Study guidelines
- 1.
The cerebral cortex is the part of the body that makes us truly human. Its structure is enormously complex, and assignment of functions to different parts is made difficult, and often unrealistic, by the multiplicity of interconnections.
- 2.
Sensory, motor, and cognitive areas of the cortex are taken in turn.
- 3.
Although damage often leads to permanent disability, the plasticity of the cortex is of special interest to all concerned with neurorehabilitation. Examples are taken from sensory and motor areas.
Structure
The cerebral cortex ( pallium; Gr. ‘shell’) varies in thickness from 2 to 4 mm, being thinnest in the primary visual cortex and thickest in the primary motor area. More than half of the total cortical surface is hidden from view in the walls of the sulci. The brain contains approximately 86 billion neurons (the cerebral cortex contains only 19% of this total but represents 81% of the brain’s mass), a similar number of neuroglial cells, and a dense capillary bed.
Microscopy reveals the cortex to have a layered or laminar appearance reflecting the arrangement of its cells and nerve fibres as well as a radial organisation of these cellular elements. The general cytoarchitectonic structure (patterns based on the appearance of cells; patterns based on the arrangement of myelinated fibres are referred to as the myleoarchitectonic structure) varies in detail from one region to another, permitting the cortex to be mapped into dozens of histologically different ‘areas’. Considerable progress has been achieved in relating these areas to ‘specific’ functions, but while conceptually useful, the relationships represent a simplification because they often represent only nodal points of more widespread functional systems with connectivity to other parts of the brain.
Laminar organisation
A laminar arrangement of neurons is apparent in sections taken from any part of the cortex. Phylogenetically ‘old elements’, including the paleocortex (olfactory cortex) and the archicortex (hippocampal formation and dentate gyrus; concerned with memory) are made up of three cellular laminae, but transition into six laminae is seen in the neocortex ( neopallium; or isocortex , which refers to its uniform cortical neurogenesis that results in these six laminae) representing the remaining 90% or the majority of the cerebral cortex.
Cellular laminae of the neocortex (Figure 29.1)
- I.
The molecular layer contains the tips of the apical dendrites of pyramidal cells (see below), and the most distal branches of axons projecting to the cortex from the intralaminar nuclei of the thalamus.
- II.
The outer granular layer contains small pyramidal and stellate cells (see below).
- III.
The outer pyramidal layer contains medium-sized pyramidal cells and stellate cells.
- IV.
The inner granular layer contains stellate cells receiving afferents from the thalamic relay nuclei. ( Stellate [granule] cells are especially numerous in the primary somatic sensory cortex, primary visual cortex, and primary auditory cortex, which all receive afferent sensory information. The term granular cortex is applied to these areas. In contrast, the primary motor cortex that gives rise to corticospinal and corticobulbar projections contains relatively few stellate cells in lamina IV and prominent pyramidal cells in laminae III and V, which further obscure individual layers. This area is called the agranular cortex .)
- V.
The inner pyramidal layer contains large pyramidal cells projecting to the striatum, brainstem, and spinal cord.
- VI.
The fusiform layer contains modified pyramidal cells projecting to the thalamus.
Columnar organisation (Figure 29.1)
While the laminar organisation of the cerebral cortex is often emphasised, there is also a radial or ‘columnar’ organisation of these cellular components. This columnar organisation of the neocortex provided the initial conceptual framework to investigate the functionality of its neuronal components in the somatic sensory cortex of animals. This radial grouping of cellular components was felt to represent discrete units with similar physiologic findings and forming the building block for more complex functions. Groups of columns could form modules, characterised by dealing with different aspects of a specific sensory modality or function. It is now clear that columns are not homogeneous across the cortex because multiple characteristics can vary, including their actual cellular constituents and their numbers, ontogeny, synaptic connectivity, and molecular markers, which all contribute to variable functional and stimulus response properties. As an organising principle the concept of a column is helpful, but it is equally useful to consider the cortex as being organised in both horizontal (laminar) and vertical (radial) dimensions. While not resulting in an analogous structure (column) with visibly discrete edges, this concept is more faithful to the underlying anatomy, observed experimental functionality, and the ‘economy’ and plasticity that exist within the cortex. Interconnectivity between groups of columns allows more complicated activities, behaviour or cognitive, to emerge.
