Role of Apoptosis Signal-Regulating Kinase in Stress-Induced Neural Cell Apoptosis in Vivo

【摘要】  Apoptosis signal-regulating kinase 1 (ASK1) is a mitogen-activated protein kinase kinase kinase that plays an important role in oxidative stress-induced apoptosis. In the present study, we used ASK1 knockout (KO) mice to examine the possibility that ASK1 is involved in the neural cell apoptosis that occurs during retinal development and ischemic injury. ASK1 was expressed in retinal neurons, including retinal ganglion cells (RGCs), but retinal structure and extent of cell death during development were normal in ASK1 KO mice. On the other hand, the strain was less susceptible to ischemic injury, and the number of surviving retinal neurons was significantly increased compared with that in wild-type mice. Interestingly, ischemia-induced phosphorylation of p38 mitogen-activated protein kinase (p38), which mediates RGC apoptosis, was almost completely suppressed in ASK1 KO mice. In such retinas, the numbers of cleaved caspase-3- and TUNEL-positive neurons were apparently decreased compared with those in wild-type mice. Furthermore, cultured RGCs from ASK1 KO mice were resistant to H2O2-induced apoptosis. Our findings suggest that ASK1 is involved in the neural cell apoptosis after various kinds of oxidative stress. Thus, inhibition of the ASK1-p38 pathway could be useful for the treatment of neurodegenerative diseases including glaucoma.
--------------------------------------------------------------------------------
Ischemic retinal injury is implicated in a number of pathological states, such as retinal artery occlusion, glaucoma, and diabetic retinopathy.1-4 Thus, it is necessary to understand the molecular regulation after ischemic injury. It has been proposed that free radicals are important mediators of damage caused by retinal ischemia.5,6 Reperfusion injury after ischemia appears paradoxical, but oxygen-derived and other free radicals are principally formed when reduced compounds that accumulate during ischemia are reoxidized.3 There is evidence that this free radical burst, produced during the early stage of reperfusion, overwhelms normal cellular antioxidant defense mechanisms, causing oxidative stress and a variety of types of tissue injury.7,8 However, the detailed molecular mechanism of oxidative stress-induced neural cell apoptosis remains to be elucidated.
Apoptosis signal-regulating kinase 1 (ASK1) is a mitogen-activated protein kinase (MAPK) kinase kinase, which is activated in response to various stimuli through distinct mechanisms.9 Overexpression of wild-type or constitutively active ASK1 induces apoptosis in various cell types through mitochondria-dependent caspase activation.10,11 On the other hand, oxidative stress and tumor necrosis factor-induced apoptosis are suppressed in ASK1-deficient cells.12 Recent studies have shown that ASK1 relays its apoptotic signals to the stress-activated MAPK family members, c-Jun N-terminal kinase (JNK) and p38 MAPK (p38).13 p38 is activated and phosphorylated by environmental stress such as H2O2 and UV-B radiation, as well as proinflammatory cytokines like interleukin-1 and tumor necrosis factor.14 Lipton and co-workers15,16 previously reported that axotomy of the optic nerve or intraocular injection of N-methyl-D-aspartate activates p38, which leads to neural cell apoptosis, especially for retinal ganglion cells (RGCs). These results suggest the possibility that the ASK1-p38 pathway induces retinal death in various pathological conditions. In the present study, we attempted to elucidate a role of ASK1 during retinal development and ischemic injury by analyzing ASK1 knockout (KO) mice. We provide evidence for a critical role of the ASK1-dependent cell death pathway in the retina.

【关键词】  apoptosis signal-regulating stress-induced apoptosis

Materials and Methods

Animals

Experiments were performed using ASK1 KO mice12 and their littermates in accordance with the Association for Research in Vision and Ophthalmology statement for the Use of Animals in Vision Research. Light intensity inside the cages ranged from 100 to 200 lux under a 12-hour light/dark cycle.

Histological Analysis

For developmental studies, mice were sacrificed at embryonic day 15 (E15), E17, postnatal day 0 (P0), P7, P14, and P90. For E15 and E17 mice, their heads were fixed in 4% paraformaldehyde (PFA) in 0.1 mol/L phosphate buffer (PB; pH 7.4) containing 0.5% picric acid for 2 hours at 4??C and embedded in paraffin wax. They were sectioned transversely at 7-µm thickness, mounted, and stained with hematoxylin and eosin. For P0, P7, P14, and P90 mice, they were anesthetized with diethylether and perfused transcardially with saline, followed by 4% PFA in 0.1 mol/L PB containing 0.5% picric acid at room temperature. The eyes were removed and postfixed in the same fixative for 2 hours at 4??C and embedded in paraffin wax. The posterior part of the eyes was sectioned saggitaly at 7-µm thickness, mounted, and stained with hematoxylin and eosin.

In Situ Hybridization

Sense and antisense cRNA probes corresponding to 431 bases (nucleotides 283 to 713) of the human Ask1 cDNA (GenBank accession no. D84476) were generated. This region is contained in the first coding exon of human Ask1 gene, which is deleted in ASK1 KO mice.12 Nucleotides 283 to 713 of the human Ask1 cDNA were amplified by PCR using human ASK1 HA cDNA in pcDNA3 as template. The following primers were designed to amplify this region: sense primer, 5'-GACGAGGGCATCACTTTCTC-3'; and antisense primer, 5'-GCGTTGTAAAAACGGTCCAG-3'. The PCR products were subcloned into pGEM-T easy Vector (Promega, Madison, WI) according to manufacturer??s protocol. Probes were synthesized from this cDNA template with SP6 (sense) or T7 (anti-sense) RNA polymerases (Roche, Basel, Switzerland) and digoxygenin-RNA labeling mix (Roche).

