Target-derived and locally derived neurotrophins support retinal ganglion cell survival in the neonatal rat retina

Brain-derived neurotrophic factor (BDNF) and neurotrophin-4/5 (NT-4/5) protein and mRNA are found in the neonatal rat retina and also in target sites such as the superficial layers of the superior colliculus. Both neurotrophins support neonatal retinal ganglion cell survival in vitro. In vivo, injections of recombinant BDNF and NT-4/5 reduce naturally occurring cell death as well as death induced by removal of the contralateral superior colliculus. In the latter case, the peak of retinal ganglion cell death occurs about 24 h postlesion. We wished to determine: whether a similar time-course of degeneration occurs after selective removal of target cells or depletion of target-derived trophic factors, and whether ganglion cell viability also depends on intraretinally derived neurotrophins. Retinal ganglion cell death was measured 24 and 48 h following injections of kainic acid or a mixture of BDNF and NT-4/5 blocking antibodies into the superior colliculus and 24 h after intraocular injection of the same antibodies. Retinotectally projecting ganglion cells were identified by retrograde labeling with the nucleophilic dye diamidino yellow. We show that collicular injections of either kainic acid or BDNF and NT-4/5 blocking antibodies significantly increased retinal ganglion cell death in the neonatal rat 24 h postinjection, death rates returning to normal by 48 h. This increase in death was greatest following collicular injections; however, death was also significantly increased 24 h following intravitreal antibody injection. Thus retinal ganglion cell survival during postnatal development is not only dependent upon trophic factors produced by central targets but may also be influenced by local intraretinal neurotrophin release.     

Susceptibility to neurodegeneration in a glaucoma is modified by Bax gene dosage

In glaucoma, harmful intraocular pressure often contributes to retinal ganglion cell death. It is not clear, however, if intraocular pressure directly insults the retinal ganglion cell axon, the soma, or both. The pathways that mediate pressure-induced retinal ganglion cell death are poorly defined, and no molecules are known to be required. DBA/2J mice deficient in the proapoptotic molecule BCL2-associated X protein (BAX) were used to investigate the roles of BAX-mediated cell death pathways in glaucoma. Both Bax+/- and Bax-/- mice were protected from retinal ganglion cell death. In contrast, axonal degeneration was not prevented in either Bax+/- or Bax-/- mice. While BAX deficiency did not prevent axonal degeneration, it did slow axonal loss. Additionally, we compared the effects of BAX deficiency on the glaucoma to its effects on retinal ganglion cell death due to two insults that are proposed to participate in glaucoma. As in the glaucoma, BAX deficiency protected retinal ganglion cells after axon injury by optic nerve crush. However, it did not protect retinal ganglion cells from N-methyl-D-aspartate (NMDA)-induced excitotoxicity. BAX is required for retinal ganglion cell death in an inherited glaucoma; however, it is not required for retinal ganglion cell axon degeneration. This indicates that distinct somal and axonal degeneration pathways are active in this glaucoma. Finally, our data support a role for optic nerve injury but not for NMDA receptor-mediated excitotoxicity in this glaucoma. These findings indicate a need to understand axon-specific degeneration pathways in glaucoma, and they suggest that distinct somal and axonal degeneration pathways may need to be targeted to save vision.     

Autophagy promotes survival of retinal ganglion cells after optic nerve axotomy in mice

Autophagy is an essential recycling pathway implicated in neurodegeneration either as a pro-survival or a pro-death mechanism. Its role after axonal injury is still uncertain. Axotomy of the optic nerve is a classical model of neurodegeneration. It induces retinal ganglion cell death, a process also occurring in glaucoma and other optic neuropathies. We analyzed autophagy induction and cell survival following optic nerve transection (ONT) in mice. Our results demonstrate activation of autophagy shortly after axotomy with autophagosome formation, upregulation of the autophagy regulator Atg5 and apoptotic death of 50% of the retinal ganglion cells (RGCs) after 5 days. Genetic downregulation of autophagy using knockout mice for Atg4B (another regulator of autophagy) or with specific deletion of Atg5 in retinal ganglion cells, using the Atg5(flox/flox) mice reduces cell survival after ONT, whereas pharmacological induction of autophagy in vivo increases the number of surviving cells. In conclusion, our data support that autophagy has a cytoprotective role in RGCs after traumatic injury and may provide a new therapeutic strategy to ameliorate retinal diseases.     

