Ocular dominance distribution of cells in area 21a correlates with perceptual deficits in amblyopic...

Schröder, J. H., Fries, P., Roelfsema, P. R., Singer, W., Engel, A. K. (2002) Ocular dominance in extrastriate cortex of strabis-mic amblyopic cats. Vision Research 42(1), 29-39. doi: 10.1016/S0042-6989(01)00263-2

1

Published in final edited form as:

Vision Research 42(1), 29-39.

doi: 10.1016/S0042-6989(01)00263-2

Ocular dominance in extrastriate cortex of strabismic amblyopic cats

Jan-Hinrich Schröder

a,*, Pascal Fries

a,b,1, Pieter R. Roelfsema

a,2, Wolf Singer

a

and Andreas K. Engela,3

a

Max Planck Institute for Brain Research, Deutschordenstraße 46, 60528 Frankfurt, Germany b

Department of Psychiatry I, Johann Wolfgang Goethe University, Heinrich-Hoffmann-Straße 10,

60528 Frankfurt, Germany

Abstract

Ocular dominance in extrastriate visual cortex of cats with behaviorally defined strabismic

amblyopia was studied using extracellular recording techniques. In area 18, the amblyopic eye

drove about as many cells as the normal one. In area posteromedial lateral suprasylvian area

(PMLS), about 60% of the cells responded exclusively to stimulation of the normal eye and

30% to stimulation of the amblyopic eye. In area 21a more than 75% of the cells were mo-

nocularly driven by the non-amblyopic eye while only 5% were monocularly driven by the

amblyopic eye. These findings suggest that ventral pathways (area 21a) are more affected in

amblyopia than dorsal pathways (area PMLS).

Keywords

Amblyopia; Area 21a; Extrastriate; Ocular dominance; Strabismus


2

Amblyopia is defined by a decrease of visual acuity for which no organic causes can be de-

tected by physical examination of the eye. In the case of strabismus, amblyopia results if sub-

jects continuously suppress the signals arriving from the same eye in order to avoid double

vision (e.g. von Noorden, 1990). In addition to the reduction in visual acuity, strabismic am-

blyopia is associated with a variety of other deficits. The image conveyed by the amblyopic

eye is temporally unstable (Altmann & Singer, 1986; Hess, Campbell, & Greenhalgh, 1978a)

and spatially distorted (Lagrèze & Sireteanu, 1991; Sireteanu, Lagrèze, & Constantinescu,

1993). Subjects tend to mislocate targets (Hess & Holliday, 1992) and make false conjunc-

tions between stimulus elements (Hess et al., 1978a). Another deficit typically encountered in

this condition is the so-called crowding phenomenon, i.e., a deterioration of the ability to dis-

criminate figures if they are surrounded by other contours (e.g. von Noorden, 1990).

Evidence from psychophysics has led to the conclusion that the deficit is situated in the visual

cortex (Hess & Pointer, 1985; Levi & Klein, 1985). In a more detailed search for the physio-

logical basis of amblyopia, several animal models have been investigated. Hubel and Wiesel

(1965a) reported that following induction of strabismus cells in cat area 17, the primary visual

cortex, had become monocular and that the two eyes were represented about equally in each

hemisphere. Later studies confirmed this result but some of them revealed an additional slight

bias in ocular dominance (OD) favoring the non-operated eye (Berman & Murphy, 1982;

Freeman & Tsumoto, 1983; Sireteanu & Best, 1992; Yinon, Auerbach, Blank, & Friesen-

hausen, 1975). Various studies have reported decreased spatial resolution of striate cortical

neurons in strabismic cats (Chino, Shansky, Jankowski, & Banser, 1983; Crewther &

Crewther, 1990). Taken together, the correlates of amblyopia in area 17 were subtle and not

consistent across studies. Note that in this respect strabismic amblyopia strongly differs from

the form caused by monocular deprivation, which leads to a almost complete decoupling of

the deprived eye (Hubel & Wiesel, 1970) and a profound loss of visual function.

In the monkey, the search for a cortical substrate of strabismic amblyopia has been confined

to area V1. Kiorpes, Kiper, O'Keefe, Cavanaugh, and Movshon (1998) found that the two

eyes were about equally effective in driving cortical cells but neurons driven by the amblyop-

ic eye had reduced spatial resolution. However, optimal spatial frequency and peak contrast

sensitivity of these neurons were better than expected from behavioral examination. From this

the authors concluded that a full account of the physiological basis of amblyopia would re-

quire the investigation of processing stages beyond striate cortex (for a review see Kiorpes &

McKee, 1999). In the present study, we have therefore measured neuronal response properties

in various areas of the visual cortex of strabismic cats which had developed amblyopia.

Two previous studies have examined the effect of strabismus on neuronal response synchro-

nization in the striate cortex of strabismic cats. The first one has demonstrated that in exo-

tropic cats synchronization between monocular cells connected to different eyes is drastically

reduced in comparison to synchronization among cells driven from the same eye (König, En-

gel, Löwel, & Singer, 1993). Synchronization among binocular cells could not be measured

due to the low incidence of these cells in strabismic animals. The second study, conducted in

esotropic amblyopic cats, revealed that response synchronization is reduced among neurons


3

driven by the amblyopic eye as compared to those driven by the normal eye, in particular

when responses are evoked by gratings with high spatial frequency (Roelfsema, König, Engel,

Sireteanu, & Singer, 1994). Quite unexpectedly, no significant differences were found in the

spatial resolution and response vigor of cells driven by the normal and the amblyopic eye,

suggesting reduced synchronization as a major cause of the amblyopic deficit. It has been

proposed that synchronization raises the saliency of responses due to enhanced summation of

excitatory synaptic potentials in target cells (Niebur, Koch, & Rosin, 1993), and recent exper-

imental (Alonso, Usrey, & Reid, 1996; Brecht, Singer, & Engel, 1998) and theoretical (Lu-

mer, Edelman, & Tononi, 1997; Stevens & Zador, 1998) evidence is compatible with this

suggestion. This predicts that changes in response synchronization at one level of processing

lead to changes in discharge rate at the next processing stage. The amblyopic deficit should

then be associated with reduced responses to the amblyopic eye at processing stages beyond

primary visual cortex. Here we examine this prediction by analyzing responses in areas 18

and 21a and posteromedial lateral suprasylvian (PMLS) of amblyopic cats. Preliminary data

have been published in abstract form (Schröder, Fries, Roelfsema, Singer, & Engel, 1998).

