- Email: [email protected]
Synaptic organization of inhibitory circuits in the pigeon's optic tectum
Brain Research, 365 (1986) 383-387
383
Elsevier BRE 21354
Synaptic organization of inhibitory circuits in the pigeon's optic tectum N. LERESCHE, O. HARDY, E. AUDINAT and D. JASSIK-GERSCHENFELD Laboratoire de Psychophysiologie Sensorielle, UniversitdPierreet Marie Curie, 75230Paris Cedex 05 (France)
(Accepted October 9th, 1985) Key words: inhibitory postsynaptic potential - - synaptic organization - - retinal afferent - - optic tectum
The synaptic organization of inhibitory systems in the pigeon's optic tectum was studied with intracellular recording techniques. An extrapolation procedure based on response latency was used to determine the synaptic delay of the postsynaptic potentials (PSPs) and the velocity of conduction of the associated retinal axons. Tectal cells receive mostly disynaptic, trisynaptic or polysynaptic inhibition from retinal ganglion cells. However, evidence was found which together with previous studies raised the possibility of the existence of a direct inhibitory retino-tectal path. Our present results also suggest that inhibition is transmitted from the retina to the tectal cells by way of both, feedforward and feedback pathways. The pigeon's optic tectum receives a major projection from the contralateral retina2, 3 (for review, see ref. 13). Previous electrophysiological studies on avian tectal neurons have dealt mainly with the properties of the visual receptive fields (for review, see ref. 8). More recently, in order to reach a better understanding of the neuronal activities in the pigeon's optic rectum, a series of experiments has been per, formed using intracellular recordings of the postsynaptic events following activation of the retinal fibers. In one such study 4 it was shown that stimulation of the contralateral optic nerve in the pigeon elicits in most tectal cells an inhibitory postsynaptic potential (IPSP) either alone or preceded by an excitatory postsynaptic potential (EPSP). Furthermore, in more recent studiesS, ~1 the synaptic organization of the excitatory pathways mediating excitation from the retina to the pigeon's tectal cells has also been analyzed. These results show that in the pigeon, excitation following optic nerve stimulation is transmitted to the efferent tectal cells by way of three different circuits involving mono-, di- or polysynaptic connections with the retinal endings. The aim of the present work was to extend this research to the analysis of the synaptic organization of the inhibitory systems in the pigeon's optic rectum. In this regard we
have determined the time involved in the synaptic transmission of the PSPs evoked in the tectal cells after activation of the retinal axons as well as the conduction velocity of the associated retinal fibers. The experiments were performed on adult pigeons (Red Carneau). The surgical procedures were carfled out under general anesthesia (Hemineurine, 10 mg/kg). After tracheotomy and wing vein cannulation the animals were placed in a head-holder. The bone covering the lateral side of the left optic tectum was drilled open and the dura was removed. The right eye was opened by an equatorial cut and the vitreous was removed by suction to expose the optic papilla. Pigeons were then paralyzed with a continuous intravenous infusion of Flaxedil (14 mg/ml; 0.95 ml/h) and artificially ventilated by unidirectional airflow9. Special care was taken to block conduction of nociceptive impulses by using local anesthetics at all pressure points and wounds. The body temperature was maintained at 41 °C, using an electric heating pad. Glass micropipettes filled with 3 M KC1 and having resistances of 30-40 M f2 were used for recording. All electrophysiological data and control signals were stored on magnetic tape. Latency measurements of PSPs were performed off-line on a small computer (bin-width 0.02-0.03 ms). Usually, corn-
Correspondence: D. Jassik-Gerschenfeld, Laboratoire de Psychophysiologie Sensorielle, Universit6 Pierre et Marie Curie, 4 Place Jussieu, 75230 Paris Cedex 05, France.