Pharmacological Inhibition of Caspase-2 Protects Axotomised Retinal Ganglion Cells from Apoptosis in Adult Rats

Severing the axons of retinal ganglion cells (RGC) by crushing the optic nerve (ONC) causes the majority of RGC to degenerate and die, primarily by apoptosis. We showed recently that after ONC in adult rats, caspase-2 activation occurred specifically in RGC while no localisation of caspase-3 was observed in ganglion cells but in cells of the inner nuclear layer. We further showed that inhibition of caspase-2 using a single injection of stably modified siRNA to caspase-2 protected almost all RGC from death at 7 days, offering significant protection for up to 1 month after ONC. In the present study, we confirmed that cleaved caspase-2 was localised and activated in RGC (and occasional neurons in the inner nuclear layer), while TUNEL⁺ RGC were also observed after ONC. We then investigated if suppression of caspase-2 using serial intravitreal injections of the pharmacological inhibitor z-VDVAD-fmk (z-VDVAD) protected RGC from death for 15 days after ONC. Treatment of eyes with z-VDVAD suppressed cleaved caspase-2 activation by >85% at 3–4 days after ONC. Increasing concentrations of z-VDVAD protected greater numbers of RGC from death at 15 days after ONC, up to a maximum of 60% using 4000 ng/ml of z-VDVAD, compared to PBS treated controls. The 15-day treatment with 4000 ng/ml of z-VDVAD after ONC suppressed levels of cleaved caspase-2 but no significant changes in levels of cleaved caspase-3, -6, -7 or -8 were detected. Although suppression of caspase-2 protected 60% of RGC from death, RGC axon regeneration was not promoted. These results suggest that caspase-2 specifically mediates death of RGC after ONC and that suppression of caspase-2 may be a useful therapeutic strategy to enhance RGC survival not only after axotomy but also in diseases where RGC death occurs such as glaucoma and optic neuritis.     

Axonal damage, autophagy and neuronal survival

In recent years autophagy modulation has been shown to reduce or increase neuronal cell death in several models of neurodegeneration. How autophagy exerts these dual effects is currently unknown. Here we review recent evidence from our laboratory demonstrating that autophagy can protect the cell soma after axonal traumatic injury. Damage in the optic nerve induces retinal ganglion cell (RGC) death in glaucoma and other retinal diseases and is often modeled by axotomy of the optic nerve in laboratory animals. Using this well-known model of RGC degeneration we show that autophagy is strongly upregulated following the insult and before cell death. Enhancement of autophagy by pharmacological treatment with rapamycin decreases the number of degenerating neurons. Conversely, axotomy in Atg4B (-/-) mice increases the number of dying cells in the retinal ganglion cell layer. Similar findings were observed in Atg5 (flox/flox) mice following specific downregulation of the autophagy regulator ATG5 in RGCs, by intravitreal injection of a cre-expressing vector. Taken together, these findings point to a cytoprotective role of autophagy following axonal damage in vivo.     

