Do IV-Methybaspartate Receptors Mediate Synaptic Responses in the Mudpuppy Retina?

The Journal of Neuroscience, December 1988, 8(12): 4728-4733

Do IV-MethybAspartate Receptors Mediate Synaptic Responses in the Mudpuppy Retina?

Paul A. Coleman and Robert F. Miller Department of Ophthalmology, Washington University School of Medicine, St. Louis, Missouri 63110

Whole-cell recordings of amacrine and ganglion cells in the of the N-methyl-D-aspartate (NMDA) receptor (Massey and superfused retina-eyecup preparation of the mudpuppy were Miller, 1985). The actions of these drugs appear sufficiently obtained in order to determine which excitatory amino acid restricted such that little pharmacologicaloverlap between the receptor (EAAR) subtype mediates the synaptic responses two has been observed. of these neurons. All third-order retinal neurons tested were r%-2-Amino-5-phosphonovalerate (APS) also hasnetwork se- depolarized by kainic acid (KA), KfIIethyl-D-SSpSrtSte lective effects. Like AP4, DL-AP5 also attenuates the photic (NMDA), and quisqualate (QQ). The responses evoked by responseof retinal neuronsassociated with the On system(Ikeda NMDA were blocked by the addition of D-2-amino-5-phos- and Sheardown, 1983; Bolz et al., 1984; Slaughter and Miller, phonovaleric acid (D-APB) and o-2-amino-7 phosphonohep- 1985). Becausethe racemic mixture of AP5 is also a potent tonoic acid (D-AP7) to the perfusate. When the actions of NMDA antagonist,it wassuggested that sustainedOn responses exogenously applied NMDA were completely blocked by of third-order neuronswere mediatedby NMDA receptors(Lu- D-AP5 and D-AP7, the light-evoked responses of inner retinal kasiewicz and McReynolds, 1985). This conclusion conflicted neurons persisted without any apparent reduction or, alter- with an earlier report by Slaughterand Miller (1983), who used natively, a slight enhancement of the response was ob- o-a-amino adipate and p-a-amino suberate to block NMDA served. Light-evoked responses of bipolar, amacrine, and receptors; these NMDA antagonists did not affect the light- ganglion cells associated with the On pathway were atten- evoked responsesof third-order neurons,except for a particular uated by L-AP5 in a manner similar to its lower-order homo- subclassof Off ganglion cells. To better understand the role of log L-2-amino-4-phosphonobutyrate (AP4); nevertheless, NMDA receptors in light-activated synaptic transmission,we L-APB was not an effective NMDA antagonist. Although syn- have usedthe separateenantiomers of AP5 and AP7 on second- aptic transmission between retinal second- and third-order and third-order neurons.We report that the D-form of AP5 and neurons appears to be mediated by EAARs, the NMDA re- AP7 blocked NMDA receptors but did not block synaptic receptor does not appear to play a prominent role under our sponsesassociated with the On system. However, L-AP5 was experimental conditions. Nevertheless, our results suggest a poor NMDA antagonistand had effects similar to the L-AP4 that the racemic mixture of AP5 should not be used as an suppressionof the On bipolar