Basic Fibroblast and Epidermal Growth Factors Stimulate Survival in Adult Porcine Photoreceptor Cell Cultures

Abstract

purpose. To investigate the effects of basic fibroblast and epidermal growth factor (FGF2 and EGF, respectively) on the survival and phenotypic expression of photoreceptors isolated from adult mammalian retinas.

methods. Primary cultures highly enriched in photoreceptors were prepared from adult domestic pig retinas and maintained in chemically defined medium. Cell culture composition was characterized through the use of specific antibody markers of retinal neurons, and neuronal survival was quantified through viability assays as a function of time in the presence or absence of different doses of FGF2 and EGF. Western blot analysis of phosphotyrosine residues was used to monitor activation of FGF2 and EGF signaling pathways.

results. Reproducible survival of adult pig rod and cone photoreceptors was obtained for approximately 2 weeks in vitro, with the continued expression of rod opsin, recoverin, S-antigen, cone arrestin, and neuron-specific enolase. Purity of cultures was routinely more than 95% photoreceptors, with a rod-cone ratio of 2:3.1. Photoreceptor survival was stable for the initial week, decreasing slowly during the second, with rapid cell loss occurring thereafter. In the presence of FGF2 (20 ng/mL), the percentage of photoreceptor survival during the second week in culture was statistically significantly different, at least two times higher than in control experiments. Photoreceptor survival correlated directly with increasing concentrations of FGF2, and also of EGF. Combined treatment with FGF2 and EGF did not induce higher survival than either factor alone. There was no detectable selective loss of rods or cones in the experimental model. Phosphotyrosine immunoblots after stimulation of cultures with FGF2 and EGF revealed time-dependent appearance of multiple immunoreactive bands.

conclusions. The adult pig photoreceptor culture in the current study exhibits reproducible neuronal survival in vitro and represents a useful novel experimental system for the study of potential neuroprotective effects and signaling pathways of neurotrophic factors such as FGF2 and EGF in fully adult higher mammalian retina.

A considerable advance in understanding the pathogenic mechanisms underlying human visual diseases has come from studies using natural and transgenic rodent animal models. For example, the construction of transgenic strains harboring different mutations in the opsin gene have revealed how distinct mutations may adversely affect cell survival and culminate in photoreceptor death. ^{1 2 3} Similarly, the naturally occurring retinal degeneration (rd) and retinal degeneration slow (rds) strains have greatly accelerated characterization of the respective gene defects in cGMP phosphodiesterase and peripherin. ^{4 5 6 7} However rodent retinas differ from those of humans in several ways—notably, in the absence of the cone-enriched region known as the macula, serving high-acuity stereoscopic vision. It has also been reported recently that genetic expression patterns are different between mouse and human—for example, the cone rod homeobox (crx) gene, which appears to control transcription of key functional and structural rod and cone protein genes in the mouse, ⁸ has been shown to be expressed with a later onset in human embryos, lagging behind that of some important genes. ⁹ Also, although histologic retinal degeneration is not observed in transgenic mice in which the RPE65 gene has been deleted, one RPE65 mutant human embryo demonstrated widespread ocular changes. ¹⁰ Such differences are especially important when attempting to extrapolate experimental data obtained on rodent and avian species toward a clinical setting. Survival-promoting effects of neurotrophic factors such as basic fibroblast growth factor (FGF2) have been observed in both in vivo ^{11 12 13} and in vitro ¹⁴ studies of rat photoreceptors. However, differences exist between these findings and those obtained in mice, in which FGF2 and many other neurotrophic factors seem to be much less effective. ^{15 16} It even has been reported that FGF2 is proapoptotic for chicken photoreceptors in vitro, ¹⁷ as it is for certain tumor cells. ¹⁸ Finally, we recently showed that human cone photoreceptors display strong FGF receptor immunoreactivity, whereas such labeling is not observed in rat retinas. ¹⁹  

