Age-Dependent Susceptibility of the Retinal Ganglion Cell Layer to Cell Death

Abstract

purpose. The purpose of this study was to determine the susceptibility of the retinal ganglion cell layer (GCL) to apoptosis after optic nerve transection and excitotoxic stimulus and to investigate the regulation of apoptosis in the GCL during development. The authors also sought to determine the role played by caspases in cell death and their expression during development.

methods. TdT-mediated dUTP nick end labeling (TUNEL) was used to identify cells undergoing apoptosis during mouse retinal development from postnatal day (P)3 to P5 and in retinal explant sections under various conditions. The expression of active caspases was determined by immunohistochemistry (IHC) using an antibody that detects the cleaved large subunit. IHC was also used to detect the expression levels of procaspase-3, procaspase-9, and Apaf-1 in P6 and P60 whole eye sections. Retinal ganglion cells at ages P6 and P60 were purified by immunopanning, total RNA was extracted, and mRNA levels of the above proteins were determined by semiquantitative PCR.

results. After optic nerve transection, a significant number of TUNEL-positive cells were seen 24 hours after lesion in P6 retinas. This death was caspase dependent, as shown by IHC and caspase inhibition with zVAD-fmk. In contrast, adult GCL was resistant to apoptosis under these conditions. Similarly, after excitotoxic stimulus, the GCL of the P6 retinas underwent apoptosis at 6 hours and was caspase dependent, whereas adult GCL was resistant. Developmental apoptosis in the GCL between P2 and P6 was shown to involve caspase-3 and caspase-9. Significant downregulation of Apaf-1 and caspase-3 was detected in the P60 GCL at both the mRNA and the protein levels.

conclusions. Adult GCL is more resistant to apoptosis than neonatal GCL after ON transection and excitotoxic stimulus. The expression of caspase-3 and Apaf-1 is significantly reduced in adult GCL. The authors suggest that age-dependent susceptibility to apoptosis may be caused by this reduced expression.

Apoptosis is a genetically controlled cell death process that plays a critical role in tissue development. ¹ Indeed, nearly half the retinal ganglion cell population dies by apoptosis during retinal development. ² Although this cell death program does not occur under normal physiological conditions in the adult retina, under conditions of degeneration such as glaucoma—the second leading cause of blindness—it has been shown that retinal ganglion cells die by apoptosis. ^{3 4}  

A family of cysteine proteases called caspases, of which there are 15 members, has emerged as the central regulator of apoptosis. Caspases are enzymes present as inactive zymogens, and they become active by proteolytic processing and dimerization. Caspases, once active, are involved in an ordered cascade that culminates in the proteolysis of key structural and nuclear components and the eventual destruction of the cell. Two pathways that activate caspase-3 have been identified, an extrinsic pathway involving death receptors and an intrinsic pathway involving the mitochondria and activation of the apoptosome. ¹ The intrinsic pathway is initiated by the release of cytochrome c from mitochondria to cytosol. In the presence of dATP, cytochrome c binds to the cytosolic adaptor protein Apaf-1. This event allows the recruitment and activation of caspase-9 and results in the formation of the apoptosome, a multisubunit enzyme complex. ⁵ Caspase-9 activates executioner caspases-3 and -7. Caspase-3, in turn, activates four other caspases (-2, -6, -8, -10) in the intrinsic pathway. ⁶ A biochemical hallmark of apoptosis is DNA fragmentation. ¹ Caspase-3 cleaves the inhibitory subunit of CAD (DNase), which, along with other events, cleaves DNA into smaller fragments. ^{7 8}  

