Nimodipine Enhancement of α2 Adrenergic Modulation of NMDA Receptor via a Mechanism Independent of Ca²⁺ Channel Blocking

Abstract

Purpose.: To further understand α2 receptor signaling in the retina and the mechanisms that mediate ocular beneficial effects of brimonidine (an α2 agonist) and nimodipine (an L-type Ca²⁺ channel blocker).

Methods.: The authors used in situ retinal ganglion cells (RGCs) in the isolated rat retina to characterize α2 modulation of NMDA receptor function and a rabbit retinal NMDA excitotoxicity model to verify in vitro findings under in vivo conditions. Electrophysiological (whole-cell patch clamp) recordings and Ca²⁺ imaging were used to characterize NMDA receptor function and to verify the effect of various Ca²⁺ channel blockers. In vivo drug application in rabbits was achieved by intravitreal injections.

Results.: Application of NMDA elicited a robust whole-cell inward current in individual in situ RGCs voltage clamped at −70 mV. Pretreatment with brimonidine significantly reduced NMDA-elicited currents in RGCs. This suppressive effect of brimonidine was substantially enhanced by background addition of nimodipine or isradipine, but not by diltiazem, verapamil, or cadmium. This effect of nimodipine was blocked by either a selective α2 antagonist, a cyclic adenosine monophosphate (cAMP) analogue, or an adenylate cyclase activator, indicating that nimodipine acts through the α2 receptor-G_αi–coupled pathway. Brimonidine protects RGCs in the rabbit excitotoxicity model. This brimonidine protection is also enhanced significantly by application of nimodipine but not of diltiazem.

Conclusions.: These in vitro and in vivo findings demonstrate a novel neural mechanism involving nimodipine enhancement of α2 signaling in RGCs. This nimodipine effect appears to be independent of its classic L-type Ca²⁺ channel-blocking action.

The α2 adrenergic receptor is a G-protein–coupled receptor ¹ expressed in retinal cells such as Müller cells ² and ganglion cells (RGCs). ³ It mediates neuroprotective effects of exogenous α2 agonists, such as brimonidine, in animal models of photoreceptor degeneration, ⁴ retinal ischemia, ^5–7 retinal excitotoxicity, ⁸ optic nerve injury, ^9,10 and glaucoma. ^8,11–13 In other parts of the central nervous system, α2 receptors modulate synaptic transmission through presynaptic modulation of neurotransmitter release by inhibiting N- or P/Q-type of Ca²⁺ channels, by activating K⁺ channels, or by reducing the number of active release sites. ^14–17 These presynaptic actions of the α2 receptor are mediated by G_iβγ subunits after G-protein activation. ^15,17,18  

Brimonidine, a selective α2 agonist, is the active ingredient of a US Food and Drug Administration–approved intraocular pressure (IOP)-lowering agent (Alphagan; Allergan, Irvine, CA) for glaucoma treatment. There are recent reports demonstrating that brimonidine increases pulsatile ocular blood flow in patients with primary open-angle glaucoma ¹⁹ and normal tension glaucoma. ²⁰ Brimonidine has also been shown to improve contrast sensitivity in glaucoma patients. ²¹  

The role of α2 adrenergic receptors in retinal information processing is largely unknown. We have shown recently that brimonidine, by activating α2 receptors, selectively modulates activity of L-type Ca²⁺ channels at the inner plexiform layer but not at the outer plexiform layer. ²² L-type Ca²⁺ channels play an important role in neurotransmitter (glutamate) release from photoreceptors and bipolar cells. ^23,24 We have also shown that brimonidine modulates NMDA-type ionotropic glutamate receptors (NMDA receptors) postsynaptically in RGCs. ⁸ This α2 modulation of NMDA receptors is mediated by G_iα through inhibition of cAMP production and is the major mechanism that underlies the neuroprotection of RGCs by exogenous α2 agonists in experimental glaucoma and retinal excitotoxicity models. ⁸  

Nimodipine is a dihydropyridine derivative that selectively blocks L-type Ca²⁺ channels. ²⁵ Nimodipine improves ocular blood flow in glaucoma patients ^26,27 and is neuroprotective in retinal ischemia/excitotoxicity models. ^28,29 These effects of nimodipine are attributed largely to its antagonistic effect on L-type Ca²⁺ channels either on retinal blood vessels or on retinal neurons. In addition, nimodipine has been shown to improve color contrast sensitivity in glaucoma patients, which is believed to be independent of its action on ocular blood flow. ²⁷  

In this study, we further investigated α2 receptor signaling in the retina and the mechanisms that underlie the retinal beneficial effects of brimonidine and nimodipine. We have demonstrated a novel effect of nimodipine to enhance α2 receptor signaling in RGCs. This effect of nimodipine appears to be independent of its L-type Ca²⁺ channel-blocking action and is mimicked by isradipine, another dihydropyridine, but not by other classes of L-type or general Ca²⁺ channel blockers such as diltiazem, verapamil, or cadmium. 