0006-8993/86/$03.50 t~) 1986 Elsevier Science Publishers B.V. (Biomedical Division)
384 puter-averaged PSPs (6-10 trials) were used to assess the response latency. Penetrations were always made in the freely accessible lateral part of the left optic tectum. Stimulation of the optic nerve was performed by means of a pair of electrodes (insulated tungsten wires with tips bared 0.5 mm) inserted into the right optic papilla under microscopical control, in addition, a concentric stimulation electrode was stereotaxically l0 introduced in the left optic tract, just behind the optic chiasm. Positioning of the stimulation electrodes was adjusted to produce tectal field potentials at low threshold. Stimuli applied were rectangular pulses, 50-100 #s in duration, with intensities usually set at twice threshold (varying in the range 1-3 mA). The stimulation site in the optic tract was marked with electrolytic lesions by passing anodal currents through the electrode. At the end of the experiments the brain was removed from the skull and the distance between the optic papilla and the site of penetration in the optic tectum was measured. It was found to be 20 mm in all animals. The distance between the optic tract stimulation and the optic tecturn, estimated from serial frontal sections of the brain, was 10 + 0.5 mm. Intracellular recordings were obtained from 44 tectal cells. The cell's membrane potentials ranged between -35 and -65 mV. Twenty-three neurons responded to the stimulation of the retinal fibers with an EPSP which usually did not give rise to action potentials, followed by an IPSP of 4 - 1 0 mV amplitude and 10-40 ms duration. Fig. 1A, C illustrates the response from one of these cells to stimulation of the right optic nerve and of the left optic tract, respectively. Another group of 7 cells exhibited a pure IPSP of characteristics similar to those described above. Finally, another 14 cells exhibited a pure EPSP. Only neurons responding with an IPSP either alone or preceded by an EPSP, were included in this study. Latency measurements were used to determine the synaptic transmission time or synaptic delay of the PSPs elicited in the tectal cells by the activation of the retinal fibers. In the cells exhibiting an E P S P IPSP sequence the IPSP may be superimposed in the preceding EPSP. For this reason, latency measurements of IPSPs were obtained in 15 of these cells in which we were able to reduce the amplitude of the preceding EPSP with depolarizing intracellular current injections (Fig. 1D) or more frequently to re-
Fig. 1. Intracellular recordings from a tectal cell responding to electrical stimulation of retinal axons with an IPSP preceded by an EPSP. A and B show responses following stimulation of the right optic nerve. In B the polarity of the IPSP was reversed by intracetlularly injected hyperpolariZing currents. C and D show responses following stimulation of the left optic tract. The records in D were taken with the cell depolarized by current injected into the cell. Membrane potential -60 inV. Horizontal scale = 3 ms in all records. Vertical scale = 2 mV in A and C, and 5 mV in B and D. A-C: superimposed traces. EPSP latency to optic nerve and to optic tract 1.5 and 1 ms, respectively. IPSP latency to optic nerve and to optic tract 5.3 and 3.4 ms, respectively.
verse the polarity of the IPSP. The polarity of the IPSP was reversed by hyperpolarizing currents or by leakage of Cl-ions out of the recording electrode on termination of depolarizing currents. A n example of an IPSP reversed by hyperpolarizing currents is given in Fig. 1B; the inflection on the rising phase of the PSP indicates the onset of the reversed IPSP. The distribution of the IPSPs' latencies measured in the cells exhibiting an E P S P - I P S P sequence, is shown in Fig. 2 (unshaded histograms); they ranged from 2.5 to 6:4 ms (average 5 ms) for the optic nerve and from 1.8 to 5.6 ms (average 3.2 ms) for the optic tract. Latency measurements of the EPSPs were also obtained in such cells. As is shown in Fig, 2 (dotted histograms), optic nerve stimulation elicited EPSPs at latencies
385
I PSP
•
i
1
,
i
i
i 3
i
i
.
EPSP
i
5
,
i
,
i
•
m~
I
4"
4
2
2
1
S m~
OT
N~ OT
kolency
3
3
5
ms'
LaLency
1
3
5
elated retino-tectal fibers were obtained from the reciprocal of the slope of the EPSP and IPSP regression lines. Fig. 3 gives examples of the data estimated with this procedure in 8 different tectal cells. Fig. 3A illustrates measurements for the IPSP. Fig. 3B, C illustrates measurements obtained for both the EPSP and the IPSP from 2 individual neurons responding with an E P S P - I P S P sequence; data for the cell in C were obtained from the records in Fig. 1. Estimation of the synaptic delay for the IPSP indicates that almost all of them were conducted through inhibitory circuits involving one or more interneurones. Indeed, they ranged in almost all the cells from 1.2 to 4.5 ms. There was, however, one exception from a cell ex-
ms
Fig. 2. Latency distribution of PSPs following electrical stimulation of the right optic nerve (ON) and of the left optic tract (OT). Left columnn histograms show latency distribution of IPSPs which were recorded from cells exhibiting a pure IPSP (hatched histograms) or an IPSP preceded by an EPSP (unshaded histograms). Dotted histograms at the right part of the figure show the latency distribution of EPSPs recorded from cells responding with an EPSP-IPSP sequence.