Models of glaucoma include IOP elevation in vivo, optic nerve transection, and excitotoxic stimulation, which together result in apoptosis in the ganglion cell layer (GCL). ^{9 10} In the present study, we used retinal explants (involving ON axotomy) to examine the role of the intrinsic pathway of apoptosis in the GCL at different stages of its development. Retinal explants are an ideal method for two reasons: first, they can be used to study mechanisms dependent on the histotypical organization of nervous tissue; second, they bridge the gap between the unstructured cultures of dissociated nerve cells (in which there is a loss of selective interactions between specific cell types) and in vivo preparations. ¹¹ It is evident from previous reports that substantial death occurs in the GCL at 24 hours in neonatal rodents but that death is delayed in the adult GCL after optic nerve axotomy. ^{12 13 14 15} We wanted to further investigate this phenomenon. It has been suggested by some that excitotoxic death seen in glaucoma is caused by glutamate-mediated calcium influx ^{10 16} and that this death could be mimicked using the calcium ionophore A231287. Hence, we chose this stimulus for our study. We also examined the expression of key proteins involved in apoptosis at different developmental stages because previous reports have suggested a developmental downregulation of caspase-3 and Apaf-1. ^{17 18} Finally, we also wanted to address whether caspases are active in the GCL during postnatal developmental cell death. 

Materials and Methods

Animal Treatment

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. C57BL/6 (wild-type) and Balb/c (wild-type) mice were obtained from Harlan UK (Bicester, UK). 

Retinal Explant Culture

Retinal organ culture was performed according to a previously described protocol. ¹⁹ Briefly, C57BL/6 (wild-type) or Balb/c (wild-type) mice were decapitated at postnatal day (P)6 and P60, and their eyes were removed. The anterior segment, vitreous body, and sclera were removed, and the retina was mounted on Millicell nitrocellulose inserts (Millipore, Billerica, MA) with the photoreceptor side down. Explants (without RPE) were cultured in 1.2 mL R16 medium (from A. Romeo Caffe, Wallenberg Retina Centre, Lund University, Sweden) without FCS. Three animals were used at each time point for each stimulus. Calcium ionophore A23187 (1 μM; Sigma, Dublin, Ireland) was used to induce apoptosis for 6 hours. Inhibition of caspases was achieved by incubating retinal explants with the broad caspase inhibitor zVAD-fmk (50 μM; Bachem, Bubendorf, Switzerland). After the indicated time points, retinal explants were fixed in formalin; this was followed by cryoprotection in 25% sucrose overnight. Frozen sections (5–7 μm) were cut, and TUNEL or IHC was carried out as described in the next section. 

Terminal dUTP Nick End Labeling

Briefly, retinal explants or enucleated eyes were fixed in 10% neutral-buffered formalin for a minimum of 4 hours; this was followed by cryoprotection in 25% sucrose overnight at 4°C. Frozen sections (7 μm) were permeabilized using 0.1% Triton X-100 (Sigma). After washes in PBS, sections were incubated with 50 μL reaction buffer containing terminal deoxynucleotidyl transferase (TdT; MSC, Dublin, Ireland) and fluorescein-12-dUTP (Roche, Lewes, UK) according to the manufacturer’s instructions. Hoechst 33342 (Sigma) was used to counterstain nuclei. Sections were incubated for 1 hour at 37°C in a humidified chamber. After several washes in PBS, the sections were mounted (Mowiol; Calbiochem, Nottingham, UK) and viewed under a fluorescence microscope (Eclipse E600; Nikon, Tokyo, Japan). 