cell describedby Slaughter and NMDA antagonist in retinal research, due to the APClike Miller (1981). Thus, our findings indicate that NMDA receptors actions of its L-enantiomer. do not mediate the major componentsof the synaptic responses of third-order retinal neurons. The initial results of these ex- The “glutamate” hypothesis of excitatory neurotransmission in perimentshave appearedin abstractform (Colemanand Miller, the retina has become widely accepted. It seems likely that pho- 1987). toreceptors and bipolar cells release glutamate, aspartate, or a closely related analog that activate second- and third-order neu- Materials and Methods rons. Advancements in this field have come about by the in- The superfused retina-eyecup of the mudpuppy, Necturus maculosus troduction of compounds having selective actions on subpop- (Miller and Dacheux, 1976), was used for the present experiments. After ulations of retinal neurons (Miller and Slaughter, 1986). An the animals were decapitated and pithed, the head was hemisected and placed on a Ringer’s soaked cotton ball. The anterior chamber of the interesting group of compounds, the a-amino-w-phosphono- eye was opened permitting the iris, lens, and the vitreous humor to be carboxcylic acids, differentially effects retinal excitatory amino removed. This preparation was then placed into a light-tight Faraday acid receptors (EAARs) depending on the stereospecificity and cage. the number of carbon atoms in the molecule. For example, L( +)- The gravity-feed perfusion system consisted of eleven syringes (10 ml) connected, with plastic valves, to a common manifold. The perfusate 2-amino-4-phosphonobutyrate (AP4) acts as an agonist for the exited the manifold by 24-gauge Teflon tubing and flowed into the EAAR of the depolarizing bipolar cell (Slaughter and Miller, eyecup through a pipette placed at the edge of the retina. The perfusate 198 l), whereas the n(-)-enantiomer of a higher-order homolog consisted of a modified Ringer’s solution containing 111 mM NaCl, 2.5 2-amino-7-heptonoate (AP7) is a potent and specific antagonist mM KCl, 1.8 mM CaCl,, 1 mM MgCl,, 5 mM dextrose, 10 mM HEPES, and titrated to pH 7.80 with NaOH. All pharmacological agents were added to the control perfusate and the pH adjusted, if necessary. The Received Jan. 7, 1988; revised Mar. 29, 1988; accepted Apr. 4, 1988. perfusate was bubbled with 100% oxygen and flowed at a rate of l-2 ml/min. This system had an exchange time between different solutions SuDDOrted bv NE1 Grant EY 03014. C&respondence should he addressed to Paul A. Coleman, Department of Phys- of approximately 20-30 set, due to dead space in the lines. iology, University of Minnesota, 6-255 Millard Hall, 435 Delaware Street S.E., Physiological data were obtained from individual neurons (n = 40) Minneapolis, MN 55455. using either conventional intracellular recordings or the whole-cell re- Copyright 0 1988 Society for Neuroscience 0270-6474/88/124728-06$02.00/0 cording (WCR) technique we recently developed for use in the intact The Journal of Neuroscience, December 1988, 8(12) 4729

