Comparison of the actions of glycine and related amino acids on isolated third order neurons from the tiger salamander retina

~

Pergamon

0306-4522(94)00399-8

Neuroscience Vol. 64, No. 1, pp. 153-164, 1995 Elsevier Science Ltd Copyright © 1994 IBRO Printed in Great Britain. All rights reserved 0306-4522/95 $9.50 + 0.00

C O M P A R I S O N OF THE A C T I O N S OF G L Y C I N E A N D RELATED AMINO ACIDS ON ISOLATED THIRD ORDER N E U R O N S F R O M THE T I G E R S A L A M A N D E R R E T I N A Z.-H. P A N * a n d M. M. S L A U G H T E R ' t Departments of Biophysical Sciences and Ophthalmology, School of Medicine, State University of New York at Buffalo, Buffalo, NY 14214, U.S.A. Al~traet--Whole cell voltage and current clamp recordings were obtained from third order neurons isolated from the salamander retina. Using cross desensitization, the structure-function relationship of short chain amino acids on the glycine receptor were examined. L-Serine, L-alanine, fl-alanine and taurine all cross desensitized with glycine, but did not show significant cross desensitization with GABA. This indicates that these amino acids act at the glycine receptor. The order of potency was glycine>>fl-alanine> taurine>>L-alanine> L-serine. TAG, a reputed selective taurine antagonist, was equally effective in blocking taurine and glycine currents. There is no evidence for distinct receptors for taurine. Amino acids with larger moieties at the alpha carbon, such as threonine and valine, produced inactive ligands. Placing a methyl group on the amine of glycine or esterification of the carboxyl group also greatly reduced activity. Based on these modifications of the glycine molecule, it appears that selectivity at the glycine receptor results in part from steric restrictions at all three sites in the glycine chain. The steric interference is most critical at the carboxyl and amino ends, and less limiting at the alpha carbon. Doses of L-serine that had only slight effects in voltage clamp experiments, nevertheless produced large effects in current clamp experiments. This indicates that several endogenous amino acids can have significant effects on membrane voltage, even when their shunting activity may be small. High concentrations of agonists produced desensitization in the voltage clamp records, but there was little evidence of desensitization in the current clamp experiments. These results indicate that several endogenous amino acids can activate the glycine receptor, but there is no evidence for a discrete receptor for taurine, fl-alanine, L-alanine or L-serine. Since all these endogenous amino acids have similar amino and acid terminals, reduction in potency results from steric interference around the alpha carbon. This graded potency may have functional significance in mediating inhibition.

G A B A a n d glycine are the principal inhibitory neurot r a n s m i t t e r s in the retina 26'4° as they are t h r o u g h o u t the CNS. B o t h t r a n s m i t t e r s activate receptors gating chloride c o n d u c t a n c e s with similar permeability a n d kinetic characteristics. 9m The similarities are so m a r k e d t h a t it has been suggested t h a t the two receptors m a y share a c o m m o n channel. 4'24A n u m b e r o f o t h e r e n d o g e n o u s a m i n o acids also activate these receptors, o p e n i n g the possibility t h a t several a m i n o acids act as i n h i b i t o r y transmitters. These o t h e r a m i n o acids include L-alanine, fl-alanine, L-serine a n d taurine. T h e literature is a m b i v a l e n t o n the relationship

*Present address: Department of Neurology, Harvard Medical School and Children's Hospital, 300 Longwood Avenue, Boston, MA 02115, U.S.A. tTo whom correspondence should be addressed at: Department of Biophysical Sciences, 120 Cary Hall, State University of New York, Buffalo, NY 14214, U.S.A. Abbreviations: EGTA, et hyleneglycol-bis-(flaminoethyl ether)-N,N'-tetraacetic acid; HEPES, N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid); TAG, 6aminomethyl- 3-methyl-4H- 1,2,4-benzothiadiazine- 1, I dioxide.

between these o t h e r e n d o g e n o u s a m i n o acids a n d the G A B A a n d glycine receptors. F o r example, taurine, serine a n d alanine a p p e a r to act o n glycine receptors in the spinal cord, while taurine is reported to act o n G A B A receptors in the cerebellar cortex. 7'15A7 In the cerebral cortex, b o t h types o f taurine mediated responses were found. TM M o s t of these conclusions were based o n studies o f susceptibility to glycine a n d G A B A a n t a g o n i s t s such as strychnine, bicuculline a n d picrotoxin. The disparity m a y be due to molecular variants in the G A B A a n d glycine receptors f o u n d in different tissues. The glycine receptor is a p e n t a m e r of a l p h a a n d beta subunits. F o u r different a l p h a subunits (ctl, ct2, ~ 2 " a n d ~3) have been identified, each with a different p h a r m a c o l o g y . In heterologous expression o f rat R N A in Xenopus oocytes, glycine receptors with ct 1 subunits are sensitive to fl-alanine a n d taurine, b u t glycine receptors with ct2 or ct3 subunits are sensitive to fl-alanine but n o t taurine. 6'25 The aim of this p a p e r was to examine the receptor specificity in the retina. Taurine, L-serine a n d L-alanine have been s h o w n to be active in the retina sA3'33 where b o t h L-serine a n d taurine are f o u n d at high concentrations. ~°'33 Based 153

