Callosal Influence on Visual Receptive Fields Has an Ocular, an Orientation-and Direction Bias

Abstract

One leading hypothesis on the nature of visual callosal connections (CC) is that they replicate features of intrahemispheric lateral connections. However, CC act also in the central part of the binocular visual field. In agreement, early experiments in cats indicated that they provide the ipsilateral eye part of binocular receptive fields (RFs) at the vertical midline (Berlucchi and Rizzolatti, 1968), and play a key role in stereoscopic function. But until today callosal inputs to receptive fields activated by one or both eyes were never compared simultaneously, because callosal function has been often studied by cutting or lesioning either corpus callosum or optic chiasm not allowing such a comparison. To investigate the functional contribution of CC in the intact cat visual system we recorded both monocular and binocular neuronal spiking responses and receptive fields in the 17/18 transition zone during reversible deactivation of the contralateral hemisphere. Unexpectedly from many of the previous reports, we observe no change in ocular dominance during CC deactivation. Throughout the transition zone, a majority of RFs shrink, but several also increase in size. RFs are significantly more affected for ipsi- as opposed to contralateral stimulation, but changes are also observed with binocular stimulation. Noteworthy, RF shrinkages are tiny and not correlated to the profound decreases of monocular and binocular firing rates. They depend more on orientation and direction preference than on eccentricity or ocular dominance of the receiving neuron's RF. Our findings confirm that in binocularly viewing mammals, binocular RFs near the midline are constructed via the direct geniculo-cortical pathway. They also support the idea that input from the two eyes complement each other through CC: Rather than linking parts of RFs separated by the vertical meridian, CC convey a modulatory influence, reflecting the feature selectivity of lateral circuits, with a strong cardinal bias.

Keywords: anticipation; binocular; interhemispheric connectivity; monocular; orientation selectivity.

Figure 1

Schematic summary of the direct geniculo-cortical and indirect callosal circuits affected by binocular and monocular stimulation during contralateral cooling deactivation. (A) Intact circuit of all direct geniculo-cortical and indirect callosal pathways during binocular stimulation. Red thicker lines indicate pathways crossing in the chiasm to emphasize that this portion of retino-geniculate fibers is larger than the non-chiasm-crossing portion (green). VM: vertical meridian, LGN, lateral geniculate nucleus, A17, area 17, A18, area 18. (B) The same circuit for different monocular and binocular stimulations before (left) and during cooling deactivation (right) of the right hemisphere. Cryoloop indicated by a light blue star positioned over the right hemisphere. For binocular stimulation, cooling deactivates both chiasm-crossing and non-chiasm-crossing inputs from the right hemisphere representing the left hemifield. For ipsilateral eye stimulation, cooling deactivates only the larger portion of chiasm-crossing (red) input passing through the corpus callosum. For contralateral stimulation, cooling deactivates only the smaller portion of non-chiasm crossing (green) input passing through the corpus callosum.

Figure 2

Changes in the ocular dominance distribution. (A) Classical ocular dominance response classes adapted from Hubel and Wiesel (1962) based on 5 equidistant intervals of the ocular dominance index. Class 1, contralateral eye only; class 2, dominated by the contralateral eye; class 3, binocularly driven; class 4, dominated by the ipsilateral eye; class 5, ipsilateral eye only. (B) original ocular dominance indices during baseline and reversible deactivation of CC. Note that the distributions do not change with absent visual callosal input.

Figure 3

Changes of RF size during deactivation of CC. Receptive field sizes estimated during cooling vs. baseline recordings for binocular (open circles), contralateral (light gray cricles) and ipsilateral (black circles) stimulation. Note that both increases and decreases occur for all three stimulation conditions. RF shrinkage is more frequent with ipsilateral stimulation.

Figure 4

Example receptive field of a TZ unit during CC deactivation for binocular and both monocular stimulations. (A) Upper left, intrinsic signal difference image to identify the 17/18 border (see Methods). Dark, optimal activation for area 18; light, optimal activation for area 17; dotted line, estimated area 17/18 border. Upper right, photograph of the recorded area on the left with matrix electrodes in position. Lower left, difference image for orthogonal (horizontal vs. vertical) stimulations. Lower right, vessel image with schematic drawing of the recording sites. (B) Receptive field extent of an example unit from the matrix on the area 18 side of the border. Left column, baseline recording, middle column, during cooling deactivation of CC, right column, recovery recording. Rows for the binocular and monocular stimulations as indicated by the symbols. Upper right: Polar plot of the mean firing rate of the unit for 16 different directions.

Figure 5

Mean modulation indices for RF size (MI_size, A) and firing rate (MI_rate, B) during CC deactivation. Values are separated for the recorded sub-areas 17, transition zone and 18 (color code on the right) and different eye stimulation (indicated below). Significant decreases (paired t-test, see Results) are indicated by stars. Note that units in the transition zone decrease in RF size in particular during ipsilateral stimulation. In contrast, firing rates decrease significantly in all zones and for all stimulations.

Figure 6

RF changes for binocular and the two monocular stimulations during CC deactivation (MI_size) are not significantly correlated with the OD index of the callosal input receiving neuron.

Figure 7

Cardinal bias in the influence of CC on receptive field size in the TZ. Mean modulation indices for RF decrease during ipsilateral stimulation separated by the receiving neuron's orientation preference. Note, that receptive fields of either horizontally or vertically preferring neurons shrink significantly more than those of obliquely preferring neurons when CC are deactivated. Stars indicate significant difference (see Results).

Figure 8

Example receptive fields obtained during ipsilateral stimulation during removal of callosal input (light blue) for neurons preferring movement into (upper left) and out of (upper right) the cooled hemifield, as well as up and down movement (lower left and right). Vessel image and conventions as in Figure 4. Note that the neuron preferring the out of movement is particularly affected. Neurons preferring horizontal contours decrease RF size independent of their preferred direction of motion and position.

Figure 9

Directional bias in the impact of CC input on receptive field size in the TZ. (A) Schematic illustration of the pathways activated through the ipsi (in green)-and contralateral (in red) eye during cooling of the two groups of neurons preferring either the direction of movement out of (black arrow) or into (gray arrow) the cooled hemifield (in blue). Cooled hemisphere indicated in blue. (B) Mean modulation indices for RF decrease separated by the receiving neuron's direction preference and stimulation of different eyes. Neurons from the two groups preferring either the direction of movement out of (black arrow, 0 ± 60 degrees) or into (gray arrow, 180 ± 60 degrees) the cooled hemifield are depicted. Stars indicate significant difference between ipsi–and contralateral stimulation (Mann-Whitney-U-test, see Results), and baseline and cooling recording (Wilcoxon Signed Rank test, see Results).