Such discrepancies make it hard to predict which neurotrophic factors will have the most potential impact in the clinic, and some studies have explored the use of higher mammals with retinas more closely resembling that of humans—namely, cats, dogs, and monkeys. ^{20 21 22} Although extremely informative, such analyses are expensive, difficult, and time-consuming. Tissue culture approaches have certain advantages over in vivo experiments, through providing a controlled environment for examining the pharmacology and mechanisms of action of growth factors. Traditionally, in vitro models have relied on the use of embryonic or immature nervous tissue, because it is widely believed that adult neurons exhibit only very limited capacities to survive in such artificial conditions. However, there are numerous differences between newborn and adult neurons in the expression of growth factors and their receptors ²³ and apoptotic mechanisms. ^{24 25 26}  

To circumvent some of these drawbacks, we have described the use of mixed cell cultures prepared from adult human, pig, and rat retinas obtained after death. ^{27 28} Such preparations allow survival and neurite regeneration of all neuronal types and have been used to explore pathogenic mechanisms involved in retinal ganglion cell toxicity. ²⁹ In the present study, we have further refined the model, preparing monodispersed cultures of enriched adult pig rod and cone photoreceptors to determine whether neurotrophic growth factors exhibit survival-promoting effects on adult higher mammalian photoreceptors. 

Materials and Methods

All cell culture media and sera used were from Life Technologies (Invitrogen-Gibco, Cergy Pointoise, France). All other chemicals for which the origin is not indicated were from Sigma-Aldrich (St. Louis, MO). 

Adult Pig Photoreceptor Cell Culture

Adult pig eyes were collected from a local slaughterhouse immediately after death and transported to the laboratory in cold CO₂-independent Dulbecco’s modified Eagle’s medium (DMEM). Dissection of the retina from globes was performed in our laboratory, as described previously. ^{28 30} Retinal fragments (∼1–2 mm²) were washed twice in warm mammalian Ringer’s solution (Ca²⁺- and Mg²⁺-free) and incubated with 1 U activated papain (Worthington Biochemical Corp., Freehold, NJ) in 0.5 mL/half retina for 20 minutes at 37°C and shaken gently once after 10 minutes. Digestion was stopped by addition of 1 mL of neurobasal medium (Nb-A; Neurobasal-A; Invitrogen) medium containing 2% fetal calf serum and DNase I. The tissue was dissociated by gentle shaking of the tubes for 10 seconds, after which the suspension was allowed to settle for 2 minutes. The supernatant containing photoreceptors was carefully decanted, fresh Nb-A added, and the gentle shaking repeated. The contents were again allowed to settle for 2 minutes, after which the supernatant was removed and added to the first. The pellet was discarded, and the pooled supernatants were centrifuged for 5 minutes at 800 rpm and finally resuspended in Nb-A supplemented with B27 mix. Viable cell numbers were estimated after trypan blue vital dye exclusion, and cells were seeded on 12-mm glass coverslips precoated with poly-d-lysine and laminin, placed into 24-well tissue culture plates, at an initial density of 1.10⁶ cells/well (5 × 10⁵ cells/cm²). For immunoblotting and tyrosine phosphorylation studies, cells were seeded into 6 × 35-mm tissue culture plates at the same density. Medium was changed 48 hours after seeding. Plates were rinsed carefully with Nb-A alone, and then the medium was replaced by chemically defined medium consisting of Nb-A supplemented with insulin (5 μg/mL), transferrin (5 μg/mL), selenium (5 μg/mL), sodium pyruvate (1mM), putrescine (100 μM), progesterone (64 nM), prostaglandin D2 (210 nM), tri-iodo-l-thyronine(31 nM), hydrocortisone (5.5 μM), taurine (3 mM), cytidine 5′-diphosphoethanolamine (2.9 μM), cytidine 5′-diphosphocholine (5.2 μM), and antibiotics (penicillin 10 U/mL; streptomycin 10 μg/mL). ³¹ Medium was renewed every 2 to 3 days. 

Human FGF2 (1–80 ng/mL) and epidermal growth factor (EGF; 1–50 ng/mL; Euromedex, Mundolsheim, France) were added into chemically defined medium at the time of media change at 48 hours and at each additional medium renewal. 

