GDNF Stimulates Rod Photoreceptors and Dopaminergic Amacrine Cells in Chicken Retinal Reaggregates

Abstract

purpose. To investigate the role(s) of glial cell line–derived neurotrophic factor (GDNF) on expression of rod photoreceptor and dopaminergic amacrine cell–specific genes in an in vitro reaggregate model of the chick retina.

methods. Retinal reaggregates derived from embryonic day (E)6 chicks (rosetted spheroids) were supplemented with 50 ng/mL GDNF, or, alternatively, endogenous GDNF expression was downregulated by transient transfection of spheroids with a pCMS-EGFP[GDNF] antisense vector. Using mainly semiquantitative RT-PCR analyses, expression of rhodopsin, four separate opsins, and tyrosine hydroxylase (THase) was analyzed after either treatment.

results. Supplementation with GDNF accelerated rhodopsin mRNA expression and sustained it at an increased level, in contrast to untreated control subjects, where rhodopsin mRNA levels were lower and unmaintained. Expression of red, green, blue, and violet opsins were unaffected. Under these conditions, GDNF also massively increased the expression of tyrosine hydroxylase, the rate-limiting enzyme in the synthesis of dopamine. The expression of endogenous GDNF was blocked in spheroids by using antisense transfections, which resulted in both a significant decrease in rhodopsin mRNA expression and a complete suppression of THase expression, as determined by RT-PCR, Western blot analysis, and immunocytochemistry.

conclusions. GDNF supports expression of both rhodopsin and THase in vitro, two critical molecules involved in the production of rod photoreceptors and dopaminergic amacrine cells, respectively; however, the presence of GDNF does not affect cone production and survival.

Glial cell line–derived neurotrophic factor (GDNF) is a distant member of the transforming growth factor (TGF)-β superfamily. The GDNF family ligands (GFLs) include GDNF and the related polypeptides neurturin (NRTN), persephin (PSPN), and artemin (ARTN). GFL signals are mediated through a multicomponent receptor system consisting of a ligand-binding GPI-linked coreceptor (GFRα) and a signal-transducing Ret tyrosine kinase (cRET). ^{1 2 3} GDNF supports development and survival of a variety of central and peripheral neurons, ^{4 5 6} including midbrain dopaminergic neurons. ^{7 8 9} GDNF is also involved in development of several non-neuronal tissues, ^{10 11} and attracts much attention because of its possible therapeutic application for treatment of various forms of neuronal degeneration. ^{12 13 14 15}  

The role of GDNF in the development of the retina is not yet well defined; however, GFLs and their receptors function in developing chicken retinas. Among several therapeutic approaches, the use of growth factors, ^{16 17 18 19} in particular GDNF, promises to protect and support the survival of rod photoreceptors in retinal diseases, such as retinitis pigmentosa (RP). ^{20 21 22 23 24} Using reaggregated retinal spheroids as in vitro assay models, we found that GDNF regulates proliferation, differentiation, and survival of chicken rod photoreceptors. ²⁵ In that study, GDNF caused a sustained increased proliferation and then promoted the onset of differentiation of rod photoreceptors, increasing their overall number and preventing programmed cell death. In particular, our attention was drawn toward the receptor subunit GFRα4, since it was strongly expressed during retinal development, and its presence correlated with the expression of the signal-transducing receptor unit cRET. ²⁶ Hence, we applied siRNA techniques to knock down this receptor subunit. ²⁷ Indeed, GFRα4-deficient retinal cultures showed a decrease in the number of amacrine cells (ACs), horizontal cells, and blue-sensitive cones. These data provided the first evidence that GDNF and/or activation of the various GFRαs can specifically affect distinct photoreceptor and other retinal cell subpopulations. 

One of the original functions proposed for GDNF involved the regulation of dopaminergic cells. ^{7 8 9} Because in our system GDNF affected not only photoreceptors, but also ACs, we also sought to determine its effect on dopaminergic ACs. Dopaminergic neurons in the chicken retina represent a very small (approximately 1%) subpopulation of ACs. ²⁸ They are recognizable by their large somata, wide dendritic trees that extend into specific sublayers of the inner plexiform layer (IPL), ^{29 30 31} and their location within the innermost row of the inner nuclear layer of the chicken retina. ^{31 32 33 34} Furthermore, they can be selectively visualized by antibodies against tyrosine hydroxylase (THase), the rate-limiting enzyme in the biosynthesis of dopamine. 

