Neuroscience Letters, 27 ( 1981) 81-87 81

Elsevier/North-Holland Scientific Publishers Ltd.

EFFECTS OF OL-a-AMINO ADIPIC ACID ON MULLER CELLS IN FROG AND CHICKEN RETINAE IN VIVO: RELATION TO ERG b WAVE, GANGLION CELL DISCHARGE AND TECTAL EVOKED POTENTIALS

N1COLE BONAVENTURE, GUY ROUSSEL and N O R M A WIOLAND

Laboratoire de Neurophysiologie, Centre de Neurochimie du CNRS, 5, Rue Blaise Pascal, 67000 Strasbourg (France)

(Received March 12th, 1981; Revised version received and accepted August 28th, 1981)

In both frog and chicken, an intravitreal injection of DL-C~-amino-adipic acid, (DL-c~aaa) provoked a

progressive depression and eventually the disappearance of the ERG b wave that was concomitant with severe damage to the M011er cells without any apparent damage to retinal neurons. Ganglion cell

discharges as well as tectal evoked potentials were still recorded, i.e. a visual message was still generated in the retina and transmitted to the optic tectum, when the Mt~ller cells had been damaged to as to provoke an abolition of the ERG b-wave. The whole of the drug-induced effects proved to be reversible.

It is generally admitted [1] that the ERG b wave is generated by a radially oriented dipole extending through the inner nuclear and both plexiform layers. MOiler neuro- glial cells are the only retinal cells that can produce the extracellular current flow giving rise to the b wave [5]. By means of intracellular recordings performed on the Necturus retina, Miller and Dowling [8] have shown a photic stimulus to induce a slow depolarization of MOiler cells, the recorded potentials being similar in wave form and latency to the ERG b wave. Moreover, a developmental study by Rager [12] revealed a close temporal correlation to exist between MOiler cell maturation and b wave development. On the other hand, DL-a-aminoadipic acid (DL-o~aaa) was found to specifically destroy glial cells without affecting neurons in newborn mouse brain [9] as well as in rat retina [11]. This damage is reversible since glial cells can be identified again 24 h after the injection. Moreover, DL-e~aaa proved to be a potent antagonist of the depolarizing action of the excitatory aminoacids in various struc- tures of the central nervous system [2, 6] including the retina [14].

Based on both orders of data, the present study aimed at specifying the morpho- logical and physiological effects on an intravitreal administration of DL-aaaa, parti- cular emphasis being put on: (1) the temporal relation between the alterations affec- ting the MOiler cells and the ERG b wave, respectively; and (2) the extent to which a severe damage to the glial cells, that might cause an abolition of the ERG b wave, would affect the generation of the visual message as evidenced by ganglion cell

0304-3940/81/0000-0000/$ 02.50 © Elsevier/North-Holland Scientific Publishers Ltd.

82

discharges and tectal evoked potentials. In a most recent paper received after completion of our own study*, Szamier et al. [13] reported that in the isolated retina of the skate, DL-c~aaa caused severe MOiler cell damage together with an abolition of the ERG b wave. The latter recovered its initial amplitude as soon as the cytoplasmic membranes of MOiler cells had undergone repair and in spite of persisting loss of cytoplasmic electron density.

Our study was carried out on curarized chickens (n = 23) and frogs (n = 30). The techniques of electrophysiological recordings as well as the histological methods used were described previously [3, 4]. Recording were achieved before and after an intravitreal administration of the drug. Thirty #1 (frog) or 50/zl (chicken) of DL-c~aaa were injected at various concentrations (0.3, 0.15 and 0.06 M) into the vitreous body. Control injections of saline never altered the ERG nor the tectal evoked potentials (TEP).

The electrophysiological effects observed with regard to ERG, ganglion cell dis- charges and TEP can be described in three successive steps (see Fig. 1).

(1) Dk-c~aaa (0.3 or 0.15 M) provoked a progressive depression of the ERG b wave which had totally disappeared 1 h after the injection. At a weaker concentration (0.06 M), it produced a decrease of the b wave amplitude but never its total aboli- tion. The a wave was not modified at that stage. At the same time, there occurred an increase in ganglion cell discharges and TEP amplitude. (2) About 1 h after the injection, i.e. when the b wave was totally abolished, the ERG a wave showed an increase in amplitude in about 15°70 of the studied animals (both frogs and chickens), most probably due to an increase or to an unmasking of Pro. At the same

?v

40(

/

/ A

-s -~ -3 _2 -~ 0 B log INTENSITY

ERG ~ - ~ before D L ~ a a a A_A a f t e r H

TEP o - - o before ~

e - - e a f t e r n

Fig. 1. Drug-induced electrophysiological (ERG and tectal evoked potentials, TEP) alterations. A and B:

before drug administration. C and D: half an hour after drug administration. E and F: 2 h after drug

administrat ion. Stimulus duration = 1 sec.

