- Email: [email protected]
Lesion of the PGO pathways in the kitten. II. Impairment of physiological and morphological maturation of the lateral geniculate nucleus
Brain Research, 485 (1989) 267-277 Elsevier
267
BRE 14418
Lesion of the PGO pathways in the kitten. II. Impairment of physiological and morphological maturation of the lateral geniculate nucleus Damien Davenne 1, Yves Frdgnac 2, Michel Imbert 3 and Jo~lle Adrien 1 IINSERM U288, CHU Piti~-Salp~tridre, Paris (France), 2UA 041121 CNRS, Universit~ Paris-Sud, Orsay (France) and3Universit~ Paris VI, Paris (France)
(Accepted 27 september 1988) Key words: Ponto-geniculo-occipital activity; Lateral geniculate nucleus; Maturation; Kitten; Extraretinal factor
Suppression of the geniculate ponto-geniculo-occipital (PGO) waves by bilateral lesions of PGO pathways at the mesencephalic level in 15-day-old kittens has been shown to induce a significant reduction of the mean discharge frequency recorded in the lateral geniculate nucleus (LGN) during paradoxical sleep. The present paper reports that one month after the bilateral lesion (i.e., 6-7 weeks of age) important deficits in the maturation of the LGN were observed: (1) electrophysiologically, the latencies of the LGN cellular responses to stimulation of the optic chiasm were significantly longer than those of age-paired controls or of unilaterally lesioned animals, and the proportion of visual cells characterized as type X by stimulation of the visual field was smaller; and (2) morphologically, the volume of the LGN and the size of its neuronal somata were smaller than those in control. These data suggest that bilateral suppression of extraretinal PGO afferents to the LGN in the kitten induces a significant delay in the development of this nucleus.
INTRODUCTION It is well established that specific sensory inputs play a crucial role in the development of the central nervous system: in particular, impairment of visual experience during development produces important alterations in both the morphological and functional properties of the visual system 13'15'18"2°'23'27"37. However, specific sensory stimulation is but one of the many factors which interact with the maturation of neuronal networks. One of these networks, the endogenous activity generated within the developing nervous system whether or not of retinal origin, might be crucial for the maturational processes of visual pathways especially during early life when the sensory receptors are not fully functional. Most endogenous stimulation to the sensory visual system is provided during paradoxical sleep (PS) 2'39, and the amount of time spent by mammals in this
sleep state while very large at early developmental stages decreases progressively as maturation proceeds 19. Therefore, it is likely that the state of PS bears a particular significance in developmental processes as a specialized temporal window during which massive endogenous activity is necessary for the correct expression of some innate program 32. Since the proposal of this hypothesis, several groups have attempted to clarify the role of PS in postnatal development by suppressing this sleep state in rats and cats at birth 26,36. However, this question remains unanswered due mainly to the fact that in immature animals such a deprivation has been achieved by the use of chronic systemic injections of pharmacological agents making it difficult to assess whether the effects on maturation are due to PS-deprivation or to the drug itself. In order to overcome this difficulty, we studied a developmental model where PS itself was not suppressed but where
Correspondence: J. Adrien, INSERM U288, 91, Bd. de l'H6pital, 75013 Paris, France.
268 the level of endogenous activity in forebrain neurons was specifically reduced during this sleep state. subsequent to lesioning the ponto-geniculo-occipital (PGO) pathways, as described in a previous paper 12. in brief, selective electrolytic lesions of the PGO pathways performed in the kitten at age 15 days, induced during the following month an important decrease of phasic unit activity in the lateral geniculate nucleus (LGN) which is only observed during PS. In agreement with the 'ontogenetic hypothesis '32, we now report that the maturation of this nucleus was impaired in the present model at both the anatomical and functional levels. Preliminary reports of this study have been presented elsewhere 1°'~. MATERIALS AND METHODS A total of 56 animals was used for 3 different types of analysis and when possible the same animals contributed to two of them (Fig. 1). At 15 days of age, the experimental kittens underwent different electrolytic mesencephalic lesions (5 mA DC current during 30 s between ground and a negative electrode made out of enameled nichrome, 250/~m diameter, insulated except at the tip for 2 mm) 1°, which allowed the distinction of 3 separate groups. In group L, the lesions bilaterally affected the mesencephalic pathways of P G O activity, i.e., the dorsolateral part of brachium conjunctivum and nearby area of mesencephalic reticular formation 1°'24'34. These lesions resulted in an immediate and total suppression of PGO waves in the LGN during at least one month after surgery. In group C, bilateral mesencephalic lesions of comparable size affecting areas surrounding the P G O pathways resulted in a normal level of PGO activity. In the last group, U, a unilateral lesion involving the right PGO pathways was performed. Different labeling technics (for review, see ref. 34) revealed that the LGNs receive from the P G O generator at least 70% of ipsilaterai fibers, the rest which are contralateral. Group U represents a control for possible effects of transneuronal degeneration in the LGN ipsilateral to the lesion, where in spite of the unilateral lesion the remaining fibers maintained a normal level of PGO activity as well in kittens 12 as in adult cats 24. After surgery, animals were housed in standard
NL15
7 .....
protocol
~ m
number
...........
NL30
of
animals
3
o
i_7 ....
,
i2 o
v",r
m]
,J
r
0
----n
1~5 Post-natal
3~0
45
age (days)
Fig. 1. Experimental protocols indicating the time of lesion and sacrifice as well as the type of study performed in each group of kittens. Arrows represent bilateral or unilateral lesions, whereas letters indicate the type of data the animals contributed (see Materials and Methods): m, morphometry; x, measure of latency after stimulation of optic tract; r, cellular response to stimulation of the receptive field
laboratory conditions and lived with mother and littermates until weaning at 5 weeks of age. At 6 weeks they were used for the electrophysiological experiment as described below. In addition to these three experimental groups, a number of intact kittens aged respectively 15 (group NL15), 30 (group NL30) and 45 days (group NL45) were also studied (Fig. 1).
(I) Electrophysiological recordings in the L G N performed under acute conditions according to two different protocols (1) Stimulation of the optic chiasm (OX) (protocol x). At age 6 weeks the animals were anesthetized with a-tesine (Glaxo, 1.2 ml/kg, i.m., 10.8 mg a-xalone/kg and 3.6 mg a-dolone acetate/kg) and positioned in a stereotaxic apparatus (Kopf). Anesthesia was maintained by hourly intravenous injections of a-tesine (1.2 ml/kg) dissolved in 0.9% NaCI while the electrocardiogram was monitored and body temperature was maintained around 38 °C by means of a heating pad driven by a rectal probe. After local application of xylocaine, the scalp was opened along the sagittal midline, and an opening in the skull was drilled over the LGN and the OX. A bipolar electrode (enameled nichrome wire, 200 # m diameter) was lowered to the stereotaxic coordinates of the OX (A = 10, L = 0) using a 10° lateral angle
269 in order to avoid the sagittal sinus. While moving down the electrode the mass potential evoked in response to visual stimuli provided by a strobe flash was recorded. The final position of the electrode was set at the point of maximum response. This electrode was used for stimulation of the OX with a homemade isolator-stimulator. Stimulating pulses (0.2-5 mA; 50-200 ps) were applied in trains of 2-10 pulses at 10-50 Hz frequency. The pupils were dilated with local application of atropine sulfate, contact lenses were placed on the cornea and an obstructor was placed in front of either eye in order to select the neurons which were monocularly activated by a flash. In the LGN, recordings of extracellular action potentials were performed by using tungsten-in-glass microelectrodes (2-6 M$2; Frederick Haer). Signals from the electrode were fed through an impedance adaptor, a variable gain amplifier and active filters for low potentials. These signals were then displayed on an oscilloscope, monitored through an audio amplifier, and fed in a window discriminator whose output was taped together with the original signal. The electrode was stereotaxically positioned above the LGN (A = 3.5; L = 7.5; V = -2) and moved down slowly. The first cell response driven by the strobe flash was considered to be an indicator of reaching the LGN region and its vertical coordinate was noted. From this point the electrode was lowered systematically by steps of 100 pro, and at each step all units encountered within a 50 pm vertical displacement were analyzed. Definite evidence for LGN localization was given by the monocularity of the cells encountered. Only the monocular cells of lamina A were selected, and their responses to the OX stimulation were examined on a storage oscilloscope (5-15 superimposed traces). Both the response of the optic fiber (exhibiting a constant latency) and that of the postsynaptic cells (with a variable latency) could be recorded simultaneously by using microelectrodes with low impedance 9.
(2) Stimulation of visual receptive field (protocol r). The animals were prepared as described above, and in addition their trachea was intubated. They were then placed in a stereotaxic head holder (Horsley-Clarke) adapted for the study of the visual system 33. Paralysis was induced by gallamine triet-
triodide (15 mg/kg; Flaxedil, Specia) and maintained by continuous intravenous perfusion of a-tesine (0.3 ml/kg.h) and galamine triettriodide (10 mg/kg.h) supplemented with glucose and saline. Respiration was driven at 30 cpm, and the volume was controlled so as to maintain the expired CO 2 around 4%. Body temperature was kept at about 38 °C by means of a heating pad. Optic disks were projected onto a tangent screen placed at 57 cm from the eyes in order to estimate the position of the area centralis 6. As described above a bipolar stimulating electrode was placed in the OX, and a microelectrode was moved down in the LGN. For each monocularly activated unit of lamina A the receptive field was mapped with a light spot flashed or moved on the screen 4. The receptive field center response (on, off, on-off) was determined. Finally a drifting sine-wave grating generated on a cathode-ray tube (Joyce, 350 cd/m 2 luminance) by a spatial stimulator (Cambridge Microsystem) was placed at the center of the receptive field, and a grating of optimum response was selected in order to classify the cells using static counter-phase stimulation14: all cells with a 'nul-position' (i.e. a position of grating pattern that elicited no response) were considered as X-cells 14"29 and all the others as non-X (Fig. 2).