In this study, we used reaggregated organotypic spheres (rosetted spheroids) from dissociated cells of the 6-day-old chick retina. ^{35 36} This culture technique imitates retinal development under three-dimensional in vitro conditions and can easily be manipulated by supplementation with soluble factors ²⁵ or by a temporal downregulation of gene expression. ^{27 37} We applied GDNF to rosetted spheroid cultures, or, alternatively, downregulated GDNF expression by applying GDNF antisense oligonucleotides. Using RT-PCR, we found that GDNF had a neuroprotective role on rod photoreceptors, but not on cones. Also, in the presence of GDNF, THase was detected earlier and was expressed at higher rates than in the absence of GDNF. Moreover, knockdown of GDNF completely inhibited THase mRNA expression. Thus, GDNF appears to exert multiple influences that support expression and survival of both rod photoreceptors and dopaminergic ACs. 

Materials and Methods

Tissue Culture

Six-day-old chicken embryos (E6, white leghorn) were used to produce rosetted retinal spheroids. The central parts of the retina were isolated and collected in F12 medium on ice. The retinal tissue was dissociated by tryptic digestion in F12 medium containing 0.05 mg/mL trypsin (Worthington Biochemicals/Cell Systems, Remagen, Germany) for 8 minutes at 37°C. The remaining cell clusters were mechanically dissociated in Hanks’ balanced salt solution (HBSS) containing 0.5 mg/mL DNase I (Worthington Biochemicals/Cell Systems, Remagen, Germany) by 30 to 35 gentle strokes with a round-bored Pasteur pipette. For generation of retinal spheroids, 2 × 10⁶ cells/mL were cultured in 35-mm dishes containing 2 mL aggregation medium (DMEM, 2% FCS, 1% l-glutamine, and 0.15% penicillin/streptomycin, all from Invitrogen-Gibco, Berlin, Germany) on a gyratory shaker in an incubator 37°C and 5% CO₂ (Heraeus Holding GmbH, Hanau, Germany), in either the presence or absence of 50 ng/mL GDNF (Sigma-Aldrich, Deisenhofen, Germany). 

Transfection with Antisense-GDNF Expression Vector

A GDNF antisense sequence (5′-GAATTCCAGCAAAAAGGGAAGGAGGAACCAAAAGGGCAAAAATCGGGGATGTGTCTTAACAGAAATACATTTAAATGTGACTGACTTGGATTTGGGATACGAAACCAAAGAAGAGCTCATCTTCCGGTATTGCAGTGGATCTTGTGATGCA-3′) was cloned into the multiple cloning site of pCMS-EGFP vector (BD Biosciences-Clontech, Heidelberg, Germany). For transient transfections, rosetted spheroids were transfected with pCMS-EGFP[GDNF] as, or, alternatively for controls, with the vector alone. At the appropriate culture day, aggregation medium was removed and replaced by 1700 μL fresh aggregation medium. The transfection medium contained 2× HBS buffer, 2.5 M CaCl₂, TE-buffer, and 10 μg DNA. The formation of bubbles by strokes with a Pasteur pipette initiated the generation of a precipitate. This transfection solution was incubated for 30 minutes at room temperature, before it was admitted to the cultures. Transfected retinal spheroids were treated for at least 12 hours at 37°C and 5% CO₂ on a gyratory shaker. After 12 hours, transfection medium was removed and replaced with fresh aggregation medium. At 48 hours posttransfection time, retinal spheroids were harvested, to isolate RNA and/or to produce cryosections. 

Cryosections and Immunostaining Procedures

For obtaining cryosections, the rosetted spheroids were harvested and fixed in 4% formaldehyde (Merck, Darmstadt, Germany) for 30 minutes at room temperature. After the fixative was removed by two washes in PBS, the rosetted spheroids were soaked in 25% sucrose (Merck) and stored at 4°C. Cryosections of 10 μm thickness were cut on a cryostat (Microm, Walldorf, Germany), mounted on gelatin-coated slides, and stored at −20°C. 

