The basic design of neocortex is similar in all mammals. Neocortex is a thick (compared to dorsal cortex) sheet of tissue of varying size but usually 1-3 mm in depth. The thickness of cortex results from a proliferation of neurons, such that the cell bodies of over 100 would be encountered by a pin stuck through its surface to the underlying axon fibers that shuttle information to and from the cortex. This contrasts to the one or two neurons that would be encountered by a pin stuck through the dorsal cortex of reptiles. Whereas the thin row of neurons in dorsal cortex both receives input from other parts of the brain and provides output, the thicker neocortex provides a more complex and fundamentally different type of processing. First, the neurons across the thickness of cortex are of different morphological types and they have different connections and functions. Second, neurons of similar appearance and connections are grouped according to depth in the cortex so that groups of cells form a stack of layers. Early investigators numbered and named the layers in various ways, but a scheme of six basic layers, stemming from the publications of Korbinian Brodmann around 100 years ago, has become a nearly universal standard (Fig. 1). Some of the layers of other early investigators are now considered to be sublayers within the standard scheme.
Layer 4, a middle layer of neurons, is the main receiving layer of neocortex (Fig. 2). Neurons in this layer receive most of the activating input from other structures, including nuclei of the dorsal thalamus and other subdivisions of the cortex. The connections of neurons in layer 4 are predominantly local within layer 4 and with neurons immediately above or below in adjoining layers. There, the information that is received in layer 4 rapidly spreads in a narrow focus across the layers. These and similar vertical connections of neurons in other layers form vertical columns of coactive neurons in neocortex. This is a highly characteristic feature of cortical organization.
Human Area 3b
Figure 1 The laminar organization of the neocortex. Part of the neocortex was cut into thin slices across its thickness, and the sections were stained to reveal cell bodies (dark ovals) but not axons and dendrites. The cell bodies vary in size, shape, and packing density across the thickness of neocortex in a laminar pattern. It is now usual to divide the neocortex into six main layers (I-VI; for historical reasons, Roman numerals are usually used to number layers) from the surface to the underlying connecting axons or white matter (WM). Sublayers are also commonly distinguished. Layers play different functional roles in the processing that goes on in the neocortex (see text). Different areas of the neocortex are specialized for different functional tasks, and such specialization may be reflected in variations in the appearance and relative thickness of the layers. The section of cortex shown here is from the primary somatosensory cortex of a human, commonly referred to as area 3b after a numbering scheme devised for cortical areas by Korbinian Brodmann over 100 years ago. This cortex is just over 2 mm thick. Area 3b and other sensory areas of neocortex are characterized by a relatively thick layer IV that is densely packed with neurons having small cell bodies. Layer III is especially thick in the large brains of humans.
For activation, layer 4 neurons depend on focused synapses from only a few axons from another structure, and they spread this information to only a few nearby neurons that are immediately deeper or more superficial in the cortex. Thus, layer 4 neurons are rather small, with radial dendrites spreading around the cell body and a restricted and local axon arbor contacting adjacent neurons in layer 4 and neurons
b c inputs
intrinsic neurons d e f outputs
Figure 2 Major input axons, output neurons, and intrinsic neurons of the neocortex. Input to cortex includes (a) the major activating input from the thalamus or other areas of cortex that terminates largely in layer IV; (b) the feedback activating connections from other areas of cortex that terminate more diffusely above and below layer IV; and (c) the diffusely distributed modulating input that terminates above layer IV. Intrinsic neurons include excitatory (+) stellate or granular cells in layer IV with local dendrites and a local axon arbor and an inhibitory (—), nonspiny class of stellate cells or local circuit neurons of various types in all layers. Output neurons include (d) layer III pyramidal cells that project to other areas of the cortex and to subcortical targets; (e) larger layer V pyramidal cells that project mainly to subcortical targets including the spinal cord; and (f) pyramidal-like neurons of several types in layer VI that provide feedback connections to the b c inputs
intrinsic neurons d e f outputs
Figure 2 Major input axons, output neurons, and intrinsic neurons of the neocortex. Input to cortex includes (a) the major activating input from the thalamus or other areas of cortex that terminates largely in layer IV; (b) the feedback activating connections from other areas of cortex that terminate more diffusely above and below layer IV; and (c) the diffusely distributed modulating input that terminates above layer IV. Intrinsic neurons include excitatory (+) stellate or granular cells in layer IV with local dendrites and a local axon arbor and an inhibitory (—), nonspiny class of stellate cells or local circuit neurons of various types in all layers. Output neurons include (d) layer III pyramidal cells that project to other areas of the cortex and to subcortical targets; (e) larger layer V pyramidal cells that project mainly to subcortical targets including the spinal cord; and (f) pyramidal-like neurons of several types in layer VI that provide feedback connections to the above and below layer 4. The small neurons in layer 4 are typically called stellate cells because their short dendrites protrude in all directions like the rays of a star. Because layer 4 cells are often so small as to appear to be no more than grains of sand, they are also called granule cells. Layer 4 is sometimes called a granular layer.