(II) Morphometry (protocol m) (1) Histology. At the end of the electrophysiological experiments, the animals were perfused intracardially with saline (100 ml in 5 min) and with formalin-saline 10% (500 ml in 45 rain). The brains were removed and left in formalin during 15-30 days for complete fixation. They were then dehydrated in a graded series of ethanol solutions and embedded in paraffin (for dehydration and embedding brains from each group were processed at the same time in order to reduce the variability due to these technics20). Frontal sections 20 pm thick through the LGN were mounted on slides and stained with Cresyl violet. (2) Quantitative analysis. The total volume of the LGN was calculated using the planigraphic method, and from the area where the electrophysiological recordings were made, the size of the LGN somata was measured using the following protocol: sections from the medial rostrocaudal part of the LGN (A --
270 3.0-3.5) were selected and the mean size of 100 LGN cells from the mediolateral extent of lamina A was measured. For this measurement the perikaryal outline in the plane of a cell with a visible nucleus ~s was traced at 1000×, and the cross-sectional area was estimated using a Tasic Analysis System (Leitz). The microscope field was moved step by step perpendicular to the dorsal border in order to cover the entire selected region. The large neurons lying in the interlaminar zone and the oblong ones near the dorsal LGN border were not taken into account 2°. (3) Statistical analysis. It is well known that there is a large variability in the mean size of LGN cells among animals of the same age reared under the same conditions 1s'2°. This fact has led previous investigators to consider the mean cell size for each animal as a single observation. The same index has been used in this study thus allowing the calculation of a group mean for all animals which had undergone the same lesion. Statistical tests were performed to evaluate the differences between the groups (Student's t-test, X2 and Kolmogoroff-Smirnoff). The significance level was taken as less than 0.01 for parametric and non-parametric tests. RESULTS
(I) Responses to stimulation of the optic chiasm A total of 849 cells responding to stimulation of the optic chiasm were recorded in the 4 groups of animals (protocol x, Table I). In all groups, the latencies of the retinal fibers were comparable (Table I), whereas those of the LGN cellular responses differed according to age and to treatment. In non-lesioned groups (NL) the mean latencies decreased with age (3.4 + 1.3 ms at 30 days vs 2,7 + 0.3 ms at 6 weeks of age). This difference was significant (P < 0.01) and was mainly due to an increase in the proportion of cells exhibiting short latencies (Fig. 3): at 30 days of age, only 12% of the cells had a latency shorter than 2 ms whereas at 6
TABLE I
Comparison of latencies between retinal fiber and LGN cell latencies (protocol x) The table indicates the experimental groups together with the proportion of kittens studied in each group, the number of LGN monocular cells recorded in these animals and the mean latencies of both the retinal fibers and the cell responses obtained after stimulation of the optic chiasm (OX).
Group
Numher of animals
Number of cells
Latenciesof the fibers (ms + S.E.M.)
Latenciesof the cell (ms ++_ S.E.M.)
NL30 C U L
3 15 6 15
65 277 142 365
1.15+0.12 1.08 + 0.08 1.07 + 0.08 1.09 + 0.04
3.4+1.3" 2.7 + 0.3 2.5 + 0.3 3.2 + 0.4*
* Indicates significant difference from group C (Student's t-test, P < 0.001).
weeks this amount reached 40%. In group U, at 6 weeks of age, the average latency of neurons recorded in the L G N ipsilateral to the lesion was statistically indistinguishable from that of age-paired kittens of group C (2.5 vs 2.7 ms, Table I and Fig. 3). In contrast, in group L where geniculate PGOs were bilaterally suppressed, the mean latency was significantly longer than that of group C and U (3.2 vs respectively 2.7 and 2.5 ms, P < 0.001, Table I) and a reduced proportion of cells with latencieg below 2 ms was found (Fig. 3). The latency histograms observed in group L at 6 weeks of age were in fact indistinguishable from those of the intact group NL30. This suggests that the lesion induced a 2 week lag in the maturation of OX latencies.
(II) Receptive field classification of geniculate cells Using protocol r, 30 monocular cells were recorded in 2 control animals and 47 were recorded in 4 lesioned animals. In agreement with what has been described above on a larger sample, the latencies of response to OX stimulation were larger in group L
Fig. 2. Typical post-stimuli time histograms (PSTH) displays of the responses in the LGN to spatial sinusoidal modulation of luminance: sine-wave gratings were reversed by changing the phase by 180° every second as indicated by the bars at the bottom of the figure. Control is the spontaneous activity of the cell for a null contrast. Six different phases are presented in a random manner (30 ° separation). Left PSTH (A) shows the response of an X-cell. Note the null position (90*) for which the grating pattern reversal elicited no response. On the right, an example (B) of a non-X cell where a response is found for every phase of the stimulation.
271
B
A 5 ~
,I,I h, k,I I,
CONTROL 10t
iI h ,
0
*, dill ,,I,,,I
k L,,I,,~,,Ill,
L,J,n,=,,l,,l=
i
30
,.l J .
l,
.I
I.
I~
t, A.tt L.
10t ~ dh, i
•~aL ,~.,L_
60
5~
i,tll h,l it,, , i,,[,
"!
h,l
t,,
90
I,I
120
I
0
I 1
l 2
*
J I
~d ~.., .sh.,,.l k J I ,
lo] 150
.=.
lit
I
I I
I
1
0
1
__J 2 sec.
272 n_-65
n_-277
100 ~
13---142
n--365
17%, Table 1I).
m
+
+ +
+
,
50.
[---] > 8 msec
+ .
4-8 2-4 ~B
NL30
C
U
<2
L ~J
Y
45 days
Fig. 3. Schematic representation of the latency of responses to optic chiasm stimulation and its evolution with age. Each bar is divided into segments representing the percentage of cells with response latencies within ranges shown in the key on the right. In lesioned animals aged 45 days (L) the relative number of cells with short latencies is significantly smaller than that of aged-paired control (C and U) and similar to that of 30-day-old non-lesioned kittens (NL30).
than in group C (3.0 vs 2.3 ms, Table II). A frequency histogram of their distribution (Fig. 4) shows that only 19% of the units in group L and 50% in group C exhibited a latency shorter than 2 ms. A c c o r d i n g to their responses to a light spot flashed at the center of their receptive field, the cells were classified as center 'on', 'off' or 'on-off'. Most cells were pure ' o n - ' or 'off-center' but the proportion of mixed response type cells ('on-off') was significantly larger in group L than in group C (26 vs
Considering the phase d e p e n d e n c y of the response to contrast reversal patterns, the cells were characterized as 'X' or 'non-X'. In group C, 44% of the cells were found to be of the X-type (Table iI). For these cells, the mean latency of response to O X stimulation was 2.5 ms and 38% of them exhibited a latency shorter than 2 ms with a distribution skewed in favor of the shortest latcncies on the sample population (Fig. 4). In contrast, in group L a significantly smaller n u m b e r of cells (32%, P < 0.01) were classified as X-cells (Table II) and their mean latency was larger than that of controls (3.4 vs 2.5 ms, P < 0.01). These cells did not contribute to the shortest latencies classes of the frequency histogram, since none of them contributed to latencies below 2 ms (Fig. 4). (III) Morphometry
A total of 39 animals were used (protocol m, Fig. 1). For each animal the body weight, the total volume of the right geniculate nucleus and the size of neuronal somas in this nucleus were m e a s u r e d (Table II1). The b o d y weight u n d e r w e n t a two-fold increase between 15 and 45 days of age in nonlesioned kittens. In all lesioned groups (C, U and L) this evolution was statistically the same. R e g a r d i n g the morphological analysis: in non-lesioned animals (groups NL15, NL30, NL45) both the total volume of the nucleus and size of geniculate neurons increased with age (Table III, Fig. 5), but the maturation of each p a r a m e t e r followed a different time course: at 15 days, the total volume of the
TABLE II Receptivefield properties of LGN neurons in control and lesioned kittens (protocol r)
This table shows the experimental groups together with the number of LGN monocular cells recorded in these animals, the mean latencies of the response to stimulation of the optic chiasm (OX), the proportion of cells classified as center 'on', 'off' or 'on-off' after stimulation of their receptive field centers and finally the number of cells characterized as 'X', i.e., presence of null position in their receptive field for contrast reversal. Group
C L
Number of animals
Number of cells
Latencies of the cells (ms + S. E. M.)
2 4
30 47
2.3 _+0.5 3.0 _+0.6*
Type of receptivefield center On (%)
Off (%)
On-off (%)
41 38
42 36
17 26*
X-type (%)
44 32*
* Significant differences from group C (latencies: Student's t-test, P < 0.01; different characteristics of the cells: X2, P < 0.01).
273
nucleus was proportionally larger than the size of its neurons. These proportions were reversed between 15 and 30 days and then r e m a i n e d so until the age 45 days, probably reflecting an almost equal increase in both the v o l u m e of the neuropile and the size of the somas. In control animals aged 6 weeks (C) the total v o l u m e of the L G N
TABLE III
Morphological analysis of the lateral geniculate nucleus (protocol m) This table indicates the groups of animals together with the number of kittens studied in each group, their mean body weight, the mean volume of the lateral geniculate nucleus (LGN) and the mean area of LGN cells.
and the size of geniculate
neurons were not significantly different from those
Group
Number of animals
Body weight Volume of (g +_ the nucleus S.E.M.) (mm3 + S.E.M.)