Cell-Viability Assay

The relative percentage of live cells during time in culture was determined by using the two-color fluorescence live-dead viability and cytotoxicity assay (Molecular Probes, Eugene, OR), as previously described by us. ¹⁴ Unfixed cells were incubated for 20 minutes at 37°C with a solution of acetoxymethyl ester calcein (2 μM) and ethidium homodimer-1 (1 μM). Calcein acetoxymethyl ester was hydrolyzed by esterase in living cells into membrane-impermeant calcein (green fluorescence), and ethidium homodimer-1 bound to the nucleic acids of damaged and dead cells (red fluorescence), detectable with an immunofluorescence microscope (Optiphot; Nikon, Tokyo, Japan). For each coverslip, cells were counted in 8 to 10 random fields (observed with ×10 or ×20 objectives) or along the entire diameter of the coverslip, if cell density was low. A minimum of 500 cells were counted on at least three separate coverslips for each treatment, and in at least three independent experiments. The absolute number of cells was then calculated for the entire coverslip by using a conversion factor calculated by previous calibration of the microscopic fields to determine the fraction coverage at a given magnification (for ×20 objectives, total number per microscope field × 240). The percentage of living cells under a given condition was standardized by comparing to the number of live cells after 24 hours (taken as 100%). Statistical significance was calculated using the Peritz “f” parametric test, ³² and significance was accepted as P < 0.05. 

Immunocytochemistry

Immunostaining was performed as previously described by us. ²⁷ After fixation with 4% paraformaldehyde in PBS (15 minutes at room temperature), cells were permeabilized with Triton X-100 (0.1% in PBS for 5 minutes) and then saturated with PBS containing 0.5% BSA and 0.1% Tween-20 for 30 minutes. Cells were incubated overnight at 4°C with selected primary antibodies diluted in blocking buffer. Secondary antibody incubation was performed at room temperature for 90 minutes with Alexa (594 or 488) goat anti-rabbit or anti-mouse IgG-conjugated antibodies (Molecular Probes) Primary antibodies used were the anti-rhodopsin monoclonal antibody Rho-4D2, ³³ the monoclonal anti-vimentin (CloneV9; Dako SA, Trappes, France), the polyclonal anti-S-antigen (gift of Yvonne de Kozak, National Institute of Health and Medical Research, Unit 450, Paris, France), polyclonal anti-recoverin (gift of Alexander Dizhoor, Northwestern University, MI), polyclonal anti-human cone arrestin (gift of Cheryl Craft, Xuemei Zhu, and Bruce Brown, Doheny Eye Institute, University of Southern California School of Medicine, Los Angeles, CA), ³⁴ and polyclonal neuron-specific enolase (NSE; Biogenesis, Abingdon, UK). Calculation of percentages of each cell type (rods, cones, nonphotoreceptor neurons, and glia) was performed as described earlier—that is, counting of cells immunolabeled with a variety of specific antibodies within randomly selected fields, then multiplication to obtain the total number per coverslip. Within the total photoreceptor population (representing 95% of the total number of cells), we also calculated the relative percentages of rods and cones. Results are expressed as mean total number ± SD, and were determined from a minimum of 10 coverslips in three independent experiments. 

