*Laboratory of Physiology of Reception, Pavlov Institute of Physiology, Russian Academy of Sciences, Nab. Makarova 6, St. Petersburg 199034, Russia
**Department of Comparative Physiology, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands

To whom correspondence should be addressed.

Abbreviations: L-GLU, L-glutamate; NMDA, N-methyl-D-aspartate; GLY, glycine; 7ClKyn, 7-chlorokynurenic acid; APV, DL-2-amino-5-phosphonovaleric acid; Ket, ketamine; KA, kainate; Q, quisqualate; AMPA, (+-)-alfa-amino-3-hydroxy-5 -methylisoxazole-4-propionic acid; DNQX, 6,7- dinitroquinoxaline-2,3-dione.

Abstract - An excitatory amino acid, most probably L-glutamate, acts as a neurotransmitter at the receptor cell - afferent fibre synapses in the ampullary electroreceptor organs of the freshwater catfish Ictalurus nebulosus. In the present study, we have used an electrophysiological approach to investigate the presence of N-methyl-D-aspartate receptors at this level. N-methyl-D-aspartate, dissolved in a Mg2+-containing (normal) solution, had no effect on afferent activity, not even at 5 mM. However, addition of 5 mM N-methyl-D-aspartate to a Mg2+-free solution evoked an enduring increase in firing rate. The application of N-methyl-D-aspartate combined with electrical sine wave stimulation produced a firing increase in the primary afferents, even in the presence of Mg2+ (1.5 mM). Glycine (0.01-0.001 mM) significantly potentiated the N-methyl-D-aspartate responses. Addition of antagonists of the actions of N-methyl-D-aspartate 7-chlorokynurenic acid, DL-2-amino-5-phosphonovaleric acid, and ketamine in concentrations of 0.5-2.0 mM led to a decrease in resting and stimulus-evoked activity. 7-chlorokynurenic acid also blocked the responses to application of N-methyl-D-aspartate. The glycine agonist D-serine (0.01 mM) prevented the 7-chlorokynurenic inhibitory effect.

      These results suggest the involvement of N-methyl-D-aspartate receptors in mediating the actions of L-glutamate at the afferent synapses of the electroreceptor organs of the catfish.

Key words: Electroreceptor organs; Afferent synapse; Neurotransmitter pharmacology; NMDA receptors

Electroreceptor organs of fish are considered to be part of the octavolateral system1 which also includes the organ of Corti, vestibular and otolith apparatus, and lateral-line organs. A great attention paid to electroreception may be accounted for both by the important role of electroreception in the life of some lower vertebrates and by a relatively simple organization and accessibility of the sensory receptors. On the other hand, the great similarity among receptor organs of the octavolateral system, the common ontogenetic origin and innervation make electroreceptor organs a convenient model system for the analysis of modes of action of other representatives of the octavolateral family.

      Most electroreceptor cells belong to the class of secondary receptor cells which make contacts to their primary afferents via a chemical synapse. No efferent innervation has been found in different kinds of electroreceptor organs. There is a reason to believe that the sensory function of the electroreceptor organs and other organs of the octavolateral system is determined not only by the structural, physiological and biochemical properties of receptor cells, but also by the mechanism of synaptic transmission and the characteristics of associated membrane receptors.

      The nature of the transmitter at the afferent synapse of this system is not yet fully understood (for review, see Ref. 1, 9, 11). Our recent studies best support the hypothesis that an amino acid, most probably L-glutamate (L-GLU) or a substance of a similar structure, may be a chemical transmitter at the afferent synapse in ampullary electroreceptor organs of the catfish Ictalurus nebulosus.[4-6] Actually, it has been shown that L-GLU and its agonists kainate (KA), quisqualate (Q), and AMPA induced facilitation of the afferent discharges when applied externally to the inner aspect of the skin. The afferent synapse of these organs was equally sensitive to AMPA and Q with the effective concentrations of each agonist inducing  50%  of  the  maximal response (EC50) of about 7 microM and 9 microM, followed by KA with EC50 of about 25 microM and L-GLU V(0.6 mM). 6,7- Dinitroquinoxaline-2,3-dione (DNQX), being a very selective antagonist for AMPA receptors, blocked the excitatory effects of AMPA.