2. Methods

All experimental procedures comply with European and NIH standards on the welfare of ani-

mals. This report is based on six cats in which convergent strabismus had been induced under

general ketamine/xylazine anesthesia (10 and 2 mg/kg, respectively) by transecting the lateral

rectus muscle of the right eye at an age of three weeks. The angle of the resulting convergent

squint was determined repeatedly during development using the corneal reflex method (Sher-

man, 1972; Von Grünau, 1979). Several flashlight snapshots of the animal's head were taken

and the ratio of the distance of the corneal reflexes over the distance of the pupils was deter-

mined on the photoprints. This ratio provides a measure of the strabismic angle.

At the age of 4–5 months, the animals were mildly food deprived and trained to discriminate

between a square wave grating and equiluminant gray (Teller Acuity Cards, contrast 82–84%,

luminance 25 cd/m2) on a jumping stand (Katz & Sireteanu, 1992; Mitchell, Griffin, Wil-

kinson, Anderson, & Smith, 1976). Jumps to the grating were rewarded (correct response). A

session continued until the cat stopped jumping spontaneously. After sufficient training, cats

were tested through the normal and the deviating eye on alternate days, the respective other

eye being occluded by an opaque contact lens during testing. Thus, it was possible to assess

visual acuity for the two eyes independently by variation of the spatial frequency of the grat-

ings (0.21–14.2 cycles per degree with intervals of 0.5 octaves). Each eye was tested on at

least three different days with a minimum of 180 jumps altogether. The spatial frequency of

the cards was continuously adjusted to the performance of the animal.

Discrimination threshold was defined as the spatial frequency at which the animal performed

at the 75% level (chance level: 50%). The psychometric functions resulting from the behav-

ioral testing were fitted by the function

( ) [ ( ) ] , where P denotes performance, x spatial frequency, a discrimination threshold and b slope at

threshold. For the discrimination thresholds of the two eyes 95% confidence intervals were

calculated using a Monte Carlo simulation. Animals were considered to be amblyopic if the


4

discrimination thresholds of the two eyes differed by at least half an octave and if the confi-

dence intervals for the respective discrimination thresholds were non-overlapping. For subse-

quent physiological measurements we selected four cats which were amblyopic according to

these criteria. The animals included in this study were chosen from a larger cohort of ambly-

opic cats and have been selected for similar and relatively small differences in acuity between

the two eyes which corresponds to mild amblyopia. Two additional strabismic cats that had

been classified as non-amblyopic after behavioral testing served as controls.

We used standard surgical and recording techniques. Anesthesia was induced by intramuscu-

lar injection of ketamine and xylaxine (10 and 2 mg/kg, respectively) and maintained by ven-

tilating the animal with a mixture of 70% N2O, 30% O2 and 0.8–1.2% halothane. After crani-

otomy and induction of paralysis (pancuronium bromide, 0.15 mg/(kg h)), the activity of sin-

gle cells and small multiunit clusters was recorded extracellularly in areas 18, 21a and the

PMLS area using tungsten microelectrodes. The areas were defined according to the stereo-

tactic coordinates of Tusa and coworkers (Palmer, Rosenquist, & Tusa, 1978; Tusa & Palmer,

1980; Tusa, Rosenquist, & Palmer, 1979). In most of the experiments, electrode tracks were

reconstructed histologically after transcardial perfusion and Nissl staining.

The electrode signals were amplified, band-pass filtered and fed through a Schmitt trigger to

obtain TTL pulses which signaled spike timing. Receptive fields (RFs) and neuronal response

properties were assessed with hand-held stimuli. In addition, the OD was examined quantita-

tively by monocular stimulation of the two eyes with moving light bars of optimal length,

width, orientation, direction and velocity. In cases where mapping with hand-held stimuli re-

vealed a clear response upon stimulation of one of the eyes but failed to reveal any response

through the respective other eye, quantitative measurements were omitted.

In order to classify recording sites with respect to OD we defined five OD classes. OD 1 and 5

were assigned to monocular responses obtained by left or right eye stimulation, respectively.

Binocular responses were classified as OD 3. If a given recording site responded to stimula-

tion through one eye at least twice as strongly as to stimulation through the other eye it was

assigned OD 2 or 4, depending on which eye was dominating. The significance of the most

prominent feature of the OD distributions, their bias favoring one of the two eyes, was tested

against the null hypothesis of an equal distribution of left and right eye monocular cells by

applying a test based on the cumulative binomial distribution. For this purpose, OD distribu-

tions were simplified by dividing recording sites in only two groups: sites dominated by the

left (OD 1 and 2) and the right (OD 4 and 5) eye. Because of the very low incidence (about

5%) of balanced binocular recording sites (OD 3) in our cats it seemed justified to exclude

these from statistical analysis. All p values given in this paper refer to this significance test.