Protein Extraction and Western-Blot Analysis

For immunodetection with anti-phosphotyrosine, -arrestin, -rod opsin, and -NSE antibodies, cultures were rinsed with PBS and collected in lysis buffer (20 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, 1 mM NaF, and 1 mM Na₃VO₄) containing a protease inhibitor cocktail (Roche Molecular Biochemicals, Meylan, France). For anti-phosphotyrosine immunoblots, cultures were stimulated with FGF2 (100 ng/mL) or EGF (100 ng/mL) for 0.5 to 10 minutes. The reaction was stopped by addition of liquid nitrogen. Cells were then collected in lysis buffer and lysed as just described. Proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. For probing with anti-phosphotyrosine antibody, the membranes were blocked with PBS, 0.1% Tween 20, 3% BSA, and 1% fat-free milk powder (pH 7.3), for 1 hour at room temperature. For all other antibodies, membranes were blocked with Tris-buffered saline, 0.2% Tween 20 and 3% fat-free milk powder for 1 hour at room temperature. Membranes were incubated with primary antibodies (each 1 μg/mL final concentration) overnight at 4°C, and then incubated with goat anti-mouse or goat anti-rabbit IgG-horseradish peroxidase secondary antibodies (0.15 μg/mL; Jackson ImmunoResearch Laboratories, West Grove, PA). Immunoreactive bands were visualized using a kit (Super Signal West Pico; Pierce, Rockford, IL) according to the manufacturer’s instructions. Molecular weights (Life Technologies, Gaithersburg, MD) were compared to prestained molecular size markers. 

Results

Characterization of Enriched Photoreceptor Cell Cultures

Pilot studies demonstrated that very gentle dissociation of papain-digested retinas and collection of just the first two supernatants produced highly enriched fractions of rod and cone photoreceptors, as determined by immunocytochemical labeling of freshly isolated cells (data not shown). Identification of both populations in vitro was determined by immunocytochemistry using retinal cell-type-specific antibodies at different times. Figure 1 shows a representative field after 5 days in vitro in which 23 cells (identified by differential interference contrast optics and DAPI nuclear staining: Figs. 1a 1b ) were immunostained with anti-rod opsin (rods only, Fig. 1c ) and anti-S antigen (rods and cones, Fig. 1d ) antibodies. Twenty-two cells were immunopositive for S antigen, of which 13 cells were also immunopositive for rod opsin. By these immunologic criteria, the field hence contained 13 rods and 9 cones. The identity of cone cells was verified by their specific immunolabeling by a human cone arrestin antibody. Figure 2 shows a field of 28 cells after 7 days in vitro (Figs. 2a 2b) , 16 being immunopositive for rod opsin (Fig. 2c) , and a different 6 cells expressing cone arrestin (Fig. 2d) . Based on quantitation from numerous independent experiments of cells immunostained with the different specific markers, rod cells represented 56% to 65% of the total photoreceptor population, appearing as rounded cells (soma diameter ∼4–5 μm) with short extensions; cones were larger rounded cells (soma diameter ∼7–10 μm) accounting for the remaining 35% to 44% (Table 1) . As assayed by immunostaining with cone-specific antibodies, the proportion of this cell type remained approximately constant throughout the culture lifespan (Table 1) . There was a small increase in cone numbers relative to rods as a function of time in culture, but the difference was not statistically significant. Based on these numbers, the cultures exhibited a rod–cone ratio of 1.3 to 2 rods to 1 cone. DAPI staining showed rod nuclei to occupy almost the entire cell body, exhibiting a mottled appearance, while cone nuclei were relatively smaller, positioned at one pole of the cell and exhibiting a uniform coloration (Figs. 1 2) . Very similar results were obtained for anti-recoverin antibody (labeling of rods and cones, data not shown). Furthermore, immunostaining with anti-NSE antibody (all neurons, including photoreceptors) labeled approximately 98% of the total cells, and anti-vimentin (glia) stained less than 2% of the total cells (data not shown). Hence purity of photoreceptors in these cultures was estimated at 95%, with the balance consisting of other neurons (3%) and glial cells (2%). 

Photoreceptor Survival In Vitro

The live-dead cytotoxicity assay is incompatible with immunologic staining, but based on the cell counts we assumed 95% of the cells to be photoreceptors. Survival of cultured cells was assessed every 2 days by using the live-dead cytotoxicity assay, compared with the number of living cells after 24 hours in vitro. In initial trials, very few cells survived beyond 24 hours in DMEM/Ham’s F-12. We therefore tested a basal medium developed for the culture of central nervous system neurons (Neurobasal-A; Invitrogen). Figure 3a shows that cells survived well under these conditions, the number of surviving cells being roughly stable for 1 week. Then the percentage of surviving cells diminished linearly and progressively for a further week, and, between 15 and 17 days in vitro, degeneration was rapid. 

