Effect of the richness of the environment on neurons in cat visual cortex. I. Receptive field properties

Developmental Brain Research, 53 (1990) 71-81

71

Elsevier BRESD 51054

Effect of the richness of the environment on neurons in cat visual cortex. I. Receptive field properties Clermont Beaulieu* and Max Cynader* Department of Psychology, Dalhousie University, Halifax, N.S. (Canada) (Accepted 7 November 1989)

Key words: Plasticity; Physiology; Environment; Area 17

In a recent study, it was demonstrated that the number of synaptic contacts associated with flat vesicles (FS synapses) is higher in the visual cortex of cats raised in an enriched environmental condition (EC) compared to those reared in an impoverished condition (IC). Moreover, the size of the FS synaptic contacts is also affected by the richness of the animal's environment during development. Based on evidence that the vast majority of FS synapses are GABAergic (~,-aminobutyricacid) and that many of the properties of visual cortex neurons are influenced by GABA-dependent mechanisms, it has been suggested that these morphological synaptic changes induced by the richness of the environment correlate with differences in cortical receptive field properties. In the present study, this has been explored by recording visual responses of area 17 cells in cats raised either in isolation (IC) or in a colony with ample environmental stimulation (EC). Enriched visual cortex contains a higher proportion of orientation selective cells and a lower proportion of orientation biased and unoriented cells. In addition, orientation tuning is significantly sharper in EC animals (mean bandwidth of responsive units is equal to 32°) than in IC cats (mean bandwidth is equal to 38°; P < 0.001). This is mostly due to the greater incidence of orientation biased units in impoverished cortex (23% in EC and 41% in IC animals; P < 0.01). Unit responsivity is significantly affected by the richness of the environment. We found that all units of the EC cortex were responsive to light stimuli. In contrast, 14% of the impoverished cells studied fail to increase their response to at least twice the standard deviation of the spontaneous activity and were judged as unresponsive. We suggest that the lower responsivity in IC visual units is related to the higher number of GABAergic synapses per IC neuron, while the broader selectivity in IC cortex might be due to a more diffuse distribution of the GABAergic inhibitory connections. INTRODUCTION

aminobutyric acid ( G A B A ) - d e p e n d e n t mechanisms 43' been suggested that plastic changes in cortical

44,46 it has The receptive fields of most n e u r o n s in the visual cortex of the normal adult cat are characterized by orientation and direction selectivity26, that is, visual n e u r o n s respond specifically to a stimulus with a given orientation and direction of motion. These properties can be e n c o u n t e r e d among cortical neurons of young kittens but a smaller proportion of cortical neurons are specifically selective and those which are selective are not as finely tuned as in the adult 8'9'11"21'37'4°. These receptive field properties are plastic and can be modified by alterations of the retinal signals during the first few months of postnatal development 15'17'24"27'34"36"41. Alterations of retinal inputs have been i m p l e m e n t e d in many different ways: cats have been raised in a total darkness ~6" 28 with m o n o c u l a r or binocular eyelid suture 54"55, in stroboscopic light ~7 or in a world of visual stimuli of only one orientation24 or one direction ~5. Since orientation and direction selectivity are specified, at least in part, by intracortical inhibitory circuits7 and influenced by y-

receptive field properties may be due, at least in part, to alterations of G A B A e r g i c , inhibitory connections34,36. It appears also that the general lack of responsiveness of cortical n e u r o n s in visually deprived animals is G A B A related ~3,35,5~. Recently, Beaulieu and C o l o n n i e r3 have demonstrated that the n u m b e r of symmetric synapses associated with flat vesicles (FS synapses) is lower in the visual cortex of cats reared in an enriched condition (EC: cats reared in a colony with many stimulative objects) compared to those reared in an impoverished condition (IC: cats reared alone in a cage). In fact, there are twice as many FS synaptic contacts per unit volume of IC visual cortex. In addition, the length of the FS synaptic differentiations is on average 25% higher in the EC cats. The lower n u m b e r of FS synaptic contacts in EC cats is not due so much to a decrease in the n u m b e r of b o u t o n s containing flat vesicles as to a decrease in the n u m b e r of synaptic contacts formed by each b o u t o n 4.

* Present address: Department of Ophthalmology, University of British Columbia, Eye Care Center, 2550 Willow St., Vancouver, B.C., Canada V5Z 3N9. Correspondence: C. Beaulieu, Department of Ophthalmology, University of British Columbia, Eye Care Center, 2550 Willow St., Vancouver, B.C., Canada V5Z 3N9. 0165-3806/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

72 Since the

vast m a j o r i t y

of FS synapses 22'5°'56 are

G A B A e r g i c a n d are thus p r o b a b l y i n h i b i t o r y in the c e r e b r a l c o r t e x 29"33'42, B e a u l i e u and C o l o n n i e r 3 h a v e s u g g e s t e d that the a n a t o m i c a l c h a n g e s i n d u c e d by the r i c h n e s s of the e n v i r o n m e n t on the G A B A e r g i c circuitry might

induce

a

modification

of the

receptive

field

p r o p e r t i e s o f visual cortical cells. This suggestion m a y at first s e e m i n c o m p a t i b l e

with t h e l i t e r a t u r e

on visual

d e p r i v a t i o n b e c a u s e retinal inputs of the cats studied h e r e w e r e n o t drastically m o d i f i e d . H o w e v e r it a p p e a r s that plasticity

o f the

receptive

field p r o p e r t i e s

of visual

cortical cells d e p e n d s n o t o n l y on retinal inputs but also on n o n - r e t i n a l signals: at least 3 non-visual input systems have

been

implicated

in visual c o r t e x

development.

T h e s e i n c l u d e : (1) the n o r a d r e n e r g i c system originating in t h e locus c o e r e l e u s 31'32 (but see refs. 2, 18, 19), (2) p r o p r i o c e p t i v e a f f e r e n t s f r o m the e x t r a o c u l a r muscles w12 a n d (3) the c h o l i n e r g i c s y s t e m o r i g i n a t i n g f r o m the m i d b r a i n r e t i c u l a r f o r m a t i o n 47-49. It is possible that the r i c h n e s s o f the e n v i r o n m e n t particularly affects o n e o f t h e s e s y s t e m s , a n d c o n s e q u e n t l y leads to a m o d i f i c a t i o n o f i n t r a c o r t i c a l G A B A e r g i c circuitry. D e s p i t e a b u n d a n t a n a t o m i c a l and b e h a v i o r a l studies TM 30,53, t h e r e has b e e n little e f f o r t d e v o t e d to the study of physiologic properties

of visual c o r t e x

neurons

as a

f u n c t i o n of e n v i r o n m e n t a l c o m p l e x i t y . In fact, t h e r e are n o studies using single unit t e c h n i q u e s to assess r e c e p t i v e field characteristics. T h i s has l e a d us to investigate the effect o f t h e richness o f t h e e n v i r o n m e n t on the r e c e p t i v e field p r o p e r t i e s o f the cat visual c o r t e x n e u r o n s .