Size of the cells (pro2 + S.E.M.)
NL15 NL30 NL45 C U L
5 3 4 8 6 13
288 + 453 + 615 + 593 + 595 + 505 +
151 + 22* 174 + 32* 285 + 29 270 + 24 252 + 9 218 + 4*
of unilaterally or non-lesioned kittens of the same age (respectively U and NL45). In group U, there was a trend (P < 0.05) for the total volume of the L G N to be smaller than that of non-lesioned animals while the size of the neurons was not affected. In contrast, in group L both the total volume and the size of neurons were significantly smaller (P < 0.01)
3.4 + 4.2 + 7.8 + 7.0 + 6.5 + 5.0 +
0.2 0.3 0.5 0.4 0.4 0.2
Significant difference from group C (Student's t-test, P < 0.01).
*
50-
20 33 20 21 49 27
%
40-
Group
C
- 300
n-30
10
30-
9 20-
8 10-
250
E
0°
r 01
~ 2 3 4 5 6 7 8
E
I)
T
o
6
I
Z
I 40-
200
.J
I
%
Group L
E ::L.
5
n-47 4
30
3
20
10O"
I I i i NL15
r
0
4
5 6 7 8
Fig. 4. Frequency histograms representing latencies of cellular response to stimulation of the optic chiasm in relation to characterization of the receptive field (protocol r). Black areas represent the percent of cells characterized as X-cells. Group L: kittens aged 6 weeks, which had bilateral mesencephalic lesions suppressing PGO activity in the LGNs. Group C: kittens aged 6 weeks which had mesencephalic lesions that did not impair PGO activity in the LGNs. Note that in group L the X-cells contributed to latencies longer than those in group C.
NL30
NL45
I150
c
Fig. 5. Graphic representation of the total volume of the geniculate nucleus (white bars) and of the cell area (hatched bars) measured in control and lesioned animals. The left part of the graph (NL15, NL30, NL45) represents the data obtained at respectively 15, 30 and 45 days after birth in unlesioned animals. The right part of the graph (C, U and L) represents the data obtained at 45 days, following the different kinds of lesions performed at 15 days (see Materials and Methods). Note that the LGN volume and the cell areas in group L are significantly smaller than those in age-paired controls (P < 0.01). In this respect, group L is comparable to group NL30 (non-lesioned kittens aged 30 days).
274 than in groups NL45, C and U. In comparison with control kittens these parameters in group L correspond to values obtained in 30-day-old kittens. This phenomenon represents a maturational deficit estimated at 45% of 'failure to grow' in the group bearing a lesion which suppresses PGO activity in the LGN. DISCUSSION It has been shown in a previous paper 12 that lesions of PGO pathways in 15-day-old kittens suppress P G O waves and PS-phasic unit activity in the LGN but do not disrupt sleep-waking states nor body weight development. It is reported in the present work that these lesions induce an important delay in the maturation of the LGN at both the electrophysiological and the morphological levels. (1) Classically, the latency value of geniculate cellular response to OX stimulation is an index of electrophysiological maturation of the LGN 9'29"37. This latency reflects the sum of the time of axonal conduction between the stimulation site and the terminal area and the delay taken by the synaptic transmission to the LGN relay cell. In the present study where both the pre- and the postsynaptic responses were recorded at the same time, only the parameter relative to the postsynaptic response was found to be affected by the lesions. The presynaptic response exhibited the same latency in all groups of animals (Table I) and was comparable to that reported in the literature 37. In contrast, the postsynaptic latencies recorded in the lesioned animals were found to be significantly longer than in controls. These results might be interpreted simply as a loss of facilitatory influence 24 due to the section of pontogeniculate pathways. However, together with the morphological data indicating a slower growth of LGN cells, they could alternatively suggest that chronic suppression of PGO inputs to this nucleus in kittens is a key factor which suppression induces a maturational deficit of synaptic transmission. One should note that C and U groups correspond also to a partial deafferentation, which makes plausible that a relevant difference between C and L is the location of the lesion itself, and not the simple removal of non-specific facilitatory input. Separating each alternative could be answered by performing similar
experiments in adult cats where acute effects of such lesions have not been studied. The functional development of the LGN has also been analyzed by recording the receptive field of geniculate neurons. The proportion of 'on-off' cells observed in the present work is consistent with previous reports 3'9. However, if the amount of X-cells found here (44%) is in agreement with the data of Norman et al. (50% at 6 weeks29), it differs from the results of others 5"16 who observed, respectively, 30% and 66% of X-cells at the same age. Nevertheless, when confining the comparison of the data obtained here in lesioned animals to those in controls under the same experimental conditions: (i) a smaller proportion of X-cells together with a larger latency of their response to OX stimulation, and (ii) a larger number of 'on-off' center cells were found. These results are comparable to those of Archer et al. 3 who have silenced the action potentials of retinal cells using tetrodotoxin which blocks sodium-dependent action potentials and have found that the lateral geniculate neurons exhibited abnormal retino-geniculate synaptic connections. The present study suggests that, in addition to the tonic level of afferent retinal activity, non-retinal P G O activity may control the development of these connections. Furthermore, the geniculate receptive field organization in the lesioned group was severely impaired as indicated by a reduced proportion of X-cells when compared to age-paired controls. Since several studies have suggested that the development of these cells at the geniculate level might not be influenced by sensory primary inputs 13'23'2s, it can be proposed that PGO activity represents a critical input for the maturation of X-cells. However, one cannot exclude that this critical role is not developmental in nature and that the proportion of X-cells might be dependent on PGO generator integrity or functioning even in the adult 35. (2) From a morphological point of view, the size of neuronal somata is a classical index of maturation, especially in studies of visual deafferentations where that of LGN neurons was measured in order to estimate developmental deficits 1s'2°. In the present work, the mean sizes of L G N neurons obtained in 15- and 45-day-old control animals are comparable to those reported by Kalil 2° who found a 100% increase from 150 p m z at 15 days to 300 p m 2 at 45
275 days of age. However, these data differ from those of Hickey TM where an increase of only 20% was observed during this period (200-240/~m 2 from 15 to 45 days, respectively). Nevertheless, the present study shows that none of the control groups NL45, C and U exhibited significant differences in cell sizes, and that LGN somata in animals bearing a PGO-suppressing lesion were of smaller size than those of controls under the same experimental conditions. With regard to LGN total volume it is interesting to notice that even though groups C and U were not statistically different from one another, there was indication of a difference (P < 0.05) in group U as compared to group NL45. In addition, in group L LGN volume was significantly affected. These trends might reflect the impairment, in all lesioned groups, of some excitatory input from the mesencephalon, though without effects on the LGN neuronal size. Considering that in group U the ipsilateral LGN receives only 30% of the PGO fibers but exhibits normal PGO activity 12, the lack of statistical effect on LGN latencies and cellular sizes indicates that transneuronal degeneration or other direct influence is not primarily responsible for the deficits observed in kittens with bilateral PGO lesions, and that the occurrence of PGO activity is required for a normal maturation of LGN neurons. (3) The question of the role of ongoing electrical activity in the development of the CNS during ontogenesis is still controversial (for review, see ref. 17). Indirect support to this hypothesis arises from the results of numerous in vivo experiments on visual deafferentation 15'~8'37. In most of these studies important impairments of the maturation of visual structures, i.e., cell growth and/or neuronal specificity, were observed after modifications of sensory inputs 13"23'27. Even more drastic impairments were obtained after injection of tetrodotoxin in one or both eyes 3,8,21 indicating that not only visual inputs but also the ongoing spontaneous activity of retinal ganglion cells is necessary for normal maturation of the LGN 3. Data of the present study suggest that the impulses provided to the cells by the PGO input are also involved in the maturation of synaptic mechanisms in the LGN. In this regard, it is interesting to give an estimate in lesioned animals of the overall deficit in spontaneous discharge level in the LGN. Knowing that the mean firing level is reduced by
about 30% specifically during PS and that this state represents about 30% of total time in kittens between 15 and 45 days of age 19, it can be estimated that the mean reduction of neuronal activity ranged around 10% of the total level. If such a modest deficit induced quite severe impairments of the LGN maturation, it should be admitted that the impulses which were suppressed in the lesioned animals bear a specific significance for maturation. Such impulses were shown to be the high frequency neuronal discharge which was drastically reduced during PS 12. Even though it persisted during slow-wave sleep (SWS), this type of discharge exhibited in group L an overall deficit which can be estimated of about 60%, knowing that in controls the absolute level of high frequency discharge is two-fold larger in PS than in SWS ~2. The present hypothesis attributing a particular maturational role to this pattern of neuronal firing is supported by experiments performed in vitro on isolated rat neural lobe, where pulses delivered in bursts induced greater hormone release than an identical number of pulses given at a constant frequency7. It can be proposed that in a somewhat similar way PGO activity stimulates the release of trophic factors in the LGN and more generally in the CNS. In the line of such interpretation, the present data provide strong support to the 'ontogenetic hypothesis '32. Other studies have attempted to test the latter by performing pharmacological deprivation of PS in newborn cats and rats using different compounds which affect PS regulation through distinct mechanisms. In the kitten 36, chronic a-methylDOPA treatment which suppressed 70% of PS from 1 to 14 days of age, induced a slight delay in motor coordination tested at 1 month. In the rat pup 26, daily injections of chlorimipramine or clonidine between 10 and 21 days of age inhibited PS during this period and induced later anatomical, biochemical and behavioral deficits observed at adult age. However, given the lack of specificity of the pharmacological suppression of PS, no definitive conclusion regarding the role of PS in maturation can be drawn from these experiments 25. Other types of studies have shown that mechanical sleep suppression, of both slow wave and paradoxical sleep 31 or of paradoxical sleep onlya°, enhanced the effects of monocular visual deprivation
276 as evidenced by morphological modifications at the L G N level. Although both experimental approaches were concerned with more complex experimental paradigms involved in plasticity of visual system at a later period of life 15 than in our study, they have been interpreted as the indication of an increased level of functional plasticity in the visual pathway following PS suppression. Our results, conducted over the period 15-45 days of age where a matura-
sponds to the ponto-geniculate pathways seemed to facilitate the effects of visual experience > . In spite of the difficulty of determining possible interaction between P G O and visual input during the critical period, the high level of neuronal activity which occurs during PS before vision takes control of the sensory pathways, bears a particular significance for the maturation of the L G N and probably of most forebrain structures. This extraretinal control in the
tion c o m p o n e n t is involved, more likely suggest a
development of the thalamic visual pathway might
rather non-specific arrest in maturational processes
be an example of more general functions of PS during ontogeneticai development of the CNS.