154

Z.-H. Pan and M. M. Slaughter

on strychnine blocking experiments, it has been concluded that all three amino acids act at the glycine r e c e p t o r ) '~3'33 One potential limitation o f this analysis is that strychnine's specificity is questionable. F o r example, there are n u m e r o u s reports indicating that strychnine can block G A B A receptors, 22~3s acetylcholine receptors, 16'~s or that it can activate a potassium conductance. 1 Glycine subunit variation can affect antagonist affinity as well as agonist affinity. F o r example, glycine receptors containing the rat c~2" subunit have a much lower affinity for strychnine than those containing the e 2 isoform. 6 Therefore, strychnine block may not offer unequivocal identification o f glycine receptor activity. It is also possible that these other a m i n o acids activate strychnine sensitive receptors which are distinct from the glycine receptor. Studies o f receptor desensitization offer an alternative m e t h o d to evaluate receptor specificity while avoiding possible complications arising from nonspecific antagonist actions. Desensitization, a decrease in c o n d u c t a n c e during prolonged ligand application, is a p r o m i n e n t feature o f inhibitory amino acids in the retina. 19'2°'35A l t h o u g h the mechanisms o f G A B A and glycine desensitization are still u n k n o w n , m a n y studies have d e m o n s t r a t e d that desensitization results from inactivation o f the receptors rather than a change in driving force across the m e m b r a n e (fading). This p h e n o m e n o n can be used to evaluate receptor specificity because when one ligand desensitizes a receptor, the response to a second ligand using the same receptor will be reduced. Results from the present receptor desensitization experiments indicate that: (1) I.-serine, taurine, calanine a n d / ~ - a l a n i n e all activate the retinal glycine, but not the G A B A , receptor; (2) there is no evidence for discrete taurine receptors in the retina; (3) nonlinearities between receptor current and m e m b r a n e voltage result in large voltage changes p r o d u c e d by weak agonists; and (4) steric hindrance a r o u n d the amine and b o t h c a r b o n groups o f glycine all contribute to the selectivity o f the glycine receptor.

EXPERIMENTAL PROCEDURES

Preparation Experiments were performed on single, enzymatically dissociated retinal cells from tiger salamander, Ambystoma tigrinum, obtained from Kons Scientific (Germantown, WI). The dissociation procedure was similar to previously reported methods.3'23 Briefly, the animal was decapitated, pithed, and the eye was removed. The retina was isolated and incubated for about 30~0 rain at room temperature (22°C) in 400/~1 of enzyme solution containing 12U/ml papain (Type IV, Sigma) and 5 mM L-cysteine in amphibian Ringer's solution (111 mM NaC1, 3 mM KC1, 2 mM CaC12, 1 mM MgCI2, 10 mM dextrose, buffered with 5 mM HEPES at pH 7.4). At the end of the incubation period the retina was transferred to a plastic tube and rinsed three times with 2 ml of amphibian Ringer's solution without the inclusion of calcium. Then the retina was vortexed in 2ml of this nominally calcium-free Ringer's solution for about 5 rain. A 200/~1 aliquot of the final cell suspension was dispensed into

a culture dish containing 2ml of oxygenated Ringer's solution. This preparation would be used within 6 h. Alternatively, the aliquot of cells was added to a mixture of L-15(90%) and L-99(10%) media supplemented with antibioti~antimycotic and penicillin streptomycin (obtained from GIBCO). These cultures were stored at 10°C in the presence of oxygen and were studied for periods up to one week. With relationship to the topics in this report, there was no significant difference in the acute and cultured preparations. However, most of the data were obtained from the acute dissociations. For electrophysiological recordings, the culture dish was mounted on the stage of an inverted microscope equipped with phase-contrast optics. Photoreceptors and bipolar cells could be readily identified based on their distinctive morphologies. Similarly, horizontal cells could usually be distinguished from third order neurons. Recordings for this report were from third order neurons, which were identified morphologically and by the presence of large inward sodium currents seen under a voltage clamp. Results are based on recordings from 205 third order cells.

Drug application A microperfusion system was developed for drug application. The perfusate consisted of a control solution of Ringer's or a Ringer's solution containing pharmacological agents. The different perfusion solutions were contained in a series of bottles which were gravity fed to a modified electrode holder through separate Teflon-lined valves and polyethylene tubing. A capillary tube was connected from this holder to one end of a three-way tubing adapter. One output of the adapter went to a short glass pipette. This pipette had a tip diameter of about 100 ~m and its output perfused individual cells. The other output of the adaptor served as drain for the perfusion solution. Most of the solution passed through this drain tube and only a small portion went through the perfusion pipette toward the cell. The amount of perfusate through the pipette could be regulated by adjusting the height of the drain tube. This arrangement allowed exchange of the solution around the cell in about 1 s. Unless a drug was being tested, control Ringer's solution was continuously running through the perfusion system and over the cell. TAG was a generous gift from Dr Huxtable, azetidine carboxylic acid and 3,4 dehydro-uL-proline were obtained from Aldrich Chemical Co., all other chemicals were obtained from Sigma Chemical Co.