      N-methyl-D-aspartate (NMDA) was ineffective in these experiments even at concentrations as high as 5 mM. These studies raised a question about the presence of NMDA receptors in Ictalurus electroreceptor organs. However, the negative data on the effect of NMDA itself could not be a proof for the absence of NMDA receptors at this synapse. Pioneer experiments on the neurones of the central nervous system have shown that NMDA receptors were activated by strong or repetitive neuronal depolarization.10 They were inhibited even by the physiological concentrations of Mg2+ (1-2 mM).[7,15] On the other hand, glycine (GLY) potentiates the effects of NMDA at its receptor site.[13] Consistently, we have used an electrophysiological approach to study whether NMDA receptors are present and functional in Ictalurus electroreceptor organs.

      In these experiments, the effects of NMDA and its antagonists 7-chlorokynurenic acid (7ClKyn), DL-2-amino-5-phosphonovaleric acid (APV) and ketamine (Ket), on resting and electrically evoked activity were studied. In order to characterize the Mg2+-, GLY-, and potential-dependency, the experiments were repeated in Mg2+-free solutions as well as in solutions containing different concentrations of GLY and under strong sinusoidal stimulation of the sensory organs.

Experimental procedures

      Experiments were performed on freshwater catfish Ictalurus nebulosus, having a total length of 15-20 cm. After induction of anaesthesia with saffan (24 mg/kg, i.p.) the fish was fixed between two rubber clamps in a preparation tray and perfused through the mouth with water at room temperature. A piece of skin of about 20 mm in diameter was removed from the dorsal surface of the skull. The dissected piece of skin was mounted in a perfusion chamber with the external (mucosal) side upwards. The internal (serosal) surface of the skin was perfused with normal saline, and the mucosal surface was perfused with tap water.

      Recording of action potentials from the primary afferents in the isolated piece of skin was performed by tungsten wire electrodes with tip diameters less than 10 micro-m. The electrodes were positioned with the tip in the opening of an electroreceptor organ by means of an electrode holder in a micromanipulator. The discharge frequency was recorded continuously during the experiment by means of a rate meter and chart recorder. Electroreceptor organs in the isolated piece of skin were stimulated with sinusoidal currents. The latter were generated by a function generator. The stimulus current was applied via a silver electrode that made contact with the mucosal compartment of the preparation.

      To study the effects of drugs on the synaptic transmission, the serosal side of the skin preparation, i.e. the basal faces of the receptor cells, was exposed to drugs by constant perfusion of normal and test solutions through the lower compartment of the perfusion chamber. At the same time, the mucosal surface of the skin was perfused by a micropump with tap water. The normal solution had the following composition in mM: NaCl 167; KCl 5; CaCl2 3; MgCl2 1.5; Tris buffer 4; pH 7.2. Test solutions were applied by switching a 5-way stopcock which replaced the constant flow of normal solution with test solution. The flow rate of the perfusate was maintained about 1.5 ml/min. Times for complete wash out appear to be ca. 5 min. The compounds utilized were the following: N-methyl-D-aspartic acid, glycine, 7-chlorokynurenic acid, D-serine, DL-2-amino-5-phosphonovaleric acid, ketamine (All from Sigma).

      The techniques for excision of the skin, for applying drugs to the serosal surface of the skin, and for recording of action potentials from primary afferents are described in detail elsewhere.[4]

Results

      The afferent fibres innervating ampullary electroreceptor organs are spontaneously active with spike frequencies depending on temperature.[17] In the catfish, the spontaneous discharge frequency, referred to as the resting frequency, was on the average between 12 and 48 imp/s at bath temperatures between 18 and 22 oC. Sinusoidal electrical stimulation evoked spike trains, the frequency of which was modulated up and down in accordance with the polarity of the sine wave. Threshold intensities of electroreceptor organs depended on the frequency of stimulation and were lowest at stimulus frequencies of about 10 Hz.

