Control of somatosensory input by cerebral cortex.

Direct stimulation of the pyramidal tract increases the size of the excitatory receptive fields of neurons in the somatosensory cortex of the cat. This effect reflects greater transmission of cutaneous information through the dorsal column nuclei as a result of the facilitation of cells in these nuclei by pyramidal tract fibers.

Doubly projecting neurons of the dorsal column nuclei.

In a sample of 1700 neurons recorded from the dorsal column nuclei of the cat, 44 (2.6%) were found to send an axon down the dorsal spinal cord. Fully 70% of these caudally projecting neurons also sent an axon to the ventral thalamus. Nearly all had small cutaneous receptive fields distally on the forelimb and displayed response properties similar to other neurons of the dorsal column nuclei. Most were isolated along the lateral and medial margins of the cuneate nucleus rostral to the obex, and many were excited or inhibited by pericruciate cerebral stimulation. A few clearly were excited monosynaptically from the contralateral cerebral cortex at a latency that required the largest pyramidal tract fibers. These neurons probably comprise an important subset that regulates the flow of sensory information in spinal and brainstem somatic sensory pathways.

Relation between brain and body growth in allied rats.

Cranial volumes were measured from museum specimens of wild-caught and laboratory-born Allied rats from eastern Australia. The relation of these volumes to body weight and body length, and also to age at death in the laboratory-reared sample, was determined. Growth of both brain and body was rapid during the first three postnatal months and slowed markedly over the next month, but appeared to continue at a very slow rate throughout life. In particular, the major surge in brain growth occurred in the first three postnatal weeks. Modified Gompertzian growth functions describe the pattern of growth quite well, though the nature of the data precluded highly sensitive fits. Three features were clear: 1) the rate of slowing of growth was about the same for all variables, 2) growth appeared to continue throughout the life of the animal, and 3) the trajectory of brain growth led that of body growth by about four days. The pattern of growth in Allied rats is similar to that of laboratory rats and probably to those in other murids.

Notes on the hypothesis of columnar organization in somatosensory cerebral cortex.

The five principles that make up the hypothesis of columnar organization for somatosensory cerebral cortex are considered in relation to their experimental foundations. Several important implications that flow from these principles are examined, and the requirements for their experimental evaluation are elaborated. A number of specific measurements are found to be lacking in necessary precision: the total amount of somatosensory cerebral tissue in each area, the sizes, shapes and degree of overlap of aggregate excitatory receptive fields, the number of distinct modalities and the modality equivalences in our different linguistic categories, the sizes and shapes of the columns and, finally, the number of columns for each modality and the total number of columns. Some possible columnar arrangements are set up, and their experimental detectability is assessed, using optimal conditions and values. It is shown that if a columnar structure exists, it defies clear detection by current neurophysiological techniques and experimental approaches. The concept of an 'elementary functional unit' is found to need clear definition; taken at face value, it is shown to yield some rather unusual predictions. The conditions under which the hypothesis of columnar organization can be distinguished from its rival, the hypothesis of topographic organization, are also reviewed.

Pyramidal tract fiber spectrum in rats, with comments on cats and man.

Most CNS fiber spectrums are unimodal and strongly positively skewed, with many small and few large fibers. This study shows that the pyramidal tract (PT) fiber spectrum of a rat can be calculated as the sum of three distributions of myelinated axons, each derived by normal Gompertzian growth from three normal distributions of protoaxons. Histological measurement of the rat PT determined the values entered into the model, thus forcing a unique solution. The model was generalized to cats and man by assuming values for which no experimental data was available; the simulated PT fiber spectrums closely matched the observed PT fiber spectrums, in both species. It is concluded that normal Gompertzian growth is sufficient to account for the specific shape of the fiber spectrum, with no recourse to morphogenetic sculpting. The overproduction of cells during growth, and death of cells during development, may regulate the total number of neurons in different areas of cortex, but plays no role in determining the specific shape of the PT fiber spectrum.

Fiber analysis of the pyramidal tract of the laboratory rat.