ranging from 1.5 to 5 ms (average 2.7 ms) whereas the EPSPs elicited by optic tract stimulation ranged from 1 to 4.5 ms (average 2.2 ms). The distribution of the latency of the IPSPs in the cells responding with a pure IPSP (Fig. 2, hatched histograms) ranged from 3 to 8.4 ms (average 5.6 ms) for optic nerve stimulation and from 2.8 to 5.3 ms (average 3.8 ms) for optic tract stimulation. The variability of the PSPs' latencies was very low (+ 0.02 ms) in the case of neurones exhibiting PSPs at latencies shorter than 4 ms. In the remaining neurones the PSPs' latencies were still fairly constant for a given neurone, although the variability was greater (from + 0.05 ms to + 0.2 ms). The synaptic delay of the PSPs was estimated by a classical extrapolation procedure based on response latency and conduction distance relationships. The latency of the responses following right optic nerve and left optic tract stimulation was plotted as a function of the distance between the stimulating and the recording sites. On the assumption of a constant conduction velocity regression lines were drawn through EPSP and through IPSP latencies and the synaptic transmission time was extrapolated from the interception of each regression line with the ordinates. Measurements of the conduction velocity of asso-
b
F
o"
OT
Distance (ram)
Fig. 3. Measurements of PSPs' synaptic delays and conduction velocities of associated retinal fibers estimated by plotting the PSPs latencies to right optic nerve (ON) and left optic tract (OT) stimulation as a function of the conduction distances (see text for more details). Distance between ON and OT stimulation sites and the recording site were respectively 20 mm and 10 mm. A: synaptic delays of IPSPs measured in 6 tectal cells (1.8
msin a, 4 ms in b, 0.6 ms in c, 1.4 ms in d, 2.1 ms in e and 1.2 ms in f). Velocity of conduction: 3 m/s in a, c; 6 m/s in b, d; 15 mJs in f; 17 m/s in e. B, C: synaptic delays of EPSP and IPSP from 2 cells responding with an EPSP-IPSP sequence. Light and filled circles represent latencies of EPSPs and IPSPs, respectively. The data in C were obtained from the cell in Fig. 1. Note that in both cells the IPSP was disynapti¢ (1.6 ms in B and C) and the EPSP monosynaptic (0.7 ms in B and C). Velocity of conduction of retinal fibers mediating the EPSP and the IPSP was roughly the same in B (10 m/s) whereas in C the IPSP was conducted by fibers having slower velocity of conduction (5 m/s) than those mediating the EPSP (20 m/s).
386 hibiting a pure IPSP whose synaptic delay was of 0.6 ms (Fig. 3Ac) , such as could be expected for a monosynaptic link. Since the intercept, from which the synaptic delay is obtained, could be very sensitive to variability in the latencies data, it is important to point out that for this particular cell the variability of the IPSP's latency to optic nerve and to optic tract stimulation was of +0.05 ms and +0.02 ms, respectively; that is to say too low to introduce a significant error in the determination of the synaptic delay. It should be added that the same was true for all the neurones presented here. The conduction velocity of the retinal fibers mediating the IPSP estimated from similar plots as those shown in Fig. 3, ranged from 3 to 25 m/s (average 6 m/s). Another interesting point in our study arises by comparing the results for the EPSP and for the IPSP, in each of the cells responding with an EPSP-IPSP sequence. In 9 of these cells estimation of synaptic delays for the EPSPs showed that they were in a range from 0.5 to 0.8 ms (average 0.6 ms), in keeping with a monosynaptic connection with the retinal fibers (see Fig. 3B, C). The synaptic delay for the IPSP in 6 of such cells was about 0.9 ms longer (ranging from 1.4 to 1.8 ms) than the one for the EPSP, indicating an additional synapse in the inhibitory path. Examples of this result are given in Fig. 3B, C for two different cells; for both cells the synaptic delays for the EPSP and for the IPSP were, respectively, 0.7 and 1.6 ms. In the remaining 3 cells which received a monosynaptic EPSP, the synaptie delay of the IPSP was found to be longer than 2 ms indicating that inhibition was transmitted by way of tri- or polysynaptic connections. Taken together these results indicate that the IPSP in these tectal cells is mediated by di-, tri- or polysynaptic paths in contrast to the monosynaptic link for the EPSP. Estimation of synaptic delay of the EPSP in the remaining 4 cells exhibiting an EPSP-IPSP sequence, showed that they were mediated through circuits involving one or more interneurones (range 1.4-4 ms). Among them, 3 cells exhibited synaptic delays for the IPSP which were longer than those for the EPSPs whereas in 1 cell they were about the same. In other words, in the former case the IPSP was mediated via pathways involving more synapses than the excitatory path while in the second case excitation and inhibition were trans-