Involvement of activated caspase-3-like proteases in N-methyl-D-aspartate-induced apoptosis in cerebrocortical neurons

Excessive activation of glutamate receptors mediates neuronal death in a number of neurodegenerative diseases. The intracellular signaling pathways that mediate this type of neuronal death are only partly understood. Following mild insults via NMDA receptor activation, central neurons undergo apoptosis, but with more fulminant insults, necrosis intervenes. Caspases are important in several forms of apoptosis in vivo and in vitro. Previously, we have demonstrated that caspases are important in excitotoxicity-mediated apoptosis of cerebrocortical neurons. To determine the possible activation of caspase-3 in NMDA-induced neuronal apoptosis, we used an affinity-labeling technique: Biotinylated N-acetyl-Asp-Glu-Val-Asp-aldehyde (DEVD.CHO) preferentially labels conformationally active caspase-3-like proteases, allowing us to visualize affinity-labeled caspases with streptavidin-fluorescein isothiocyanate under confocal microscopy. NMDA-induced apoptosis of cerebrocortical neurons was associated with a time-dependent increase in conformationally active caspase-3-like proteases. The activation of caspases was apparent within 20 min of NMDA stimulation and was localized primarily in the cytosol. However, following incubation of neurons for 18-24 h, conformationally active caspase-3-like proteases were also detectable in nuclei. Double labeling with propidium iodide to detect chromatin condensation indicated that affinity-labeled caspase-3-like proteases were specifically expressed in apoptotic cells. To further confirm this, we used an antibody specific for the conformationally active fragment of caspase-3 and found largely concordant results. Moreover, preincubation with DEVD.CHO prevented NMDA-induced apoptosis. Our results suggest that caspase-3-like proteases play a major role in excitotoxin-induced neuronal apoptosis.     

Spatial differences in active caspase-8 defines its role in T-cell activation versus cell death

Caspase-8, a cysteine-protease, initiates apoptosis when activated by death receptors. Caspase-8 is also essential for initiating T lymphocyte proliferation following T-cell antigen receptor (TCR) signaling. Given these disparate functions of caspase-8, we sought to determine whether this represented only a difference in the magnitude of caspase-8 activation, or different intracellular locations of active caspase-8. We demonstrate by high-resolution multicolor confocal laser scanning microscopy an aggregation of active caspase-8 within membrane lipid rafts in T cells stimulated with anti-CD3. This suggests that following TCR stimulation active caspase-8 physically interacts with lipid raft proteins, possibly to form a signaling platform. In contrast, Fas stimulation of T cells resulted in a much more profound activation of caspase-8 that was exclusively cytosolic. These confocal microscopic findings were confirmed using discontinuous sucrose gradient ultracentrifugation to isolate lipid raft versus cytosolic components. This sequestration model of caspase-8 activation was further supported by the observation that a classic caspase-8 substrate, BID, was not cleaved in CD3-stimulated T cells, but was cleaved after Fas engagement. Our data support a model that the location of active caspase-8 may profoundly influence its functional capacity as a regulator of either cell cycling or cell death.     

Activation of BMP-Smad1/5/8 signaling promotes survival of retinal ganglion cells after damage in vivo

While the essential role of bone morphogenetic protein (BMP) signaling in nervous system development is well established, its function in the adult CNS is poorly understood. We investigated the role of BMP signaling in the adult mouse retina following damage in vivo. Intravitreal injection of N-methyl-D-aspartic acid (NMDA) induced extensive retinal ganglion cell death by 2 days. During this period, BMP2, -4 and -7 were upregulated, leading to phosphorylation of the downstream effector, Smad1/5/8 in the inner retina, including in retinal ganglion cells. Expression of Inhibitor of differentiation 1 (Id1; a known BMP-Smad1/5/8 target) was also upregulated in the retina. This activation of BMP-Smad1/5/8 signaling was also observed following light damage, suggesting that it is a general response to retinal injuries. Co-injection of BMP inhibitors with NMDA effectively blocked the damage-induced BMP-Smad1/5/8 activation and led to further cell death of retinal ganglion cells, when compared with NMDA injection alone. Moreover, treatment of the retina with exogenous BMP4 along with NMDA damage led to a significant rescue of retinal ganglion cells. These data demonstrate that BMP-Smad1/5/8 signaling is neuroprotective for retinal ganglion cells after damage, and suggest that stimulation of this pathway can serve as a potential target for neuroprotective therapies in retinal ganglion cell diseases, such as glaucoma.     