(produced by PS suppression) whose effects add to the deficit produced by eyelid closure. However, involvement of P G O activity in blocking or triggering plasticity in visual pathways appears to still be a controversial issue. At the cortical level one study has reported that intraventricular 6-hydroxydopamine ( 6 - O H D A ) - - which suppresses P G O activity - - reduced plasticity= (but see ref. 1), and electrical stimulation of the mesencephalic area which corre-
REFERENCES 1 Adrien, J., Blanc, G., Buisseret, P., Fregnac, Y., GaryBobo, E., Imbert, M., Tassin, J.P. and Trotter, Y., Noradrenaline and functional plasticity in kitten visual cortex: a re-examination, J. Physiol. (Lond.), 367 (1985) 73-98. 2 Adrien, J. and Roffwarg, H.P., The development of phasic activity in the lateral geniculate nucleus of the kitten, Exp. Neurol., 43 (1974) 261-275. 3 Archer, S.M., Dubin, M.W. and Stark, L.A., Abnormal development of kitten retino-geniculate connectivity in the absence of action potentials, Science, 217 (1982) 743-745. 4 Barlow, H.B., Trigger feature, adaptation and economy of impulses. In K. Leibovic (Ed.), Information Processing in the Nervous System, Springer, New York, 1969, pp. 209-226. 5 Berardi, N. and Morrone, C., Development of gammaaminobutyric acid mediated inhibition of X-cells of the cat lateral geniculate nucleus, J. Physiol. (Lond.), 357 (1984) 525-537. 6 Bishop, P.O., Kozak, W. and Vakkur, G.J., Some quantitative aspects of the cat's eye: axis and plane of reference, visual field co-ordinates and optics, J. Physiol. (Lond.), 163 (1962) 466-502. 7 Cazalis, M., Govindan, D. and Nordmann, J.J., The role of patterned burst and interburst interval on the excitationcoupling mechanism in the isolated rat neuronal lobe, J. Physiol. (Lond.), 369 (1985) 45-60. 8 Chapman, B., Jacobson, M.D., Reiter, O.H. and Stryker, M.P., Ocular dominance shift in kitten visual cortex caused by imbalance in retinal electrical activity, Nature (Lond.), 324 (1986) 154-156. 9 Daniels, J.D., Pettigrew, J.D. and Norman, J.L., Development of single-neuron responses in kitten's lateral
ACKNOWLEDGEMENTS D. D a v e n n e was supported by a grant from la Fondation Singer-Polignac. The authors thank Pr. J.L. Binet and C. Lestie for their assistance and opportunity of using their Tasic Analysis System, and S. Osborne-Orsoni for correcting the English.
geniculate nucleus, J. Neurophysiol., 41 (1978) 1373-1393. 10 Davenne, D. and Adrien, J., Suppression of PGO waves in the kitten: anatomical effects on the lateral geniculate nucleus, Neurosci. Lett., 45 (1984) 33-38. 11 Davenne, D. and Adrien, J., Maturation 61ectrophysiologique des neurones du noyau g6nicul~ lat6ral apr6s 16sion des voies ponto-geniculo-occipitales (PGO) chez le chaton, C.R. Acad. Sci., 300 (1985) 59-64. 12 Davenne, D. and Adrien, J., Lesion of the pontogeniculo-occipital pathways in kittens. I. Effects on sleep and on unitary discharge of the lateral geniculate nucleus, Brain Research, 409 (1987) 1-9. 13 Derrington, A.M. and Hawken, M.J., Spatial and temporal properties of cat geniculate neurons after prolonged deprivation, J. Physiol. (Lond.), 314 (1981) 107-120. 14 Enroth-Cugell, C. and Robson, J.C., The contrast sensitivity of retinal ganglion cells of the cat, J. Physiol. (Lond.), 187 (1966) 517-552. 15 Fregnac, Y. and Imbert, M., Development of neuronal selectivity in primary visual cortex of cat, Physiol. Rev., 64 (1984) 325-434. 16 Hamasaki, D.I. and Sutija, V.C., Development of X- and Y-cells in kittens, Exp. Brain Res., 35 (1979) 9-23. 17 Harris, W.A., Neural activity and development, Annu. Rev. Physiol., 43 (1981) 689-710. 18 Hickey, T.L., Development of the dorsal lateral geniculate nucleus in normal and visually deprived cats, J. Comp. Neurol., 189 (1980) 467-481. 19 Jouvet-Mounier, D., Astic, L. and Lacote, D., Ontogenesis of the states of sleep in rat, cat and guinea pig during the first postnatal month, Dev. Psychobiol., 2 (1970) 216-239. 20 Kalil, R.E., A quantitative study of the effects of monocular enucleation and deprivation on cell growth in the dorsal geniculate nucleus of the cats, J. Comp. Neurol., 189
277 (1980) 483-524. 21 Kalil, R.E., Dubin, M.W., Scott, G. and Stark, L.A., Elimination of action potentials blocks the structural development of retinogeniculate synapses, Nature (Lond.), 323 (1986) 156-158. 22 Kasamatsu, T. and Pettigrew, J.D., Depletion of brain catecholamines: failure of ocular dominance shift after monocular occlusion in kittens, Science, 194 (1976) 206209. 23 Kratz, K.E., Spatial and temporal sensitivity of lateral geniculate cells in dark-reared cats, Brain Research, 251 (1982) 55-63. 24 Laurent, J.P., Cespuglio, R. and Jouvet, M., D61in6ation des voies ascendantes de l'activit6 PGO chez le chat, Brain Research, 65 (1974) 29-52. 25 Mirmiran, M., The role of REM sleep in development, Sleep Res., 16 (1987) 183. 26 Mirmiran, M., Scholtens, J., Van de Poll, N.E., Uylings, H.B.M., Van der Gugten, J. and Boer, G.J., Effects of experimental suppression of active REM sleep during early development upon adult brain and behavior in the rat, Dev. Brain Res., 7 (1983) 277-286. 27 Movshon, J.A. and Van Sluyters, R.C., Visual neuronal development, Annu. Rev. Psychol., 32 (1981) 477-522. 28 Mower, G.C., Burchfield, J.L. and Dully, EL., The effects of dark-rearing on the development and plasticity of the lateral geniculate nucleus, Dev. Brain Res., 1 (1981) 418-424. 29 Norman, J.L., Pettigrew, J.D. and Daniels, J.D., Early development of X-cells in kitten lateral geniculate nucleus, Science, 198 (1977) 202-204. 30 Oksenberg, A., Marks, G., Farber, G., Cobbey, K., Speciale, S. and Roffwarg, H.P., Effects of REM sleep deprivation during the critical period of cat visual system
development, Soc. Neurosci. Abstr., 13 (1986) 439. 31 Pompeiano, O. and Corvaja, N., Effects of sleep deprivation on the postnatal development of cells in the lateral geniculate body, Sleep Res., 12 (1983) 336. 32 Roffwarg, H.P., Muzio, J.N. and Dement, W., Ontogenetic development of sleep-dream cycles, Science, 152 (1966) 604-619. 33 Rose, G.H., A stereotaxic apparatus for the developing kitten brain, Physiol. Behav., 5 (1970) 539-540. 34 Sakai, K., Some anatomical and physiological properties of ponto mesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep in the cat. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, Raven, New York, 1980, pp. 427-447. 35 Salinger, W.L., Garraghty, P.E. and Schwartz, M.A., Response of the mature lateral geniculate nucleus to monocular paralysis: contributions of non retinal and retinal components, Brain Research, 192 (1980) 255-260. 36 Saucier, D. and Astic, L., Effets de l'alpha-methyl-L-dopa sur le sommeil du chat nouveau-n6. Evolution comportementale au cours du ler mois post-natal, Psychopharmacology, 42 (1975) 299-303. 37 Sherman, S.M. and Spear, P.D., Organisation of the visual pathways in normal and visually deprived cats, Physiol. Rev., 62 (1982) 738-855. 38 Singer, W. and Rauschecker, J.P., Central core control of developmental plasticity in the kitten visual cortex: electrical activation of mesencephalic and diencephalic projections, Exp. Brain Res., 47 (1982) 223-233. 39 Steriade, M. and Hobson, J.A., Neuronal activity during the sleep-waking cycle. In G.A. Kerkut and J.W. Phillis (Eds.), Progress in Neurobiology, Vol. 6, Elsevier, Amsterdam, 1976, pp. 155-376.