Recording apparatus Patch electrodes were pulled from omega dot borosilicate glass tubing (0.6 mm i.d., 1.2 mm o.d.) with a vertical pipette puller (Narishige PP-83) and fire polished. The pipette resistances were about 5 M~ when filled with a standard internal solution. This internal solution contained 100 mM KCl/K-gluconate (a spectrum of chloride concentrations were used ranging from 100mM KC1/0mM K-gluconate to 0 mM KC1/100 mM K-gluconate), 5 mM NaCI, 2 mM MgCI2, 5 mM EGTA, 5 mM HEPES, pH 7.4. In the figures of this paper, the cells were always clamped at - 70 mV and the chloride reversal potential was always more positive than the holding potential. Cell attached seal resistances were more than 5 Gf~. Calculated series resistances were generally about 15 MfL Whole cell membrane currents and voltages were recorded using a patch clamp amplifier (Model l-C, Axon Instruments) and the amplified signals were recorded on a Tektronix digital storage oscilloscope and a Vetter video cassette recorder. RESULTS Figure 1 illustrates a potential problem in using strychnine as the sole d e t e r m i n a n t in concluding that agonists act on the glycine receptor. In this cell,

Retinal glycine receptors voltage clamped at - 70 mV, both GABA and glycine (100 #M) produced inward currents, since the chloride reversal potential was more positive than the holding potential. When the cell was then treated with strychnine (20/~M) the glycine current was completely eliminated, but the GABA current was suppressed. We did not find a concentration of strychnine that was sufficient to completely block 100 # M glycine which did not also show a suppression of the GABA effect. This crossover of strychnine antagonism in the retina has been reported previously when high antagonist concentrations were used, 1°'35 although antagonist specificity was observed when low concentrations were tested. 19'34 We therefore used desensitization to probe glycine receptor ligand sensitivity, c-serine, c-alanine, taurine and /~-alanine were also found to invoke inward currents from holding potentials of - 7 0 mV in isolated cells. The chemical formulae of these other amino acids are very similar to GABA and glycine, as illustrated in Fig. 2. The currents produced by these amino acids had the same reversal potential as that of glycine and GABA, suggesting that they open the same ionic channel, which is chloride selective. 8:L35 In addition, all the amino acids produced a pronounced desensitization at high agonist concentrations. As reported by others, this desensitization was rapid, having a time constant of a few seconds and reached a steady state within 10-20s. n'2°'35 Desensitization was characterized by a reduction in the agonist induced current during constant drug application. Small voltage steps indicated that the

155

decline in current with time correlated with a decrease in membrane conductance (data not illustrated). This indicates that the observed phenomenon was desensitization and not fading resulting from a reequilibration of chloride ions across the cell. L-Serine, taurine, c-alanine, //-alanine all cross desensitized with glycine. For example, after desensitization produced by 100/~M glycine (Fig. 3A), the response to 500/~M/~-alanine was almost completely suppressed. Conversely, after desensitization produced by 500/~M fl-alanine, 100/~M glycine produced a very small current. We did not observe a similar cross desensitization between these amino acids and GABA. For example, in the cell shown in Fig. 3B, 500pM /~-alanine produced a large inward current. But after the prolonged application of 100/~M GABA, which caused desensitization of GABA receptors, /~-alanine still produced a large inward current. Similar experiments with c-alanine, L-serine and taurine indicated that none showed significant cross desensitization with the GABA receptor. Both c-serine and L-alanine mimic the effect of glycine in the amphibian retinal eyecup preparation) 5 These amino acids also cross desensitized with glycine (Fig. 4), although their potency was much less than that of//-alanine. Figure 4A shows a neuron in which 500/~M L-serine induced a small inward current. After full recovery, application of 100#M glycine produced a larger inward current, which desensitized the response to 500/~M L-serine. There was a similar cross desensitization between glycine and L-alanine 100 /~M GABA ~ /

100 ~M CLYCINE 20 ~M STRYCHNINE

200 pA

20

s

Fig. 1. Strychnine (20/~M) completely blocked the glycine induced current but also partially blocked the GABA current of a third order retinal neuron clamped at --70 mV. Currents are plotted in time from left to right with downward deflectionscorresponding to inward current. The duration of drug application is marked by a coded bar above the current trace.

156

Z.-H. Pan and M. M. Slaughter

N~+ ~.~,\

too-

\

/ooo \CH 2

NH;\

/,'

\ \ __ CH 2

GLYClNE

-- O

CH 3

%/

/ CH 2

CH 2


-\\

\\

GLYCINE METHYL

ESTER

",oH

~.~- NH~\ / c o o

CHg0H

CH 3

L SERINE

L ALANINE

SARCOSINE

~H\

C -- NHg

oH2

GLYCINAMIDE

GABA

"~"

,

/

NH:~o

+ NH~

S03

liI

COO

/

\

/.

CH 2 TAURINE

/ <

+ N\

\

/ COO

"cH/

,

H•, / HYPOTAURINE

/

.

CO0

CH 2

,6'- ALANINE

AZETIDINECARBOXYLIC ACID

DEHYDRO- DL-PROLINE

Fig. 2. The chemical formulae of endogenous amino acids that are putative glycine agonists and other glycine analogs are illustrated.

c~. 100 ~zbl G1,YCINI< L]T/TETI 5 0 0 t~M ~

ALAN[NE

~Z#/TAS~ 100 f*M GAF~A

A)

O1L[] -

.

WF~H

I r rTl]7~77qq]l

20 s

E]]II~ frill i

VI/SZ~Z./S/JSJI'~/~////~/~d

Fig. 3. Comparison of the effects of glycine and GABA desensitization on fl-alanine induced currents in third order neurons, clamped at - 7 0 mV. (A) 100 p M glycine produced a peak current that desensitized. After glycine desensitization, 500/~ M fl-alanine produced only a small current. After glycine was removed, fl-alanine produced a large inward current similar to the glycine current. After fl-alanine desensitization, 100/aM glycine produced a small inward current. (B) 500#M fl-alanine produced a large inward current. Although 100#M GABA produced a large inward current that desensitized, 500/~M fl-alanine still produced a large current in the presence of this GABA desensitization. The peak inward currents were clipped for this record.

