α2 Adrenergic Receptor–Mediated Modulation of Cytosolic Ca⁺⁺ Signals at the Inner Plexiform Layer of the Rat Retina

Abstract

purpose. Compelling evidence suggests that α2 agonists, such as brimonidine, protect retinal ganglion cells (RGCs) from injury in a wide range of animal models. However, the mechanism of action for this protection and the physiological role of the α2 adrenergic system in the retina is not well understood. A major goal of this work was to explore the role of the α2 adrenergic system in the modulation of cytosolic Ca²⁺ signaling at retinal synaptic layers, particularly the inner plexiform layer (IPL), where communication between RGCs and their presynaptic cells takes place.

methods. Functional Ca²⁺ imaging at the inner plexiform layer (IPL) and outer plexiform layer (OPL) of living rat retinal slices was conducted with a high-speed confocal system. The relative changes of cytosolic free Ca²⁺ were monitored with the fluorescent Ca²⁺ dye fluo-4. The Ca²⁺ signal was elicited by membrane depolarization produced by a high K⁺ (40 mM) Ringer solution that was delivered rapidly and briefly to the test regions of the retinal slice by a custom-made multichannel local perfusion system.

results. A brief application (8 seconds) of high K⁺ Ringer elicited a robust cytosolic Ca²⁺ increase at the IPL and OPL. In both cases, this Ca²⁺ signal was eliminated by nimodipine, a selective L-type voltage-gated Ca²⁺-channel blocker, or when the extracellular Ca²⁺ in the Ringer was replaced with equal molar EGTA. At IPL, the Ca²⁺ signal was also suppressed in a dose-dependent manner by brimonidine and other α2 receptor agonists, such as medetomidine. The suppressive action of brimonidine and medetomidine was completely blocked by classic α2 receptor antagonists, such as yohimbine, rauwolscine, and atipamezole. Interestingly, the α2 receptor agonists had no effect on the high K⁺ Ringer-elicited cytosolic Ca²⁺ signal at OPL. Blocking the N-methyl-d-aspartate (NMDA) type of ionotropic glutamate receptor with D-AP5 attenuated this high K⁺–elicited Ca²⁺ signal by approximately 20% at IPL. D-AP5 had no effect on the Ca²⁺ signal at OPL.

conclusions. These findings provide the first direct evidence of α2 receptor–mediated modulation of L-type Ca²⁺ channel activity in the CNS (the retina is part of the CNS). This α2 modulation appears to occur at the IPL but not at the OPL of the retina. These findings suggest that a physiological function of the retinal α2 system is the regulation of synaptic transmission at IPL and that brimonidine and other α2 agonists may protect RGCs under disease conditions by preventing abnormal elevation of cytosolic free Ca²⁺ either in RGCs, in their presynaptic cells, or in both.

Retinal ganglion cells (RGCs) are the output neurons that carry visual information to higher visual centers of the brain. They are particularly vulnerable to injury in certain eye diseases such as glaucoma. ¹ RGCs receive inputs from bipolar cells and amacrine cells and perhaps even glial cells at the inner plexiform layer (IPL), a synaptic region composed primarily of the neural processes of bipolar, amacrine, and ganglion cells. The other synaptic region of the retina, the outer plexiform layer (OPL), is the site of synaptic communication between photoreceptors and second-order neurons—the bipolar cells and horizontal cells. ²  

Membrane depolarization–elicited increases in the cytosolic free Ca²⁺ in neurons are, in most instances, essential for neurotransmitter release from presynaptic terminals. However, compelling evidence indicates that under disease conditions, an abnormal elevation of intracellular free Ca²⁺ can lead to neuronal cell injury and contribute to the pathophysiology of acute and chronic neurologic disorders. ^{3 4 5}  

The α2 adrenergic receptors are G-protein–coupled receptors (GPCRs). ⁶ It has been shown in a number of previous studies that α2 agonists are neuroprotective in the retina. For example, brimonidine (UK14304), a classic α2 agonist, protects RGCs in animal models of glaucoma, ⁷ acute retinal ischemia, ^{8 9 10 11} and optic nerve injury. ¹² This neuroprotective action of brimonidine appears to be retinal in origin and is unrelated to its IOP (intraocular pressure)-lowering effect. Retinal α2 receptors have been demonstrated in a number of mammalian species, including the rat and human. ^{13 14 15} However, the mechanism that underlies neuroprotection of RGCs by α2 agonists in these animal models is not well understood, primarily because the physiological function of the retinal α2 adrenergic system is largely unknown. 

