Calcineurin cleavage is triggered by elevated intraocular pressure, and calcineurin inhibition blocks retinal ganglion cell death in experimental glaucoma

Abstract

Increased intraocular pressure (IOP) leads, by an unknown mechanism, to apoptotic retinal ganglion cell (RGC) death in glaucoma. We now report cleavage of the autoinhibitory domain of the protein phosphatase calcineurin (CaN) in two rodent models of increased IOP. Cleaved CaN was not detected in rat or mouse eyes with normal IOP. In in vitro systems, this constitutively active cleaved form of CaN has been reported to lead to apoptosis via dephosphorylation of the proapoptotic Bcl-2 family member, Bad. In a rat model of glaucoma, we similarly detect increased Bad dephosphorylation, increased cytoplasmic cytochrome c (cyt c), and RGC death. Oral treatment of rats with increased IOP with the CaN inhibitor FK506 led to a reduction in Bad dephosphorylation and cyt c release. In accord with these biochemical results, we observed a marked increase in both RGC survival and optic nerve preservation. These data are consistent with a CaN-mediated mechanism of increased IOP toxicity. CaN cleavage was not observed at any time after optic nerve crush, suggesting that axon damage alone is insufficient to trigger cleavage. These findings implicate this mechanism of CaN activation in a chronic neurodegenerative disease. These data demonstrate that increased IOP leads to the initiation of a CaN-mediated mitochondrial apoptotic pathway in glaucoma and support neuroprotective strategies for this blinding disease.

Glaucoma affects >66 million people worldwide and is the leading cause of irreversible blindness (1). Increased intraocular pressure (IOP) is the principal risk factor for developing glaucoma (2, 3), but the way in which elevated IOP leads to retinal ganglion cell (RGC) death is unknown. The primary treatment for glaucoma is lowering of the IOP, but many patients continue to lose vision despite aggressive treatment. Neuroprotective strategies have been heralded as an adjunctive approach to lowering of IOP, but target candidates have been lacking.

Although the triggers of RGC death are not clear, they have been shown to die by apoptosis in both animal models (4-6) and human disease (7, 8). Activation of two initiator caspase pathways, caspase 8 (9, 10) and caspase 9 (5, 9), has been reported in experimental glaucoma. Inhibition of the effector caspase, caspase 3, leads to protection of RGC in experimental glaucoma (11), but upstream mechanisms triggering apoptosis remain unknown.

Calcineurin (CaN) is a Ca²⁺ calmodulin-dependent protein phosphatase that is highly expressed in the central nervous system (12) and retina (13, 14). The 60-kDa CaN A subunit contains the catalytic domain (amino acids 20-340), the binding site for calmodulin, and a C-terminal autoinhibitory domain. A constitutively active form of CaN can be formed by cleavage or truncation of the regulatory domains, including the autoinhibitory domain (15, 16). This truncated form leads to apoptosis of cultured neurons, and cell death is inhibited by FK506, an immunophilin ligand that inhibits CaN (16).

Ca²⁺-induced apoptosis has been reported to depend on CaN-mediated dephosphorylation of the proapoptotic Bcl-2 family member, Bad. Dephosphorylated Bad translocates from the cytoplasm to the mitochondria, where it complexes with Bcl-2 or Bcl-x_L (17, 18), leading to cytochrome c (cyt c) release, changes in mitochondrial membrane potential, caspase activation, and apoptotic cell death (17, 19). Our current studies show that increased IOP is associated with the formation of a truncated form of CaN in the retina that is analogous to constitutively active constructs that have been characterized in in vitro experimental systems (15, 16, 20). Bad dephosphorylation, cyt c release, and RGC death ensue. CaN inhibition by oral FK506 prevents each of these effects and is neuroprotective for RGC and the optic nerve (ON) in eyes with elevated IOP. These results imply that both activated and cleaved CaN are mediators of apoptosis resulting from increased IOP and suggest neuroprotective strategies based on CaN inhibition for this chronic blinding disease.