      In the following, the effects of various substances on the afferent activity of single nerve fibres are described, since it is well documented that spike activity in the primary afferents is due to neurotransmitter release in the preceding synapse between receptor cells and primary afferents.[9]

Effects of NMDA in Mg2+-free solutions

      NMDA and Mg2+ interactions were studied by comparing the effects of NMDA in either Mg2+-containing or Mg2+-free solutions. In accordance with our recent data [5] NMDA, dissolved in the Mg2+-containing (normal) solution, had no effect, not even at 5 mM. The results of a typical experiment with a Mg2+-free solution are shown in Fig. 1A. As can be clearly seen, the beginning of perfusion with the Mg2+-free solution did not change firing in the afferent nerve fibre. However, addition of 5 mM NMDA to the Mg2+-free solution induced a pronounced and enduring increase in the firing rate and a sustained level of resting discharge. The excitatory effect of NMDA was moderate, and the frequency increase did not exceed 35% of the original level of the resting discharge. Interestingly, subsequent perfusion with a Mg2+-free solution alone did not restore firing frequency. On the contrary, the afferent fibre maintained its enhanced resting discharge at the sustained level during the period of application of the Mg2+-free solution. The original level of resting activity restored only after the start of perfusion with normal solution. In contrast to, as shown in Fig. 1B, if the Mg2+-free NMDA-containing solution was replaced directly by normal solution, the resting activity decreased its frequency just after the change of solutions, with a latency due diffusion.



Fig. 1. Effects of NMDA on excitation in afferent fibres under Mg2+-free conditions. The horizontal bars indicate the duration of drugs application. Original recordings of two experiments with 5 mM NMDA which raised the resting frequency. A. Mg2+-free solution did not stop the NMDA effect. It persisted during perfusion with the Mg2+-free solution. B. Only a normal solution restrained the frequency increase under NMDA. Ordinate: spike frequency (imp/s); abscissa: time (min.).



Fig. 2. Modulatory effects of sine wave electric stimulation on NMDA responses in the primary afferents. Original recordings of two experiments with electrical stimulation which enhanced the average discharge frequency only in the presence of NMDA. A. 9 Hz stimuli (70 nA/cm2) were delivered under NMDA and control conditions. B. NMDA was applied during constant background electric stimulation.

Electrical stimulation

      To determine whether NMDA responses in the primary afferents are modulated by repetitive stimulation, the applications of NMDA were combined with strong electric sine wave stimulation of the electroreceptor organs. As shown in Fig. 2A, 9 Hz trains were delivered firstly in the presence of NMDA and secondly under control conditions. NMDA added to Mg2+-containing solution did not impact meaningfully on the resting discharge of the afferent fibre. However, a relatively strong sinusoidal stimulation (70 nA/cm 2) evoked an increase in the firing rate and a tonic response with slow adaptation in the presence of 5 mM NMDA. Control sinusoidal stimulation with the same stimulus parameters in the end of the records evoked spike bursts but there was no any significant increase in average frequency of the afferent discharge. In the experiment shown in Fig. 2B, NMDA was applied during the constant background electric stimulation of the sense organs. In such a case, the electric stimulation itself, applied both prior and after NMDA application, did not evoke any changes in the average discharge level. However, under stimulation conditions addition of NMDA produced a clear-cut firing increase of the primary afferent fibre.

Influence of glycine on NMDA effects

      GLY and NMDA interactions were studied with 0.01-0.001 mM GLY using the Mg 2+-free solution as a vehicle. GLY itself had no influence on afferent spike frequency. In these experiments the preparation was at first perfused with 5 mM NMDA and the effect observed was compared with that in the presence of GLY. In these conditions, GLY significantly potentiated the responses evoked by administration of NMDA emphasizing the frequency increase and lengthening of their duration (Fig. 3). The statistical examination of the results of six experiments revealed that the potentiating effect of GLY on NMDA responses was 37.5(+ , - )5.0% of the control discharge.