*A preliminary report of some of our data was presented at the 1SCEV meeting (Amsterdam, May 1980)

under the title 'Morphological and physiological effects of DL-aaaa on the ret ina ' .

83

time, the rate of ganglion cell discharges decreased, especially at OFF. The pattern of the TEP became simplified and their amplitude decreased, again especially at OFF. (3) At 3 -4 h after injection, the ERG was no longer altered, but ganglion cell discharges and TEP had disappeared in about 25% of the studied animals.

Following an injection of DL-c~aaa, the OFF components of both ERG and TEP were transiently abolished in the frog, while they remained unchanged in the chicken, suggesting a different origin of these OFF components in the two species.

The intensity-voltage curves for both the ERG b wave and the first positive wave of the TEP were established before and after drug administration (Fig. 2). In the control animals, the response threshold was nearly the same for both b wave and TEP. One hour after drug administration, however, the b wave threshold was increased by two log units, whereas the TEP threshold was increased by 0.5 log unit only. The I /V curve for the TEP was merely shifted by about one log unit to the right, while that for the ERG b wave was much more compressed. Half an hour later, the ERG b wave had totally disappeared, whereas the TEP did not show any further alteration. The physiological effects of DL-c~aaa proved to be totally reversible: 24 h after injection, the ERG, ganglion cell activity and TEP no longer differed from those initially recorded before drug administration.

Chicken and frog eyes were removed at various delays after injection of DL-aaaa, just after having undergone an electrophysiological control of the effect of the injected drug. DL-ctaaa provoked marked histological lesions that could be seen already under the light microscope and that were shown more clearly with the

CHICKEN FROG ERG TEP ERG TEP

|

I 3oopV

!

Fig. 2. Intensi ty-vol tage curves of ERG b wave and first positive component o f TEP, before and 1 h (A) and 1.5 h (B) after drug administration, in the chicken.

84

electron microscope (Figs. 3 and 4). Miiller cells exhibited very marked alterations, the most striking effect consisting in a massive oedematous swelling of the cell bodies and of their processes. The plasma membranes appeared disrupted. A loss of electron density and an internal destruction of the cell were observed, with a vacuo- lization of the cytoplasm and a swelling of the cytoplasmic organelles. The cell nuclei were swollen and chromatin was no longer homogeneously distributed. It was condensed in irregular masses in the nucleoplasm, which appeared very clear. Whenever the Mtiller cell destruction was incomplete, it seemed to prevail in the distal rather than in the proximal part of the retina (Fig. 3B). Likewise, the cell alterations were found to develop first close to the external limiting membrane. No retinal element other than Mtiller cells was affected by Dt.-c~aaa. It thus appeared

that OL-c~aaa had a gliotoxic but not a neurotoxic effect. About 7 h later, however, Mtiller cells could be recognized again and they appeared quite normal. In other words, OL-c~aaa did not provoke an actual necrosis of the cells, but it altered them only transiently. Thus, the drug-induced morphological effects, too, proved to be reversible. Just like Szamier et al. [13], we observed the progressive disappearance

c

Fig. 3. Compared morphology and electroretinography of chicken retina. A; before drug injection; MOiler cell bodies and processes exhibit a characteristic optical density. Normal ERG. B: 1 h after injection of Di.-cxaaa (0.06 M); morphological damage is limited to the distal part of the retina; MOiler celt

processes are emptied at the photoreceptor level. ERG b wave is decreased. C: 1 h after injection of t)l - ~aaa (0.3 M); Mtiller cell bodies and processes are totally emptied. The arrow indicates the clear Mtiller

cell nuclei. The b wave is totally abolished. 360 x .

85

Fig. 4. Electron micrographs of the internal nuclear layer of the chicken retina. A: intact MOiler cell body (M) with this dense processes in control animal. B: damaged M011er cell (nucleus and processes) after injection of 50 ~1 of OL-c~aaa (0.3 M). 3500 ×.

o f the E R G b wave to be c o n c o m i t a n t with a l te ra t ions affect ing the MOiler cells. But

c o n t r a r y to their obse rva t ions m a d e on the isola ted re t ina , we f o u n d a close para l le l -

ism to exist be tween the reduc t ion o f b wave a m p l i t u d e and the degree o f glial cell

des t ruc t ion . As a ma t t e r o f fact , whenever the cell a l te ra t ions were l imi ted to the

d is ta l pa r t o f the re t ina , the b wave was merely depressed (Fig. 3B). This resul t adds

a fur ther a rgumen t to the hypothes is tha t re t inal glial cells are deeply involved in the

gene ra t ion o f the E R G b wave [8, 12].