Light and electron microscopic study of the pyramidal tract of the laboratory rat at a midbulbar level revealed the total number of myelinated fibers on one side to be about 200,000. They ranged from 0.2 micron to more than 5 microns, but clustered strongly in the neighborhood of 1.0 micron (mode of 0.9 micron and mean of 1.2 micron), forming the highly skewed fiber spectrum so familiar for mammalian pyramidal tracts and other central fiber pathways. Numerous small clusters of unmyelinated axons were found scattered throughout the tract, adding another 100,000 axons to the estimated number. Not only were the fibers exceedingly small, but also the degree of myelination relative to axon diameter varied widely, suggesting that conduction speed within the tract is not optimal for all fibers. In fact, about half of the fibers in the pyramidal tract would, in theory, conduct faster if they had no myelin wrapping.

Pattern of myelination in the pyramidal tract of the rat.

The size and myelination of midbulbar pyramidal tract axons were measured by electron microscopy in the rat. We found that myelin thickness did not increase linearly with fiber size; rather, it took on certain preferred thicknesses almost independently of fiber size. This pattern of growth and development is fundamentally different from that of peripheral nerve and may be important for the physiology of the pyramidal tract.

Superposition of antidromic responses in pyramidal tract cell clusters.

Large-seeing-distance microelectrodes were used to record simultaneously the activity of several pyramidal tract neurons in cerebral cortex. When activated antidromically, these neurons responded simultaneously, forming a "stack" of superimposed spikes, rather than responding at different times within the 10-ms time interval during which they might be expected to respond. Using a variety of spike collision tests, we found that these individual spikes arose from separate sources and reflected the activity of individual, albeit neighboring, pyramidal tract neurons. The collaterals of neurons within a stack projected to different structures, further verifying that separate neurons were involved. Such synchrony of antidromic activity among neighboring pyramidal tract neurons is an exceedingly low-probability event, if neighboring fibers conduct independently of one another. Our results imply that fibers from small clusters of neurons in the cortex assemble to form synchronously conducting bundles of fibers within the pyramidal tract.

Influence of molecular layer on pyramidal tract neurons.

Activity in layer I increases the excitability of pyramidal tract (PT) neurons, the effect being stronger on slow than on fast PT neurons. Extracellular recordings were made from lateral postcruciate cortex of domestic cats, using antidromic activation from medullary pyramid to identify and classify PT neurons. Their responses to contralateral forepaw (CFP) and direct cortical (Ctx) stimulation, 3 to 4 mm caudal to the recording site, were determined before and after placement of vertical cuts between the Ctx stimulating and recording sites. These cuts had a minor effect on the responses of PT neurons to CFP stimulation, but a strong effect on the responses to Ctx stimulation. Cuts through layers I and II markedly delayed the responses of slow PT neurons, but had no effect on fast PT neurons. After deeper cuts (II/III through V/VI), half the fast and half the slow PT neurons failed to respond to Ctx stimulation. Of those that did, fast PT responses were markedly delayed, but slow PT responses were only mildly affected. The Ctx-CFP interactions showed the familiar facilitation-depression sequence. The period of depression was unaffected by any of the vertical cuts, but disappeared after undercutting the stimulus site below layer VI. The period of facilitation depended primarily on layer I for its production, although deeper layers also contributed to the facilitation of fast PT neurons.

Antidromic response to medullary pyramid stimulation in rats and its relation to that in cats.

The response evoked in the cerebral cortex of laboratory rats after stimulation of the medullary pyramid is surface-positive. It begins 0.9-1.6 ms after the stimulus, attains peak amplitude (up to 2 mV) in 0.8-1.2 ms and lasts 2-4 ms. It occurs throughout the anterior two-thirds of the dorsal cortex and is largest lateral to bregma, with a secondary maximum in the somatosensory area II. Although it depends on antidromic conduction in pyramidal tract fibers for its production, it varies in amplitude, configuration and latency at different recording sites and at the same sites on repeated trials. It reverses polarity deep in the cortex to become a large, negative wave deep in layer V, and maintains that polarity into the white matter. Current source density analysis reveals a strong sink in layer V, with a strong source just superficial to that sink and a weaker source in layer VI. The antidromic response disappears during spreading depression, but recovers more rapidly than the primary response evoked by skin stimulation. It decreases progressively in amplitude with continuous 200-Hz iterative stimulation, and recovers slowly at the end of stimulation. The primary response evoked by contralateral forepaw and hindpaw stimulation is highly localized, being entirely within the antidromic response distribution. The antidromic response in laboratory rats consists of a small, surface-positive component analogous to the pure antidromic response of cats, and of a large, surface-positive response analogous to that found in woodchucks, rabbits, opossums and slow lorises. It is argued that this latter response results from synaptic action in pyramidal tract axon collaterals, probably onto cells in layer V, rather than being a purely antidromic event.