In this study, we conducted functional Ca²⁺ imaging at the IPL and OPL of living rat retinal slices with the use of a high-speed confocal imaging system. A major goal of this work was to explore the role of the α2 adrenergic system in modulating intracellular Ca²⁺ signaling at retinal synaptic layers, particularly at the IPL, where communication between RGCs and their presynaptic cells takes place. 

Methods

Preparation of Living Retinal Slices

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. Rat retinal slices were prepared using procedures similar to those described previously. ¹⁶ Briefly, Brown Norway rats were deeply anesthetized using intramuscular injection of ketamine (75 mg/kg) and xylazine (10 mg/kg). After enucleation of both eyes, animals were humanely killed immediately with intracardial injection of (120 mg/kg; Eutha-6, Western Medical Supply, Arcadia, CA). Retinas were carefully isolated, and a small piece of retina (approximately 2 × 6 mm) was placed vitreal side down on a piece of black filter paper (catalog no. habp04700; Millipore, Bedford, MA). The retina and filter paper were then radially sliced at approximately 250-μm intervals. Retinal slices were carefully transferred to recording chambers and were securely positioned, after a 90° turn, 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., Melville, NY). 

Solutions and Test Agents

The dye-loading medium was prepared using Hanks balanced salt solution without sodium bicarbonate but with 20 mM HEPES and 2 μM fluo-4 AM (Molecular Probes, Eugene, OR). Final pH was adjusted to 7.4, with 1 M NaOH. 

The bathing medium (rat Ringer) contained 120 mM NaCl, 3 mM KCl, 1.2 mM CaCl₂, 1.2 mM MgSO₄, 0.5 mM KH₂PO₄, 10 mM glucose, and 26 mM NaHCO₃. The Ringer was bubbled continuously with 95% O₂ and 5% CO₂. The high K⁺ Ringer was prepared by replacing 37 mM NaCl in the normal rat Ringer with equal molar KCl, yielding a total K⁺ concentration of 40 mM. The 0 Ca²⁺ Ringer was prepared by replacing CaCl₂ with equal molar EGTA. 

The following were purchased for the study: brimonidine (UK 14,304), yohimbine, rauwolscine, and nimodipine (Sigma, St. Louis, MO); medetomidine (Tocris, Ellisville, MO), and atipamezole (Antisedan; Novartis Animal Health, Basel, Switzerland). 

Loading Membrane Permeable Fluorescent Ca²⁺ Dye and Confocal Ca²⁺ Imaging

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. One of the chambers containing dye-loaded retinal slices was mounted on the microscope stage. Slices were continuously perfused with normal rat Ringer through multichannel bath and local perfusion systems. Flow rates were 3.5 and 0.3 mL/min for the bath and local perfusion systems, respectively. Imaging experiments were conducted at room temperature (approximately 20°C). 

Cytosolic Ca²⁺ signals were elicited by a brief (8-second) local perfusion of the high K⁺ Ringer that perfused the entire retinal slice being imaged. The light path shutter, high K⁺ perfusion, and Ca²⁺ imaging were precisely controlled (P-Clamp 8 software; Molecular Devices Corp., Sunnyvale, CA). Ca²⁺ imaging was conducted with a spinning disc confocal system (Nipkow; Solamere Technology Group, Salt Lake City, UT) equipped with a high-sensitivity, high-speed intensified charge-coupled device (CCD) camera (XR/Mega 10; Stanford Photonics Inc., Palo Alto, CA). Retinal Ca²⁺ images were acquired at 2 to 4 frame/s through a 60× long-working distance water immersion objective (LUMPlan FI 60×/0.90 W; Olympus America Inc, Melville, NY). 

Data Analysis

Imaging data were first measured and analyzed with QED software (MediaCybernetics, Silver Spring, MD). The measured data were then further analyzed and plotted with Origin (OriginLab, Northampton, MA) and Microsoft Excel. Statistical data were expressed as mean ± SD. 

