Effects of digesting chondroitin sulfate proteoglycans on plasticity in cat primary visual cortex

Abstract

Monocular deprivation (MD) during a critical period of postnatal development produces significant changes in the anatomy and physiology of the visual cortex, and the deprived eye becomes amblyopic. Extracellular matrix molecules have a major role in restricting plasticity such that the ability to recover from MD decreases with age. Chondroitin sulfate proteoglycans (CSPGs) act as barriers to cell migration and axon growth. Previous studies showing that degradation of CSPGs by the bacterial enzyme chondroitinase can restore plasticity in the adult rat visual cortex suggest a potential treatment for amblyopia. Here MD was imposed in cats from the start of the critical period until 3.5 months of age. The deprived eye was reopened, the functional architecture of the visual cortex was assessed by optical imaging of intrinsic signals, and chondroitinase was injected into one hemisphere. Imaging was repeated 1 and 2 weeks postinjection, and visually evoked potentials (VEPs) and single-cell activity were recorded. Immunohistochemistry showed that digestion of CSPGs had been successful. After 2 weeks of binocular exposure, some recovery of deprived-eye responses occurred when chondroitinase had been injected into the hemisphere contralateral to that eye; when injected into the ipsilateral hemisphere, no recovery was seen. Deprived-eye VEPs were no larger in the injected hemisphere than in the opposite hemisphere. The small number of neurons dominated by the deprived eye exhibited poor tuning characteristics. These results suggest that despite structural effects of chondroitinase in adult cat V1, plasticity was not sufficiently restored to enable significant functional recovery of the deprived eye.

Figure 1.

Immunohistochemical evidence for CSPG digestion. Photomicrographs show coronal sections through the primary visual cortex near the apex of the postlateral gyrus of one cat in which the left hemisphere, contralateral to the deprived eye, had been injected with chondroitinase ABC. Arrows indicate orientation (D, dorsal; V, ventral; L, lateral). The first two columns show images of the injected hemisphere at low magnification (scale bar, 500 μm) and high magnification (scale bar, 50 μm). The third and fourth columns show images of the noninjected hemisphere, again at low and high magnification (scale bars, 500 and 50 μm, respectively. The top row of images shows the extent of ChABC digestion as revealed by labeling the CSPG stubs with the monoclonal antibody 1B5.While there is very little staining of either neuropil or perineuronal nets in the noninjected hemisphere (right) the injected hemisphere (left) exhibits dense labeling throughout all cortical layers (densest in layer 4; approximate layer boundaries are indicated), as well as intense labeling of PNNs surrounding individual neurons. In contrast, staining for WFA (second row) and aggrecan (third row) both show strong labeling of neuropil and PNNs in the noninjected hemisphere but weak labeling in the injected hemisphere.

Figure 2.

Ocular dominance maps in an animal in which the hemisphere contralateral to the deprived eye had been injected with chondroitinase. The top row shows the surface view of the dorsal part of the primary visual cortex on the day when the deprived right eye was reopened and the left hemisphere was injected (stars mark injection sites). It then shows the ocular dominance map for the nondeprived left eye followed by the ocular dominance map for the deprived right eye. Dark regions indicate areas of response through the stimulated eye. Note that the left and right eye maps are negatives of each other; they are both shown to facilitate visual comparison. The middle row illustrates cortical surface and the left and right eye ocular dominance maps 1 week (wk) later; the bottom row shows the same 2 weeks (wks) later. Arrows indicate orientation (A, anterior; R, right). Scale bar, 2 mm.

Figure 3.

Cortical areas responding to deprived and nondeprived eye stimulation in both hemispheres of V1. A, Mean areas (± SEM) of left (L) and right (R) eye responses before chondroitinase injection into the left hemisphere (contralateral to the deprived eye) 1 week (wk) later and 2 weeks later. Gray bars give results for the left (injected) hemisphere (LH), and white bars give results for the right (untreated) hemisphere (RH). B, Mean areas (± SEM) of left and right eye responses at the time of reopening of the deprived eye (“end MD”), one week later and 2 weeks later. C, Mean areas (± SEM) of left and right eye responses before chondroitinase injection into the right hemisphere (ipsilateral to the deprived eye) 1 week later and 2 weeks later.

Figure 4.