~OJJM 00 -~OJJM KA ~OJJM NMDA _

Figure 1. WCR from an On sustained 2SEC DIFFUSE LIGHT ganglion cell in the mudpuppy retina. 125uM D-AP7 Application of each of the EAAR agonists is indicated by the vertical solid 50uM KA ~OOJJM NMDA line above the voltage record, whereas the antagonist is indicated by the dashed line. This cell was depolarized by KA (50 PM), NMDA (200 PM), and QQ (10 PM). D-AP7 (125 PM) only blocked the effects of NMDA. However, the light- 20mVI evoked response persisted in the pres- 20SEC ence of the NMDA antagonist. retina (P. A. Coleman and R. F. Miller, unpublished observations). The WCR electrodes were made with thin-wall borosilicate, Omega-dot glass tubing (O.D. = 1.2 mm, I.D. = 0.86 mm; Glass Company of America) Results and had tips that were l-2 pm in diameter and were 8-12 Mfl in resistance when filled with the intracellular electrolyte solution. This Figure 1 shows that bath application of the three EAAR agonists solution consisted of KCH,SO,, 90 mM; NaCH,SO,, 5 rnhr; MgSO,, 10 KA (50 PM), NMDA (200 PM), and QQ (10 PM) to an On ganglion mM; CaCl,, 1 mM; EGTA, 11 mM; HEPES, 5 rnhi; glucose, 2 mM; cell resultedin a depolarizing response,accompanied by action glutathione, 1 mM; ATP-Mg*+, 1 mM; phenol red, 0.0004%. The solution potentials. Thesedrugs activate EAARs intrinsic to third-order was titrated to a pH of 7.2 with KOH, which brought the final K+ concentration to approximately 125 mM. neurons (Slaughter and Miller, 1983; Lukasiewicz and Mc- Conventional intracellular electrodes were made using medium-walled Reynolds, 1985; Coleman et al., 1986). The perfusate bathing Omega-dot glass tubing (O.D. = 1.2 mm, I.D. = 0.6 mm; Glass Company the retina was switched to one containing 125 PM of the D-en- of America) and fabricated on an ultrafine micropipette puller (Camp- antiomer of AP7 and the agonistsreapplied. Both KA and QQ den Instruments). When filled with 1.5 M KC2H,02, the electrodere- sistance was 150-200 Mfl. The intracellular potentials were monitored depolarized the cell but NMDA was ineffective. This indicated with a Dagan 8 1OO- 1 amplifier. Conventional intracellular records were that D-AP7 was a selective NMDA antagonist for mudpuppy obtained in order to confirm that no artifacts were introduced by the inner retinal neurons. WCR technique. Identical pharmacological results were acquired using Although D-AP7 successfullyblocked the depolarizing effects either technique. ERG electrodes were made as conventional intracel- of bath applied NMDA, Figure 1 also shows that the light- lular electrodes, with the tips broken back and beveled until they were approximately 5 MO when filled with Ringer’s solution. The D.C. re- evoked responseof the neuron persisted.For other cells DL- cordings of the ERG were monitored using a Grass P- 16 amplifier. AP7 was used;which also was a potent NMDA antagonistbut A dual-beam light bench projected focused images of a spot (lOO- also did not block the light-evoked responseof any neuron 500 pm), a slit (130 pm x 10 mm) or diffuse light onto the retina tested. The results with AP7 contrasted sharply with the obser- (irradiance = l-2 mW/m*). The light pulses were 2 set in duration and controlled with electromechanical shutters (Uniblitz, Vincent Associ- vations obtained with the NMDA antagonist DL-APS. Other ates). Intensity was controlled using two counterbalanced neutral density studieshave shownthat the light-evoked responsesof ganglion wedges (Kodak). cells (especially On cells) were attenuated by DL-AP5 (Ikeda

500uM NMDA

PSEC ..a 40 DIFFUSE mV I 20 500j~M NMDA set 250j~M DAP5

Figure 2. WCR from anOn sustained ganglion cell which was depolarized by NMDA (500 UM). ADDlication of D-AP5 (250 PM) completely blocked the effects of NMDA, nevertheless, no effect on 500gM NMDA the light-evoked response was ob- 250)~M LAP5 served. In contrast, L-AP5 (250 PM) reduced the light-evoked response of this cell but could not block the depolarization and increase in spike activity caused !!!!!l!!!l-!-II& by NMDA. 4730 Coleman and Miller - Blockage of Nonsynaptic NMDA Receptors

;yM 250,uM DHS - - -

2 SEC LIGHT 30mV

BOSEC 30,uM ~OOJJM 250~~M NMDA DHS 250)~M DAP7 fi ------______------

Figure 3. WCR from an Off ganglion cell in which the effects of NMDA (500 PM) and DHS (250 PM) are contrasted. Althouah the deoolarization evoked bv KA (36~~) was unaffected by D-AP7 (250 PM), the responses evoked by 30mV NMDA and DHS were essentially elim- 1 inated. BOSEC and Sheardown, 1983; Bolz et al., 1984; Lukasiewicz and the depolarizing effectsof NMDA. As happenedwith the higher- McReynolds, 1985; Coleman and Miller, 1987). In order to order homolog D-AP7, D-AP5 proved to be a potent NMDA discern the basis for this apparent discrepancy, the separate antagonist but left the light-evoked responseessentially un- enantiomers of AP 5 were tested. The sustainedOn ganglion changed. cell illustrated in Figure 2 wasdepolarized by NMDA (500 FM). The bottom trace of Figure 2 showsresults with the L(+)- The samecell was exposedto D-APS (250 PM), which blocked enantiomer of AP5. The light-evoked responsefrom this On ganglioncell wasseverely attenuated when L-AP5 (250 PM) was added to the perfusate. This reduction in the light-evoked re-