Materials and Methods

Isolated Retina and In Situ RGCs

We used in situ RGCs in the isolated rat retina, an ex vivo model, to investigate α2 modulation of NMDA receptor function. The present study was conducted in accordance with guidelines outlined in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and was approved by an institutional animal care and use committee. Male Brown Norway rats (275–300 g) were deeply anesthetized by intramuscular injection of ketamine (75 mg/kg) and xylazine (10 mg/kg). Immediately after enucleation of both eyes, rats were euthanatized by intracardial injection of sodium pentobarbital (120 mg/kg; Eutha-6; Western Medical Supply Co., Arcadia, CA). Retinas were carefully isolated, and a small piece (approximately 3 × 6 mm) was placed photoreceptor-side down in a recording chamber. A glass pipette filled with normal Ringer solution was used to expose the somas of in situ RGCs by mechanically removing a small portion of the inner limiting membrane and cleaning the surface of the cell membrane for whole-cell patch clamp. 

Whole-Cell Patch Clamp and Drug Application to In Situ RGCs

The isolated retina was perfused continuously with normal Ringer by both whole chamber and local perfusion systems to accelerate delivery and removal of the test agents. To better monitor NMDA receptor activity, we recorded simultaneously both the NMDA-induced whole cell current and the cytosolic Ca²⁺ signal. The normal Ringer contained 120 mM NaCl, 3 mM KCl, 2 mM CaCl₂, 1.2 mM MgSO₄, 0.5 mM KH₂PO₄, 10 mM d-glucose, 26 mM NaHCO₃, 0.005 mM strychnine, and 0.02 mM SR95530. In those experiments in which the effect of nimodipine on α2 signaling was evaluated, nimodipine Ringer (10 μM nimodipine added to the normal Ringer) was used in both whole-chamber and local perfusion. Conventional methods for whole-cell patch clamp were used. RGCs were voltage-clamped at −70 mV. After whole-cell patch clamp was established, normal Ringer was replaced with a 0 Mg²⁺ Ringer, which was made by removing MgSO₄ from and adding 0.02 mM ascorbic acid to the normal Ringer. The normal intracellular (patch pipette) solution contained (in mM): 125 mM CsCH₃SO₃, 1 mM MgCl₂, 15 mM TEA Cl, 10 mM HEPES, 4 mM ATP-Mg, 0.5 mM GTP-Na₃, 12 mM phosphocreatine, 5 mM QX-314, and 0.1 mM fluo-4. NMDA (100 μM NMDA + 5 μM glycine) was applied for 4 to 8 seconds through a multichannel local perfusion system. In experiments using brimonidine, atipamezole, or both, these two agents were added to both the whole-chamber and local perfusion. Ca²⁺ images were obtained with a spinning disc confocal system (Nipkow; Solamere Technology Group, Salt Lake City, UT) mounted on a fixed-stage upright microscope (BX51WI; Olympus America Inc., Melville, NY). 

Confocal Ca²⁺ Imaging in Live Rat Retinal Slices

After retinas were carefully isolated, a small piece of retina (approximately 2 × 6 mm) was placed vitreal side down onto a piece of black filter paper (catalog no. habp04700; Millipore Corp., Bedford, MA). The retina and filter paper were then sliced at approximately 250-μm intervals. Retinal slices were carefully transferred (cross-section facing up) to recording chambers and were securely positioned by placing the filter paper to which the retinal slices were attached on two tracks of vacuum grease, so that all retinal layers, including OPL and IPL, could be viewed and imaged with a fixed-stage microscope (BX50WI; Olympus America Inc.). 

After retinal slices were mounted in the perfusion chambers, the chambers were filled with dye-loading medium and placed on a shaker (model 1304; Laboratory-line Instruments, Melrose Park, IL) for 50 minutes at room temperature. Then one of the chambers containing dye-loaded retinal slices was mounted on the microscope stage. The slices were continuously perfused with normal rat Ringer through both multichannel bath and local perfusion systems. Flow rates were 3.5 mL/min and 0.3 mL/min for the bath and local perfusion systems, respectively. 

