This document contains text automatically extracted from a PDF or image file.

Formatting may have been lost and not all text may have been recognized.
To remove this note, right-click and select "Delete table".

Exp. Eye Res. (2001) 72, 393±401 doi:10.1006/exer.2000.0968, available online at http://www.idealibrary.com on

Voltage-dependent Calcium Channels in the Rat Retina: Involvement in NMDA-stimulated In¯ux of Calcium
JOSE MELENA
AND

NEVILLE N. OSBORNE* Nuf®eld Laboratory of Ophthalmology, University of Oxford, Walton Street, Oxford, OX2 6AW, U.K. (Received Cleveland 12 June 2000, accepted in revised form 29 November 2000 and published electronically 8 February 2001)
Rises in intracellular Ca2 induced by activation of glutamate receptors are of ultimate importance for neuronal excitability and pathophysiological processes. In the present study, we aimed to elucidate the types of voltage-dependent Ca2 channels involved in the NMDA-stimulated in¯ux of Ca2 into the isolated rat retina by using selective blockers. Additionally, the number of binding sites for radioligands labelling L- ([3H]nitrendipine), N- ([125I]o-conotoxin MVIIA) and P/Q-type ([125I]o-conotoxin MVIIC) Ca2 channels was assessed in the rat retina and, for further comparison, in the rat cortex. Incubation of isolated rat retinas with 100 m M NMDA produced a three-fold increase in the in¯ux of 45Ca2 that was completely blunted by MK-801, a NMDA receptor antagonist, and partially attenuated (approximately 20%) by tetrodotoxin, a Na channel blocker. The L-type Ca2 channel blocker nifedipine reduced NMDA-stimulated Ca2 in¯ux in a dose-related fashion, with a maximum reduction of approximately 50%. Similar effects were observed with verapamil and diltiazem. Blockers of N- and P/Q-type Ca2 channels had no signi®cant effect on the in¯ux of Ca2 evoked by NMDA. Co2 , a non-speci®c Ca2 channel blocker, caused an inhibition of NMDA-stimulated Ca2 in¯ux similar to that of nifedipine. Therefore, of all voltage-dependent Ca2 channels, L-type channels appear to make the greatest contribution (up to 50%) to the NMDA-stimulated in¯ux of Ca2 into the isolated rat retina. This ®nding contrasts with evidence obtained in brain neurones supporting a role for L-, N- and P/Q-type channels in NMDA-evoked Ca2 signals. A comparison of the number of radioligand binding sites associated with L-, N- or P/Q-type Ca2 channels in the rat cortex and retina revealed that such a difference cannot be ascribed to a distinct expression pattern of these channels in both tissues, although some variations were found. Interestingly, a different af®nity of [3H]nitrendipine for L-type Ca2 channels in the rat retina and cortex was observed which may re¯ect the expression of different classes of L-type channels in these tissues. The ability of L-type Ca2 channel blockers to attenuate NMDA-stimulated Ca2 in¯ux may underlie their neuroprotective effects in the retina. # 2001 Academic Press Key words: voltage-dependent calcium channel; NMDA; calcium in¯ux; retina; neuroprotection; ischaemia; glaucoma; rat.

1. Introduction The in¯ux of Ca2 through voltage-dependent Ca2 channels (VDCCs) into cells mediates a variety of processes which include neurotransmitter release, excitability, excitation-contraction coupling, secretion and gene expression. VDCCs are classi®ed into L-, N-, P/Q-, R- and T-types based on their functional properties, such as time- and voltage-dependent kinetics, and sensitivity to different pharmacological agents (Birnbaumer et al., 1994; Dunlap, Luebke and

Turner, 1995). Ca2 currents through L-type VDCCs, exhibited both in excitable and nonexcitable cells, are high-voltage activated, long-lasting and very sensitive to dyhydropyridines (e.g. nifedipine), as well as phenylalkylamines (e.g. verapamil) and benzothiazepines (e.g. diltiazem). The high voltage-activated Nand P/Q-type VDCCs are largely restricted to neurons and selectively blocked by o-conotoxins GVIA (o-CgTx-GVIA) and MVIIA (o-CgTx-MVIIA) and by o-agatoxins IVA (o-AgaTx-IVA) and TK (o-AgaTxTK), respectively (Birnbaumer et al., 1994; Dunlap et al., 1995). Neuronal high-voltage activated VDCCs also include the R-type, characterized by its resistance to both dihydropyridines and toxins and its susceptibility to blockade by low concentrations of Ni2 (Avery and Johnston, 1996; Imaizumi, Kocsis and Waxman, 1999). T-type VDCCs are low-threshold, transient channels which are implicated in repetitive ®ring and pacemaker activity. T-type VDCCs show resistance to dihydropyridines and toxins, but are sensitive to mibefradil, ¯unarizine and high concentrations of Ni2 (Birnbaumer et al., 1994; Dunlap et al., 1995). Rises in intracellular calcium following activation of a-amino-3-dihydro-5-methyl-isoxazol-4-propionic acid/kainate (AMPA/KA) and N-methylD