Increased expression and activation of poly(ADP-ribose) polymerase (PARP) contribute to retinal ganglion cell death following rat optic nerve transection

Excessive activation of the nuclear enzyme poly(ADP-ribose) polymerase (PARP) by free-radical damaged DNA mediates necrotic cell death in injury models of cerebral ischemia-reperfusion and excitotoxicity. We recently reported that secondary retinal ganglion cell (RGC) death following rat optic nerve (ON) transection is mainly apoptotic and can significantly but not entirely be blocked by caspase inhibition. In the present study, we demonstrate transient, RGC-specific PARP activation and increased retinal PARP expression early after ON axotomy. In addition, intravitreal injections of 3-aminobenzamide blocked PARP activation in RGCs and resulted in an increased number of surviving RGCs when compared to control animals 14 days after ON transection. These data indicate that secondary degeneration of a subset of axotomized RGCs results from a necrotic-type cell death mediated by PARP activation and increased PARP expression. Furthermore, PARP inhibition may constitute a relevant strategy for clinical treatment of traumatic brain injury.     

Apoptosis in the developing visual system

Programmed cellular death is a widespread phenomenon during development of the nervous system. Two classes of molecules are particularly important in the context of apoptosis control in the nervous system: intracellular effectors homologous to the Caenorhabditis elegans Ced-3, -4, and -9 proteins, which in mammals correspond to the proteases of the caspase family, Apaf-1, and the members of the Bcl-2 protein family, and neurotrophic factors. Retinal ganglion cells lend a convenient model system with which to investigate apoptosis in central neurons during development as well as after injury. In this review, we discuss the role of these molecules in the control of programmed cellular death in the retinotectal system. Transgenic animal models and expression studies have shown that caspases, Bcl-2, Bax, and possibly Bcl-X are necessary players for the control of programmed cellular death in retinal ganglion cells. Bax and caspase 3 expression in retinal ganglion cells is upregulated after injury, and inhibition of Bax or caspase 3 increases the survival of injured retinal ganglion cells. Neurotrophins can support the survival of injured retinal ganglion cells, but this effect is transient. The physiological role of neurotrophins in the development of the retinocollicular system seems more related to the topographic refinement of retinocollicular projections, a process that is mediated, at least partially, by selective elimination of retinal ganglion cells making inappropriate topographic projections.     

A Critical role for Bim in retinal ganglion cell death

Optic nerve transection results in the death of retinal ganglion cells (RGCs) by apoptosis. Apoptosis is regulated by the Bcl-2 family of proteins, of which the Bcl-2 homology (BH3) -only proteins forms a subset. As BH3-only proteins have been shown to play a significant role in regulating cell death in the central nervous system, we wished to investigate the role of Bcl-2 interacting mediator of cell death (Bim), a prominent member of this protein family in the regulation of cell death in the RGC layer using in vitro retinal explants. In this study, we use an innovative retinal shaving procedure to isolate the cells of the ganglion cell layer to use for western blotting. Members of the BH3-only protein family are down-regulated during retinal development and are not normally expressed in the adult retina. Using this procedure, we demonstrate that Bim is re-expressed and its expression is increased over time following axotomy. Expression of Bad and Bik decreases over the same time course, whereas there is no indication that Bid and Puma are re-expressed. We show that explants from Bim knockout mice are resistant to axotomy-induced death when compared with their wild-type counterparts. Genetic deletion of Bim also prevents caspase 3 cleavage. The activity of Bim can be negatively regulated by phosphorylation. We show that the decrease of Bim phosphorylation correlates with a decrease in expression of survival kinases such as pAkt and pERK over the same time course. These results implicate Bim re-expression as being essential for axotomy-induced death of RGCs and that phosphorylation of Bim negatively regulates its activity in RGCs.     