This underlying ‘circuitry’ of cortical organisation can result in the cells of each column becoming modality (functionally) specific as their components ‘process’ information. However, the ultimate response of the projection neurons within those columns can differ significantly based on varying stimulus parameters and inputs for each neuron. For example, a given column may respond to movement of a particular joint but not to stimulation of the overlying skin; however, if circumstances differ, so may their ultimate response.
Cell types
Cortical neurons morphologically comprise two broad groups. The majority, 60 to 85%, are pyramidal neurons (referring to their shape) that are the sole output (and major input) of the cortex and the rationale for their alternate name, cortical projection neurons; their projections are excitatory and glutamatergic. The remaining 15 to 40% are nonpyramidal or interneurons and, while their connectivity remains local within the cortex, they significantly influence and modulate cortical activity; they are predominantly inhibitory and γ-amino- butyric acid (GABA)ergic. Within each group, multiple subtypes can be distinguished based on morphology, connectivity, electrophysiological properties, cellular lineage, physiologic properties, molecular markers, and so on. (Examples of the principal morphologic and functional cell types include pyramidal cells, spiny stellate cells [modified pyramidal cells], and the group of nonpyramidal inhibitory interneurons [ Figure 29.2 ].)
- •
Pyramidal cells have a pyramid-like shape with the apex pointed toward the surface and with cell bodies ranging in height from 20 to 30 μm in laminae II and III to more than twice that height in lamina V. Largest of all, at 80 to 100 μm, are the giant pyramidal cells of Betz in the motor cortex. The single apical dendrite of each pyramidal cell reaches out to lamina I, often ending in a tuft of dendrites. Several basal dendrite branches arising from the basal ‘corners’ of the cell extend radially within their respective laminae. The apical and basal dendrites branch freely and are studded with dendritic spines. Most pyramidal cells are found in cortical layers II to III and V to VI. The axons of pyramidal cells arise from their base and give off recurrent branches, capable of exciting neighbouring pyramidal cells, before entering the underlying white matter.
- •
Spiny stellate cells are one type of atypical pyramidal cell, located in laminae IV and abundant in the primary sensory cortex. Their spiny dendrites are limited to laminae IV, but their axons may ascend or descend, making excitatory glutamatergic synaptic contacts with pyramidal cells. They receive most of the thalamic afferent input to laminae IV and likely have a major role in its radial propagation.
The nonpyramidal, inhibitory interneurons share GABA as their neurotransmitter, but otherwise are morphologically heterogeneous and classified in various ways. (Neocortical neurons are represented by a complicated and evolving nomenclature. Smooth stellate [or granule] cells are found in all cortical layers; their dendrites radiate in all directions, and their axons form an arborisation locally within that same territory, often referred to as local plexus neurons. However, neurogliaform , chandelier, and basket cells are all considered special classes of stellate cells despite their unique morphologic characteristics. Our only advice is when encountering the term granule or smooth stellate cell, conceptually substitute the word interneuron to guide your reading and comprehension.) For our organisational purposes they can be subdivided into three large families based on markers their interneurons express: parvalbumin, somatostatin, and the serotonin (5-hydroxytryptamine; 5HT) 3a receptor (5HT3aR).