3. Results

3.1. Squint induction

The corneal reflex method indicated that induction of convergent strabismus had been effec-

tive in all our animals. The ratio of the distance of the pupils over the distance of the reflexes

ranged from 0.97 to 1.03 (Table 1) which has to be compared to 0.954±0.007 (SD), the value

for normal animals more than two months old (Von Grünau, 1979).


5

Table 1 Behavioral data of the six cats used in the study

Cat Ratio (reflex-es/pupils)

Threshold of the non-deviated eye

Threshold of the deviated eye

Acuity differ-ence (oc-taves)

1 1.03±0.01 2.07 (1.66–2.79) 1.19 (0.87–1.86) 0.80 2 0.99±0.01 2.19 (1.91–2.71) 1.36 (1.2–1.6) 0.69 3 1.01±0.002 1.68 (1.36–2.23) 1.04 (0.88–1.36) 0.69 4 1.02±0.005 1.47 (1.35–1.65) 0.92 (0.79–1.07) 0.68 5 1.00±0.03 2.35 (1.91–3.13) 2.86 (2.3–3.93) -0.28 6 0.97±0.005 2.1 (1.7–2.6) 1.9 (1.6–2.4) 0.14

In the second column, the ratios of reflex distance over pupil distance on the photoprints are reported. In the third and fourth column the acuities measured as discrimination thresholds in the jumping stand procedure are given in cycles per degree for the left (non-deviated) and the right (deviated) eye. The numbers in brackets indicate 95% confidence intervals as de-termined by a Monte Carlo simulation. The difference in acuity is expressed in octaves in the fifth column.

3.2. Assessment of amblyopia

The behavioral performance of cats 1–4 is shown in Fig. 1. All animals were able to discrimi-

nate higher spatial frequencies with the non-operated eye than with the deviating eye. The

discrimination thresholds differed by about 0.7 octaves. With the exception of cat 1, the 95%

confidence intervals of the discrimination thresholds were non-overlapping. Thus, cats 2, 3

and 4 satisfied our criteria of significance (Table 1) and were classified as amblyopic. It might

be argued that cat 1 cannot be considered amblyopic because of the slight overlap of the con-

fidence intervals. However, as the OD of recording sites in cat 1 was consistent with those in

cats 2–4 we decided not to exclude these data from our sample. In the control animals (cats 5

and 6), by contrast, the interocular difference in visual acuity was much smaller than in cats

1–4 and far from significant (see Fig. 4a and b).


6

Figure 1 Behaviorally determined visual acuity of amblyopic cats. Performance with the non-operated left eye and the operated right eye is shown by squares and triangles, respectively. Error bars indicate the standard error of the mean. Performance was fitted by a sigmoid function. The dashed line refers to the 75% level of performance, which was taken as a criterion for the discrimination threshold. Light and dark rectangles on the abscissa denote the 95% confidence intervals of discrimination thresholds for the non-operated and the operated eye which were determined by a Monte Carlo simulation.

3.3. Ocular dominance

Data were obtained from 511 multi-unit clusters or single units along 109 independent pene-

trations distributed over three areas (18, PMLS, 21a) in all eight hemispheres of the four am-

blyopic cats. RFs were typically located within the central 15° of visual angle. In most cases

cells responded strongly to visual stimuli and, therefore, orientation preference, direction

preference and, most importantly, OD could be determined reliably. Parameters like RF loca-

tion and size, orientational tuning for light bars, preferred direction of moving stimuli and OD

typically changed along the electrode tracks, sometimes continuously, sometimes abruptly.

This reflects the fact that electrodes penetrated obliquely to the cortical surface, thus traveling

through different columns. An example of the RF sequence in a typical penetration is illus-

trated in Fig. 2.


7

Figure 2 Example of a sequence of RFs along an electrode track in area 21a, left hemisphere. Numbers in the

RFs denote the penetration depth of the electrode and the OD at the penetration site. Arrows indicate the pre-

ferred direction of movement of a visual stimulus. Note the variability of RF properties and OD along the elec-

trode track. ACL and ACR indicate the projection of the area centralis of the left and right eye, respectively. The

distance of the areae centrales was much larger than shown here: After induction of paralysis this cat (3) showed

a squint angle of about 31°. The inset on the lower right illustrates the changes of OD along the recording tracks.

In general, about 80% of the recording sites in all areas under study responded monocularly

(Fig. 3a). In area 18 of the amblyopic cats 1–4, the vast majority (approximately 75%) of the

cells in both hemispheres responded exclusively to visual stimulation of the contralateral eye

(Fig. 3a). Superimposed on this effect was a weak bias towards the normal eye (the left eye in

all cats). In the hemisphere contralateral to the normal eye (right hemisphere) this bias was

adding to and in the other hemisphere it was reducing the prevalence of the respective contra-

lateral eye.

Figure 3 OD distributions summed up across cats 1–4. (a) Numbers on the ordinate indicate the per-centage of cells in each OD class. L and R, left and right hemisphere. Note the low incidence of binoc-


8

ular responses. In areas 18 and PMLS the respective contralateral eye is overrepresented, while in area 21a there is a bias favoring the left, non-amblyopic eye. (b) Control for the independence of sites recorded within the same penetration (details see Section 3). Note the similarity between the OD graphs in a and b.

The relative strength of these two effects was somewhat shifted in area PMLS, the trend to-

wards the normal eye being slightly more pronounced than in area 18. The normal eye domi-

nated all of the monocular cells in the contralateral (right) hemisphere and about 40% of them

in the ipsilateral (left) hemisphere (Fig. 3a).

In area 21a, we obtained a strikingly different result: In both hemispheres, more than 75% of

the cells responded exclusively to visual stimulation of the normal (left) eye (OD class 1)

without displaying any response to stimulation of the amblyopic eye. In this area, only 5% of

the recording sites were monocularly driven by the amblyopic (right) eye (OD class 5). The

remaining 20% of the neurons in area 21a responded binocularly (Fig. 3a).

