Protection by Eliprodil against in Cultured Rat Retinal Ganglion Cells

Iok-Hou Pang,1 Eric M. Wexler,2 Scott Nawy,2 Louis DeSantis,1 and Michael A. Kapin1

PURPOSE. TO test whether eliprodil (SL 82.0715), a unique antagonist for the N-methyl-D-aspartate (NMDA) receptor, is protective in the glutamate-induced cytotoxicity model in cultured rat retinal ganglion cells (RGCs). METHODS. TWO to four days after a fluorescent dye, Di-I, was injected near the superior colliculi, neonatal rats were killed, and retinal cells were dissociated and cultured. Survival of RGCs after drug treatment was assayed by counting Di-I fluorescent cells.

RESULTS. In rat RGCs, glutamate-induced toxicity with a mean EC50 of 10.7 JUM. Only 47% of RGCs survived after a 3-day treatment with 100 JUM glutamate. Studies using selective agonists and antagonists indicated that the glutamate-induced toxicity was mediated largely by the NMDA receptor. Pretreatment with eliprodil protected against such toxicity. Eliprodil exhibited a mean IC50 of 1.0 nM (log [IC50] = -9.00 ± 0.01, mean ± SEM, n = 3; against cell death produced by 100 /xM glutamate). At 1 /xM, eliprodil was maximally protective; cell survival in the presence of 100 piM glutamate challenge was 100% ± 5% (n = 3). This protective effect of eliprodil may be related to its reduction (by 78%) of NMDA-induced currents recorded under patch-clamp recording in these cells. CONCLUSIONS. Eliprodil is protective against glutamate cytotoxicity in retinal neurons. It may be a useful novel compound for the treatment of retinopathies including glaucoma in which excitotox- icity has been implicated. (Invest Ophthalmol Vis Sci. 1999;40:1170-1176)

any pathologic changes in the retina and optic nerve treatment, a single intravitreal injection of glutamate or N- head, such as damage related to glaucoma or isch- methyl-D-aspartate (NMDA) in adult rats induces a dose-depen- emia, are thought to be mediated in part by excita- dent loss of cells in the GCL, without obvious alterations in the M 6 8 tory amino acids (EAA). For example, high intraocular pressure more distal retinal layers. " Such toxicity is blocked by has been regarded as an important risk factor for glaucomatous MK801, an NMDA receptor antagonist.6 In vitro studies using optic neuropathy. In ocular hypertensive monkeys, glutamate chick retina or isolated rat retinal ganglion cells (RGCs) also concentration in the vitreous was found to significantly in- showed that EAAs produce cytotoxicity that can be antago- crease compared with ocular normotensive eyes.1 Similarly, nized by pretreatment with antagonists for EAA receptors.9""' elevated vitreal glutamate concentration was also observed in In addition to glaucoma, retinal ischemia can also lead to open-angle glaucoma patients.' Excessive EAA is toxic to cells EAA-mediated retinal cell injury.15 In rabbits, retinal ischemia, in the retina, especially ganglion cells. Lucas and Newhouse2 induced by raising intraocular pressure, causes a reduction in demonstrated that systemic administration of glutamate in electroretinogram b-wave amplitude. Pretreatment with dex- young mice, whose blood-retina barrier is yet to be completed, tromethorphan, a weak NMDA receptor antagonist, enhances destroyed the inner retinal layers, primarily the ganglion cell the postischemic recovery of b-wave amplitude, suggesting the layer (GCL). More recent light microscopic and electron mi- involvement of NMDA receptors in this functional change of croscopic studies corroborated such findings.3"5 Azuma et al.3 the retina.l6 In agreement with this finding, a twofold increase further reported that treatment of neonatal rats with subcuta- of glutamate release in the rabbit vitreous has been detected neous glutamate caused a "deeply excavated" optic disc, a during retinal ischemia, and a sevenfold increase in glutamate condition characteristic of glaucoma. In addition to systemic release has been observed during initial reperfusion after re- lease of intraocular pressure.17 For the time being, the origin and mechanisms of this glutamate release are still unclear. However, it is believed that prolonged deprivation of oxygen From the 'Alcon Laboratories, Fort Worth, Texas; the depart- ment of Ophthalmology and Visual Sciences, Albert Einstein College of and nutrients to neuronal tissues leads to cell membrane de- Medicine, Bronx, New York. polarization, which, in turn, increases synaptic release of EAA Supported by Alcon Laboratories, Fort Worth, Texas. and reduces EAA reuptake, together resulting in a buildup of Presented in part at the annual meeting of the Association for extracellular EAA, causing lethal injury to sensitive neuronal Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 18 1997, and of the Society for Neuroscience, San Diego, California, cell types. November 1996. Eliprodil (SL 82.0715) is a potent, noncompetitive NMDA Submitted for publication March 27, 1998; revised July 29, 1998; receptor antagonist.19'20 It has been shown to be protective accepted September 21, 1998. against ischemia-induced neuronal death, while lacking many Proprietary interest category: E. Reprint requests: Iok-Hou Pang, Alcon Laboratories, R3-24, 6201 of the central nervous system side effects characteristic of 19 20 South Freeway, Fort Worth, TX 76134. other EAA antagonists, ' making it a good candidate for