-aspartate (NMDA) ionotropic glutamate receptors have been associated with critical physiological processes, such as synaptictransmission,synapticplasticityandneuronal differentiation (Collingridge and Singer, 1990; Ozawa, Kamiya and Tsuzuki, 1998). Additionally, abnormally
0014-4835/01/040393 09 $35.00/0 # 2001 Academic Press * Address correspondence to: Neville N. Osborne, Nuf®eld Laboratory of Ophthalmology, Oxford University, Walton Street, Oxford, OX2 6AW, U.K. E-mail: neville.osborne@eye.ox.ac.uk

high levels of glutamate released during ischaemia can produce overstimulation of ionotropic glutamate receptors, leading to neuronal cell death triggered by a large in¯ux of Ca2 into cells mainly via NMDA receptors and secondary opening of VDCCs (Lee, Zipfel and Choi, 1999). Therefore, substances able to reduce Ca2 overload either by interacting with NMDA receptors and/or VDCCs should theoretically alleviate ischaemic/excitotoxic damage to the retina. Accordingly, a number of NMDA receptor antagonists have been shown to protect retinal cells in models of ischaemia/excitotoxicity, although their potential therapeutic use is questioned because of their major side effects (Osborne et al., 1999b; Ritch, 2000). An alternative way to reduce Ca2 overload following an ischaemic insult and hence to ameliorate neuronal cell death would theoretically include the blockade of VDCCs. Little is known, however, about the types of VDCCs which are involved in the glutamate-evoked entry of extracellular Ca2 into retinal cells. Therefore, the aim of the present study was to determine which kind of VDCCs are involved in the NMDA-stimulated in¯ux of Ca2 into isolated rat retinas by applying selective blockers. In addition, a comparative assessment of the number of L-, N- and P/Q-type VDCCs in the rat cortex and retina was performed by using speci®c radioligands. 2. Materials and Methods Determination of NMDA-stimulated In¯ux of Ca into Isolated Rat Retinas Adult Wistar rats (250±350 g) bred in our laboratory were killed by decapitation and retinas rapidly removed and preincubated (378C for 25 min) in magnesium-free (138 m
M

NaCl, 5 . 6 Krebs±Ringer m
M

bicarbonate KCl, 1 m
M

CaCl buffer NaHCO and 20 3 ,m

1M m
M

HEPES NaH
2

buffer, PO
4

-Na
2

pH HPO 7
4

. , 4) 10 continuously m
2 M

, 11 m
M

glucose perfused then added with and 95% the O retinas 2 /5% CO incubated 2 . Antagonists were for a further 10 min. Calcium in¯ux was initiated by the addition of ®nal 1 mCi volume 45CaCl of 2 2 and ml. 100 m
M

NMDA After 15 min or buffer in a incubation, the reaction 10 m
M

washed was EGTA twice halted per 0 by . 9 % the NaCl. addition Retinas of in 2 ml ice-cold 1 m
M

EGTA/0 2 ml ice-cold were . 9% NaCl then and sonicated in 1 ml distilled water. Radioactivity was determined by liquid scintillation spectrometry in 5 ml of Insta-gel Plus (Packard, Groningen, The Netherlands) and protein concentration with a bicinchoninic acid protein assay kit (Sigma, Poole, U.K.) using bovine serum albumin (BSA) as standard.

Tissue Preparation for Radioligand Binding Studies Adult Wistar rats were killed by decapitation. The cerebral cortices and retinas were dissected over ice and homogenized in 15 volumes of ice-cold 50 m
M 394 J. MELENA AND N. OSBORNE

. Tris±HCl buffer, pH 7 4 (for [3H]nitrendipine binding), or 50 m
M

. 4 (for [125I]o-CgTxMVIIA and [125I]o-CgTx-MVIIC binding). The homogenate was centrifuged at 1000 g for 10 min at 48C and the resulting supernatant was washed by means of two consecutive centrifugation cycles (48 000 g, 10 min, 48C) with intermittent resuspension of the pellet in fresh buffer. The ®nal pellet was resuspended in 50 m
M

HEPES buffer, pH 7 Tris±HCl (for [3H]nitrendipine binding) or 50 m
M

HEPES (for [125I]o-CgTx-MVIIA and [125I]oCgTx-MVIIC binding) buffer to yield an original tissue concentration of 100 mg mlÀ1 and stored at À708C until use. Protein concentration was determined as described above. Radioligand Binding to Voltage-dependent Ca2 Channels [3H]Nitrendipine was used for labelling L-type VDCCs. Saturation binding assays were performed by incubating aliquots of cortical (100 mg protein per tube) or retinal (250 mg protein per tube) membranes for 90 min at 258C in 300 ml of 50 m
M

Tris±HCl buffer containing 0 . 05±2 n
M [3H]nitrendipine. At the

end of the incubation, samples were diluted with 4 ml of ice-cold 50 m
M

Tris±HCl buffer (pH 7

. 4), immediately vacuum ®ltered through Whatman GF/B ®lters and washed twice with 4 ml of ice-cold buffer. Radioactivity on the ®lters was measured by liquid scintillation spectrometry in 5 ml of Insta-gel Plus. Non-speci®c binding of [3H]nitrendipine was determined in the presence of 1 m
M

nifedipine. These experiments were carried out in subdued light to minimize [3H]nitrendipine and nifedipine degradation. N-type and P/Q-type VDCCs were labelled with [125I]o-conotoxin MVIIA (o-CgTx-MVIIA) and [125I]o-conotoxin MVIIC (o-CgTx-MVIIC), respectively. Saturation binding assays were performed by incubating aliquots of cortical (1 mg protein per tube) or retinal (5 mg protein per tube) membranes with 1±20 p
M [125I]o-CgTx-MVIIA or 1±20 p M [125I]o-

CgTx-MVIIC for 60 min at 258C in 300 ml of 20 m
M

HEPES buffer containing 75 m
M

. 1m
M

EDTA, 0 NaCl, 0 . 1m
M

EGTA and 0 . 1% BSA (pH 7 . 2). At the end of the incubation, samples were diluted with 4 ml of icecold washing buffer (20 m
M

NaCl, and 0 HEPES, 125 m
M

.