Ocular dominance maps in an animal in which the hemisphere ipsilateral to the deprived eye had been injected with chondroitinase. The top row shows the surface view of the dorsal part of the primary visual cortex on the day when the deprived right eye was reopened and the right hemisphere was injected with chondroitinase (white stars mark injection sites) while the left hemisphere was injected with saline (black star). It then shows the ocular dominance map for the nondeprived left eye followed by the ocular dominance map for the deprived right eye. Dark regions indicate areas of response through the stimulated eye. Note that the left and right eye maps are negatives of each other; they are both shown to facilitate visual comparison. The bottom row illustrates cortical surface and left and right eye ocular dominance maps 2 weeks later. Arrows indicate orientation (A, anterior; R, right). Scale bar, 2 mm.

Figure 5.

Orientation maps in an animal in which the left hemisphere contralateral to the deprived eye had been injected with chondroitinase. A, Maps obtained at the end of the deprivation period. The top row shows the surface view of the dorsal part of the primary visual cortex on the day when the deprived right eye was reopened and the left hemisphere was injected (stars mark injection sites). It then shows iso-orientation maps for right (deprived) eye responses to horizontal (0°) gratings and vertical (90°) gratings normalized by subtraction of the response to a gray screen (“blank”). Dark regions indicate areas of response. On the left the second row shows the “polar” orientation preference map obtained by vectorial addition of responses to gratings of 0, 45, 90, and 135°, where the hue indicates each pixel's preferred orientation, and the brightness indicates its selectivity for orientation. The middle and right images show iso-orientation maps for right (deprived) eye responses to 0 and 90° gratings normalized by subtraction of the response to the cocktail blank (the sum of responses to all orientations through that eye). The third row shows, on the left, the “polar” orientation preference map for the left (nondeprived) eye. The second and third images show iso-orientation maps for left eye responses to 0 and 90° gratings normalized by subtraction of the response to the cocktail blank. B, Maps obtained from the same animal 2 weeks later; all conventions as in A. Arrows indicate orientation (A, anterior; R, right). Scale bar, 2 mm.

Figure 6.

VEPs in response to contrast-reversing gratings of different spatial frequencies. A, Amplitude of VEPs recorded from one cat in which the right hemisphere ipsilateral to the deprived eye had been injected with chondroitinase. The responses through the left nondeprived eye (LE) exhibit a cutoff at 2.26 c/deg for both the left hemisphere (LH) and the right hemisphere (RH). Responses through the right eye (RE) were not significantly different from those to a blank screen in either hemisphere. B, VEP amplitudes averaged across five animals in which the right hemisphere ipsilateral to the deprived eye had been injected with chondroitinase. Mean amplitudes (± SEM) are plotted against the spatial frequency of the test gratings for both the left, nondeprived eye and the right, deprived eye. C, VEP amplitudes averaged across seven animals in which the left hemisphere contralateral to the deprived eye had been injected with chondroitinase; all conventions as in B.

Figure 7.

Orientation tuning of single neurons recorded in the injected hemisphere. A, Responses of a simple cell dominated by the left, nondeprived (non-depr.) eye (LE) recorded from the left hemisphere. The polar plot displays the response (in spikes/s) against the direction of drift of the grating stimulus (0° denotes a horizontal grating drifting up). Error bars represent SEM. This cell had a tuning half-width of 24°. B, Responses of a complex cell dominated by the right deprived (depr.) eye (RE) recorded from the left hemisphere. This cell had a tuning half-width of 29°. C, Responses of a binocular simple cell, recorded from the right hemisphere. Note that the responses through the left (nondeprived) eye (filled circles) were quite sharply tuned for orientation (half-width, 26°), while the responses through the right (deprived) eye (open circles) were nonselective.

Figure 8.

Spatial frequency tuning of single neurons recorded in the injected hemisphere. A, Responses of a complex cell dominated by the left nondeprived (non-depr.) eye (LE) recorded from the left hemisphere. The filled circles represent the responses through the left eye, the open circles those through the right (deprived) eye. The open diamond on the right represents the spontaneous activity level in the presence of a blank screen. Error bars represent SEM. B, Responses of a complex cell dominated by the right (deprived) eye (RE) recorded from the left hemisphere. Conventions are as in A.