Immunohistochemistry

Retinal explants or enucleated eyes were fixed in 10% neutral-buffered formalin for 4 hours, followed by cryoprotection in 25% sucrose overnight at 4°C. After antigen retrieval in 10 mM sodium citrate buffer and quenching of endogenous peroxidase activity in 0.3% hydrogen peroxidase, frozen sections (7 μm) were incubated with anti–cleaved caspase-3 (9661, 1:100) or anti–cleaved caspase-9 (9509, 1:100), whereas whole eye sections (7 μm) were incubated with anti–caspase-3 (9662, 1:100), anti–caspase-9 (9504, 1:100; Cell Signaling, Beverley, MA), or anti–Apaf-1 (APO-20A-061, 1:50; Apotech, Epalinges, Switzerland) overnight at 4°C. Washes in PBS with 0.01% Tween 20 were followed by incubation with secondary antibody, biotinylated anti–rabbit IgG (caspase-3, -9), or anti–rat IgG (Apaf-1) for 1 hour at room temperature. Antibody detection was achieved with a stain kit and reagent (VectaStain Elite ABC Kit and DAB reagent; Vector Laboratories, Peterborough, UK). Sections were counterstained with hematoxylin (BDH, Poole, UK) to facilitate tissue orientation and mounted in DPX (BDH). 

Immunopanning of Retinal Ganglion Cells

Immunopanning of retinal ganglion cells was carried out according to a previously described protocol. ²⁰ Briefly, dissected retinas were enzymatically dissociated in papain (Sigma) in Dulbecco phosphate-buffered saline (Gibco, Carlsbad, CA) to create a single-cell suspension. Retinal ganglion cells were isolated from this suspension using sequential immunopanning using two subtraction plates coated with rabbit α-mouse macrophage antibody (Research Diagnostics, Inc., Flanders, NJ) and one selection plate coated with α-Thy 1.2 IgM antibody (Sigma). 

Semiquantitative PCR

Total RNA was isolated at each time point using reagent according to the manufacturer’s conditions (Trizol; Bio Sciences Ltd., Dun Laoghaire, Ireland). This was then reverse transcribed to cDNA using M-MLV reverse transcriptase (200 U/μL) and oligo (dT) (Promega). Products were amplified with the use of 5 U/μL polymerase (Taq; Promega). To ensure linear signals, all products were optimized for cycle number and product amplification. Equal loading was achieved using α-tubulin as a loading control. Primers (1 mM each) used were as follows: caspase-3—forward, 5′ GTGGACTCTGGGATCTATCTGGACA 3′; reverse, 5′ CACGGGATCTGTTTCTTTGCGTGG 3′; caspase-9—forward, 5′ GCTTCTCTGCCACTGTACTACTGA 3′; reverse, 5′ GGAAGAATTAGCCCTTCTGGTAAC 3′; Apaf-1—forward, 5′ AGCTGATGGGAAGACACTGATTTC 3′; reverse, 5′ GGATTTCTCCATTGTCATCTCCAG 3′; tubulin—forward, 5′ TCGTATCCACTTCCCTCTGG 3′; reverse, 5′ AGCTTGGGGTCTCTGTCAAA 3′ (MWG Biotech); thy 1—forward, 5′ TGCCTGGTGAACCAGAACCTT 3′; reverse, 5′ TGACAGAGAAATGAAGTCCGTGGC 3′; chx 10—forward, 5′ AAATCCGAGACAGTGGCCAA 3′; reverse, 5′ CATTATGCCATCCTTGGCAGA 3′; hpc 1—forward, 5′ AGATCCGGGGCTTTATTGACA 3′; reverse, 5′ GATG-ATGATGCCCAGAATCACA 3′; rhodopsin—forward, 5′ CTCTTCTACACCAGCAACCAA 3′; reverse, 5′ AGCCAAGGTCTAGAATACCAGA 3′; blue cone opsin—forward, 5′ TGAGTCATATGGTGGTGGTGA 3′; reverse, 5′ TGGGAGAGGCCACATAAGGA 3′. 

PCR products were separated on a 1% agarose gel, stained with ethidium bromide, and visualized under UV light. 

Statistical Analysis

Data are given as mean ± SD. Statistical significance was evaluated by Student’s t-test for comparisons between groups. Differences were considered significant for P < 0.05. 