For immunostaining, sections were dried at 37°C and preincubated in blocking solution, containing 3% BSA, 0.1% Triton-X-100 (all from Sigma-Aldrich) in PBS for 30 minutes at room temperature. The tissues were then incubated with the primary antibody for 75 minutes, followed by three washes in PBS. The rod-specific monoclonal antibody rho4D2 (a generous gift from David Hicks, Louis Pasteur University, Strasbourg, France) was used at a dilution of 1:1000 in blocking solution. The red- and green-specific antibody CERN906 (a generous gift of Willem DeGrip, University of Nijmegen, The Netherlands) was used at a dilution of 1:500. For detection of the primary antibody, sections were incubated with donkey anti-mouse-conjugated Cy3 and goat-anti-rabbit (both 10 μg/mL; Dianova, Hamburg, Germany) for 1 hour. Between the last washes, cell nuclei were stained with DAPI (0.1 mg/mL 4′,6-diamidine-2-phenylindol-dihydrochloride in PBS; Merck). Finally, sections were dried and embedded in Kaiser’s glycerin gelatin (Merck). 

Reverse Transcription-PCR

For semiquantitative RT-PCR, total RNA was isolated from retinal cultures at appropriate stages and transcribed in cDNA by using a reverse transcription system from Promega (Mannheim, Germany). Target cDNA was amplified by PCR with specific primers. PCR was performed using parameters of 94°C for 1 minute, 57°C for 1 minute, and 72°C for 1 minute for 29 cycles. Primers for PCR were 5′-TGCCAGAGGATTACCCAGAT-3′ and 5′-AGGTCATCGTCATAGGCTGT-3′ for GDNF, 5′-GGACAGTTCAAGGGCACTGT-3′ and 5′-CTTCTGTGTGGCTGTGATGG-3′ for GAPDH, 5′-TCACCATCCAGCACAAGAAA-3′ and 5′-GACCAGGTTCCCATAGCAGA-3′ for rhodopsin; 5′-ATCAACCAGATCTCGGGGTA-3′ and 5′-CAGAAGCAGTAGGCCACGAT-3′ for red opsin; 5′-CATGATGGGGATAGCTTTCA-3′ and 5′-TGGTGATCATGCAATTACGG-3′ for green opsin; 5′-CGCTCTGAGTGTGATGAGGA-3′ and 5′-AGCTCCGTTCCCACCTTAAT-3′ for blue opsin; 5′-CGTGGGCCTTCTACCTACAG-3′ and 5′-TACTCGCTGCGGTATTTCGT-3′ for violet opsin; and 5′-AGCCTGTATCCAACCCAT-3′ and 5′-GCTTTGACTATCCCATTCTGT-3′ for THase. The relative expression of mRNA was quantified as a percentage of GAPDH band intensity at the appropriate stages. 

Western Blot Analysis

For Western blot analyses, retinal spheroids from the appropriate stages were stored in homogenization buffer (1 mM NaHCO₃, 0.2 mM MgCl₂ × 6 H₂O, 0.2 mM CaCl₂ × H₂O, 1 mM spermidine; pH 8.0; all from Merck) and homogenized. Appropriate amounts of homogenates of spheroids were boiled in Laemmli buffer (Bio-Rad Laboratories, Munich, Germany) with 5% β-mercaptoethanol (Merck), fractionated by SDS polyacrylamide gel electrophoresis and then blotted to nitrocellulose membranes (GE Healthcare, Braunschweig, Germany). Blots were probed with specific primary monoclonal antibody against THase (Chemicon, Hampshire, Great Britain) at a dilution of 1:1000 for 12 hours and then incubated with peroxidase-conjugated anti-IgG (Dianova, Hamburg, Germany) at a dilution of 1:5000 for 90 minutes. Then, blots were developed with chemifluorescence (ECF Western detection system; GE Healthcare) and visualized using exposure to x-ray film for 4 hours. 