Mice were perfused transcardially with ice-cold 0.1 mol/L phosphate buffered saline (PBS; pH 7.4) and then with 4% PFA in 0.07 mol/L PB. Eyes were immediately enucleated and immersed in the same fixative for 2 hours at 4??C, followed by immersion in a sucrose solution (30% in PB) overnight at 4??C. Eyes were embedded in optimal cutting temperature compound (Tissue-Tek, Tokyo, Japan) and frozen on dry ice. The posterior parts of the eyes were sectioned at 10-µm thickness, collected on MAS-coated slides (Matsunami, Osaka, Japan), and processed as previously reported.17,18 Briefly, the sections were washed with PBS and treated with 2 µg/ml Proteinase K for 30 minutes at 37??C. They were fixed with 4% PFA in PB for 10 minutes, treated with 0.2 mol/L HCl for 10 minutes, and then treated with 0.25% acetic anhydride in 0.1 mol/L triethanol amine for 10 minutes. Hybridization was performed overnight at 50??C with 500 ng/ml probe in 3x standard saline citrate, 50% formamide, 0.12 mol/L PB, 1x Denhardt??s solution, 125 µg/ml tRNA, 100 µg/ml salmon sperm DNA, and 10% dextran sulfate. After washing probe, the sections were incubated with an alkaline phosphatase-conjugated anti-digoxygenin antibody (Roche) diluted 1:1000 in 0.1 mol/L Tris-HCl buffer containing 0.15 mol/L NaCl and 0.01% Tween-20. After washing, visible signal was detected with 5-bromo-4-chloro-3-indoxyl phosphate and nitro blue tetrazolium chloride substrate system (DAKO, Carpinteria, CA).

Immunohistochemistry

Sections were prepared as described for histological analysis section. They were treated with 0.3% H2O2 for blocking endogenous peroxidase. After washing in PBS, they were blocked with PBS containing 0.1% normal horse serum and 0.4% Triton-X 100 for 1 hour at room temperature. They were incubated overnight at 4??C with a rabbit polyclonal antibody against active caspase-3 (1:10,000; R&D Systems, Minneapolis, MN), active p38 (1:100; Promega), a goat polyclonal antibody against Brn3b (1.0 µg/ml; Santa Cruz Biotechnology, Santa Cruz, CA), a mouse monoclonal antibody against calretinin (1.0 µg/ml; Chemicon, Temecula, CA), calbindin (2.0 µg/ml; Sigma, Saint Louis, MO), protein kinase C (2.0 µg/ml; Sigma), glutamine synthetase (1.0 µg/ml; Chemicon), and glial fibrillary acidic protein (5.0 µg/ml; Progen, Heidelberg, Germany). The sections were then incubated with goat anti-rabbit or anti-mouse immunoglobulins conjugated to peroxidase labeled-dextran polymer (DAKO EnVision; DAKO), or biotinylated rabbit anti-goat IgG (Nichirei, Tokyo, Japan) and peroxidase-conjugated streptavidin (Nichirei) for 30 minutes and visualized with diaminobenzidine substrate kit (DAKO).

Ischemic Retinal Injury

Ischemia was achieved, and the animals were treated essentially as previously described.1,19-21 Briefly, we instilled sterile saline into the anterior chamber of the left eye at 120 cm of H2O pressure for 20 minutes while the right eye served as a nonischemic control. Seven days after reperfusion, animals were sacrificed, and the posterior part of the eyes was sectioned sagittally at 7-µm thickness through the optic nerve, mounted, and stained with hematoxylin and eosin. Ischemic damage was quantified in three ways. First, the thickness of the inner retinal layer (IRL; between the internal limiting membrane and the interface of the outer plexiform layer and the outer nuclear layer) was measured. Second, in the same sections, the number of neurons in the ganglion cell layer (GCL) was counted from one ora serrata through the optic nerve to the other ora serrata. Third, RGCs were retrogradely labeled from the superior colliculus (SC) with Fluoro-Gold (FG; Fluorochrome, Englewood, CO)15,22 soon after ischemic injury. Each midbrain was exposed, and after removing the skull using a micro drill, 2 µl of 4% FG was injected into the SC to mark the RGCs by retrograde axonal transport. Seven days after FG application, the eyes were enucleated and the retinas were detached and prepared as flattened whole mounts in 4% PFA in 0.1 mol/L PBS solution. GCL was examined in whole-mounted retinas with fluorescence microscopy to determine the RGC density. Four standard areas (0.04 mm2) of each retina at the point of 0.1 mm from the optic disk were randomly chosen, labeled cells were counted by observers blinded to the identity of the mice, and the average number of RGCs per square millimeter was calculated. For statistical analysis, six animals were used for each wild-type (WT) and ASK1 KO group. The changes in RGC number after ischemia were expressed as a percentage of the WT control eyes.

Terminal Deoxynucleotidyl Transferase Biotin-dUTP Nick End Labeling (TUNEL) Staining

Paraffin sections were treated with 10 µg/ml Proteinase K and then incubated in 0.26 U/µl terminal deoxynucleotidyl transferase in the supplied 1x buffer (Invitrogen, Carlsbad, CA) and 20 µmol/L biotinylated-16-dUTP (Roche) for 1 hour at 37??C. The sections were washed three times in PBS and blocked for 30 minutes with 10% normal donkey serum in PBS. The sections were then incubated with dichlorotriazinyl amino fluorescein-conjugated streptavidin (Jackson Immunoresearch, West Groove, PA) and diluted 1:500 in PBS for 30 minutes.