Results

Differential Susceptibility to Optic Nerve Transection

As several previous reports have shown, optic nerve transection results in the death of retinal ganglion cells. ^{12 13 14 15} We wanted to investigate the difference in susceptibility to apoptosis of the GCL to cell death in postnatal and adult mice. Using a retinal explant culture system, we enucleated eyes from P6 and P60 mice and mounted the retinas in inserts. These were left in culture for up to 24 hours for P6 retinas and up to 96 hours for P60 retinas. Apoptosis was detected using TUNEL, a technique that labels cells with DNA strand breaks. It is clear the number of dead cells in the GCL of P6 retinas increased significantly at each time point (P < 0.05) up to 24 hours (Fig. 1e [see also Fig. 4 ]), with 23.68% of the GCL showing TUNEL-positive staining at 24 hours. In contrast, the P60 GCL shows no TUNEL-positive cells at the same time points (Fig. 1j[see also Fig. 4 ]), but a significant number of TUNEL-positive cells (P < 0.05) appeared in the GCL at 72 hours (13.42%), which increased to 23.9% at 96 hours. 

We wanted to determine whether death caused by optic nerve transection was dependent on caspases. In explant sections, staining for cleaved caspase-3 (34.9%) and cleaved caspase-9 (30.9%) was apparent after 24 hours in the GCL at P6 (Figs. 2c 2d [see also Fig. 5 ]). In contrast, cleaved caspase-3 and caspase-9 staining was completely absent from P60 sections at the same time point (Figs. 2g 2h[see also Fig. 5 ]). Secondary antibody-only controls indicated that there was no nonspecific binding by the secondary antibody (Figs. 2i 2j) . Cleaved caspase-3 (9.9%) and cleaved caspase-9 (11.1%) staining appeared at 72 hours and increased significantly (P < 0.05) to 21% and 16.8%, respectively, at 96 hours (see Fig. 5 ). Caspase-dependent death was confirmed by the addition of the broad-spectrum caspase inhibitor zVAD-fmk to P6 explant cultures for 24 hours. In Figure 1f(see also Fig. 4 ), it is evident that the inhibition of caspases protects the GCL and the other layers from apoptosis, showing that death in the GCL as a result of optic nerve transection is dependent on caspases. C57 and Balb/c strains of mice subjected to these treatments responded similarly. Data for both strains were compared using Student’s t-test, and we calculated P > 0.05 for all the time points and treatments; no significant strain difference was observed. 

Differential Susceptibility to Excitotoxic Stimuli

Excitotoxicity, particularly calcium toxicity, is thought to play a role in glaucoma. ¹⁰ Therefore, we treated retinal explants from P6 and P60 mice with the calcium ionophore A23187, which causes calcium influx through the cell membrane. Explants were treated for 6 hours—at this time point, a minimal amount of death caused by optic nerve transection occurs in the GCL. A significant number (32.2%) of TUNEL-positive cells were present in the P6 GCL at the 6-hour treatment with 1 μM A23187 (Figs. 3b 4) , but none were present in the P60 GCL (Figs. 3d 4) . IHC staining of the ionophore-treated explants demonstrated that Ca²⁺-induced death is dependent on caspase-3 with 35.49% staining in GCL and on caspase-9 with 38.2% staining in GCL (Figs. 3e 3f 5) ; however, there was no evidence of caspase activation in the P60 sections (Figs. 3g 3h 5) , correlating with the absence of TUNEL-positive cells. Again we calculated P > 0.05 when we compared both strains of mice in these experiments; no significant strain difference was observed. 