The OD distributions of area 18 (left and right hemisphere), area PMLS (right) and area 21a

(left and right) differed from an even distribution at the level of p<0.001 and the distribution

in area PMLS of the left hemisphere at a level of p=0.024.

In area 18 (Cynader, Gardner, & Mustari, 1984) and in lateral suprasylvian cortex (Sireteanu

& Best, 1992) of strabismic cats some of the few remaining binocular cells have been found

to exhibit anomalous retinal correspondence. The retinal positions of the RFs in the eyes were

shifted so that they remained superimposed despite the strabismic angle. We have examined

our data for the occurrence of this rare phenomenon but found no case of anomalous corre-

spondence in any of the analyzed areas. A likely reason is that the animals examined here had

developed amblyopia which implies that they had a strong preference for one eye and did not

strive to preserve binocular functions. Other reasons could be differences in the amplitude of

squint angle and in eye motility between the present and those previous studies.

3.4. Controls

For the quantification of OD distributions we counted the recording sites with a certain OD

across all penetrations and expressed these numbers in percent of the total number of sites

analyzed in a particular area and hemisphere (Fig. 3a). To control for the independence of the

results from different sites along the same electrode tracks, we also determined separately OD

distributions for single electrode tracks and normalized them to the number of analyzed sites.

Subsequently, distributions were calculated across penetrations and across all amblyopic cats

(cats 1–4) by summing up the normalized single-track distributions, thereby assigning the

same weight to each penetration. OD distributions obtained in this way did not differ substan-

tially from those obtained with the standard procedure (Fig. 3b). Thus, results from different

recording sites can be regarded as being sufficiently independent from each other.

In all amblyopic cats (1–4) it was the operated eye that had become amblyopic. In order to

exclude an effect of surgery per se on OD distributions we examined two cats (5 and 6) which

had undergone unilateral surgery like the others but had not become amblyopic (Fig. 4a and

b). These cats, too, exhibited a significant overrepresentation of the non-operated eye in area

21a of both hemispheres (p<0.01; Fig. 4c). However, in both cats, this OD shift was much


9

less pronounced than in the amblyopic animals (Fig. 4d). The proportion of cells dominated

by the operated eye (OD 4 and 5) was about three times the one found in amblyopic cats (28%

vs. 9%). From our results in cats 1–4 we were able to predict the proportions of cells that

should have been dominated by the operated and the normal eye in cats 5 and 6 assuming the

null hypothesis that the latter had exhibited the same degree of amblyopia. The deviation of

the measured OD distributions from this prediction was highly significant (p<0.001), suggest-

ing that the skewed OD in area 21a reflects the degree of amblyopia and not only unilateral

surgery.

Figure 4 Visual acuity and OD distribution in area 21a in cats 5 and 6, classified as non-amblyopic. (a, b) Visual acuity of the two eyes (conventions as in Fig. 1). These cats were classified non-amblyopic because in both of them the acuity of their two eyes differed by less than 0.3 octaves and the 95% confidence intervals were largely overlapping. (c) OD distributions from area 21a. Note the relatively weak bias in favor of the non-operated eye in both hemispheres. (d) Comparison between non-amblyopic cats 5 and 6 and amblyopic cats 1–4. OD distributions for area 21a were summed up across the two hemispheres. The overrepresentation of the left, non-operated eye was significantly more expressed in the amblyopic cats than in non-amblyopic cats 5 and 6.

In all cases where histological reconstructions of electrode tracks were obtained these con-

firmed that recording sites had indeed been located in the targeted areas.


10

4. Discussion

4.1. Visual acuity

The visual acuity values obtained for our cohort of esotropic cats are similar to those deter-

mined in several previous studies (Holopigian & Blake, 1983; Jacobson & Ikeda, 1979; Mow-

er & Duffy, 1983; Von Grünau & Singer, 1980). The acuity values for the non-deviating eye

(averaging 1.94 cycles/degree) were lower than values reported for a group of normal cats that

had been tested in the same apparatus (3.1 cycles/degree in Katz & Sireteanu, 1992). This is

in accordance with other published demonstrations of reduced visual acuity in strabismic an-

imals (Holopigian & Blake, 1983). However, the acuity values reported in the present study

are also below those found by Crewther, Crewther, and Cleland (1985) and Mitchell, Ruck,

Kaye, and Kirby (1984) in cats with strabismic amblyopia. With all likelihood, this difference

can be ascribed to differences in the testing and reinforcement paradigms. We varied spatial

frequencies rapidly and always included suprathreshold stimuli, while Crewther et al. (1985)

and Mitchell et al. (1984) presented the same spatial frequencies on blocks of trials, allowing

the animal to adjust its effort to the task difficulty.

4.2. Ocular dominance in amblyopic and normal cats

In this study OD was assessed from multi-unit rather than single-unit data and this has most

likely led to an overestimation of binocular units. A recording site got only classified as mo-

nocular if all of the simultaneously recorded neurons were monocular. By contrast, only one

or a few binocular cells would have led to a classification of the site as binocular even if the

majority of the cells had been monocular.

In the present study, most neurons in the examined areas were monocular, which is consistent

with earlier findings in areas 18 (Chino, Ridder, & Czora, 1988; Cynader et al., 1984) and

PMLS (Sireteanu & Best, 1992) of strabismic cats. In contrast to area 17, where OD distribu-

tions tend to be U shaped in both hemispheres of strabismic cats (e.g. Hubel & Wiesel,

1965a), OD in area 18 exhibited a strong bias towards the respective contralateral eye, that

was superimposed on a weak trend towards the normal eye. Results in area PMLS resembled

those in area 18 except that the trend favoring the non-amblyopic eye was more pronounced.