Quantification of Retinal Cell Apoptosis

Cell death was quantified by an ELISA (Roche) using a combination of antibodies recognizing histones and DNA, allowing the quantification of soluble nucleosomes in cell lysates.23 A mouse retina was homogenized in 100 µl of PBS containing 1 mmol/L phenylmethylsulfonyl fluoride and centrifuged at 15,000 x g for 10 minutes. A portion of the supernatant was used to quantify protein concentration, and the rest was processed. Absorbance values ranged between 0.1 and 1.0 and were normalized with respect to the values obtained with control retinas.

Immunoblot

Whole retinas were quickly removed at 3, 6, or 24 hours after ischemic injury. Retinas were homogenized in 200 µl of ice-cold 50 mmol/L Tris-HCl containing 150 mmol/L NaCl, 10 mmol/L NaF, 1 mmol/L Na3VO4, and the protease inhibitor cocktail (Roche). The protein concentration of the samples was determined by the Bio-Rad protein assay kit (Bio-Rad, Hercules, CA). Three micrograms of the protein were separated on an 11% sodium dodecyl sulfate-polyacrylamide gel and subsequently transferred to an Immobilon-P filter (Millipore, Billerica, MA). Membranes were incubated with anti-p38 (1:1000; Cell Signaling, Beverly, MA) or anti-phospho-p38 monoclonal antibody (1:1000; Cell Signaling). They were then incubated with a horseradish peroxidase-labeled anti-mouse IgG antibody (Amersham Bioscience, Piscataway, NJ) and visualized using an ECL Western blotting system (Amersham Bioscience).

Isolation of RGCs

RGCs were isolated by using a previously described two-step panning method.24-26 Briefly, P6 mice were killed to obtain approximately 20 enucleated eyes for each experiment. Isolated retinas were incubated in calcium- and magnesium-free Hanks?? balanced salt solution containing 0.5 mg/ml papain, 0.24 mg/ml cysteine, and 0.5 mmol/L EDTA for 10 minutes, and dissociated cells were incubated with rabbit anti-mouse macrophage antibody (Inter-Cell Technologies, Hopewell, NJ) for 5 minutes. Cell suspensions were treated for 30 minutes in 100-mm petri dishes coated with goat antibody against rabbit IgG heavy and light chains (Southern Biotechnology Associates, Birmingham, AL). Suspensions containing the cells that did not adhere to the petri dish were treated for 1 hour in 100-mm petri dishes coated with anti-mouse Thy 1.2 antibody (Serotec, Oxford, UK). Cells that adhered to the second petri dish were collected after separation by a 5-minute incubation with 0.05% trypsin and 0.53 mmol/L EDTA and were incubated in 96-well plates for cell death assay or in 8-well chamber slides for identifying the purity of isolated RGCs.

Culture plates or chamber slides were coated with 0.1 mg/ml poly-D-lysine (Sigma, St. Louis, MO) for at least 4 hours. They were then additionally coated overnight with 5 µg/ml Engelbreth-Holm-Swarm-laminin (Sigma). Medium developed by Politi et al27 for monolayer culture of mixed mouse retinal neurons, as modified for use in this experiment, consisted of Dulbecco??s modified Eagle??s medium with the additions of 1.6 µmol/L insulin, 40 nmol/L progesterone, 60 nmol/L selenite, 125 nmol/L transferrin, 200 µmol/L putrescine, 100 nmol/L hydrocortisone, 5.2 µmol/L cytidine-5'-diphosphocholine, 2.9 µmol/L cytidine-5'-diphosphoethanolamine, 80 ng/ml each of brain-derived neurotrophic factor and ciliary neurotrophic factor, 10 µmol/L forskolin, and 1% fetal bovine serum. Cells were incubated at 37??C in humidified 5% CO2 and 95% air.

Identification of Isolated RGCs

Mice were anesthetized by hypothermia, and RGCs were retrogradely labeled with 2 µl of 4% FG at P3. After 3 days, RGCs were purified and seeded in culture as described in the former section. In some experiments, we prepared whole-mounted retinas from FG-injected mice and determined whether injected FGs are really transported into the RGCs in a retrograde manner. FG-labeled RGCs were observed to be uniformly distributed in retinal tissue (data not shown). Three days after preparation, cultured RGCs were fixed with 10% formalin in PBS labeled with propidium iodide and assessed for double-labeling with FG.

Assessment of H2O2-Induced Cell Death in Cultured RGCs

WT and ASK1 KO RGCs were seeded at density of 5 x 104 cells/well and cultured with 0.1 ml of medium on a 96-well culture plate. After 2 days, they were unstimulated (control), or stimulated with 2.5 mmol/L H2O2 for 16 hours. After incubation, 50 µl of culture medium was collected and analyzed using a lactate dehydrogenase (LDH) cytotoxic test kit (Wako, Osaka, Japan), to estimate cell damage. LDH activity after H2O2 stimulation was normalized by subtracting control values from unstimulated cells.

Statistics

Data are presented as means ?? SD except as noted. When statistical analyses are performed, Student??s t-test or Mann-Whitney??s U-test were used to estimate the significance of the results. Statistical significance was accepted at P < 0.05.

Results

ASK1 Expression during Retinal Development

We first examined ASK1 mRNA expression during mouse retinal development by in situ hybridization analysis. ASK1 expression was obscure at E15 (Figure 1A) and E17 (data not shown) but clearly visible in the retinal GCL, which is still composed of multiple layers, at P0 (Figure 1B) . In postnatal mice, ASK1 expression was observed in the mature GCL and weakly in the inner nuclear layer (INL) (Figure 1, C and D) . In contrast, we did not detect staining in retinal sections from WT mice treated with control sense probe (Figure 1E) or those from ASK1 KO mice treated with antisense probe (Figure 1F) . These results demonstrate that ASK1 is expressed in retinal neurons including RGCs after birth and suggest that ASK1 may be involved in postnatal retinal development.