Neuroprotective Effect of FGF2 and EGF on Photoreceptor Survival In Vitro

FGF2 (20 ng/mL) was added to cultures at 48 hours after seeding and throughout the culture period. No statistically significant difference in cell number was detected between FGF2-treated and untreated cells up to 10 days in vitro. Thereafter, the percentage of surviving cells was significantly higher in the presence of FGF2 than in untreated cultures, and, at 14 days, the number of cells in FGF2-treated cultures were two times that in control cultures (Fig. 3a) . This increased number was due to a delay in cell death rather than any increase in cell proliferation, because FGF2-treated cells also eventually died. The survival-promoting effect of FGF2 was dose dependent, with the degree of survival augmented by increasing concentrations of FGF2 (Fig. 3b) . Addition of increasing concentrations of EGF also led to detectable and significant enhancement of survival, approximately 2.5 times control levels after 12 days in vitro (Fig. 4a) . Addition of FGF2 and EGF together did not stimulate cell survival above levels observed with either factor used alone (Fig. 4b) . In a separate series of experiments, we plotted the number of immunolabeled cones throughout the culture time span, in the presence or absence of FGF2 or EGF. As shown in Table 1 , this number remained very stable, independent of culture time or growth factor supplement, with approximately 60% rods and 40% cones. Immunocytochemical labeling of cultures after 12 to 14 days revealed that FGF- or EGF-treatment did not normally lead to increases in glial the number of cells (data not shown). Cultures in which the number of glial cells exceeded 2% were excluded from analysis of photoreceptor survival. 

Intracellular Signaling after FGF2 and EGF Stimulation

Phosphotyrosine immunoblots of photoreceptor-enriched cultures stimulated for different times with a fixed amount of FGF2 or EGF showed time-dependent changes in tyrosine phosphorylation, distinct for each growth factor. In FGF2-treated cells, the band that showed the largest increase in tyrosine phosphorylation compared with nonstimulated control cells was approximately 110 kDa and was activated to its maximum as early as 30 seconds. This activation was maintained at high levels until 2 minutes, then decreased by 5 minutes, and reached background by 10 minutes (Fig. 5a , arrow). Another band at ∼64 kDa showed similar kinetics, although its activation appeared more transient, because it had already returned to baseline by 1 minute. Higher bands at ∼120 and ∼140 kDa showed more extended kinetics reaching maximum activation toward 5 minutes. This was also true for lower molecular mass doublets at ∼42/44 kDa and ∼32/34 kDa. 

In EGF-treated cultures, the phosphotyrosine profile was very different. A major phosphotyrosine band was seen at ∼180 kDa, becoming activated to its maximum at 2 to 5 minutes, then decreasing by 10 minutes (Fig. 5b , arrow). Additional major bands were observed at ∼70, ∼34/32 (doublet) kDa and ∼30/29 (doublet) kDa, with delayed-activation kinetics compared with the EGF receptor. Fainter bands were also detected at ∼220, ∼110, and ∼80 kDa. 

Western Blot Analysis of Pig Photoreceptor Cultures

Western blot analysis of pig photoreceptor-enriched culture extracts at different times revealed continued expression of NSE, S-antigen, and rod opsin, at least for the first week, with a profile similar to that of in vivo retinal extracts (Fig. 6) . NSE was visible as a single intense band of 46 kDa (Fig. 6a) ; S-antigen appeared slightly larger, ∼48 kDa (Fig. 6b) ; and rod opsin was present both as dimeric forms at ∼74 kDa and multimeric forms of higher molecular mass (Fig. 6c) . Whereas expression levels of rod opsin, S-antigen, and NSE were approximately equivalent in untreated and FGF2-treated cultures at 2 and 6 days in vitro, expression of these proteins was more intense in growth factor-treated cells at 13 days (Fig. 6) . 