Effect of optic nerve stimulation on neurons in pericruciate cortex of cats.

The distribution of optic chiasm input to different types of neurons in pericruciate cortex of cats agreed with previous work using light flashes. Neuron response times served to differentiate the input pathways to pericruciate cortex, and the types of neurons they influence. Input from the optic chiasm arrived in three distinct surges: the first via the superior colliculus, the second via an unidentified pathway, and the third via the visual cortex. A fourth, diffuse surge arrived in the postcruciate cortex via some unidentified pathway. Stimulation of the contralateral side of the optic chiasm had a weaker effect than stimulation of the ipsilateral side; it evoked activity at a higher threshold, with fewer spikes per response, and at a longer latency. The difference in response latency between the two sides was largest on neurons responding to the first surge, decreasing in later surges, and being least on those neurons responding to the last surge. About 2.3% of the postcruciate and 15% of the precruciate neurons responded only to optic chiasm stimulation; they were isolated in the granular layers, and their responses could not be influenced by prior cutaneous input. It is suggested that much of the visual input to pericruciate cortex serves to modulate on-going cortical output and, thereby, the behavior of the animal.

Cerebral response to pyramidal tract stimulation in wood rats and its relation to laboratory rats.

The cerebral response evoked by stimulation of the bulbar pyramidal tract in wood rats, like that of laboratory rats, consisted of a small alpha wave, almost obscured by a very large, superimposed r wave. The alpha wave behaved like a purely antidromic response, whereas the r wave behaved like a postsynaptic response, including a marked variability in amplitude on repeated trials. The contralateral forepaw and hindpaw motor sites mapped onto the somatic sensory foci for these two paws; further examination showed that the somatic sensory and motor representations were largely superimposed. An incipient sagittal fissure 1.5 mm lateral to the midline marked the boundary between limbic and neocortex. Because of their structural similarities and their differences in somatic sensory and motor organization, wood rats and laboratory rats are prime subjects for comparative study of the role of amalgamated and separate sensory and motor cortices in regulating movement and behavior.

Spatial distribution of modalities and receptive fields in sensorimotor cortex of awake cats.

A sample of 504 single neurons isolated in three curvilinear arrays of 10 closely spaced tracks in primary somatosensory and in pericruciate sensorimotor cortex was studied in two awake, restrained domestic cats. Modality sensitivity and receptive field size and location were assessed for each neuron, along with response adaptation rate and state of arousal at the time of recording. Reconstruction of the spatial distribution of these response properties failed to show any simple organization, beyond general somatotopy. The spatial distribution of modality sensitivities was quantitatively tested in relation to a strict columnar model and to a random model; the data could not be clearly distinguished from the random model, in any of the three recording arrays. Observations made on two or more neurons isolated simultaneously at the same recording site revealed that few shared both modality and receptive field (RF) in common. Among the simultaneously recorded neurons, five-ninths showed disparate modality sensitivities and two-thirds showed limited or no RF overlap. Many pairs of neurons showing the same modality sensitivity showed limited or no RF overlap, and many pairs showing partial or complete RF overlap showed disparate modality sensitivities. Hence, the data failed to support any model of cerebral organization that features local, bounded regions within which all neuron response properties are the same and, in particular, the model of columnar organization. On the other hand, models that feature intermingled local clusters of neurons (a cluster consists of neurons that share the same response properties) are not excluded by the data.

Intrinsic features contributing to spike train patterning in proprioceptive cuneate neurons.