Our

results s h o w that t h e m o r p h o l o g i c a l synaptic c h a n g e s i n d u c e d by t h e c o m p l e x i t y o f the e n v i r o n m e n t are r e l a t e d to d i f f e r e n c e s in cortical r e c e p t i v e field p r o p e r t i e s .

MATERIALS AND METHODS

En vironment Six cats were used in the present study. They were paired by litter and by sex. The two environmental conditions were identical to those used by Beaulieu and Colonnier3'4. Briefly, kittens were born in the laboratory and stayed with their mother until weaning at 6 weeks of age. At this age, kittens were matched by sex and assigned randomly to either an enriched or an impoverished environmental condition. The enriched condition (EC) consisted of a large room equipped with wooden platforms, play furniture and toys. In the EC room, a large number of cats and kittens lived together. These cats were regulary visited by many persons who played with the animals. In addition, a variety of stimulative objects were added and removed regularly in the colony room. In contrast, impoverished kittens were placed alone in separate cages in another room. The sides of the cage were opaque except for the front door which faced a blank wall. All EC and IC kittens had the same type of food and water ad libitum. All animals were treated according to the standards set by the Canadian Council of Animal Care.

Physiological recording Conventional procedures described in other publications from this laboratory were used for single-unit recording 1. Cats were anesthe-

tized with an intravenous injection of a 2.5c~ solution of sodium thiopental. An intravenous injection of 0.15 mg atropine sulfate was also given at this time. All subsequent surgical wounds were infused with 0.25% Marcaine, a long lasting local anesthetic. A tracheal cannula was inserted and the animal was placed in a stereotaxic apparatus. Three small holes were drilled in the cranium; two ot these were over the frontal pole to record EEG and the third hole (diameter of about 2-3 ram) was drilled at stereotaxic coordinates P4/L2 overlying the visual cortex. In order to preserve good stability during single cell recording, care was taken to not puncture the dura and a solution of 0.5% agar-agar was placed in the hole over the visual cortex. The animal was paralyzed with an intravenous administration of gallamine triethiodide (10 mg/kg). During recording, the animal was artificially respired with a mixture of 70% N:,O and 30% O z, End-tidal CO 2 was monitored with a Beckman Medical gas analyser (LB-2) and the respirator stroke volume and rate were adjusted to maintain a peak end-tidal CO 2 concentration of about 3.8~/,. Throughout the experiment, the animal received a continuous infusion of a mixture of gallamine triethiodide (10 mg/kg/h), 5% dextrose in lactated Ringer (2 ml/h) and sodium pentobarbital (1 mg/kg/h). In addition, 50 ml of 5% dextrose in lactated Ringer was injected subcutaneously about every 12 h. Rectal temperature, EEG, and heart rate were monitored continuously throughout the recording period. Atropine sulfate (1%) and phenylephrine hydrochloride (10%) eyedrops were used to dilate the pupils and retract the nictitating membranes. The cornea was then covered by a neutral contact lens. Using a slit retinoscope, refraction was performed on each eye to determine the refractive correction appropriate to bring the eyes to focus on a tangent screen at 1.0 m distant. Contact lenses of appropriate power with 3 mm diameter artificial pupils were then placed on the corneas. The locations of the optic disks and the areae centralis were determined using a reversing ophthalmoscope and plotted on the tangent screen. In those instances in which the area centralis was difficult to locate accurately, its known relative displacement from the optic disk was used to estimate its position. The location of the area centralis was verified periodically during the experiment. Glass-coated platinum-iridium microelectrodes were used, allowing for direct penetration of the intact dura. All penetrations in the cortex were perpendicular to the surface of the lateral gyrus. Penetrations were continued to a depth of up to 3 mm along the medial bank of area 17. Receptive fields of cells encountered were at retinal eccentricities of less than 10°. Eccentricity was defined as the euclidian distance between the receptive field center and the area centralis. Signals from the electrode were amplified (x 1000), displayed on an oscilloscope, reproduced on an audio monitor and fed into a window discriminator which produced digital signals that could be registered by the computer. To ensure accurate isolation of single units, the waveform producing an output from the window discriminator was displayed on the oscilloscope and the shape and time course of the spike was monitored. In order to prevent sampling bias, the electrode was advanced at least 100,~m before the next isolated unit was recorded. Receptive fields of isolated units were plotted with manually controlled moving and flashed stimuli. The ocular dominance was determined and all subsequent stimuli were delivered only to the dominant eye. Receptive field size, preferred orientation and velocity were then determined. Because of the low responsivity of some cells, these parameters could be difficult to estimate with the hand-held light stimulus. In these units, we first determined the approximate center of the receptive field and the preferred orientation of the neural background at the location of the isolated unit. This formed the initial estimates for quantitative stimulations. For quantitative measurements, an optic bench was used to project light bar onto the tangent screen. Light stimuli had a luminance of about 120 cd/m 2, against a constant large field background light of about 35 cd/m 2 provided by a commercial incandescent lamp. The stimuli were drifted across the receptive

73 ficld under control of a PDP-I1/34 computer, using a 12-bit digital-to-analog converter driving a servo controlled mirror galvanometer (time constant, 5 ms). The PDP 11/34 computer was used to control stimulus movement, record responses, and analyze data. Stimulus presentation and spike records were scheduled to a 1 ms time resolution. Moving stimuli of varying velocity were produced by varying the amount of incremental mirror galvanometer movement on each 1 ms update. The orientation tuning curves of individual cells were determined using a light bar (stimulus dimension 6° × 0.5 °) moving across the receptive field in 18 equally spaced different directions (every 20°). Care was taken to ensure that the beginning and the end of the sweep were outside the boundaries of the hand plotted receptive field and that the center of the sweep corresponded closely to the center of the field. Typically the record time of each sweep was about 750 ms with a wait period of at least 100 ms. Fifteen different complete sets of sweeps were used to determine the orientation tuning curve of the cell being studied. The number of spikes at each angle and direction of movement of the stimulus were stored on computer disks for later analysis. Responses were quantified by counting the number of spikes recorded for a given stimulus angle during the 15 sweeps or by computing the peak firing rate over a 50 ms period during the sweep. Evoked responses were obtained by estimating the mean response per sweep and subtracting the spontaneous activity. An estimate of the spontaneous activity was obtained by counting the number of spikes during a 30 s period in which only background levels of illumination were present. We operationally defined a unit as 'responsive' if, for a given stimulus orientation or direction, the cell's firing rate was increased to at least twice the standard deviation of the spontaneous activity. By plotting neural evoked responses against the orientation of the stimulus, several parameters could be measured directly from the resulting orientation tuning curve. Fig. 1 represents an orientation tuning curve of cell A26 and illustrates 5 different parameters that we have estimated. These included: the bandwidth (A; half width at half height), the preferred orientation and direction of the stimulus (C), the mean number of spikes evoked at preferred orientation and direction (B), the firing evoked 90° from the preferred orientation (D: mean of D1 and D2) and at the direction of motion 180° away from the preferred direction (E). From these parameters, we derived two additional measures: (1) the dynamic range and (2) the index of directionality. The dynamic range is obtained by the equation: ( ( B - D)/(B + D))*IO0