Figure 1. Expression of ASK1 during retinal development. ACF: In situ hybridization analysis of wild-type (ACE) and ASK1 KO (F) mice retina at E15 (A), P0 (B), P7 (C), and P90 (DCF). ASK1 expression was obscure at E15 (A) but appeared in the inner retina after P0 (BCD). Retinal sections from wild-type mice incubated with control sense probe (E) and those from ASK1 KO mice incubated with antisense probe (F) did not show staining. NBL, neuroblast layer; INL, inner nuclear layer. Scale bar = 50 µm.

Effect of ASK1 on Programmed Cell Death and Retinal Morphology during Development

Because ASK1 may be involved in programmed cell death (PCD) in the developing retina, we next examined whether the development of the retina is perturbed in ASK1 KO mice. Figure 2 illustrates the retinal development of ASK1 KO and their WT littermates. The time course of retinal development in ASK1 KO mice (Figure 2, B, D, and F) was normal compared with that in WT mice (Figure 2, A, C, and E) . In addition, retinal thickness and construction in adult mice (Figure 3A) were also normal in ASK1 KO mice (Figure 3B) . We further examined WT and ASK1 KO mouse retinas with various cell type-specific markers (data not shown). The expression pattern of Brn3b (ganglion cells), calretinin (ganglion and amacrine cells), calbindin (horizontal cells), protein kinase C (bipolar cells), glutamine synthetase (M?ller glial cells), and glial fibrillary acidic protein (astrocytes) showed no remarkable change between WT and ASK1 KO mice. We next quantified PCD in the whole retina at E15, E17, P0, and P7 using an ELISA method.23 Consistent with the normal retinal development, extent of retinal cell death in the developing retina was normal in ASK1 KO mice (data not shown). Thus, the absence of ASK1 signaling did not seem to be involved in PCD, differentiation, or migration of retinal cells or their assembly into defined layers.

Figure 2. Retinal development in ASK1 KO mice. ACF: Hematoxylin and eosin staining of retinal sections at E15 (A and B), P0 (C and D), and P14 (E and F) in WT (A, C, and E) and ASK1 KO (B, D, and F) littermates. Asterisks in A and B indicate lens. NBL, neuroblast layer; INL, inner nuclear layer; ONL, outer nuclear layer. Scale bar = 50 µm.

Figure 3. Quantitative analysis of ischemic retinal injury in ASK1 KO mice. ACD: Hematoxylin and eosin staining of control (A and B) and ischemic (C and D) retinal sections from WT (A and C) and ASK1 KO (B and D) mice. E and F: Thickness of the IRL (E) and the number of cells in the GCL (F) before and after ischemic injury as a percentage of the WT control eyes. Results of six independent experiments are presented as means ?? SD. Note the decreased neural cell loss in ASK1 KO mice. INL, inner nuclear layer; ONL, outer nuclear layer. Scale bar = 50 µm.

Effect of ASK1 on Ischemia-Induced Neural Cell Death

Because ASK1 may mediate stress-induced neural cell apoptosis in adult retina, we next examined the effect of ischemic injury, which is severe oxidative stress,3,20 in WT and ASK1 KO mice. Histological evaluation demonstrated a significant decrease in ischemic damage in ASK1 KO mice (Figure 3D) compared with their WT littermates (Figure 3C) . The thickness of the IRL after ischemia was decreased to 65 ?? 2% (n = 6) in WT, but 81 ?? 2% (n = 6) in ASK1 KO mice (Figure 3E) . Second, the percentage of surviving cells in the GCL was also increased in ASK1 KO (81 ?? 1%; n = 6) compared with WT mice (71 ?? 3%; n = 6) (Figure 3F) . These results suggest that inner retinal neurons are less susceptible to ischemic injury in ASK1 KO mice. Because GCL may contain RGCs, displaced amacrine cells, and other minor cell types,28 we next performed retrograde labeling of RGCs with FG and tried to determine the effect of ASK1 on RGC survival. Figure 4, ACH , shows the representative results of RGC labeling in ASK1 KO mice and their WT littermates. Consistent with the results of GCL counting (Figure 3F) , the extent of ischemia-induced RGC death seemed to be mild in ASK1 KO mice (Figure 4, D and H) compared with their littermates (Figure 4, C and G , respectively). Quantitative analysis revealed that FG-labeled (surviving) RGCs after ischemia were significantly increased in ASK1 KO mice (82 ?? 2%; n = 6) compared with their WT littermates (65 ?? 1%; n = 6) (Figure 4I) . These results demonstrate that loss of endogenous ASK1 has a protective effect on inner retinal neurons, especially for RGCs, after ischemic injury.

Figure 4. Quantitative analysis of retinal ganglion cell number in ASK1 KO mice. ACH: Retrograde labeling of retinal ganglion cells (RGCs) 7 days after ischemic injury (C, D, G, and H) and nonischemic fellow eyes (A, B, E, and F, respectively) in WT (A, C, E, and G) and ASK1 KO (B, D, F, and H) mice. E through H are enlarged images of A through D at approximately the same distance (0.1 mm) from the optic disk (eg, asterisk in A), respectively. I: RGC number before and after ischemic injury as a percentage of the WT control eyes. The number of labeled RGCs within four fields of identical size (0.04 mm2) at approximately the same distance (0.1 mm) from the optic disk were counted, and RGCs/mm2 were calculated. Results of six independent animals are presented as means ?? SD. Note the increased RGC resistance against ischemic injury in ASK1 KO mice. Scale bar = 500 µm in A through D and 100 µm in E through H.