Cytosolic Ca²⁺ signals were elicited by a brief (5- to 8-second) local perfusion of a high K⁺ Ringer. The light path shutter, high K⁺ perfusion, and Ca²⁺ imaging were controlled precisely by software (P-Clamp 8; Molecular Devices Corp., Sunnyvale, CA). Ca²⁺ imaging was conducted with a spinning disc confocal system (Nipkow; Solamere Technology Group) equipped with a high-sensitivity, high-speed intensified CCD camera (XR/Mega 10; Stanford Photonics Inc., Palo Alto, CA). Retinal Ca²⁺ images were acquired at two to four frames per second by a 60× long-working distance water-immersion objective (LUMPlan FI 60×/0.90 W; Olympus America Inc.). 

Rabbit Retinal Excitotoxicity Model

In vivo excitotoxic insult to RGCs was induced by a single intravitreal injection of 3.6 μmol NMDA (in 50-μL vehicle containing 50% wt/vol (2-hydroxypropyl)-β-cyclodextrin solution). The effects of brimonidine (3.6 nmol), atipamezole (24 nmol), and nimodipine (24 nmol), or a combination of these agents, on NMDA-induced RGC injury were evaluated by coinjection with NMDA. In addition, these agents were also injected at 1 hour before and 24 hours after NMDA injection. In the NMDA alone group, 50 μL vehicle was also injected at 1 hour before and 24 hours after NMDA injection to control for any injection-induced effects on cell survival. Nimodipine was first dissolved in 100% dimethyl sulfoxide (DMSO) as a 100-mM stock solution and then diluted with vehicle to an injection solution of 480 μM. Fifty microliters of the injection solution was intravitreally injected. The final vitreal DMSO concentration was estimated to be approximately 0.02%. 

Thirteen days after intraocular NMDA application, RGCs were fluorescence labeled by intravitreal injection of the fluorescent nuclear dye 4′-6-diamidino-2-phenylindole (DAPI; 1 mM, 50 μL). Twenty-four hours later, rabbits were euthanatized, and the retinas were isolated (see Fig. 5 legend). RGC injury was evaluated using a fixed-stage upright microscope (BX51WI; Olympus America Inc.) equipped with an automated microscope stage (H101A; Prior Scientific Inc., Cambridge, UK), a high-sensitivity CCD camera (Orcaer; Hamamatsu Photonics, Hamamatsu, Japan), and imaging software (Image-Pro Plus; Media Cybernetics, Bethesda, MD). Neurons in the ganglion cell layer were counted within an 8-mm diameter sample of the central retina extending inferiorly from the inferior rim of the optic nerve head. All retinal samples from different eyes were obtained from the same region, and neurons were counted within 25 fields composing a 5 × 5 array (see Fig. 5A). 

Statistical Analysis

Group data are expressed as mean ± SEM. Statistical comparisons were made using a two-population student t-test (Origin; OriginLab, Northampton, MA). P < 0.05 was chosen to indicate a statistically significant difference. 

Results

Nimodipine Enhances α2 Receptor–Mediated Suppression of NMDA-Elicited Responses in In Situ RGCs

In the mammalian retina, the NMDA receptor is expressed mostly in RGCs. ^30,31 Brief application of 100 μM NMDA (with 5 μM glycine) elicited a robust inward current (Fig. 1) that was blocked by D-AP5, a specific NMDA receptor antagonist (data not shown). (For simplicity, we use “NMDA elicited” current instead of “NMDA/glycine elicited” current in the text and the figure legends.) NMDA responses were sensitive to Mg²⁺ block and required glycine as a co-agonist (data not shown). Pretreatment with brimonidine caused a significant reduction (P < 0.01) of both NMDA-induced currents and Ca²⁺ signals (Fig. 1), similar to what was observed in our earlier study. ⁸ The effects of the test agents on the NMDA-elicited current and Ca²⁺ signal are similar. For simplicity, we show only the effect on NMDA-elicited currents here. Perfusion for 15 minutes with the Ringer solution containing 10 μM nimodipine enhanced significantly (Fig. 1B; P < 0.01) the suppressive effect of brimonidine on the NMDA-elicited current (Fig. 1). Interestingly, this enhancement was more pronounced on the Ca²⁺ signal compared with the whole-cell current. With nimodipine in the background, brimonidine application almost completely suppressed NMDA-elicited Ca²⁺ signals (data not shown). 

