Clustered organization of cortical connectivity

Long-range corticocortical connectivity in mammalian brains possesses an intricate, nonrandom organization. Specifically, projections are arranged in ‘small-world’ networks, forming clusters of cortical areas, which are closely linked among each other, but less frequently with areas in other clusters. In order to delineate the structure of cortical clusters and identify their members, we developed a computational approach based on evolutionary optimization. In different compilations of connectivity data for the cat and macaque monkey brain, the algorithm identified a small number of clusters that broadly agreed with functional cortical subdivisions. We propose a simple spatial growth model for evolving clustered connectivity, and discuss structural and functional implications of the clustered, small-world organization of cortical networks.     

Neurotoxic lesions of ventrolateral prefrontal cortex impair object-in-place scene memory

Disconnection of the frontal lobe from the inferotemporal cortex produces deficits in a number of cognitive tasks that require the application of memory-dependent rules to visual stimuli. The specific regions of frontal cortex that interact with the temporal lobe in performance of these tasks remain undefined. One capacity that is impaired by frontal-temporal disconnection is rapid learning of new object-in-place scene problems, in which visual discriminations between two small typographic characters are learned in the context of different visually complex scenes. In the present study, we examined whether neurotoxic lesions of ventrolateral prefrontal cortex in one hemisphere, combined with ablation of inferior temporal cortex in the contralateral hemisphere, would impair learning of new object-in-place scene problems. Male macaque monkeys learned 10 or 20 new object-in-place problems in each daily test session. Unilateral neurotoxic lesions of ventrolateral prefrontal cortex produced by multiple injections of a mixture of ibotenate and N-methyl-D-aspartate did not affect performance. However, when disconnection from inferotemporal cortex was completed by ablating this region contralateral to the neurotoxic prefrontal lesion, new learning was substantially impaired. Sham disconnection (injecting saline instead of neurotoxin contralateral to the inferotemporal lesion) did not affect performance. These findings support two conclusions: first, that the ventrolateral prefrontal cortex is a critical area within the frontal lobe for scene memory; and second, the effects of ablations of prefrontal cortex can be confidently attributed to the loss of cell bodies within the prefrontal cortex rather than to interruption of fibres of passage through the lesioned area.     

Immunocytochemical localization of glycine receptors in the mammalian retina

The distribution of glycinergic synapses in the mammalian retina was studied with monoclonal antibodies against glycine receptors and a glycine receptor-related protein (gephyrin). Monoclonal antibody 2b is specific for the α1 subunit of the glycine receptor; monoclonal antibody 4a is specific for all known α subunits and the β subunit, and monoclonal antibody 7a is specific for gephyrin. The three antibodies were applied to the retina of cat, macaque monkey, rat, and rabbit. The general staining pattern is comparable in all these species and it is similar but distinct with all of the three antibodies. Labeling is characterized by a punctate appearance indicating that it occurs at synapses. In the inner plexiform layer, labeling is concentrated in two bands. One band is located close to the inner nuclear layer; the other band is located in the middle of the inner plexiform layer. In the outer plexiform layer, sparse punctate labeling is seen. The distribution of gephyrin was also studied at the ultrastructural level in cat and monkey retina. Gephyrin is present on the postsynaptic membrane of amacrine cells and ganglion cells. The presynaptic profile to gephyrin immunoreactivity is always of an amacrine cell. The AII amacrine cell, the crucial glycinergic interneuron of the rod pathway, is presynaptic to gephyrin immunoreactivity in the OFF-sublamina and is itself gephyrin-positive at an input synapse from another (possibly GABAergic) amacrine cell in the ON-sublamina. © 1993 Wiley-Liss, Inc.     