Target‐derived and locally derived neurotrophins support retinal ganglion cell survival in the neonatal rat retina

Brain‐derived neurotrophic factor (BDNF) and neurotrophin‐4/5 (NT‐4/5) protein and mRNA are found in the neonatal rat retina and also in target sites such as the superficial layers of the superior colliculus. Both neurotrophins support neonatal retinal ganglion cell survival in vitro. In vivo, injections of recombinant BDNF and NT‐4/5 reduce naturally occurring cell death as well as death induced by removal of the contralateral superior colliculus. In the latter case, the peak of retinal ganglion cell death occurs about 24 h postlesion. We wished to determine: whether a similar time‐course of degeneration occurs after selective removal of target cells or depletion of target‐derived trophic factors, and whether ganglion cell viability also depends on intraretinally derived neurotrophins. Retinal ganglion cell death was measured 24 and 48 h following injections of kainic acid or a mixture of BDNF and NT‐4/5 blocking antibodies into the superior colliculus and 24 h after intraocular injection of the same antibodies. Retinotectally projecting ganglion cells were identified by retrograde labeling with the nucleophilic dye diamidino yellow. We show that collicular injections of either kainic acid or BDNF and NT‐4/5 blocking antibodies significantly increased retinal ganglion cell death in the neonatal rat 24 h postinjection, death rates returning to normal by 48 h. This increase in death was greatest following collicular injections; however, death was also significantly increased 24 h following intravitreal antibody injection. Thus retinal ganglion cell survival during postnatal development is not only dependent upon trophic factors produced by central targets but may also be influenced by local intraretinal neurotrophin release. © 2004 Wiley Periodicals, Inc. J Neurobiol 60: 319–327, 2004     

Metabotropic glutamate receptors are differentially regulated under elevated intraocular pressure

Glaucoma is a leading cause of blindness, ultimatively resulting in the apoptotic death of retinal ganglion cells. However, molecular mechanisms involved in ganglion cell death are poorly understood. While the involvement of ionotropic glutamate receptors has been extensively studied, virtually nothing is known about its metabotropic counterparts. Here, we compared the retinal gene expression of metabotropic glutamate receptors (mGluR) in eyes with normal and elevated intraocular pressure (IOP) of DBA/2J mice, a model for secondary angle-closure glaucoma using RT-PCR and immunohistochemistry. Elevated IOP in DBA/2J mice significantly increased retinal gene expression of mGluR1a, mGluR2, mGluR4a, mGluR4b, mGluR6 and mGluR7a when compared to C57BL/6 control animals, while mGluR5a/b and mGluR8a were decreased and no difference was observed for mGluR3 and mGluR8b. Specific antibodies detected an increase of mGluR1a and mGluR5a/b in both synaptic layers and in the ganglion cell layer of the retina under elevated IOP. Because ganglion cell death in DBA/2J mice occurs most likely by apoptotic mechanisms, we demonstrated up-regulation of mGluRs in neurons undergoing apoptosis. In summary, we support the idea that the specific gene regulation of mGluRs is a part of the glaucoma-like pathological process that develops in the eyes of DBA/2J mice.     

The Wnt Signaling Pathway Protects Retinal Ganglion Cell 5 (RGC-5) Cells from Elevated Pressure