250~~M 250uM DAPS sponsewas not accompanied by the blockade of the NMDA LAP5 receptorssince NMDA still evoked a significant depolarization. Though not illustrated here, the effects of L-AP5 were completely reversible. Theseexperiments show that the light-evoked responseof On ganglion cells did not significantly dependupon NMDA receptors. To further characterize the ganglion cell NMDA receptors, another EAAR agonist, D-homocysteinesulphinic acid (DHS), was tested. In other regionsof the brain, DHS appearsto be a selective NMDA receptor agonist(Mewett et al., 1983). Figure 3 showsthe depolarizing effects of KA (30 MM), NMDA (500 250pM DAP7 : XM PM), and DHS (250 PM) on an Off ganglion cell. As seenin the lower trace, when theseagents were tested againstD-AP7 (250 PM), the effectsof DHS were largely blocked in a manner similar to NMDA itself. Nevertheless,careful examination of the bottom trace showsthat DHS slightly decreasedthe amplitude of the light-evoked responsein the presenceof D-AP7. This in- ..“.,?‘I -.,.....,,...,...... ~.~~.--,r,,~~,~s-s-,~~-.,~~-,-.-~--.-.~~----~----- dicates that DHS may have someweak agonist effects on non- . . . ..l...L...... NMDA receptors. As a comparison, we challengedthe effects of another closely related EAAR agonist, D-homocysteic acid ~SOJJM LAP7 15 mV (DHC), againstthe NMDA antagonists.The depolarizing effects I of DHC were unaffected by the NMDA antagonists.In accord with previous results(Mewett et al., 1983),DHS wasan effective NMDA agonist, whereasthe related analog, DHC, activated a broader spectrum of EAARs. Figure 3 also showsthe light-evoked responsesof Offganglion cells, in our sample,were not reduced by NMDA antagonists. In our experience there has been little observed effect on the Figure 4. Conventional intracellular recordings from On bipolar light-evoked PSPsof the mudpuppy inner retinal neuronsduring cells in which L-AP5 (250 PM) was applied. Like its lower-order homolog bath application of either D-AP5 or D-AP7; however, under L-AP4 (Slaughter and Miller, 1981), L-AP5 reduced the amplitude of different stimulus conditions the action of NMDA receptors the photic response of this class of second-order neurons. When the perfusate was changed to one containing an NMDA antagonist (either may be more apparent. Nonetheless,since the On responsesof D-AP5 or D-AP7), the amplitude of the light-evoked response returned inner retinal neuronswere attenuated by L-APS, a poor NMDA to control levels. antagonist, we concluded that its effects were mediated by a The Journal of Neuroscience, December 1988, 8(12) 4731

250blM !l AP4 ', Ji \ ! <,P- - .