Discussion

Long-term survival and regeneration of adult pig retina has been reported by our laboratory for a mixed population of retinal neurons and glial cells, ²⁷ as well as short-term survival of purified young rat retinal photoreceptors. ¹⁴ The present study represents an advance in these two models that permits routine harvesting and in vitro maintenance of more than 95% pure photoreceptors from fully adult large mammalian retina. These cells (a mixture of rods and cones) survived in vitro in the absence of a feeder layer of supporting Müller glial cells for approximately 2 weeks under defined culture conditions. Addition of either FGF2 or EGF led to delayed cell death, and hence increased survival, in the final week of culture. 

The apparent selectivity of papain digestion for photoreceptor isolation is probably a consequence of the stratification of the retina in cell-type specific layers. The photoreceptor layer is fully exposed to the enzyme solution, which seems to penetrate more readily through the outer limiting membrane. The vitreal surface of the retina seems less sensitive to enzymatic digestion, probably due to the inner limiting membrane’s creating an initial barrier. A previous study ³⁵ first reported that more gentle papain conditions preferentially releases photoreceptors, whereas more extended treatments favor inner retinal neurons. 

Nb-A medium is a serum-free medium derived from the original DMEM/Ham’s F12 formula by modifying (glutamine, sodium bicarbonate) or eliminating (glutamate, aspartate, ferrous sulfate) certain components, and by diminishing the osmolarity. ³⁶ It was difficult to determine the absolute cell recovery of this preparation, because we could not distinguish between the efficiency of cell attachment to the substrate and the percentage of cells surviving or dying before attachment. After approximately 24 hours in vitro, we estimated living attached cells to represent 5% of those initially seeded (2 to 3 × 10⁴ cells per coverslip). Thereafter, cell survival and expression of photoreceptor proteins remained constant for 1 week before beginning a slow decline. Both rods and cones continued to express S-antigen and recoverin, whereas rods expressed rhodopsin and cones expressed cone arrestin. In our earlier study, ¹⁴ photoreceptors did not survive in vitro when isolated from rats older than 2 weeks. The difference in the technical approaches may partly explain these results. Vibratome isolation may be more stressful to cell survival and photoreceptors may have greater difficulty in recuperating from the mechanical damage. Indeed, we observed that pig photoreceptors survived only very poorly when isolated by vibratome and cultivated without feeder layers. ²⁸ Enzyme digestion does not lead to such severe perturbations. Another possible explanation is that adult pig photoreceptors possess higher endogenous levels of FGF2, which allows survival up to 15 days without any further addition. We have shown that endogenous retinal levels of FGF1 and FGF2 increase dramatically (more than 100-fold) during late postnatal stages. ^{37 38} There is growing evidence that adult neurons exhibit different responses to neurotrophic factors than do newborn or embryonic tissue. ^{25 39} There are changes in expression and recruitment of both pro- and antiapoptotic proteins in mature compared with immature cells. ^{40 41} Eventually, endogenous growth factor levels seem to be depleted in our culture conditions and are not sufficient to allow continued cell survival. We have no direct evidence that photoreceptors die by apoptosis in these experimental conditions, but a large body of literature suggests that photoreceptors die by apoptosis in numerous retinal degenerations in vivo ^{42 43} and in dispersed primary culture. ⁴⁴ Comparison of relative percentages of rods and cones during culture showed little variation, either as a function of in vitro age or the addition of growth factors. These values (approximately 35%–40% cones and 60%–65% rods) are somewhat higher than the same percentages in vivo (15%–20% cones, 80%–85% rods ¹² ), which may result from preferential release of the more superficially located cones during isolation. Given that there is experimental evidence for the existence of rod-derived cone survival factors, ^{45 46} differential survival might have been expected. One plausible explanation is that sufficient numbers of rods remain viable in these cultures to assure cone survival. Unambiguous testing of this possibility necessitates purification of the rod and cone populations currently being tested. 