Investigative Ophthalmology & Visual Science, May 1999, Vol. 40, No. 6 1170 Copyright © Association for Research in Vision and Ophthalmology

Downloaded from jov.arvojournals.org on 09/27/2021 IOVS, May 1999, Vol. 40, No. 6 Eliprodil Protects Retinal Ganglion Cells 1171

chronic retinoprotective therapy. Therefore, we assessed the Di-I-labeled fluorescent cells in 20 microscopic fields were effects of eliprodil on EAA-induced neuronal conductance and counted and averaged. delayed neuronal death of cultured rat RGCs. Electrophysiology of RGCs The cell-culturing technique used in the electrophysiological studies was slightly different from that involved in the survival MATERIALS AND METHODS studies. Retinae from newborn rats were isolated and incu- RGCs isolated from neonatal Sprague-Dawley rats were used in bated for 1 hour in 5 U/ml papain dissolved in DMEM contain- ing 25 mM Hepes and supplemented with 1.5 mM cysteine, 0.5 survival and electrophysiological studies. The treatment of an- 2+ imals and procedures used in these studies conformed to the mM EDTA, and 2 mM Ca . They were then gently triturated ARVO Statement on the Use of Animals in Ophthalmic and through a fire-polished Pasteur pipette and resuspended in Vision Research. growth medium containing DMEM, 2 mM , 17 mM KC1, 5% heat-inactivated fetal calf serum, and serum supple- ment (MITO; Collaborative Research, Grand Island, NY). Cells RGC Survival Assay were grown on glass coverslips coated with poly-D-lysine in Techniques for the isolation and culture of RGCs were adapted 24-well plates. 21 from those reported by Takahashi et al. The procedure in- Patch pipettes were pulled from hematocrit glass, fire- volved retrograde labeling of ganglion cells by injecting a polished to a resistance of 2 Mil to 4 Mil, and filled with a fluorescent dye, Di-I, into the superior colliculi. Two to 4 days solution containing NaH2PO4 120 mM, NaCl 10 mM, Hepes 10 later, retinal cells were dissociated. Cultured RGCs were iden- mM, EGTA 10 mM, and MgATP 4 mM. The extracellular solu- tified by sufficient Di-I fluorescence to allow them to be tion contained NaCl 160 mM, CaCl2 2 mM, and Hepes 5 mM. observed visually using a fluorescence microscope. The de- The pH values of both solutions were adjusted to 7.4 with tailed procedures follow. NaOH. Recordings were obtained using an Axopatch ID am- Neonatal Sprague-Dawley rats 2 to 5 days old were anes- plifier. Drugs were applied via a series of computer-controlled thetized by chilling on ice for 10 minutes, after which, a 2-mm flow pipes (outer diameter, 358 /am). midline opening was made in the scalp just caudal to the Whole-cell recordings were obtained from RGCs cultured traverse sinus. The tip of the injection needle (30-gauge) was for 10 to 14 days. Ganglion cells were distinguished from inserted 6 mm below top of the skull, and a 5-JLII Di-I solution, amacrine cells, the other class of neurons with regenerative containing 3 mg/ml Di-I (l,l'-dioctadecyl-3,3,3',3'-tetrameth- sodium currents, by their greater size and larger resting poten- ylindo-carbocyanine perchlorate; Molecular Probes, Eugene, tial, and a more slowly accommodating action potential.22 OR) in 90% ethanol and 10% dimethyl sulfoxide, was injected. Virtually every ganglion cell examined at this age responded to