Involvement of ASK1 Signaling in Ischemia-Induced Neuronal Apoptosis

We next analyzed apoptotic cells in the retina by TUNEL staining 1 day after ischemic injury. Control animals showed practically no signals in both WT and ASK1 KO mice (data not shown). In ischemic retinas, many TUNEL-positive cells were observed in the inner retina in WT (Figure 5A) but not in ASK1 KO mice (Figure 5B) . We previously demonstrated that ischemia-induced retinal cell apoptosis in the inner retina is executed mainly by caspase-3.20 In fact, immunohistochemical analysis showed many cleaved caspase-3-positive cells in the inner retina in WT mice (Figure 5C) , but the number was apparently decreased in ASK1 KO mice (Figure 5D) . Previous studies have suggested a possibility that ASK1 could activate p38 MAPK pathway, leading to oxidative stress-induced RGC apoptosis.13,15,16 To determine this possibility, we examined total and phosphorylated/activated p38 immunoreactivity after ischemic injury in WT and ASK1 KO mice. Immunoblot analysis revealed that the levels of phosphorylated p38 in nonischemic control retinas were very low in both strains (Figure 6A , top panel). However, ischemic injury strongly induced phosphorylation of p38 at 3, 6, and 24 hours after ischemic injury in WT retina (Figure 6B) . On the other hand, phosphorylated p38 immunoreactivity at 3 hours was low and essentially disappeared by 24 hours after ischemia in ASK1 KO mice (Figure 6, A and B) . Nonetheless, total p38 immunoreactivity (including both phosphorylated and nonphosphorylated forms) did not change after ischemia in both WT and ASK1 KO mice (Figure 6A , bottom panel). We also examined the localization of phosphorylated p38 immunoreactivity in the mouse retina. In control retinas, phosphorylated p38-positive cells were absent (data not shown). In contrast, 3 hours after ischemia, phosphorylated p38-like immunoreactivity was observed in the inner retina in WT (Figure 5E) but almost absent in ASK1 KO (Figure 5F) mice. These results suggest that ischemic injury induces neural cell apoptosis, especially for RGCs, through the activation of p38 MAPK and caspase-3 under the regulation of ASK1.

Figure 5. Localization of apoptotic cells, cleaved caspase-3, and phosphorylated p38 immunoreactivity in ischemic retina. A and B: Representative pictures of TUNEL staining in WT (A) and ASK1 KO (B) mice retina. CCF: Immunohistochemical analysis of cleaved caspase-3 (C and D) and phosphorylated p38 (E and F) 3 hours after ischemic injury. INL, inner nuclear layer; ONL, outer nuclear layer. Scale bar = 90 µm in A and B and 50 µm in C through F.

Figure 6. Immunoblot analysis of p38 immunoreactivity in ischemic retina. A: Three micrograms of retinal protein were prepared from ischemic eyes and nontreated fellow eyes and separated in 11% sodium dodecyl sulfate-polyacrylamide gel followed by immunoblot analysis using anti-p38 and anti-phspho-p38 antibodies. Representative image of three independent experiments are shown. B: Quantitative analysis of phosphorylated p38 immunoreactivity at 3, 6, and 24 hours after ischemic injury. Results of three independent experiments are presented as means ?? SD. Note the decreased phosphorylation of p38 in ASK1 KO mice. *P < 0.01.

Involvement of ASK1 in RGC Apoptosis

We next assessed the requirement of ASK1 for oxidative stress-induced apoptosis by using cultured RGCs from ASK1 KO mice. To determine the purity of our culture cells, mouse RGCs were retrogradely labeled with FG at P3, and culture cells were prepared at P6. Three days after preparation, these cultured cells were labeled with propidium iodide (PI) and assessed for double-labeling with FG. As shown in Figure 7 , most cultured cells (89 ?? 1%; n = 4) were double-labeled (yellow in Figure 7C ) with FG (green in Figure 7A ) and propidium iodide (red in Figure 7B ). By using such relatively pure cultured RGCs, we examined the effect of H2O2 on RGC death. H2O2-induced apoptosis was clear in WT RGCs, as determined by examining extracellular LDH activities (Figure 7D) . On the other hand, RGCs from ASK1 KO mice were more resistant to H2O2 compared with WT RGCs (Figure 7D) . These results clearly indicate a role of ASK1 in RGC apoptosis under stress conditions.

Figure 7. Quantitative analysis of H2O2-induced RGC death in ASK1 KO mice in vitro. ACC: Cultured RGCs were prepared from mouse retinas that had been retrogradely labeled with FG. Most cultured cells (89 ?? 1%, n = 4) were double labeled (yellow in C) with FG (green in A) and propidium iodide (PI; red in B), thus confirming the purity of our cultured RGCs. Scale bar = 100 µm. D: Extent of RGC death after stimulation with 2.5 mmol/L H2O2 was quantified by examining extracellular LDH activities. Note the decrease of H2O2-induced death in RGCs from ASK1 KO mice. Results of three independent experiments are presented as means ?? SD. *P < 0.05.