Age-Dependent Downregulation of Apoptotic Mediators

This study clearly shows a marked difference in the response of cells in the GCL to optic nerve transection and excitotoxic stimuli. We wanted to investigate potential causes of this phenomenon. Expressions of the key apoptotic mediators in the GCL of P6 and P60 were compared. A population of retinal ganglion cells was purified by sequential immunopanning, ²⁰ and total RNA was extracted. This was then reverse transcribed to cDNA, and semiquantitative PCR was performed using sequence-specific primers and α-tubulin as a loading control. As shown in Figure 6(a) , the levels of transcript of caspase-3 and Apaf-1 decreased between both ages, whereas the level of caspase-9 was reduced slightly. Measuring levels of inner nuclear layer (INL), outer nuclear layer (ONL), and GCL markers confirmed the purity of the immunopanned population. As shown in Figure 6(a) , the INL (gfap, chx 10) and ONL (blue cone opsin, rhodopsin) markers were absent in the immunopanned population (IP) at P60 (IP P60) and at P6 (data not shown), whereas Thy1 was present, confirming a pure ganglion cell population. We wanted to confirm our results at the protein level. IHC staining of whole eye sections was carried out using antibodies specific for the proform of the caspase proteins and Apaf-1. As shown in Figure 6(b) , staining was obvious in the GCL of all three proteins at P6 (Figs. 6(b) a–c), but this staining was not detectable at P60 for caspase-3 and Apaf-1 (Figs. 6(b) d, f). Staining for caspase-9 remained, consistent with the mRNA result (Fig. 6(b) e). Absence of staining in secondary antibody-only controls showed that there was no nonspecific binding by the secondary antibodies (Figs. 6(b) g–i). Both figures comprehensively showed that caspase-3, the main effector of apoptosis, and Apaf-1, a key mediator in caspase-3 activation, are downregulated during the development of the mouse retina to maturity. 

Developmental Cell Death

Given that key apoptotic modulators are present in neonatal mice, we sought to determine whether these ganglion cells operated a caspase-dependent cell death program during development. The peak of retinal ganglion cell (RGC) death was previously reported to occur between P2 and P5 ²¹ , which prompted us to examine this period. Death in the GCL of P3 to P5 mice was visualized using TUNEL of whole eye sections (Figs. 7d 7e 7f) , and nuclei in the same field were counterstained with Hoechst (Figs. 7a 7b 7c) . In Figures 7d 7e 7f , it is evident there were few dying cells in the GCL per field at each age, consistent with previous reports. ² We then examined the sections of the same ages for caspase activity using IHC staining with antibodies specific for cleaved forms of caspases-3 and -9. Cleaved caspase-3 and caspase-9 were present during developmental cell death, as shown by punctate brown staining of the dying cells in the GCL (Figs. 7g 7h 7i 7j 7k 7l) . 

Discussion

In agreement with previous reports, ^{12 13 14 15} our results show that after optic nerve (ON) axotomy of neonatal rodents, a significant number of cells in the GCL become apoptotic within 24 hours. Developmental apoptosis in the neuroblastic layer (NBL) occurs as usual in vitro and is visible at P6. This death is accompanied by the activation of caspase-3 and caspase-9. Death in the GCL caused by ON axotomy is visibly greater than death caused by development. It has been reported that overexpression of Bcl-2, a protein shown to affect the release of cytochrome c and eventual apoptosome formation, ^{22 23} resulted in the rescue of nearly all postnatal RGCs 24 hours after axotomy, compared with a 50% loss of wild-type cells at the same time. ²⁴ This report supports the key role of the intrinsic pathway in the axotomy model. 

It has been reported, however, that axotomy of the ON in adult rodents results in slower activation of the apoptotic process. ^{12 13} Our results show that 24 hours after axotomy, no TUNEL-positive or pyknotic nuclei were present in the GCL. Cleaved caspase-3 and caspase-9 immunoreactivity was also absent at this time. This result indicates an age-dependent susceptibility to this particular death stimulus. We extended the incubation time and found a significant increase in TUNEL-positive cells in the P60 GCL 72 hours after axotomy. These results are in agreement with previous findings showing that pyknotic nuclei begin to appear 3 days after lesions. Numbers of RGCs are reduced to 10% after 2 weeks, ¹² reaching a peak of apoptosis at 7 days. ¹³  