267
BRE 14418
Lesion of the PGO pathways in the kitten. II. Impairment of physiological and morphological maturation of the lateral geniculate nucleus Damien Davenne 1, Yves Frdgnac 2, Michel Imbert 3 and Jo~lle Adrien 1 IINSERM U288, CHU Piti~-Salp~tridre, Paris (France), 2UA 041121 CNRS, Universit~ Paris-Sud, Orsay (France) and3Universit~ Paris VI, Paris (France)
(Accepted 27 september 1988) Key words: Ponto-geniculo-occipital activity; Lateral geniculate nucleus; Maturation; Kitten; Extraretinal factor
Suppression of the geniculate ponto-geniculo-occipital (PGO) waves by bilateral lesions of PGO pathways at the mesencephalic level in 15-day-old kittens has been shown to induce a significant reduction of the mean discharge frequency recorded in the lateral geniculate nucleus (LGN) during paradoxical sleep. The present paper reports that one month after the bilateral lesion (i.e., 6-7 weeks of age) important deficits in the maturation of the LGN were observed: (1) electrophysiologically, the latencies of the LGN cellular responses to stimulation of the optic chiasm were significantly longer than those of age-paired controls or of unilaterally lesioned animals, and the proportion of visual cells characterized as type X by stimulation of the visual field was smaller; and (2) morphologically, the volume of the LGN and the size of its neuronal somata were smaller than those in control. These data suggest that bilateral suppression of extraretinal PGO afferents to the LGN in the kitten induces a significant delay in the development of this nucleus.
INTRODUCTION It is well established that specific sensory inputs play a crucial role in the development of the central nervous system: in particular, impairment of visual experience during development produces important alterations in both the morphological and functional properties of the visual system 13'15'18"2°'23'27"37. However, specific sensory stimulation is but one of the many factors which interact with the maturation of neuronal networks. One of these networks, the endogenous activity generated within the developing nervous system whether or not of retinal origin, might be crucial for the maturational processes of visual pathways especially during early life when the sensory receptors are not fully functional. Most endogenous stimulation to the sensory visual system is provided during paradoxical sleep (PS) 2'39, and the amount of time spent by mammals in this
sleep state while very large at early developmental stages decreases progressively as maturation proceeds 19. Therefore, it is likely that the state of PS bears a particular significance in developmental processes as a specialized temporal window during which massive endogenous activity is necessary for the correct expression of some innate program 32. Since the proposal of this hypothesis, several groups have attempted to clarify the role of PS in postnatal development by suppressing this sleep state in rats and cats at birth 26,36. However, this question remains unanswered due mainly to the fact that in immature animals such a deprivation has been achieved by the use of chronic systemic injections of pharmacological agents making it difficult to assess whether the effects on maturation are due to PS-deprivation or to the drug itself. In order to overcome this difficulty, we studied a developmental model where PS itself was not suppressed but where
Correspondence: J. Adrien, INSERM U288, 91, Bd. de l'H6pital, 75013 Paris, France.
268 the level of endogenous activity in forebrain neurons was specifically reduced during this sleep state. subsequent to lesioning the ponto-geniculo-occipital (PGO) pathways, as described in a previous paper 12. in brief, selective electrolytic lesions of the PGO pathways performed in the kitten at age 15 days, induced during the following month an important decrease of phasic unit activity in the lateral geniculate nucleus (LGN) which is only observed during PS. In agreement with the 'ontogenetic hypothesis '32, we now report that the maturation of this nucleus was impaired in the present model at both the anatomical and functional levels. Preliminary reports of this study have been presented elsewhere 1°'~. MATERIALS AND METHODS A total of 56 animals was used for 3 different types of analysis and when possible the same animals contributed to two of them (Fig. 1). At 15 days of age, the experimental kittens underwent different electrolytic mesencephalic lesions (5 mA DC current during 30 s between ground and a negative electrode made out of enameled nichrome, 250/~m diameter, insulated except at the tip for 2 mm) 1°, which allowed the distinction of 3 separate groups. In group L, the lesions bilaterally affected the mesencephalic pathways of P G O activity, i.e., the dorsolateral part of brachium conjunctivum and nearby area of mesencephalic reticular formation 1°'24'34. These lesions resulted in an immediate and total suppression of PGO waves in the LGN during at least one month after surgery. In group C, bilateral mesencephalic lesions of comparable size affecting areas surrounding the P G O pathways resulted in a normal level of PGO activity. In the last group, U, a unilateral lesion involving the right PGO pathways was performed. Different labeling technics (for review, see ref. 34) revealed that the LGNs receive from the P G O generator at least 70% of ipsilaterai fibers, the rest which are contralateral. Group U represents a control for possible effects of transneuronal degeneration in the LGN ipsilateral to the lesion, where in spite of the unilateral lesion the remaining fibers maintained a normal level of PGO activity as well in kittens 12 as in adult cats 24. After surgery, animals were housed in standard
NL15
7 .....
protocol
~ m
number
...........
NL30
of
animals
3
o
i_7 ....
,
i2 o
v",r
m]
,J
r
0
----n
1~5 Post-natal
3~0
45
age (days)
Fig. 1. Experimental protocols indicating the time of lesion and sacrifice as well as the type of study performed in each group of kittens. Arrows represent bilateral or unilateral lesions, whereas letters indicate the type of data the animals contributed (see Materials and Methods): m, morphometry; x, measure of latency after stimulation of optic tract; r, cellular response to stimulation of the receptive field
laboratory conditions and lived with mother and littermates until weaning at 5 weeks of age. At 6 weeks they were used for the electrophysiological experiment as described below. In addition to these three experimental groups, a number of intact kittens aged respectively 15 (group NL15), 30 (group NL30) and 45 days (group NL45) were also studied (Fig. 1).
(I) Electrophysiological recordings in the L G N performed under acute conditions according to two different protocols (1) Stimulation of the optic chiasm (OX) (protocol x). At age 6 weeks the animals were anesthetized with a-tesine (Glaxo, 1.2 ml/kg, i.m., 10.8 mg a-xalone/kg and 3.6 mg a-dolone acetate/kg) and positioned in a stereotaxic apparatus (Kopf). Anesthesia was maintained by hourly intravenous injections of a-tesine (1.2 ml/kg) dissolved in 0.9% NaCI while the electrocardiogram was monitored and body temperature was maintained around 38 °C by means of a heating pad driven by a rectal probe. After local application of xylocaine, the scalp was opened along the sagittal midline, and an opening in the skull was drilled over the LGN and the OX. A bipolar electrode (enameled nichrome wire, 200 # m diameter) was lowered to the stereotaxic coordinates of the OX (A = 10, L = 0) using a 10° lateral angle
269 in order to avoid the sagittal sinus. While moving down the electrode the mass potential evoked in response to visual stimuli provided by a strobe flash was recorded. The final position of the electrode was set at the point of maximum response. This electrode was used for stimulation of the OX with a homemade isolator-stimulator. Stimulating pulses (0.2-5 mA; 50-200 ps) were applied in trains of 2-10 pulses at 10-50 Hz frequency. The pupils were dilated with local application of atropine sulfate, contact lenses were placed on the cornea and an obstructor was placed in front of either eye in order to select the neurons which were monocularly activated by a flash. In the LGN, recordings of extracellular action potentials were performed by using tungsten-in-glass microelectrodes (2-6 M$2; Frederick Haer). Signals from the electrode were fed through an impedance adaptor, a variable gain amplifier and active filters for low potentials. These signals were then displayed on an oscilloscope, monitored through an audio amplifier, and fed in a window discriminator whose output was taped together with the original signal. The electrode was stereotaxically positioned above the LGN (A = 3.5; L = 7.5; V = -2) and moved down slowly. The first cell response driven by the strobe flash was considered to be an indicator of reaching the LGN region and its vertical coordinate was noted. From this point the electrode was lowered systematically by steps of 100 pro, and at each step all units encountered within a 50 pm vertical displacement were analyzed. Definite evidence for LGN localization was given by the monocularity of the cells encountered. Only the monocular cells of lamina A were selected, and their responses to the OX stimulation were examined on a storage oscilloscope (5-15 superimposed traces). Both the response of the optic fiber (exhibiting a constant latency) and that of the postsynaptic cells (with a variable latency) could be recorded simultaneously by using microelectrodes with low impedance 9.
(2) Stimulation of visual receptive field (protocol r). The animals were prepared as described above, and in addition their trachea was intubated. They were then placed in a stereotaxic head holder (Horsley-Clarke) adapted for the study of the visual system 33. Paralysis was induced by gallamine triet-
triodide (15 mg/kg; Flaxedil, Specia) and maintained by continuous intravenous perfusion of a-tesine (0.3 ml/kg.h) and galamine triettriodide (10 mg/kg.h) supplemented with glucose and saline. Respiration was driven at 30 cpm, and the volume was controlled so as to maintain the expired CO 2 around 4%. Body temperature was kept at about 38 °C by means of a heating pad. Optic disks were projected onto a tangent screen placed at 57 cm from the eyes in order to estimate the position of the area centralis 6. As described above a bipolar stimulating electrode was placed in the OX, and a microelectrode was moved down in the LGN. For each monocularly activated unit of lamina A the receptive field was mapped with a light spot flashed or moved on the screen 4. The receptive field center response (on, off, on-off) was determined. Finally a drifting sine-wave grating generated on a cathode-ray tube (Joyce, 350 cd/m 2 luminance) by a spatial stimulator (Cambridge Microsystem) was placed at the center of the receptive field, and a grating of optimum response was selected in order to classify the cells using static counter-phase stimulation14: all cells with a 'nul-position' (i.e. a position of grating pattern that elicited no response) were considered as X-cells 14"29 and all the others as non-X (Fig. 2).