mitted to the same cell through paths involving a sin> ilar number of synapses. As is also shown in Fig. 3B, the slopes of the EPSP and the IPSP lines for this neuron were roughly the same indicating that the EPSP and the IPSP are mediated by retinal fibers having about the same conduction velocity. Similar results were obtained from another 8 cells. On the contrary, the slopes of the EPSP and the IPSP lines were rather different in the case of the cell illustrated in Fig. 3C, indicating that the EPSP and the IPSP were mediated by fibers having different conduction velocities. Similar results were obtained in a total of 5 cells. It was a common observation that the EPSP from such cells was mediated by fibers conducting at faster velocities (from 20 to 25 m/s) than those mediating the IPSP (from 3 to 6 m/s). For example, in the case of the celt in Fig. 3C, the retinal fibers responsible for the EPSP had a conduction velocity of 20 m/s while those responsible for the IPSP had a conduction velocity of 5 m/s. In conclusion, the results presented here demonstrate that inhibition following retinal activation is transmitted to the tectal cells mainly by way of di-, tri- or polysynaptic circuits. However, there was an exception to this rule from a cell in which the estimated synaptic delay of the IPSP was in a monosynaptic range (0.6 ms). Our present data also show that such result cannot be attributed to variability in the IPSP's latencies. There is, however, another way in which an erroneous reading of the synaptic delay can be produced if there are two paths mediating inhibition, one relatively faster than the other. In such a case a spuriously small synaptic delay will be obtained if the optic tract electrode excites the fast input and the optic nerve electrode excites the slow. A variant of this will occur if the optic nerve electrode excites an indirect path while the optic tract electrode activates a relatively direct path. It is evident that such possibilities should be tested in future experiments. Nevertheless, it is interesting to point out that in a previous study 5 based on intracellular recording and H R P labeling of individual tectal cells, we have shown that IPSPs able to follow high frequency rates of optic nerve stimulation without changes either in latency or in amplitude, could be recorded from neurons located in the region of optic fiber endings2,3,6,7,12 of the pigeon's optic tectum. This finding and our present observation raise the possibility of
387 the existence of a direct inhibitory path from the retina to the optic tectum. A n o t h e r interesting result from our research arises from the study of cells exhibiting an E P S P - I P S P sequence. The data from these cells show that excitation and inhibition could be transmitted to the same tectal neuron by retinal axons having either different or similar velocities of conduction. In the former case, it is most likely that IPSPs are mediated through one or more interneurones in a feedforward system, since in such cells EPSPs and IPSPs are conveyed by way of different populations of retinal fibers. On the contrary, in the second case EPSP and IPSP are mediated to the same cell by the same population of retinal fibers. One possibility is that in such cells the
IPSP is conveyed by way of recurrent axon collaterals and interneurones in a feedback system. Although, this last suggestion requires further confirmation it is interesting to point out that histological studies 1 have revealed that many tectal cells have recurrent axon collaterals which terminate within the region of the cell body. Thus, this histological observation suggests that the circuitry needed to support a recurrent inhibitory system is available in the pigeon's optic tectum.
1 Angaut, P. and Reperant, J., Fine structure of the optic fibre layers in the pigeon optic tectum: a Golgi and electromicroscope study, Neuroscience, 1 (1976) 93-105. 2 Cajal, S., Ramon y, Histologie du Systdme Nerveux de l'Homme et des Vertebras, Tome 2, Ch. X, Maloine, Paris, 1911. 3 Cowan, W.M., Adamson, L. and Powell, T.P.S., An experimental study of the avian visual system, J. Anat., 95 (1961) 545-563. 4 Hardy, O., Leresche, N. and Jassik-Gerschenfeld, D., Postsynaptic potentials in neurons of the pigeon's optic rectum in response to afferent stimulation from the retina and other visual structures: an intracellular study, Brain Research, 311 (1984) 65-74. 5 Hardy, O., Leresche, N. and Jassik-Gerschenfeld, D., Morphology and laminar distribution of electrophysioiogically identified cells in the pigeon's optic tectum: an intracellular study, J. Comp. Neurol., 233 (1985) 390-404. 6 Hayes, B.P. and Webster, K.E., An electron microscope study of the retino-receptive layers of the pigeon optic tectum, J. Comp. Neurol., 162 (1975) 447-465. 7 Hunt, S.P. and Webster, K.E., The projection of the retina upon the optic tectum of the pigeon, J. Comp. Neurol., 162 (1975) 433-446.
8 Jassik-Gerschenfeld, D. and Hardy, O., The avian optic tectum: Neurophysiology and Behavioral Correlations. In H. Vanegas (Ed.), Comparative Neurology of the Optic Tectum, Plenum Press, New York and London, 1984, pp. 649-686. 9 Jassik-Gerschenfeld, D., Minois, F. and Conde-Courtine, F., Receptive field properties of directionally selective units in the pigeon's optic tectum, Brain Research, 24 (1970) 407-421. 10 Karten, H.J. and Hodos, W., A Stereotaxic Atlas of the Brain of the Pigeon (Columba Livia), The Johns Hopkins Press, Baltimore, MD, 1967. 11 Leresche, N., Hardy, O., Audinat, E. and Jassik-Gerschenfeld, D., Synaptic transmission of excitation from the retina to cells in the pigeon's optic tectum, Brain Research, in press. 12 Reperant, P. and Angaut, P., The retinotectal projections in the pigeon. An experimental optical and electron microscope study, Neuroscience, 2 (1977) 119-140. 13 Webster, K.E., Changing concepts of the central visual pathways in birds. In R. Bellairs and E.G. Gray (Eds.), Essays on the Nervous System, Claredon Press, Oxford, 1974, pp. 258-298.
The authors are greatly indebted to A. Dalbera for her excellent technical assistance. This research was supported by CNRS ( U A 662), I N S E R M (no. 84 60 13) and M R T (no. 84.C.1309).