Delayed pyknosis has been shown elsewhere to be accompanied by the activation of caspase-3 and caspase-9. ^{25 26} We have also found this to be the case in our explant studies. Caspase-8 activity was also detected in one study, ²⁷ hinting at the possibility that the extrinsic pathway of apoptosis may also be involved. However, this possibility was discounted because injection of TNFα and Fas ligand has no effect on RGC numbers. Moreover, caspase-8 has previously been shown to be activated by caspase-3 in the intrinsic pathway, ⁶ and this probably accounted for the detected activity. Our results show that adding the broad caspase inhibitor zVAD-fmk reduced the appearance of TUNEL-positive nuclei to below control levels. This is in agreement with findings of a previous study in which caspase inhibitors were injected in vivo, increasing RGC survival up to 14 days after axotomy in adult rodents and supporting the necessity of caspase activation in the axotomy model. ²⁸  

The differential susceptibility seen with the axotomy model was also seen with an excitotoxic stimulus, the calcium ionophore A23187. This study shows a substantial number of TUNEL-positive cells in the GCL and in the NBL of the P6 retina 6 hours after treatment. Caspase activation was also seen. However, as with the axotomy model, death was not observed in the GCL of the P60 mouse for the same incubation time, nor was there evidence of caspase activation. Increasing the concentration of ionophore (5 μM) had no effect on the presence of TUNEL-positive cells or caspase immunoreactivity at P60 (data not shown). As mentioned earlier, calcium is thought to play a key role in glutamate toxicity. A recent report showed that removing calcium from the media reduced glutamate toxicity by 70%. ²⁹ In the same report, it was also shown that in dissociated cultures, there was a differential susceptibility to glutamate toxicity between neonatal and adult mice. Work conducted in our laboratory shows a differential susceptibility to various stimuli in the GCL and in other layers of the retina as well. There is an age-dependent increase in resistance to stimuli which cause widespread death in neonatal retina (Donovan et al., unpublished results, 2005). 

It has yet to be shown definitively why there is a delay in death after axotomy of adult rodents and a difference in susceptibility of RGCs to glutamate toxicity. It has been suggested that the delay in the appearance of pyknotic cells in the axotomy model is caused by a brief upregulation of the key survival protein phosphorylated-Akt (P-Akt) because inhibition of this event results in premature nuclear fragmentation. ³⁰ In this study, the relative levels of P-Akt double 1 day after axotomy and return to control levels at 3 days, at approximately the time of significant activation of caspases and TUNEL-positive nuclei. Our data seem to correlate with these earlier findings. ³⁰ We have shown previously, using retinal lysates, that there is a downregulation of the key proteins involved in the intrinsic pathway as the retina develops. ¹⁷ In addition, an age-dependent susceptibility of the rat cortex to traumatic brain injury has been shown in another study. ¹⁸ The authors suggest that this was caused by the downregulation of caspase-3 and Apaf-1. ¹⁸ This prompted us to measure the levels of caspase-3, caspase-9, and Apaf-1 in neonatal and adult mice. In agreement with the reports described, we showed a significant downregulation of caspase-3 and Apaf-1 at both the mRNA and the protein levels and a slight downregulation of caspase-9 in the GCL. This is the first report, to our knowledge, that shows a decrease in their expression in the GCL. The mechanism for this event is unknown but is under investigation. 

We have shown, along with others, ^{9 10} the importance of the intrinsic pathway in axotomy and excitotoxicity-induced death. Given the developmental downregulation of the key proteins in this pathway, it is not possible for mature retinal cells to execute this pathway immediately, which could explain a delay in activation. Reentry into the S-phase of the cell cycle first might be necessary for apoptosis to occur, as has been shown to occur in other postmitotic neurons. ^{31 32} The eventual activation of caspases in adult rodents after axotomy (as shown in previous reports ^{25 26 27 28} ) correlates with the increase in cell death seen at equivalent time points and suggests reexpression of these genes for the execution of death pathways. There may also be a caspase-independent component of RGC death, as has been suggested by a previous report in neonatal rats. ³³ Reexpression of key apoptotic mediators could take place by a number of mechanisms. Apaf-1 and caspase genes (casp3) are transcriptional targets of the E2F1 transcription factors. ^{34 35} p53 has also been shown to target the Apaf-1 gene. ³² It has been suggested that an increase in p53 resulting from injury could sensitize cells to apoptosis by increasing Apaf-1 levels. ¹⁸  