Fig. 3. Potentiating effect of GLY on the responses evoked by application of NMDA in Mg2+-free solution. Time courses of the responses in this Fig. as well as in Figs. 5 and 6 are displayed in plots of spike frequency versus elapsed time. In these cases, spike frequency was determined over sequental 2 min epochs from the corresponding chart records. The plots represent consecutive records from the same preparation. The middle graph shows that addition of GLY (0.01 mM) significantly potentiated the responses to NMDA as compared to the control applications of NMDA (top and bottom graphs).

Effects of NMDA antagonists

      A variety of NMDA antagonists were effective at the afferent synapse of ampullary electroreceptor organs of the catfish. Among them were Ket, APV, and 7ClKyn. Addition of these antagonists in concentrations from 0.5 to 2.0 mM led to a decrease of both resting and stimulus-evoked activity which was sustained throughout the period of application. Effects of various antagonists were plotted as suppression ratio of the resting discharge (Fig. 4). At 1 mM the relative potencies of Ket, 7ClKyn and APV were 64.5(+, - )9.9, 30.5(+, - )11.3 and 24.0(+, -)4.6%, respectively (n=6-8).



Fig. 4. Dose-response relationships for NMDA antagonists Ket, APV and 7ClKyn. Drugs effects were plotted as suppression ratio of the resting discharge (%). Each point represents the mean + S.E.M. of at least 6 experiments.

      A question arose as to whether the NMDA-induced responses in Mg 2+ - free solution could be blocked with the aid of antagonists for NMDA. 7ClKyn was chosen from the NMDA antagonists as being the glycine site antagonist so far tested. The experiments showed that initial perfusion with Mg 2+ - free did not produce any changes in the resting discharge (Fig. 5). However, addition of 5 mM NMDA induced a marked increase in firing frequency and a sustained level of resting discharge. Subsequent application of 0.01 mM 7ClKyn to the NMDA-containing Mg 2+ - free solution strongly decreased the resting discharge frequency. Thus, 7ClKyn blocked the responses to NMDA completely. The effects observed were fully reversible following perfusion with drug free solution.

      To characterize NMDA receptors in ampullary electroreceptor organs, the antagonism of 7ClKyn and the GLY site agonist D-serine was used (Fig. 6). The experiments showed that 5 mM NMDA being applied to the Mg2+-free solution caused a clear-cut frequency increase. Subsequent perfusion with 0.01 mM 7ClKyn plus 0.01 mM D-serine did not change firing in the afferent nerve fibre.       Thus, application of D-serine prevented the effects of 7ClKyn (n=6) and restored the responses to NMDA.



Fig. 5. Inhibitory action of 7ClKyn on responses evoked by NMDA in Mg2+-free solution. One experiment is shown with 5 mM NMDA which increased the frequency of the resting discharge. Subsequent application of 1 mM 7ClKyn stopped the NMDA effect and decreased resting frequency. If a normal solution was applied resting frequency recovered with a considerable delay.



Fig. 6. Inhibitory action of 7Clkyn was reversed by D-serine. Excitatory effects of NMDA in Mg2+-free solution persisted, despite the presence of 7ClKyn if D-serine was added. Normal solution stopped the NMDA effect with a delay due to perfusion.

Discussion

In accordance with the results of our previous study,5 5 mM NMDA dissolved in the Mg 2+ -containing (normal) solution had no effect on resting activity in the afferent fibres of the electroreceptor organs. These data raised the question about the existence of NMDA receptor subtypes at this synapse. On the other hand, it is well known from the literature that responses elicited by NMDA in other neuronal structures appeared to be most susceptible to experimental manipulations and controlled stimulations (for review, see Ref. 1,10,19). For example, in the central nervous system, Mg2+ evidently blocked the NMDA ion channels in a voltage-dependent manner.[15] An increase of Mg2+-concentration in the bath solution from 1 to 3 mM drastically depressed the afferent discharge of the semicircular canals and the responses to application of NMDA.[18] In the afferent synapse of electroreceptor organs of the ampullae of Lorenzini, the excitatory responses to NMDA were depressed completely in solutions with elevated concentration of Mg2+.[2,3] Thus, if electroreceptor organs of the catfish are not unusual in pharmacology of their glutamate receptors, special experimental conditions should be applied todetect NMDA receptors.

