Primate S1 cortex
|Jon H. Kaas (2013), Scholarpedia, 8(6):8238.
|revision #150514 [link to/cite this article]
Introduction All mammals have a region of somatosensory cortex that is called primary somatosensory cortex or S1. The term stands for the cortical representation of touch that was discovered and described first. The name became useful when other cortical representations of touch were discovered, starting with the second representation, S2. Early investigators used Roman numerals for cortical areas, but now Arabic numerals have become common. S1 or SI has also been referred to as SmI by Clinton Woolsey and co-workers to recognize that somatosensory area I also has a small motor (m) component. S1 was first described in cats in 1941 as a single, systematic representation of the contralateral body surface located in parietal cortex of each cerebral hemisphere. The experimental results, obtained by recording the activities of neurons in many locations throughout the representation with penetrating electrodes, indicated that medial recording sites in S1 are activated by touching the hind foot, more lateral sites by the forefoot, and the most lateral sites by the face. The disclosed representation of the contralateral body surface was described as somatotopic because neurons in S1 were activated from tail to tongue in a mediolateral sequence across cortex. From this beginning, S1 has been considered to be a single, roughly somatotopic representation of the contralateral body surface that has a characteristic mediolateral orientation and internal organization across mammal species.
Subsequent studies have associated S1 across species with a strip of cortex that has a distinctive laminar appearance, as cortical layer 4 is densely packed with small neurons that receive activating inputs via axons from the ventroposterior nucleus of the somatosensory thalamus. S1 sends somatosensory information to other areas of somatosensory cortex, including S2. Neurons in S1 (3b) are activated by light touch on the skin. Each neuron is typically activated by inputs from a small region of skin, such as part of the tip of a single finger, and this activating surface is called the receptive field. Touch outside the activating receptive field usually reduces the response to touch within the receptive field, or has little effect. After damage to S1, small objects may be unrecognized by touch and ignored, and grasping with the hand may be difficult.
The S1 Region of Monkeys and Humans In early studies of parietal cortex in monkeys and humans, S1 was not correctly identified. In brief, the region often called S1 in monkeys, apes and humans mistakenly included four different architectonic fields that are called Brodmann's areas 3a, 3b, 1 and 2 after the early studies of Brodmann over 100 years ago. Each of these areas has its own representation of the body, but only the area 3b representation has the defining characteristics of S1 as described in cats, rats and other mammals. These include having a representation that is almost exclusively responsive to the activation of touch receptors, a characteristic internal somatotopic organization that is coextensive with a cytoarchitectonic field that has a layer 4 that is more densely packed with small neurons than adjoining areas, and more dense inputs from the ventroposterior nucleus to layer 4 than any other area. In monkeys and humans, areas 3a, 1 and 2 contain additional representations of the body surface that parallel S1 in somatotopic organization. While area 3a is dominated by inputs from muscle spindle receptors for proprioception, and area 2 also has major inputs from proprioceptors, neurons in area 1 and at least much of area 2 respond well to light touch. The responsiveness of these other cortical areas to touch was part of the problem of defining S1 in early recording studies of somatosensory cortex in monkeys and humans. The other part of the problem was that most of the recordings in early experiments were done with surface electrodes on the exposed surface of somatosensory cortex, and most of area 3b is hidden on the caudal bank of the deep central sulcus of Old World monkeys, apes and humans. Thus, most of the recordings were from areas 1 and 2, as they were largely on the cortical surface. As areas 1 and 2 paralleled area 3b in their mediolateral somatotopic sequence, recordings from areas 1 and 2 were considered to be from S1. Furthermore, as areas 3b and 3a were also recognized as responsive to touch, all four architectonic strips, each now known to contain a separate representation of the body, were considered to be parts of a single representation, S1, by early investigators. This misnomer persists in many reports today, as S1 is often used to refer broadly to anterior parietal cortex in current studies in humans and monkeys. Since each of the four architectonic fields has its own representation and functions, this careless misuse creates confusion and impedes further progress and understanding.