The wound was then covered with a drop of Flexible Collo- NMDA. dion (Amend Drug & Chemical, Irvington, NJ). Rats were returned to their mother after warming and recovery from Test Compounds anesthesia. Glutamate, quisqualate, NMDA, and MK801 () were Two to 4 days after Di-I injection, rats were anesthetized obtained from Research Biochemicals (Natick, MA). Eliprodil by hypothermia and killed by decapitation. Their eyes were HC1 was provided by Synthelabo Recherche (Bagneux, enucleated and placed in Dulbecco's modified Eagle's medium: France). All other chemicals were obtained from Sigma Chem- Nutrient mixture F12 (1:1, DMEM/F12; Gibco, Grand Island, ical. NY). The retina from each eye was detached and isolated. Retinal cells were dissociated by combining 12 retinae with 5 ml of papain solution, containing 10 mg papain (34 units/ml; RESULTS Sigma, St. Louis, MO), 2 mg DL-cysteine (33 mM; Sigma), and 2 mg bovine serum albumin (0.4 mg/ml; Sigma) in 5 ml of Retrograde uptake of the fluorescent dye Di-I after its injection DMEM/F12, for 25 minutes at 37°C, then washed 3 times with into the superior colliculi should label only ganglion cells in the 5 ml RGC medium (DMEM [Gibco], supplemented with 10% retina. Thus, the presence of Di-I fluorescence distinguished fetal bovine serum [Hyclone, Logan, UT], 4 mM glutamine RGCs from other retinal cells in culture (Fig. 1). After 3 days in [Gibco], 100 U/ml penicillin, and 100 ju-g/ml streptomycin culture, the surviving RGCs appeared to be morphologically [Sigma]). Additional RGC medium was added to the retinal healthy. Ganglion cells adhered to poly-D-lysine-coated cul- pieces to a final total volume of 40 ml. Retinal pieces were ture dishes with flat cell bodies and occasional neuritelike triturated by passing through a disposable pipette several times extensions (Figs. 1A, IB). Treatment with glutamate for 3 days until cells were dispersed. Cell suspension (1.5 ml; containing was toxic to RGCs. It increased the probability of cells with an approximately 4.5 X 106 cells) was placed into each of the aberrant appearance, such as the presence of many vacuole- poly-D-lysine coated glass-bottomed culture dishes. The cells like structures in the cytoplasm (Figs. IE, IF); and cell frag- were cultured for 3 days in 95% air/5% CO2 at 37°C. ments (Figs. 1C, ID). Additionally, glutamate treatment also In experiments assessing excitotoxicity, glutamate or decreased the total number of fluorescent cells. Under control other excitatory amino acid analogues were added to the cells conditions, there were 11.4 ± 0.9 Di-I-labeled cells/micro- 30 minutes after isolation and incubated for 3 days at 37°C with scopic field (mean ± SEM of 30 studies, 20 individual fields 5% CO2. For experiments involving MK801 or eliprodil, the were counted in each study). Glutamate decreased the number antagonist was added to the cells 30 minutes before the addi- of surviving cells in a concentration-dependent fashion. A rep- tion of glutamate. Three days later, the cells were observed resentative concentration-response curve for glutamate-in- with a fluorescence microscope at 200X magnification, and duced excitotoxicity is shown in Figure 2A. The mean EC50 for

Downloaded from jov.arvojournals.org on 09/27/2021 1172 Pang et al. IOVS, May 1999, Vol. 40, No. 6