1 % BSA, pH 7 . 2), immediately vacuum ®ltered through Whatman GF/B ®lters and washed twice with 4 ml of ice-cold buffer. The ®lters were presoaked in 0.6 % polyethylenimine for 2 hr to reduce radioactivity trapping on the ®lters. Radioactivity on the ®lters was measured using a gamma counter (NE 1600, Nuclear Enterprises Ltd, Edinburgh, U.K.). Nonspeci®c binding was determined in the presence of 20 n
M

o-CgTx-MVIIA or 100 n
M

o-CgTx-MVIIC. Under these conditions, [125I]o-CgTx-MVIIC selectively labels P/Q-type VDCCs (Kristipati et al., 1994).

Materials [3H]Nitrendipine (72 . 4 Ci mmolÀ1), [125I]o-CgTxMVIIA (2200 Ci mmolÀ1) and [125I]o-CgTx-MVIIC (2200 Ci mmolÀ1) were obtained from NEN Research Products from Amersham (Stevenage, (Amersham, U.K.) and U.K.). 45CaCl o-CgTx-MVIIA,
2

(2 mCi mlÀ1) o-CgTx-MVIIC and o-AgaTx-TK were purchased from TCS Biologicals Ltd (Botolph Claydon, U.K.) and MK-801 and tetrodotoxin from Semat Technical Ltd (St. Albans, U.K.). NMDA, diltiazem, nifedipine and verapamil came from Sigma (Poole, U.K.). Analysis of Data Saturation binding data nonlinear regression method were (GraphPad analysed Prism using 1 . 0). a All experiments were performed in duplicate and data are expressed as mean+
S.E.(M.)

. Statistical analyses were performed by using an unpaired Student's t-test and one-way analysis of variance Bonferroni's post-test. Values of (ANOVA) P 5 0 . 05 using were the considered statistically signi®cant. 3. Results Role of Voltage-dependent Ca2 Channels in the NMDAstimulated In¯ux of Ca2 into Isolated Rat Retinas As shown in Fig. 1, NMDA (100 m
M

) induced approximately a three-fold increase in the in¯ux of 45Ca2 into isolated rat retinas from 3082+ 283 cpm mgÀ1 486 cpm mgÀ1 protein protein (n 15) (n 10329+ . 001 by unpaired Student's t-test). Such an increase in
CALCIUM CHANNELS IN THE RAT RETINA 395

17, P to 5 0

.F IG 1. In¯ux of 45Ca2 into isolated rat retinas. NMDA (100 m M (control) which was completely blunted by MK-801 and ) partially inhibited caused by a signi®cant TTX (P 5 increase 0 . 001 and in the 0 . 05, basal respectively, in¯ux of calcium by oneway ANOVA followed by parentheses. *Values signi®cantly Bonferroni's different post-test). (P 5 0 Data . 001) represent from control the mean+ S.E.(M.) of group by one-way the number of experiments shown in ANOVA followed by Bonferroni's posttest.

.
45Ca2

in¯ux was completely blunted (P 5 0 001 by one-way ANOVA followed by Bonferroni's post-test) by the NMDA receptor antagonist MK-801 (Fig. 1). Tetrodotoxin (TTX), at a (1 (P 5 m
M

0 ), . 05 produced by saturating . concentration an small (19 8%), albeit signi®cant one-way ANOVA followed by Bonferroni's post-test), inhibition of NMDA-stimulated Ca2 in¯ux into the rat retina (Fig. 1), indicating a role for voltage-sensitive Na channels in this effect. The dihydropyridine nifedipine, a speci®c L-type VDCC blocker, was found to decrease the NMDAstimulated in¯ux of Ca2 into the isolated rat retina in a dose-related fashion, with a maximum reduction of approximately 50% (Fig. 2). Nonlinear ®tting of the effects of nifedipine on the in¯ux of Ca2 into the rat retina evoked by NMDA, considering as maximum effect that observed an and EC a 50 slope value factor of 0 . of 88 with 0 m .

92. M 100 (ÀlogEC The m
M

benzothiazepine nifedipine, 50 6 . 06+0 revealed . 16) and the phenylalkylamine L-type VDCC blockers diltiazem and verapamil, respectively, produced similar reductions in the NMDA-stimulated in¯ux of Ca2 into the isolated rat retina (Fig. 2). Both o-CgTx-MVIIA, a speci®c blocker of N-type VDCCs, and o-AgaTx-TK, a P/Q-type VDCC blocker, had no effect on the NMDA-induced in¯ux of Ca2 into isolated rat retinas at a saturating concentration of 100 n
M

(Fig. 3). No effect was also observed with a higher concentration (500 n
M

) of o-CgTx-MVIIA (data not shown). Moreover, combination of 100 n
M

o-CgTx-MVIIA or 100 n
M

o-AgaTx-TK with 1 m
M

nifedipine did not result in an inhibition of Ca2 in¯ux signi®cantly higher than that observed with 1 m
M

nifedipine alone (data not shown). Ni2 was used to
45Ca2+ influx (cpm mg protein)

explore the potential role of T-type and R-type VDCCs in the NMDA-stimulated in¯ux of Ca2 into rat retinas. At low concentrations (530 m
M