Layer 5 neurons, just under layer 4, are activated by layer 4 and other neurons in the superficial layers. The major neuron type is the large pyramidal neuron (Fig. 2). The cell body is triangular in shape with the apex pointing outward as it gives rise to a long apical dendrite that ascends toward or to the surface of layer 1. The base of the triangular cell body also gives rise to basilar dendrites that are much shorter and spread out horizontally away from the vertical cell column and ascending apical dendrite. The dendrites are covered with synaptic contacts from other neurons. The arrangement of the dendrites indicates that layer 5 pyramidal cells are designed to gather information somewhat horizontally in layer 5 via the basilar dendrites, and more locally from the more superficial layers via the long ascending apical dendrite, which has short radiating branches along its ascending stem. Thus, layer 5 pyramidal neurons have access to input from a vertical array of neurons, including those of a vertical row along the apical dendrite and those in adjacent rows contacted by the basilar dendrites or the branches of the apical dendrite. The layer 5 neuron sums all this information, and it provides the major output of the neocortex. In addition to forming a local axonal arbor, most layer 5 pyramidal neurons project over quite long axons to nuclei in the dorsal thalamus, structures in the subcortical basal ganglia, the mid-brain, and other brain stem structures, and even the spinal cord. Thus, the output of neocortex that most directly affects behavior and performance stems from layer 5 pyramidal neurons. Some layer 5 neurons also send information to more distant locations in the neocortex, either in the same cerebral hemisphere or in the opposite cerebral hemisphere via the corpus callosum. Thus, layer 5 neurons help to inform other areas of neocortex about ongoing computations. In brief, a major function of layer 5 neurons is to send information to other parts of the brain. Because these targets can be relatively far from the neocortex and distance is time for the nervous system, layer 5 neurons have long, thick axons. Because thicker axons have faster conduction rates, the transfer of information is speeded up. Neurons with long, thick axons and long, branching dendrites need large cell bodies to maintain these structures, and layer 5 pyramidal cells are large. They also vary in size according to axon length and thickness, so that the large Betz cells of the motor cortex project all the way to the lower spinal cord, whereas the large Meynert cells of the visual cortex project to the opposite hemisphere, to a distant visual area in the neocortex, and to the brain stem.
The other main output layer of the neocortex is layer 3, and it also contains pyramidal cells with basilar and apical dendrites (Fig. 2). However, layer 3 pyramidal cells generally are not as large as layer 5 pyramidal cells, because their apical dendrites are shorter and, more importantly, their axons are shorter, traveling to other areas of cortex rather than to more distant subcortical structures. Thus, neurons in any part of the neocortex send information to other parts of the neocortex, and this is done largely by layer 3 pyramidal neurons. In a manner comparable to layer 5 pyramidal neurons, layer 3 pyramidal neurons receive local input along the vertical apical dendrite extending and branching throughout the depth of layer 3 and the basilar dendrites extending horizontally, sometimes to adjacent cell columns. In general, deeper layer 3 pyramidal cells are larger than more superficial layer 3 pyramidal cells. Layer 3 often has obvious sublayers. Layer 3 neurons send information to layer 4 neurons of other regions of the cortex or directly to other layer 3 neurons. The great computational power of the neocortex stems from these layer 3 pyramidal cells because they transmit information from one region of neocortex to another. Each local group or column of neurons performs a similar and somewhat simple operation or computation, but it is the consequence of a series of simple computations that gives the cortex its great computational power, a power beyond that of the input-output operations of the dorsal cortex of reptiles.
Other layers add to the functional complexity of neocortex. Feedback about computational outcomes is important in information processing systems, and neocortex is designed to provide feedback both to the subcortical thalamic nuclei that provide all sensory information to the neocortex and to other cortical areas that provide information for further processing. Layer 6 is the main feedback layer (Fig. 2). Neurons in layer 6 typically provide feedback projections to the same structure that provides the major activating input to a cortical area or region. If, for example, the major activating input for a subdivision of visual cortex is from a nucleus in the visual thalamus, then layer 6 neurons project back to that same nucleus. If layer 3 neurons in one area of cortex provide the major activating input to layer 4 of another area of cortex, then layer 6 neurons from the target area typically project back to the activating region. Other neurons in other layers may also participate in providing some feedback. The layer 6 neurons receive rather direct information from upper layers, especially layer 4, and they even receive a small amount of direct input from branches or collaterals of axons terminating in layer 4. Thus, input from the visual thalamus to the visual cortex terminates largely in layer 4, but it also provides a few direct branches to layer 6. Layer 6 neurons in turn project back to the sending neurons in the visual thalamus to inform these neurons about the state of the cortex and to influence the next burst of information being sent to the cortex. Layer 6 neurons, by concen trating on responding to local layer 6 connections and input from more superficial neurons, have basilar dendrites and a variable but often long apical dendrite. Thus, they are considered to be pyramidal cells or modified pyramidal cells, although layer 6 is sometimes referred to as the internal granular layer, with layer 4 being the external granular layer. However, layer 6 neurons differ from layer 4 neurons by projecting to distant structures, and thus layer 6 neurons generally are larger. Besides having projections to thalamic nuclei and other cortical areas, layer 6 neurons project to a subcortical structure, the claustrum, which in turn projects back to layer 6. This reciprocal pattern of connections functions to modify the response properties of layer 6 neurons.
The other two layers of neocortex are layers 1 and 2. Layer 2 is a thin layer of densely packed small neurons with local connections and a role in modifying local processing via contacts on apical dendrites of deeper pyramidal cells. Layer 1 is a fiber layer with few neurons that largely consists of the ends of apical dendrites and axons that course along the brain surface and contact these dendrites. Thus, layer 1 is reminiscent of the outer fiber layer of the dorsal cortex of reptiles where input contacts dendrites. Some of the input to layer 1 is modulating input from brain stem neurons. Other input is feedback connections from other cortical areas.
Was this article helpful?
Whenever a doctor informs the parents that their child is suffering with Autism, the first & foremost question that is thrown over him is - How did it happen? How did my child get this disease? Well, there is no definite answer to what are the exact causes of Autism.