Figure 5. Recordings of the electroretinogram (ERG) during applications of the separate enantiomers of AP4, AP5, and API. The b-wave in these 250LlM R* records appeared as a large downward J! API 0" deflection due to the intraretinal place- ment of the ERG electrode below the b-wave isopotential point. The b-wave was most effected by L-AP4 and L-AP5. \ 500 m.rc In contrast, the NMDA antagonists D-APS and D-AP7 had no obvious ef- I \ fect on any of the components of the c ERG. non-NMDA receptor. It is unlikely that L-APS blocks the light- plied, the amplitude of the On bipolar light-evoked response evoked responsesof ganglion and amacrine cells by acting as returned to levels observedin the control perfusate.Application an antagonist to KA or QQ receptors, since concentrations of of L-AP7 had essentiallyno effect on the light-evoked response DL-AP5 sufficient to reduce light-evoked activity did not block of On bipolars (bottom trace) or any other inner retinal neuron the depolarizing effects of KA and QQ (Lukasiewicz and tested. McReynolds, 1985). The neuronal basis of the b-wave of the electroretinogram Previous studieshave shown that DL-APS is an agonist for (ERG) is the On bipolar cell (Dick and Miller, 1985).Therefore, the On bipolar EAAR (Slaughterand Miller, 1985).This unique drugs that reduce the light-evoked responseof the On bipolar receptor is often referred to as the “APB receptor” becauseof cell should also reduce the amplitude of the b-wave. In this the high affinity with which 2-amino-4-phosphonobutyrateserves sense,the ERG is a useful tool becauseit allows a comparison as an agonist for the activation of the receptor. It would not be to be made between many drugs with little concern about the surprising that the L-enantiomer of the one carbon chain longer stability of the recording. Figure 5 showsthe effects of the D- homolog, AP5, had a similar mode of action. Therefore, since and L-forms of AP4, APS, and AP7 (250 PM) on the b-wave of the D-enantiomer did not reduce the light-evoked responseof the ERG. The b-wave, seenas the large downward transient ganglioncells and the L-enantiomer did, we tested the possibility response,was completely eliminated by application of L-AP4. that the effects of L-AP5 were localized to the On bipolar cell. This effect was stereo specificin that the b-wave was only par- Intracellular recordingswere maintainedfrom On bipolar cells tially attenuated with an equal concentration of the D-enan- during L-AP5 perfusion. Figure 4 showsthat L-AP5 (250 PM) tiomer. Application of L-AP5 also causeda large reduction in reduced the amplitude of the light-evoked responseof the On the b-wave, leaving only a smallremnant of the response,where- bipolar cells. To further demonstratethat NMDA receptorswere as L-AP7 had little if any effect on b-wave amplitude. Both the not involved in the On bipolar response,the perfusate was D-enantiomersof AP5 and AP7 had virtually no effect on the switchedto one containing either D-AP5 (upper trace) or D-AP7 b-wave or any other component of the ERG. Since the ERG (middle trace). Regardlessof which NMDA antagonistwas ap- largely reflects the activity of the outer retina, this result indi- 4732 Coleman and Miller * Blockage of Nonsynaptic NMDA Receptors