Discussion

In the present study, we demonstrated that the ASK1 MAPK kinase kinase signaling pathway is involved in neural cell apoptosis after ischemic retinal injury in vivo (Figures 3 and 4) . In addition, we found that both p38 MAPK and caspase-3 activation are clearly suppressed in ASK1 KO mice (Figures 5 and 6) . Our results are consistent with previous studies describing that activation of p38 induces RGC apoptosis after axotomy of the optic nerve or mediated by glutamate neurotoxicity.15,16 Ischemic injury is associated with excessive concentrations of glutamate, which results in overactivation of glutamate receptors and initiates a cascade of events that leads to necrosis and/or apoptosis. Consistently, selective inhibition of glutamate receptor antagonists2,3 or N-acetylated--linked-acidic dipeptidase, an enzyme responsible for the hydrolysis of neuropeptide N-acetyl-aspartyl-glutamate to N-acetyl-aspartate and glutamate, protects neurons in animal models of stroke and ischemic retinal injury.19,29 Furthermore, we previously showed that glutamate transporter KO mice are susceptible to ischemic retinal injury.1 Therefore, as well as p38 inhibition, appropriate control of glutamate concentration and/or glutamate receptor activation could be useful against ischemic injury.3,30

In addition to typical glutamate neurotoxicity, reactive oxygen species, such as superoxide and H2O2, are important mediators in damage caused by retinal ischemia.5,6 Our present study demonstrated that H2O2-induced RGC death is attenuated in ASK1 KO mice in vitro (Figure 7) . Under stress conditions, several stress proteins, such as ubiquitin, play important roles as heat shock- and stress-regulated proteins.31 However, recent studies have shown that ubiquitin promotes either cell survival or apoptosis.32-35 In neural cells, excessive ubiquitin induction after ischemic injury may rather lead to neural cell apoptosis by the degradation of antiapoptotic proteins and suppression of the transcription of prosurvival proteins.20 These results suggest a possibility that modulation of stress-responsive systems can become a "double-edged sword." Although our present results suggest its usefulness, further investigations revealing the precise effect of complete ASK1 inhibition in vivo will be needed. On the other hand, ASK1 activity and consequent activation of downstream signaling molecules can be negatively regulated by stimulating phosphoinositide-3 kinase-Akt pathway. It was reported that Akt-mediated phosphorylation of ASK1 inhibits its ability to activate JNK/p38 and protects cells against H2O2-induced apoptosis.36 Another possible factor that negatively regulates ASK1 activity is thioredoxin (Trx).37 In resting cells, ASK1 constantly forms an inactive complex with Trx, whereas on stimulation, ASK1 is dissociated from Trx and activated by conformational changes and covalent modifications. Recent studies have suggested the possibility that Trx is a kind of neurotrophic factor released from retinal pigmented epithelial cells and plays a crucial role in maintaining photoreceptors from light damage.38 Because other trophic factors such as ciliary neurotrophic factor and glial cell line-derived neurotrophic factor supplied from M?ller glial cells rescue neural cells,4,23,39,40 application of Trx, in combination with these trophic factors, may stimulate multiple cellular targets and activate separate mechanisms to rescue RGCs as well as photoreceptors during retinal degeneration. Further studies determining potential mechanisms other than ASK1-p38 pathway, which may be involved in the regulation of ischemia-induced neural cell death, will be needed.

In contrast to the dramatic alternation in adult neuronal apoptosis, extent of PCD, developmental process, structure, and cell type-specific population were all normal in ASK1 KO retina (Figure 2) . There are two periods of cell death in the developing retina.41 The first period occurs during E15 to E17 that is the main onset of neurogenesis, neural migration, and initial axon growth. In this period, at least part of retinal apoptosis is regulated by the low-affinity neurotrophin receptor p75, and most dying cells are observed in the neuroepithelium of the central retina, close to the optic nerve exit.42 We could detect ASK1 mRNA after P0, but not at E15 and E17. Thus, it seems to be natural that early PCD was normal in ASK1 KO mice. The second period coincides with the phase of tectal and thalamic innervation and synapse formation.43 In the rat, 50% of the total RGC population dies in the first postnatal week soon after reaching their target axons, the SC, and the lateral geniculate nucleus.44 However, loss of ASK1 did not change the extent of RGC death in this period (Figures 2C4) . Since ASK1 is a stress-responsive protein, ASK1-JNK/p38 pathway may not be active in the developing retina. In contrast, Bax, a member of the BclII family of cell death regulators, is involved in the control of developmental cell death, but not of photoreceptor degeneration.45 There is also another possibility that RGCs are regulated by multiple molecules in addition to ASK1, and single loss of ASK1 has no effect on RGC development. For example, previous studies have shown that brain-derived neurotrophic factor (BDNF) or neurotrophin-4/5 (NT-4/5) injected into the SC promotes the survival of neonatal RGCs.46,47 However, BDNF KO and NT-4/5 KO mice have normal RGC numbers.21,48 On the other hand, BDNF and NT-4/5 double KO mice showed delayed RGC development.21 Thus, there may be the functional redundancy between ASK1 and other molecules, as well as BDNF and NT-4/5, during development. Taken together, multiple survival and death signals seem to be differentially involved in RGC apoptosis according to the developmental stage and pathological conditions. Further studies determining the detailed mechanisms of RGC number control may provide important information for the progress of therapeutic methods in RGC protection and regeneration.

Ischemic retinal injury is now implicated in a number of pathological states, such as retinal artery occlusion, glaucoma, and diabetic retinopathy.1-4 Although further studies are needed to identify the precise effect of ASK1 inhibition in vivo, our findings raise intriguing possibilities for the management of neural degeneration, as well as infectious diseases and inflammation,49 by modifying the activity of the ASK1-p38 pathway.