(Fig. 4B). At concentrations < 1 mM, L-serine and L-alanine never completely desensitized the response to 100 p M glycine. A typical result is illustrated in Fig. 5A, where 500/t M L-serine only partially desensitized the response to 100/~M glycine. The lack of full cross desensitization most likely reflects the low potency of serine. This is supported by data such as illustrated in Fig. 5B, which shows that low doses of glycine (20/~ M, which approximates the effectiveness of 500 # M L-serine) produced a similar, partial desensitization of its own receptor population. Taurine has often been suggested as a possible transmitter in the retina and elsewhere in the CNSj3,29 3~ We have found that taurine fully cross desensitizes with glycine. For example, in the cell illustrated in Fig. 6A, 100/~M glycine rapidly produced a peak inward current which then gradually decayed to a steady state level due to desensitization. After a steady state of desensitization had been reached, addition of 1 m M taurine did not evoke an additional inward current. This suggests that the taurine responsive receptors were totally desensitized by glycine. After the drugs were washed out and the receptors were allowed to recover from desensitization, application of 1 m M taurine evoked an inward current which peaked and plateaued, like the response to glycine. When desensitization reached a steady state, application of 100/tM glycine produced no additional inward current. If these receptors were totally independent, these currents would have been additive. Figure 6B demonstrates that hypotaurine is much less effective than taurine in activating the glycine receptor. The observation that taurine and glycine fully cross desensitize suggests that each activates the full set of the other's receptors. However, there are numerous

Retinal glycine receptors

157 IKVIF] 5 0 0 /~M L - S E R I N E ~///////z

5 0 0 p~M L - A L A N I N E 1 0 0 ~ M GLYCINE

A)

B)

50 pA lOs

V

10 s

Fig. 4. Glycine cross desensitizes with L-serine and L-alanine. (A) 500 # M L-serine produced a small inward current, which was suppressed after glycine desensitization. (B) L-alanine (500 p M ) produced an inward current that was also blocked for glycine desensitization.

am<~\~ 20 /zM GLYCINE

100 /zM GLYCINE ~,\x\'~,,r 1 mM TAURINE

t:KaXE~ 100 #M GLYCINE filEVT~ 5 0 0 /zM L SERINE

muurmuuuLl[HIIIIrlLkktktLtllkhALI

s

10 s

uuuuut 500 p.M HYPOTAURINE 500 #M TAUIRINE [3)

am~

lOs

Fig. 5. (A) L-serine (500 p M) only partially desensitized the response to 100/tM glycine. (B) Low (10 p M ) glycine concentrations, which are approximately equivalent to 500 p M t.-serine, produced a very similar partial desensitization.

Fig. 6. (A) Evidence that taurine acts exclusively at the glycine receptor, since 100#M glycine and 1 m M taurine fully cross desensitize. (B) Hypotaurine produced a much smaller current than an equimolar concentration of taurine.

Z.-H. Pan and M. M. Slaughter

158 ~

500 #M TAURINE

1

2 5 0 /xM TAG

1 0 0 JiM GLYCINE

A I

1 0 0 pA los

/

/

/

i

, / 1 0 0 pA 10

J

Fig. 7. TAG is not a selective taurine antagonist in the retina. (A) 250 pM TAG blocked approximately 50% of the taurine induced current. (B) However, 250 #M TAG also blocked about 60% of the glycine induced current. reports in the literature that taurine has its own unique receptor. One basis for this suggestion is the finding that TAG is a selective blocker of taurine, but not glycine, in the C N S . 29'30"39 We did not find this selectively. Figure 7A shows a cell in which 500 pM taurine produced a large inward current. Approximately 50% of this taurine current was blocked in the presence of 250pM TAG. However, Fig. 7B shows that TAG was also an effective glycine antagonist, blocking about 60% of the current induced by 100 pM glycine. The third application of glycine (far right) shows that prolonged desensitization was not a factor in the TAG mediated block of the glycine response. TAG alone produced no current. In general, we found that TAG was equally effective in blocking glycine and taurine induced currents. There was a lot of variability, from cell to cell, in the amount of current produced by glycine and its analogs, presumably representing differences in receptor number. However, a comparison of the currents in any one cell always showed that glycine was the most potent, followed by fl-alanine, then taurine, then L-alanine, while L-serine was the least potent. Series, such as that shown in Fig. 8, were used to establish relative potency. Four hundred micromolar fl-alanine and 1 mM taurine were about equipotent with 100/~M glycine. Equipotent concentrations of L-serine and L-alanine exceeded 1 mM and were not pursued.