FIG~IRE1. Photomicrographs of cultured rat retinal cell$. (A. C. E) Photographs taken under pl1;t.x optics; (B, D. F) IluoreKence pliotogccphs of the s;lnic cells. (A. R) (:ells culrc~retluntlcr control conditions. Scale bar. 50 Fm. (C,D, E. F) Ex;~niplesof cells with ahemnt n~orpliology;~fter they wcrc trr~tetlwith 100 ph4 glut:~rnatefor 3 days. Scale har. IS pm.

glutaniate was calcul;~tedto be 10.7 pM (log IE<:,,,1 = -4.97 2 sible for the majority of glutamate-induced toxicity in these 0.13, mean 5 SEM, t~ = 3). Compared to controls. only 47'X ? cells. To further confirm this finding. a selective NMIlA recep 2'6 (n = 14) of cells survived after a 3-day treatment of 100 pM tor antagonist. MK801, was tested for its ability to antagonize of glutamate. Currently, we do not know whether cells that the excitotoxic effects of glutamate. MKS01 was highly prtr sunriwd the glutamate treatment were different from those tective against the glutamate cytotoxicity. with maximal prtr th~tdid not. However, treating the cells with glutanrate for 5 tection of 100%. h representative cot~centcition-respo~ise clays did not further decrease the survival rate (data not curve for MKHOl protection is shown in Figure 3A. The mean shown). In initial pilot studies in which retinal cells were IC,,, value of MKHO1 was 20.9 nM (log IIC,,,l = -7.68 ? 0.4 1. cocultured with additional Miiller cells. glutamate was much 11 = 3) in the presence of 100 pM glutamatc, which translates less potent; its apparent EC,,, tlecreasecl by 10- to 100-fold (data to an apparent pKi for MKS01 of 8.69 ? 0.41. These values not shown). agree well with the published affinity of MKHOl for NMDA (;lutamate interacts with many receptor subtypes." To receptors. define the contribution of v;~riousglutamatr receptor sub- Pretreatment of the retinal cells with eliprodil for 30 type(~)to glutamate-induced damage, we compared the exci- minutes protected cells against the toxic effects of glutamatc. totoxicity of a variety of subtype-selective glutamate agonists. Figure 3R illustntes the concentrdtion-effect relation for elip- NMDA (selective to NMDA receptors), kainate (selective to rodil. The protective effect of eliprodil exhibited a mean IC,,, ionotropic AMPA/kainate receptors), and quisqualate (interacts of 1.0 nM (log IIC,,,] = -9.00 + 0.01. mean -+ SEM. PI = 3) with both the ionotropic AMPWkainatc receptors and metabo- against 100 pM glutamate. At 1 pM. eliprodil was maximally tropic receptors) were tested on nt RCCs. As shown in Figure protective; cell survival was 100011 + 5% (PI = 3) despite thr 2H. at 100 pM, only NMDA ;mcl glutamate c;iused toxicity. presence of glutanrate. Kainate and quisqualate were ineffective. NMDA-induced tox- To investigate potential mechanism(s) involved in the icity was also conc-entmtion-depenclent (Fig. LC), with a mean protective action of eliprodil, we directly measured the effect EC,,, of 1.82 pM (log [E<:,,,l - -5.74 f 0.55. mean 2 SEM, of eliprodil on EAA-induced inward currents in KGCs. Figure 4 11 = 3).'l'hcsc results suggest that NMDA receptors are respon- shows inward currents evoked in RGCs by npid application of

Downloaded from jov.arvojournals.org on 09/27/2021 IOVS, May 1999, Vol. 40, No. 6 Eliprodil Protects Retinal Ganglion Cells 1173