Dynamical consequences of lesions in cortical networks

To understand the effects of a cortical lesion it is necessary to consider not only the loss of local neural function, but also the lesion-induced changes in the larger network of endogenous oscillatory interactions in the brain. To investigate how network embedding influences a region's functional role, and the consequences of its being damaged, we implement two models of oscillatory cortical interactions, both of which inherit their coupling architecture from the available anatomical connection data for macaque cerebral cortex. In the first model, node dynamics are governed by Kuramoto phase oscillator equations, and we investigate the sequence in which areas entrain one another in the transition to global synchrony. In the second model, node dynamics are governed by a more realistic neural mass model, and we assess long-run inter-regional interactions using a measure of directed information flow. Highly connected parietal and frontal areas are found to synchronize most rapidly, more so than equally highly connected visual and somatosensory areas, and this difference can be explained in terms of the network's clustered architecture. For both models, lesion effects extend beyond the immediate neighbors of the lesioned site, and the amplitude and dispersal of nonlocal effects are again influenced by cluster patterns in the network. Although the consequences of in vivo lesions will always depend on circuitry local to the damaged site, we conclude that lesions of parietal regions (especially areas 5 and 7a) and frontal regions (especially areas 46 and FEF) have the greatest potential to disrupt the integrative aspects of neocortical function.     

Prediction of the main cortical areas and connections involved in the tactile function of the visual cortex by network analysis

We explored the cortical pathways from the primary somatosensory cortex to the primary visual cortex (V1) by analysing connectional data in the macaque monkey using graph-theoretical tools. Cluster analysis revealed the close relationship of the dorsal visual stream and the sensorimotor cortex. It was shown that prefrontal area 46 and parietal areas VIP and 7a occupy a central position between the different clusters in the visuo-tactile network. Among these structures all the shortest paths from primary somatosensory cortex (3a, 1 and 2) to V1 pass through VIP and then reach V1 via MT, V3 and PO. Comparison of the input and output fields suggested a larger specificity for the 3a/1-VIP-MT/V3-V1 pathways among the alternative routes. A reinforcement learning algorithm was used to evaluate the importance of the aforementioned pathways. The results suggest a higher role for V3 in relaying more direct sensorimotor information to V1. Analysing cliques, which identify areas with the strongest coupling in the network, supported the role of VIP, MT and V3 in visuo-tactile integration. These findings indicate that areas 3a, 1, VIP, MT and V3 play a major role in shaping the tactile information reaching V1 in both sighted and blind subjects. Our observations greatly support the findings of the experimental studies and provide a deeper insight into the network architecture underlying visuo-tactile integration in the primate cerebral cortex.     

Sensory enrichment after peripheral nerve injury restores cortical, not thalamic, receptive field organization

Sensory perception can be severely degraded after peripheral injuries that disrupt the functional organization of the sensory maps in somatosensory cortex, even after nerve regeneration has occurred. Rehabilitation involving sensory retraining can improve perceptual function, presumably through plasticity mechanisms in the somatosensory processing network. However, virtually nothing is known about the effects of rehabilitation strategies on brain organization, or where the effects are mediated. In this study, five macaque monkeys received months of enriched sensory experience after median nerve cut and repair early in life. Subsequently, the sensory representation of the hand in primary somatosensory cortex was mapped using multiunit microelectrodes. Additionally, the primary somatosensory relay in the thalamus, the ventroposterior nucleus, was studied to determine whether the effects of the enrichment were initiated subcortically or cortically. Age-matched controls included six monkeys with no sensory manipulation after median nerve cut and regeneration, and one monkey that had restricted sensory experience after the injury. The most substantial effect of the sensory environment was on receptive field sizes in cortical area 3b. Significantly greater proportions of cortical receptive fields in the enriched monkeys were small and well localized compared to the controls, which showed higher proportions of abnormally large or disorganized fields. The refinements in receptive field size and extent in somatosensory cortex likely provide better resolution in the sensory map and may explain the improved functional outcomes after rehabilitation in humans.     