Figure 1.

 

Nimodipine enhances brimonidine-induced suppression of NMDA receptor function in in situ RGCs. (A) Brimonidine (brimo, 3 μM) suppressed the NMDA (100 μM)-elicited whole cell current in the absence and presence of nimodipine (nimo, 10 μM). Vertical and horizontal calibration bars represent 100 pA and 5 seconds, respectively. (B) Normalized group data from seven RGCs. To have more accurate comparison between nimo control and nimo+brimo, nimo+brimo and wash data sets were renormalized against nimo control and plotted in dark gray on the right. **P < 0.01 (P = 0.00329) between brimo alone and nimo+brimo.

All data shown in Figure 1 were from the same group of RGCs. Similar enhancement was also observed between separate RGC groups (brimonidine in normal Ringer vs. brimonidine in nimodipine Ringer). Because the success rate was significantly lowered to measure the effects of brimonidine under two conditions (>40 minute stable whole-cell recordings were required), we used different RGC groups in the subsequent experiments for other Ca²⁺ channel blockers. 

Nimodipine Effect Requires α2 Receptor Activation and Is Blocked by Agents That Preserve Intracellular cAMP

The nimodipine-induced enhancement of the brimonidine effect on NMDA responses seems to be independent of its classic action as an L-type Ca²⁺ channel blocker. Because nimodipine was applied for an extended time (15 minutes), its Ca²⁺ channel-blocking action was already present in the control response (Fig. 1A, lower panel). To determine the mechanism for this action of nimodipine, we tested the effect of atipamezole in the presence of nimodipine. Pretreatment with atipamezole completely eliminated nimodipine enhancement (Figs. 2A, 2C), indicating that nimodipine effect requires activation of the α2 receptor. 

Figure 2.

 

Nimodipine's enhancement requires α2 receptor activation and the cAMP pathway intracellular signaling. (A, C) Atipamezole (a selective α2 antagonist) abolished nimodipine's enhancement of brimonidine (brimo, 3 μM)–induced suppression of NMDA receptor function in in situ RGCs. (A) Traces of whole cell currents recorded under two different experimental conditions: nimodipine (nimo, 10 μM)+atipamezole (ati, 10 μM) and nimo+ati+brimo. (C) Group data from 8 RGCs (n = 8). (B) Forskolin (an adenylate cyclase activator) abolished nimodipine's enhancement of brimonidine-induced suppression of NMDA receptor function in in situ RGCs. Traces of whole cell currents recorded under two different experimental conditions: nimodipine (nimo, 10 μM) in combination with intracellular forskolin (forsk, 10 μM) and nimodipine (nimo, 10 μM)+intracellular forskolin+brimonidine (brimo, 3 μM). (D) Normalized group data obtained in the absence (n = 9) and presence of intracellular forskolin (10 μM, n = 4) or intracellular Sp-cAMPS (cAMP, 200 μM, n = 8). **P < 0.01 between nimodipine control and other data sets. (A, B) Vertical and horizontal calibration bars represent 100 pA and 5 seconds, respectively.

The α2 receptor is coupled to G_i/G_o. It can signal through a number of effector mechanisms, including modulation of adenylate cyclase (AC) activity, Ca²⁺ and K⁺ channels, and Na⁺/H⁺ exchangers. ^14–17 The inhibition of AC occurs through G_iα/G_oα (a cAMP-dependent signaling pathway), whereas modulation of Ca²⁺ and K⁺ channels or the number of active release sites is believed to be through G_iβγ/G_oβγ-mediated direct action (cAMP-independent pathways). ^17,18  

We showed previously ⁸ that α2 modulation of NMDA receptor function is mediated by G_iα/G_oα through the inhibition of AC. To further understand the mechanism that underlies nimodipine action, we tested agents known to affect AC activity or intracellular cAMP concentration. To limit the activity of these agents to the recorded RGCs, the agents were delivered intracellularly through the patch electrode. Intracellular application of forskolin, an AC stimulator, abolished completely the nimodipine-induced enhancement of the brimonidine effect on both the NMDA-elicited whole cell current and the cytosolic Ca²⁺ signal (Figs. 2B, 2D). Similarly, the nimodipine effect was also completely eliminated by intracellular application of Sp-cAMPS, a hydrolysis-resistant cAMP analogue (Fig. 2D). 