120 200 /LtM NMDA plus 1 /LLM . (This concentration of glycine was added to mimic the saturating concentrations of glycine found in culture media.) Eliprodil reduced the NMDA- 100 H evoked response in a dose-dependent fashion. Both the peak and steady state components of the response were reduced. The peak responses versus eliprodil concentrations are plotted in the lower panel of Figure 4. The maximal blockade of the NMDA response by eliprodil was approximately 78%, with a calculated IC50 of 2.3 ptM. Incomplete block of NMDA currents by eliprodil has been observed previously in cortical neurons2'1 and by , an eliprodil analogue, on hippocampal neu- rons.25 The onset of inhibition of NMDA inward currents by -oo .7 -6 -5 -4 -3 eliprodil was relatively rapid, as is illustrated in Figure 5. The left trace of the upper panel shows a current response to log [Glutamate] (M) NMDA alone. Starting with the next trace, NMDA was ap- plied for 2 seconds at 16-second intervals, and eliprodil was applied continuously to the cell. The lines above each 120 record indicate NMDA application. Almost all the total blockable current was inhibited with the first application of NMDA (middle versus right traces). In contrast, recovery of (0 | CO

£ •y jb ^ i

-co

log [NMDA] (M) FIGURE 2. Effect of excitatory amino acids on cell survival in rat RGCs. 20 Cells were treated with the indicated amounts of compounds, and survival -00 -10 was determined by counting the survived, Di-I-labeled cells. Survival with vehicle alone defines 100%. (A) A representative concentration-response log [Eliprodil] (M) curve of glutamate. Each symbol represents a single datum point. Similar results were obtained in three independent studies each in duplicate. (B) FIGURE 3. Effect of NMDA receptor antagonists on glutamate (100 Effect of various compounds (100 /iM each) on survival of rat RGCs. Each /u,M>induced cytotoxicity of nit RGCs. Representative concentration- bar represents mean result of each treatment ± SEM. Numbers of studies response curves of MK801 (A) and eliprodil (B) are shown. Survival are shown at the bottom of each bar. (C) A representative concentration- with vehicle alone defines 100%. Each symbol represents a single response curve of NMDA. Each symbol represents a single datum point. datum point. Similar results were obtained in 3 independent studies, Similar results were obtained in 3 independent studies, each in duplicate. each in duplicate.

Downloaded from jov.arvojournals.org on 09/27/2021 1174 Pang et al. IOVS, May 1999, Vol. 40, No. 6

Eliprodil 0.2 |iM 0 LI 20 (iM fSSSSSSSSSS/SSSSSSSSSSSSSSSj 'SSSSSSSSSSSSSSSSSSSSSSSSSSJ fSSSSSSSSSSSSSSSSSSSSSSSSSSJ. WSJ. NMDA 200

-00 -7 -6 -4 log [Eliprodil] (M)

FIGURE 4. Current evoked by NMDA is reduced in the presence of eliprodil. Top: four sample records showing response to 200 yM NMDA alone and in the presence of 3 concentrations of eliprodil. Here eliprodil was applied before and during exposure to NMDA, but results were similar when they were coapplied. Holding potential: -50 mV. Bottom: peak NMDA response for 4 cells at 3 concentrations of eliprodil. Data were fitted by the equation: (1 + Max) (1 + ([EliprodilJ/ICj,,)^) + Max, where Max is the

fraction of residual current in the presence of saturating concentrations of eliprodil; IC50 is the concen- tration of eliprodil that produced 50% inhibition; Eliprodil is the concentration of eliprodil; and N is the

Hill coefficient. The curve is obtained with Max = 0.22, IC50 = 2.3 /xM, N = 1.16.

inhibition after washout of eliprodil occurred much more with application of glutamate, only approximately half of the slowly. The first application of NMDA in the absence of RGCs were susceptible to the toxic insult. In our experimental eliprodil revealed only slight recovery (top right versus conditions, we never observed a complete loss of the cells. It bottom left), and only approximately 50% of the current was is currently unclear whether the neurons that survived the restored after 5 applications (bottom panel). A slow off rate insult were phenotypically different from those that did not. of eliprodil has been reported previously.25 One possible explanation is suggested by the report of Dreyer et al.,10 that ganglion cells of larger size are more sensitive to NMDA receptor-mediated death. Studies are being conducted DISCUSSION to test this and other hypotheses. We have shown that in cultured rat RGCs, EAA markedly The RGC culture model used in the survival assay was 21 decreased survival. Although functional glutamate receptors of almost identical to that described by Takahashi et al., with NMDA and non-NMDA subtypes are expressed in these one major difference. In their study, a confluent monolayer of cells,26'27 EAA-induced toxicity appears to be mediated by glial cells (retinal Muller cells or cortical astrocytes) was NMDA receptors, because only glutamate and NMDA, but not present in the retinal cell culture to assist and sustain the quisqualate or kainate, were cytotoxic. Furthermore, the selec- growth of RGCs, whereas, in our studies, the feeder cells were tive NMDA antagonist MK801 was potent and efficacious in eliminated. We found that, by doubling the density of retinal preventing the excitotoxic effects of glutamate. Interestingly, cells in culture to approximately 3 X 106 cells/ml, sufficient