In the case o f to ta l d a m a g e to Mtil ler cells, the gangl ion cell d ischarges and the

T E P were still r eco rded in response to a phot ic s t imula t ion . In o ther words , a visual

message was still genera ted in the re t ina and t r ansmi t t ed to the opt ic t ec tum, though

86

with a simplified pat tern when the MUller cells had been damaged so as to provoke

an abol i t ion of the ERG b wave. This is of interest with regard to the ques t ion raised

by Szamier et al. [13] as to whether the observed b wave suppression might not be

due simply to an interference with t ransmiss ion between photoreceptors and second

order neurons . The ERG b wave seems to be generated ' in parallel ' to the genera t ion

of the actual visual message. Thus, its abol i t ion may tell little about possible

al terat ions of the various retinal mechanisms.

Fur thermore , DL-c~aaa is considered to be an antagonis t of aspartate and the

quest ion arises as to the respective roles played by gliotoxicity and possible

interact ions with aspartate in the origin of the drug- induced electrophysiological

effects. In the spinal cord [6, 7], the D-isomer of c~aaa was found to be more potent

than the racemic form in blocking the excitatory synaptic receptors. Conversely the

L-isomer is to be considered as an agonist of g lutamate and aspartate. This L-isomer

induces striking gliotoxic and neurotoxic changes within the arcuate hypothalamic

nucleus in in fan t mice, while the D-isomers induces only mild gliotoxic changes [10].

It would be of interest, therefore, to carry out a further compara t ive morphological

and electrophysiological analysis with the D- and e-isomers, respectively. Such a

compar i son should help to clarify to what extent the observed bioelectrical effects

are due to gliotoxicity of De-c~aaa a n d / o r to interference with excitatory neurot rans-

mission in the retina.

The authors express their grat i tude to Miss B. J a r d o n and to Miss G. Rudo l f for

their technical assistance. This work was supported by a grant f rom INSERM (ATP

68.78.100, Cont ra t 12).

1 Arden, G.B. and Brown, K.T., Some properties of components of the cat electroretinogram revealed by local recording under oil, J. Physiol. (Lond.), 176 (1965) 429-461.

2 Bergey, (Lond.), Martin, M.R. and Hermes, M., Effects of D,L-alpha-aminoadipate on postsynaptic amino acid responses in cultured mouse spinal cord neurons, Brain Res., 193 (1980) 199-207.

3 Bonaventure, N., Wioland, N. and Mandel, P. Antagonists of the putative inhibitory transmitter effects of taurine and GABA in the retina, Brain Res., 80 (1974) 281-290.

4 Bonaventure, N., Wioland, N. and Roussel, G., Effects of some amino-acids (GABA, glycine, taurine) and of their antagonists (picrotoxin, strychnine)on spatial and temporal features of frog retinal ganglion cell responses, Pfl0gers Arch., 385 (1980) 51-64.

5 Faber, D.S., Analysis of the slow transretinal potentials in response to light. Ph.D. Thesis, State University of New York at Buffalo, 1969.

6 Lodge, D., Headley, P.M. and Curtis, D.R., Selective antagonism by D-a-aminoadipate of amino acid and synaptic excitation of cat spinal neurons, Brain Res., 152 (1978) 603-608.

7 McLennan, H. and Hall, J.G., The action of D-a-aminoadipate on excitatory aminoacid receptors of rat thalamic neurons, Brain Res., 149 (1978) 541-545.

8 Miller, R.F. and Dowling J.E., Intracetlular responses of the M~ller (glials) cells of mud-puppy retina: their relation to the b wave of the ERG, J. Neurophysiol., 33 (1970) 323-341.

90lney, J.W., Ho, O.L. and Rhee, V., Cytotoxic effects of acidic and sulphur containing amino acids on the infant mouse central nervous system, Exp. Brain Res., 14 (1971) 61-76.

87

10 Olney, J.W., De Gubareff, T. and Collins, J.F., Stereospecificity of the gliotoxic and antineurotoxic actions of alpha-aminoadipate, Neurosci. Lett., 19 (1980) 277-282.

11 Pedersen, O. and Karlsen, R.L., Destruction of Mt~ller cells in the adult rat by intravitreal injection of O,L-o~-aminoadipic acid. An electron microscopic study, Exp. Eye Res., 28 (1979) 569-575.

12 Rager, G., The cellular origin of the b-wave in the electroretinogram. A developmental approach, J. comp. Neurol., 188 (1979) 225-245.

13 Szamier, R.B., Ripps, H. and Chappell, R.L., Changes in ERG b-wave and Mtiller cell structure induced by c~-aminoadipic acid, Neurosci. Lett., 21 (1981) 307-312.

14 Wu, S.M. and Dowling, J.E., L-Aspartate: evidence for a role in cone photoreceptor synaptic transmission in carp retina, Proc. nat. Acad. Sci. (Wash.), 75 (1978) 5205-5210.