By themselves, Sp-cAMPS and forskolin did not have a significant effect on the NMDA-elicited current or Ca²⁺ signal. ⁸ Taken together, the results shown in Figure 3 demonstrate strongly that the nimodipine effect requires activation of α2 receptors and is mediated by a reduction of intracellular cAMP production. This is consistent with a G_αi-mediated inhibition of AC. Preserving intracellular cAMP concentration either by directly stimulating AC or by adding an exogenous cAMP analogue can block the nimodipine effect. 

Figure 3.

 

Isradipine, but not diltiazem, verapamil, or cadmium (Cd²⁺), mimicked the effect of nimodipine in in situ RGCs. (A) Traces of whole cell currents recorded in the presence of isradipine (israd, 10 μM) and isradipine+brimonidine (brimo, 3 μM). Vertical and horizontal calibration bars represent 100 pA and 5 seconds, respectively. (B) Normalized group data obtained with brimonidine (brimo, 3 μM, n = 7) alone and in combination with either nimodipine (nimo, 10 μM, n = 9), isradipine (israd, 10 μM, n = 6), diltiazem (DTZ, 20 μM, n = 9), verapamil (verap, 10 μM, n = 6), or cadmium (Cd²⁺, 100 μM, n = 9). **P < 0.01 between brimonidine control and other data sets.

Nimodipine Effect Is Mimicked by Isradipine but Not by Other Classes of L-type or General Ca²⁺ Channel Blockers

Most commonly used L-type Ca²⁺ channel blockers can be classified into three chemical groups: dihydropyridines (nimodipine, isradipine), benzothiazepines (diltiazem), and phenylalkylamines (verapamil). ²⁵ To determine whether nimodipine-induced enhancement of α2 signaling is shared by other L-type Ca²⁺ channel blockers, we repeated the same experiment using other L-type blockers. The results are summarized in Figure 3. The nimodipine effect was mimicked only by isradipine, another dihydropyridine, but not by diltiazem or verapamil. We also used cadmium (Cd²⁺), a broad-spectrum Ca²⁺ channel blocker, to test whether it could enhance the brimonidine effect on NMDA responses. In cadmium Ringers, brimonidine was equally effective in suppressing the NMDA-induced inward current compared with that obtained with normal Ringer solutions (see Fig. 3 legend for details). 

To confirm that the concentrations of the Ca²⁺ channel blockers used in Figure 3 can indeed effectively block Ca²⁺ signal in the rat retina, we determined the effect of these agents on high K⁺ (40 mM) Ringer-elicited intracellular free Ca²⁺ signals at the inner plexiform layer (IPL) with confocal Ca²⁺ imaging. Brief application of high K+ Ringer elicited a robust Ca²⁺ signal at IPL. This Ca²⁺ signal was completely abolished with 0 Ca²⁺ Ringer or by Cd²⁺ (Fig. 4), indicating that the signal was produced by an influx of Ca²⁺ from the extracellular space through voltage-activated Ca²⁺ channels. Application of nimodipine, isradipine, diltiazem, and verapamil can all effectively block this Ca²⁺ signal (Fig. 4). Taken together, results shown in Figures 1 to 4 suggest strongly that the effect of nimodipine is independent of its classic action as an L-type Ca²⁺ channel blocker but depends on the dihydropyridine core structure. 

Figure 4.

 

Effects of 0 Ca²⁺ Ringer, nimodipine (nimo, 10 μM, n = 4), isradipine (israd, 10 μM, n = 6), diltiazem (DTZ, 20 μM, n = 5), verapamil (verap, 10 μM, n = 4), and cadmium (Cd²⁺, 100 μM, n = 4) on high K⁺ elicited increase of intracellular Ca²⁺ signal at the inner plexiform layer (IPL) of live rat retinal slices. (A) Traces of Ca²⁺ signals measured with confocal Ca²⁺ imaging. The gray traces were recorded in normal Ringer and black traces were recorded either with 0 Ca²⁺ Ringer or in the presence of the various Ca²⁺ channel blockers indicated. The thick black bars above the traces indicate the duration (6 sec) of high K⁺ Ringer application. (B) Normalized group data sets. **P < 0.01 between Ca²⁺ signals obtained under normal Ringer and in 0 Ca²⁺ Ringer or in the presence of various Ca²⁺ channel blockers.