Downloaded from jov.arvojournals.org on 09/27/2021 JOVS, May 1999, Vol. 40, No. 6 Eliprodil Protects Retinal Ganglion Cells 1175

Control First Tenth application application

Eliprodil

100 PA| 1 sec

First Fifth Tenth application application application

*1 if*"1** Wash

FIGURE 5. Recovery from eliprodil inhibition of NMDA current is slow. Top records: responses to 200 JUM NMDA in the absence (Jeff) and presence (middle, rigbf) of 20 JLLM eliprodil. Time between applications was 16 seconds. Inhibition was nearly maximal during the first application of NMDA. Lower records: the time course of recovery. Holding potential: —50 mV.

Di-I-prelabeled and morphologically normal cells were avail- brief exposure used in the electrophysiological experiments. It able for quantification at the time of assay. More important, in would be interesting to measure NMDA responses in ganglion the absence of a feeder layer, the potency of glutamate toxicity cells that have undergone long-term exposure to eliprodil. increased by at least 10-fold.28 It is probable that Miiller cells Thus, we demonstrated that eliprodil was effective in and cortical astrocytes can interfere with the actions of gluta- blocking both glutamate-induced inward currents and cyto- mate by uptake into glial cells, decreasing its effective concen- toxicity in RGCs. These results agree with previous reports tration for the ganglion cells. that eliprodil is an antagonist affecting the polyamine site of Our electrophysiological results provide direct evidence NMDA receptors (cf. Ref. 19). Eliprodil and its analogues that ganglion cells in these cultures express NMDA receptors have been suggested to interact selectively with one subtype and that currents elicited by activation of these receptors are of NMDA receptor (i.e., one comprised of heteroligomeric 29 inhibited by eliprodil. These data provide a potential mecha- complexes of NR1A and NR2B). Pharmacologically, it has nism for the neuroprotective properties of eliprodil demon- been shown that eliprodil blocks NMDA-induced responses strated here, although a higher concentration of eliprodil is in a variety of neuronal cell types. For example, eliprodil required to block the NMDA current than to provide neuro- inhibits the depolarizing effects of NMDA on cultured protection. One possible explanation is the dramatic difference mouse spinal neurons,30 reduces the NMDA-induced inward in exposure times in these two experiments. The 3 days of current in cultured rat cortical neurons,24 suppresses exposure to eliprodil that are required for the neviroprotection NMDA-stimulated production of cGMP in cerebellar slices, experiments might have induced an additional high-affinity and blocks acetylcholine release in striatal slices.30 In addi- mechanism of NMDA block that is not apparent during the tion to excitotoxicity, eliprodil was also effective in reduc-

Downloaded from jov.arvojournals.org on 09/27/2021 1176 Pang et al. IOVS, May 1999, Vol. 40, No. 6