), Ni2 is considered to selectively block R-type VDCCs, whereas at higher concentrations it would also inhibit T-type channels (Avery and Johnston, 1996; Imaizumi et al., 1999). At 10 m
M

concentration, Ni2 was found to
. 2. Effects of L-type Ca2 channel blockers nifedipine, verapamil and diltiazem on the NMDA-stimulated in¯ux of 45Ca2 into isolated rat retinas. Data represent the mean+ S.E.(M.) signi®cantly different from control group by one-way ANOVA F IG of the followed number of by Bonferroni's experiments post-test: shown *P 5 in 0 . 01, parentheses. **P 5 0 . 001. Values 396 J. MELENA AND N. OSBORNE .F IG 3. Effects of Ca2 channel blockers on the NMDA-stimulated in¯ux of 45Ca2 into isolated rat retinas. Data represent the mean+ S.E.(M.) of way ANOVA the followed number by of Bonferroni's experiments post-test: shown *P 5 in 0 . parentheses. 01, **P 5 0 . Values 001. signi®cantly different from control group by one-

signi®cantly enhance the NMDA-stimulated Ca2 in¯ux into the isolated rat retina, while no signi®cant effect was observed at 100 m
M

concentration (Fig. 3). Finally, the non-speci®c Ca2 channel blocker Co2 , at a saturating concentration (500 m
M

), inhibited the NMDA-stimulated in¯ux of Ca2 into the rat retina by approximately 60% (Fig. 3). The in¯ux of Ca2
12000 10000 8000 6000 4000 20 100 10 10 100 Dil apam Ver 7) Control l 6000 4000 1 8000 - 6000 4000 2000 0

Co2 measured in the presence of 500 m
M

(4276 +313, n 4) was, however, not signi®cantly different from that observed with 10 m
M

nifedipine (5163 +334, n 4). Radioligand Binding to Voltage-dependent Ca2 Channels Speci®c binding of [3H]nitrendipine to rat retinal homogenates was saturable and consistent with the existence of a single class of binding sites, as indicated by values the calculated linear Scatchard by nonlinear plot [Fig. analysis 4(A)]. for K
D

the and speci®c B
max

binding of [3H]nitrendipine to both rat cortical and retinal homogenates are shown in Table I. [3H]Nitrendipine showed a signi®cantly lower af®nity (a higher K cortex D value) (Table for binding I). The number sites in the of binding retina [3H]nitrendipine in the cortex was than in the rat approximately sites (B
max

) for three-fold higher than in the rat retina (Table I). Speci®c binding of [125I]o-CgTx-MVIIA to N-type VDCCs in rat cortical homogenates was saturable and of high af®nity (Table I). Speci®c binding of [125I]oCgTx-MVIIA to rat retinal membranes showed similar characteristics and no evidence for multiple binding sites, given the linearity of the Scatchard plot [Fig. 4(B)]. [125I]o-CgTx-MVIIA showed a similar af®nity for binding sites in both the rat cortex and retina, while the number of binding sites was approximately 11-fold lower in the retina (Table I). Speci®c binding of [125I]o-CgTx-MVIIC to rat retinal membranes was saturable and consistent with the existence of a single class of binding sites, as revealed by the linearity and CgTx-MVIIC B
max

values to of the Scatchard for the speci®c binding plot [Fig. of 4(C)]. [125I]oK
D

both rat cortical and retinal homogenates are shown in Table I. No differences were found in the af®nity of [125I]o-CgTx-MVIIC for radioligand binding sites in the rat cortex and retina, while the binding sites in the retina were signi®cantly (six-fold) less abundant (Table I). 4. Discussion In the present study we attempted to identify the types of VDCCs involved in the NMDA-stimulated
CALCIUM CHANNELS IN THE RAT RETINA 397

T
ABLE

I Af®nity (K
D

) and number of binding sites (B
max

) for radioligands labelling VDCCs in the rat cortex and retina
Radioligand Cortex Retina K D (fmol mgÀ1 protein) n K D (fmol mgÀ1 protein) n [3H]Nitrendipine 216 (p M )B max (p M )B max . 0+ 20 . 5 119 . 4+ 3 . 1 3 533 . 2+ 28 . 9* 40 . 1+2 .

[125I]o-Conotoxin MVIIA [125I]o-Conotoxin MVIIC 2 5 . . 7+0 4+0 . . 2 4 240 448 . . 2+ 8+ 20 10 . . 9 0 9* 4 433 4 . . 4+ 9+ 0 0 . . 4 2 22 74 . . 4+2 2+6 . . 6* 1* 4 4 Values are mean + S.E.(M.) of n independent experiments performed in duplicate. *Signi®cantly different (P 5 0 . 001) from corresponding value in cortex by unpaired Student's t-test.

in¯ux of Ca2 into the isolated rat retina as well to assess the number of retinal L-, N- and P/Q-type VDCCs by radioligand binding. Incubation of isolated rat retinas with 100 m
M

NMDA caused a three-fold increase in the basal total cellular Ca2 which was blunted by MK-801, a speci®c NMDA receptor antagonist. Such an increase is likely to re¯ect the in¯ux of Ca2 into amacrine, ganglion and MuÈller

cells, the main types of retinal cells shown to express functional NMDA receptors (Thoreson and Witkovsky, 1999; Fletcher et al., 2000). It has not been demonstrated that photoreceptors, bipolar and horizontal cells express functional NMDA receptors (Thoreson and Witkovsky, 1999; Fletcher et al., 2000). The speci®city of the in¯ux of Ca2 into the isolated rat retina evoked by NMDA and the absence of functional NMDA receptors in photoreceptors suggest that non-speci®c uptake of Ca2 by photoreceptor cells was negligible in our preparation. TTX, a blocker of voltage-dependent sodium channels (VDSCs), was found to signi®cantly reduce (approximately 20%) the NMDA-stimulated in¯ux of Ca2 into the rat retina, revealing a role for VDSCs in this effect. It is likely that membrane depolarization following activation of NMDA receptors leads to opening of VDSCs, resulting in further membrane depolarization. Blockade of VDSCs by TTX could decrease the extent of membrane depolarization triggered by NMDA receptor activation and hence reduce the magnitude of other voltage-dependent cellular processes, such as the entry of Ca2 through VDCCs. Alternatively, TTX could reduce spontaneous spiking activity of retinal neurones, thus reducing membrane depolarization and increasing the probability of voltage-dependent block by Mg2 of NMDA receptors (Mayer, Westbrook and Guthrie, 1984). The increase in total intracellular Ca2 in the isolated rat retina induced by NMDA was reduced by the L-type VDCC blocker nifedipine (dihydropyridine class) 0 . 88 in a m
M