383
Elsevier BRE 21354
Synaptic organization of inhibitory circuits in the pigeon's optic tectum N. LERESCHE, O. HARDY, E. AUDINAT and D. JASSIK-GERSCHENFELD Laboratoire de Psychophysiologie Sensorielle, UniversitdPierreet Marie Curie, 75230Paris Cedex 05 (France)
(Accepted October 9th, 1985) Key words: inhibitory postsynaptic potential - - synaptic organization - - retinal afferent - - optic tectum
The synaptic organization of inhibitory systems in the pigeon's optic tectum was studied with intracellular recording techniques. An extrapolation procedure based on response latency was used to determine the synaptic delay of the postsynaptic potentials (PSPs) and the velocity of conduction of the associated retinal axons. Tectal cells receive mostly disynaptic, trisynaptic or polysynaptic inhibition from retinal ganglion cells. However, evidence was found which together with previous studies raised the possibility of the existence of a direct inhibitory retino-tectal path. Our present results also suggest that inhibition is transmitted from the retina to the tectal cells by way of both, feedforward and feedback pathways. The pigeon's optic tectum receives a major projection from the contralateral retina2, 3 (for review, see ref. 13). Previous electrophysiological studies on avian tectal neurons have dealt mainly with the properties of the visual receptive fields (for review, see ref. 8). More recently, in order to reach a better understanding of the neuronal activities in the pigeon's optic rectum, a series of experiments has been per, formed using intracellular recordings of the postsynaptic events following activation of the retinal fibers. In one such study 4 it was shown that stimulation of the contralateral optic nerve in the pigeon elicits in most tectal cells an inhibitory postsynaptic potential (IPSP) either alone or preceded by an excitatory postsynaptic potential (EPSP). Furthermore, in more recent studiesS, ~1 the synaptic organization of the excitatory pathways mediating excitation from the retina to the pigeon's tectal cells has also been analyzed. These results show that in the pigeon, excitation following optic nerve stimulation is transmitted to the efferent tectal cells by way of three different circuits involving mono-, di- or polysynaptic connections with the retinal endings. The aim of the present work was to extend this research to the analysis of the synaptic organization of the inhibitory systems in the pigeon's optic rectum. In this regard we
have determined the time involved in the synaptic transmission of the PSPs evoked in the tectal cells after activation of the retinal axons as well as the conduction velocity of the associated retinal fibers. The experiments were performed on adult pigeons (Red Carneau). The surgical procedures were carfled out under general anesthesia (Hemineurine, 10 mg/kg). After tracheotomy and wing vein cannulation the animals were placed in a head-holder. The bone covering the lateral side of the left optic tectum was drilled open and the dura was removed. The right eye was opened by an equatorial cut and the vitreous was removed by suction to expose the optic papilla. Pigeons were then paralyzed with a continuous intravenous infusion of Flaxedil (14 mg/ml; 0.95 ml/h) and artificially ventilated by unidirectional airflow9. Special care was taken to block conduction of nociceptive impulses by using local anesthetics at all pressure points and wounds. The body temperature was maintained at 41 °C, using an electric heating pad. Glass micropipettes filled with 3 M KC1 and having resistances of 30-40 M f2 were used for recording. All electrophysiological data and control signals were stored on magnetic tape. Latency measurements of PSPs were performed off-line on a small computer (bin-width 0.02-0.03 ms). Usually, corn-
Correspondence: D. Jassik-Gerschenfeld, Laboratoire de Psychophysiologie Sensorielle, Universit6 Pierre et Marie Curie, 4 Place Jussieu, 75230 Paris Cedex 05, France.