Despite the fact that 500 pM L-serine or L-alanine produced only small currents in the voltage clamp experiments described above, they nevertheless produced very significant voltage responses in current clamp experiments. We found that 500 pM L-serine or L-alanine produced almost the same voltage response as the more potent 500 p M fl-alanine or 100pM glycine. This phenomenon is illustrated in Fig. 9, which compares the effects of glycine and its analogs on membrane current and voltage. Figure 9A shows the response of a cell to sequential application of L-alanine, taurine and fl-alanine. The upper trace shows the current responses, when the cell was clamped at - 70 mV, the lower trace in Fig. 9A shows the voltage responses evoked by these glycine analogs in the same cell. The chloride reversal potential was near 0 mV, so that the effect of all these agents on voltage could be discriminated. L-Alanine, the weakest of the three agonists, produced a voltage response that was almost equivalent to that of the more potent agonists. The current clamp data do not reflect the large differences in potency revealed in the voltage clamp data. The difference in potency is reflected more in the speed and duration of depolarization, rather than amplitude. This is more clearly seen in Fig. 9B, which shows the voltage response to 500#M L-serine, 500/~M taurine and 100pM glycine. Glycine produced the largest voltage change, followed by taurine, then L-serine. This follows the potency series observed in the voltage clamp experiments. Nevertheless, L-serine produced a very significant voltage response, almost equivalent to that of the more potent analogs. In contrast with the voltage clamp data, there was no evidence of desensitization in current clamp when high concentrations of agonist were used. This can be explained by the non-linear relationship between current and voltage. Specifically, the desensitized current was still sufficient to produce a large depolarization. The selectivity of the glycine receptor is presumably based, at least in part, on the small size of the glycine molecule. The fact that L-alanine is significantly less potent than glycine, while L-serine is even weaker, and threonine is without effect (not shown), suggests that steric hindrance around the alpha carbon is a factor in the selectivity of the receptor. We also investigated the importance of steric hindrance around the amino group (sarcosine) and the carboxyl group (glycine methyl ester) and the effect of incorporating the amino terminus in a ring structure (azetinecarboxylic acid and dehydro-proline; see Fig. 2). These modifications to the glycine molecule resulted in a dramatic reduction in activity, as illustrated in Fig. 10. Formation of a ring that includes the amine produced agents with a marginally detectable activity at concentrations up to 1 mM. At a concentration of 5 mM, these ring structures produced small currents (Fig. 10A). If a methyl moiety was added to the amine (sarcosine), the activity was also dramatically reduced (Fig. 10A, B). The current induced by sarcosine

Retinal glycine receptors (N-methyl glycine) was much less than L-alanine, where the methyl group was added to the alpha carbon. In contrast, when a methyl group was added to the carboxyl terminal, the resulting glycine methyl ester apparently remained an effective glycine agonist (Fig. 10B, C). This action of glycine methyl ester, and even glycine ethyl ester, has been noticed previously36 and led to the suggestion that the hydroxyl group of glycine, and not the anionic oxygen, is important for receptor binding. This implies that receptor binding at the carboxyl end of glycine involves hydrogen bonding rather than electrostatic interactions. Consequently, we tested glycinamide, a glycine analog that should be even more effective than glycine methyl ester in making hydrogen bonds. Glycinamide was ineffective at concentrations up to l raM. Dr Jo Cunningham kindly carried out a high pressure liquid chromatography analysis of our glycine methyl ester solutions and found that they contained as much as 10% glycine. This contamination, occurring either in factory synthesis or in hydrolysis in solution, fully accounts for the apparent effectiveness of glycine methyl ester.

159

tively desensitized the responses to these short chain amino acids, while GABA (100 #M) was ineffective. The relative potency is: glycine >>fl-alanine > taurine >>L-alanine > L-serine. This is equivalent to the potency series originally reported in cat spinal cord, 15 where extracellular spike activity was measured, and to rat isolated ventromedial hypothalamic neurons, where desensitization was measured. 36There was no indication that any of these ligands acts on distinct receptors. There have been numerous suggestions that taurine is a neurotransmitter in the retina and acts at its own receptor. 31 In the spinal cord, TAG has been found to block taurine and fl-alanine but not glycine or GABA, 29'3° the strongest evidence for discrete receptors. We did not find that TAG expressed selectivity in blocking taurine compared with glycine. Furthermore, cross desensitization between 100/~M glycine and 1 mM taurine showed that both agents were able to fully desensitize the response to the other, indicating that they activate fully overlapping receptor populations.

Comparisons with intact retina pharmacology DISCUSSION

Agonist action at the glycine receptor These desensitization experiments confirm that Lserine, L-alanine, taurine and fl-alanine all activate the glycine receptor. Glycine (IO0#M) very effec-

The results in this study correlate well with experiments in the intact amphibian retina, where it was found that taurine, ~3 L-alanine and L-serine 33 mimicked the effects of glycine and were blocked by strychnine but not by GABA antagonists. However, there are several interesting distinctions between the 500 /*M L-ALANINE rT~

500 /~M TAURINE

500 t.tM fl-ALANINE V//7/7/A500 /zM L-SERINE

A)

/ /

40 pA lOs

i

200 pA 10 s

Fig. 8. Examples of test series used to determine the relative potency of glycineagonists. (A) At equimolar concentrations, L-serinewas less potent than g-alanine, while fl-alanine was more potent. (B) Taurine was more potent than L-alanine, but less potent than fl-alanine.

Z.-H. Pan and M. M. Slaughter

160

, J 500 ~,\~\~ 5 0 0 ............. 5 0 0 500 100

A)

~

~

/~M #M ¢zM #M /zM

L-ALANINE TAURINE /3-ALANINE L SERINE GLYCINE

nnn~

0

p A lOs

10 s -2~

GK2~

KK~

2

,['--h t I

,I

i

't

f

'~

~

20 r;/V

l_ 10s

Fig. 9. Comparison of the voltage and current responses evoked by a number of amino acids. (A) The top trace shows the currents evoked (in voltage clamp, - 70 mV) by equimolar concentrations of Dalanine, taurine and/~-alanine. The trace below shows the depolarization produced by all these agents on the same cell, in current clamp (zero current, membrane potential approximately - 7 0 mV). Note that the voltage response amplitudes are very similar, although the currents are significantlydifferent. (B) This trace, from another cell, shows that 500/~M L-serine, 500/IM taurine and 100/iM glycine produced very similar membrane depolarizations. The weakest agonist, L-serine, produced almost as large a potential change as the strongest agonist, glycine. There are differences in the rise time and duration of action. This cell's rest membrane potential was about 70 mV.