was dose-related calculated for fashion. nifedipine An in EC inhibiting 50 value the of in¯ux of Ca2 into isolated rat retinas evoked by 100 m
M

NMDA. Nifedipine caused a maximal inhibition of NMDA-stimulated in¯ux of Ca2 of approximately 50 %. Similar reductions were also observed

398 J. MELENA AND N. OSBORNE .F IG 4. Speci®c binding of [3H]nitrendipine (A), [125I]o-CgTx-MVIIA (B) and [125I]o-CgTx-MVIIC (C) to rat retinal homogenates. Scatchard plots of the data are shown as insets. The results represent data from a single typical experiment performed in duplicate. Mean K D and B max values for four experiments are given in Table I. endipine bound mg pr otein) A 0.10 0.00 0 IO 20 30 40 0 80 70 60 50 Free (nM) o Bound/Free 1.5 2.0 0.3 0 0.2 Free (PM) 0.00 0 20 40 60 80 Free (PM)

after incubation with L-type VDCC blockers diltiazem (benzothiazepine class) and verapamil (phenylalkylamine class). Blockers of N-type (o-CgTx-MVIIA) and P/Q-type (o-AgaTx-TK) VDCCs did not, however, signi®cantly inhibit the in¯ux of Ca2 into the isolated rat retina stimulated by NMDA, suggesting that both N- and P/Q-type VDCCs do not play a major role in this effect. When Ni2 was applied at a low concentration (10 m
M

), at which it is considered to selectively block R-type VDCCs (Avery and Johnston, 1996; Imaizumi et al., 1999), a signi®cant enhancement of the NMDA-stimulated in¯ux of Ca2 into the rat retina was found. However, when Ni2 was applied at a higher concentration (100 m
M

), able to block T-type VDCCs (Avery and Johnston, 1996; Imaizumi et al., 1999), no signi®cant effect was observed. This biphasic effect of Ni2 on NMDA-induced in¯ux of Ca2 could be ascribed to a direct potentiation (at low concentrations) or inhibition (at high concentrations) of the NMDA receptor activity, as previously reported in cultured rat cerebellar granule cells (Eimerl and Schramm, 1993). The unavailability of other speci®c blockers of R- and T-type VDCCs (compounds used as T-type antagonists such as ¯unarizine and mibefradil can also block other VDCCs) prevented us from exploring in more detail the possible role of these VDCCs in the NMDA-evoked in¯ux of Ca2 into the rat retina. Nevertheless, the fact that the non-speci®c VDCC blocker Co2 caused an inhibition of NMDAstimulated in¯ux of Ca2 into the rat retina almost identical to that elicited by nifedipine strongly suggests that only L-type VDCCs play a major role in the in¯ux of Ca2 into the rat retina triggered by NMDA. The ®nding that L-type VDCCs predominantly contribute (up to 50 %) to the NMDA-stimulated in¯ux of Ca2 into the rat retina contrasts with evidence obtained in brain neurones about the involvement of VDCCs in glutamate-induced rises in intracellular Ca2 . In cultured rat cerebellar granule neurones, L-, N- and P/Q-type VDCCs have been shown to be involved in intracellular Ca2 signals

evoked by 50±200 m
M

NMDA (Qiu et al., 1998; Netzeband et al., 1999). In rat embryonic lumbar motoneurones in situ, blockers of L-, N- and P/Q-type produced approximately a 25 % inhibition of glutamate-induced rises in intracellular Ca2 in each case, a 50% inhibition being observed with the joint application of these drugs (Metzger et al., 2000). Furthermore, both L- (approximately 52 %) and N-type (approximately 33 %) VDCCs have been reported to mediate the in¯ux of Ca2 stimulated by kainate in cultured rat hippocampal neurones (AmbroÂsio et al., 1999). In order to elucidate whether differences in the involvement of VDCCs in intracellular Ca2 responses mediated by glutamate receptor may re¯ect a distinct distribution of VDCCs in the brain and retina, the number of L-, N- and P/Q-type VDCCs in both the rat
CALCIUM CHANNELS IN THE RAT RETINA 399

cortex and retina was assessed by using radioligand binding techniques. Binding characteristics of [3H]nitrendipine and [125I]o-CgTx-MVIIA to L- and N-type VDCCs, respectively, in rat cortical membranes were in accordance with data previously published (Ehlert et al., 1982; Gould, Murphy and Snyder, 1984; Stoehr and Dooley, 1993; Hisamoto et al., 1998). Although no data are available about the binding of [125I]o-CgTx-MVIIC to rat cortical membranes, the ®nding of approximately a two-fold higher number of binding sites for [125I]oCgTx-MVIIC than for [125I]o-CgTx-MVIIA in the rat cortex is in good agreement with the results obtained by Kristipati et al. (1994) in bovine and rat cortical synaptosomes. Under the radioligand binding conditions employed here, [125I]o-CgTx-MVIIC selectively labels P/Q-type VDCCs (Kristipati et al., 1994). In the rat retina, we found a reduction of the number of radioligand binding sites associated with L- (threefold), N- (ten-fold) and P/Q-type (six-fold) VDCCs when compared to the cortex. The most abundant retinal VDCC was the P/Q-type, as in the cortex, followed by the L-type and, ®nally, by the N-type, which is much reduced in comparison with its levels in the cortex. A similar distribution pattern of VDCCs in the rat retina was reported by Kamphuis and Hendriksen (1998) by analysing levels of mRNAs encoding the various a