The cortical connections of area V6: an occipito-parietal network processing visual information

The aim of this work was to study the cortical connections of area V6 by injecting neuronal tracers into different retinotopic representations of this area. To this purpose, we first functionally recognized V6 by recording from neurons of the parieto-occipital cortex in awake macaque monkeys. Penetrations with recording syringes were performed in the behaving animals in order to inject tracers exactly at the recording sites. The tracers were injected into the central or peripheral field representation of V6 in different hemispheres. Irrespective of whether injections were made in the centre or periphery, area V6 showed reciprocal connections with areas V1, V2, V3, V3A, V4T, the middle temporal area /V5 (MT/V5), the medial superior temporal area (MST), the medial intraparietal area (MIP), the ventral intraparietal area (VIP), the ventral part of the lateral intraparietal area and the ventral part of area V6A (V6AV). No labelled cells or terminals were found in the inferior temporal, mesial and frontal cortices. The connections of V6 with V1, and with all the retinotopically organized prestriate areas, were organized retinotopically. The connection of V6 with MIP suggests a visuotopic organization for this latter. Labelling in V6A and VIP after either central or peripheral V6 injections was very similar in location and extent, as expected on the basis of the nonretinotopic organization of these areas. We suggest that V6 plays a pivotal role in the dorsal visual stream, by distributing the visual information coming from the occipital lobe to the sensorimotor areas of the parietal cortex. Given the functional characteristics of the cells of this network, we suggest that it could perform the fast form and motion analyses needed for the visual guiding of arm movements as well as their coordination with the eyes and the head.     

Entorhinal cortex of the monkey: IV. Topographical and laminar organization of cortical afferents

The nonhuman primate entorhinal cortex is the primary interface for information flow between the neocortex and the hippocampal formation. Based on previous retrograde tracer studies, neocortical afferents to the macaque monkey entorhinal cortex originate largely in polysensory cortical association areas. However, the topographical and laminar distributions of cortical inputs to the entorhinal cortex have not yet been comprehensively described. The present study examines the regional and laminar termination of projections within the entorhinal cortex arising from different cortical areas. The study is based on a library of 51 (3)H-amino acid injections that involve most of the afferent regions of the entorhinal cortex. The range of termination patterns was broad. Some areas, such as the medial portion of orbitofrontal area 13 and parahippocampal areas TF and TH, project widely within the entorhinal cortex. Other areas have a more focal and regionally selective termination. The lateral orbitofrontal, insular, anterior cingulate, and perirhinal cortices, for example, project only to rostral levels of the entorhinal cortex. The upper bank of the superior temporal sulcus projects mainly to intermediate levels of the entorhinal cortex, and the parietal and retrosplenial cortices project to caudal levels. The projections from some of these cortical regions preferentially terminate in the superficial layers (I-III) of the entorhinal cortex, whereas others project more heavily to the deep layers (V-VI). Thus, some of the cortical inputs may be more influential on the cortically directed outputs of the hippocampal formation or on gating neocortical information flow into the other fields of the hippocampal formation rather than contributing to the perforant path inputs to other hippocampal fields.     

Primate spinothalamic pathways: I. A quantitative study of the cells of origin of the spinothalamic pathway

In six monkeys spinothalamic (STT) cells were retrogradely labeled by injecting 2% wheat germ agglutinin-conjugated horseradish peroxidase into the somatosensory thalamus. Following a 5-day survival period, the animals were perfused and the tissue was removed and processed with the tetramethyl benzidine technique. In all animals there were HRP-labeled STT cells in all segments of the spinal cord. In one old world monkey, the injection included most of the thalamus and resulted in 18.235 estimated total number of STT cells. Of this total, 35% were located in the upper cervical segments (C1-C3), 18% were located in C4-C8, 19% were in the thoracic spinal cord with most found in T1-T3; 6% were in L1-L3, 13% were in L4-L7, and 7% were in the coccygeal segments. Of the total labeled STT cells, 17% were found in the spinal cord ipsilateral to the thalamic injections; 53% of these cells were located in C1-C3 primarily in lamina VIII. The percentage of label found in the contralateral lower cervical region laminae I-III (43-50%), IV-VI (33-48%), and VII-X (8-17%) was similar among three animals with similar thalamic injections. The distributions of the shapes of the labeled STT cells were similar for each lamina between the lower cervical and lower lumbar regions. The mean diameter of the labeled STT cells varied with spinal cord segment and lamina. The lamina I STT cells were the smallest. In the cervical spinal cord, lamina VIII STT cells had the largest diameters, while in the lumbar region laminae IV-VI had the largest STT cells.     