Results

High K⁺ Elicited Intracellular Ca²⁺ Signal at IPL

In order to explore the role of the α2 adrenergic system in modulation of intracellular Ca²⁺ signaling at IPL, we have developed a method for functional Ca²⁺ imaging in living mammalian retinal slice, an ex vivo preparation that has many properties resembling those seen in vivo. In the acutely prepared rat retinal slice (Fig. 1A) , all major retinal layers, including IPL and OPL, are clearly visible and accessible for functional imaging. The yellow rectangle encloses the area in which intracellular Ca²⁺ signals were measured. 

We used local perfusion of a high K⁺ (40 mM) Ringer solution to produce membrane depolarization of IPL cell processes. Figure 1Bshows that an 8-second application of high K⁺ Ringer elicited a robust intracellular Ca²⁺ signal at IPL (red trace). This signal was completely abolished during perfusion with 0 Ca²⁺ Ringer (green trace). The Ca²⁺ signal was fully recovered after re-perfusion with normal Ca²⁺ solution (blue trace). This high K⁺-elicited Ca²⁺ signal was also virtually abolished after application of normal Ringer containing 10 μM nimodipine, a selective L-type Ca²⁺ channel blocker (Fig. 1C) . Thus, these results indicate that the high K⁺-elicited elevation of cytosolic free Ca²⁺ was produced mainly by Ca²⁺ influx through L-type, voltage-gated Ca²⁺ channels on the plasma membrane. 

α2 Receptor-Mediated Suppression of High K⁺-Elicited Ca²⁺ Signal at IPL

Next, we tested whether brimonidine could modulate this depolarization-induced intracellular Ca²⁺ increase. After a high K⁺-elicited Ca²⁺ response was recorded under control conditions (Fig. 2A , red trace), the slices were superfused with normal Ringer solution containing 600 nM brimonidine for 3 minutes before the high K⁺ solution (containing the same concentration of brimonidine as the bath solution) was applied. Brimonidine treatment abolished the high K⁺-elicited Ca²⁺ signal (Fig. 2A , green trace). The brimonidine effect was reversible. Full recovery of the high K⁺-elicited Ca²⁺ signal was usually observed after 3- to 5-minute washing with the brimonidine-free Ringers (Fig. 2A , blue trace). Brimonidine pretreatment did not seem to have a significant effect on the basal Ca²⁺ signal. Figure 2Bshows the dose-response relationship for brimonidine block of the high K⁺-elicited Ca²⁺ signal at IPL. The suppressive effect of brimonidine could be completely blocked by coapplication of either yohimbine or rauwolscine, two classic α2 adrenergic receptor antagonists (Figs. 2C 2D) . 

To further characterize the retinal α2 receptors, we also tested another widely used α2 receptor agonist, medetomidine. Figure 3Ashows that medetomidine suppressed the high K⁺-elicited Ca²⁺ signal at IPL, an effect that could also be completely blocked by yohimbine or atipamezole (a more potent and selective α2 antagonist than either yohimbine or rauwolscine). Figure 3Bshows the dose-response relationship of the medetomidine effect. Although medetomidine appeared to be more potent than brimonidine (lower IC₅₀ than that of brimonidine), the maximal block of IPL Ca²⁺ signal was only approximately 75%, a result consistent with medetomidine acting as a partial agonist. Brimonidine is able to block 100% of the IPL Ca²⁺ signal and appears to act as a full agonist at the α2 receptor in this case. 