The importance of caspase activation, in particular the intrinsic pathway of apoptosis in the development of the CNS, is visible by the phenotypes of caspase-3^–/–, ³⁶ caspase-9^–/–, ^{37 38} and Apaf-1^–/–, ^{39 40} with caspase-9 and apaf null mutations perinatally lethal and with caspase-3 null mutants dying within the first few weeks of birth. The development of the retina appears to be affected by these mutations. Although the same layering organization and neuronal cell types are seen in the caspase-3 ³⁶ and Apaf-1 ⁴⁰ null mutants, an increase in thickness of retinal layers indicated a lack of developmental apoptosis. 

Our results confirmed TUNEL-positive cells in the GCL during postnatal development. Although more than 50% of retinal ganglion cells are reported to die during postnatal development, instantaneous detection of <1% of degenerating profiles occurs, ² possibly because phagocytosis of dead retinal cells has a clearance time of only a few hours. ⁴¹  

Cleaved caspase-3 and caspase-9 immunoreactivity in the GCL from P3 to P5 indicate that the intrinsic pathway is necessary for the appearance of TUNEL-positive cells and pyknotic nuclei given the nuclear localization in many instances of the active caspases. Previous findings indicate that caspases are necessary for retinal ganglion cell development in the embryonic period. ⁴² Along with results shown in this study, there is strong evidence that caspases are essential for the normal development of the GCL. 

Strong evidence from in vitro studies also shows that survival of the neonatal GCL might be regulated by growth factors and electrical activity. ^{43 44} Previous work in vivo in which BDNF was injected into the superior colliculus of developing hamsters showed a large reduction in developmental apoptosis in the GCL. ⁴⁵ It has been suggested that immature neurons are competent to die by an active program of cell death unless they receive the essential trophic support to actively prevent it. In contrast, adult neurons must survive for the lifetime of the organism; as a result, the developmental default pathway is shut down. Consequently, growth factors are no longer needed to prevent death. ^{46 47} The results presented in this report certainly show that the key apoptotic machinery of the intrinsic pathway is “shut down ” during the development of the GCL. Consequently, susceptibility to death stimuli is reduced significantly compared with that in the immature GCL, in which the cell death program is active. 

 Supported by the Irish Research Council for Science Engineering and Technology under the Embark Initiative, the Health Research Board, and Science Foundation Ireland.

 Submitted for publication April 27, 2005; revised September 25, 2005; accepted January 19, 2006.

 Disclosure: D.P. McKernan, None; C. Caplis, None; M. Donovan, None; C.J. O’Brien, None; T.G. Cotter, 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.

 

Cell death occurs in the GCL of neonatal retinal explants after optic nerve transection. Hoechst-stained nuclei of P6 (a-c) and P60 (g, h) retinal explant sections are shown. DNA fragmentation in P6 (d-f) and P60 (i, j) was measured using TUNEL staining of same sections. After optic nerve transection, retinal explants were left for 0 hours (a, d, g, i), 24 hours (b, e, h, j), or 24 hours in the presence of 50 μM zVAD-fmk (c, f). White arrowheads: TUNEL-positive nuclei.

Figure 2.