1

subunits of the VDCCs. These data indicate that the apparent lack of contribution of N- and P/Q-type VDCCs to the NMDA-stimulated Ca2 in¯ux into the isolated rat retina cannot be explained by an overall reduced presence of these channels in the retina with regard to the L-type channel. While the expression of L-type VDCCs in the retina is relatively increased as compared with the cortex (29% of the total of L-, Nand P/Q-type channels in the retina vs 15 % in the cortex), the relative levels of P/Q-type channels are the same in both tissues (55 %), and only N-type channels are markedly reduced in the retina (16 % in the retina vs 30% in the cortex). Taken together, these data suggest that L-type VDCCs play a predominant role in the total intracellular Ca2 rise evoked by NMDA in the isolated rat retina. The possibility exists, however, that N- or P/Qtype VDCCs may contribute to NMDA-elicited Ca2 signals in certain retinal cells, as for example the ganglion cells, and that potential contribution could not be detected in our experimental setting. N-type VDCCs have been shown to signi®cantly contribute (approximately 40 %) to the whole-cell Ca2 current in adult rat ganglion cells (Guenther et al., 1994; Schmid and Guenther, 1999), which express functional NMDA receptors (Thoreson and Witkovsky, 1999). The limitations of the technique used, able to detect changes in the total intracellular Ca2 but not in free intracellular Ca2 concentration, could have led to an underestimation of the role of N- or P/Q-type VDCCs in the NMDA-induced Ca2 in¯ux into the isolated rat retina. The apparent lack of effect of N- and

P/Q-type VDCC blockers on the in¯ux of Ca2 evoked by NMDA could also be ascribed to potential dif®culties of these high molecular weight peptides in reaching their sites of action in the isolated rat retina. Studies on hippocampal slices indicate, however, that these toxins do readily penetrate into wholetissue preparations (Fox, 1994; Keith et al., 1995). Interestingly, [3H]nitrendipine showed a signi®cantly lower af®nity for binding sites in the rat retina than in the cortex, which may result from a different expression in these tissues. of a
1

Nitrendipine subunit isoforms of the L-type channel the L-type VDCC and hence binds alterations to the a in 1 subunit particular of amino binding acid characteristics(Striessnig sequences of the a
1

subunit et al., 1998). can modify Nitrenits dipine exhibits, for example, a much lower af®nity for the L-type VDCC of skeletal muscle, which expresses the cerebral a
1S

isoform, cortex, than for the L-type channel of the Kamphuis and formed Hendriksen by the (1998) a
1D

(Gould reported et al., that 1984). the main isoform of L-type channel the rat retina isoform in the is cortex the a
1D

so , which that is a
1

also subunit the predominant expressed in differences in [3H]nitrendipine af®nity would seem unlikely. However, a new member isoform, of has the been L-type recently VDCC a identi®ed 1 subunit family, in the the retina a
1F

(Bech-Hansen et al., 1998; Strom et al., 1998). This a demonstrated 1F isoform, absent in the from the brain and so far only

retina, shows some amino acid variations in the transmembrane domains critical for conferring dihydropyridine sensitivity that may account for the different af®nity of [3H]nitrendipine for cortical and retinal L-type channels. Ca2 signaling mediated by NMDA receptors appears to be of ultimate importance for retinal synaptic transmission and development (Thoreson and Witkovsky, 1999; Fletcher et al., 2000; GruÈnder et al., 2000). Additionally, overstimulation of NMDA receptors plays a key role in ischaemic/excitotoxic insults to the retina, as may occur in glaucoma (Dreyer, 1998; Osborne et al., 1999a), leading to neuronal cell death due to intracellular Ca2 overload (Sucher, Lei and Lipton, 1991; Kitano, Morgan and Caprioli, 1996; Zhang et al., 2000). The ability of L-type VDCC blockers to decrease NMDA-stimulated Ca2 in¯ux into the retina could therefore be responsible for the neuroprotective effects of these drugs in paradigms of ischaemia/excitotoxicity (Crosson, Willis and Potter, 1990; Sucher et al., 1991; Takahashi et al., 1992; Block and Schwarz, 1998; Toriu et al., 2000). To date, no evidence is available supporting a retinal neuroprotective effect of blockers of N- or P/Q-type VDCC in models of ischaemia/excitotoxicity, which may be related to the apparent inability of these drugs to inhibit NMDAevoked in¯ux of Ca2 into the retina reported here. It should be noted, however, that N- or P/Q-type VDCC blockers may theoretically exert neuroprotective effects by reducing glutamate release during or after
400 J. MELENA AND N. OSBORNE

retinal ischaemia, as described in the brain (Kimura et al., 1998; Kobayashi and Mori, 1998). Nevertheless, in contrast to the brain, L-type VDCCs also appear to play a major role in controlling the release of glutamate in the retina (Thoreson and Witkovsky, 1999), thus providing a potential additional mechanism for explaining the retinal neuroprotective effects of the blockers of these channels. In conclusion, this study demonstrates that NMDAstimulated in¯ux of Ca2 into the isolated rat retina can be markedly attenuated by the blockade of L-type VDCCs. On the contrary, other types of VDCCs do not seem to play a major role in the in¯ux of Ca2 evoked by NMDA. This different involvement of VDCCs in the