Lack of Suppressive Effect of Brimonidine on High K⁺-Elicited Ca²⁺ Signal at OPL

In addition to IPL, photoreceptors communicate with bipolar cells and horizontal cells at another synaptic layer in the retina, the outer plexiform layer (OPL). Therefore, we were interested in determining whether a similar α2 modulation of depolarization-induced Ca²⁺ signals operates at OPL as well. Application of high K⁺ Ringer solution elicited a robust Ca²⁺ signal at OPL (Fig. 4A)and IPL (Fig. 4B)in the same retinal slice. As seen for IPL, the Ca²⁺ signal at OPL could also be eliminated by 0 Ca²⁺ Ringer and nimodipine (Fig. 4C) , indicating that it is also produced predominantly by Ca²⁺ influx through the L-type, voltage-gated Ca²⁺ channels. However, the application of brimonidine had no effect on the high K⁺-elicited Ca²⁺ signal at OPL. In the same slice, a typical brimonidine-induced suppression of the Ca²⁺ signal was observed at IPL (Fig. 4C , third pair from left). These results demonstrate that brimonidine suppression of the depolarization-induced Ca²⁺ signal is region specific: it potently suppresses the Ca²⁺ signal at IPL but not at OPL. 

High K⁺ depolarization–induced Ca²⁺ influx through L- and other types of Ca²⁺ channels may trigger glutamate release from photoreceptor and bipolar cell synaptic terminals. Glutamate may activate NMDA receptors, which have a high Ca²⁺ permeability and are expressed predominantly, if not exclusively, on the dendrites or processes of third-order retinal neurons, particularly ganglion cells. If the high K⁺-elicited Ca²⁺ signal is mediated mainly by NMDA receptors (secondary Ca²⁺ signal) and if the α2 agonists also modulate NMDA receptor activity, this could explain the observed region-specific α2 modulation of high K⁺-elicited Ca²⁺ signal. We used a selective NMDA receptor antagonist, D-AP5, to test this possibility. At a concentration of 50 μM, D-AP5 completely blocked NMDA-induced whole-cell currents in retinal neurons (data not shown) and did have a small effect on the high K⁺-elicited Ca²⁺ signal at IPL but not at OPL. It reduced the signal by 17% ± 6% (Fig. 4C , right pair). Thus, suppression of the high K⁺-elicited Ca²⁺ signal at IPL by α2 agonists appeared to have been caused mainly by the modulation of L-type Ca²⁺ channels. 

Discussion

Our results have provided the first direct evidence in the retina for α2 receptor–mediated modulation of cytosolic free Ca²⁺ signal induced by membrane depolarization. To our knowledge, this is also the first evidence of α2 modulation of L-type Ca²⁺ channels in the CNS. We have demonstrated that this α2 modulation occurs only at the IPL, not at the OPL. Because the IPL consists primarily of neural processes of RGCs and their presynaptic cells, this unique region-specific α2 modulation of cytosolic Ca²⁺ signals suggests a novel retinal mechanism that could contribute significantly to the reduction of RGC injury associated with α2 agonist treatment in animal models of glaucoma and acute retinal ischemia. 

Retinal α2 Adrenergic System

Unlike many other neurotransmitter systems in the retina, such as glutamate, GABA, glycine, acetylcholine, and dopamine, whose physiological function in the retina has been studied extensively, the normal function of the retinal α2 adrenergic system is largely unknown, ^{17 18} even though compelling evidence indicates that α2 receptors are expressed in the retina. ^{13 14 15} On the other hand, strong evidence indicates that α2 agonists are neuroprotective in a wide range of models for injury to retinal neurons. For example, α2 agonists, such as brimonidine, can reduce RGC injury in various experimental retinal disease models, including acute retinal ischemia, ^{8 9 10 11} glaucoma, ¹⁴ and optic nerve injury. ¹² However, the underlying mechanism(s) for this α2-mediated protection is (are) not well understood. 

It has been demonstrated in other neural tissues that presynaptic α2 receptors regulate transmitter release by modulating voltage-gated K⁺ and Ca²⁺ (N or P/Q subtypes) channels. ^{19 20 21 22} Our results not only provide the first direct evidence in the retina that the α2 mechanism modulates activity of voltage-gated Ca²⁺ channels, they suggest that this modulation is mainly on the L-type because the depolarization-elicited elevation of the cytosolic free Ca²⁺ is almost fully blocked by the selective L-type channel blocker nimodipine (Fig. 4) . The L-type Ca²⁺ channel is a major subtype in the retina that plays a critical role in neurotransmitter release from photoreceptors and from bipolar cells. ^{18 23} Given that α2 modulation occurred only at IPL, our results suggest that a normal function of the α2 mechanism in the retina is the modulation of synaptic transmission between second- and third-order retinal neurons. Under disease conditions, this negative modulation of Ca²⁺ influx (which would be expected to limit the release of neurotransmitters, particularly glutamate) may significantly reduce glutamatergic excitotoxic injury to RGCs. Consistent with this is a previous finding that, in acute retinal ischemia, brimonidine treatment not only reduces ischemic injury to RGCs, it significantly reduces ischemia-induced elevation of vitreal glutamate levels. ⁹  