Cell Counting and Statistical Analysis

To determine the number of immunostained positive cells, we stained frozen sections with DAPI, rho4D2, and CERN906 antibody, the percentage of immunolabeled cells per section of a single spheroid were calculated in relation to DAPI-positive cells of the same spheroid. At least six cryosections of different spheroids derived from two individual experiments were analyzed. Data are presented as the mean ± SD and compared by a two-tailed Student’s t-test. Note that the SD represents the difference between individual spheroids. To determine the relative expression of mRNA, the percentage was calculated in relation to the GAPDH band intensity for each culture stage. Each experiment was performed at least three times. 

Microscopy and Photography

Photomicrographs of sections were taken with a microscope (Axiophot; Carl Zeiss, Jena, Germany) combined with a charge-coupled device, three-color (CCD-3) digital camera (Intas, Göttingen, Germany). Photomicrographs were processed on a computer with documentation and analysis software (Diskus Histologie MAN F70B; Hilgers, Königswinter, Germany; Multianalysts 2.0, Bio-Rad; Adobe Photoshop CS, Adobe Systems, Inc., San Jose, CA; and Excel, Microsoft, Redmond, WA). 

Results

Effect of GDNF Supplementation on rho4D2⁺ and Rhodopsin mRNA–Expressing Cells

When 50 ng/mL GDNF was added to the rosetted spheroids from E6 chick retinas under serum-reduced culture conditions, the number of rho4D2⁺ cells (immature rods) increased significantly in a time-dependent manner (Fig. 1 ; and see the introduction). At the same time, the expression of endogenous rhodopsin mRNA showed a parallel increase (Fig. 2) . Specifically, the initial expression of rhodopsin mRNA in GDNF-treated cultures was low at 6% (relative to GAPDH controls), but increased rapidly to 21% at day in culture (dic)4 and remained high until dic10 (Figs. 2B 2C) . By contrast, control cultures maintained in serum-reduced medium demonstrated a relative rhodopsin mRNA expression level of 7% at dic2, increasing to 17% at dic6 before declining to 4% by dic10 (Figs. 2A 2C) . 

As opposed to its affects on rod photoreceptors, the number of CERN906⁺ cells (immature red and green cones) were not affected by treatment with 50 ng/mL GDNF, since immunostaining revealed no significant differences between treated and control cultures (Figs. 1B 1D 1H) . Semiquantitative RT-PCR for red, green, blue, and violet cone opsin expression confirmed this finding (Fig. 3) . 

Downregulation of GDNF Expression Using GDNF Antisense Oligonucleotides

By transfecting rosetted spheroids with a pCMS-EGFP(GDNF) antisense vector, we then tested the consequences of downregulation of endogenous GDNF expression. We tested four different GDNF antisense constructs and then used the sequence that achieved the highest transfection efficiency (approximately 30%). The temporal expression of endogenous GDNF was analyzed by semiquantitative RT-PCR (Fig. 4) . Control cultures showed a somewhat variable pattern of GDNF expression, with the highest signals of mRNA transcripts found at dic9 and 11 (Fig. 4A) . After treatment with the GDNF antisense vector, the intensity of GDNF mRNA expression indeed was very low, with expression of GDNF mRNA transcripts detectable only during the first two culture days (Fig. 4B) . Quantitatively (Fig. 4C) , the highest relative GDNF mRNA expression occurred at dic5 (5%), thereafter diminishing gradually to 1% at dic13. By contrast, control cultures maintained expression levels of 8 to 22% between dic5 and 13. 

Knockdown of GDNF Decreases Rhodopsin Expression

The consequence of transfection with GDNF antisense oligonucleotides on the expression of endogenous rhodopsin mRNA was analyzed by semiquantitative RT-PCR (Fig. 5) . Control cultures showed a strong expression of rhodopsin mRNA transcripts at dic5 to 9 48 hours after transfection (17%–20% relative expression), and then a significant decrease at dic11 and 13 (6%; Figs. 5A 5C ). During the first culture days, rhodopsin mRNA expression after GDNF downregulation was high at levels of 8% to 12%. This level subsequently declined to approximately 2% by dic13 (Figs. 5B 5C) . 

