23 Jun Varieties of Nerve Cells
Did you know that not all brain cells are created equally? I think they are more different from one another than snowflakes and some of the differences are astounding. The heterogeneity of neurons indicates that modeling brain functions may require a heterogeneous approach (as opposed to many neural networks that are homogeneous). Besides the differences in branching patterns shown in the illustration, there are differences in sizes, in number of branches, in the lengths of the branches, in the isolation within a brain area or extension across brain areas, and many other structural differences.
The differences in morphology and arborization of neurons suggests adaptation to different purposes or different roles in the chain of cognition. The layering of neurons in each specific area of the cortex further suggests differences from cell to cell. The location on the receiving neuron of the incoming signal creates variation in the characteristics of the signal (Sejnowski 1996). What can we safely infer about these differences?
In this post, we will examine what some of the differences are, and we’ll deal with the modeling assumptions in later posts.
|Understanding Context Cross-Reference|
|Click on these Links to other posts and glossary/bibliography references|
|Prior Post||Next Post|
|The Coming Revolution||A World of Chaos|
|brain cells||Kuffler 1884 Mountcastle 1978|
|cerebrum||Corsi 1991 Sejnowski 1996|
|neural networks||Kandel 1985|
In the cerebrum, there are six layers (the allocortex and heterogenetic cortex have only three layers). There are five major and three minor cell types. Cerebral cell types are pyramidal, stellate (also called granular or Golgi type II), fusiform (polymorphic multi-form), horizontal (of Cajal), and cells of Martinotti (ascending axon cells). Neurogliaform and basket type cells are also present in the cerebrum. The illustration shows a example of the complexity of brain circuitry and the distribution of cells and their fibers in the cerebrum. This distribution is fairly regular throughout the specialized areas of the cerebrum although some areas have slightly different characteristics due to their functions and different varieties of nerve cells.
Pyramidal and Stellate Cells
Pyramidal cells are the most numerous type of cells in the cortex. The name “pyramidal” arises from the distinct apical (forming an apex) profile of these cells. Most pyramidal cells have profuse apical dendrite branching in layers adjacent to the cell body. The branching reaches toward the cortical surface. Non-apical dendrites of pyramidal cells extend horizontally in the same stratum as the cell body. Pyramidal cell axons normally extend into sub-cortical white matter. The presence of efferent fibers indicates an output function.
Stellate (also called small granule) cells are the second most numerous type in the cortex. They are star-shaped or polygonal, ranging in size from 4 to 8 micrometers. Their omnidirectional radial processes (branches) are short; the axon termini are generally near the cell body. Their function is correlative. This correlative function is critical to advanced cognitive functions because we need to find what relationships exist between the huge amounts of world knowledge distributed throughout the entire brain. Without correlative cells, our ability to process information would be reduced and maybe profoundly limited.
The presence of the stellate cells and pyramidal cells, which are not present in lower biological forms, indicates the advanced functions of the human brain.
Horizontal cells are spindle shaped. They have large nuclei but little cytoplasm. They are only found in Layer I. Unlike pyramidal and fusiform cells, their processes (branches) extend horizontally, as the name suggests, so they are laterally parallel to the surface of the cortex. Their processes are also polar to the cell body and remain in the top layer of the gray matter. Their function is correlative. In other words, they directly process no input or output – they interact only with other local cells. Implications of this functionality are considered in Section 4, 5 and 6 of this blog.
Fusiform cells, as the name suggests, have varied shapes and fiber distributions. Although they are most frequently spindle shaped, they may also be pyramidal or ovoid. They are only in the deepest layer of the gray cortex (Layer IV), and their arborization is more polar than other types of neurons: dendrites extend only from the top toward the cortical surface and from the bottom to form connections in Layer VI. The center of the soma spindle is mostly free of processes (branching). The axon extends between dendrite processes into white matter.
Cells of Martinotti
Cells of Martinotti send their axons up in the opposite direction of pyramidal and fusiform cells and perpendicular to horizontal and
granule cells. Martinotti cells are small. Although they are usually polygonal in shape, they may also be round or triangular. They are found in the lower five layers of the cerebral cortex.
They have short dendritic processes and longer longitudinal axons that extend lateral subfibers throughout the tangential layer to form synapses with cells and dendrites there. As the afferent cells of other areas are the only other source of axons reaching toward the surface of the cortex, it is reasonable to assume a special supporting function for cells of Martinotti: they assist afferent fibers in propagating action potential toward intended targets in the more superficial layers of the cortex.
Is this function important to learning? As new associations between previously learned entities are formed, Martinotti cells may act as the bridge that allows cells or bundles of cells that store information to be distributed longitudinally in the gray matter.
Are these differences between cell structure, location and function important to cybernetic modeling of brain functions? I think so, thus I took time to talk about it. At the same time, as we’ll explore in Section 3 on the human neural network, the electrical transmissions and transmission patterns are remarkably consistent throughout the brain. We’ll explore more on the subject of specialization in future posts.
|Click below to look in each Understanding Context section|
|4||Perception and Cognition||5||Fuzzy Logic||6||Language and Dialog||7||Cybernetic Models|
|8||Apps and Processes||9||The End of Code||Glossary||Bibliography|