The emergence and properties of mutual synchronization in in vitro coupled cortical networks

We have studied the emergence of mutual synchronization and activity propagation in coupled neural networks from rat cortical cells grown on a micro-electrode array for parallel activity recording of dozens of neurons. The activity of each sub-network by itself is marked by the formation of synchronized bursting events (SBE) - short time windows of rapid neuronal firing. The joint activity of two coupled networks is characterized by the formation of mutual synchronization, i.e. the formation of SBE whose activity starts at one sub-network and then propagates to the other. The sub-networks switch roles in initiating the mutual SBE. However, spontaneous propagation (initiation) asymmetry emerges - one of the sub-networks takes on the role of initiating substantially more mutual SBE than the other, despite the fact that the two are engineered to be similar in size and cell density. Analysis of the interneuron correlations in the SBE also reveals the emergence of activity (function) asymmetry - one sub-network develops a more organized structure of correlations. We also show activity propagation and mutual synchronization in four coupled networks. Using computer simulations, we propose that the function asymmetry reflects asymmetry between the internal connectivity of the two networks, whereas the propagation asymmetry reflects asymmetry in the connectivity between the sub-networks. These results agree with the experimental findings that the initiation and function asymmetry can be separately regulated, which implies that information transfer (activity propagation) and information processing (function) can be regulated separately in coupled neural networks.     

Photoreceptor topography of the retina in the adult pigtail macaque (Macaca nemestrina)

In spite of the crucial role retinal photoreceptors play in mapping optical images into a pattern of neural excitation, there are no complete studies of photoreceptor topography in any primate retina. We have measured the spatial density and inner segment areas of cones and rods across the whole mounted retinas of three adult pigtail macaques (Macaca nemestrina) and constructed maps of photoreceptor density and inner segment diameter. These retinas contain an average of 3.1 million cones (2.8-3.3 million), with an average peak foveal cone density of 210,000 cones/mm2 (190,000-260,000 cones/mm2). Cone density falls steeply with increasing eccentricity, to 100,000 cones/mm2 at 200 microns from the fovea, and to 50,000 cones/mm2 at 750 microns. Imposed on this gradient is a "streak" of higher cone density along the horizontal meridian. At equivalent eccentricities, cone density is higher in nasal and inferior retina. Cone inner segments increase in diameter from 2.3 microns at the foveal center to 11 microns in far temporal retina and 10 microns in far nasal retina. These retinas contain an average of 60.1 million rods (44.9-75.3 million). Rod density is zero within 20 microns of the foveal center, increases to the crest of a "rod ring" at the eccentricity of the optic disk, and then declines. Central rod topography is asymmetric, with higher densities in superior retina. Density along the crest of the rod ring peaks in superior retina at 177,000 rods/mm2, dips as low as 120,000 rods/mm2 along the horizontal meridian, and increases to about 150,000 rods/mm2 in inferior retina. Far peripheral rod topography is relatively symmetric around the fovea. Rod inner segment diameter ranged from 1.5 microns in the fovea to 4 microns at the temporal edge and 3.4 microns at the nasal edge of the retina. At eccentricities exceeding 6 mm, rod inner segment diameter was greater temporally than nasally. Cone inner segments cover 85-90% of the central fovea, with extrareceptor space accounting for the remainder. Cone coverage declines with increasing eccentricity to 20% at the temporal edge and 35% at the nasal edge of the retina. In contrast, rod coverage increases from zero at the foveal center to a maximum of 65% in temporal retina and 50% in nasal retina. The photoreceptor topography of the pigtail macaque is qualitatively similar to that of other macaques and to humans. Photoreceptor topography is formed by a complex interaction between regional changes in cone and rod density and inner segment diameter.     