voltage clamp data obtained in isolated cells and the current clamp data in the intact retina. In isolated cells, GABA antagonists block part of the glycine response, while glycine antagonists block part of the GABA response. Paradoxically, in the intact retina strychnine, bicuculline and picrotoxin seem to be fairly specific. 5'2833 Our experiments comparing voltage and current clamp data may explain this discrepancy. The voltage clamp data indicate that high concentrations of strychnine, which totally block the glycine receptor, partially block GABA currents. But current clamp data show that the remaining, unblocked GABA current is still sufficient

to produce a near maximal voltage response. Therefore, in voltage recordings from intact retina, strychnine crossover to GABA receptors is not apparent. A related observation is that agonists produce a peak and plateau in the current records but only a sustained response in the voltage records. The small sustained plateau current is sufficient to maintain the cell near its peak depolarized state. This can be seen by comparing the voltage and current clamp responses to e-serine and glycine in Figs 4a and 9. While the early spike of current induced by glycine adds little to the amplitude of the depolarization, it does quicken the voltage response rise time.

Retinal glycine receptors

Interpretation of desensitization Desensitization can be either receptor specific (homologous) or affect a spectrum of receptor types (heterologous). The observation that the analogs tested fully cross desensitize with glycine but not GABA suggests that the process examined in this study is homologous desensitization. However, there is no way of excluding the possibility that there are discrete receptors for taurine, L-serine, or other amino acids that show complete heterologous desensitization with the glycine receptor, but not the GABA receptor. Consequently, these cross desensitization experiments alone are not conclusive. However, by corroborating the conclusions derived using antagonists, our results strengthen the conclusion that all of these short chain amino acids act on the glycine receptor.

Structure-activity relationships The inhibitory glycine receptor is thought to bind to the amine and the carboxyl ends of glycine, both of which are ionized at physiological pH. Since several amino acids contain the same amine and carboxyl groups in an identical conformation, the specificity of the glycine receptor is believed to be based on interacting with the small glycine molecule while excluding interactions with larger ligands. Our results indicate that steric hindrance is greatest at the amino and carboxyl terminals, and less significant at the alpha carbon at the center of the glycine molecule. At the

161

alpha carbon, potency declined as the side chain size increased from a hydrogen (glycine), to a methyl (alanine), to a methanol (serine). The decrease in potency is not due to efficacy since high concentrations of weak agonists could generate large currents. Larger additions at this alpha carbon (leucine, threonine, valine) made the ligands ineffective at concentrations of 1 mM. The selectivity of the receptor for glycine over GABA apparently results from steric hindrance around glycine's alpha carbon region, where GABA must fold to match the glycine pharmacophore. Modification of the amino acid sequence of the glycine ~t1 receptor subunit indicates that hydroxyl groups on amino acids 159 and 161 are important in restricting ligand interactions with the glycine receptor. 25 If the phenylalanine at position 159 and the tyrosine at position 161 are interchanged, then the resulting receptor is much more sensitive to ~-alanine, which becomes more potent than glycine itself. In addition, D-serine and GABA also activate this modified receptor, although they were inactive in the wild type. Thus, these two amino acids are at least partially responsible for the steric restrictions at the alpha carbon of glycine. Compared with modifications at the alpha carbon of glycine, there is much less tolerance at the amino and carboxyl moieties. Placing a methyl moiety on the amino group (sarcosine) reduces activity much more than placing a methyl group on the alpha carbon (L-alanine). Steric influences at the carboxyl

1 mM SARCOSINE ~\\\\\'~ 5 mM SARCOSINE P777/~ 5 mM DEHYDRO DL PROLINE i ~ 5 mM AZETINECARBOXYLIC ACID i

1 mM GLYCINE METHYL ESTER

I - - I 1 mM /3-ALANINE

E5 pA [ _ _ 15s

B)

i

L~ 100 pA I 10 s

c)

[ZZZ3

i

100 pA 10 s

Fig. 10. shows the effect of modification of the amine or carboxyl group of glycine. (A) Additions to the amino terminus greatly reduced activity. (B) Addition of a methyl group to the carboxyl end of glycine, glycine methyl ester, produced an agent that appears to retain activity. (C) Glycine methyl esters appears to be slightly more potent than an equimolar dose of fl-alanine. The effect of the glycine methyl ester can be fully accounted for by glycine contamination.

162

Z.-H. Pan and M. M. Slaughter

group are difficult to examine because of the lack of compounds with an appropriate anionic oxygen, other than a carboxyl group. The restriction at the carboxyl end is illustrated by the difference in potency between taurine and hypotaurine. Hypotaurine, an analog of taurine in which one oxygen is removed from the sulphur group, is much weaker than taurine. The low activity of glycine methyl ester indicates that the ionized oxygen in the carboxyl group is essential for binding. This conflicts with the findings of Tokutomi et al. 36 who reported that glycine methyl ester was an effective agonist. This led them to conclude that glycine binding to the receptor involves hydrogen bonds, not ionic interactions, at the carboxyl group. However, they also used high concentrations of glycine methyl ester before seeing an effect on the glycine receptor. Thus, they may have been misled by contamination problems that were similar to our own. It is possible that both oxygens on the carboxyl group are involved in ionic interactions with the receptor, as suggested by molecular modeling/ The inhibitory glycine receptor is distinctly different from the allosteric glycine site on the NMDA receptor? t Methylation of the amine or carboxyl terminus block activity at both sites, but methylation at the alpha carbon appears to reduce activity more at the NMDA receptor. 27 Lengthening the carbon chain (/~-alanine or taurine) is also more detrimental at the NMDA receptor. The NMDA receptor site also shows a steric preference for the o- form of agonists such as alanine or serine. Interestingly, McBain et al/7 suggest that the hydroxyl group of o-serine binds to an auxiliary site on the receptor protein, while there have been some suggestions that hydrogen bonding may also be important at the inhibitory glycine receptor. 36'37 Contrasting the pharmacologies of the inhibitory and the N M D A glycine receptor demonstrates that even a molecule as simple as glycine can interact with two very distinct receptors. Functional significance