Materials and Methods

Animals. All procedures concerning animals were in accordance with the statement of the Association for Research in Vision and Ophthalmology for the use of animals in research. Adult male Brown Norway rats (300-450 g, Charles River Laboratories) and 7-month-old female DBA/2J mice (The Jackson Laboratory) were housed in covered cages, fed with a standard rodent diet ad libitum, and kept on a 12-h light/dark cycle.

Backlabeling of RGCs. Anesthesia was induced by using a mixture of acepromazine maleate (1.5 mg/kg), xylazine (7.5 mg/kg), and ketamine (75 mg/kg) (all from Webster Veterinary Supply, Sterling, MA). Anesthetized rats were put in a stereotaxic apparatus (Kopf Instruments, Tujunga, CA), and the skull was exposed and leveled by using the lambda and bregma sutures as landmarks. The skull was thinned, and an injector was lowered into the superior colliculus 5.3 mm posterior to bregma, 1.5 mm lateral to midline, and 4.8 mm ventral to the skull surface. Two microliters of a 3% Fluorogold (Fluorochrome, Denver) solution in PBS with 10% DMSO was then injected over 10 min and repeated on the contralateral side of the brain (21). Animals were allowed 7 days for retrograde transport of Fluorogold before further experimental interventions.

Experimentally Induced Glaucoma. Unilateral elevation of IOP was produced by injecting hypertonic saline into aqueous veins of Brown Norway rats, as described by Morrison et al. (22). Briefly, hypertonic 1.9 M saline was injected into limbal aqueous humor collecting veins of the left eye. The right eye served as a control. In cases where the IOP was not elevated within 2 weeks, reinjection was performed in a different episcleral vein. A maximum of three injections were performed.

ON Crush. Rats were deeply anesthetized, and the ON was identified as it exited the globe. The ON was crushed three times for 10 sec each with a jeweler's forceps ≈2.5-3.0 mm posterior to the globe (23). Any rat with evidence of vascular compromise was excluded from further analysis. Sham-operated fellow eyes were used as controls.

IOP Determination. All IOP measurements were performed in conscious rats between 10:00 a.m. and 2:00 p.m. to minimize diurnal variability in IOP. IOP was measured by using a TonoPen XL tonometer (Medtronic Ophthalmics, Jacksonville, FL) (24). Fifteen readings were taken for each eye and averaged. Baseline IOP was obtained before the first saline injection and three times per week thereafter. Animals were subjected to a 5- or 10-day period of elevated IOP exposure. The beginning of this interval was defined as the day before the first recorded pressure elevation. As a measure for IOP exposure, we integrated IOP by calculating the area under the pressure-time curve (experimental eye-control eye), beginning with the day of the first saline injection.

Measuring IOP in DBA/2J mice was performed by direct cannulation as described (25). After anesthesia with a mixture of 90 mg/kg ketamine and 9 mg/kg xylazine, a glass micropipette tip with an external diameter of 50 μm and a length of 2 mm was introduced into the anterior chamber, taking care to avoid the lens, and connected to plastic tubing. A pressure transducer (CyQ low-pressure transducer, Cybersence, Nicholasville, KY) and a signal conditioner (CyQ104 signal conditioner, Cybersence) were used to obtain pressure recordings for a 1-min period when the eye had stabilized after microneedle insertion.

Drug Administration. Daily administration of FK506 (Fujisawa Pharmaceutical, Deerfield, IL) was performed by gavage in awake rats. FK506 was dissolved in PBS and dosed at 5 mg/kg per day. This dose was chosen because it has previously been shown to be effective after both peripheral nerve injury (26) and central nervous system injury (23).

Tissue Preparation. For RGC counting and ON grading, animals were killed by CO₂ inhalation followed by intracardiac perfusion with PBS and 4% paraformaldehyde (PFA). Eyes and ON were removed, and the ON were placed into fixative consisting of 2.5% glutaraldehyde and 2% formaldehyde in 0.1 M cacodylate buffer with 0.08 M CaCl₂ at 4°C. The eyes were postfixed in 4% PFA for 1 h, cryoprotected with serial sucrose dilutions, frozen in optimal cutting temperature compound (Tissue-Tek, Miles Diagnostic Division, Elkhart, IN), sectioned in their entirety at 16 μm, mounted on Superfrost Plus slides (VWR Scientific, West Chester, PA), and stored at -80°C. ON segments were washed in 0.1 M cacodylate buffer and postfixed in 2% aqueous OsO₄. The segments were dehydrated in graded alcohols and embedded in epon. One-micrometer sections were cut and stained with 1% toluidine blue in 1% borate buffer.