0006-8993/86/$03.50 t~) 1986 Elsevier Science Publishers B.V. (Biomedical Division)
384 puter-averaged PSPs (6-10 trials) were used to assess the response latency. Penetrations were always made in the freely accessible lateral part of the left optic tectum. Stimulation of the optic nerve was performed by means of a pair of electrodes (insulated tungsten wires with tips bared 0.5 mm) inserted into the right optic papilla under microscopical control, in addition, a concentric stimulation electrode was stereotaxically l0 introduced in the left optic tract, just behind the optic chiasm. Positioning of the stimulation electrodes was adjusted to produce tectal field potentials at low threshold. Stimuli applied were rectangular pulses, 50-100 #s in duration, with intensities usually set at twice threshold (varying in the range 1-3 mA). The stimulation site in the optic tract was marked with electrolytic lesions by passing anodal currents through the electrode. At the end of the experiments the brain was removed from the skull and the distance between the optic papilla and the site of penetration in the optic tectum was measured. It was found to be 20 mm in all animals. The distance between the optic tract stimulation and the optic tecturn, estimated from serial frontal sections of the brain, was 10 + 0.5 mm. Intracellular recordings were obtained from 44 tectal cells. The cell's membrane potentials ranged between -35 and -65 mV. Twenty-three neurons responded to the stimulation of the retinal fibers with an EPSP which usually did not give rise to action potentials, followed by an IPSP of 4 - 1 0 mV amplitude and 10-40 ms duration. Fig. 1A, C illustrates the response from one of these cells to stimulation of the right optic nerve and of the left optic tract, respectively. Another group of 7 cells exhibited a pure IPSP of characteristics similar to those described above. Finally, another 14 cells exhibited a pure EPSP. Only neurons responding with an IPSP either alone or preceded by an EPSP, were included in this study. Latency measurements were used to determine the synaptic transmission time or synaptic delay of the PSPs elicited in the tectal cells by the activation of the retinal fibers. In the cells exhibiting an E P S P IPSP sequence the IPSP may be superimposed in the preceding EPSP. For this reason, latency measurements of IPSPs were obtained in 15 of these cells in which we were able to reduce the amplitude of the preceding EPSP with depolarizing intracellular current injections (Fig. 1D) or more frequently to re-
Fig. 1. Intracellular recordings from a tectal cell responding to electrical stimulation of retinal axons with an IPSP preceded by an EPSP. A and B show responses following stimulation of the right optic nerve. In B the polarity of the IPSP was reversed by intracetlularly injected hyperpolariZing currents. C and D show responses following stimulation of the left optic tract. The records in D were taken with the cell depolarized by current injected into the cell. Membrane potential -60 inV. Horizontal scale = 3 ms in all records. Vertical scale = 2 mV in A and C, and 5 mV in B and D. A-C: superimposed traces. EPSP latency to optic nerve and to optic tract 1.5 and 1 ms, respectively. IPSP latency to optic nerve and to optic tract 5.3 and 3.4 ms, respectively.
verse the polarity of the IPSP. The polarity of the IPSP was reversed by hyperpolarizing currents or by leakage of Cl-ions out of the recording electrode on termination of depolarizing currents. A n example of an IPSP reversed by hyperpolarizing currents is given in Fig. 1B; the inflection on the rising phase of the PSP indicates the onset of the reversed IPSP. The distribution of the IPSPs' latencies measured in the cells exhibiting an E P S P - I P S P sequence, is shown in Fig. 2 (unshaded histograms); they ranged from 2.5 to 6:4 ms (average 5 ms) for the optic nerve and from 1.8 to 5.6 ms (average 3.2 ms) for the optic tract. Latency measurements of the EPSPs were also obtained in such cells. As is shown in Fig, 2 (dotted histograms), optic nerve stimulation elicited EPSPs at latencies
385
I PSP
•
i
1
,
i
i
i 3
i
i
.
EPSP
i
5
,
i
,
i
•
m~
I
4"
4
2
2
1
S m~
OT
N~ OT
kolency
3
3
5
ms'
LaLency
1
3
5
elated retino-tectal fibers were obtained from the reciprocal of the slope of the EPSP and IPSP regression lines. Fig. 3 gives examples of the data estimated with this procedure in 8 different tectal cells. Fig. 3A illustrates measurements for the IPSP. Fig. 3B, C illustrates measurements obtained for both the EPSP and the IPSP from 2 individual neurons responding with an E P S P - I P S P sequence; data for the cell in C were obtained from the records in Fig. 1. Estimation of the synaptic delay for the IPSP indicates that almost all of them were conducted through inhibitory circuits involving one or more interneurones. Indeed, they ranged in almost all the cells from 1.2 to 4.5 ms. There was, however, one exception from a cell ex-
ms
Fig. 2. Latency distribution of PSPs following electrical stimulation of the right optic nerve (ON) and of the left optic tract (OT). Left columnn histograms show latency distribution of IPSPs which were recorded from cells exhibiting a pure IPSP (hatched histograms) or an IPSP preceded by an EPSP (unshaded histograms). Dotted histograms at the right part of the figure show the latency distribution of EPSPs recorded from cells responding with an EPSP-IPSP sequence.