The Wnt pathway is an essential signaling cascade that regulates survival and differentiation in the retina. We recently demonstrated that retinal ganglion cells (RGCs) have constitutively active Wnt signaling in vivo. However, the role of Wnt in RGC viability or function is unknown. In this study, we investigated whether Wnt protects the retinal ganglion cell line RGC-5 from elevated pressure, oxidative stress, and hypoxia injuries. Expression of RGC marker genes in the RGC-5 cultures was confirmed by immunocytochemistry and PCR. We demonstrated that the Wnt3a ligand significantly reduced pressure-induced caspase activity in RGC-5 cells (n = 5, P = 0.03) and decreased the number of TUNEL-positive cells (n = 5, P = 0.0014). Notably, Wnt3a-dependent protection was reversed by the Wnt signaling inhibitor Dkk1. In contrast, Wnt3a did not protect RGC-5 cells from oxidative stress or hypoxia. Furthermore, Wnt3a significantly increased growth factor expression in the presence of elevated pressure but not in the presence of oxidative stress and hypoxia. These results indicate that Wnt3a induces injury-specific survival pathways in RGC-5 cells, potentially by upregulating neuroprotective growth factors. Therefore, activation of the Wnt pathway by Wnt3a could be investigated further as a tool to develop novel molecular therapeutic strategies for the prevention of RGC death in retinal disease.     

Caspase-independent photoreceptor apoptosis in vivo and differential expression of apoptotic protease activating factor-1 and caspase-3 during retinal development

Apoptosis is the mode of photoreceptor cell death in many retinal dystrophies. Exposure of Balb/c mice to excessive levels of light induces photoreceptor apoptosis and represents an animal model for the study of retinal degenerations. Caspases have emerged as central regulators of apoptosis, executing this tightly controlled death pathway in many cells. Previously we have reported that light-induced photoreceptor apoptosis occurs independently of one the key executioners of apoptosis, caspase-3. This present study extends these results reporting on the lack of activation of other caspases in this model including caspases-8, -9, -7, and -1. Furthermore, photoreceptor apoptosis cannot be inhibited with the broad range caspase inhibitor zVAD-fmk indicating that light-induced retinal degeneration is caspase-independent. We demonstrate that cytochrome c does not translocate from mitochondria to the cytosol during photoreceptor apoptosis. We also show that during retinal development apoptotic protease activating factor (Apaf-1) protein levels are markedly decreased and this is associated with the inability to activate the mitochondrial caspase cascade in the mature retina. In addition, there is also a significant reduction in expression of caspases-3 and -9 during retinal maturation and these levels do not increase following light exposure. Finally, we show that the calcium-dependent proteases calpains are active during light-induced retinal degeneration and establish that the calcium channel blocker D-cis-diltiazem completely inhibits photoreceptor apoptosis.     

Cell-specific caspase expression by different neuronal phenotypes in transient retinal ischemia

Emerging evidence supports an important role for caspases in neuronal death following ischemia-reperfusion injury. This study assessed whether cell specific caspases participate in neuronal degeneration and whether caspase inhibition provides neuroprotection following transient retinal ischemia. We utilized a model of transient global retinal ischemia. The spatial and temporal pattern of the active forms of caspase 1, 2 and 3 expression was determined in retinal neurons following ischemic injury. Double-labeling with cell-specific markers identified which cells were expressing different caspases. In separate experiments, animals received various caspase inhibitors before the induction of ischemia. Sixty minutes of ischemia resulted in a delayed, selective neuronal death of the inner retinal layers at 7 days. Expression of caspase 1 was not detected at any time point. Maximal expression of caspase 2 was found at 24 h primarily in the inner nuclear and ganglion cell layers of the retina and localized to ganglion and amacrine neurons. Caspase 3 also peaked at 24 h in both the inner nuclear and outer nuclear layers and was predominantly expressed in photoreceptor cells and to a lesser extent in amacrine neurons. The pan caspase inhibitor, Boc-aspartyl fmk, or an antisense oligonucleotide inhibitor of caspase 2 led to significant histopathologic and functional improvement (electroretinogram) at 7 days. No protection was found with the caspase 1 selective inhibitor, Y-vad fmk. These observations suggest that ischemia-reperfusion injury activates different caspases depending on the neuronal phenotype in the retina and caspase inhibition leads to both histologic preservation and functional improvement. Caspases 2 and 3 may act in parallel in amacrine neurons following ischemia-reperfusion. These results in the retina may shed light on differential caspase specificity in global cerebral ischemia.