The intrinsic processes contributing to the three discharge patterns of proprioceptive cuneate neurons described by Surmeier and Towe were studied experimentally and with computer simulation. Examination of the alterations in excitability produced by antidromic activation suggested that a prolonged inhibition was a concomitant of discharge in proprioceptive cuneate neurons. Computer simulation was performed to test the possible roles of inhibitory hyperpolarizing processes in governing the observed discharge patterns. These simulations used two constant threshold models. The simplest model linearly integrated synaptic potentials until the spike threshold was reached. After the discharge, synaptic potentials that preceded the spike were ignored (i.e., the model was "reset"). The second model was similar to the first except that following a spike two hyperpolarizing processes were activated and preceding events continued to play a role in membrane potential. Simulation of class A spike trains that possessed positive correlations between nearby intervals was successful only with a resetting model. This suggested that class A neurons have fast, no-memory postspike conductance changes, which effectively shunt synaptic charge. Simulation of class B spike trains was possible with the nonresetting model. At least two periodic inputs, which evoked brief, relatively large EPSPs, were required. In addition, a prominent, fast, spike-dependent hyperpolarization and a small-amplitude, slow hyperpolarization were required. Simulation of class C spike trains was also possible with the nonresetting model. Several periodic inputs were required; one input had to evoke a slow suprathreshold EPSP. In contrast to class B simulations, class C spike train simulation required that a large-amplitude, slow hyperpolarization, as well as a brief hyperpolarization, following spike initiation. The results of class B and C simulations suggested that these two groups differed primarily in the amplitude of a slow, hyperpolarizing, postspike conductance. Some role may also be played by the time course of the driving EPSPs.

Properties of proprioceptive neurons in the cuneate nucleus of the cat.

Fifty-two slowly adapting proprioceptive neurons in the cuneate nucleus of chloralose-anesthetized cats were studied. Recordings were made from 3 mm rostral to the obex to 5 mm caudal. The highest densities of proprioceptive neurons were found above and more than 3 mm caudal to the obex. Analysis of the spike trains produced with the forelimb held fixed revealed three basic periodic patterns. Neurons exhibiting these patterns were partitioned into three groups, referred to as the A, B, and C classes. Class A neurons (42%; 22/52) produced regular spike trains that were qualitatively similar to muscle spindle fibers. Interval distributions for this class were typically unimodal and slightly positively skewed. Adjacent intervals were frequently positively correlated. Spectral analysis suggested that 91% of class A spike trains had one to two periodic components. Class B neurons (21%; 11/52) had additional spikes interposed in their periodic discharge; these "interrupting" spikes did not significantly alter the timing of the dominant periodic discharge. Interval distributions were typically bimodal and adjacent intervals were negatively correlated. Spectral analysis suggested that two or more periodic components were present in their spike trains. Class C neurons (36%; 26/52) had spike trains with a basic rhymicity, but when this specific discharge was interrupted, the subsequent interval was near modal length; thus, they were "reset." Interval distributions were usually multimodal and adjacent intervals were frequently negatively correlated. Spectral analysis suggested that C spike trains usually had four or more periodic components. Estimates of information-carrying capacity of each class using a mean rate code and those of primary muscle spindle fibers suggested that a sizable information loss may occur in synaptic transmission. This potential loss was smaller for A-neurons (40%) than for B- (69%) or C-neurons (64%). Electrical stimulation of cutaneous structures influenced 55% (22/52) of the sample. All were members of the B and C classes. Responses were typically biphasic. The cutaneous receptive fields nearly always included a portion of the forepaw. No relationship was found between movement sensitivity and receptive field topography. Contralateral input was found in half (10/20) the neurons tested.

Pattern of pyramidal tract collateralization to medial thalamus, lateral hypothalamus and red nucleus in the cat.

Stimulating electrodes were placed in the red nucleus, lateral hypothalamus and medial thalamus in order to determine whether pyramidal tract (PT) neurons send collaterals to those sites. The red nucleus projections are well-known, but it was discovered that PT neurons also project into the other two sites. All of the fibers that sent collaterals to all three sites originated from fast PT neurons. Those that responded to stimulation of the skin and that sent collaterals to two or three sites were predominantly fast PT neurons. Those neurons that responded only to cerebral peduncle stimulation were predominantly slowly-conducting, when compared with the set of PT neurons in response to cerebral peduncle stimulation. The patterns of collateral branching to red nucleus and to lateral hypothalamus were similar, suggesting a similar synaptic effect of the pyramidal system in the two sites. Measurement of the speed of conduction from three sites along the length of corticospinal fibers revealed large changes on some, but not all, fibers; there was no evident pattern to these changes that might be associated with collateral branching. A new hypothesis concerning the functional role of fast PT neurons in regulating movement is presented.