Y(x) = A e (x ~'2)/2a2

(2)

A cortical unit is highly selective for direction of motion when this index tends toward 100. In contrast, an unselective unit has a value of close to zero with this scale. In addition, we have placed responsive cells in three different classes, according to their relative responsivity at the orthogonal orientation. This has been obtained by dividing the mean number of spikes at the orthogonal orientation (D) by the number of spikes at the preferred direction (B). This represents the relative decrease in responsivity at the stimulus orientation orthogonal to the preferred orientation. A cortical unit was classified as orientation selective when the number of spikes evoked at the orthogonal

(3)

Where A is the amplitude,/~ is the center position and 3 the width of the curve. We have assumed, as in earlier studies, that twice the standard deviation (23) is the width of the receptive field 1. The levels of statistical significance reported in the present paper, comparing cells of animals reared in the two different environments were calculated using the Mann-Whitney U-test. An analysis of variance (ANOVA) was also used and gave similar results. Coefficients of correlation were calculated between orientation bandwidth and the measurements of the responsivity. When splitting responsive units into 3 categories, only one enriched cell was classified as non-oriented. Thus, it was impossible to apply statistical tests to compare unoriented units in the two groups of cats.

6-

(1)

The dynamic range thus represents the relative evoked response at the preferred orientation and direction compared with that at the orthogonal stimulus orientation. It is thus a measure of how much modulation of unit firing can be produced by stimuli at the preferred and orthogonal orientation. The index of directionality represents the difference between the evoked response at the preferred direction and the response to the stimulus at the same orientation but moving in the opposite direction. The degree of direction selectivity of the unit under study is obtained from the following equation: ((B - E)/(B + E)),100

orientation was less than 10% of that at the preferred orientation. Visual units in which orthogonal responsivity fell between 10% and 50% of the preferred direction response were classified as orientation biased. All other units in which the relative decrease of responsivity did not reach 50% of the total were placed in a third category called unoriented cells. (By definition, it was impossible to estimate the orientation bandwidth and the responsivity at the preferred and non-preferred orientation of these cells.) The reliability of neural response was also evaluated. The reliability is the reciprocal of the variability39. According to Saul and Daniels3s the response reliability can be estimated as the mean number of spikes per sweep divided by the standard deviation across the stimulus sweeps for a given direction. In the present study, we have measured the response reliability at the preferred direction of the stimulus over the 15 sweeps used. We have also quantitatively analysed the width of the receptive field by flashing a light bar at 15 equally spaced intervals across the receptive field. Each flash lasted 250 ms. Only the width of the receptive field at the preferred orientation of the stimulus was investigated. Care was taken to ensure that the region sampled by the flashing bars extended outside the receptive field. Response was quantified by counting the number of evoked spikes in all subregions of the receptive field. The number of spikes was plotted and data were fitted by a Gaussian function:

5

[] B

C E L L #A 26

rn 4

3 2

i

O E

1

:

D1

D2

C

STIMULUS ORIENTATION

Fig. 1. Orientation tuning curve of an unit (cell no. A26) recorded in an enriched visual cortex. This figure illustrates 5 different parameters that we have estimated. These included: the bandwidth (A; half width at half height), the preferred orientation and direction of the stimulus (C), the mean number of spikes evoked at preferred orientation and direction (B), the firing evoked 90° from the preferred orientation (D: mean of D1 and D2) and at the direction of motion 180" away from the preferred direction (E). The stippled area represents the spontaneous activity.

74 RESULTS

General Cortical response properties were quantitatively characterized in 86 cells of the enriched cats and in 94 neurons of the impoverished animals. A unit was judged to be 'responsive', if for some stimulus orientation, the cell response increased to at least twice the standard deviation of the spontaneous activity. Using this criterion, all units of the EC visual cortex were responsive to the light slits used. In contrast, 14% of the impoverished cells studied failed to increase their evoked response to at least twice the standard deviation of the spontaneous activity and were thus classified as unresponsive. Responsive cells were placed in three different categories using the classification defined in the Methods section. In the enriched visual cortex, 76% of units encountered were orientation selective, 23% were orientation biased and only 1% were unoriented. In contrast, only 49% of the responsive units in the impoverished visual cortex were orientation selective, 41% were biased and 10% were unoriented (Table I). Thus, the richness of the environment affects the relative distribution of the degree of orientation selectivity in the cortical unit population. Orientation and direction selectivity assessment Bandwidth of the orientation tuning curve. The frequency distribution of the bandwidth (half width at half height) of the orientation tuning curves is presented for orientation-selective, orientation-biased and non-oriented cells of the two experimental groups in Fig. 2. The bandwidth can be estimated only in units in which the relative responsivity at the orthogonal orientation represents no more than 50% of the number of spikes at the preferred orientation. On the 86 units studied in enriched cats, it is impossible to estimate the bandwidth of only 1 unit (Fig. 2). Among oriented cells, it appears that enriched cells

tend to have a narrower bandwidth than impoverished units. The mean bandwidth among oriented cells is 16% lower in enriched cats (32 °) than in impoverished animals (38°). This difference is highly significant with P < 0.001 on a U-test and on an ANOVA. It is interesting to note also that this difference in the bandwidth of oriented cells is mainly due to a significant difference in the incidence and bandwidth of orientation biased units (37 ° in EC cats and 41 ° in IC cats; P < 0.01); the difference in the bandwidth among orientation selective cells in the two groups of cats is smaller and not significant (31 ° and 33 ° in EC and IC cortex respectively; P i> 0.1). Thus, the impoverished cats have a greater incidence of cells which are more broadly tuned for orientation than in the enriched animals. Bandwidth of orientation tuning curves thus seems to be affected by the richness of the environment. Index of directionality. The index of directionality represents the difference between the evoked response at the preferred direction and the response to the stimulus at the same orientation but moving in the opposite

BANDWIDTH OF THE ORIENTATION TUNING CURVES

30

ENRICHEDCATS

i

20

-

10

6b

2b

sb

Bandwidth (des.)