The fact that all these amino acids act on glycine receptors but glycine is the most potent suggests that the actions of these other ligands are an epiphenomenon of no physiological significance. On the other hand, the levels of serine and taurine in the retina are very high, making them difficult to ignore. ~°~33 In addition, current clamp data show that these agents can produce very significant voltage responses and studies in the intact retina show that exogenous taurine, L-serine, and L-alanine can suppress synaptic activity. 8'~3'33If these agents do have a role in activat-

ing the glycine receptor what might it be? One possibility is that they are neuromodulators which enhance the potency of glycine. If glycine receptor activation requires cooperative binding of two or more glycine molecules (e.g. Tokutomi et al. 36 finds a Hill coefficient of 1.8), then taurine or L-serine could bind at one site, thus requiring glycine itself to bind at only one site. This would enhance the potency of glycine at the synapse. A similar mechanism has been proposed in the rat cerebellum, where taurine and GABA both activate the GABA receptor. 7't7 Both amino acids have been colocalized to Purkinje cell terminals and corelease has been postulated. Another possibility is that these glycine analogs are independent neurotransmitters. In this case, our data would suggest that they could produce significant potential changes, but much smaller conductance changes than glycine. There may be conditions when these small conductances could be significant. For example, third order retinal neurons have very high input resistances 12and are under tonic inhibitionYs Under these conditions, agents such as taurine could produce significant voltage responses and moderate inhibition. But would not lower concentrations of glycine accomplish the same function? One might speculate that the non-linearity in transmitter release plus the non-linearity in receptor interaction (cooperativity) may mean that it is more effective to use alternative transmitters of very different potencies to increase dynamic range. If these analogs do have a functional role, then it adds another permutation at the synapse: not only may one transmitter act on various postsynaptic receptors, but several different transmitters can act at a single receptor. CONCLUSIONS

Several amino acids, including L-serine, L-alanine, taurine and fl-alanine, activate the glycine receptor. Glycine fully desensitizes responses to these amino acid analogs, giving no evidence for other, discrete receptors for any of these amino acids. Relative to glycine, these analogs produce small currents. However, they can produce significant voltage changes suggesting that they could function as inhibitory transmitters. Structure-activity comparisons indicate that the inhibitory glycine receptor binds to charged sites on the amine and carboxyl groups of the glycine molecule. Steric interference at these putative binding sites of glycine reduces activity more significantly than does steric interference at the alpha carbon.

Acknowledgements--This work was supported by the National Eye Institute grant No. EY05725.

REFERENCES

1. Aibara K., Oonuma M. and Akaike N. (1991) Strychnine-induced potassium current in isolated dorsal root ganglion cells of the rat. Br. J. Pharmac. 102, 492-496. 2. Aprison M. H. and Lipkowitz K. B. (1992) Muscimol and N,N, dimethylmuscimol:from a GABA agonist to a glycine antagonist. J. Neurosci. Res. 31, 166-174.