This is in accordance with the results of Sireteanu and Best (1992), who had examined

prestriate areas in strabismic cats, however, without testing for amblyopia. Von Grünau

(1982), by contrast, reported that binocularity was largely conserved in the LS areas of stra-

bismic cats. The reason for this discrepancy is most likely that in his study squint was induced

at the age of six to nine weeks rather than at three weeks as in our study.

In sharp contrast to the OD distributions in areas 17 (e.g. Roelfsema et al., 1994) and 18

which are either U shaped or biased towards the contralateral eye, the distributions in area 21a

were strongly biased towards the normal eye. As suggested by the much more symmetrical

distribution in the squinting control cats that had no interocular difference in visual acuity,

this marked loss of responses to stimulation of the deviated eye is most likely due to the de-

velopment of amblyopia. Our data do not allow us to determine whether the weak bias in the


11

control cats is related to mild amblyopia that was not accompanied by detectable acuity

changes or whether it is solely related to strabismus.

OD differs across areas also in normal cats. First, the percentage of monocular cells decreases

from 15–30% to 5–10% as one proceeds from primary areas 17 (Dreher, Michalski, Ho, Lee,

& Burke, 1993; Hubel & Wiesel, 1965a; Sireteanu & Best, 1992) and 18 (Cynader et al.,

1984; Dreher et al., 1993; Hubel & Wiesel, 1965b) to prestriate areas PMLS (Hubel &

Wiesel, 1969; Sireteanu & Best, 1992; Spear & Baumann, 1975) and 21a (Dreher et al.,

1993). Second, while the two eyes are represented equally in area 17, the contralateral eye

dominates the majority of the cells in areas 18 and PMLS. In area PMLS, there is an addition-

al eccentricity gradient: the proportion of monocular cells driven by the contralateral eye in-

creases to about 20% beyond 10° of eccentricity (Hubel & Wiesel, 1969). Third, OD in area

21a favors the ipsilateral eye: About 50% of the cells respond more strongly to the ipsilateral

eye's input while 30% of them prefer the contralateral eye (Dreher et al., 1993).

4.3. Putative mechanisms

Singer, von Grünau, and Rauschecker (1980) have described prolonged latencies and reduced

response amplitudes of supragranular layers of striate cortex in response to stimulation of the

amblyopic eye. They hypothesized that these might be caused by weaker intraareal transmis-

sion from layer IV to supragranular layers. Eschweiler and Rauschecker (1993) proposed that

this reduced impact of the amblyopic eye might be due to disturbed synchronicity of excitato-

ry inputs to the cortex.

In the primary visual cortex of cats strabismic amblyopia is associated with reduced response

synchronization among neurons connected to the amblyopic eye (Roelfsema et al., 1994).

This could account for the massive failure of neurons in area 21a to respond to stimulation of

the amblyopic eye. Reduced synchrony has been shown both experimentally (Alonso, Usrey,

& Reid, 1996) and in simulation studies (Lumer et al., 1997; Stevens & Zador, 1998) to im-

pair transmission of activity from one processing stage to the next because it renders summa-

tion of synaptic potentials less effective. The reduced saliency of responses evoked from the

amblyopic eye could have impaired responses in area 21a in two ways: First, through actual

transmission failure of activity conveyed from area 17 to area 21a and, second, through activi-

ty-dependent long-term changes in synaptic efficacy. In the visual cortex, activity-dependent

long-term modifications of synaptic efficacy follow Hebb's rule (Hebb, 1949). Synapses are

strengthened if the probability is high that they are active in temporal contiguity with the

postsynaptic target cell, but destabilize if they are inactive while their target is driven by other

inputs (Miller, Keller, & Stryker, 1989; Rauschecker & Singer, 1979). Prior to squint induc-

tion, most cells in area 21a should have been binocular as in normal animals (Dreher et al.,

1993) because the efferent projection cells in area 17 are binocular. Because of squint and

resulting competition among afferents from the two eyes, cells in area 17 become monocular

(Hubel & Wiesel, 1965a). This condition is, in turn, likely to lead to competition in area 21a

among the afferents from area 17 that are now monocular. In this competition the afferents

conveying activity from the amblyopic eye are likely to lose, because their activity is less


12

synchronized and will drive cells in area 21a less effectively than afferents representing the

normal eye.

The amblyopia-related changes in OD were much less pronounced in area PMLS than in area

21a. One reason could be differences in the organization of subcortical inputs. Areas 18 and

PMLS receive direct input from the LGN complex but area 21a does not (review: Orban,

1984). Moreover, the input to areas 18 and PMLS is mainly derived from Y-type retinal gan-

glion cells while input to area 21a is supplied predominantly by the X-system (Burke, Dreher,

& Wang, 1998). We think that the difference in OD distributions is mainly due to the lack of

direct input to area 21a because OD distributions in amblyopic cats are similar in all areas

receiving direct LGN input, irrespective of whether this input is from the X-system (area 17)

or the Y-system (area 18, PMLS).

It seems that area 21a is more sensitive to the loss of binocular congruency than area PMLS.

This correlates well with area 21a being specialized for central vision by covering a strip be-

tween 10° above and 5° below the horizontal meridian (Dreher, Wang, Turlejski, Djavadian,

& Burke, 1996). Moreover, cells in this area have small RFs which, in early development,

might easily lead to a misalignment of the two monocular fields. In contrast, RFs of PMLS

cells are distributed throughout the visual field and are significantly bigger than those of 21a

cells at comparable eccentricities (Dreher et al., 1996).