Brimonidine Protection of RGCs in a Rabbit Retinal Excitotoxicity Model Is Enhanced by Nimodipine but Not by Diltiazem

We showed previously that brimonidine application protects retinal ganglion cells in a rabbit retinal NMDA excitotoxicity model. ⁸ This protective effect is mediated by α2 receptors through modulation of NMDA receptor function. ⁸ To test whether nimodipine can enhance α2 signaling under in vivo conditions, we compared the efficacy of brimonidine in the presence and absence of nimodipine or diltiazem in this in vivo rabbit model. 

The rabbit eye and vitreal space are much larger than those of the rat, which makes intravitreal drug application substantially easier. A single intravitreal injection of 3.6 μmol NMDA resulted in loss of approximately 40% of the neurons in the ganglion cell layer at 2 weeks after injection (Fig. 5C). In the rabbit retina, neurons at the ganglion cell layer are mostly (approximately two-thirds) RGCs. The remaining one-third are displaced amacrine cells (ACs), predominantly starburst ACs. ³² These displaced starburst ACs are particularly sensitive to DAPI, and they can be selectively labeled with a very low dose of DAPI. ³³ Although intravitreal application of NMDA (3.6 μmol) caused approximately 40% cell loss at the ganglion cell layer (Figs. 5C, 5D), the same NMDA treatment did not cause any cell loss in the displaced starburst ACs (selectively labeled with very low doses of DAPI ³³ ; see Fig. 5 legend). Our results are consistent with results of an earlier study that the NMDA application produces the smallest functional responses in displaced ACs and the largest responses in RGCs among all third-order retinal neurons. ³¹ Thus, our results indicate that the lost neurons at the ganglion cell layer were predominantly, if not exclusively, RGCs. This NMDA-induced cell loss can be completely blocked by pretreatment with the selective NMDA receptor blocker MK-801, confirming that neuronal cell loss was caused by NMDA receptor overactivation. ⁸ Pretreatment with brimonidine (3.6 nmol) was associated with an approximately 50% reduction in NMDA-induced cell loss (from 40.2% to 20.4%; Fig. 5D). Pretreatment with nimodipine was also neuroprotective by itself (cell loss was reduced from 40.2% to 27.5%; Fig. 5D; P < 0.01). Conceivably, nimodipine-mediated neuroprotection is produced by the blocking of excessive Ca²⁺ influx through the L-type Ca²⁺ channel on RGCs because intravitreally injected NMDA is expected to depolarize the RGCs by activating the NMDA receptor and activating secondarily voltage-gated L-type Ca²⁺ channels. Pretreatment with the combination of nimodipine and brimonidine produced significantly more protection (cell loss was reduced from 27.5% to 9.9%, a 64% reduction in cell loss; P < 0.05; Fig. 5D). Pretreatment with diltiazem was also neuroprotective by itself (cell loss was reduced from 40.2% to 28.1%; Fig. 5D). Conceivably, diltiazem-mediated RGC protection is also produced by blocking excessive Ca²⁺ influx through the L-type Ca²⁺ channel on RGCs because diltiazem is also an L-type Ca²⁺ channel blocker. The amount of protection with diltiazem was comparable with that seen with nimodipine (27.5% [nimodipine] vs. 28.1% [diltiazem]), but no enhancement of brimonidine effect was observed when diltiazem was applied with brimonidine (20.6% [with diltiazem] vs. 20.4% [without diltiazem]). These in vivo findings thus are consistent with our results from in situ RGCs that nimodipine, but not other classes of L-type Ca²⁺ channel blockers, enhances α2 signaling (Fig. 3). 

Figure 5.

 