 

Immunohistochemistry shows caspase activation in GCL of neonatal explants after optic nerve transection. Hematoxylin-stained sections of P6 (a-d) and P60 (e-h) retinal explant sections are shown. IHC of cleaved caspase-3 was carried out on explants after 0 hours (a, e) and 24 hours (c, g). IHC of cleaved caspase-9 was carried out on explants after 0 hours (b, f) and 24 hours (d, h). Secondary antibody-only controls were carried out on P6 explants for cleaved caspase-3 (i) and cleaved caspase-9 (j) antibodies to rule out nonspecific binding of the secondary antibody. Black arrows: cleaved caspase-positive cells.

Figure 3.

 

Cell death and caspase activation occur in the retinal GCL of neonatal explants 6 hours after treatment with 1 μM calcium ionophore A23187. Hoechst-stained nuclei (a, c), TUNEL staining (b, d), and hematoxylin-stained nuclei (e-h) of retinal explant sections are shown. After 6-hour treatment with A23187, P6 (b) and P60 (d) explants were analyzed for DNA fragmentation using TUNEL. IHC of cleaved caspase-3 was carried out on P6 (e) and P60 (g) explants after treatment. IHC of cleaved caspase-9 was carried out on P6 (f) and P60 (h). White arrowheads: TUNEL-positive nuclei; black arrows: cleaved caspase-positive cells.

Figure 4.

 

Percentage of TUNEL-positive cells in the retinal GCL of P6 and P60 explants in culture. (1) P6 0 hours; (2) P6 6 hours; (3) P6 12 hours; (4) P6 24 hours; (5) P6 24 hours + 50 μM zVAD; (6) P6 6 hours + 1 μM A23187; (7) P60 24 hours; (8) P60 48 hours; (9) P60 72 hours; (10) P60 96 hours; (11) P60 6 hours + 1 μM A23187.

Figure 5.

 

Percentage of cleaved caspase-3– and cleaved caspase-9–positive cells in the retinal ganglion cell layer of P6 and P60 explants in culture. (1) P6 0 hours; (2) P6 6 hours; (3) P6 12 hours; (4) P6 24 hours; (5) P6 6 hours + 1 μM A23187; (6) P60 24 hours; (7) P60 48 hours; (8) P60 72 hours; (9) P60 96 hours; (10) P60 6 hours + 1 μM A23187.

Figure 6.

 

Downregulation of caspase-3 and Apaf-1 expression in the GCL occurs during development. (a) Levels of caspase-3, caspase-9, and Apaf-1 mRNA were measured using semiquantitative PCR of P6 and P60 retinal ganglion cells that were purified by sequential immunopanning; tubulin was used as a loading control. Markers of INL (gfap, chx 10), ONL (bco, rhodopsin), and GCL (Thy 1) were measured to confirm the purity of the population. Ret indicates whole retinal population; IP, immunopanned population. (b) Levels of (pro)caspase-3 (a, d), (pro)caspase-9 (b, e), and Apaf-1 (c, f) protein in the GCL were measured in P6 (a–c) and P60 (d–f) whole eye sections using immunohistochemistry. Secondary antibody-only controls were carried out on P6 whole eye sections for caspase-3 (g), caspase-9 (h), and Apaf-1 (i) antibodies to rule out nonspecific binding of the secondary antibody.

Figure 7.

 

A caspase-dependent pathway is executed in the GCL during the peak of developmental apoptosis. Hoechst-stained nuclei (a–c) and hematoxylin-stained nuclei (g–l) of P3 (a, d, g, j), P4 (b, e, h, k), and P5 (c, f, i, l) whole eye sections are shown. Cleaved caspase-3 (g --i) and cleaved caspase-9 (j–l) were measured using immunohistochemistry. White arrowheads: TUNEL-positive nuclei; black arrows: cleaved caspase-positive cells.

The authors thank Theo Van Veen for the opportunity to learn retinal explantation techniques in his laboratory and Romeo Caffe for his training and advice in the subject. They thank members of the laboratory for helpful discussions. They also thank Science Foundation Ireland and the Health Research Board for funding this study. DPM is an IRCSET scholar funded under the EMBARK Initiative.