ON BIPOLAR in the input resistanceof this cell with the NMDA antagonist. But consideringthe small relative changein R, (< 3%) and that only half of the cells tested in this manner (n = 4) showedan increasedR, with the application of the NMDA antagonists,it seemslikely that additional mechanismsmay be involved in the enhancement phenomena. A similar observation was recently reported (Forsythe and Westbrook, 1988) that non- NMDA, EAAR-mediated synaptic EPSPsbetween cultured hip- pocampal neurons were also enhanced after the blockage of LIGHT NMDA receptors. I CICR 5mV L Discussion 500mSEC Virtually all third-order neurons of the mudpuppy retina re- OFF GANGLION CELL spond to KA, NMDA, and QQ. Evidence presentedelsewhere suggeststhat theseagonists activate separatereceptors (Coleman et al., 1986). Determining the functional role of the individual EAARs is almost impossible without the use of selective antagonists since retinal amacrine and ganglion cells are all depolarized by eachof theseEAAR agonists.Certainly the NMDA receptor antagonistsare the most selective, and in other regions of the CNS, NMDA antagonistshave beenuseful in establishing the functional role of this classof EAAR (Salt, 1986). LIGHT In this paper, we have demonstrated that only the n-enan- I tiomer of AP5 worked as an effective NMDA antagonist,while WCR the L-form exhibited very little antagonismto exogenouslyap- 1OmV I plied NMDA. Similar results werealso obtained with AP7. Yet, 500mSEC when we examined the suppressiveeffects of AP5 on the light- evoked responsesof ganglion cells, we observed the reverse Figure 6. Superimposedrecords from an On bipolarcell and an Off sensitivity. Relatively high levels of D-APS, which completely ganglioncell beforeand after applicationof D-APS and D-AP7. As shownhere, if theseNMDA antagonistshad an effecton synaptically eliminated responsesto exogenousNMDA, did not block the evokedresponses, a slight enhancement was observed. light responseof On cells, whereasthe L-enantiomer did. Thus, we conclude that the DL-APS-mediated reduction in the On responsesofganglion cellspreviously reported (Ikeda and Shear- cated that NMDA receptors exerted no measurableinfluence down, 1983; Bolz et al., 1984; Lukasiewicz and McReynolds, upon signal transmission between photoreceptors and second- 1985) was not due to a block of NMDA receptors but to the order neurons. These results, using the ERG, mirrored our in- non-NMDA action of the L-enantiomer of the racemic mixture tracellular On bipolar cell data using L-APS and the NMDA used. antagonists.In addition, there is little evidence suggestingthe Analysis of On bipolars showsthat L-AP5 blocked On bipolar presenceof NMDA receptors among the neurons of the outer cell light responsesand can therefore account for the suppression retina; application of NMDA, even in high concentrations, had of On responsesof ganglion cells when the racemic mixture of little influence on photoreceptors, horizontal, and bipolar cells AP5 was employed. The effects of L-AP5 are similar to those (Slaughter and Miller, 1983; Massey and Miller, 1987). of L-AP4, which has proven to be a highly selective agonistfor Although the NMDA receptor antagonistsdo not reduce the the On bipolar synaptic receptors (Slaughterand Miller, 1981). light-evoked responseof retinal neurons, on occasionthey have It is interesting to note that L-AP7, two carbons longer than beenobserved to slightly increasethe light-evoked PSPs.Figure AP5, had essentiallyno effect as an agonistfor the synaptic On 6 (top) shows 2 superimposedresponses from the On bipolar bipolar EAAR. Thus, the racemicmixture of AP7 could be used illustrated in Figure 4. If D-APS wasobserved to have any effect, as an NMDA antagonist,whereas the racemic mixture of AP5 it slightly increasedthe PSPsof theseneurons. The sameeffect is contraindicated asan NMDA antagonistbecause of the “AP4- was observed for D-AP7 on other bipolar cells. A similar en- like” actions of the L-enantiomer. hancement wasalso seenin some third-order neurons, as illus- It is quite perplexing that no major portion of the light-evoked trated with the Off ganglion cell illustrated at the bottom of responseswas blocked with the NMDA antagonistssimply be- Figure 6. In the presenceof D-AP7, the light-evoked response causeof the ubiquitous nature of thesereceptors among inner from this Off ganglion cell was slightly enhanced.A slight en- retinal neurons. Similar results have been found in this labo- hancementwas also observed on occasionin someOn ganglion ratory on the rabbit retina, in that there was no alteration in cells also (not shown). the receptive field properties of the ganglion cells tested (S. C. One possible reasonfor larger PSPs in some neurons could Massey and R. F. Miller, personalcommunication). However, have been that NMDA antagonistsincreased the input resis- in contrast to the presentresults, a small reduction in the number tance (R,) of thesecells. Figure 7 showsthe Z-V relationship of of light-evoked action potentials wasobserved in many ganglion the Off ganglion cell presentedin the previous figure, evaluated cells using AP7 as an antagonist. The significanceof this dis- in the absenceor presenceof D-AP7. The R, wasfit to the data crepancy is not clear, but it does not undermine the major using a least-squaresapproximation of the linear regression. agreementbetween these and earlier studies(Slaughter and Mil- When measuredin the dark, there wasa slight increase(27 MB) ler, 1983) that NMDA receptors do not subservethe major The Journal of Neuroscience, December 1988, 8(12) 4733

-I CONTROL

+-+ CONTROL * *-* o-AP7 (250 p&o ” -80.0 /

-160.