Under the experimental conditions reported herein, FGF2 exhibited a survival-promoting effect on adult pig photoreceptors similar to the effect previously observed in young rat cells. ¹⁴ This was observable as both an increase in cell number, as well as stimulation of the expression of phototransduction-related proteins (rod opsin and S-antigen) and metabolic enzymes (NSE) in the surviving cells. This effect was due to a reduction in the rate of cell death rather than a stimulation in the number of new cells. In contrast to the previous observations on rat photoreceptors, ¹⁴ EGF was also an effective neurotrophic agent for adult pig photoreceptors. EGF has been shown to promote survival of hippocampal neurons in vivo and in vitro, ⁴⁷ whereas it had no effect on photoreceptor survival in rats subjected to phototoxicity. ⁴⁸ Previous studies have demonstrated that activation of the EGF receptor triggers second-messenger activation in intact retina, although no signaling cascades were detectable in photoreceptors themselves. ⁴⁹ Both FGF2 and EGF are known to be important regulators of apoptosis, ^{50 51} although the detailed pathways by which this is achieved is not known for mature CNS tissue. Combined addition of FGF2 and EGF did not show cumulative effects on photoreceptor survival, despite activating separate signaling pathways, as evidenced by the distinct phosphotyrosine profiles. This lack of additive effect may be explained by both factors converging on a similar target, such as a component of the apoptosis mechanism. These species differences raise the possibility that the use of EGF as a therapeutic approach to photoreceptor degeneration may be feasible in humans. 

Western blot analysis data have revealed that FGF2 and EGF treatment leads to activation of intracellular signaling cascades resembling those observed in young rat photoreceptors and different from those observed in inner retinal cells from the same age. ⁵² Verification of the identity of second-messenger molecules through immunologic means was not possible because commercially available antibodies do not readily cross-react with porcine tissue. By comparison with published data, the most likely candidate band for FGFR is at 110 kDa, within the molecular mass ranges for FGFR-1, -2, and -4. ⁵³ The higher molecular mass proteins may also represent FGFR, although the ∼140-kDa band could be phospholipase Cγ1. ⁵⁴ Identifying the other phosphotyrosine bands necessitates further experiments, although the p42/44 doublet is almost certainly mitogen-activated protein (MAP) kinase. ⁵⁵ The ∼64 kDa protein was not observed in rat photoreceptor cultures but has the correct mass to represent adaptor proteins such as tyrosine receptor phosphatase (SH-PTP2) or a higher mass isoform of Shc. ^{56 57} One intriguing possibility is that because the pig cultures contain a much higher fraction of cones (approximately 35% cones, 65% rods) than do the almost pure-rod rat cultures, some of these phosphotyrosine bands may be restricted to cones. We are currently attempting to further separate these preparations into rod or cone-enriched fractions to address this question. 

In conclusion, this culture model describes highly enriched dispersed preparations of adult large mammal retinal photoreceptors with the purpose of using them as screens for examining putative neuroprotective effects of candidate trophic factors. The results indicate that FGF2 and EGF can promote survival and phenotypic expression under these conditions, underlining the potential clinical interest of these neurotrophic agents. 

 Supported by the Foundation Fighting Blindness (VT, LG, DH), the British Retinitis Pigmentosa Society (DH), and ProRetina Germany (NK).

 Submitted for publication May 13, 2003; revised June 6 and June 10, 2003; accepted June 12, 2003.

 Disclosure: V. Traverso, None; N. Kinkl, None; L. Grimm, None; J. Sahel, None; D. Hicks, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Immunocytochemicalcharacterization of adult pig photoreceptor cultures. This representative field of 23 cells, as visualized by differential interference contrast optics (a) and nuclear staining with DAPI (b) revealed abundant double immunolabeling of rod opsin (c) and S-antigen (d). All rod opsin-immunoreactive cells were also S-antigen immunopositive (small star, a sample rod, all panels), whereas many S-antigen-immunopositive cells were opsin immunonegative (large star, a sample cone, all panels). Scale bar, 10 μm.

Figure 2.

 

Immunocytochemicalcharacterization of adult pig photoreceptor cultures. This representative field of 28 cells, as visualized by differential interference contrast optics (a) and nuclear staining with DAPI (b) revealed abundant double immunolabeling of rod opsin (c) and cone arrestin (d). Rod opsin- and cone arrestin-immunoreactivities were mutually exclusive. (★) Sample rod (a); sample cone (b). The remaining unlabeled cells were nonphotoreceptor neurons. Scale bar, 10 μm.