Prefrontal unit activity of macaque monkeys during auditory and visual reaction time tasks

During a simple reaction time task using auditory or visual stimuli, a total of 96 single units were recorded from the dorsolateral prefrontal cortex of macaque monkeys. These monkeys were trained to depress a lever for a fixed period which produced a tone burst or a small spot of light. After a variable period, the stimulus intensity changed, and then, the monkey released the lever. Eighty-one cue-related units were classified into 3 types according to their decay time; that is, phasic, tonic and mixed. Phasic units (n = 19) showed a transient increase of discharge rate with a relatively short peak latency (70−300 ms). Of these, 17 units responded exclusively to either visual or auditory stimuli and two to both. Tonic units (n = 55) showed enhanced or suppressed activity, with longer latencies, which was sustained as long as the cue period continued. The temporal pattern of the discharge in 23 tonic units was found to be similar for both the auditory and visual cues. Seven mixed-type units showed combined phasic and tonic patterns. Lever release-related units (n = 15) were activated only during the period of lever release with no distinction in cue modality.It is suggested that the dorsolateral prefrontal cortex receives sensory inputs fairly discretely on the phasic-type neurons and that these sensory activities are transmitted to the tonic-type neurons which lead to an initiation of the lever release behavior.     

The organization of visual object representations: a connectionist model of effects of lesions in perirhinal cortex

We have developed a simple connectionist model based on the idea that perirhinal cortex has properties similar to other regions in the ventral visual stream, or 'what' pathway. The model is based on the assumption that representations in the ventral visual stream are organized hierarchically, such that representations of simple features of objects are stored in caudal regions of the ventral visual stream, and representations of the conjunctions of these features are stored in more rostral regions. We propose that a function of these feature conjunction representations is to help to resolve 'feature ambiguity', a property of visual discrimination problems that can emerge when features of an object predict a given outcome (e.g. reward) when part of one object, but predict a different outcome when part of another object. Several recently reported effects of lesions of perirhinal cortex in monkeys have provided key insights into the functions of this region. In the present study these effects were simulated by comparing the performance of connectionist networks before and after removal of a layer of units corresponding to perirhinal cortex. The results of these simulations suggest that effects of lesions in perirhinal cortex on visual discrimination may be due not to the impairment of a specific type of learning or memory, such as declarative or procedural, but to compromising the representations of visual stimuli. Furthermore, we propose that attempting to classify perirhinal cortex function as either 'perceptual' or 'mnemonic' may be misguided, as it seems unlikely that these broad constructs will map neatly onto anatomically defined regions of the brain.     

Impairments in visual discrimination after perirhinal cortex lesions: testing 'declarative' vs. 'perceptual-mnemonic' views of perirhinal cortex function

Two experiments tested the predictions of 'declarative' vs. 'perceptual-mnemonic' views of perirhinal cortex function. The former view predicts that perirhinal cortex lesions should impair rapidly learned, but not more slowly learned, visual discriminations, whereas the latter view predicts that impairments should be related not to speed of learning but to perceptual factors. It was found that monkeys with perirhinal cortex lesions were impaired in the acquisition and performance of slowly learned, perceptually difficult greyscale picture discriminations, but were not impaired in the acquisition of rapidly learned, perceptually easier discriminations. In addition, these same monkeys were not impaired in the acquisition or performance of difficult colour or size discriminations, indicating that the observed pattern of impairments was not due to ceiling effects or difficulty per se. These findings, taken together, are consistent with the 'perceptual-mnemonic' view that the perirhinal cortex is involved in both perception and memory, but are not consistent with the 'declarative' view that the perirhinal cortex is important exclusively for declarative memory, having little or no role in perception. Moreover, the results are consistent with the more specific proposal that the perirhinal cortex contributes to the solution of complex visual discriminations with a high degree of 'feature ambiguity', a property of visual discrimination problems that can emerge when features of an object are rewarded when part of one object, but not when part of another. These and other recent findings suggest the need for a revision of prevailing views regarding the neural organization of perception and memory.     