4.4. Involvement of cortical pathways in amblyopia

It has been suggested that, in the cat, area 21a might be the gateway to a form-processing ven-

tral stream while area PMLS could be regarded as the entrance to a distinct motion-processing

dorsal stream (Burke et al., 1998; Dreher et al., 1996). Given the homologies between cat area

PMLS and monkey area MT (Payne, 1993) as well as between cat area 21a and monkey areas

VP and/or V4 (Dreher et al., 1993; Payne, 1993) our data predict that in amblyopic monkeys

the weak eye should be less represented in areas VP and V4 than in area MT.

Psychophysical results suggest that the cortex is the neural site of amblyopia in humans (e.g.

Hess & Pointer, 1985; Levi & Klein, 1985) and the nature of some of the deficits indicates

anomalies in primary visual cortex. However, there is also evidence for deficits at higher lev-

els of processing (Sharma, Levi, & Klein, 2000). Non-invasive imaging studies show that the

relative reduction of responses to the amblyopic eye increases from striate to extrastriate areas

(Anderson, Holliday, & Harding, 1999; Barnes, Hess, Dumoulin, Achtman, & Pike, 2001;

Goodyear, Nicolle, Humphrey, & Menon, 2000; Imamura et al., 1997; Muckli et al., 1998).

However, it is still unclear whether dorsal and ventral processing streams are affected differ-

entially (Hess, Demanins, & Bex, 1997). The stronger deficit in the perception of color as

compared to luminance stimuli (Mullen, Sankeralli, & Hess, 1996) suggests a stronger im-

pairment of the ventral stream and measurements of contrast sensitivity indicated normal mo-

tion processing (Hess & Anderson, 1993; Hess, Howell, & Kitchin, 1978b). However, Wood

and Kulikowski (1978) found motion and pattern vision to be equally affected and Rentschler,

Hilz, and Brettel (1981) came to the same conclusion, studying apparent motion. Impaired


13

motion processing is also suggested by the observations that the velocity of gratings moving

in temporal direction is underestimated (Tychsen & Lisberger, 1986) and that motion afteref-

fects are weaker in amblyopes (Hess et al., 1997). Thus, psychophysical data suggest impair-

ment of functions attributed to both dorsal and ventral pathways.

Measurements of cortical evoked potentials in amblyopic children (Kubova, Kuba, Juran, &

Blakemore, 1996) revealed a reduced amplitude and prolonged latency of responses to pattern

reversal whereas motion onset responses were not affected suggesting a relative sparing of the

dorsal pathway. This result is in line with imaging data (fMRI; Muckli et al., 1998) which

suggests that activation deficits in prestriate visual areas are more pronounced along the tem-

poral pathway from area VP to the fusiform gyrus than along the dorsal pathway from V2 to

posterior intraparietal areas. Thus, in agreement with the present study, measurements of neu-

ronal activity favor the view that the ventral pathway is more affected in amblyopia than the

dorsal pathway.

Acknowledgements

We like to thank Ruxandra Sireteanu for her collaboration in the behavioral testing of the cats.

We are grateful also to Carmen Selignow, Petra Janson, Hanka Klon-Lipok, Sandra

Schwegmann, and Maren Kurschat for excellent technical assistance, and to Renate Ruhl and

Selina Völsing for their help in preparing the figures. Supported by the Heisenberg Program

of the DFG, the Minna-James-Heineman Foundation and the Max Planck Society.

References

Alonso, J. M., Usrey, W. M., & Reid, R. C. (1996). Precisely correlated firing in cells of the

lateral geniculate nucleus. Nature, 383, 815–819.

Altmann, L., & Singer, W. (1986). Temporal integration in amblyopic vision. Vision Re-

search, 26, 1959–1968.

Anderson, S. J., Holliday, I. E., & Harding, G. F. A. (1999). Assessment of cortical dysfunc-

tion in human strabismic amblyopia using magnetoencephalography (MEG). Vision Research,

39, 1723– 1738.

Barnes, G. R., Hess, R. F., Dumoulin, S. O., Achtman, R. L., & Pike, G. B. (2001). The corti-

cal deficit in humans with strabismic amblyopia. Journal of Physiology, 533.1, 281–297.

Berman, N., & Murphy, E. H. (1982). The critical period for alteration in cortical binocularity

resulting from divergent and convergent strabismus. Developmental Brain Research, 2, 181–

202.

Brecht, M., Singer, W., & Engel, A. K. (1998). Correlation analysis of corticotectal interac-

tions in the cat visual system. Journal of Neurophysiology, 79, 2394–2407.


14

Burke, W., Dreher, B., & Wang, C. (1998). Selective block of conduction in Y optic nerve

fibres: significance for the concept of parallel processing. European Journal of Neuroscience,

10, 8–19.

Chino, Y. M., Ridder, W. H. III, & Czora, E. P. (1988). Effects of convergent strabismus on

spatio-temporal response properties of neurons in cat area 18. Experimental Brain Research,

72, 264– 278.

Chino, Y. M., Shansky, M. S., Jankowski, W. L., & Banser, F. A. (1983). Effects of rearing

kittens with convergent strabismus on development of receptive-field properties in striate cor-

tex neurons. Journal of Neurophysiology, 50, 265–286.

Crewther, S. G., Crewther, D. P., & Cleland, B. G. (1985). Convergent strabismic amblyopia

in cats. Experimental Brain Research, 60, 1–9.

Crewther, D. P., & Crewther, S. G. (1990). Neural site of strabismic amblyopia in cats: spatial

frequency deficit in primary cortical neurons. Experimental Brain Research, 79, 615–622.