For protein analysis, animals were killed by CO₂ inhalation, and the retinas were immediately removed and homogenized in buffer containing 1 mM EDTA/EGTA/DTT, 10 mM Hepes (pH 7.6), 0.5% Igepal (Sigma), 42 mM KCl, 5 mM MgCl₂, 1 mM PMSF, and one tablet of protease inhibitors (Complete Mini, Roche Diagnostics) per 10-ml buffer. After a 15-min incubation on ice, samples were centrifuged at 21,000 rpm at 4°C for 30 min. The supernatant was stored at -80°C. The pellet containing the mitochondrial fraction was resuspended by adding 100 μl of buffer as above.

Western Blot. Protein concentrations were determined by using the Bio-Rad D_C Protein Assay. Proteins were separated by SDS/PAGE (Tris·HCL Ready-Gels, Bio-Rad) followed by transfer to polyvinylidene difluoride membranes (Immobilon-P, Millipore) and then blocked with 5% nonfat dry milk in Tris-buffered saline with Tween. The following primary antibodies were used: monoclonal mouse anti-CaN (1:250; BD Transduction Laboratories), polyclonal rabbit anti-CaN (1:1,000, Chemicon), monoclonal mouse antiphospho-Bad Ser-112 (1:1,000; Cell Signaling Technology, Beverly, MA), and monoclonal mouse anti-cyt c (1:1,000; BD Pharmingen). Secondary antibodies were rabbit peroxidase-conjugated (1:20,000; Jackson ImmunoResearch) and mouse peroxidase-conjugated (1:20,000; Jackson ImmunoResearch). After overnight incubation at 4°C, membranes were washed with Tris-buffered saline with Tween and incubated for 1 h in secondary antibody at room temperature. SuperSignal reagent (Pierce) was used to detect labeled protein, and membranes were exposed to HyperFilm (Amersham Pharmacia Biosciences). Anti α-tubulin (1:2,000; Abcam, Inc., Cambridge, MA.) and anti-COXIV antibody (1:1,000, Molecular Probes) were used as loading controls. Densitometry was carried out by using imagequant 1.2 (Molecular Dynamics).

Stereological Quantification of RGC. Each eye was sectioned in its entirety, and every eighteenth section was used for counting. To visualize RGC, sections were incubated in 1% BSA for 1 h at room temperature and then overnight at 4°C with an antibody specific for Fluorogold (1:200, Fluorochrome). Sections were rinsed 3× in PBS, incubated with secondary antibody [goat biotinylated anti-rabbit IgG (1:500, Vector Laboratories)] for 1 h at room temperature, rinsed 3× in PBS, and incubated in avidin-biotin-peroxidase complex (Vector Laboratories) in PBS for 30 min at room temperature. Coloration was performed in distilled deionized H₂O containing diaminobenzidine and hydrogen peroxide.

The total number of RGC in each retina was estimated by using unbiased stereology with the optical fractionator (27-29). Sections were selected systematically after a constant sampling intensity of every eighteenth section. RGC were counted manually by using the Olympus C.A.S.T. System (Version 2.3.1.2; Olympus, Albertslund, Denmark). A stereological algorithm was used to calculate the number of positive cells (27). Five percent of the area of the RGC layer was counted in control retinas and 10% of the area of the RGC layer in experimental glaucoma retinas on each sampled section to achieve an acceptable coefficient of error (CE) (30) and coefficient of variation (CV) (31).

ON Grading. ONs were assessed by using a modification of a previously reported grading classification (32). Sections for evaluation were taken from ≈2 mm posterior to the globe. Damage was assessed on a 1 (normal) to 5 (swollen and degenerating axons comprising nearly all of the ON) scale. A stereologically informed sampling scheme was used. An average of 20 regions per ON cross section viewed at ×60 were graded. Each region was photographed and graded by three masked independent observers. The grade for each ON for each observer was determined and an average for each ON calculated.