- •
Parvalbumin expressing interneurons have nonspiny dendrites and receive excitatory inputs from the thalamus and cortex but inhibitory inputs from other similar interneurons. They are believed to play a role in stabilising the activity of cortical networks of cells. As is the case in the cerebellar cortex ( Chapter 25 ), these neurons exert a focusing action in the cerebral cortex by silencing weakly active cell columns. Chandelier cells (so named because of the candle-shaped clusters of axoaxonic boutons) are most common in lamina II, synapse on the initial axon segment of pyramidal cells, and principally affect cortico–cortical connections. Basket cells are predominantly found in laminae II and V, and as their name implies their axons form pericellular baskets around pyramidal cell bodies, but also their distal dendrites and axons and other basket cells ( Figure 29.3 ).
- •
Somatostatin-expressing interneurons are exemplified by the Martinotti cells , located in laminae V and VI, which send their axons into lamina I. Receiving input from pyramidal cells, they can cause lateral restriction of activation and may serve to integrate nonsensory input that results in behaviour-dependent control of dendritic integration of stimuli by their pyramidal cells.
- •
5HT3a-expressing interneurons are a heterogeneous group, but represent the most numerous interneurons in superficial cortical laminae. They may have a role in learning through their effects on cortical circuits via their cortical and thalamic inputs. Dendrites of neurogliaform cells (spiderweb cells) , one prominent type of interneuron in laminae II and III, spread radially but are unique because they establish synapses with each other and other interneuron types; this suggests a prominent role in synchronising cortical circuits. Another morphologically diverse group of interneurons releases vasoactive intestinal polypeptide (VIP) in addition to GABA; other interneurons in this group coexpress cholecystokinin (CCK) and other peptide receptors.
Afferents
Afferents to a given region of the cortex can be derived from four sources (primarily from the cortex) and have distinctive patterns of termination:
- 1.
Long and short association fibres from small- and medium-sized pyramidal cells in laminae II and III in other parts of the ipsilateral cortex.
- 2.
Commissural fibres from medium-sized pyramidal cells in laminae II and III project through the corpus callosum from matching or modality-related (homotopic) areas of the opposite hemisphere.
- 3.
Thalamocortical fibres from the appropriate specific or association nucleus, for example fibres from the ventral posterior thalamic nucleus to the somatic sensory cortex and from the dorsomedial thalamic nucleus to the prefrontal cortex (defined below) terminate in lamina IV. Nonspecific thalamocortical fibres from the intralaminar nuclei terminate in all laminae.
- 4.
Cholinergic and aminergic fibres from the basal forebrain, hypothalamus, and brainstem. These fibres are represented in green in Figure 29.1 , and while they innervate the entire cortex, this does not result in a generalised or nonspecific response. Their anatomic specificity (cortical, laminar, and cell) results in activation or inhibition of limited collections of neurons. The relevant nuclei of origin, and the transmitters/modulators involved, are:
- •
nucleus basalis of Meynert (basal forebrain nuclei), acetylcholine;
- •
tuberomammillary nucleus (posterior hypothalamus), histamine;
- •
substantia nigra pars compacta (ventral midbrain tegmentum), dopamine;
- •
raphe nuclei (midbrain and rostral pons), serotonin;
- •
locus ceruleus (rostral pons), norepinephrine.
These five sets of neurons are of particular relevance to psychiatry and are considered in Chapter 34 .
Efferents
All efferents from the cerebral cortex are axons of pyramidal cells, and all are excitatory in nature. Axons of some pyramidal cells contribute to short or long association fibres, others form commissural or projection fibres; association and commissural fibres make up the bulk of white matter of the cerebral hemisphere.
- •
Examples of short association fibre pathways (pass between adjacent cortical areas through the superficial white matter as U fibres ) are those entering the motor cortex from the sensory cortex and vice versa ( Figure 29.1 ). Examples of long association fibre pathways are projections between the prefrontal cortex (cortex anterior to the motor areas) and sensory association areas. Pyramidal cells that primarily reside in laminae II and III are the source of these fibres.