1 1 . 1 O-AP7 (250 pM> -. 000

+-+ Rn-961 MOHW t -136. +--Y Rn=989 HOHH -80.0

-160. -160. 1 25.0 75.0 IiS. . 175. 225. maec

Figure 7. Z-V relationship measured in the presence or absence of D-AP7 (250 PM) in the Off ganglion cell of the previous figure. The dark R. increased from 96 1 to 988 MB in the presence of the NMDA antagonist. Also shown are the voltage responses to the 8 current pulses (255 msec), which ranged from - 100 to +40 pA. Except for the fact that no spikes were present when a +20 pA current pulse was injected during the application of D-AP7, no obvious distinction between the 2 sets of curves was evident. No change in the dark resting potential was observed (- 58 mV) in the presence of D-AP7. portion of light-evoked responsesin the retina. Therefore, it Massey, S C., and R. F. Miller (1987) Excitatory amino acid receptor must be left to future studies to determine the functional im- of rod- and cone-driven horizontal cells in the rabbit retina. J. Neu- rophysiol. 57: 645-659. portance of NMDA receptors in the vertebrate retina. Mewett, K. N., D. J. Oakes, H. J. Olverman, D. A. S. Smith, and J. C. Watkins (1983) Pharmacology of the excitatory actions of sulphonic References and sulphinic amino acids. In CNS Receptors-From Molecular Phar- Bolz, J., H. Wlssle, and P. Thier (1984) Pharmacological modulation macolonv to Behavior, P. Mandel and F. V. DeFeudis, eds., Raven, of ON and OFF ganglion cells in the cat retina. Neuroscience 12: New Y&k. 875-885. Miller, R. F., and R. F. Dacheux (1976) Synaptic organization and Coleman, P. A., and R. F. Miller (1987) Do NMDA receptors mediate ionic basis of ON and OFF channels in the InUdDUDDY retina. I. light responses in the retina? Invest. Vis. Sci. Suppl. 28: 403. Intracellular analysis of chloride-sensitive electrogenic properties of Coleman, P. A., S. C. Massey, and R. F. Miller (1986) Kynurenic acid receptor, horizontal cells, bipolar cells and amacrine cells. J. Gen. distinguishes kainate and quisqualate receptors in the vertebrate ret- Physiol. 67: 639-659. ina. Brain Res. 381: 172-175. Miller, R. F., and M. M. Slaughter (1986) Excitatory amino acid Dick, E., and R. F. Miller (1985) Extracellular K+ activity changes receptors of the retina: Diversity of subtypes and conductance mech- related to electroretinogram components. J. Gen. Physiol. 85: 885- anisms. Trends Neurosci. 9: 2 1 l-2 18. 909. Salt, T. E. (1986) Mediation of thalamic sensory input by both NMDA Forsythe, I. D., and G. L. Westbrook (1988) Slow excitatory postsyn- receptors and non-NMDA receptors. Nature 322: 263-265. aptic currents mediated by N-methyl-D-aspartate receptors on cul- Slaughter, M. M., and R. F. Miller (1981) 2-Amino-4-phosphono- tured mouse central neurons. J. Physiol. (Lond.) 396: 5 15-533. butyric acid: A new pharmacological tool for retinal research. Science Ikeda, H., and M. J. Sheardown (1983) Functional transmitters at 211: 182-185. retinal ganglion cells in the cat. Vision Res. 23: 116 l-l 174. Slaughter, M. M., and R. F. Miller (1983) The role ofexcitatory amino Lukasiewicz, P. D., and J. S. McReynolds (1985) Synaptic transmis- acid transmitters in the mudpuppy retina: An analysis with kainic sion at N-methyl-D-aspartate receptors in the proximal retina of the acid and N-methyl aspartate. J. Neurosci. 3: 170 l-l 7 11. mudpuppy. J. Physiol. (Lond.) 367: 99-l 15. Slaughter, M. M., and R. F. Miller (1985) Characterization of an Massey, S. C., and R. F. Miller (1985) AP-7 is a potent and specific extended glutamate receptor of the bipolar neuron in the vertebrate NMDA antagonist in the rabbit retina. Invest. Vis. Sci. Suppl. 26: retina. J. Neurosci. 5: 224-233. 117.