Statistical Analysis. ncss (NCSS Statistical Software, Kaysville, UT) was used to perform all statistical analyses except ON grading. Results are expressed as mean ± SD. Paired comparisons were performed by using a Student's t test, and multiple comparisons were performed by using a one-way ANOVA followed by a Fisher's probable least-squares difference post hoc test. Significance was assessed at the 0.05 level. For ON grading, the Mann-Whitney U test was performed with significance assessed at the 0.05 level (http://eatworms.swmed.edu/~leon/stats/utest.html).

Results

Elevated IOP Leads to Cleavage of CaN. IOP was surgically elevated in one eye of each rat with the fellow eye serving as a control. Peak IOP and integrated IOP were determined for all rats (Table 1, which is published as supporting information on the PNAS web site). For inclusion in the 5-day group, rats needed a peak IOP >40 mmHg (1 mmHg = 133 Pa). The criteria for inclusion in the 10-day group were peak IOP (>40-mmHg) and integrated IOP (>280-mmHg) days. Two hundred-mmHg days have been reported to be the point at which RGC loss begins and after which RGC loss progresses rapidly (33). Rats were killed after 5 or 10 days of elevated IOP for retinal protein analysis. Immunoblot analysis was performed by using a well characterized antibody that recognizes full-length (60 kDa) and cleaved (45 kDa) CaN (15, 34). The truncated product of CaN (45 kDa) was clearly seen in half of the eyes with elevated IOP for 5 days and in all eyes with elevated IOP for 10 days (Fig. 1 A and C).

Fig. 1.

CaN cleavage in experimental glaucoma. (A) Rats with elevated IOP in one eye for 5 or 10 days show the presence of the full-length CaN in all eyes and cleaved CaN (45 kDa) only in eyes with high IOP. Similar results were obtained in the DBA/2J mice. (B) Retinal protein from eyes of rats after ON crush. No cleaved CaN is observed in any eyes 3, 5, or 8 days after ON crush. (C) Rat summary data of the 45-kDa cleaved CaN band (n = 6 in all groups). Cleaved CaN was significantly increased in eyes with high IOP when compared with their fellow control eyes (†; 5 days, P < 0.02; 10 days, P < 0.005). In eyes with high IOP, the increase in cleaved CaN was significantly greater after 10 than after 5 days (^*, P < 0.02).

To investigate this finding in a second model of glaucoma, we studied the DBA/2J mouse model of spontaneous glaucoma. A total of 18 eyes (nine mice) of 7-month-old female DBA/2J mice were screened. Three mice had one eye with normal IOP and one eye with IOP >17. Another mouse had both eyes with IOP >17. This IOP distribution is similar to what has been reported in the literature (25). These four mice (eight eyes) were harvested for analysis for CaN cleavage. Eyes classified as having normal IOP (10 ± 1.5 mmHg, n = 3) or elevated IOP (18.6 ± 0.9 mmHg, n = 5) were evaluated by Western blot. Cleaved CaN was evident only in eyes with high IOP (Fig. 1 A).

We next examined another type of apoptosis-inducing injury to RGC, ON crush, to determine whether CaN cleavage was a nonspecific apoptosis-related event or due to elevated IOP. Rats underwent ON crush and were killed 3, 5, and 8 days later with processing of their retinas for protein analysis. By 8 days, there is substantial apoptosis and RGC loss (35). Fig. 1B shows immunoblot analysis at all three time points with no evidence of CaN cleavage at any time after ON crush. Taken together, these data support the idea that CaN cleavage is a part of the pathological process that develops in eyes with elevated IOP.

To verify that the observed 45-kDa band represented cleaved CaN, we used a second well characterized antibody recognizing a different region of the protein (15, 36). The first antibody (Transduction Laboratories, no. C26920) was made to an epitope in the middle of the peptide (amino acid 247-449), and the second (Chemicon, no. AB1696) was made to an epitope at the far N terminus (amino acid 8-18). Fig. 6, which is published as supporting information on the PNAS web site, shows results using the two different antibodies and demonstrates identical bands in rat eyes with elevated IOP.