IMPO~

°

U

CATS

All lI P00rimntttion mlectlve ot'im~d mfim

33°

I

TABLE I

The proportion (and the number) of orientation selective, orientation biased, unoriented and unresponsive units in the visual cortex of EC (enriched) and IC (impoverished) cuts See methods for the definition of each class of cells.

EC

IC

o

o

Orientation selective Orientation biased Unoriented Responsive Unresponsive

76% (65) 23% (20) 1% (1)

49% (40) 41% (33) 10% (8)

100% (86) 0% (0)

86% (81) 14% (13)

2b

4b

"

eb

fro

Bmdwld~ (des.)

Fig. 2. Frequency distribution of the bandwidth (half-width at half-height) in enriched and impoverished cats. Represented by the letter A in Fig. 1. This parameter is significantly affected by the richness of the environment for all units (P < 0.001) and orientation biased cells (P < 0.01).

75 THE NUMBER OF SPIKES AT THE PREFERRED

INDEX OF DIRECTIONALITY

DIRECTION AND ORIENTATION ENRICHED CATS 15

N

All ~

15

ENRI~

10

k-'lOrimmation r . l c ¢ ~ I [ ] OrimSazion biased

All uni~

I0

20

CATS

46.8

40

Indexo f ~

60

80

i00

0

((B-E)/(B+E)X 100)

N X

.

0

10

,

,

.

15.1

. . . . .

20

. 30

,

.

40

15.3 14.3

50

60

Number at ~ p~fmcd orientation(spi~s/swe=p) IMPOVERISI~ED CATS 15

20

IMPOVERISHED CATS .~dl ~ t J

10

36 0

15

i

o

/o

~o

Index of d i ~ c t l o n ~

do

s'0

160

((B-E)/(B+E) X 100)

Fig. 3. Frequency distribution of the index of directionality (from Fig. 1, index directionality = ((B - E)/(B + E) × 100)) in enriched and impoverished cats. This parameter is significantly affected by the richness of the environment for all units (P < 0.05). Note however, that the proportion of impoverished units at the low end of the graph (<20%) is much higher than the proportion of enriched cells in the same part of the distribution graph.

direction. The exact formula used is described in the M e t h o d s section, with high values representing strongly direction-selective neurons and values near zero representing unselective cells. For o r i e n t e d cells, the index of directionality is about 47% in the enriched cats and 40% in the i m p o v e r i s h e d animals (Fig. 3). This difference is significant at a P < 0.05 on an A N O V A but not statistically different on a U-test. M o r e o v e r , no significant difference can be found for the subpopulations of orientation selective or orientation biased cells in the two e x p e r i m e n t a l groups of cats. It thus seems that the index of directionality, though slightly different in the two groups of animals, is much less affected by the environmental complexity than the bandwidth of the orientation tuning curves.

Responsivity Responsivity at the preferred orientation and direction. We have estimated the responsivity of cortical units at the p r e f e r r e d stimulus orientation and direction by measur-

~

9.7

i r~ o , ~ , i , ~

12.0

~o,~,~

,

0

69

.

.

.

.

.

.

.

10

.

-

20

.

30

.

.

.

40

.

.

50

.

60

Number at the preferredorientation (spilms/swecp)

Fig. 4. Frequency distribution of the number of spikes per sweep at the preferred orientation and direction of the stimulus in enriched and impoverished cats. This is represented by the letter B in Fig. 1 in enriched and impoverished cats. This parameter is significantly affected by the richness of the environment for all units (P < 0.001), and orientation biased cells (P < 0.001).

ing (1) the n u m b e r of spikes p e r sweep (count analysis), (2) the p e a k firing rate during the 50 ms p e r i o d of the sweep evoking the greatest response and (3) the reliability of the e v o k e d firing. For an average unit, a light bar moving across the receptive field at the p r e f e r r e d orientation and direction elicits a b o u t 15 spikes/sweep in an enriched cell and 10 spikes/sweep in an impoverished unit (Fig. 4). This difference of 50% is highly significant ( P < 0.001). The n u m b e r of spikes at the p r e f e r r e d direction is twice as great in orientation biased E C units (mean of 14 and 7 spikes per sweep in E C and IC orientation biased units respectively; P < 0.001). However, for orientation selective cells, this p a r a m e t e r is not significantly different in enriched and impoverished units (15 and 12 spikes p e r sweep; P > 0.05). Thus, cells in the impoverished visual cortex are less responsive to visual stimuli and this difference is mostly due to lower responsivity of orientation biased units• We have also estimated the cell responsivity by calculating the p e a k firing rate during the 50 ms period

76 RELIABILITY

NUMBER

O F S P I K E S P E R S W E E P AT 90 °

FROM THE PREFERRED

DIRECTION

ENRICHEDCATS 20

All an.i= k-I~ o n

mele~ve bilued

I~ ~ t i o n

15

2.7 2.6 2.7

ENRICHEDCATS

2O %~ N~

15

[

Alluni= !"1~

0.4 -0.3

mk~ive

10 5

0

5

o

•

i

•

i

I

i

•

w

•

i

°

w

2 3 Reliability

,

,

w

•

i

4

•

0

=

5

20

Allunim I I~ Orkmationlelective

2.0

legOaemtimbiazd ,

1.S

~

~

i'q ",t .

.

•

.

,

0

.

,

2

WIm .

4

,

.

,

6

-

8

15

% % N [ ] IMPOVERISE[~CATS

IO

%

2.2

!

z

,

10

" ~ ' ~ ' ' ~ ' ' ~ Rdi~ility

Ill I ~ o . = , ~ = ~

o.2 2

' 5

5

"'i

,

-2

:

Numb= ,,r 90* from tl~ l ~ k ' m d direction (Spik~=~-'p)

IMPOVERISI-~DCATS

15

,

-4

25

1

~

,

~

~

o ~. -4

.r,.,

.

.

.

.

.

.

-2 O 2 4 Numb~ -' 90" fxom the laefened ~

.

.

.

.

.

.

.

6 8 (Splaa/~V)

l0

Fig. 5. Frequency distribution of the reliability (reciprocal of the standard deviation divided by the mean) at the preferred orientation and direction of the stimulus in enriched and impoverished cats. This parameter is significantly affected by the richness of the environment for all units (P < 0.001), orientation-selective (P < 0.05) and orientation-biased (P < 0.001) cells,

Fig. 6. Frequency distribution of the number of spikes per sweep 90° from the preferred orientation and direction of the stimulus in enriched and impoverished cats. This is represented by the letter D in Fig. 1. This parameter is significantly affected by the richness of the environment for all units (P < 0.001).

with the greatest response. The mean peak firing is about 43 spikes/s in EC visual cortex and 31 spikes/s in IC brain. This difference of 38% is significant (P < 0.005) and also indicates that enriched visual neurons are more responsive. The reliability of the evoked response at the preferred direction was assessed by calculating the reciprocal of the mean evoked response divided by the standard deviation over the 15 sweeps (Fig. 5). This parameter provides another valuable estimate of the responsivity of the unit because it takes into account units' variability in calculating the signal-to-noise ratio. Enriched visual cells are, on average, more reliable than impoverished cortical neurons (2.7 and 2.0 respectively; P < 0.001). That is, the signal-to-noise ratio tends to be higher in enriched cortex. This significant difference in the two groups of cats can be found for the subpopulations of orientationselective (P < 0.05) and orientation-biased cells (P < 0.001). This indicates that for these two classes of responsive units, the reliability and signalling ability of enriched units is higher than that of impoverished cells.