Primate spinothalamic pathways: II. The cells of origin of the dorsolateral and ventral spinothalamic pathways

The cells of origin of the dorsolateral (DSTT) and the ventral (VSTT) spinothalamic tracts were studied in 11 monkeys. The spinothalamic tract cells were retrogradely labeled by horseradish peroxidase (HRP) injected in the thalamus. All animals also received a midthoracic spinal cord lesion on the side ipsilateral to the thalamic injections. The distribution of labeled cells found in these animals throughout the cervical segments was similar to animals with no spinal cord lesions. Five animals had ventral quadrant lesions to demonstrate the cells of origin of the DSTT. In macaques with complete ventral quadrant lesions, more than 80% of the HRP label in the contralateral L4-L7 segments was located in lamina I, while in squirrel monkeys, the label in the contralateral lower lumbar region was distributed between laminae I-III and IV-VI. Few labeled cells were found in laminae VII-X. Six animals received dorsolateral funiculus lesions to demonstrate the cells of origin of the VSTT. In animals with adequate lesions, 84-99% of the contralateral HRP label in L4-L7 was located in laminae IV-X. Macaques had a larger percentage of labeled cells located in lamina I than squirrel monkeys. The results indicate the existence of two spinothalamic pathways in the primate. The DSTT was calculated to compose about one fourth of the total spinothalamic population.     

Primate spinothalamic pathways: III. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways

The termination sites of the dorsolateral (DSTT) and ventral (VSTT) spinothalamic pathways were determined by using anterograde transport of horseradish peroxidase from the lumbar spinal cord in primates. One animal had no spinal cord lesion, while of two other animals, one received a midthoracic dorsolateral funiculus lesion, and the other received a midthoracic ventral quadrant lesion contralateral to the injection. The thalamic label in the animal with no spinal cord lesion was much less than the label in the two animals with spinal lesions. Moreover, in the animals with spinal lesions, HRP-labeled cells were found within the thalamus. Therefore, the remaining six animals received ipsilateral hemisections and bilateral dorsal column lesions, irrespective of the contralateral lesions. The thalamic label in the animals without contralateral lesions were assumed to represent the total spinothalamic input to the diencephalon. In these animals, label was located mainly in suprageniculate and pulvinar oralis, caudal and oral divisions of ventral posterior lateral nucleus, the lateral half of ventral posterior inferior nucleus, and zona incerta, while in the medial thalamus label was primarily in two distinct bands in medial dorsal nucleus and in the posterior dorsal portion of central lateral nucleus. Scattered lighter labeling was found in other thalamic nuclei. The pattern of terminal labeling observed in the ventral posterior lateral region was arranged in patches, while elsewhere in the thalamus a more uniform labeling pattern was observed. The thalamic label in animals with contralateral ventral quadrant lesions represented the terminations of the DSTT, while the label in animals with contralateral dorsolateral funiculus lesions represented VSTT terminations. The labeling pattern was similar between these two groups. However, there were small differences between them. These results indicate that DSTT and VSTT terminations largely overlap and innervate the lateral and medial thalamamus.     

Tempo of neurogenesis and synaptogenesis in the primate cingulate mesocortex: Comparison with the neocortex

In the neocortex, the onset of the rapid phase (phase 3) of synaptogenesis occurs after the end of neurogenesis. However, we still do not know whether or not these two developmental events are causally related. The present study compares the time-course and tempo of neurogenesis and synaptogenesis in the anterior cingulate cortex (area 24 of Brodmann) and in the primary visual cortex (area 17) in a series of pre- and postnatal rhesus monkeys. Autoradiographic analysis of animals fetally injected with ³H-thymidine showed that all neurons destined for area 24 are generated by embryonic day 70, which is 30 days earlier than in area 17. The rapid phase of synaptogenesis in area 24 starts during the third embryonic month and continues at the same rate through the remainder of gestation and the first 2 months after birth, as has been seen in neocortical areas examined previously. Statistical analysis of the linear portions of the rapid phase indicates that, although neurogenesis in area 24 is completed 1 month earlier than in area 17, the rapid phase of synaptogenesis occurs 41 days later. Moreover, the tempo of synaptic accretion was remarkably similar to that in motor, somatosensory, visual, or associational areas. All were grouped within the same time window of about 40 days, centered at birth. After the second postnatal month, synaptic density in area 24 remains at a high level until sexual maturity. This work shows that the rapid phase of synaptogenesis in the cingulate mesocortex is not linked temporally to the end of neurogenesis. We suggest that it is regulated by the same genetic or humoral factors that control synaptogenesis in the phylogenetically newer neocortical areas. © 1995 Wiley-Liss, Inc.