(II) Morphometry (protocol m) (1) Histology. At the end of the electrophysiological experiments, the animals were perfused intracardially with saline (100 ml in 5 min) and with formalin-saline 10% (500 ml in 45 rain). The brains were removed and left in formalin during 15-30 days for complete fixation. They were then dehydrated in a graded series of ethanol solutions and embedded in paraffin (for dehydration and embedding brains from each group were processed at the same time in order to reduce the variability due to these technics20). Frontal sections 20 pm thick through the LGN were mounted on slides and stained with Cresyl violet. (2) Quantitative analysis. The total volume of the LGN was calculated using the planigraphic method, and from the area where the electrophysiological recordings were made, the size of the LGN somata was measured using the following protocol: sections from the medial rostrocaudal part of the LGN (A --
270 3.0-3.5) were selected and the mean size of 100 LGN cells from the mediolateral extent of lamina A was measured. For this measurement the perikaryal outline in the plane of a cell with a visible nucleus ~s was traced at 1000×, and the cross-sectional area was estimated using a Tasic Analysis System (Leitz). The microscope field was moved step by step perpendicular to the dorsal border in order to cover the entire selected region. The large neurons lying in the interlaminar zone and the oblong ones near the dorsal LGN border were not taken into account 2°. (3) Statistical analysis. It is well known that there is a large variability in the mean size of LGN cells among animals of the same age reared under the same conditions 1s'2°. This fact has led previous investigators to consider the mean cell size for each animal as a single observation. The same index has been used in this study thus allowing the calculation of a group mean for all animals which had undergone the same lesion. Statistical tests were performed to evaluate the differences between the groups (Student's t-test, X2 and Kolmogoroff-Smirnoff). The significance level was taken as less than 0.01 for parametric and non-parametric tests. RESULTS
(I) Responses to stimulation of the optic chiasm A total of 849 cells responding to stimulation of the optic chiasm were recorded in the 4 groups of animals (protocol x, Table I). In all groups, the latencies of the retinal fibers were comparable (Table I), whereas those of the LGN cellular responses differed according to age and to treatment. In non-lesioned groups (NL) the mean latencies decreased with age (3.4 + 1.3 ms at 30 days vs 2,7 + 0.3 ms at 6 weeks of age). This difference was significant (P < 0.01) and was mainly due to an increase in the proportion of cells exhibiting short latencies (Fig. 3): at 30 days of age, only 12% of the cells had a latency shorter than 2 ms whereas at 6
TABLE I
Comparison of latencies between retinal fiber and LGN cell latencies (protocol x) The table indicates the experimental groups together with the proportion of kittens studied in each group, the number of LGN monocular cells recorded in these animals and the mean latencies of both the retinal fibers and the cell responses obtained after stimulation of the optic chiasm (OX).
Group
Numher of animals
Number of cells
Latenciesof the fibers (ms + S.E.M.)
Latenciesof the cell (ms ++_ S.E.M.)
NL30 C U L
3 15 6 15
65 277 142 365
1.15+0.12 1.08 + 0.08 1.07 + 0.08 1.09 + 0.04
3.4+1.3" 2.7 + 0.3 2.5 + 0.3 3.2 + 0.4*
* Indicates significant difference from group C (Student's t-test, P < 0.001).
weeks this amount reached 40%. In group U, at 6 weeks of age, the average latency of neurons recorded in the L G N ipsilateral to the lesion was statistically indistinguishable from that of age-paired kittens of group C (2.5 vs 2.7 ms, Table I and Fig. 3). In contrast, in group L where geniculate PGOs were bilaterally suppressed, the mean latency was significantly longer than that of group C and U (3.2 vs respectively 2.7 and 2.5 ms, P < 0.001, Table I) and a reduced proportion of cells with latencieg below 2 ms was found (Fig. 3). The latency histograms observed in group L at 6 weeks of age were in fact indistinguishable from those of the intact group NL30. This suggests that the lesion induced a 2 week lag in the maturation of OX latencies.
(II) Receptive field classification of geniculate cells Using protocol r, 30 monocular cells were recorded in 2 control animals and 47 were recorded in 4 lesioned animals. In agreement with what has been described above on a larger sample, the latencies of response to OX stimulation were larger in group L
Fig. 2. Typical post-stimuli time histograms (PSTH) displays of the responses in the LGN to spatial sinusoidal modulation of luminance: sine-wave gratings were reversed by changing the phase by 180° every second as indicated by the bars at the bottom of the figure. Control is the spontaneous activity of the cell for a null contrast. Six different phases are presented in a random manner (30 ° separation). Left PSTH (A) shows the response of an X-cell. Note the null position (90*) for which the grating pattern reversal elicited no response. On the right, an example (B) of a non-X cell where a response is found for every phase of the stimulation.
271
B
A 5 ~
,I,I h, k,I I,
CONTROL 10t
iI h ,
0
*, dill ,,I,,,I
k L,,I,,~,,Ill,
L,J,n,=,,l,,l=
i
30
,.l J .
l,
.I
I.
I~
t, A.tt L.
10t ~ dh, i
•~aL ,~.,L_
60
5~
i,tll h,l it,, , i,,[,
"!
h,l
t,,
90
I,I
120
I
0
I 1
l 2
*
J I
~d ~.., .sh.,,.l k J I ,
lo] 150
.=.
lit
I
I I
I
1
0
1
__J 2 sec.
272 n_-65
n_-277
100 ~
13---142
n--365
17%, Table 1I).
m
+
+ +
+
,
50.
[---] > 8 msec
+ .
4-8 2-4 ~B
NL30
C
U
<2
L ~J
Y
45 days
Fig. 3. Schematic representation of the latency of responses to optic chiasm stimulation and its evolution with age. Each bar is divided into segments representing the percentage of cells with response latencies within ranges shown in the key on the right. In lesioned animals aged 45 days (L) the relative number of cells with short latencies is significantly smaller than that of aged-paired control (C and U) and similar to that of 30-day-old non-lesioned kittens (NL30).
than in group C (3.0 vs 2.3 ms, Table II). A frequency histogram of their distribution (Fig. 4) shows that only 19% of the units in group L and 50% in group C exhibited a latency shorter than 2 ms. A c c o r d i n g to their responses to a light spot flashed at the center of their receptive field, the cells were classified as center 'on', 'off' or 'on-off'. Most cells were pure ' o n - ' or 'off-center' but the proportion of mixed response type cells ('on-off') was significantly larger in group L than in group C (26 vs
Considering the phase d e p e n d e n c y of the response to contrast reversal patterns, the cells were characterized as 'X' or 'non-X'. In group C, 44% of the cells were found to be of the X-type (Table iI). For these cells, the mean latency of response to O X stimulation was 2.5 ms and 38% of them exhibited a latency shorter than 2 ms with a distribution skewed in favor of the shortest latcncies on the sample population (Fig. 4). In contrast, in group L a significantly smaller n u m b e r of cells (32%, P < 0.01) were classified as X-cells (Table II) and their mean latency was larger than that of controls (3.4 vs 2.5 ms, P < 0.01). These cells did not contribute to the shortest latencies classes of the frequency histogram, since none of them contributed to latencies below 2 ms (Fig. 4). (III) Morphometry
A total of 39 animals were used (protocol m, Fig. 1). For each animal the body weight, the total volume of the right geniculate nucleus and the size of neuronal somas in this nucleus were m e a s u r e d (Table II1). The b o d y weight u n d e r w e n t a two-fold increase between 15 and 45 days of age in nonlesioned kittens. In all lesioned groups (C, U and L) this evolution was statistically the same. R e g a r d i n g the morphological analysis: in non-lesioned animals (groups NL15, NL30, NL45) both the total volume of the nucleus and size of geniculate neurons increased with age (Table III, Fig. 5), but the maturation of each p a r a m e t e r followed a different time course: at 15 days, the total volume of the
TABLE II Receptivefield properties of LGN neurons in control and lesioned kittens (protocol r)
This table shows the experimental groups together with the number of LGN monocular cells recorded in these animals, the mean latencies of the response to stimulation of the optic chiasm (OX), the proportion of cells classified as center 'on', 'off' or 'on-off' after stimulation of their receptive field centers and finally the number of cells characterized as 'X', i.e., presence of null position in their receptive field for contrast reversal. Group
C L
Number of animals
Number of cells
Latencies of the cells (ms + S. E. M.)
2 4
30 47
2.3 _+0.5 3.0 _+0.6*
Type of receptivefield center On (%)
Off (%)
On-off (%)
41 38
42 36
17 26*
X-type (%)
44 32*
* Significant differences from group C (latencies: Student's t-test, P < 0.01; different characteristics of the cells: X2, P < 0.01).
273
nucleus was proportionally larger than the size of its neurons. These proportions were reversed between 15 and 30 days and then r e m a i n e d so until the age 45 days, probably reflecting an almost equal increase in both the v o l u m e of the neuropile and the size of the somas. In control animals aged 6 weeks (C) the total v o l u m e of the L G N
TABLE III
Morphological analysis of the lateral geniculate nucleus (protocol m) This table indicates the groups of animals together with the number of kittens studied in each group, their mean body weight, the mean volume of the lateral geniculate nucleus (LGN) and the mean area of LGN cells.
and the size of geniculate
neurons were not significantly different from those
Group
Number of animals
Body weight Volume of (g +_ the nucleus S.E.M.) (mm3 + S.E.M.)