ranging from 1.5 to 5 ms (average 2.7 ms) whereas the EPSPs elicited by optic tract stimulation ranged from 1 to 4.5 ms (average 2.2 ms). The distribution of the latency of the IPSPs in the cells responding with a pure IPSP (Fig. 2, hatched histograms) ranged from 3 to 8.4 ms (average 5.6 ms) for optic nerve stimulation and from 2.8 to 5.3 ms (average 3.8 ms) for optic tract stimulation. The variability of the PSPs' latencies was very low (+ 0.02 ms) in the case of neurones exhibiting PSPs at latencies shorter than 4 ms. In the remaining neurones the PSPs' latencies were still fairly constant for a given neurone, although the variability was greater (from + 0.05 ms to + 0.2 ms). The synaptic delay of the PSPs was estimated by a classical extrapolation procedure based on response latency and conduction distance relationships. The latency of the responses following right optic nerve and left optic tract stimulation was plotted as a function of the distance between the stimulating and the recording sites. On the assumption of a constant conduction velocity regression lines were drawn through EPSP and through IPSP latencies and the synaptic transmission time was extrapolated from the interception of each regression line with the ordinates. Measurements of the conduction velocity of asso-
b
F
o"
OT
Distance (ram)
Fig. 3. Measurements of PSPs' synaptic delays and conduction velocities of associated retinal fibers estimated by plotting the PSPs latencies to right optic nerve (ON) and left optic tract (OT) stimulation as a function of the conduction distances (see text for more details). Distance between ON and OT stimulation sites and the recording site were respectively 20 mm and 10 mm. A: synaptic delays of IPSPs measured in 6 tectal cells (1.8
msin a, 4 ms in b, 0.6 ms in c, 1.4 ms in d, 2.1 ms in e and 1.2 ms in f). Velocity of conduction: 3 m/s in a, c; 6 m/s in b, d; 15 mJs in f; 17 m/s in e. B, C: synaptic delays of EPSP and IPSP from 2 cells responding with an EPSP-IPSP sequence. Light and filled circles represent latencies of EPSPs and IPSPs, respectively. The data in C were obtained from the cell in Fig. 1. Note that in both cells the IPSP was disynapti¢ (1.6 ms in B and C) and the EPSP monosynaptic (0.7 ms in B and C). Velocity of conduction of retinal fibers mediating the EPSP and the IPSP was roughly the same in B (10 m/s) whereas in C the IPSP was conducted by fibers having slower velocity of conduction (5 m/s) than those mediating the EPSP (20 m/s).
386 hibiting a pure IPSP whose synaptic delay was of 0.6 ms (Fig. 3Ac) , such as could be expected for a monosynaptic link. Since the intercept, from which the synaptic delay is obtained, could be very sensitive to variability in the latencies data, it is important to point out that for this particular cell the variability of the IPSP's latency to optic nerve and to optic tract stimulation was of +0.05 ms and +0.02 ms, respectively; that is to say too low to introduce a significant error in the determination of the synaptic delay. It should be added that the same was true for all the neurones presented here. The conduction velocity of the retinal fibers mediating the IPSP estimated from similar plots as those shown in Fig. 3, ranged from 3 to 25 m/s (average 6 m/s). Another interesting point in our study arises by comparing the results for the EPSP and for the IPSP, in each of the cells responding with an EPSP-IPSP sequence. In 9 of these cells estimation of synaptic delays for the EPSPs showed that they were in a range from 0.5 to 0.8 ms (average 0.6 ms), in keeping with a monosynaptic connection with the retinal fibers (see Fig. 3B, C). The synaptic delay for the IPSP in 6 of such cells was about 0.9 ms longer (ranging from 1.4 to 1.8 ms) than the one for the EPSP, indicating an additional synapse in the inhibitory path. Examples of this result are given in Fig. 3B, C for two different cells; for both cells the synaptic delays for the EPSP and for the IPSP were, respectively, 0.7 and 1.6 ms. In the remaining 3 cells which received a monosynaptic EPSP, the synaptie delay of the IPSP was found to be longer than 2 ms indicating that inhibition was transmitted by way of tri- or polysynaptic connections. Taken together these results indicate that the IPSP in these tectal cells is mediated by di-, tri- or polysynaptic paths in contrast to the monosynaptic link for the EPSP. Estimation of synaptic delay of the EPSP in the remaining 4 cells exhibiting an EPSP-IPSP sequence, showed that they were mediated through circuits involving one or more interneurones (range 1.4-4 ms). Among them, 3 cells exhibited synaptic delays for the IPSP which were longer than those for the EPSPs whereas in 1 cell they were about the same. In other words, in the former case the IPSP was mediated via pathways involving more synapses than the excitatory path while in the second case excitation and inhibition were trans-