These findings were independently confirmed by immunostaining cryosections using the rod-specific antibody rho4D2 (Fig. 6) . Compared with a nontransfected (Figs. 6A 6C)and a transfected control (Figs. 6D 6F) , the GDNF antisense-treated cultures revealed a lower number of rho4D2⁺ cells (Figs. 6G 6I) . For quantification, we determined the ratios of rho4D2⁺ cells in control and GDNF-treated cultures based on the total number of DAPI-positive cells. In transfected control cultures at dic7 and 9 (48 hours after transfection), the number of rho4D2⁺ cells reached 6% to 7%, while in nontransfected cultures, the percentage was approximately 1% higher. In contrast, cultures after knockdown of GDNF possessed less than or equal to half this number of rho4D2⁺ cells. In fact, at the first and last time points tested after GDNF downregulation (dic5 and 13), rho4D2⁺ cells were not detectable at all, while in control cultures rho4D2⁺ cells were readily apparent. 

The rod-specific effect mediated by GDNF was verified by immunostaining for red and green cones using the specific antibody CERN906 (Figs. 6B 6E 6H) . Only at dic7, 9, and 11 (48 hours after transfection) was a weak decrease in the number of red and green cones detected in GDNF antisense transfected cultures (e.g., 0.5% fewer CERN906⁺ cells; Fig. 6K ) compared to control cultures (Fig. 6J) . Therefore, while rod photoreceptors were clearly diminished after downregulation of endogenous GDNF, cone photoreceptors were not significantly affected. 

Supplementation with GDNF Enhanced THase Expression

Since GDNF has been reported to support the expression of THase, we analyzed the effect of GDNF supplementation on THase expression in rosetted spheroids using semiquantitative RT-PCR at different culture stages (Fig. 7) . Under serum-reduced control conditions, initial expression occurred at dic8 (Fig. 7A) , and remained constant until dic10. In contrast, addition of exogenous 50 ng/mL GDNF caused a significant temporal shift of THase expression (Fig. 7B) , appearing at least 2 days earlier than in control cultures. Moreover, the amount of THase expression was significantly increased in response to GDNF supplementation. Quantitatively (Fig. 7C) , the degree of THase mRNA expression increased gradually both under control and GDNF-supplemented culture conditions. However, while in control cultures relative THase expression increased from 5% to 6.5% between dic8 and 10, application of GDNF accelerated this expression. Specifically, at dic6, relative expression was 4%, increasing to 6.5% at dic8 and 9% at dic10. 

The effect of GDNF on THase expression was confirmed by Western blot analysis of GDNF-treated and control cultures at dic6, 8, and 10. Figure 8shows immunoblots probed with an antibody specific for THase. The GDNF-treated samples show initial THase protein expression at dic8, whereas in control cultures THase protein expression was first detected at dic10. By this stage, the immunoblots of GDNF-treated rosetted spheroids revealed a much higher intensity of THase expression than nontreated samples. 

Effect of Knockdown of GDNF on Expression of THase

Final confirmation of the dependence of THase on GDNF was illustrated by downregulating GDNF by using antisense transfection (Fig. 9) . As expected, this treatment suppressed expression of THase mRNA transcripts completely (Figs. 9B 9C) , while in control cultures, THase transcripts were detected from dic7 onward (Fig. 9A) , consistent with previous data. 

Discussion

There is tremendous interest in the possibility of using GDNF therapeutically for maintenance of photoreceptors in blinding diseases. Therefore, a precise knowledge of its actions on development and maintenance of retinal tissue and its various cell types is essential. Using GDNF supplementation or knockdown techniques and RT-PCR analyses, we offer further support for a specific effect of GDNF on rhodopsin expression, and therefore on rod photoreceptors, during in vitro development of chick retinal tissue (3-D retinal spheroids). This result had been elucidated by other, mainly histologic, techniques in an earlier report. ²⁵ In addition, we applied antisense techniques to downregulate endogenous expression of GDNF to establish firmly the role of GDNF in rod development. Moreover, we detected a pronounced effect of GDNF on THase expression and thus on development of dopaminergic ACs of the retina. Two topics will be discussed in relation to these findings: first, the individual specificity and mode of GDNF action on rods and dopaminergic ACs, and second, the possibility of a common link between the actions of GDNF on these two different cell types. 