CaN Inhibition Leads to Increased RGC Survival in Experimental Glaucoma. We next tested the hypothesis that CaN activation contributes to RGC death in experimental glaucoma. RGC loss after 10 days of elevated IOP was estimated with unbiased stereology. Fluorogold-positive RGC were counted from an average of 21 sections for each eye. After 10 days of elevated IOP, there were 56,815 ± 4,288 RGC (range, 51,000-63,029, mean CE = 0.06; CV = 0.075) compared with the normal IOP eyes that had 85,589 ± 5,818 RGC (range 78,000-93,600; mean CE = 0.07; CV = 0.068), indicating a 33.6 ± 2% loss of RGC (n = 9). Oral treatment with the CaN inhibitor FK506 (started at the time of the first measured elevation of IOP) did not affect the IOP (Table 1; P = 0.91) but did result in a significant increase in survival of RGC (Fig. 2; P = 0.00001). In FK506-treated rats after 10 days of elevated IOP, there were 75,301 ± 4,943 RGC (range, 65,850-81,000, mean CE = 0.06; CV = 0.066) compared with the normal IOP eyes that had 87,772 ± 4,894 RGC (range, 80,300-95,063; mean CE = 0.058; CV = 0.056), indicating a 16.7 ± 3% loss of RGC in animals treated with FK506 (n = 10).

Fig. 2.

Cell loss in experimental glaucoma and protection by FK506. Summary data showing the percentage of cells lost compared to the fellow eye after 10 days of elevated IOP in control animals (n = 9) and animals receiving daily FK506 (n = 10; ^*, P = 0.00001).

CaN Inhibition Leads to ON Preservation in Experimental Glaucoma. In a subset of rats, ONs were graded to assess for the degree of damage (Fig. 3). After 10 days of elevated IOP, significant ON damage had occurred (average grade 3.34; average peak IOP, 43.9 ± 1.7 mmHg; integrated IOP, 336.9 ± 37.7 mmHg days; n = 3). In rats receiving FK506, significantly less ON damage was observed after 10 days of elevated IOP (average grade, 1.21; average peak IOP, 43.7 ± 1.4 mmHg; integrated IOP, 338.7 ± 22.5; n = 4). This represents a significant protection of the ON when compared with rats not receiving FK506 (P < 0.03).

Fig. 3.

ON injury in experimental glaucoma and protection by FK506. (A) ON cross section from a normal IOP rat eye. The architecture of the ON and the axons is normal. (B) ON cross section from a rat eye with high IOP for 10 days. There are widespread degenerative changes in the ON. (C) ON cross section from a rat with normal IOP given daily FK506 for 10 days with normal appearance. (D) ON cross section from a rat eye with high IOP for 10 days given daily FK506 with significant preservation of ON tissue.

Elevated IOP Leads to Bad Dephosphorylation Through a CaN-Dependent Mechanism. To determine the mechanism by which CaN inhibition led to protection of RGC, we examined a known downstream effect of activated CaN, the dephosphorylation of Bad. Phosphorylated Bad (pBad) is sequestered in its inactive form in the cytosol by binding to 14-3-3, whereas dephosphorylated Bad is targeted to the mitochondria, where it causes cell death by binding Bcl-x_L and Bcl2 (17). Cytosolic retinal protein from rat eyes with and without elevated IOP was analyzed by using an antibody that recognizes only pBad. After 5 and 10 days, cytosolic levels of pBad were significantly reduced in eyes with elevated IOP (Fig. 4). Treatment with FK506 resulted in significantly higher levels of pBad in the cytosol of eyes with elevated IOP when compared with high IOP eyes of rats not receiving FK506 (P < 0.01; n = 6).

Fig. 4.

Bad is dephosphorylated in eyes with high IOP, and this effect is blunted by FK506. (A) pBad levels are decreased in eyes with elevated IOP for 5 and 10 days. (B) Summary data showing pBad in eyes with normal IOP (open bars) and high IOP (black bars). There is significantly less pBad in eyes with high IOP, but this effect is blunted by the administration of FK506. (n = 6 in all groups; ^*, P < 0.01 comparing pBad in high IOP rats eyes after 10 days with and without daily FK506; †, comparing pBad levels between normal and high IOP eyes at 5 days, P < 0.02; 10 days, P < 0.005; 10 days with FK506, P < 0.01).