Coefficients of correlation have been calculated between the orientation bandwidth and the three parameters estimating the cell's responsivity. We have found that for impoverished cats the bandwidth is significantly correlated with the number of spikes per sweep (r = -0.308; P < 0.01), the peak firing rate (r = -0:364; P < 0.01) and the reliability (r = -0.157; P < 0.05). In contrast, no significant correlation between these parameters can be found for enriched animals. This indicates that for IC units, the narrower the orientation bandwidth, the greater the responsivity. Responsivity at the orientation 90° from the preferred orientation and the dynamic range. The frequency distribution of the number of spikes evoked at the orthogonal orientation is presented for the two environmental groups in Fig. 6. For oriented cells, the responsivity at the orthogonal orientation is in the order of 0.44 spikes per sweep for the enriched units and 1.14 spikes per sweep for the impoverished cells. This difference is significant at a P < 0.001 on a U-test and a two-way ANOVA. Note that for some units, the response is negative because at

77 SIZE OF THE R E C E P T I V E FIELD

DYNAMIC RANGE 2°1

18,

L

102 112

All aaron ~.lcctive

12,

0

~

ENRICHEDCATS

w

70

0

.IKI

sb

100 150 200 r m ~ ((B-D)/(B+D) XlO0)

0

1

250

20,

2

3 4 5 6 Size receptive field(~g.)

7

8

9

I0

IMPOVERISHEDCATS

IMPOVERISHEDCATS

12

X

!+

All uni~ [] Orim~titmRle.aive

15

80 102 53

~8

~

[] o m ~ t i ~ ~ a i ~

~]ma

m otim~o~ breed

z5 Z3°

X X X

0 ,

o

'

0

.N

too t~o ~o Dymmie ranF ((B-D)/(B+D)XlO0)

2~o

3~0

Fig. 7, Frequency distribution of the dynamic range (from Fig. 1, dynamic range = ((B - D)/(B + D) x 100)) in enriched and impoverished cats. This parameter is significantly affected by the richness of the environment for all units (P < 0.001) and orientation biased cells (P < 0.001).

the orthogonal orientation the responsivity is reduced below the level of spontaneous activity. In fact, 46% of the EC cells decrease their activity below that of spontaneous activity. In contrast, for IC units only 23% of the responsive cells decrease their activity below spontaneous levels and the peak of the frequency distribution for IC units is clearly greater than 0. We have also calculated the relative size of the evoked response at the preferred orientation (and direction) compared with that at the orthogonal stimulus orientation. This yields a dynamic range (Fig. 7) for a unit (see formula in the Methods section). For oriented cells, the average dynamic range is in the order of 102% for EC and 80% for IC units. We see in Fig. 7 that the peak frequency distribution of the dynamic range in EC units is around 100% and that remaining units are symmetrically distributed on each side of the peak. For IC units, the peak is lower and the remaining units are much less evenly distributed. Among orientation selective cells, there is no significant difference between the EC and IC cats (112% and 102% respectively). However, the dy-

LxtNIN.INd\IU"--I~.. ~

l-~

ix~rxP~xJ

S ~ receptivefield(deg.) Fig. 8. Frequency distribution of the size of the receptive field in enriched and impoverished cats. This parameter is not significantly affected by the richness of the environment in any of the groups of cells studied,

namic ranges of orientation biased units are significantly higher in EC cats than in IC cortex (70% and 53%; P < 0.001).

Neural firing evoked 180° away from the preferred direction. The number of spikes per sweep evoked when an optimally oriented slit is moved in the direction opposite to that preferred by the cell has been measured. The evoked response in this case is 7 spikes per sweep on average for the enriched units and 4 spikes per sweep for the impoverished cells. This difference is slightly significant (P < 0.05). Size of the receptive field and eccentricity We have quantitatively estimated the size of the receptive field of visually responsive units by flashing bars at equally spaced intervals throughout the receptive field of the cell being studied (see Methods). The average size of the receptive fields determined by fitting a gaussian curve to the evoked response is 2.0 _+ 0.1 ° for EC units and 2.3 + 0.3 ° for IC cells (Fig. 8), No significant difference can be found in the two environmental groups. For orientation selective and biased cells, there are also no statistical differences in the size of the

78

SPONTANEOUS ACTIVITY

20~

Allunits

/~ _

1.70

Imo,~,~ [ao~a~ b~ laun ,

151~

/tq

1.85 1.23

o.,3

DISCUSSION

z

0

2

4 6 S Sp,'wnnaOUltactivity(~kea/sec)

18 l i

1'0

IMPOVERISHEDCATS Allmaim

~ Z

responsive cells could be activated by either eye. In facl. only 5% of the units in each condition have been classified as exclusively monocular (class 1 and 7 according to Hubel and Wiesel's scale). This indicates that the richness of the environment does not greatly affect the degree of binocular excitatory convergence onto visual cortical cells.

I I~ Orlmtatioubimed [MUntriented

6~

2.17 1.79 2.33

o s ~ i

~tivity(~e~ec)

Fig. 9. Frequency distribution of the spontaneous activity in enriched and impoverished cats. This parameter is not significantly affected by the richness of the environment for any of the groups of cells studied.

receptive field. Thus, the richness of the environment seems to have little effect on the size of the receptive field. The eccentricity of centers of the hand-plotted receptive fields of the cells studied in the present study were always lower than 10° from the area centralis. An analysis of variance demonstrated that the eccentricity of units was not significantly different in the two environmental conditions. The mean eccentricity was 5.1 ° for EC units and 5.6 ° for IC cells.