Nimodipine, but not diltiazem, enhanced brimonidine protection of RGCs in the rabbit retinal NMDA excitotoxicity model. (A) Diagram illustrating 25 fields in the rabbit retina in which the number of neurons in the ganglion cell layer was counted. Each field measured 220 μm × 220 μm. (B) Representative images from row 1 (visual streak area) that show control and the effect of NMDA treatment alone (single intravitreal injection of 50 μL saline containing 3.6 μmol NMDA) or cotreatment with brimonidine or brimonidine+nimodipine on neuronal survival at 2 weeks after NMDA injection. Each individual fluorescent dot in an image is the DAPI (22.8 μg/50 μL PBS/eye)–labeled nucleus of a neuron in the ganglion cell layer. NMDA was injected either alone or in various combinations with brimonidine (brimo, 3.6 nmol), atipamezole (ati, 24 nmol), nimodipine (nimo, 24 nmol), or diltiazem (DTZ, 48 nmol), as indicated. In addition, test compounds were injected (at the same dose) 1 hour before and 24 hours after NMDA injection. In the NMDA alone group, 50 μL vehicle was also injected 1 hour before and 24 hours after NMDA injection to control for any injection-induced changes on cell survival. (C) Group statistical data showing that displaced starburst amacrine cells (dsAC, the predominant subtype of non-ganglion cells in the ganglion cell layer of the rabbit retina) are resistant to NMDA excitotoxicity. dsACs were selectively labeled with very low doses of DAPI (0.1 μg/50 μL PBS/eye). ³³ The y-axis is the total cell count from all 25 fields illustrated in (A). Left two columns: for all cells (control, 19 retinas; NMDA, 12 retinas), cell counts are shown using a normal dose of DAPI (22.8 μg/50 μL PBS/eye); the cells were counted 24 hours after DAPI injection. Right two columns: for dsACs (control, 6 retinas; NMDA, 6 retinas), cell counts are shown using very low dose of DAPI that selectively labeled dsACs; the cells were counted 48 hours after DAPI injection. **P < 0.01 between control and NMDA-treated data sets. (D) Group statistical data showing the effect of NMDA alone or in various combinations with other agents on neuronal survival at the ganglion cell layer. Both nimodipine and diltiazem were neuroprotective (P < 0.01 between NMDA and NMDA+nimo or between NMDA and NMDA+DTZ, 12 retinas for NMDA alone and 6 retinas for both NMDA+nimo and NMDA+DTZ). However, only nimodipine, but not diltiazem, significantly enhanced brimonidine protection (*P < 0.05 between NMDA+brimo and NMDA+nimo+brimo).

Discussion

Our ex vivo and in vivo results have demonstrated a novel neural mechanism involving nimodipine enhancement of α2 signaling in RGCs. This previously undescribed effect of nimodipine appears to be unrelated to its classic action as an L-type Ca²⁺ channel blocker (Fig. 3) because this effect is not mimicked by other L-type Ca²⁺ channel blockers, such as diltiazem and verapamil, or the broad-spectrum Ca²⁺ channel blocker Cd²⁺, all of which could effectively block depolarization-induced intracellular Ca²⁺ signals in the rat retina, as expected (Fig. 4). 

The exact mechanism that underlies nimodipine enhancement is unknown. However, though nimodipine enhancement is not mimicked by other L-type Ca²⁺ channel blockers that are benzothiazepine (diltiazem) or phenylalkylamine (verapamil) derivatives ²⁵ or an inorganic (Cd²⁺) Ca²⁺ channel blocker, it is mimicked by another dihydropyridine, isradipine (Fig. 3). Thus, this novel effect of nimodipine appears to depend on a dihydropyridine core structure. Because the nimodipine effect can be completely blocked by a specific α2 antagonist (Fig. 2), it is possible that nimodipine works as a positive modulator of the α2 receptor to enhance the effectiveness of the α2 agonist. More work is needed to elucidate the exact mechanism that underlies this novel nimodipine effect. 

Both nimodipine and diltiazem protect RGCs when applied alone in the rabbit NMDA excitotoxicity model (Fig. 5). This protection is likely produced by blocking excessive Ca²⁺ influx through voltage-gated L-type Ca²⁺ channels on RGCs because continuous activation of NMDA receptors on RGCs by intravitreally injected NMDA is expected to cause prolonged membrane depolarization that can secondarily activate L-type Ca²⁺ channels. In agreement with the ex vivo results (Fig. 3), pretreatment with nimodipine, but not diltiazem, in combination with brimonidine produced significantly better (P < 0.05) neuroprotection of RGCs than pretreatment with brimonidine alone (Fig. 5). One might argue that the neuroprotective effects of nimodipine and brimonidine are simply additive (Fig. 5D). However, this additive effect was not observed when diltiazem was used instead of nimodipine (Fig. 5D). There is evidence that Ca²⁺ overload by NMDA receptors is significantly more toxic than that by voltage-gated Ca²⁺ channels (source-specific toxicity). ³⁴ We speculate that though prolonged activation of L-type Ca²⁺ channels or NMDA receptors can contribute to RGC injury, overactivation of NMDA receptors may play a predominant role in RGC death. This can explain why the degree of neuroprotection produced by a combination of brimonidine and diltiazem is not significantly better than that produced by brimonidine alone (Fig. 5D) because brimonidine protection is mediated mainly by suppression of the NMDA receptor. ⁸ Nimodipine, but not diltiazem, in addition to blocking L-type Ca²⁺ channels, enhances brimonidine suppression of NMDA receptor function (Fig. 1). Therefore, it enhances brimonidine neuroprotection (Fig. 5D). 