Elevated IOP Leads to Mitochondrial Release of cyt c, and This Release Is Inhibited by FK506. Previous in vitro studies have shown that dephosphorylated Bad can translocate to the mitochondria and result in cyt c release and apoptosis. To further delineate the sequence of events occurring with elevated IOP, we examined the cytoplasmic and mitochondrial levels of cyt c. Fig. 5 shows mitochondrial and cytoplasmic levels of cyt c after 5 and 10 days of elevated IOP. Cyt c levels are seen to decrease in the mitochondrial fraction and increase in the cytoplasmic fraction of eyes with elevated IOP. Comparing levels between rats with and without treatment with FK506, there is significantly less release of cyt c into the cytoplasm (P = 0.001) and maintenance of mitochondrial cyt c levels (P < 0.01) in rats receiving FK506 (n = 6 in all groups).

Fig. 5.

cyt c is released into the cytoplasm in eyes with high IOP, and this effect is blunted by FK506. (A) A decrease in mitochondrial cyt c and an increase in cytoplasmic cyt c in eyes with elevated IOP. FK506 significantly diminishes the amount of cyt c released from the mitochondria into the cytoplasm. (B) Summary data for mitochondrial and cytoplasmic levels of cyt c in eyes with normal and high IOP for 5 or 10 days (with and without FK506). (n = 6 in all groups). ^*, comparing cyt c levels between eyes with high IOP in rats with and without FK506 in both mitochondrial (P = 0.01) and cytoplasmic (P = 0.001) fractions. †, comparing cyt c levels between eyes with normal and high IOP in the mitochondrial fraction (5 days, P < 0.01; 10 days, P = 0.005; 10 days FK506, P = 0.03) and cytoplasmic fraction (5 days, P < 0.01; 10 days, P < 0.005; 10 days FK506, P < 0.02).

Discussion

The findings reported in this manuscript provide information regarding the molecular events that accompany RGC death in experimental glaucoma. We show here that (i) CaN is cleaved under conditions of elevated IOP in two different models but not after ON crush to form a known constitutively active form of the phosphatase; (ii) increased IOP leads to Bad dephosphorylation, mitochondrial cyt c release, and RGC death; and (iii) CaN inhibition with oral FK506 [which is known to inhibit the phosphatase activity of both the full length and cleaved forms of CaN (16, 20)] blunts Bad dephosphorylation and cyt c release and results in significant protection of RGC and ONs in eyes with elevated IOP.

The three model systems used in these studies share apoptosis as the mechanism by which RGC die. The rat model of glaucoma is well established as a chronic model with characteristic sustained elevation of IOP, ON atrophy, and apoptotic RGC death with apoptosis starting ≈1 week after elevation of IOP (22, 37, 38). The spontaneous model of glaucoma in the DBA/2J mouse has also been demonstrated to have chronic elevation of IOP, ON atrophy, and RGC loss that occurs over several months (25, 39). Although these two models share elevated IOP as the proximate cause of RGC death, the way in which the IOP is raised is distinct in each model. The induction of elevated IOP in the rat usually occurs within 2-4 weeks of surgery, and in the mouse the IOP becomes spontaneously elevated between 6 and 9 months of age (25, 39). These two models differ in their tempo of RGC death, with death occurring at a slower rate in the DBA/2J mouse than in the rat. The third model, ON crush, is an acute model of RGC axon injury where RGC apoptosis is observed to start at 4 days and with >80% of RGC dying within 14 days (35). Our observation that CaN is cleaved in both models of elevated IOP but not after ON crush supports our hypothesis that CaN cleavage is an IOP-mediated cellular response that is not initiated by axon injury alone or as a nonspecific component of apoptotic cell death in the retina.