      In our present experiments, addition of 5 mM NMDA to the Mg2+-free solution induced frequency increase that varied in the range of 25-35% of the resting discharge. Interestingly, that in contrast to responses to application of L-GLU, Q, KA and AMPA, which were characterized by a complete desensitization, addition of NMDA resulted in the frequency increase but responses were not desensitized even by a prolonged action of NMDA. Moreover, the sustained response under NMDA continued even after cessation of the perfusion with NMDA in the Mg2+-free solution.

      The blockade of NMDA responses by Mg2+ is now known to be maximal at negative potentials and to decrease with depolarization.[7] In the electroreceptor organs, the depolarization conditions may correspond to sine wave stimulation of high intensities. Accordingly, in our experiments applications of NMDA in Mg2+-containing solution evoked a considerable frequency increase in the afferent nerve under 9 Hz stimulation. Thus, it is conceivable that sine wave stimulation could alter the properties of membrane receptors or activate additional receptors. Similarly, it has been recently shown that strong and repetitive sound stimulation is required to activate the NMDA receptors in the cochlea16 (for review, see Ref.11). In those experiments, a significant reduction of the amplitude of the compound action potentials and an increase of the N1 latency were observed in the presence of NMDA only at high-intensity burst stimulation.

      Of particular interest in identifying NMDA membrane receptors were effects of specific NMDA-preferring antagonists. Among them were the competitive NMDA antagonist APV, the non-competitive antagonist Ket, and the glycine site antagonist 7ClKyn. These substances did not differ greatly in their ability to suppress firing of the afferent nerve fibres. All of them were effective in mM concentrations. The fact that NMDA antagonists of different mode of action were active at this synapse is an important argument in favour of the existing NMDA receptors in the sensory synapse of the electroreceptor organs.

       A remarkable property of the responses to NMDA observed in Mg2+-free solutions was that they were potentiated by low concentrations of GLY. The effects of GLY were abolished by a specific GLY site antagonist 7ClKyn. Moreover, the antagonistic action of 7ClKyn on the resting discharge and on the response evoked by NMDA was prevented by the GLY agonist D-serine. These results suggest that receptor cells release GLY in vivo that can enhance responses to NMDA and which can be washed away with perfusion in isolated preparation. The results can also account for the competitive action between 7Clkyn and D-serine indicative of GLY site binding as a possible requirement for NMDA receptor activation. This GLY dependency is a characteristic property of the NMDA receptors.[8,13] It is likely that in intact preparation the GLY site is fully occupied since in brain slices and in vivo, GLY and D-serine did not potentiate NMDA responses.[12,14] In contrast, artificial solution which is used in our experiment is a solution of ions without GLY. This fact could possibly explain the futile searching for NMDA receptors in our previous experiments.[5]

Conclusion

      The above body of data discussing the effects of NMDA and related substances has led us to propose that NMDA receptors are actually present at the afferent synapse of electroreceptor organs. It is only possible to speculate whether this subtype of membrane receptors is functional at this synapse because the experiments have shown that involvement of NMDA receptors is highly dependent upon the experimental conditions used. Among these important factors are: the extracellular Mg2+, the level of electrical stimulation of the sense organs, and the presence of the potentiating agent GLY. We have also found that non-NMDA receptors may mediate the synaptic effects of L-GLU at this synapse.[5,6] Thus, NMDA- and non-NMDA receptors may function as a dual receptor system. It is conceivable that non-NMDA receptors may be utilized during resting activity and at low levels of stimulation, whereas NMDA receptors are activated by high intensity stimulation. The activity-dependent process of activation of the dual receptor system may endow such important synaptic processes as synaptic plasticity, temporal integration and rhythmic firing.

Acknowledgements - This work was supported by the Netherlands Research Organization grant 713-008 (to R.C.P. and G.N.A.).
We gratefully recall Huub G. Gall©'s electrode donation.

