins in the rat retina by Western blot analyses. Blots of detergent-solubilized retinal extracts were generated and probed with TrkB_in, BDNF (N-20), BDNF_168–177, and BDNF_111–123. Immunoblots probed with TrkB_in revealed an ∼145-kDa band (Fig. 4A ) corresponding to full-length TrkB receptor protein. ^{34 42} This immunoreactivity was eliminated by competition with recombinant TrkB protein (Fig. 4A) , but not by preadsorption with recombinant TrkA and TrkC (not shown). A nonspecific lower molecular weight band (∼62.5 kDa) remained unchanged throughout all our competitive experiments. These results indicate that TrkB_in specifically recognizes the isoform of the TrkB receptor containing the catalytic domain in solubilized retinal extracts. 

Western blot analysis of retinal extracts probed with anti-BDNF N-20 identified an ∼14-kDa protein, a band having identical mobility with human recombinant BDNF (hrBDNF; Fig. 4B ). Competition experiments using excess recombinant BDNF eliminated the ∼14-kDa immunoreactive band, but did not alter a nonspecific ∼42.5-kDa band that was also detected. Similar experiments in which anti-BDNF was preadsorbed with recombinant NT-3, NT-4/5, or NGF did not alter this staining pattern. Identical results were obtained when blots were probed with BDNF_168–177 or BDNF_111–123 (not shown). Our data support the conclusion that the protein migrating at ∼14 kDa is BDNF and that the antibodies used in this study accurately recognize BDNF proteins on Western blot analysis and sections of retinal tissue. 

The rod-dominant rat retina contains only approximately 1% of cones, of which green-red cones are the predominant type—approximately 93% of the entire cone population. ³⁸ To determine the cellular distribution of TrkB and BDNF proteins, we performed colocalization studies using antibodies against cone or rod visual pigments. Double-labeling with TrkB_in and anti-green-red cone opsin (COS-1) demonstrated that all green-red cone outer segments were immunoreactive for full-length TrkB (Figs. 5A 5B 5C ). This staining pattern was observed throughout the dorsal and ventral retina. In contrast, double-labeling with TrkB_in and anti-blue-UV cone opsin (OS-2) did not show any correspondence in the distribution of these two markers (Figs. 5D 5E 5F) . Rod outer segments, visualized with anti-rhodopsin (4D2), were not stained with TrkB_in (not shown). Similar double-labeling experiments using BDNF_168–177 demonstrated BDNF immunoreactivity in all green-red–sensitive cone outer segments (Figs. 6A 6B 6C ), whereas none of the blue-UV–sensitive cones (Figs. 6D 6E 6F) or rods was positively labeled. 

Discussion

We have examined the cellular distribution of NTs and their signaling receptors in photoreceptors of the adult rat retina. Our finding that TrkB and BDNF proteins were exclusively present in all green-red cones is novel. The validity of our results depends critically on the specificity of the antibodies used. Several lines of evidence support the conclusion that the immunoreactivity observed corresponds to the distribution of TrkB and BDNF proteins in green-red cones: Identical labeling patterns were observed using different antibodies that recognize nonoverlapping TrkB or BDNF epitopes; Western blot analysis demonstrated that these antibodies recognize full-length TrkB or BDNF proteins in solubilized extracts of rat retinas; competitive experiments using excess recombinant BDNF or Trk proteins showed that recognition with each antibody is epitope specific; and several of the antibodies used have been extensively characterized in previous studies describing the distribution of TrkB and BDNF in the adult rat brain. ^{31 34 42}  

The anti-TrkB antibodies used here recognize the intracellular catalytic domain of TrkB, an integral membrane protein receptor. Thus, positive TrkB immunostaining in green-red cones suggests that this receptor is likely to be expressed and synthesized by these cells. However, BDNF is a secreted protein that is produced in other layers of the retina, including cells in the INL and GCL. ²² The present study does not resolve whether BDNF immunoreactivity in green-red cones reflects their capacity to synthesize and secrete the NT or their ability to bind BDNF, produced elsewhere, through TrkB membrane receptors. Of note, two independent studies have demonstrated the presence of mRNAs for TrkB ²⁴ and BDNF ²⁵ in photoreceptors of the chick retina, suggesting that these molecules are expressed by these cells. However, previously reported in situ hybridization studies from our laboratory ¹⁹ and others ^{22 43 44} have failed to detect TrkB or BDNF mRNAs in photoreceptors of the rat retina. Several explanations may account for the differences between these protein and mRNA distributions; for example, the low-density (<1%) of green-red cones in the rat retina ³⁸ compared with the chicken retina, in which cones account for up to 85% of all photoreceptors. ^{45 46} In addition, the low sensitivity and spatial resolution of the radiolabeled probes used in these studies, added to the possibility that TrkB and BDNF mRNA levels may be low, could make the detection of these mRNAs difficult. 

The absence of NTs and their receptors in blue-UV cones and rods contrasts with the selective localization of TrkB and BDNF proteins in green-red cones. Although the precise neurotrophic factor dependency of photoreceptors has yet to be elucidated, these results suggest that each photoreceptor type may have different neurotrophic factor requirements and may also depend on changing sources of endogenous and exogenous trophic support. For example, the survival time of dissociated rod outer segments in vitro was enhanced in the presence of ciliary neurotrophic factor, glial cell line–derived neurotrophic factor, and basic fibroblast growth factor, but did not change with BDNF. ⁴⁷ This supports our finding that rods do not express TrkB receptors and are probably unresponsive to BDNF. In vivo studies, however, have demonstrated that intraocular administration of BDNF results in the survival of both rod and cone populations in the adult retina. ^{12 13} The rescuing effect of BDNF on rods may be attributable to an indirect effect through cell–cell interactions between rods and other retinal cells, such as green-red cones or cells in the INL, that express TrkB and are likely to respond to this NT. 