【参考文献】
  Harada T, Harada C, Watanabe M, Inoue Y, Sakagawa T, Nakayama N, Sasaki S, Okuyama S, Watase K, Wada K, Tanaka K: Functions of the two glutamate transporters GLAST and GLT-1 in the retina. Proc Natl Acad Sci USA 1998, 95:4663-4666

Lipton SA: Retinal ganglion cells, glaucoma and neuroprotection. Prog Brain Res 2001, 131:712-718

Osborne NN, Casson RJ, Wood JP, Chidlow G, Graham M, Melena J: Retinal ischemia: mechanisms of damage and potential therapeutic strategies. Prog Retin Eye Res 2004, 23:91-147

Mitamura Y, Harada C, Harada T: Role of cytokines and trophic factors in the pathogenesis of diabetic retinopathy. Curr Diabet Rev 2005, 1:73-81

Pellegrini-Giampietro DE, Cherici G, Alesiani M, Carla V, Moroni F: Excitatory amino acid release and free radical formation may cooperate in the genesis of ischemia-induced neuronal damage. J Neurosci 1990, 10:1035-1041

Bonne C, Muller A, Villain M: Free radicals in retinal ischemia. Gen Pharmacol 1998, 30:275-280

Gilgun-Sherki Y, Rosenbaum Z, Melamed E, Offen D: Antioxidant therapy in acute central nervous system injury: current state. Pharmacol Rev 2002, 54:271-284

Liang D, Dawson TM, Dawson VL: What have genetically engineered mice taught us about ischemic injury? Curr Mol Med 2004, 4:207-225

Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, Moriguchi T, Takagi M, Matsumoto K, Miyazono K, Gotoh Y: Induction of apoptosis by ASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways. Science 1997, 275:90-94

Hatai T, Matsuzawa A, Inoshita S, Mochida Y, Kuroda T, Sakamaki K, Kuida K, Yonehara S, Ichijo H, Takeda K: Execution of apoptosis signal-regulating kinase 1 (ASK1)-induced apoptosis by the mitochondria-dependent caspase activation. J Biol Chem 2000, 275:26576-26581

Kanamoto T, Mota M, Takeda K, Rubin LL, Miyazono K, Ichijo H, Bazenet CE: Role of apoptosis signal-regulating kinase in regulation of the c-Jun N-terminal kinase pathway and apoptosis in sympathetic neurons. Mol Cell Biol 2000, 20:196-204

Tobiume K, Matsuzawa A, Takahashi T, Nishitoh H, Morita K, Takeda K, Minowa O, Miyazono K, Noda T, Ichijo H: ASK1 is required for sustained activations of JNK/p38 MAP kinases and apoptosis. EMBO Rep 2001, 2:222-228

Matsuzawa A, Ichijo H: Stress-responsive protein kinases in redox-regulated apoptosis signaling. Antioxid Redox Signal 2005, 7:472-481

Woodgett JR, Avruch J, Kyriakis J: The stress activated protein kinase pathway. Cancer Surv 1996, 27:127-138

Kikuchi M, Tenneti L, Lipton SA: Role of p38 mitogen-activated protein kinase in axotomy-induced apoptosis of rat retinal ganglion cells. J Neurosci 2000, 20:5037-5044

Manabe S, Lipton SA: Divergent NMDA signals leading to proapoptotic and antiapoptotic pathways in the rat retina. Invest Ophthalmol Vis Sci 2003, 44:385-392

Wada R, Sakata I, Kaiya H, Nakamura K, Hayashi Y, Kangawa K, Sakai T: Existence of ghrelin-immunopositive and -expressing cells in the proventriculus of the hatching and adult chicken. Regul Pept 2003, 111:123-128

Maseki Y, Nakamura K, Iwasawa A, Zheng J, Inoue K, Sakai T: Development of gonadotropes in the chicken embryonic pituitary gland. Zoolog Sci 2004, 21:435-444

Harada C, Harada T, Slusher BS, Yoshida K, Matsuda H, Wada K: N-acetylated--linked-acidic dipeptidase inhibitor has a neuroprotective effect on mouse retinal ganglion cells after pressure-induced ischemia. Neurosci Lett 2000, 292:134-136

Harada T, Harada C, Wang YL, Osaka H, Amanai K, Tanaka K, Takizawa S, Setsuie R, Sakurai M, Sato Y, Noda M, Wada K: Role of ubiquitin carboxy terminal hydrolase-L1 in neural cell apoptosis induced by ischemic retinal injury in vivo. Am J Pathol 2004, 164:59-64

Harada C, Harada T, Quah HM, Namekata K, Yoshida K, Ohno S, Tanaka K, Parada LF: Role of neurotrophin-4/5 in neural cell death during retinal development and ischemic retinal injury in vivo. Invest Ophthalmol Vis Sci 2005, 46:669-673

Yoshida K, Behrens A, Le-Niculescu H, Wagner EF, Harada T, Imaki J, Ohno S, Karin M: Amino-terminal phosphorylation of c-Jun regulates apoptosis in the retinal ganglion cells by optic nerve transection. Invest Ophthalmol Vis Sci 2002, 43:1631-1635

Harada T, Harada C, Nakayama N, Okuyama S, Yoshida K, Kohsaka S, Matsuda H, Wada K: Modification of glial-neuronal cell interactions prevents photoreceptor apoptosis during light-induced retinal degeneration. Neuron 2000, 26:533-541

Barres BA, Silverstein BE, Corey DP, Chun LL: Immunological, morphological, and electrophysiological variation among retinal ganglion cells purified by panning. Neuron 1988, 1:791-803

Kashiwagi K, Iizuka Y, Araie M, Suzuki Y, Tsukahara S: Effects of retinal glial cells on isolated rat retinal ganglion cells. Invest Ophthalmol Vis Sci 2001, 42:2686-2694

Kashiwagi K, Iizuka Y, Mochizuki S, Tsumamoto Y, Mishima HK, Araie M, Suzuki Y, Tsukahara S: Differences in nitric oxide production: a comparison of retinal ganglion cells and retinal glial cells cultured under hypoxic conditions. Brain Res Mol Brain Res 2003, 112:126-134