Table 1.

 

Percentage of Immunolabeled Cones as a Function of Time in Culture and Growth Factor Treatment

Days In Vitro	Cones in Different Conditions (%)
		CDM	FGF2	EGF
2	35 ± 2.8	—	—
6	36.2 ± 2.0	35.3 ± 3.7	39.2 ± 2.4
9	42.0 ± 3.2	44.0 ± 6.8	39.1 ± 5.8
12	44.7 ± 13.2	38.9 ± 8.1	43.2 ± 6.5

Figure 3.

 

FGF2 increased pig photoreceptor survival in culture. (a) During the first 10 days, there was no difference in survival between nontreated control and FGF2-treated cells (20 ng/mL). Between 10 and 15 days, FGF2 treatment significantly reduced photoreceptor degeneration. At 12 days, FGF2 induced a 45% increase in cell number, and at 14 days the increase was 100%. Data expressed as means ± SD, relative to percentage of cells surviving at 24 hours in three independent experiments for each data point. (b) When photoreceptor survival was measured after 14 days in vitro, FGF2 showed a dose-dependent effect on viability. The effect was first detectable at doses of 20 ng/mL, increasing at 40 and 80 ng/mL to induce five and seven times the number of cells than in control cultures. Data are expressed as absolute number of cells per coverslip (×40 objective) in three independent experiments for each data point (*P < 0.05, **P < 0.01).

Figure 4.

 

EGF increases pig photoreceptor survival in culture. (a) When photoreceptor survival was measured after 12 days in vitro, EGF showed a dose-dependent effect on viability. The effect was statistically significant at doses of 50 ng/mL, inducing a 2.6 times higher number of cells than in control cultures. (b) The survival-inducing effects of FGF2 and EGF were not cumulative, because simultaneous addition of both factors did not stimulate survival levels above those observed with either factor alone. Results expressed as absolute cell number per coverslip (×40 objective) in three independent experiments for each data point (*P < 0.05).

Figure 5.

 

Phosphotyrosine immunoblot of cultured pig photoreceptors after treatment with FGF2 or EGF (both 100 ng/mL). (a) FGF2-induced phosphorylation: notice the time-dependent phosphorylation of prominent bands at ∼140, ∼120, ∼110 (arrow: FGF receptor), ∼64, ∼42/44, and ∼32/34 kDa. Numbers above lanes refer to times of FGF2 stimulation: lane 1, 0 min (untreated control); lane 2, 30 seconds; lane 3, 1 minute; lane 4, 2 minutes; lane 5, 5 minutes; lane 6, 10 minutes. (b) EGF-induced phosphorylation: notice the time-dependent phosphorylation of bands at ∼180 (arrow: EGF receptor), ∼70, ∼34/32, and ∼30/29 kDa. Numbers above lanes refer to times of EGF stimulation: lane 1, 0 min (untreated control); lane 2, 1 minute; lane 3, 2 minutes; lane 4, 5 minutes; lane 5, 10 minutes. Left: molecular masses for both blots.

Figure 6.

 

Immunoblot detection of phototransduction proteins in cultured pig photoreceptors. Cultures were either left untreated (left) or were treated with 20 ng/mL FGF2 (right) and were assayed at 2, 6, and 13 days in vitro for the following proteins: (a) NSE, (b) S-antigen, and (c) rod opsin. NSE and S-antigens were present as single 46- and 48-kDa bands, respectively, whereas rod opsin was present as multimeric forms at ∼74 and ∼100 kDa and more than 180 kDa. Qualitative expression levels between control and FGF2-treated cultures were similar at 2 and 6 days, but, in all three blots, FGF2-treated cultures showed more intense protein expression at 13 days (with loading of equal total protein concentrations for each lane). Left: molecular masses for all three blots.

The authors thank Valérie Forster for expert technical assistance.