Giorgio M. Innocenti

General Organization of Callosal Connections in the Cerebral Cortex

The necessity of interhemispheric connections, and the nature of this necessity, are demonstrated by the following hypothetical event. An intelligent being from outer space lands on earth and is asked to design the brain of a cat. The being is intrigued to find that the body of a cat is bilaterally symmetric (he looks himself rather like a multieyed and multiwhiskered octopus).

Exuberant projection into the corpus callosum from the visual cortex of newborn cats

In the first postnatal week, neurones projecting into the corpus callosum can be identified in kittens visual cortex by retrograde transport of HRP. The neurones are located in layers III, IV, and VI. The region of cortex which gives rise to the callosal projection extends beyond its adult boundaries over most of area 17, 18, 19 and in the suprasylvian sulcus.

Exuberance in the development of cortical networks.

The cerebral cortex is the largest and most intricately connected part of the mammalian brain. Its size and complexity has increased during the course of evolution, allowing improvements in old functions and causing the emergence of new ones, such as language. This has expanded the behavioural and cognitive repertoire of different species and has determined their competitive success. To allow the relatively rapid emergence of large evolutionary changes in a structure of such importance and complexity, the mechanisms by which cortical circuitry develops must be flexible and yet robust against changes that could disrupt the normal functions of the networks.

Postnatal shaping of callosal connections from sensory areas

SummaryHorseradish peroxidase (HRP) was injected unilaterally into the first and second visual areas (VI and V2; areas 17 and 18) of 20 kittens aged between 2 and 90 days and into the second somatosensory area (S2) of 16 kittens aged between 1 and 52 days. The radial and tangential (normal and parallel to the pial surface, respectively) distributions of neurones giving origin to callosal axons (callosal neurones) were studied. In adult cats, callosal efferent zones (CZs) are defined by the distribution of callosal neurones. CZs occupy, in the visual cortices, tangentially and radially restricted parts of areas 17, 18, 19 of the lateral suprasylvian gyms and in the somatosensory cortices, parts of SI and S2. At birth, callosal neurones are distributed throughout the tangential extent of visual and somatosensory areas; they are also more widespread in depth than in the adult. During the first postnatal month, as a result of the gradual disappearence of callosal neurones from parts of the visual and somatosensory areas, the adult CZs emerge. The CZ in areas 17 and 18 undergoes a further tangential reduction during the second and third postnatal months.

Bilateral transitory projection to visual areas from auditory cortex in kittens

A transitory projection from primary and secondary auditory areas to the contralateral and ipsilateral areas 17 and 18 exists in newborn kittens. Distinct neuronal populations project to ipsilateral areas 17-18, contralateral areas 17-18 and contralateral auditory cortex; they are at different depth in layers II, III, and IV. By postnatal day 38 the auditory to visual projections have been lost, apparently by elimination of axons rather than by neuronal death. While it was previously reported that the elimination of transitory axons is responsible for focusing the origin of callosal connections to restricted portions of sensory areas it now appears that similar events play a more general role in the organization of cortico-cortical networks. Indeed, the elimination of juvenile projections is largely responsible for determining which areas will be connected in the adult.

The postnatal development of visual callosal connections in the absence of visual experience or of the eyes.

SummaryCounts of callosal neurons retrogradely labeled by horseradish peroxidase (visualized using multiple substrates) were obtained in areas 17 and 18 of five kittens reared with their eyelids bilaterally sutured and of three kittens which had undergone bilateral enucleation on postnatal days 1–4. These counts were compared with those obtained in normal adult cats.The normal adult distribution of the callosal neurons results from the gradual postnatal reduction of a more widespread juvenile population. Binocular visual deprivation by lid suturing dramatically decreases the final number of callosal neurons and narrows their region of distribution (callosal zone) in areas 17 and 18. A less severe reduction in the final number of callosal neurons is caused by bilateral enucleation, which also increases the width of the callosal zone compared to that of normal cats. Thus, visual experience is necessary for the normal stabilization of juvenile callosal connections. However, since some callosal neurons form connections in the absence of vision, other influences capable of stabilizing juvenile callosal neurons also exist. These influences are probably antagonized by destabilizing influences or inhibited, when the eyes are intact.

Is there a genuine exuberancy of callosal projections in development? A quantitative electron microscopic study in the cat.

The main tract of interhemispheric connections, the corpus callosum, is now suspected to contain more axons at birth than in adulthood. This notion is based on results obtained with retrograde pathway tracing techniques, but this indirect approach has several shortcomings. Since the elimination of projections during development now seems to be a general phenomenon, probably a crucial one in the establishment of connections, we have examined the development of the corpus callosum using quantitative electron microscopy. An average of 70% of the callosal axons present at birth are eliminated by adulthood in the cat. We have also calculated a new figure of 23 million axons in the adult cat corpus callosum, which is over 4 times greater than the currently accepted figure.

Assessment of EEG synchronization based on state-space analysis.

Cortical computation involves the formation of cooperative neuronal assemblies characterized by synchronous oscillatory activity. A traditional method for the identification of synchronous neuronal assemblies has been the coherence analysis of EEG signals. Here, we suggest a new method called S estimator, whereby cortical synchrony is defined from the embedding dimension in a state-space. We first validated the method on clusters of chaotic coupled oscillators and compared its performance to that of other methods for assessing synchronization. Then nine adult subjects were studied with high-density EEG recordings, while they viewed in the two hemifields (hence with separate hemispheres) identical sinusoidal gratings either arranged collinearly and moving together, or orthogonally oriented and moving at 90 degrees . The estimated synchronization increased with the collinear gratings over a cluster of occipital electrodes spanning both hemispheres, whereas over temporo-parietal regions of both hemispheres, it decreased with the same stimulus and it increased with the orthogonal gratings. Separate calculations for different EEG frequencies showed that the occipital clusters involved synchronization in the beta band and the temporal clusters in the alpha band. The gamma band appeared to be insensitive to stimulus diversity. Different stimulus configurations, therefore, appear to cause a complex rearrangement of synchronous neuronal assemblies distributed over the cortex, in particular over the visual cortex.

Evolution amplified processing with temporally dispersed slow neuronal connectivity in primates

The corpus callosum (CC) provides the main route of communication between the 2 hemispheres of the brain. In monkeys, chimpanzees, and humans, callosal axons of distinct size interconnect functionally different cortical areas. Thinner axons in the genu and in the posterior body of the CC interconnect the prefrontal and parietal areas, respectively, and thicker axons in the midbody and in the splenium interconnect primary motor, somatosensory, and visual areas. At all locations, axon diameter, and hence its conduction velocity, increases slightly in the chimpanzee compared with the macaque because of an increased number of large axons but not between the chimpanzee and man. This, together with the longer connections in larger brains, doubles the expected conduction delays between the hemispheres, from macaque to man, and amplifies their range about 3-fold. These changes can have several consequences for cortical dynamics, particularly on the cycle of interhemispheric oscillators.

Morphological correlates of visual field transformation in the corpus callosum.

Horseradish peroxidase (HRP) injections at the boundary of areas 17 and 18 in cat visual cortex resulted in retrograde labelling of neurones located in contralateral areas 17 and 18. The HRP-positive neurones were spread over a region located around the representation of the vertical meridian; they were most numerous in the locus for area centralis. Layer III pyramidal cells and stellate cells in upper layer IV were identified as the principal neurones of origin of the visual callosal pathway.