Cynader, M., Gardner, J. C., & Mustari, M. (1984). Effects of neonatally induced strabismus

on binocular responses in cat area 18. Experimental Brain Research, 53, 384–399.

Dreher, B., Michalski, A., Ho, R. T., Lee, C. W. F., & Burke, W. (1993). Processing of form

and motion in area 21a of cat visual cortex. Visual Neuroscience, 10, 93–115.

Dreher, B., Wang, C., Turlejski, K. J., Djavadian, R. L, & Burke, W. (1996). Areas PMLS

and 21a of cat visual cortex: two functionally distinct areas. Cerebral Cortex, 6, 585–595.

Eschweiler, G. W., & Rauschecker, J. P. (1993). Temporal integration in visual cortex of cats

with surgically induced strabismus. European Journal of Neuroscience, 5, 1501–1509.

Freeman, R. D., & Tsumoto, T. (1983). An electrophysiological comparison of convergent

and divergent strabismus in the cat: electrical and visual activation of single cortical cells.

Journal of Neurophysiology, 49, 238–253.

Goodyear, B. G., Nicolle, D. A., Humphrey, G. K., & Menon, R. S. (2000). BOLD fMRI re-

sponse to early visual areas to perceived contrast in human amblyopia. Journal of Neurophys-

iology, 84, 1907–1913.

Hebb, D. O. (1949). The organization of behavior. New York: Wiley. Hess, R. F., & Ander-

son, S. J. (1993). Motion sensitivity and spatial undersampling in amblyopia. Vision Re-

search, 33, 881–896.


15

Hess, R. F., Campbell, F. W., & Greenhalgh, T. (1978a). On the nature of neural abnormality

in human amblyopia: neural aberrations and neural sensitivity losses. Pflügers Archiv, 377,

201–207.

Hess, R. F., Demanins, R., & Bex, P. J. (1997). A reduced motion aftereffect in strabismic

amblyopia. Vision Research, 37, 1303–1311.

Hess, R. F., & Holliday, I. E. (1992). The spatial localization deficit in amblyopia. Vision

Research, 32, 1319–1339.

Hess, R. F., Howell, E. R., & Kitchin, J. (1978b). On the relationship between pattern and

movement perception in strabismic amblyopia. Vision Research, 18, 375–377.

Hess, R. F., & Pointer, J. S. (1985). Differences in the neural basis of human amblyopia: the

distribution of the anomaly across the visual field. Vision Research, 25, 1577–1594.

Holopigian, K., & Blake, R. (1983). Spatial vision in strabismic cats. Journal of Neurophysi-

ology, 50, 287–296.

Hubel, D. H., &Wiesel, T. N. (1965a). Binocular interactions in striate cortex of kittens reared

with artificial squint. Journal of Neurophysiology, 28, 1041–1059.

Hubel, D. H., & Wiesel, T. N. (1965b). Receptive fields and functional architecture in two

nonstriate visual areas (18 and 19) of the cat. Journal of Neurophysiology, 28, 229–289.

Hubel, D. H., & Wiesel, T. N. (1969). Visual area of the lateral suprasylvian gyrus (Clare-

Bishop area) of the cat. Journal of Physiology, 202, 251–260.

Hubel, D. H., & Wiesel, T. N. (1970). The period of susceptibility to the physiological effects

of unilateral eye closure. Journal of Physiology, 206, 419–436.

Imamura, K., Richter, H., Fischer, H., Lennerstrand, G., Franzèn, O., Rydberg, A., Andersson,

J, Schneider, H., Onoe, H., Watanabe, Y., & Lângström, B. (1997). Reduced activity in ex-

trastriate visual cortex of individuals with strabismic amblyopia. Neuroscience Letters, 225,

173–176.

Jacobson, S. G., & Ikeda, H. (1979). Behavioural studies of spatial vision in cats reared with

convergent squint: is amblyopia due to arrest of development? Experimental Brain Research,

34, 11–26.

Katz, B., & Sireteanu, R. (1992). Development of visual acuity in kittens: a comparison be-

tween jumping stand and Teller Acuity Card test. Clinical Vision Science, 7, 219–224.


16

Kiorpes, L., Kiper, D. C., O’Keefe, L. P., Cavanaugh, J. R., & Movshon, J. A. (1998). Neu-

ronal correlates of amblyopia in the visual cortex of macaque monkeys with experimental

strabismus and anisometropia. Journal of Neuroscience, 18, 6411–6424.

Kiorpes, L., & McKee, S. (1999). Neural mechanisms underlying amblyopia. Current Opinion

in Neurobiology, 9, 480–486.

König, P., Engel, A. K., Löwel, S., & Singer, W. (1993). Squint affectssynchronization of

oscillatory responses in cat visual cortex. European Journal of Neuroscience, 5, 501–508.

Kubova, Z., Kuba, M., Juran, J., & Blakemore, C. (1996). Is the motion system relatively

spared in amblyopia? Evidence from cortical evoked responses. Vision Research, 36, 181–

190.

Lagrèze, W. D., & Sireteanu, R. (1991). Two-dimensional spatial distortions in human stra-

bismic amblyopia. Vision Research, 31, 1271–1288.

Levi, D. M., & Klein, S. (1985). Vernier acuity, crowding and amblyopia. Vision Research,

25, 979–991.

Lumer, E. D., Edelman, G. M., & Tononi, G. (1997). Neural dynamics in a model of the

thalamocortical system. II. The role of neural synchrony tested through perturbations of spike

timing. Cerebral Cortex, 7, 228–236.

Miller, K. D., Keller, J. B., & Stryker, M. P. (1989). Ocular dominance column development:

analysis and simulation. Science, 245, 605–615.

Mitchell, D. E., Griffin, F., Wilkinson, F., Anderson, P., & Smith, M. L. (1976). Visual reso-

lution in young kittens. Vision Research, 16, 363–366.