As an L-type Ca²⁺ channel blocker, nimodipine could protect RGCs under disease conditions, such as glaucoma and acute retinal ischemia, by preventing glutamate over release from bipolar cells because L-type Ca²⁺ channels play a critical role in glutamate release at bipolar cell axon terminals. However, in the rabbit retinal NMDA excitotoxicity model used in this study (Fig. 5), RGC injury was caused directly by exogenously applied NMDA and can be completely prevented by coapplication of MK-801. ⁸ Therefore, the contribution of nimodipine's effect on other retinal cells (such as bipolar cells) to RGC protection is probably small, if any, because bipolar cells do not express functional NMDA receptors and are not directly affected by exogenously applied NMDA. Certain subtypes of amacrine cells do express NMDA receptors. However, amacrine cells are predominantly inhibitory neurons. Activation of NMDA receptors on these cells by exogenous NMDA is expected to produce a net inhibition on RGCs both indirectly through feedback inhibition to bipolar cell terminals (which causes a reduction in glutamate release from bipolar cells) and directly through feedforward inhibition on RGCs by GABA or glycine receptors. Thus, nimodipine's effect on amacrine cells is unlikely to play a significant role in RGC protection. 

The nimodipine-induced enhancement of α2 (brimonidine) modulation of NMDA receptor function can be blocked by either forskolin or an exogenous nonhydrolyzable analogue of cAMP (Sp-cAMPS; see Fig. 3). This indicates that nimodipine-induced enhancement also relies on the cAMP signaling pathway, a classic second-messenger signaling pathway that is associated with the α2 receptor. Although in this case the NMDA receptor was a downstream target of this α2 signaling pathway in RGCs, it is conceivable that, in different ocular tissues or cells, the cAMP pathway could modulate different downstream targets. Given that both brimonidine and nimodipine have a neuroprotective effect in the retina and that the results of the present work suggest that nimodipine can enhance brimonidine's neuroprotective effect, our findings raise the possibility that nimodipine's neuroprotective effects ²⁹ may occur, in part, by enhancing the effect of endogenous adrenergic ligands on α2 receptors in retinal neurons. 

Nimodipine has been shown to improve ocular blood flow in glaucoma patents. ^26,27 This effect of nimodipine is attributed to its antagonistic action on vascular L-type Ca²⁺ channels. Interestingly, brimonidine has also been shown to increase ocular blood flow in glaucoma patients. ^19,20 Although IOP lowering by itself could lead to improved ocular blood flow, the two do not necessarily have a positive and linear relationship. For example, topical betaxolol also lowered IOP but was associated with a reduced, instead of an increased, ocular blood flow. ²⁰ On the other hand, although trabeculectomy surgery was three times more effective in IOP lowering than topical brimonidine in patients with normal tension glaucoma, these two treatments produced virtually identical levels of increase in ocular blood flow. ²⁰ Additional mechanisms might have contributed to brimonidine's effect on ocular blood flow. We have recently reported that brimonidine can downmodulate L-type Ca²⁺ channel activity in the retina. ²² Because neuronal and vascular L-type Ca²⁺ channels have very similar pharmacologic profiles and are both sensitive to nimodipine blockade, our results raise the possibility of brimonidine's modulation of vascular L-type Ca²⁺ channel activity as a mechanism that contributes to enhanced ocular blood flow after brimonidine application in glaucoma patients. ^19,20  

In summary, the results of the present work have demonstrated a novel neural mechanism involving nimodipine enhancement of α2 signaling in RGCs. This effect of nimodipine is unrelated to its classic L-type Ca²⁺ channel-blocking action and may be partially responsible for nimodipine's retinal/ocular beneficial effects in patients and in animal models of retinal diseases. 

Footnotes

 Disclosure: C.-J. Dong, Allergan (F, I, E); Y. Guo, Allergan (F, I, E); P. Agey, Allergan (F, I, E); L. Wheeler, Allergan (F, I, E); W.A. Hare, Allergan (F, I, E)