Politi LE, Lehar M, Adler R: Development of neonatal mouse retinal neurons and photoreceptors in low density cell culture. Invest Ophthalmol Vis Sci 1988, 29:534-543

Perry VH: Evidence for an amacrine cell system in the ganglion cell layer of the rat retina. Neuroscience 1981, 6:931-944

Slusher BS, Vornov JJ, Thomas AG, Hurn PD, Harukuni I, Bhardwaj A, Traystman RJ, Robinson MB, Britton P, Lu XC, Tortella FC, Wozniak KM, Yudkoff M, Potter BM, Jackson PF: Selective inhibition of NAALADase, which converts NAAG to glutamate, reduces ischemic brain injury. Nat Med 1999, 5:1396-1402

Tanaka K: Functions of glutamate transporters in the brain. Neurosci Res 2000, 37:15-19

Hershko A, Ciechanover A, Varshavsky A: The ubiquitin system. Nat Med 2000, 6:1073-1081

Dimmeler S, Breitschopf K, Haendeler J, Zeiher AM: Dephosphorylation targets Bcl-2 for ubiquitin-dependent degradation: a link between the apoptosome and the proteasome pathway. J Exp Med 1999, 189:1815-1822

Breitschopf K, Haendeler J, Malchow P, Zeiher AM, Dimmeler S: Posttranslational modification of Bcl-2 facilitates its proteasome-dependent degradation: molecular characterization of the involved signaling pathway. Mol Cell Biol 2000, 20:1886-1896

Daino H, Matsumura I, Takada K, Odajima J, Tanaka H, Ueda S, Shibayama H, Ikeda H, Hibi M, Machii T, Hirano T, Kanakura Y: Induction of apoptosis by extracellular ubiquitin in human hematopoietic cells: possible involvement of STAT3 degradation by proteasome pathway in interleukin 6-dependent hematopoietic cells. Blood 2000, 95:2577-2585

Yang Y, Fang S, Jensen JP, Weissman AM, Ashwell JD: Ubiquitin protein ligase activity of IAPs and their degradation in proteasomes in response to apoptotic stimuli. Science 2000, 288:874-877

Kim AH, Khursigara G, Sun X, Franke TF, Chao MV: Akt phosphorylates and negatively regulates apoptosis signal-regulating kinase 1. Mol Cell Biol 2001, 21:893-901

Saitoh M, Nishitoh H, Fujii M, Takeda K, Tobiume K, Sawada Y, Kawabata M, Miyazono K, Ichijo H: Mammalian thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK) 1. EMBO J 1998, 17:25596-25606

Tanito M, Kwon YW, Kondo N, Bai J, Masutani H, Nakamura H, Fujii J, Ohira A, Yodoi J: Cytoprotective effects of geranylgeranylacetone against retinal photooxidative damage. J Neurosci 2005, 25:2396-2404

Harada T, Harada C, Kohsaka S, Wada E, Yoshida K, Ohno S, Mamada H, Tanaka K, Parada LF, Wada K: Microglia-M?ller glia cell interactions control neurotrophic factor production during light-induced retinal degeneration. J Neurosci 2002, 22:9228-9236

Harada C, Harada T, Quah HM, Maekawa F, Yoshida K, Ohno S, Wada K, Parada LF, Tanaka K: Potential role of glial cell line-derived neurotrophic factor receptors in M?ller glia cells during light-induced retinal degeneration. Neuroscience 2003, 122:229-235

Bähr M: Live or let die: retinal ganglion cell death and survival during development and in the lesioned adult CNS. Trends Neurosci 2000, 23:483-490

Harada C, Harada T, Nakamura K, Sakai Y, Tanaka K, Parada LF: Effect of p75NTR on the regulation of naturally occurring cell death and retinal ganglion cell number in the mouse eye. Dev Biol, in press

Cellerino A, Bähr M, Isenmann S: Apoptosis in the developing visual system. Cell Tissue Res 2000, 301:53-69

Perry VH, Henderson Z, Linden R: Postnatal changes in retinal ganglion cell and optic axon populations in the pigmented rat. J Comp Neurol 1983, 219:356-368

Mosinger-Ogilvie J, Deckwerth TL, Knudson CM, Korsmeyer SJ: Suppression of developmental retinal cell death but not of photoreceptor degeneration in Bax-deficient mice. Invest Ophthalmol Vis Sci 1998, 39:1713-1720

Cui Q, Harvey AR: At least two mechanisms are involved in the death of retinal ganglion cells following target ablation in neonatal rats. J Neurosci 1995, 15:8143-8155

Ma YT, Hsieh T, Forbes ME, Johnson JE, Frost DO: BDNF injected into the superior colliculus reduces developmental retinal ganglion cell death. J Neurosci 1998, 18:2097-2107

Cellerino A, Carroll P, Thoenen H, Barde YA: Reduced size of retinal ganglion cell axons and hypomyelination in mice lacking brain-derived neurotrophic factor. Mol Cell Neurosci 1997, 9:397-408

Matsuzawa A, Saegusa K, Noguchi T, Sadamitsu C, Nishitoh H, Nagai S, Koyasu S, Matsumoto K, Takeda K, Ichijo H: ROS-dependent activation of the TRAF6-ASK1Cp38 pathway is selectively required for TLR4-mediated innate immunity. Nat Immunol 2005, 6:587-592

作者单位：From the Department of Molecular Neurobiology,* Tokyo Metropolitan Institute for Neuroscience, Fuchu, Tokyo; the Laboratory of Molecular Neuroscience, School of Biomedical Science and Medical Research Institute, Tokyo Medical and Dental University, Tokyo; the Department of Ophthalmology, School of M