The functional significance of expression of full-length TrkB and BDNF proteins in green-red–sensitive cones remains to be elucidated. The observation that TrkB is restricted to outer segments suggests that this may be the site of NT action. Moreover, the colocalization of TrkB and BDNF immunoreactivity in green-red cones suggests a possible role for BDNF, acting through a paracrine and/or autocrine loop, in the maintenance of these neurons. A model of retinal development has been proposed in which extrinsic, often diffusible, factors influence the choice of cell fate. ^{48 49} Consistent with this, a recent study demonstrated that a secreted factor or factors increases the production of cones by progenitor cells in rat retinal cultures. ⁵⁰ Although the identity of these factors remains to be determined, it is possible that neurotrophic factors, such as BDNF, participate in the differentiation of the green-red cone lineage. Of interest, abnormalities in the neural retina of TrkB knockout mice have been identified, including delayed development of photoreceptors, shortened outer segments, and altered electroretinogram responses. ⁵¹ Further analysis of the phenotype of animals without TrkB and/or BDNF function may provide insight into their potential roles in the development and survival of the green-red cone population. 

²  Present address: Département de Pathologie et Biologie Cellulaire, Faculté de Médecine, Université de Montréal, Quebec, Canada.

 Supported by a Young Investigator Award from the Foundation Fighting Blindness (ADP) and a Medical Research Council of Canada grant (AJA, GMB).

 Submitted for publication December 8, 1999; revised May 15, 2000; accepted July 5, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Localization of full-length TrkB receptor protein in cones of the adult rat retina. (A) Retinal radial cryosection immunostained with anti-TrkB_in. Robust staining was found in cones. TrkB-positive cells were also observed in the GCL and INL. (B) Detail showing TrkB labeling restricted to cone outer segments. Note the absence of staining in inner segments or photoreceptor nuclei. (C) Staining with pan-Trk203, an antibody that recognizes a different epitope of TrkB, produced an identical labeling pattern. OS, outer segments; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer. Scale bars, (A) 50 μm; (B, C) 10 μm.

Figure 2.

 

Competitive experiments with recombinant Trk receptor proteins. TrkB immunoreactivity in the retina was eliminated when the antibody TrkB_in was preadsorbed with TrkB protein expressed in Sf9 cells (A), but not when this antibody was preincubated with TrkA and TrkC (B). OS, outer segments; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer. Scale bars, 20 μm.

Figure 3.

 

BDNF immunoreactivity in cones. (A) Distribution of BDNF immunoreactivity in cone outer segments as well as cells in the GCL with the antibody BDNF_168–177. (B) BDNF_168–177 preadsorbed with recombinant BDNF did not produce the immunostaining pattern observed in (A). (C) Labeling identical with that produced with BDNF_168–177 was observed with BDNF_111–123, an antibody that recognizes a different BDNF epitope. OS, outer segments; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bars, (A, B) 50 μm; (C) 8 μm.

Figure 4.

 

Western blot analysis of TrkB and BDNF in solubilized extracts of adult rat retinas. (A) The antibody TrkB_in recognized an ∼145-kDa band corresponding to full-length TrkB receptor present in the rat retina. Preadsorption of TrkB_in with recombinant TrkB completely eliminated this band. (B) An ∼14-kDa band corresponding to BDNF was detected in retinal extracts using anti-BDNF N-20. This band was specifically eliminated by preadsorbing N-20 with excess recombinant BDNF. This antibody also recognized hrBDNF.

Figure 5.

 

Full-length TrkB immunoreactivity in green-red cone outer segments. (A) Confocal fluorescent images of retinal sections show anti-TrkB_in immunostaining; (B) anti–green-red cone opsin antibody (COS-1) labeling within identical fields. and (C) superimposition of the images in (A) and (B). Adjacent sections stained independently with each of these antibodies under the same conditions as those used in (A) and (B) did not show bleed-through between filters (not shown). (D) Anti-TrkB_in immunoreactivity; (E) anti–blue-UV cone opsin (OS-2) staining in the same field; and (F) superimposition of (D) and (E). Note that there is no overlap between TrkB_in- and OS-2-immunopositive cone outer segments in (F). In Figures 5 and 6 , magnification, ×40; numeric aperture, 1.3. Scale bars, 10μ m.

Figure 6.

 

BDNF immunoreactivity in green-red cone outer segments. (A) Anti-BDNF immunoreactivity in cone outer segments observed with a fluorescein-isothiocyanate filter; (B) anti–green-red cone opsin immunostaining observed with a Cy3 filter in the same field; and (C) superimposition of (A) and (B). Complete overlap between BDNF- and COS-1–positive cone outer segments was found (C). No bleed-through between filters was observed in adjacent sections labeled independently with each of these antibodies. (D) BDNF-positive cone outer segments; (E) outer segment stained with anti–blue-UV cone opsin in the same field; and (F) superimposition of (D) and (E) showing absence of overlap between immunostained cone outer segments.

The authors thank Wendy Wilcox for technical assistance, Danny Baranes for assistance with confocal microscopy, and Timothy Kennedy for critical comments on the manuscript.