intracellular Ca2 rises elicited by NMDA in the retina and in the brain does not appear to result from a distinct expression pattern of L-, N- or P/Q-type VDCCs in both tissues, although some variations were found. Interestingly, a different af®nity of [3H]nitrendipine for L-type VDCCs in the rat retina and cortex was observed which may re¯ect the expression of different classes of L-type channels in these tissues. This ability of L-type VDCC blockers to attenuate NMDA-stimulated Ca2 in¯ux may underlie their neuroprotective effects in the retina. Future work is needed to elucidate the role of VDCCs in the release of glutamate in the ischaemic retina. Acknowledgements
J. Melena was supported by a post-doctoral Marie Curie grant (TMR programme, European Commission).

References
AmbroÂsio, A. F., Silva, A. P., Malva, J. O., Soares-da-Silva, P., Carvalho, A. P. and Carvalho, C. M. (1999). Carbamazepine inhibits L-type Ca2 channels in cultured rat hippocampal neurones stimulated with glutamate receptor agonists. Neuropharmacol. 38, 1349±59. Avery, R. B. and Johnston, D. (1996). Multiple channel types contribute to the low-voltage-activated calcium current in hippocampal CA3 pyramidal neurons. J. Neurosci. 16, 5567±82. Bech-Hansen, N. T., Naylor, M. J., Maybaum, T. A., Pearce, W. G., Koop, B., Fishman, G. A., Mets, M., Musarella, M. A. and Boycott, K. M. (1998). Loss-of-function mutations Xp11 . 23 in cause a incomplete calcium-channel X-linked a 1 -subunit gene in congenital stationary night blindness. Nature Genet. 19, 264±7. Birnbaumer, L., Campbell, K. P., Catterall, W. A., Harpold, M. M., Hofmann, F., Horne, W. A., Mori, Y., Schwartz, A., Snutch, T. P., Tanabe, T. and Tsien, R. W. (1994). The naming of voltage-gated calcium channels. Neuron 13, 505±6. Block, F. and Schwarz, M. (1998). The b-wave of the electroretinogram as an index of retinal ischemia. Gen. Pharmacol. 30, 281±7. Collingridge, G. L. and Singer, W. (1990). Excitatory amino acid receptors and synaptic plasticity. Trends Pharmacol. Sci. 11, 290±6. Crosson, C. E., Willis, J. A. and Potter, D. E. (1990). Effect of the calcium antagonist, nifedipine, on ischemic retinal dysfunction. J. Ocul. Pharmacol. 6, 293±9.

Dreyer, E. B. (1998). A proposed role for excitotoxicity in glaucoma. J. Glaucoma 7, 62±7. Dunlap, K., Luebke, J. I. and Turner, T. J. (1995). Exocytotic Ca2 channels in mammalian central neurons. Trends Neurosci. 18, 89±98. Ehlert, F. J., Roeske, W. R., Itoga, E. and Yamamura, H. I. (1982). The binding of [3H]nitrendipine to receptors for calcium channel antagonists in the heart, cerebral cortex, and ileum of rats. Life Sci. 30, 2191±202. Eimerl, S. and Schramm, M. (1993). Potentiation of 45Ca uptake and acute toxicity mediated by the N-methylD aspartate receptor: the effect of metal binding and transition metal ions. J. Neurochem. 61, 518±25. Fletcher, E. L., Hack, I., BrandstaÈtter, J. H. and WaÈssle, H. (2000). Synaptic localization of NMDA receptor subunits in the rat retina. J. Comp. Neurol. 420, 98±112. Fox, J. A. (1994). Novel omega-conopeptides reduced ®eld potential amplitudes in the rat hippocampal slice. Neurosci. Lett. 165, 157±60. Gould, R. J., Murphy, K. M. M. and Snyder, S. H. (1984). Tissue heterogeneity of calcium channel antagonist binding sites labeled by [3H]nitrendipine. Mol. Pharmacol. 25, 235±41. GruÈnder, T., Kohler, K., Kaletta, A. and Guenther, E. (2000). The distribution and developmental regulation of NMDA receptor subunit proteins in the outer and inner retina of the rat. J. Neurobiol. 44, 333±42. Guenther, E., Rothe, T., Taschenberger, H. and Grantyn, R. (1994). Separation of calcium currents in retinal ganglion cells from postnatal rat. Brain Res. 633, 223±35. Hisamoto, M., Inaoka, Y., Sakaida, Y., Kagazaki, T., Enokida, R., Okazaki, T., Haruyama, H., Kinoshita, T. and Matsuda, K. (1998). A-53930A and B, novel N-type Ca2 channel blockers. J. Antibiot. 51, 607±17. Imaizumi, T., Kocsis, J. D. and Waxman, S. G. (1999). The role of voltage-gated Ca2 channels in anoxic injury of spinal cord white matter. Brain Res. 817, 84±92. Kamphuis, W. and Hendriksen, H. (1998). Expression patterns of voltage-dependent subunits Res. 55, 209±20. (a 1A ±a 1E ) mRNA in calcium channel rat retina. Mol. Brain. a 1 Keith, R. A., Mangano, T. J., Lampe, R. A., DeFeo, P. A., Hyde, M. J. and Donzanti, B. A. (1995). Comparative