Spontaneous activity and binocularity The frequency distribution of the spontaneous activity of the cells studied in the two environmental conditions is presented in Fig. 9. The spontaneous activity of responsive units in enriched cats (1.70 spikes/s) is not significantly different from those in impoverished animals (2.17; P > 0.1 on a U-test and on an ANOVA). Additionally, no significant differences were found in any of the 3 classes of responsive cells in the two groups of cats. The spontaneous activity thus appears largely unaffected by the richness of the environment. In the two environmental conditions, most of the

The results obtained show that differences in orientation sensitivity and in responsivity are present in the visual cortex of cats raised in an impoverished condition compared to animals reared in an enriched milieu. Despite an abundant literature on anatomical and behavioral changes induced by the richness of the environment 3°'53, this is the first study which analyses the effects of environmental complexity on the physiology of the visual cortex. Our main result is that rearing a cat alone in a cage induced a greater incidence of units which were more broadly tuned for orientation and increased the incidence of unresponsive or poorly responsive visual cortex cells. The changes in receptive field properties induced by the richness of the environment are similar (though much less severe) than those obtained with dark-rearing or binocular eyelid suture 21'4~. However, it is difficult to compare our results with these studies because the impoverished condition does not imply a drastic modification of the retinal inputs as does visual deprivation. Recent evidence however makes it clear that more than simply retinal signals are important for the induction of experience dependent modifications 47-49. It is well known that, for instance, retinal signals cannot induce any changes in the visual cortex functions if the animals are paralyzed or anaesthetized 1°'~2"2°'47. In these cases, light stimuli elicit action potentials in visual cortex cells but no changes in ocular dominance or orientation selectivity can be obtained. In addition, retinal stimuli can induce experience dependent modifications in only one hemisphere after an unilateral lesion of the thalamic intralaminar nuclei 48'49. Other studies have pointed to the importance of motori interaction with the visual environment for experience dependent modifications. We thus suggest that since the amount of visual retinal signals is probably similar between our two groups of experimental animals, the observed differences in the physiology of the visual cortex cells are due mainly to the fact that enriched animals must pay greater attention to their visual inputs and use them more extensively for control of their behavior. This is consistent with evidence that non-retinal inputs can modulate visual cortex cell activity. In the present study it is difficult to determine which

79 non-retinal system is involved in the differential changes in receptive field properties. However, to date at least two different chemically-specific afferent systems have been implicated in the normal maturation of receptive field properties. These include (1) the cholinergic projection 47-49 from the midbrain reticular formation and (2) the noradrenergic afferents originating in the locus coereleus 31'32 (but see refs. 2, 18, 19). It has been demonstrated that the cholinergic system can exert facilitory actions on transmission in the striate cortex, associated with arousal 45'47. Moreover, Singer 47 demonstrated that the facilitory action on transmission in the striate cortex produced by the midbrain reticular formation stimulation, might involve cholinergic input to the cortex. Assuming that the physiological changes in enriched and impoverished visual cortex result from differences in attention and arousal mechanisms, it is possible that the cholinergic system may play an important role in producing the physiological differences reported in the present study. In addition, Hohman et a l Y have also demonstrated that in mouse, lesions of cholinergic nuclei in the basal forebrain lead to a rapid decrease of the amount of acetylcholine in the visual cortex followed later on, by a gradual decrease in the amount of G A D (GABA-synthesizing enzyme). Cholinergic afferences might thus play a role in the regulation of the intracortical GABAergic system. Cholinergic system inputs might not modify directly receptive field properties but would modulate excitability of the target cells. In fact, iontophorese applications of acetylcholine result in a variety of physiological changes, one of these being a slow hyperpolarization which facilitates signal to noise ratio in cortical cells. A powerful regulation of the GABAergic intracortical system by cholinergic afferents has been substantiated recently by the anatomical demonstration that cholinergic synapses show a strong preference to target GABAergic intracortical cells 6. This preference suggests that acetylcholine can modify receptive field properties via selective excitation of G A B A e r gic intracortical cells. Some evidence suggests that the noradrenergic afferent system from the locus coereleus may be also implicated in the normal and experience dependent maturation of visual cortex function 31'32. However, the involvement of the noradrenergic system in visual cortex properties is unclear with some unresolved issues remaining 2'18'~9. Whether the noradrenergic system is implicated or not in normal cortical maturation, it is possible that one of several other neurochemicallyspecific modulatory systems can cause a disruption of the intracortical GABAergic circuitry. Anatomical studies of the effects of differential rearing provide valuable data that appear to correlate with

physiological differences found in the present study. In the impoverished cortex, we demonstrate a population of units which are unresponsive to light stimulus. (For any orientation of a light slit, the cell's response is not increased to at least twice the standard deviation of the spontaneous activity.) This contrasts with enriched cortex in which all units are responsive. We have also demonstrated that impoverished cells that are responsive to light stimuli show a lower responsivity than enriched units. This difference is mostly due to a greater incidence of low responsive unoriented and orientation biased units in impoverished cortex. This decrease in the responsiveness can be explained by either lower excitation or greater inhibition of impoverished cortical neurons (or both). It is obviously very difficult to directly link physiological and anatomical data, especially in a complex system like the cat visual cortex. We believe however a possible link exists between anatomical studies on the visual cortex of enriched and impoverished cats and changes in the responsiveness of visual cortex. It is well known that variations in environmental complexity induce changes in the anatomy of the brain 233°'53. Differential changes range from modifications of the gross anatomy of the brain to the fine alterations at the synaptic level. At the synaptic level, there is an alteration of the number of asymmetric synapses per neuron, but their number per unit volume is unaffected 3"52. In addition, a higher number of symmetric synaptic contacts associated with flat vesicles per neuron (FS synapses) was found in visual cortex of impoverished cats 3. Since the vast majority of flat symmetrical (FS) synapses are GABAergic 5'5° and thus inhibitory 29"33"42 and assuming that round asymmetrical (RA) contacts are excitatory ~4, the higher number of FS synapses and the lower number of RA contacts on an average impoverished visual cortex cell suggest that the excitatory/inhibitory equilibrium is shifted towards more inhibition and less excitation on an average cortical unit. This could lead, in itself, to the observed decrease in responsivity of units in the impoverished visual cortex reported in the present study and is the hypothesis we currently favour. It must be noted however that these changes in the unit's responsivity can also be due to a change in the function of the lateral geniculate nucleus or even in the retina. However, no data are presently available on the physiology and on the anatomy of these structures in EC and IC cats. We also show that orientation sensitivity is clearly affected by the richness of the environment: in impoverished cats, the bandwidth of orientation tuning curves tends to be larger and a shift of the distribution of responsive cells toward a higher proportion of orientation biased and unoriented cells was observed. Thus, impoverished visual cortex has a greater incidence of cells