- •
The commissural fibres of the brain are entirely composed of pyramidal cell axons running through the corpus callosum, and posterior and anterior commissures (and in other, minor commissures) to corresponding cortical areas in the opposite hemisphere (e.g. the primary cortical area projecting to its equivalent contralateral association area) and noncorresponding areas (These commissural connections are lacking between the primary visual cortex [area 17] and the primary somatosensory and motor cortex representing the distal part of the upper limb.) Pyramidal cells that primarily reside in laminae II and III are the source of these fibres.
- •
Projection fibres from the primary sensory and motor cortex form the largest input to the basal ganglia ( Chapter 33 ). The thalamus receives projection fibres from all parts of the cortex. Other major projection systems are corticopontine (to the ipsilateral pontine nuclei), corticonuclear (to contralateral motor and somatic sensory cranial nerve nuclei in the pons and medulla), and corticospinal. Pyramidal cells that primarily reside in laminae V and VI (preferentially projecting to specific thalamic relay nuclei) are the source of these fibres.
Cortical areas
The most widely used reference map is that of Brodmann, who divided the cortex into 44 areas (his numbering scheme extended to 52, but not all numbers were used) on the basis of cytoarchitectonic characteristics. Most of these areas are shown in Figure 29.4 , but ‘sharp’ borders between areas do not exist. (These numbers are often used to refer to functional areas, although Brodmann rejected any such correlation.) Coloured in Figure 29.4 are the three principal primary sensory areas (somatic, visual, and auditory) and the single primary motor area. The adjacent cortex to each primary sensory or motor cortical area is the association cortex referred to as the unimodal association area (same modality). The rest of the neocortex consists of multimodal (polymodal) association areas receiving fibres from more than one unimodal association area (e.g. receiving tactile and visual inputs, or visual and auditory) and other multimodal or paralimbic areas.
Investigating functional anatomy
The term connectome was coined to represent a ‘comprehensive map of neural connections whose purpose is to illuminate brain function’. However, the desire to develop a complete functional map of the human brain necessitates collection of empirically derived data on that structural connectivity, but much still remains unknown. Contemporary approaches are providing unique opportunities to achieve this goal through advances in computing capabilities and data storage, neuropsychologic testing, and magnetic resonance imaging (MRI) capabilities that allow imaging of the living human brain.
These new advances in understanding the brain are accompanied by a shift from looking at individual areas of the cortex to considering all areas and their interactions at once. New methodical or theoretical frameworks have been deployed to describe and predict these complex system dynamics through the use of network analysis and a mathematical approach based on graph theory . Network models use collections of ‘elementary’ cortical units and their connections to demonstrate how functions can emerge dynamically or ‘capture the brain in action’. These models remain limited by the known connections between areas, but some connections are inferred to exist based on primate studies. However, models may also predict functional relationships unsupported by a known structural basis or pathways predicted to exist based on a performed behaviour. While imaging of living brain pathways and connectivity will advance through the use of these neuroradiologic techniques and mathematical modelling, the development of new and continued use of ‘old’ techniques of neuroanatomy are required to provide the structural evidence for these emerging pathways and systems of cortical activation.
Two dominant methods are in use for ‘identification and localisation’ of functions in the human brain. Both techniques depend upon the local increases in blood flow that meet the additional oxygen demand imposed by localised neural activity.
Positron emission tomography
Positron emission tomography (PET) measures oxygen consumption following injection of water labelled with oxygen-15 ( 15 O) into a forearm vein. 15 O is a positron-emitting isotope of oxygen; the positrons react with nearby electrons in the blood to create γ rays, which are counted by γ-ray detectors. Alternatively, fluorine-18-labelled deoxyglucose ( 18 F-deoxyglucose) may be used to measure glucose consumption. 18 F-deoxyglucose is taken up by neurons as readily as glucose.
Image subtraction and image averaging are required for meaningful interpretation of PET studies, as explained in the caption to Figure 29.5 , and a similar signal extraction process is deployed in functional MRI (fMRI).