Here, S1, sometimes called S1 proper, refers only to the area 3b representation. Electrophysiological evidence that area 3b of monkeys contains a complete representation of touch receptors of the contralateral body surface was first obtained in New World owl and squirrel monkeys, as these monkeys had most of area 3b exposed on the surface of the brain. Owl monkeys have only a shallow dimple in cortex rather than a deep central sulcus, and squirrel monkeys have only a short, shallow central sulcus. Thus, it was possible to visually place microelectrodes into hundreds of sites in area 3b of these monkeys, record the responses of neurons at these sites to touch on the body, and determine where on the body touch activated neurons for each site. In this manner, receptive fields were determined for neurons at each site. By outlining the cortical territories of sites where neurons were activated by the same body part, such as digit 1 (thumb), maps of where body parts are represented in area 3b were constructed (Fig. 1). The results indicate that the area 3b representation (S1) extends from the medial wall of the cerebral hemisphere lateral to the margin of the lateral sulcus where it curves rostrally to end near the ventral margin of the hemisphere. Part of the posterior leg and tail are represented on cortex of the medial wall, followed in a mediolateral sequence by representation of the toes and foot, anterior leg, genitals and trunk, arm, and hand and digits. In cortex just lateral to the hand representation, parts of the face are represented, followed as the representation curves anteriorly by the teeth, tongue, and finally the ipsilateral teeth and tongue as these important mouth parts are represented in S1 of both cerebral hemispheres. The cortical territories devoted to the teeth are activated by sensitive receptors around the roots of the teeth that respond to when the teeth are touched. Large proportions of area 3b (S1) that respond to touch on the digits of the hand, the teeth and the tongue, show how important these parts of the body are in providing touch information to the cortex.
A similar area 3b (S1) representation has been revealed in all studied primates. This representation was examined first in microelectrode mapping studies in other New World monkeys, such as marmosets and titi monkeys, and in prosimian galagos, where a central sulcus is absent, but area 3b has been explored in some detail in Old World macaque monkeys with electrodes that penetrated into the cortex along the posterior bank of the central sulcus. In addition, a similar organization of area 3b has been determined in less detail in humans using functional magnetic imaging.
Area 3b (S1) is bordered along its anterior border in all primates by a narrow strip of cortex, area 3a that is activated mainly by muscle spindle receptors and other proprioceptors that respond while the body moves and provide a sense of the position of body parts. This information is relayed to area 3a from a separate nucleus in the somatosensory thalamus called the ventroposterior superior nucleus (VPS). A similar proprioceptive strip of cortex is found along the anterior border of S1 in at least most mammals. As area 3a receives inputs from area 3b (S1), neurons in area 3a also responds to touch, but such responses are often not apparent in anesthetized animals. Area 3a provides an important source of proprioceptive information to the anteriorly adjoining motor cortex. The posterior border of area 3b (S1) is formed by area 1 in primates. Area 1 receives direct inputs from area 3b, and it appears to be dependent on area 3b for most of its neural responsiveness. Thus, area 1 is a second order cortical area, much like the second visual area, V2, or the second somatosensory area, S2. Nevertheless, area 1 does receive somatosensory inputs directly from the ventroposterior nucleus of the somatosensory thalamus, but to layer 3 rather than layer 4. In most monkeys neurons in area 1 are highly responsive to touch, even in anesthetized animals, and area 1 contains a complete somatotopic representation of the contralateral body that parallels that in area 3b (S1), thus accounting for the confusion of including area 1 in S1 in early studies on monkeys. However, the internal order of the somatotopy in area 1 differs from that in area 3b in that area 1 basically forms a mirror reversal of the somatotopy in area 3b. For example, the digits tips are represented anteriorly in area 3b and posteriorly in area 1. A variously named strip of cortex with inputs from S1 exists along the posterior border of S1 in most mammals, but this cortex is not highly responsive to touch in anesthetized animals. This is also the case in prosimian primates, and even in marmoset monkeys, where the area 1 cortex is unresponsive or poorly responsive to touch in anesthetized animals. Thus, in these primates, and in most mammals, there would be little chance of mistakenly including the area 1 territory in S1 in recording experiments.
In a similar manner, a strip of cortex along the posterior border of area 1 is variably responsive to touch across primates. At least in parts of this area 2 field neurons are highly responsive to touch in macaque monkeys, and this responsive region was included in S1 in early studies in Old World macaque monkeys. Area 2 receives proprioceptive information directly from the ventroposterior superior (VPS) nucleus of the thalamus, and touch information from area 3b (SI) and area 1. Neurons in area 2 are not very responsive to touch in some New World monkeys, and in prosimian primates, and even in Old World monkeys, neurons in all parts of the architectonically defined field are not highly responsive to touch. Area 2 provides an important source of somatosensory information to posterior parietal cortex.
Kaas, J H (1983). What, if anything is SI? Organization of first somatosensory area of cortex. Physiological Reviews 63: 206-230.
Qi, H-X; Preuss, T M and Kaas, J H (2007). Somatosensory areas of the cerebral cortex: Architectonic characteristics and modular organization. In: E Gardner and J Kaas (Vol. Eds.), The Senses: A Comprehensive Reference, Vol. 6 Somatosensation (pp. 142-169). London: Elsevier.