Rod Photoreceptors as a Major Target for GDNF

The loss of photoreceptors caused by cell degeneration leads to retinitis pigmentosa (RP), a primary cause of blindness worldwide. Trophic factors capable of protecting neurons from cell death provide a promising strategy for neuroprotective intervention, as indicated by retinal degeneration studies in various animal models. ^{16 17 38} A survival-promoting effect of GDNF has been reported for ganglion cells and photoreceptors ^{39 40 41} and more specifically for rod survival. ^{42 43} GDNF was shown to delay degeneration of rod photoreceptors in mouse and in a transgenic rat model of human retinitis pigmentosa. ^{20 21} Moreover, the survival time and functional maintenance of rod outer segments were significantly improved by GDNF. ¹⁸  

In an earlier study, we demonstrated multiple effects of GDNF on rod photoreceptors during in vitro development of the chicken retina, such as the ability of GDNF to affect proliferation, differentiation, and survival of rod photoreceptors. ²⁵ In that study, the role of GDNF as a mitogenic factor was established by demonstrating dose-dependent, cell-specific increases in rod proliferation. We also applied exogenous GDNF to chicken retinal spheroids and documented its effects on photoreceptors by immunohistochemistry. ²⁵ In the present study, we extended the analysis to examine the mRNA levels of rhodopsin using rod-specific RT-PCR probes. Adding 50 ng/mL GDNF to retinal spheroids enhanced the expression of rhodopsin mRNA significantly. In contrast to control cultures, a very high expression level of rhodopsin mRNA was sustained over the whole culture period; therefore, GDNF appears to promote the survival of rod photoreceptors. 

Using a loss-of-function approach, we achieved an almost complete gene knockdown and concurrent reduction in rhodopsin mRNA expression to almost half that of control cultures. Immunohistochemistry also revealed a decreased number of rho4D2⁺ cells. 

As established earlier, ²⁵ GDNF effects are mostly restricted to rod photoreceptors. Cone photoreceptors are not much affected after exogenous application of GDNF or its downregulation, as shown in this study. Such rod-specific effects could be explained by differences in the development of rods and cones. The importance of neurotrophic factors in the development of specific cell types of the retina is well known. ^{44 45 46 47 48} In particular, several studies have underlined the importance of extrinsic signals for development of rod photoreceptors. ⁴⁹ As opposed to rods, whose development depends on microenvironmental changes and spatial proximity of cones, ^{48 50} cones are assumed to be more autonomous. An alternative explanation to the rod-specific effect of GDNF is based on the fact that cones are born earlier than rods. Hence, it could be postulated that some cones were already postmitotic and not responsive anymore to GDNF at embryonic day (E)6, the age when our cultures were started. However, a specific effect of GDNF on rods is more likely. 

GDNF in Dopaminergic Cells in the Retina

Dopaminergic neurons in the retina possess a typical morphology that is consistent across different vertebrate classes. Because GDNF had been reported to prevent dopaminergic neurons from cell death, ^{7 8 9 51} and since an effect of knockdown of GFRα4 on ACs was demonstrated, ²⁷ we analyzed its role on dopaminergic cells in retinal spheroids. This approach was technically difficult because of the very small subpopulation of dopaminergic ACs in the inner nuclear layer of the normal chicken retina. 

The influence of GDNF on dopaminergic ACs of the retina has not been extensively studied, whereas the effectiveness of brain-derived neurotrophic factor (BDNF) has been firmly established. In the nervous system, BDNF can regulate the release of several neurotransmitters, including dopamine. The dopamine synthesis and release is modulated by light, ⁵² as is the synthesis of BDNF, ^{53 54} establishing a link between photoreceptors and dopaminergic cells. Dopaminergic ACs present a strong immunoreactivity for TrkB, the specific receptor for BDNF. ^{52 55 56 57} This factor can modulate the dopaminergic system of the retina and also promises to become a therapeutic tool in degenerative eye diseases (e.g., diabetic retinopathy). In this disease, the density of dopaminergic neurons is decreased, ⁵⁸ and changes in the dopaminergic system have been recognized as a first step in its progress, ^{59 60} verifying that a growth factor acting on cells of the inner retina can affect the survival of photoreceptors. This fact underscores the necessity to gain better understanding of mutual relationships between photoreceptors and cells of the inner retina and their regulation by various growth factors. Therefore, we also studied the regulation of dopaminergic ACs by GDNF. 