Abnormal elevation of cytosolic free Ca²⁺, produced mainly by Ca²⁺ influx through voltage- or ligand-gated Ca²⁺ channels, or both, has been demonstrated to contribute significantly to neuronal injury in many neural disorders. ^{3 4 24} Therefore, it is conceivable that a direct negative modulation of depolarization-elicited Ca²⁺ influx in RGCs (in this case, the dendrites of RGCs) by α2 agonists may also contribute to neuroprotective activity. Indeed, previous studies have shown that α2 receptors are expressed in IPL and RGC layers. ^{13 14 15} Thus, the α2 system could protect RGCs under disease conditions by presynaptic (preventing the excess release of glutamate) and postsynaptic (preventing the overactivation of L-type Ca²⁺ channels on RGCs) mechanisms. More work is needed on specific types of retinal cells (such as bipolar cells and RGCs themselves) to identify directly these presynaptic and postsynaptic mechanisms. 

Primary and Secondary Ca²⁺ Signals

Our results suggest that the high K⁺-elicited increase of cytosolic Ca²⁺ predominantly represents Ca²⁺ influx through voltage-gated Ca²⁺ channels (particularly the L-type channels). Because membrane depolarization and voltage-gated Ca²⁺ channel activation could trigger the release of excitatory neurotransmitters such as glutamate, a secondary Ca²⁺ signal may be generated as a result of the activation of Ca²⁺-permeable ionotropic glutamate receptors (such as NMDA receptors, which are highly Ca²⁺ permeable and are sometimes called ligand-gated Ca²⁺ channels). However, direct testing with a selective NMDA receptor antagonist D-AP5 showed that this NMDA receptor–mediated secondary contribution was relatively small (Fig. 4C , right pair). In cultured rat retinal neurons, it has been shown that brimonidine attenuates a glutamate-elicited increase of cytosolic Ca⁺⁺. ²⁵ Therefore, a similar α2 modulation of NMDA receptor function could make a small contribution to the overall α2 modulation of Ca²⁺ signals at IPL. 

Region-Specific Modulation of Ca²⁺ Signals by α2 Agonists

It is surprising that α2 agonists selectively modulated Ca²⁺ signals at IPL but not at OPL, despite the fact that the Ca²⁺ signal at both synaptic layers was mediated mainly by L-type Ca²⁺ channels (Fig. 4) . α2 Receptors are expressed in outer retina, albeit at a significantly lower level. ¹⁵ One possible explanation is that α2 receptors in outer retina may not be coupled to the L-type Ca²⁺ channels but make use of a different intracellular signaling pathway. For example, it has been reported that α2 agonists are neuroprotective to photoreceptors. This effect is mediated by the induction of basic fibroblast growth factor, a unique α2-mediated action found in retinal photoreceptors but nowhere else in the CNS. ²⁶ Alternatively, α2 agonists may have a differential (opposite) effect on L-type Ca²⁺ channels at rod and cone terminals, which could explain the absence of an α2 effect on the overall Ca²⁺ signal at OPL. This kind of differential effect of dopamine and somatostatin has been reported earlier by others. ^{27 28}  

Although functional NMDA receptors are expressed predominantly, if not exclusively, on third-order retinal neurons, this differential expression pattern cannot be the main reason for this region-specific modulation of Ca²⁺ signals by α2 agonists because NMDA receptor contribution to the overall high K⁺-elicited Ca²⁺ signal at IPL is relatively small (Fig. 4C) . 

Conclusions

Our findings have provided the first evidence that α2 receptors can modulate the activity of L-type Ca²⁺ channels in the CNS. They also show that this α2 modulation is region specific in the retina in that it occurs at the IPL but not at the OPL. These findings suggest that a physiological function of the retinal α2 system is the regulation of synaptic transmission at IPL and that brimonidine and other α2 agonists may protect RGCs from experimental injury by preventing abnormal elevation of cytosolic free Ca²⁺ either in RGCs, in their presynaptic cells, or in both. 