Size of the cells (pro2 + S.E.M.)
NL15 NL30 NL45 C U L
5 3 4 8 6 13
288 + 453 + 615 + 593 + 595 + 505 +
151 + 22* 174 + 32* 285 + 29 270 + 24 252 + 9 218 + 4*
of unilaterally or non-lesioned kittens of the same age (respectively U and NL45). In group U, there was a trend (P < 0.05) for the total volume of the L G N to be smaller than that of non-lesioned animals while the size of the neurons was not affected. In contrast, in group L both the total volume and the size of neurons were significantly smaller (P < 0.01)
3.4 + 4.2 + 7.8 + 7.0 + 6.5 + 5.0 +
0.2 0.3 0.5 0.4 0.4 0.2
Significant difference from group C (Student's t-test, P < 0.01).
*
50-
20 33 20 21 49 27
%
40-
Group
C
- 300
n-30
10
30-
9 20-
8 10-
250
E
0°
r 01
~ 2 3 4 5 6 7 8
E
I)
T
o
6
I
Z
I 40-
200
.J
I
%
Group L
E ::L.
5
n-47 4
30
3
20
10O"
I I i i NL15
r
0
4
5 6 7 8
Fig. 4. Frequency histograms representing latencies of cellular response to stimulation of the optic chiasm in relation to characterization of the receptive field (protocol r). Black areas represent the percent of cells characterized as X-cells. Group L: kittens aged 6 weeks, which had bilateral mesencephalic lesions suppressing PGO activity in the LGNs. Group C: kittens aged 6 weeks which had mesencephalic lesions that did not impair PGO activity in the LGNs. Note that in group L the X-cells contributed to latencies longer than those in group C.
NL30
NL45
I150
c
Fig. 5. Graphic representation of the total volume of the geniculate nucleus (white bars) and of the cell area (hatched bars) measured in control and lesioned animals. The left part of the graph (NL15, NL30, NL45) represents the data obtained at respectively 15, 30 and 45 days after birth in unlesioned animals. The right part of the graph (C, U and L) represents the data obtained at 45 days, following the different kinds of lesions performed at 15 days (see Materials and Methods). Note that the LGN volume and the cell areas in group L are significantly smaller than those in age-paired controls (P < 0.01). In this respect, group L is comparable to group NL30 (non-lesioned kittens aged 30 days).
274 than in groups NL45, C and U. In comparison with control kittens these parameters in group L correspond to values obtained in 30-day-old kittens. This phenomenon represents a maturational deficit estimated at 45% of 'failure to grow' in the group bearing a lesion which suppresses PGO activity in the LGN. DISCUSSION It has been shown in a previous paper 12 that lesions of PGO pathways in 15-day-old kittens suppress P G O waves and PS-phasic unit activity in the LGN but do not disrupt sleep-waking states nor body weight development. It is reported in the present work that these lesions induce an important delay in the maturation of the LGN at both the electrophysiological and the morphological levels. (1) Classically, the latency value of geniculate cellular response to OX stimulation is an index of electrophysiological maturation of the LGN 9'29"37. This latency reflects the sum of the time of axonal conduction between the stimulation site and the terminal area and the delay taken by the synaptic transmission to the LGN relay cell. In the present study where both the pre- and the postsynaptic responses were recorded at the same time, only the parameter relative to the postsynaptic response was found to be affected by the lesions. The presynaptic response exhibited the same latency in all groups of animals (Table I) and was comparable to that reported in the literature 37. In contrast, the postsynaptic latencies recorded in the lesioned animals were found to be significantly longer than in controls. These results might be interpreted simply as a loss of facilitatory influence 24 due to the section of pontogeniculate pathways. However, together with the morphological data indicating a slower growth of LGN cells, they could alternatively suggest that chronic suppression of PGO inputs to this nucleus in kittens is a key factor which suppression induces a maturational deficit of synaptic transmission. One should note that C and U groups correspond also to a partial deafferentation, which makes plausible that a relevant difference between C and L is the location of the lesion itself, and not the simple removal of non-specific facilitatory input. Separating each alternative could be answered by performing similar
experiments in adult cats where acute effects of such lesions have not been studied. The functional development of the LGN has also been analyzed by recording the receptive field of geniculate neurons. The proportion of 'on-off' cells observed in the present work is consistent with previous reports 3'9. However, if the amount of X-cells found here (44%) is in agreement with the data of Norman et al. (50% at 6 weeks29), it differs from the results of others 5"16 who observed, respectively, 30% and 66% of X-cells at the same age. Nevertheless, when confining the comparison of the data obtained here in lesioned animals to those in controls under the same experimental conditions: (i) a smaller proportion of X-cells together with a larger latency of their response to OX stimulation, and (ii) a larger number of 'on-off' center cells were found. These results are comparable to those of Archer et al. 3 who have silenced the action potentials of retinal cells using tetrodotoxin which blocks sodium-dependent action potentials and have found that the lateral geniculate neurons exhibited abnormal retino-geniculate synaptic connections. The present study suggests that, in addition to the tonic level of afferent retinal activity, non-retinal P G O activity may control the development of these connections. Furthermore, the geniculate receptive field organization in the lesioned group was severely impaired as indicated by a reduced proportion of X-cells when compared to age-paired controls. Since several studies have suggested that the development of these cells at the geniculate level might not be influenced by sensory primary inputs 13'23'2s, it can be proposed that PGO activity represents a critical input for the maturation of X-cells. However, one cannot exclude that this critical role is not developmental in nature and that the proportion of X-cells might be dependent on PGO generator integrity or functioning even in the adult 35. (2) From a morphological point of view, the size of neuronal somata is a classical index of maturation, especially in studies of visual deafferentations where that of LGN neurons was measured in order to estimate developmental deficits 1s'2°. In the present work, the mean sizes of L G N neurons obtained in 15- and 45-day-old control animals are comparable to those reported by Kalil 2° who found a 100% increase from 150 p m z at 15 days to 300 p m 2 at 45
275 days of age. However, these data differ from those of Hickey TM where an increase of only 20% was observed during this period (200-240/~m 2 from 15 to 45 days, respectively). Nevertheless, the present study shows that none of the control groups NL45, C and U exhibited significant differences in cell sizes, and that LGN somata in animals bearing a PGO-suppressing lesion were of smaller size than those of controls under the same experimental conditions. With regard to LGN total volume it is interesting to notice that even though groups C and U were not statistically different from one another, there was indication of a difference (P < 0.05) in group U as compared to group NL45. In addition, in group L LGN volume was significantly affected. These trends might reflect the impairment, in all lesioned groups, of some excitatory input from the mesencephalon, though without effects on the LGN neuronal size. Considering that in group U the ipsilateral LGN receives only 30% of the PGO fibers but exhibits normal PGO activity 12, the lack of statistical effect on LGN latencies and cellular sizes indicates that transneuronal degeneration or other direct influence is not primarily responsible for the deficits observed in kittens with bilateral PGO lesions, and that the occurrence of PGO activity is required for a normal maturation of LGN neurons. (3) The question of the role of ongoing electrical activity in the development of the CNS during ontogenesis is still controversial (for review, see ref. 17). Indirect support to this hypothesis arises from the results of numerous in vivo experiments on visual deafferentation 15'~8'37. In most of these studies important impairments of the maturation of visual structures, i.e., cell growth and/or neuronal specificity, were observed after modifications of sensory inputs 13"23'27. Even more drastic impairments were obtained after injection of tetrodotoxin in one or both eyes 3,8,21 indicating that not only visual inputs but also the ongoing spontaneous activity of retinal ganglion cells is necessary for normal maturation of the LGN 3. Data of the present study suggest that the impulses provided to the cells by the PGO input are also involved in the maturation of synaptic mechanisms in the LGN. In this regard, it is interesting to give an estimate in lesioned animals of the overall deficit in spontaneous discharge level in the LGN. Knowing that the mean firing level is reduced by
about 30% specifically during PS and that this state represents about 30% of total time in kittens between 15 and 45 days of age 19, it can be estimated that the mean reduction of neuronal activity ranged around 10% of the total level. If such a modest deficit induced quite severe impairments of the LGN maturation, it should be admitted that the impulses which were suppressed in the lesioned animals bear a specific significance for maturation. Such impulses were shown to be the high frequency neuronal discharge which was drastically reduced during PS 12. Even though it persisted during slow-wave sleep (SWS), this type of discharge exhibited in group L an overall deficit which can be estimated of about 60%, knowing that in controls the absolute level of high frequency discharge is two-fold larger in PS than in SWS ~2. The present hypothesis attributing a particular maturational role to this pattern of neuronal firing is supported by experiments performed in vitro on isolated rat neural lobe, where pulses delivered in bursts induced greater hormone release than an identical number of pulses given at a constant frequency7. It can be proposed that in a somewhat similar way PGO activity stimulates the release of trophic factors in the LGN and more generally in the CNS. In the line of such interpretation, the present data provide strong support to the 'ontogenetic hypothesis '32. Other studies have attempted to test the latter by performing pharmacological deprivation of PS in newborn cats and rats using different compounds which affect PS regulation through distinct mechanisms. In the kitten 36, chronic a-methylDOPA treatment which suppressed 70% of PS from 1 to 14 days of age, induced a slight delay in motor coordination tested at 1 month. In the rat pup 26, daily injections of chlorimipramine or clonidine between 10 and 21 days of age inhibited PS during this period and induced later anatomical, biochemical and behavioral deficits observed at adult age. However, given the lack of specificity of the pharmacological suppression of PS, no definitive conclusion regarding the role of PS in maturation can be drawn from these experiments 25. Other types of studies have shown that mechanical sleep suppression, of both slow wave and paradoxical sleep 31 or of paradoxical sleep onlya°, enhanced the effects of monocular visual deprivation
276 as evidenced by morphological modifications at the L G N level. Although both experimental approaches were concerned with more complex experimental paradigms involved in plasticity of visual system at a later period of life 15 than in our study, they have been interpreted as the indication of an increased level of functional plasticity in the visual pathway following PS suppression. Our results, conducted over the period 15-45 days of age where a matura-
sponds to the ponto-geniculate pathways seemed to facilitate the effects of visual experience > . In spite of the difficulty of determining possible interaction between P G O and visual input during the critical period, the high level of neuronal activity which occurs during PS before vision takes control of the sensory pathways, bears a particular significance for the maturation of the L G N and probably of most forebrain structures. This extraretinal control in the
tion c o m p o n e n t is involved, more likely suggest a
development of the thalamic visual pathway might
rather non-specific arrest in maturational processes
be an example of more general functions of PS during ontogeneticai development of the CNS.