It was difficult to detect an effect of GDNF on dopaminergic cells in reaggregated chicken retinal spheroids by immunohistochemistry, since less than 1% of ACs are dopaminergic in vivo and in vitro. However, by using RT-PCR to analyze THase mRNA levels in developing spheroids, we showed a very strong effect of GDNF on THase expression. Although this does not prove that differentiation and survival of these cells is sustained by GDNF, the results support such a notion. Amplification of the rate-limiting enzyme THase for dopamine synthesis revealed an earlier and much increased expression in GDNF-treated cultures. These findings were further confirmed by Western blot analysis, revealing a similar protein expression pattern as that shown for mRNA expression. Even more impressively, the downregulation of GDNF in chicken retinal spheroids suppressed expression of THase mRNA completely. Therefore, it is very likely that the differentiation of dopaminergic ACs is affected by GDNF. 

Why does GDNF affect both rods and dopaminergic ACs? These latest results, together with our earlier findings, strongly suggest that GDNF supports differentiation and survival of both dopaminergic ACs and rod photoreceptors. This notion is in line with other reports that show that GDNF enhances differentiation of various cell types in the nervous system. ^{61 62 63} In the retina, not much is known about the effect of GDNF on other cell types, such as ganglion, bipolar, amacrine, horizontal, or Müller glia cells (MGCs). Given the present results, we have to ask how GDNF achieves similar effects on two separate cell types in the developing retina. GDNF has been suggested to act on photoreceptor precursors in a direct manner, ⁴³ whereas some other trophic factors mediate their protective effects indirectly by activation of MGCs. ^{64 65} In porcine retina, GFRα1, -2, and -3 and cRET were expressed on MGCs but not on photoreceptors, strongly suggesting that MGCs could represent a primary GDNF target. Indeed, a rescue of photoreceptors by GDNF was mediated by retinal MGCs, possibly due to upregulation and release of FGF-2, which in turn could act on photoreceptors. ⁶⁵ This appealing mechanistic scheme puts MGCs at the center of GDNF action in the retina and certainly deserves further attention. With its processes extending across the entire retinal width, MRCs could well be cellular mediators for the effects of GDNF on both photoreceptors and on dopaminergic ACs. 

Altogether, we have shown that GDNF possesses different functions during in vitro development of the chicken retina. GDNF strongly affected both rhodopsin and THase expression; thus, GDNF and its receptor(s) are likely to function in the differentiation and protection of rod photoreceptors and dopaminergic ACs. For a further understanding of the function of GDNF and its possible interaction with MGCs and the rest of the retina, additional investigations are needed. However, results from these studies offer more evidence that GDNF and its receptors could become therapeutic tools and/or targets for extending the lifespan of degenerating photoreceptors. 

²  Present affiliation: Biotechnologisch-Biomedizinisches Zentrum, Universität Leipzig, Leipzig, Germany.

 Supported by European Union SENSPESTI (Tissue Engineering of Living Biosensors to Evaluate Risks for Health by Pesticides Affecting the Cholinergic Neurotransmitter System) Consortium QLK4-CT-2002-02264 and Deutsche Forschungsgemeinschaft Grant DFG-La379/12-1.

 Submitted for publication March 14, 2007; revised April 30 and June 26, 2007; accepted September 11, 2007.

 Disclosure: K.N. Volpert, None; A. Rothermel, None; P.G. Layer, 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.

 

GDNF increased the number of rod (G) but not cone (H) photoreceptors in rosetted spheroids after treatment with 50 ng/mL GDNF. Dashed circles: spheroids; ipl, histotypic area homologous to the inner plexiform layer. Cryosections were immunostained with the red and green cone–specific antibody CERN906 (B, E) and the rod-specific antibody rho4D2 (C, F). For quantification, cell nuclei were stained with DAPI (A, C). The percentage of CERN906⁺ cells and rho4D⁺ cells was calculated in relation to DAPI⁺ cells (G, H). Each data point represents the mean ± SD of multiple spheroid sections (n = 6). *P < 0.01.