actions of synthetic omega-grammotoxin SIA and synthetic omega-Aga-IVA on neuronal calcium entry and evoked release of neurotransmitters in vitro and in vivo. Neuropharmacology 34, 1515±28. Kimura, M., Sawada, K., Miyagawa, T., Kuwada, M., Katayama, K. and Nishizawa, Y. (1998). Role of glutamate receptors and voltage-dependent calcium and sodium channels in the extracellular glutamate/ aspartate accumulation and subsequent neuronal injury induced by oxygen/glucose deprivation in cultured hipocampal neurons. J. Pharmacol. Exp. Ther. 285, 178±85. Kitano, S., Morgan, J. and Caprioli, J. (1996). Hypoxic excitotoxic damage to cultured rat retinal ganglion cells. Exp. Eye Res. 63, 106±12. Kobayashi, T. and Mori, Y. (1998). Ca2 channel antagonists and neuroprotection from cerebral ischemia. Eur. J. Pharmacol. 363, 1±15. Kristipati, R., NaÂdasdi, L., Tarczy-Hornoch, K., Lau, K., Miljanich, G. P., Ramachandran, J. and Bell, J. R. (1994). Characterization of the binding of omegaconopeptides to different classes of non-L-type neuronal calcium channels. Mol. Cell. Neurosci. 5, 219±28. Lee, J.-M., Zipfel, G. J. and Choi, D. W. (1999). The changing landscape of ischaemic brain injury mechanisms. Nature 399, A7±14. CALCIUM CHANNELS IN THE RAT RETINA 401 Mayer, M. L., Westbrook, G. L. and Guthrie, P. B. (1984). Voltage-dependent block by Mg2 of NMDA responses in spinal cord neurones. Nature 309, 261±3. Metzger, F., Kulik, A., Sendtner, M. and Ballanyi, K. (2000). Contribution of Ca2 -permeable AMPA/KA receptors to glutamate-induced Ca2 rise in embryonic lumbar motoneurons in situ. J. Neurophysiol. 83, 50±9. Netzeband, J. G., Conroy, S. M., Parsons, K. L. and Gruol, D. L. (1999). Cannabinoids enhance NMDA-elicited Ca2 signals in cerebellar granule neurons in culture. J. Neurosci. 19, 8765±77. Osborne, N. N., Ugarte, M., Chao, M., Chidlow, G., Bae, J. H., Wood, J. P. M. and Nash, M. S. (1999b). Neuroprotection in relation to retinal ischemia and relevance to glaucoma. Surv. Ophthalmol. 43, S102±S8. Osborne, N. N., Wood, J. P. M., Chidlow, G., Bae, J.-H., Melena, J. and Nash, M. S. (1999a). Ganglion cell death in glaucoma: what do we really know?. Br. J. Ophthalmol. 83, 980±6. Ozawa, S., Kamiya, H. and Tsuzuki, K. (1998). Glutamate receptors in the mammalian central nervous system. Prog. Neurobiol. 54, 581±618. Qiu, Z., Sweeney, D. D., Netzeband, J. G. and Gruol, D. L. (1998). Chronic interleukin-6 alters NMDA receptor-

mediated membrane responses and enhances neurotoxicity in developing CNS neurons. J. Neurosci. 18, 10445±56. Ritch, R. (2000). Neuroprotection: is it already applicable to glaucoma therapy?. Curr. Opin. Ophthalmol. 11, 78±84. Schmid, S. and Guenther, E. (1999). Voltage-activated calcium currents in rat retinal ganglion cells in situ: changes during prenatal and postnatal development. J. Neurosci. 19, 3486±94. Stoehr, S. J. and Dooley, D. J. (1993). Characteristics of [125I]o-conotoxin MVIIA binding to rat neocortical membranes. Neurosci. Lett. 161, 113±6. Striessnig, J., Grabner, M., Mitterdorfer, J., Hering, S., Sinnegger, M. J. and Glossmann, H. (1998). Structural basis of drug binding to L Ca2 channels. Trends Pharmacol. Sci. 19, 108±15. Strom, T. M., Nyakatura, G., Apfelstedt-Sylla, E., Hellebrand, H., Lorenz, B., Weber, B. H. F., Wutz, K., Gutwillinger, N., RuÈther, K., Drescher, B., Sauer, C., Zrenner, E., Meitinger, T., Rosenthal, A. and Meindl, A. (1998). An L-type calcium-channel gene mutated in incomplete X-linked congenital stationary night blindness. Nature Genet. 19, 260±3. Sucher, N. J., Lei, S. Z. and Lipton, S. A. (1991). Calcium channel antagonists attenuate NMDA receptormediated neurotoxicity of retinal ganglion cells in culture. Brain Res. 297, 297±302. Takahashi, K., Lam, T. T., Edward, D. P., Buchi, E. R. and Tso, M. O. M. (1992). Protective effects of ¯unarizine on ischemic injury in the rat retina. Arch. Ophthalmol. 110, 862±70. Thoreson, W. B. and Witkovsky, P. (1999). Glutamate receptors and circuits in the vertebrate retina. Prog. Retin. Eye Res. 18, 765±810. Toriu, N., Akaike, A., Yasuyoshi, H., Zhang, S., Kashii, S., Honda, Y., Shimazawa, M. and Hara, H. (2000). Lomerizine, a Ca2 channel blocker, reduces glutamate-induced neurotoxicity and ischemia/reperfusion damage in the rat retina. Exp. Eye Res. 70, 475±84. Zhang, S., Kashii, S., Yasuyoshi, H., Honda, Y., Ujihara, H., Sasa, M., Tamura, Y. and Akaike, A. (2000). Involvement of NMDA-receptor in kainate-induced neurotoxicity in cultured fetal retinal neurons. Graefe's Arch. Clin. Exp. Ophthalmol. 238, 243±8.