mitted to the same cell through paths involving a sin> ilar number of synapses. As is also shown in Fig. 3B, the slopes of the EPSP and the IPSP lines for this neuron were roughly the same indicating that the EPSP and the IPSP are mediated by retinal fibers having about the same conduction velocity. Similar results were obtained from another 8 cells. On the contrary, the slopes of the EPSP and the IPSP lines were rather different in the case of the cell illustrated in Fig. 3C, indicating that the EPSP and the IPSP were mediated by fibers having different conduction velocities. Similar results were obtained in a total of 5 cells. It was a common observation that the EPSP from such cells was mediated by fibers conducting at faster velocities (from 20 to 25 m/s) than those mediating the IPSP (from 3 to 6 m/s). For example, in the case of the celt in Fig. 3C, the retinal fibers responsible for the EPSP had a conduction velocity of 20 m/s while those responsible for the IPSP had a conduction velocity of 5 m/s. In conclusion, the results presented here demonstrate that inhibition following retinal activation is transmitted to the tectal cells mainly by way of di-, tri- or polysynaptic circuits. However, there was an exception to this rule from a cell in which the estimated synaptic delay of the IPSP was in a monosynaptic range (0.6 ms). Our present data also show that such result cannot be attributed to variability in the IPSP's latencies. There is, however, another way in which an erroneous reading of the synaptic delay can be produced if there are two paths mediating inhibition, one relatively faster than the other. In such a case a spuriously small synaptic delay will be obtained if the optic tract electrode excites the fast input and the optic nerve electrode excites the slow. A variant of this will occur if the optic nerve electrode excites an indirect path while the optic tract electrode activates a relatively direct path. It is evident that such possibilities should be tested in future experiments. Nevertheless, it is interesting to point out that in a previous study 5 based on intracellular recording and H R P labeling of individual tectal cells, we have shown that IPSPs able to follow high frequency rates of optic nerve stimulation without changes either in latency or in amplitude, could be recorded from neurons located in the region of optic fiber endings2,3,6,7,12 of the pigeon's optic tectum. This finding and our present observation raise the possibility of
387 the existence of a direct inhibitory path from the retina to the optic tectum. A n o t h e r interesting result from our research arises from the study of cells exhibiting an E P S P - I P S P sequence. The data from these cells show that excitation and inhibition could be transmitted to the same tectal neuron by retinal axons having either different or similar velocities of conduction. In the former case, it is most likely that IPSPs are mediated through one or more interneurones in a feedforward system, since in such cells EPSPs and IPSPs are conveyed by way of different populations of retinal fibers. On the contrary, in the second case EPSP and IPSP are mediated to the same cell by the same population of retinal fibers. One possibility is that in such cells the
IPSP is conveyed by way of recurrent axon collaterals and interneurones in a feedback system. Although, this last suggestion requires further confirmation it is interesting to point out that histological studies 1 have revealed that many tectal cells have recurrent axon collaterals which terminate within the region of the cell body. Thus, this histological observation suggests that the circuitry needed to support a recurrent inhibitory system is available in the pigeon's optic tectum.
1 Angaut, P. and Reperant, J., Fine structure of the optic fibre layers in the pigeon optic tectum: a Golgi and electromicroscope study, Neuroscience, 1 (1976) 93-105. 2 Cajal, S., Ramon y, Histologie du Systdme Nerveux de l'Homme et des Vertebras, Tome 2, Ch. X, Maloine, Paris, 1911. 3 Cowan, W.M., Adamson, L. and Powell, T.P.S., An experimental study of the avian visual system, J. Anat., 95 (1961) 545-563. 4 Hardy, O., Leresche, N. and Jassik-Gerschenfeld, D., Postsynaptic potentials in neurons of the pigeon's optic rectum in response to afferent stimulation from the retina and other visual structures: an intracellular study, Brain Research, 311 (1984) 65-74. 5 Hardy, O., Leresche, N. and Jassik-Gerschenfeld, D., Morphology and laminar distribution of electrophysioiogically identified cells in the pigeon's optic tectum: an intracellular study, J. Comp. Neurol., 233 (1985) 390-404. 6 Hayes, B.P. and Webster, K.E., An electron microscope study of the retino-receptive layers of the pigeon optic tectum, J. Comp. Neurol., 162 (1975) 447-465. 7 Hunt, S.P. and Webster, K.E., The projection of the retina upon the optic tectum of the pigeon, J. Comp. Neurol., 162 (1975) 433-446.
8 Jassik-Gerschenfeld, D. and Hardy, O., The avian optic tectum: Neurophysiology and Behavioral Correlations. In H. Vanegas (Ed.), Comparative Neurology of the Optic Tectum, Plenum Press, New York and London, 1984, pp. 649-686. 9 Jassik-Gerschenfeld, D., Minois, F. and Conde-Courtine, F., Receptive field properties of directionally selective units in the pigeon's optic tectum, Brain Research, 24 (1970) 407-421. 10 Karten, H.J. and Hodos, W., A Stereotaxic Atlas of the Brain of the Pigeon (Columba Livia), The Johns Hopkins Press, Baltimore, MD, 1967. 11 Leresche, N., Hardy, O., Audinat, E. and Jassik-Gerschenfeld, D., Synaptic transmission of excitation from the retina to cells in the pigeon's optic tectum, Brain Research, in press. 12 Reperant, P. and Angaut, P., The retinotectal projections in the pigeon. An experimental optical and electron microscope study, Neuroscience, 2 (1977) 119-140. 13 Webster, K.E., Changing concepts of the central visual pathways in birds. In R. Bellairs and E.G. Gray (Eds.), Essays on the Nervous System, Claredon Press, Oxford, 1974, pp. 258-298.
The authors are greatly indebted to A. Dalbera for her excellent technical assistance. This research was supported by CNRS ( U A 662), I N S E R M (no. 84 60 13) and M R T (no. 84.C.1309).