Retinal glycine receptors

163

3. Bader C. R., MacLeish P. R. and Schwartz E. A. (1979) A voltage-clamp study of the light response in solitary rods of the tiger salamander. J. Physiol., Lond. 296, 1-26. 4. Barker J. L. and McBurney R. N. (1979) GABA and glycine may share the same conductance channel on cultured mammalian neurones. Nature 277, 234-236. 5. Belgum J. H., Dvorak D. R. and McReynolds J. S. (1984) Strychnine blocks transient but not sustained inhibition in mudpuppy retinal ganglion cells. J. Physiol. 354, 273-286. 6. Betz H. (1991) Glycine receptors: heterogeneous and widespread in the mammlian brain. Trends Neurosci. 14, 458-461. 7. Billard J. M. (1990) Taurine in deep cerebellar nuclei of the rat. In vivo comparison to GABA inhibitory effect. Brain Res. 514, 155-158. 8. Bolz J., Thier T., Voigt T. and W/issle H. (1985) Action and localization of glycine and taurine in the cat retina. J. Physiol., Lond. 362, 395-413. 9. Bormann J., Hamill O. P. and Sakmann B. (1987) Mechanism of anion permeation through channels gated by glycine and ?-aminobutyric acid in mouse cultured spinal neurones. J. Physiol., Lond. 276, 243-286. 10. Cohen A, I., McDaniel M. and Orr H. (1973) Absolute levels of some free amino acids in normal and biologically fractionated retinas. Invest. Ophthal. 12, 686q593. 11. Cohen B. N., Fain G. L. and Fain M. J. (1989) GABA and glycine channels in isolated ganglion cells from the goldfish retina. J. Physiol. 417, 53-82. 12. Coleman P. A. and Miller R. F. (1989) Measurement of passive membrane properties with whole cell recording from neurons in the intact amphibian retina. J. Neurophysiol. 61, 218-230. 13. Cunningham R. A. and Miller R. F. (1980) Electrophysiological analysis of taurine and glycine action on neurons of the mudpuppy retina--I. Intracellular recording. Brain Res. 197, 123 138. 14. Curtis D. R., Duggan A. W., Felix D., Johnston G. A. R. and McLennan H. (1971) Antagonism between bicuculline and GABA in the cat brain. Brain Res. 33,57 73. 15. Curtis D. R., Hosli L. and Johnston G. A. R. (1968) A pharmacological study of the depression of spinal neurons by glycine and related amino acids. Expl Brain Res. 6, 1-18. 16. Doi T. and Ohmori H. (1993) Acetylcholine increases intracellular Ca 2÷ concentration and hyperpolarizes the guinea-pig outer hair cell. Hearing Res. 67, 179 188. 17. Frederickson R. C. A., Neuss M., Morzorati S. L. and McBride W. J. (1978) A comparison of the inhibitory effects of taurine and GABA on identified Purkinje cells and other neurons in the cerebellar cortex of the rat. Brain Res. 145, 117-126. 18. Fuchs P. A. and Murrow B. W. (1992) A novel cholinergic receptor mediates inhibition of chick cochlear hair ceils. Proc. R. Soc. Lond. B. 248, 35-40. 19. Gilberton T. A., Borges S. and Wilson M. (1991) The effects of glycine and GABA on isolated horizontal cells from the salamander retina. J. Neurophysiol. 66, 2002-2013. 20. Ishida A. T. and Cohen B. N. (1988) GABA-activated whole-cell currents in isolated retinal ganglion ceils. J. Neurophysiol. 60, 381-396. 21. Johnson J. W. and Ascher P. (1987) Glycine potentiates the N M D A response in cultured mouse brain neurons. Nature 325, 529-531. 22. Kaneda M., Wakamori M. and Akaike N. (1989) GABA-induced chloride current in rat isolated Purkinje cells. Am. J. Physiol. 256, 1153-1159. 23. Lam D. M.-K. (1972) Biosynthesis of acetylcholine in turtle photoreceptors. Proc. natn. Acad. Sci. U.S.A. 69, 1987 1991. 24. Lewis C. A. and Faber D. S. (1993) GABA responses and their partial occlusion by glycine in cultured rat medullary neurons. Neuroscience 52, 83 96. 25. Malosio M. L., Grenningloh G., Kuhle J., Schmieden V., Schmitt B., Prior P. and Betz H. (1991) Alternative splicing generates two variants of the alpha 1 subunit of the inhibitory glycine receptor. J. biol. Chem. 266, 2048-2053. 26. Marc R. C. (1989) Role of glycine in the mammalian retina. Prog. Retinal Res. 8, 67 107. 27. McBain C. J., Kleckner N. W., Wyrick S. and Dingledine R. (1989) Structural requirements for activation of the glycine coagonist site of N-methyl-D-aspartate receptors expressed in Xenopus oocytes. Molec. Pharmac. 36, 556-565. 28. Miller R. F., Frumkes T. E., Slaughter M. and Dacheux R. F. (1981) Physiological and pharmacological basis of GABA and glycine action on neurons of mudpuppy retina--II. Amacrine and ganglion cells. J. Neurophysiol. 45, 764-782. 29. Okamoto K., Kimura H. and Sakai Y. (1983) Evidence for taurine as an inhibitory neurotransmitter in cerebellar stellate interneurons: selective antagonism by TAG (6-aminomethyl-3-methyl-4H-1,2,4-benzothiadiazide-l,l-dioxide). Brain Res. 265, 13 168. 30. Padjen A. L., Mitsoglou G. M. and Hassessian H. (1989) Further evidence in support of taurine as a mediator of synaptic transmission in the frog spinal cord. Brain Res. 488, 288596. 31. Pasantes-Morales H. (1986) Current concepts on the role of taurine in the retina. Prog. Retinal Res. 5, 207-229. 32. Schmieden V., Kuhse J. and Betz H. (1993) Mutation of glycine receptor subunit creates fl-alanine receptor responsive to GABA. Science 262, 256-258. 33. Slaughter M. M. and Miller R. F. (1991) Characterization of serine's inhibitory action on neurons in the mudpuppy retina. Neuroscience 41, 817 825. 34. Suzuki S., Tachibana M. and Kaneko A. (1990) Effects of glycine and GABA on isolated bipolar cells of the mouse retina. J. Physiol. 421, 645~62. 35. Tauck D. L., Frosch M. P. and Lipton S. A. (1988) Characterization of GABA- and glycine-induced currents of solitary rodent retinal ganglion ceils in culture. Neuroscience 27, 193-203. 36. Tokutomi N., Kaneda M. and Akaike N. (1989) What confers specificity on glycine for its receptor site? Br. J. Pharmac. 97, 353 360. 37. Vandenberg R. J., Handford C. A. and Schofield P. R. (1992) Distinct agonist- and antagonist-binding sites on the glycine receptor. Neuron 9, 491-496.

164

Z.-H. Pan and M. M. Slaughter

38. Yakushiji T., Tokutomi N., Akaike N. and Carpenter D. O. (1987) Antagonists of GABA responses, studied using internally perfused frog dorsal root ganglion neurons. Neuroscienee 22, 1123-1133. 39. Yarbrough G. G., Singh D. K. and Taylor D. A. (1981) Neuropharmacological characterization ofa taurine antagonist. J. Pharmac. exp. Ther. 219, 604-613. 40. Yazulla S. (1986) GABAergic mechanisms in the retina. Prog. Retinal Res. 5, 1 52. (Accepted 21 June 1994)