ing neuronal damage due to ischemia/anoxia, both in 14. Tung NN, Morgan IG, Ehrlich D. A quantitative analysis of the vivo20,31,32 and in hippocampal slices in vitro.; effects of excitatory neurotoxins on retinal ganglion cells in the The in vitro protective effects of eliprodil on retinal cells chick. Vis Neurosci. 1990;4:217-223. 15. Bresnick GH. Excitotoxins: a possible new mechanism for the we demonstrate here predicted that the compound may be pathogenesis of ischemic retinal damage. Arch Ophthalmol. 1989; protective against EAA-induced toxicity in the retina in vivo. 107:339-341. 34 Indeed, as demonstrated in the accompanying manuscript, 16. Yoon YH, Marmor MF. protects retina against eliprodil is effective in a dose-dependent fashion in protecting ischemic injury in vivo. Arch Ophthalmol. 1989;107:409-4ll. the rat retina from damages due to NMDA in vivo. Further- 17. Louzada-Junior P, Dias JJ, Santos WF, Lachat JJ, Bradford HF, more, it also minimized retinal functional changes caused by Coutinho-Netto J. Glutamate release in experimental ishaemia of pressure-induced ischemia in the rabbit eye. the retina: an approach using microdialysis. / Neurochem. 1992; Hence, our data show that eliprodil reduces EAA-evoked 59:358-363. 18. Drejer J, Benveniste H, Diemer NH, Schousboe A. Cellular origin of membrane conductance and ultimate cell loss in cultured rat 34 ischemia-induced glutamate release from brain tissue in vivo and RGCs. Taken together with our in vivo data, this suggests that vitro. / Neurochem. 1985;45:l45-151. eliprodil may be a useful compound for the treatment of 19- Scatton B, Avenet P, Benavides J, et al. Neuroprotective potential various retinopathies, especially those related to glaucoma, of the polyamine site-directed NMDA receptor antagonists—ifen- and/or retinal ischemia. prodil and eliprodil. In: Palfreyman MG, Reynolds IJ, Skolnick P, eds. Direct and Allosteric Control of Glutamate Receptors. Boca Acknowledgments Raton, FL: CRC Press; 1994:139-154. 20. Scatton B, Giroux C, Thenot JP, et al. Eliprodil hydrochloride. The authors thank Bernard Scatton of Synthelabo for insightful discus- Drugs Future. 1994;19:905-909. sion on the design of the studies and for reviewing the finaldraf t of this 21. Takahashi N, Cummins D, Caprioli J. Rat retinal ganglion cells in manuscript; Patrick K. Stanton and Robert J. Collier for their critical culture. Exp Eye Res. 1991;53:565-572. review and valuable comments on the manuscript; Joseph Caprioli and 22. Taschenberger H, Grantyn R. Several types of Ca2+ channels me- Yoshi Koseki for providing hands-on training on the isolation and diate synaptic responses to activation of single Thy- culturing of rat RGCs used in the viability studies. 1-immunolabeled rat retinal ganglion neurons. / Neurosci. 1995; 15:2240-2254. References 23. Nakanishi S, Masu M. Molecular diversity and functions of gluta- mate receptors. Annu Rev Biophys Biomol Struct. 1994;23:319- 1. Dreyer EB, Zurakowski D, Schumer RA, Podos SM, Lipton SA. 348. Elevated glutamate levels in the vitreous body of humans and 24. Biton B, Granger P, Carreau A, Depoortere H, Scatton B, Avenet P. monkeys with glaucoma. Arch Ophthalmol. 1996;ll4:299-3O5. The NMDA receptor antagonist eliprodil (SL 82.0715) blocks volt- 2. Lucas DR, NewhouseJP. The toxic effect of sodium L-glutamate on 2+ age-operated Ca channels in rat cultured cortical neurons. Eur the inner layers of the retina. Arch Ophthalmol. 1957;58:193-201. J Pharmacol. 1994;257:297-301. 3. Azuma N, Kawamura M, Kohsaka S. Morphological and immuno- histochemical studies on degenerative changes of the retina and 25. Legendre P, Westbrook GL. Ifenprodil blocks N-methyl-D-aspartate the optic nerve in neonatal rats injected with monosodium-L- receptors by a two-component system. Mol Pharmacol. 