(produced by PS suppression) whose effects add to the deficit produced by eyelid closure. However, involvement of P G O activity in blocking or triggering plasticity in visual pathways appears to still be a controversial issue. At the cortical level one study has reported that intraventricular 6-hydroxydopamine ( 6 - O H D A ) - - which suppresses P G O activity - - reduced plasticity= (but see ref. 1), and electrical stimulation of the mesencephalic area which corre-
REFERENCES 1 Adrien, J., Blanc, G., Buisseret, P., Fregnac, Y., GaryBobo, E., Imbert, M., Tassin, J.P. and Trotter, Y., Noradrenaline and functional plasticity in kitten visual cortex: a re-examination, J. Physiol. (Lond.), 367 (1985) 73-98. 2 Adrien, J. and Roffwarg, H.P., The development of phasic activity in the lateral geniculate nucleus of the kitten, Exp. Neurol., 43 (1974) 261-275. 3 Archer, S.M., Dubin, M.W. and Stark, L.A., Abnormal development of kitten retino-geniculate connectivity in the absence of action potentials, Science, 217 (1982) 743-745. 4 Barlow, H.B., Trigger feature, adaptation and economy of impulses. In K. Leibovic (Ed.), Information Processing in the Nervous System, Springer, New York, 1969, pp. 209-226. 5 Berardi, N. and Morrone, C., Development of gammaaminobutyric acid mediated inhibition of X-cells of the cat lateral geniculate nucleus, J. Physiol. (Lond.), 357 (1984) 525-537. 6 Bishop, P.O., Kozak, W. and Vakkur, G.J., Some quantitative aspects of the cat's eye: axis and plane of reference, visual field co-ordinates and optics, J. Physiol. (Lond.), 163 (1962) 466-502. 7 Cazalis, M., Govindan, D. and Nordmann, J.J., The role of patterned burst and interburst interval on the excitationcoupling mechanism in the isolated rat neuronal lobe, J. Physiol. (Lond.), 369 (1985) 45-60. 8 Chapman, B., Jacobson, M.D., Reiter, O.H. and Stryker, M.P., Ocular dominance shift in kitten visual cortex caused by imbalance in retinal electrical activity, Nature (Lond.), 324 (1986) 154-156. 9 Daniels, J.D., Pettigrew, J.D. and Norman, J.L., Development of single-neuron responses in kitten's lateral
ACKNOWLEDGEMENTS D. D a v e n n e was supported by a grant from la Fondation Singer-Polignac. The authors thank Pr. J.L. Binet and C. Lestie for their assistance and opportunity of using their Tasic Analysis System, and S. Osborne-Orsoni for correcting the English.
geniculate nucleus, J. Neurophysiol., 41 (1978) 1373-1393. 10 Davenne, D. and Adrien, J., Suppression of PGO waves in the kitten: anatomical effects on the lateral geniculate nucleus, Neurosci. Lett., 45 (1984) 33-38. 11 Davenne, D. and Adrien, J., Maturation 61ectrophysiologique des neurones du noyau g6nicul~ lat6ral apr6s 16sion des voies ponto-geniculo-occipitales (PGO) chez le chaton, C.R. Acad. Sci., 300 (1985) 59-64. 12 Davenne, D. and Adrien, J., Lesion of the pontogeniculo-occipital pathways in kittens. I. Effects on sleep and on unitary discharge of the lateral geniculate nucleus, Brain Research, 409 (1987) 1-9. 13 Derrington, A.M. and Hawken, M.J., Spatial and temporal properties of cat geniculate neurons after prolonged deprivation, J. Physiol. (Lond.), 314 (1981) 107-120. 14 Enroth-Cugell, C. and Robson, J.C., The contrast sensitivity of retinal ganglion cells of the cat, J. Physiol. (Lond.), 187 (1966) 517-552. 15 Fregnac, Y. and Imbert, M., Development of neuronal selectivity in primary visual cortex of cat, Physiol. Rev., 64 (1984) 325-434. 16 Hamasaki, D.I. and Sutija, V.C., Development of X- and Y-cells in kittens, Exp. Brain Res., 35 (1979) 9-23. 17 Harris, W.A., Neural activity and development, Annu. Rev. Physiol., 43 (1981) 689-710. 18 Hickey, T.L., Development of the dorsal lateral geniculate nucleus in normal and visually deprived cats, J. Comp. Neurol., 189 (1980) 467-481. 19 Jouvet-Mounier, D., Astic, L. and Lacote, D., Ontogenesis of the states of sleep in rat, cat and guinea pig during the first postnatal month, Dev. Psychobiol., 2 (1970) 216-239. 20 Kalil, R.E., A quantitative study of the effects of monocular enucleation and deprivation on cell growth in the dorsal geniculate nucleus of the cats, J. Comp. Neurol., 189
277 (1980) 483-524. 21 Kalil, R.E., Dubin, M.W., Scott, G. and Stark, L.A., Elimination of action potentials blocks the structural development of retinogeniculate synapses, Nature (Lond.), 323 (1986) 156-158. 22 Kasamatsu, T. and Pettigrew, J.D., Depletion of brain catecholamines: failure of ocular dominance shift after monocular occlusion in kittens, Science, 194 (1976) 206209. 23 Kratz, K.E., Spatial and temporal sensitivity of lateral geniculate cells in dark-reared cats, Brain Research, 251 (1982) 55-63. 24 Laurent, J.P., Cespuglio, R. and Jouvet, M., D61in6ation des voies ascendantes de l'activit6 PGO chez le chat, Brain Research, 65 (1974) 29-52. 25 Mirmiran, M., The role of REM sleep in development, Sleep Res., 16 (1987) 183. 26 Mirmiran, M., Scholtens, J., Van de Poll, N.E., Uylings, H.B.M., Van der Gugten, J. and Boer, G.J., Effects of experimental suppression of active REM sleep during early development upon adult brain and behavior in the rat, Dev. Brain Res., 7 (1983) 277-286. 27 Movshon, J.A. and Van Sluyters, R.C., Visual neuronal development, Annu. Rev. Psychol., 32 (1981) 477-522. 28 Mower, G.C., Burchfield, J.L. and Dully, EL., The effects of dark-rearing on the development and plasticity of the lateral geniculate nucleus, Dev. Brain Res., 1 (1981) 418-424. 29 Norman, J.L., Pettigrew, J.D. and Daniels, J.D., Early development of X-cells in kitten lateral geniculate nucleus, Science, 198 (1977) 202-204. 30 Oksenberg, A., Marks, G., Farber, G., Cobbey, K., Speciale, S. and Roffwarg, H.P., Effects of REM sleep deprivation during the critical period of cat visual system
development, Soc. Neurosci. Abstr., 13 (1986) 439. 31 Pompeiano, O. and Corvaja, N., Effects of sleep deprivation on the postnatal development of cells in the lateral geniculate body, Sleep Res., 12 (1983) 336. 32 Roffwarg, H.P., Muzio, J.N. and Dement, W., Ontogenetic development of sleep-dream cycles, Science, 152 (1966) 604-619. 33 Rose, G.H., A stereotaxic apparatus for the developing kitten brain, Physiol. Behav., 5 (1970) 539-540. 34 Sakai, K., Some anatomical and physiological properties of ponto mesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep in the cat. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, Raven, New York, 1980, pp. 427-447. 35 Salinger, W.L., Garraghty, P.E. and Schwartz, M.A., Response of the mature lateral geniculate nucleus to monocular paralysis: contributions of non retinal and retinal components, Brain Research, 192 (1980) 255-260. 36 Saucier, D. and Astic, L., Effets de l'alpha-methyl-L-dopa sur le sommeil du chat nouveau-n6. Evolution comportementale au cours du ler mois post-natal, Psychopharmacology, 42 (1975) 299-303. 37 Sherman, S.M. and Spear, P.D., Organisation of the visual pathways in normal and visually deprived cats, Physiol. Rev., 62 (1982) 738-855. 38 Singer, W. and Rauschecker, J.P., Central core control of developmental plasticity in the kitten visual cortex: electrical activation of mesencephalic and diencephalic projections, Exp. Brain Res., 47 (1982) 223-233. 39 Steriade, M. and Hobson, J.A., Neuronal activity during the sleep-waking cycle. In G.A. Kerkut and J.W. Phillis (Eds.), Progress in Neurobiology, Vol. 6, Elsevier, Amsterdam, 1976, pp. 155-376.