80 which are m o r e b r o a d l y tuned for orientation than in the enriched animals. Since cortical orientation sensitivity is specified at least in p a r t by intracortical circuits and is G A B A d e p e n d e n t 7' 43,44,46 , we believe that there is also a possible link b e t w e e n physiological differences induced by the richness of the e n v i r o n m e n t and a structural change in the inhibitory G A B A e r g i c connections of the visual cortex, in the following m a n n e r : the n u m b e r and size of flat symmetrical synapses (which are mostly G A B A e r g i c ) , are altered by the richness of the environment. In i m p o v e r i s h e d cats, the area of each synaptic contact is smaller and the n u m b e r of FS contacts per bouton is higher than in the enriched cats 3'4. The smaller contact area implies both a n a r r o w e r release site and a smaller n u m b e r of r e c e p t o r s on the postsynaptic side, suggesting that each synapse is less effective. The higher n u m b e r of contacts p e r b o u t o n in an impoverished c o m p a r e d to an e n r i c h e d cortex implies a different a r r a n g e m e n t of the contact sites: for e x a m p l e , an average enriched terminal b o u t o n forming 2 large contacts on only two postsynaptic e l e m e n t s might form 3 small contacts on 3 postsynaptic e l e m e n t s in i m p o v e r i s h e d animals. Clearly, such an a r r a n g e m e n t could result in less selective inhibition of the key postsynaptic elements. It might be that this loss of selectivity of inhibition would result in m o r e broadly t u n e d cells in the i m p o v e r i s h e d visual cortex. Again, these suggestions do not exclude the possibility that the lateral geniculate nucleus and/or the retina is affected and REFERENCES 1 Baker, C.L. and Cynader, M., Spatial recePtive field properties of direction selective neurons in cat striate cortex, J. Neurophysiol., 55 (1986) 1136-1152. 2 Bear, M.E and Singer, W., Modulation of visual cortical plasticity by acetylcholine and noradrenaline, Nature (Lond.), 320 (1986) 172-176. 3 Beaulieu, C. and Colonnier, M., The effect of the richness of the environment on cat visual cortex, J. Comp. Neurol., 266 (1987) 478-494. 4 Beaulieu, C. and Colonnier, M., Richness of environment affects the number of contacts formed by boutons containing fiat vesicles but does not alter the number of these boutons per neuron, J. Comp. Neurol., 274 (1988) 347-356. 5 Beaulieu, C. and Somogyi, P., Targets and quantitative distribution of GABAergic synapses in the visual cortex of cat, Eur. J. Neurosci., in press. 6 Beaulieu, C. and Somogyi, P., Neurochemical properties and postsynaptic targets of cholinergic synapses in cat visual cortex, Soc. Neurosci. Abstr., 15 (1989) 1107. 7 Benevento, L.A., Creutzfeldt, O.D. and Kuhnt, U., Significance of intracortical inhibition in the visual cortex, Nature New Biol., 238 (1972) 124-126. 8 Blakemore, C. and Van Sluyters, R.C., Innate and environmental factors in the development of kitten's visual cortex, J. Physiol. (Lond.), 248 (1975) 663-716. 9 Bonds, A.B., Development of orientation tuning in the visual cortex of kittens. In R.D. Freeman (Ed.), Developmental Neurobiology of Vision, Vol. 27, Plenum, New York, 1979, pp.

might contribute to some extent to the physiological changes observed. Changes in responsiveness and in receptive field properties similar to those found in the impoverished cortex were also described in i m m a t u r e and visually deprived cortical cells c o m p a r e d with ' n o r m a l ' mature cellsS.9,11,21,37.40. Because most of the visually responsive cells in visually deprived cats have no orientation selectivity, even though their responses to visual stimuli were clearly e n h a n c e d by removal of G A B A - m e d i a t e d inhibition with bicuculline, Tsumoto and F r e e m a n 5~ suggest that two different types of inhibitory synapses must be present in the visual cortex: one of these would be r e t a r d e d or dysfunctional as a consequence of deprivation while the o t h e r would be developed. We suggest that the impoverished cortex is less responsive because there are m o r e inhibitory synapses p e r neuron. The loss of selectivity is achieved, in spite of this increase, because the n u m e r o u s smaller synaptic contacts are not as selectively distributed on key postsynaptic elements. If this suggestion is correct, visual deprivation m a y cause changes in orientation and direction selectivity and in responsivity by affecting the same inhibitory G A B A e r g i c system as that considered here. Acknowledgements. This work was supported by a grant from the Medical Research Council of Canada to M.C. (PG-29) and by a postdoctoral fellowship (C.B.) from the Fonds de la recherche en sant6 du Qu6bec. We wish to thank Dr. M. Colonnier for his comments on the paper. 31-49. 10 Buisseret, P., Gary-Bobo, E. and Imbert, M., Ocular motility and recovery of orientational properties of visual cortical neurones in dark-reared kittens, Nature (Lond.), 272 (1978) 816-817. 11 Buisseret, P. and M.I., Visual cortical cells: their developmental properties in normal and dark-reared kittens, J. Physiol. (Lond.), 255 (1976) 511-525. 12 Buisseret, P. and Singer, W., Proprioceptive signals from extraocular muscles gate experience-dependent modifications of receptive fields in the kitten visual cortex, Exp. Brain Res., 51 (1983) 443-450. 13 Burschfiel, J.L. and Duffy, F.H., Role of intracortical inhibition in deprivation amblyopia: reversal by microiontophoresis of bicuculline, Brain Res., 206 (1981) 479-484. 14 Colonnier, M., The electron microscopic analysis of the neuronal organization of the cerebral cortex. In F.O. Schmitt, EG. Worden and S.D. Dennis (Eds.), Organization of the Cerebral Cortex, MIT Press, Cambridge, 1981, pp. 125-151. 15 Cynader, M., Berman, N. and Hein, A., Cats raised in one directional world: effects on receptive fields in visual cortex and superior colliculus, Exp. Brain Res., 22 (1975) 267-280. 16 Cynader, M., Berman, N. and Hein, A., Recovery of function in cat visual cortex following prolonged visual deprivation, Exp. Brain Res., 25 (1976) 139-156. 17 Cynader, M. and Chernenko, G., Abolition of direction selectivity in the visual cortex of the cat, Science, 193 (1976) 504-505. 18 Daw, N.W., Robertson, T.W., Rader, R.K., Videen, T.O. and Coscia, C.J., Substantial reduction of cortical noradrenaline by lesions of adrenergic pathway does not prevent effects of