1991 ;40: glutamate. Ada Soc Ophthalmol Jpn. 1989;93:72-79- 289-298. 26. Leinders-Zufall T, Rand MN, Waxman SG, Kocsis JD. Differential 4. Karlsen RJL, Fonnum F. The toxic effect of sodium glutamate on rat 2+ retina: changes in putative transmitters and their corresponding role of two Ca -permeable non-NMDA glutamate channels in rat enzymes. / Neurochem. 1976;27:l437-l44l. retinal ganglion cells: kainate-induced cytoplasmic and nuclear 5. Olney JW. Glutamate-induced retinal degeneration in neonatal Ca2+ signals. / Neurophysiol. 1994;72:2503-25l6. mice: electron microscopy of the acutely evolving lesion. / Neu- 27. Zhang D, Sucher NJ, Lipton SA. Co-expression of AMPA/kainate ropathol Exp Neurol. 1969;28:455- 474. receptor-operated channels with high and low Ca2+ permeability 6. Siliprandi R, Canella R, Carmignoto G, et al. Ar-methyl-D-aspartate- in single rat retinal ganglion cells. Neuroscience. 1995;67:177-188. induced neurotoxicity in the adult rat retina. Vis Neurosci. 1992; 28. Kitano S, Morgan J, Caprioli J. Hypoxic and excitotoxic damage to 8:567-573. cultured rat retinal ganglion cells. Exp Eye Res. 1996;63:105-112. 7. Sisk DR, Kuwabara T. Histological changes in the inner retina of 29. Williams K. Ifenprodil discriminates subtypes of the N-methyl-D- albino rats following intravitreal injection of monosodium L-gluta- aspartate receptor: selectivity and mechanisms at recombinant mate. GraefesArch Gin Exp Ophthalmol. 1985;223:250-258. heteromeric receptors. Mol Pharmacol. 1993;44:851-859. 8. Vorwerk CK, Lipton SA, Zurakowski D, Hyman BT, Sabel BA, 30. Carter C, Benavides J, Legendre P, et al. Ifenprodil and SL-82.0715 Dreyer EB. Chronic low-dose glutamate is toxic to retinal ganglion as cerebral anti-ischemic agents, II: evidence for yV-methyl-i>aspar- cells: toxicity blocked by . Invest Ophthalmol Vis Sci. tate receptor antagonist properties. / Pharmacol Exp Ther. 1988; 1996;37:l6l8-l624. 247:1222-1232. 9. Blanks JC, Reif-Lehrer L, Casper D. Effects of monosodium gluta- 31. Gotti B, Duverger D, Bertin J, et al. Ifenprodil and SL 82.0175 as mate on the isolated retina of the chick embryo as a function of cerebral anti-ischemic agents, I: evidence for efficacy in models of age: a morphological study. Exp Eye Res. 1981;32:105-124. 10. Dreyer EB, Pan Z-H, Storm S, Lipton SA. Greater sensitivity of focal cerebral ischemia. / Pharmacol Exp Ther. 1988;247:1211- larger retinal ganglion cells to NMDA-mediated cell death. Neuro- 1221. report. 1994;5:629-631. 32. Gotti B, Benavides J, MacKenzie ET, Scatton B. The pharmacother- 11. Reif-Lehrer L, Bergenthal J, Hanninen L. Effects of monosodium apy of focal cortical ishaemia in the mouse. Brain Res. 1990;522: glutamate on chick embryo retina in culture. Invest Ophthalmol. 290-307. 1975;14:114-124. 33. Reyes M, Reyes A, Opitz T, Kapin MA, Stanton PK. Eliprodil, a 12. Sattayasai J, Ehrlich D. Morphology of quisqualate-induced neuro- non-competitive, NR2B-selective NMDA antagonist, protects pyra- toxicity in the chicken retina. Invest Ophthalmol Vis Sci. 1987; midal neurons in hippocampal slices from hypoxic/ischemic dam- 28:106-117. age. Brain Res. 1998;782:212-218. 13- Sucher NJ, Aizenman E, Lipton SA. N-methyl-D-aspartate antago- 34. Kapin MA, Doshi R, Scatton B, DeSantis LM, Chandler ML. Neuro- nists prevent kainate neurotoxicity in rat retinal ganglion cells in protective effects of eliprodil in retinal excitotoxicity and isch- vitro. / Neurosci. 1991; 11:966 -971. emia. Invest Ophthalmol Vis Sci. 1999;40:l 177-1182.

Downloaded from jov.arvojournals.org on 09/27/2021