References

1. Akoev G.N. and Andrianov G.N. (1993) Sensory Hair Cells. Synaptic Transmission, 194 p.p.Springer, Berlin etc.
2. Akoev G.N., Andrianov G.N., Bromm B. and Szabo T. (1991) Neuropharmacological analysis of transmission in afferent synapses of electroreceptors (ampullae of Lorenzini) in the marine skate, Raja clavata. J. Comp. Physiol. 168A, 639-646.
3. Akoev G.N., Andrianov G.N., Szabo T. and Bromm B. (1991) Effects of excitatory amino acid antagonists on synaptic transmission in the ampullae of Lorenzini of the skate, Raja clavata. J. Comp. Physiol. 158A, 647-652.
4. Andrianov G.N., Bretschneider F., Peters R.C. and Teunis P.M.F. (1992) In vitro electroreceptor organs for pharmacological studies. J. Neurosci. Meth. 44, 1-6.
5. Andrianov G.N., Bretschneider F. and Peters R.C. (1992) The actions of L-glutamate and its agonists on the ampullary electroreceptor organs of the catfish Ictalurus nebulosus. Comp. Biochem. Physiol. 103C, 65-71.
6. Andrianov G.N., Peters R.C. and Bretschneider F. (1994) Identification of AMPA receptors in catfish electroreceptor organs. NeuroReport 5, 1056-1058.
7. Ascher P. and Nowak L. (1988) The role of divalent cations in the N-methyl-D-aspartate responses of mouse central nervous system. J. Physiol. (Lond.) 399, 227-246.
8. Bertolino M., Vicini S., Mazzetta J. and Costa E. (1988) Phenciclidine and glycine modulate NMDA-activated high conductance cationic channels by acting at different sites. Neurosci. Lett. 84, 351-355.
9. Bretschneider F. and Peters R.C. (1992) Transduction and transmission in ampullary electroreceptors of catfish Comp. Biochem. Physiol. 103A, 245-252.
10. Collingridge G.L. and Lester R.A.J. (1989) Excitatory amino acid receptors in the vertebrate central nervous system. Pharmacol. Rev. 40, 143-210.
11. Eybalin M. (1993) Neurotransmitters and neuromodulators of the mammalian cochlea. Physiol. Rev. 73, 309-373.
12. Fletcher E.J. and Lodge D. (1988) Glycine reverses antagonism of N-methyl-D-aspartate (NMDA) by 1-hydroxy-3-aminopyrrolidone-2 (H-966) but not D-2-amino-5-phosphonovalerate (D-AP5) on rat cortical slices. Eur. J. Pharmacol. 151, 161-162.
13. Johnson J.W. and Ascher P. (1987) Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature 325, 529-531.
14. Kemp J.A., Foster A.C., Leeson P.D., Priestley T., Tridgett R., Iversen L.L. and Woodroof G.N. (1988) 7-Chlorokynurenic acid is a selective antagonist at the glycine modulatory site of the N-methyl-D-aspartate receptor complex. Proc. Natl. Acad. Sci. USA 85, 6547-6550.
15. Nowak L., Bregestowski H., Ascher P., Herbert A. and Prochianz Z. (1984) Magnesium gates glutamate-activated channels in mouse central neurons. Nature 307, 462-465.
16. Puel J.-L., Ladrech S., Chabert R., Pujol R. and Eybalin M. (1991) Electrophysiological evidence for the presence of NMDA receptors in the guinea pig cochlea. Hear. Res. 51, 255-264.
17. Sch¤fer K., Braun H.A., Bretschneider F., Teunis P.F.M. and Peters R.C. (1990) Ampullary electroreceptors in catfish (Teleostei): temperature dependence of stimulus transduction. Pflìgers Arch. 417, 100-105.
18. Soto E., Flores A., Erãstegui C. and Vega R. (1994) Evidence for NMDA receptor in the afferent synaptic transmission of the vestibular system. Brain Res. 633, 289-296. 19. Watkins J.C., Krogsgaard-Larsen P. and Honor© T. (1990) Structure-activity relationships in the development of excitatory amino acid receptor agonists and competitive antagonists. Trends Pharmacol. Sci. 11, 25-33.

---

Contact address: Andrianov Yu.N. Pavlov Institute of Physiology RAS,
nab. Makarova, 6 199034 St.Peterburg
E-mail: andryu@infran.ru

---

Previous page                                                 Home