Mitchell, D. E., Ruck, M., Kaye, M. G., & Kirby, S. (1984). Immediate and long-term effects

on visual acuity of surgically induced strabismus in kittens. Experimental Brain Research, 55,

420–430.

Mower, G. D., & Duffy, F. H. (1983). Animal models of strabismic amblyopia comparative

behavioral studies. Behavioural Brain Research, 7, 239–251.

Muckli, L., Tonhausen, N., Goebel, R., Lanfermann, H., Zanella, F. E., Singer, W., & Sire-

teanu, R. (1998). Reduced fMRI response to monocular amblyopic eye stimulation in ex-

trastriate visual areas. NeuroImage, 7, 4, p. 19 (abstract).

Mullen, K. T., Sankeralli, M. J., & Hess, R. F. (1996). Color and luminance vision in human

amblyopia: shifts in isoluminance, contrast sensitivity losses, and positional deficits. Vision

Research, 36, 645–653.


17

Niebur, E., Koch, C., & Rosin, C. (1993). An oscillation-based model for the neuronal basis

of attention. Vision Research, 33, 2789–2802.

Orban, G. A. (1984). Neuronal operations in the visual cortex. Berlin: Springer. Palmer, L. A.,

Rosenquist, A. C., & Tusa, R. J. (1978). Retinotopic organization of lateral suprasylvian visu-

al areas in the cat. Journal of Comparative Neurology, 177, 237–256.

Payne, B. R. (1993). Evidence for visual cortical area homologos in cat and monkey. Cerebral

Cortex, 3, 1–25.

Rauschecker, J. P., & Singer, W. (1979). Changes in the circuitry of the kitten visual cortex

are gated by postsynaptic activity. Nature, 280, 58–60.

Rentschler, I., Hilz, R., & Brettel, H. (1981). Amblyopic abnormality involves neural mecha-

nisms concerned with movement processing. Investigative Ophthalmology and Visual Sci-

ence, 20, 695–700.

Roelfsema, P. R., König, P., Engel, A. K., Sireteanu, R., & Singer, W. (1994). Reduced syn-

chronization in the visual cortex of cats with strabismic amblyopia. European Journal of Neu-

roscience, 6, 1645–1655.

Schröder, J.-H., Fries, P., Roelfsema, P. R., Singer, W., & Engel, A. K. (1998). Correlates of

strabismic amblyopia in cat extrastriate visual areas. European Journal of Neuroscience, 10

(suppl. 10), 2 37 (abstract).

Sharma, V., Levi, D. M., & Klein, S. A. (2000). Undercounting features and missing features:

evidence for a high-level deficit in strabismic amblyopia. Nature Neuroscience, 3, 496–501.

Sherman, S. M. (1972). Development of interocular alignment in cats. Brain Research, 37,

187–203. Singer, W., von Grünau, M., & Rauschecker, J. P. (1980). Functional amblyopia in

kittens with unilateral exotropia: I. Electrophysiological assessment. Experimental Brain Re-

search, 40, 294–304.

Sireteanu, R., & Best, J. (1992). Squint-induced modification of visual receptive fields in the

lateral suprasylvian cortex of the cat: binocular interaction, vertical effect, and anomalous

correspondence. European Journal of Neuroscience, 4, 235–242.

Sireteanu, R., Lagrèze, W. D., & Constantinescu, D. H. (1993). Distortions in two-

dimensional visual space perception in strabismic observers. Vision Research, 33, 677–690.

Spear, P. D., & Baumann, T. P. (1975). Receptive-field characteristics of single neurons in

lateral suprasylvian visual area of the cat. Journal of Neurophysiology, 38, 1403–1420.


18

Stevens, C. F., & Zador, A. M. (1998). Input synchrony and the irregular firing of cortical

neurons. Nature Neuroscience, 1, 210–217.

Tusa, R. J., & Palmer, L. A. (1980). Retinotopic organization of areas 20 and 21 in the cat.

Journal of Comparative Neurology, 193, 147–164.

Tusa, R. J., Rosenquist, A. C., & Palmer, L. A. (1979). Retinotopic organization of areas 18

and 19 in the cat. Journal of Comparative Neurology, 185, 657–678.

Tychsen, L., & Lisberger, S. G. (1986). Maldevelopment of visual motion processing in hu-

mans who had strabismus with onset in infancy. Journal of Neuroscience, 6, 2495–2508.

Von Grünau, M. W. (1979). The role of maturation and visual experience in the development

of eye alignment in cats. Experimental Brain Research, 37, 41–47.

Von Grünau, M. W. (1982). Comparison of the effects of induced strabismus on binocularity

in area 17 and the LS area of the cat. Brain Research, 246, 325–329.

Von Grünau, M. W., & Singer, W. (1980). Functional amblyopia in kittens with unilateral

exotropia: II. Correspondence between behavioral and electro-physiological assessment. Ex-

perimental Brain Research, 40, 305–310.

von Noorden, G. K. (1990). Binocular vision and ocular motility. St. Louis: C.V. Mosby.

Wood, I. C. J., & Kulikowski, J. J. (1978). Pattern and movement detection in patients with

reduced visual acuity. Vision Research, 18, 331–334.

Yinon, U., Auerbach, E., Blank, M., & Friesenhausen, J. (1975). The ocular dominance of

cortical neurons in cats developed with divergent and convergent strabismus. Vision Re-

search, 15, 1251–1256

Ocular dominance distribution of cells in area 21a correlates with perceptual deficits in amblyopic...

Documents

Transcript of Ocular dominance distribution of cells in area 21a correlates with perceptual deficits in amblyopic...