81 monocular deprivation, J. Neurosci., 4 (1984) 1354-1360. 19 Daw, N.W., Videen, T.O., Parkinson, D. and Rader, R.K., (N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine) depletes noradrenaline in kitten visual cortex without altering the effects of monocular deprivation, J. Neurosci., 5 (1985) 1925-1933. 20 Freeman, R.D. and Bonds, A.B., Cortical plasticity in monocularly deprived immobilized kittens depends on eye movements, Science, 206 (1979) 1093-1095. 21 Fr6gnac, Y. and Imbert, M., Early development of visual cortical cells in normal and dark-reared kittens: relationship between orientation selectivity and ocular dominance, J. Physiol. (Lond.), 278 (1978) 27-44. 22 Freund, T.E, Martin, K.A.C., Smith, A.D. and Somogyi, P., Glutamate decarboxylase-immunoreactive terminals of Golgiimpregnated axoaxonic cells and on presumed basket cells in synaptic contact with pyramidal neurons of the cat's visual cortex, J. Comp. Neurol., 221 (1983) 263-278. 23 Greenough, W.T., Enduring brain effects of differential experiences and training. In M.R. Rosenzweig and E.L. Bennett (Eds.), Neural Mechanisms of Learning and Memory, MIT Press, Cambridge, 1976, pp. 255-277. 24 Hirsch, H.V.B. and Spinelli, D.N., Visual experience modifies distribution of horizontally and vertically oriented receptive fields in cats, Science, 168 (1970) 869-871. 25 Hohmann, C.E, Bear, M.E and Ebner, F.E, Glutamic acid decarboxylase activity decreases in mouse neocortex after lesions of the basal forebrain, Brain Res., 333 (1986) 165-168. 26 Hubel, D.H. and Wiesel, T.N., Receptive fields, binocular interaction and functional architecture in the cat's visual cortex, J. Physiol. (Lond.), 160 (1962) 106-154. 27 Hubei, D.H. and Wiesel, T.N., Binocular interaction in striate cortex of kittens reared with artificial squint, J. Neurophysiol., 28 (1965) 1041-1059. 28 Imbert, M. and Buisseret, P., Receptive field characteristics and plastic properties of visual cortical cells in kittens reared with or without visual experience, Exp. Brain Res., 22 (1975) 25-36. 29 Iversen, L.L., Mitchell, J.E and Srinivasan, V., The release of gamma aminobutyric acid during inhibition in the cat visual cortex, J. Physiol. (Lond.), 212 (1971) 519-534. 30 Jones, D.G. and Smith, B.J., The hippocampus and its response to differential environment, Prog. Neurobiol., 15 (1980) 19-69. 31 Kasamatsu, T. and Pettigrew, J.D., Depletion of brain catecholamines: failure of ocular dominance shift after monocular occlusion in kittens, Science, 194 (1976) 206-209. 32 Kasamatsu, T., Pettigrew, J.D. and Ary, M.L,, Restoration of visual cortical plasticity by local microperfusion of N.E., J. Comp. Neurol., 185 (1979) 163-182. 33 Krnjevic, K. and Schwartz, S., The action of gamma aminobutyric acid on cortical neurons, Exp. Brain Res., 3 (1967) 320-336. 34 Leventhal, A.G. and Hirsch, H.V.B., Receptive field properties of different classes of neurons in visual cortex of normal and dark-reared cats, J. NeurophysioL, 43 (1980) 1111-1132. 35 Mower, G.D., Christen, W.G. and Duffy, EH., Evidence of an enhanced role of G A B A inhibition in amblyopic cats, Soc. Neurosci. Abstr., 11 (1985) 463. 36 Pearson, H.E., Frequency specific effects of stroboscopic rearing in the visual cortex of the rabbit, Dev. Brain Res., 7 (1983) 187-196. 37 Pettigrew, J.D., The effect of visual experience on the development of stimulus specificity by kitten cortical neurons, J.

Physiol. (Lond.), 237 (1974) 49-74. 38 Saul, A. and Daniels, J.D., Personal communication. 39 Schiller, P.H., Finlay, B.L. and Volman, S.F., Short-term response variability of monkey striate neurons, Brain Res., 105 (1976) 347-349. 40 Sherk, H. and Stryker, M.P., Quantitative study of cortical orientation selectivity in visually inexperienced kitten, J. Neurophysiol., 39 (1976) 63-70. 41 Sherman, S.M. and Spear, P.D., Organization of visual pathways in normal and visually deprived cats, Physiol. Rev., 62 (1982) 738-755. 42 Sillito, A.M., The contribution of inhibitory mechanisms to the receptive field properties of neurons in the striate cortex of the cat, J. Physiol. (Lond.), 250 (1975) 305-329. 43 Sillito, A.M., The effectiveness of bicuculline as an antagonist of GABA and visually evoked inhibition in the cat's striate cortex, J. Physiol. (Lond.), 250 (1975) 287-304. 44 Sillito, A.M., Inhibitory processes underlying the directional specificity of simple, complex and hypercomplex cells in cat's visual cortex, J. Physiol. (Lond.), 271 (1977) 699-720. 45 SiUito, A.M. and Kemp, J.A., Cholinergic modulation of the functional organization of the cat visual cortex, Brain Res., 289 (1983) 143-155. 46 Sillito, A.M., Kemp, J.A., Wilson, J.A. and Berardi, N,, A re-evaluation of the mechanisms underlying simple cell orientation selectivity, Brain Res., 194 (1980) 517-520. 47 Singer, W., Central core control of visual cortex functions. In EO. Smitt and EG. Worden (Eds.), The Neurosciences, Fourth Study Program, MIT Press, Cambridge, 1979, pp. 1093-1109. 48 Singer, W., Central core control of developmental plasticity in the kitten visual cortex: I. Diencephalic lesions, Exp. Brain Res., 47 (1982) 209-222. 49 Singer, W. and Rauschecker, J.P., Central core control of developmental plasticity in the kitten visual cortex: II. Electrical activation of mesencephalic and diencephalic projections, Exp. Brain Res., 47 (1982) 223-233. 50 Somogyi, P. and Hodgson, A.J., Antiserum to y-aminobutyric acid: III. Demonstration of GABA on Golgi-impregnated neurons and in conventional electron microscopy sections of cat's striate cortex, J. Histochem. Cytochem., 33 (1985) 249-257. 51 Tsumoto, T. and Freeman, R.D., Dark-reared cats: responsivity of cortical cells influenced pharmacologically by an inhibitory mechanism, Exp. Brain Res., 65 (1987) 666-672. 52 Turner, A.M. and Greenough, W.T., Differential rearing effects on rat visual cortex synapses. I - - Synaptic and neuronal density and synapses per neuron, Brain Res., 329 (1985) 195-203, 53 Walsh, R.N., Effects of environmental complexity and deprivation on brain anatomy and histology: a review, Int. J. Neurosci., 12 (1981) 33-51. 54 Wiesel, T.N. and Hubel, D.H., Single cell responses in striate cortex of kittens deprived of vision in one eye, J. Neurophysiol., 26 (1963) 1003-1017. 55 Wiesel, T.N. and Hubel, D.H., Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kittens, J. Neurophysiol., 28 (1965) 1029-1040. 56 Wolff, J.R., Balcar, V.J., Zetzsche, T., Bottcher, H., Schmechel, D.E. and Chronwall, B.M., Development of GABA-ergic system in rat visual cortex. In J.M. Lauder and P.G. Nelson (Eds.), Gene Expression and Cell-Cell Interaction, Plenum, New York, 1984, pp. 215-239.