Figure 2.

 

Temporal expression of endogenous rhodopsin mRNA in rosetted spheroids in serum-reduced culture conditions (A) or when treated with 50 ng/mL GDNF (B). Rhodopsin mRNA was analyzed by semiquantitative RT-PCR at different days in culture. For quantification, the percentage of rhodopsin mRNA was calculated in relation to expression of GAPDH mRNA (C). Densitometric quantification of PCR products showed a gradual increase of rhodopsin mRNA from dic2 to 6, remaining high in presence of GDNF, while decreasing again in its absence. *P < 0.01; **P < 0.001; ***P < 0.0001.

Figure 3.

 

Temporal expression of endogenous opsin mRNA in rosetted spheroids in serum-reduced culture conditions (A) or when treated with 50 ng/mL GDNF (B). Opsin mRNA for red, green, blue, and violet cones was analyzed by semiquantitative RT-PCR at different days in culture. For quantification, the percentage of opsin mRNA was calculated in relation to expression of GAPDH mRNA (C, D). Densitometric quantification of PCR products revealed no significant difference between control and GDNF-treated cultures.

Figure 4.

 

Expression of GDNF mRNA in anti-GDNF-transfected rosetted spheroids. After the indicated culture periods, spheroids were transfected under control conditions (A) or with an antisense-probe specific for GDNF (B). Transfection continued for 12 hours, and harvest was 48 hours after transfection. Agarose gel electrophoresis of RT-PCR products revealed that GDNF mRNA expression was effectively downregulated (B, cf. A). Densitometric measurements document a successful GDNF knockdown (C). *P < 0.01; **P < 0.001; ***P < 0.0001.

Figure 5.

 

Temporal expression of rhodopsin mRNA in anti-GDNF–transfected rosetted spheroids. Spheroids were transfected at different culture days in control conditions (A) or with a GDNF antisense probe (B), analyzed by semiquantitative RT-PCR, and quantified by comparison to the expression of GAPDH mRNA (C). Electrophoresis of RT-PCR products revealed a similar expression pattern under both conditions (A, B), but note the different expression intensities (C). *P < 0.01; **P < 0.001.

Figure 6.

 

GDNF antisense transfection decreased the number of rod photoreceptors, whereas cones remained unaffected. Cryosections of spheroids were immunostained with the cone-specific antibody CERN906 (B, E, H) and the rod-specific antibody rho4D2 (C, F, I), and cell nuclei were stained with DAPI (A, D, G). Rho4D2⁺ cells decreased after transfection with the GDNF antisense probe (I, J), as quantified in relation to DAPI⁺ cells (G). The number of CERN906⁺ cells was unaffected by transfection with the GDNF antisense probe (H, K). Each data point represents the mean ± SD of multiple spheroid sections (n = 6). **P < 0.001. Scale bar, 100 μm.

Figure 7.

 

Expression of THase mRNA accelerated and increased in rosetted spheroids treated with GDNF (50 ng/mL) (B, cf. control cultures in A), as analyzed by semiquantitative RT-PCR at different culture days and densitometric quantification of PCR products (C). Note that, in the presence of GDNF, RT-PCR products for THase were detected earlier than in control conditions (lane 6), and their amounts were higher at all time points. **P < 0.001.

Figure 8.

 

Protein expression of THase during culture of retinal spheroids under serum-reduced culture conditions, as analyzed by Western blot of control and GDNF-treated cultures at dic6, 8, and 10. Immunoblots were probed with an anti-THase–specific antibody. GDNF-treated cultures showed expression of THase at dic8, whereas in control cultures initial expression was detectable only at dic10. Moreover, GDNF-treated cultures showed a higher intensity of THase expression than did control cultures.

Figure 9.

 

Expression of THase mRNA after transfection of rosetted spheroids with a GDNF antisense probe was undetectable (B, cf. A).

The authors thank Colin Barnstable, David Gamm, David Hicks, and Joyce Tombran-Tink for helpful discussions and Jutta Huhn and Meike Stotz-Reimers for expert technical assistance.