 Submitted for publication July 31, 2006; revised October 13, 2006; accepted December 26, 2006.

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

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Functional Ca²⁺ imaging in living rat retinal slice. (A) Bright-field image of a representative living rat retinal slice. The yellow rectangle encompasses the area in which the Ca²⁺ signal (B) was measured. (B) High K⁺-elicited elevation of cytosolic free Ca²⁺ at IPL. This high K⁺-elicited Ca²⁺ signal was abolished during perfusion of 0 Ca²⁺ Ringer. The y-axis is normalized fluorescent intensity. The black horizontal bar indicates the duration of local perfusion with high K⁺ Ringer. (C) Statistical data for the effects of 0 Ca²⁺ (2 ± 3; n = 7) and 10 μM nimodipine (8 ± 21; n = 4) on IPL Ca²⁺ signal. Data are plotted as mean ± SD.

Figure 2.

 

Suppression by brimonidine of high K⁺-elicited Ca²⁺ signal at IPL. (A) Effect of 600 nM brimonidine on the Ca²⁺ signal. (B) Dose-response relationship of brimonidine effect on the Ca²⁺ signal at IPL. (C) The brimonidine effect is blocked by the α2 receptor antagonist yohimbine. (D) Statistical data for the effects of 600 nM brimonidine (0.1 ± 0.8; n = 6) and two α2 receptor antagonists, 10 μM yohimbine (98 ± 4; n = 6) and 10 μM rauwolscine (91 ± 9; n = 6) on IPL Ca²⁺ signal. Data are plotted as mean ± SD.

Figure 3.

 

Suppression by medetomidine of high K⁺-elicited Ca²⁺ signal at IPL. (A) Effect of 300 nM medetomidine (control for yohimbine group: 32 ± 14, n = 7; control for atipamezole group: 37 ± 19, n = 6) applied alone or in combination with 10 μM yohimbine (79 ± 12; n = 7) or 3 μM atipamezole (98 ± 9; n = 6). (B) Dose-response relationship for medetomidine effect (blue) on the Ca²⁺ signal at IPL. Statistical numbers at each concentration are: 0.03 μM: 84 ± 6, n = 7; 0.1 μM: 63 ± 9, n = 9; 0.3 μM: 34 ± 16, n = 13; 0.6 μM: 28 ± 19, n = 7. For comparison, the dose-response curve for brimonidine (red, from Fig. 2B ) is also plotted. Data are plotted as mean ± SD. The two sets of dashed lines indicate the approximate IC₅₀ for medetomidine (100 nM) and brimonidine (approximately 260 nM), respectively.

Figure 4.

 

Region-specific modulation by α2 agonists of high K⁺-elicited Ca²⁺ signals. High K⁺ elicited a robust cytosolic Ca²⁺ increase at both OPL (A) and IPL (B) at the same retinal slices (n = 7). (C) Effects of 0 Ca²⁺, nimodipine (a selective L-type, voltage-gated Ca²⁺ channel blocker), brimonidine, and D-AP5 (a selective NMDA receptor antagonist) on high K⁺-elicited Ca²⁺ signals at OPL and IPL. Perfusion with 0 Ca²⁺ Ringer abolished the Ca²⁺ response at both OPL (2% ± 2%; n = 7) and IPL (2% ± 3%; n = 7). Application of 10 μM nimodipine also substantially attenuated the Ca²⁺ signal at both OPL (14% ± 18%; n = 4) and IPL (8% ± 21%; n = 4). Brimonidine 300 nM selectively suppressed the Ca²⁺ signal at IPL but not at OPL in the same retinal slices (IPL, 41% ± 12%; OPL, 95% ± 8%; n = 6). D-AP5 50 μM had a small but statistically significant effect (83% ± 6%, n = 4, P = 0.01, paired t-test) at IPL. It did not affect the Ca²⁺ signal at OPL (96% ± 4%, n = 4, P = 0.12). Data are plotted as mean ± SD.