The exact mechanism of CaN cleavage in the retina after increased IOP is unknown. A truncated ≈45-kDa CaN band has been observed in the hippocampus 24 h after i.p. injection of kainate, corresponding to the period when hippocampal neurons are undergoing apoptosis (16). MALDI-TOF analysis showed cleavage of CaN at several sites, including a putative calpain cleavage site. Calpain inhibitors could blunt CaN cleavage and were neuroprotective. On the other hand, in vitro studies in cultured cells and with purified proteins have shown that a similarly sized truncated CaN can be produced by the action of either caspase 3 or caspase 7 (15). These cleaved forms of CaN lack the calmodulin-binding and autoinhibitory domains, leading to a constitutively active form requiring only physiologic amounts of Ca²⁺ and not requiring calmodulin binding for full activity (16, 20). We hypothesize that this process prolongs and enhances the actions of CaN in RGC under conditions of elevated IOP.

However CaN is activated, our data strongly suggest this truncated form of constitutively active CaN is present in the retina after exposure to increased IOP, leading to the hypothesis that CaN pathways play a role in RGC death in glaucoma. One prediction of this model is that CaN-mediated pBad dephosphorylation would occur with increased IOP, and we observe decreased pBad in the cytoplasm of retinas with increased IOP. A second prediction of this hypothesis is that inhibition of CaN would diminish pBad dephosphorylation and would be neuroprotective. The blunting of pBad changes and neuroprotection afforded by FK506 support this hypothesis and suggest CaN inhibition as an approach to protect RGC in glaucoma. Although our data with CaN inhibition indicate a significant neuroprotective effect after 10 days of elevated IOP, it will be important to determine whether the effect extends beyond the 10-day period. Given our observation that CaN is cleaved in the DBA/2J mice that have a much more chronic elevation of IOP, we postulate that long-term inhibition of CaN would continue to confer neuroprotection significantly beyond the initial period of IOP elevation. CaN activation is known to dephosphorylate nuclear factor of activated T cells (40), leading to changes in transcription levels and the initiation of signal transduction cascades involving the activation of caspases (41), which themselves could be additional mediators of RGC stress responses. Inhibition of these pathways could potentially contribute to neuroprotection as well.

Immunophilins are proteins that act as receptors for the immunosuppressant drugs FK506 and cyclosporin A. The immunosuppressant properties of FK506 are believed to be due primarily to CaN inhibition (42). FK506 binds FKBP12, leading to inhibition of CaN. We have demonstrated (23) that FKBP12 is present in RGC, and that FK506 reaches the retina after oral dosing, resulting in a significant reduction in retinal CaN activity (43). FK506 crosses the blood-brain barrier and has been shown to be neuroprotective in a wide range of models of acute apoptotic neuronal death in the central nervous system, including axotomy (23, 44), stroke (45, 46), cerebral ischemia (47-51), kainate-induced excitotoxicity (16), spinal ischemia (52), and spinal trauma (18, 53). Neuroprotection of FK506 has been suggested to be due in part to CaN inhibition and the consequent effects on pBad dephosphorylation. Other studies, using nonimmunosuppressant analogues of FK506, suggest that the neuroprotective effects could also be due to other mechanisms, such as induction of heat-shock proteins (54, 55). In the current studies, we have used FK506 as a CaN inhibitor and observe protection from pBad dephosphorylation and cyt c release, but we cannot rule out the possibility that part of the neuroprotective effect of FK506 is due to non-CaN-mediated mechanisms.

Although previous studies have focused directly on the activation of caspases in RGC after elevation of IOP, our findings in two in vivo models of experimental glaucoma provide evidence for the idea that pBad dephosphorylation and mitochondrial release of cyt c are important elements of this RGC apoptotic program. Our data suggest a novel mechanism: CaN activation due to generation of a constitutively active truncated form appears to be a critical intermediate in neuronal apoptosis. We hypothesize that CaN cleavage leads to a prolonged and enhanced effect of CaN activation in the cell. Our results are consistent with the findings of Wu et al. (16), who find CaN cleavage after acute excitotoxic neuronal injury in vivo. Taken together, these data suggest that CaN activation and cleavage may be important mechanisms in both acute and chronic neurodegeneration. Most importantly, these results point to a specific molecular pathway in RGC death in glaucoma that may be amenable to therapeutic intervention.