Copyright © 1999 Massachusetts Institute of Technology
The neuron is the main type of cell in the nervous system that, in association with neuroglial cells, mediates the information processing that underlies nervous function. As the main building block of the brain, the nerve cell is fundamental to the neural basis of cognitive abilities. Much of the research in contemporary neuroscience focuses on neuronal structure, function, and pharmacology, using modern techniques of molecular and cell biology. Based on these results, computational models of neurons and neuronal circuits are being constructed to provide increasingly powerful insights into how the brain mediates cognitive functions.
In their speculations on the mind, the ancients knew nothing of cells or neurons, nor did DESCARTES, Locke, or KANT, or any other scientist or natural philosopher until the nineteenth century. The first step toward this understanding was the cell theory of Schwann, in 1839, which stated that all body organs and tissues are composed of individual cells. A nerve cell was recognized to consist of three main parts: the cell body (soma), containing the nucleus; short processes (dendrites); and a single long process (axon) that connects to the dendrites or somata of other nerve cells within the region or of cells in other regions. However, the branches of dendrites and axons could not be clearly visualized, leading to the belief among some workers that nerve cells are different from other cells in that their finest branches form a continuous functional network (called the reticular theory). It was also recognized that numerous non-neuronal cells, called neuroglia, surround the neurons and contribute to their functions.
Led by Ramón y CAJAL, neuroanatomists in the 1880s and 1890s, using the GOLGI stain, showed that most regions of the brain contain several distinctive types of nerve cell with specific axonal and dendritic branching patterns. Dendrites and axons were found not to be continuous, so that the nerve cell belongs under the cell theory, as summarized in the neuron doctrine. How then are signals transferred between neurons? Sherrington in 1897 suggested that this occurs by means of specialized junctions which he termed synapses. Electron microscopists in the 1950s showed that such contacts between neurons exist. Since that time, neuroanatomists have elucidated the ultrastructure of the synapse as well as the patterns of synaptic connections between the different types of neurons. The patterns of connections within a region are called canonical circuits, mediating the main types of information processing within each region. Neurons and their canonical circuits are organized at the next higher level into neural modules, such as the columns found in the CEREBRAL CORTEX (see also VISUAL CORTEX). At a still higher organization level, the patterns of neuronal connections between regions are called distributed circuits, which constitute the pathways and systems underlying behavior (e.g., see VISUAL ANATOMY AND PHYSIOLOGY).
Physiological studies of neuron properties have paralleled these anatomical developments. The axon generates a nerve impulse (action potential), a wave of depolarization of the surface membrane, which propagates rapidly along the membrane from its site of initiation (the axon hillock) to the axon terminals. Already in 1850 a finite rate of propagation (approximately 100 m per second in the largest axons) was established; this overturned the historical assumption of a mysterious instantaneous "nervous force" underlying the mind.
In the 1950s a slower potential (synaptic potential) was discovered by Katz at the synapse. It was found that an action potential invading a synapse causes it to secrete small vesicles containing a chemical neurotransmitter, which diffuses across the cleft from the presynaptic to the postsynaptic cell. There it acts on a membrane receptor to bring about an opening of membrane channels; this lets electrically charged ions flow across the membrane to change the membrane potential of the postsynaptic site. Membrane proteins that contain both receptor sites and ionic channels are called ionotropic receptors. Depending on which ions flow, the membrane may be depolarized (excitatory postsynaptic potential, EPSP) or hyperpolarized (inhibitory postsynaptic potential, IPSP). Among the cells of the body, chemical synapses are unique to neurons. Thus the study of synapses lies at the heart of the study of brain function at the neuronal level. In addition to these chemical synapses, neurons, like other cells, may be interconnected by gap junctions (electrical synapses), which permit electric currents and small molecules to pass directly between cells. These also connect neuroglial cells, and are especially prevalent during development.
Knowledge of NEUROTRANSMITTERS and how they generate EPSPs and IPSPs began with the study of the autonomic nervous system around 1900. The first neurotransmitter to be identified was acetylcholine, shown by Loewi in the 1920s to mediate the slow action of the vagus nerve in slowing the heart rate, and by Katz in the 1950s to mediate the rapid action of motor nerves in exciting skeletal muscles. Cannon in the 1920s established epinephrine (Adrenaline) as the neurohormone mediating "flight-or-fight" responses, and other biogenic amines acting on autonomic organs were revealed in the 1930s. The introduction of the neuroleptics chlorpromazine and reserpine for the treatment of schizophrenia in the 1950s shifted interest in the pharmacology of neurotransmitters to the central nervous system. The resultant growth of the field of psychopharmacology has generated mechanistic hypotheses for all the major mental disorders. For example, a common action of neuroleptic (antipsychotic) drugs is a blockade of D2 receptors at dopaminergic synapses. And current research on antidepressant drugs is focused on their ability to act as selective serotonin reuptake inhibitors (SSRIs).
Many of the neurotransmitters can activate not only ionotropic receptors but also metabotropic receptors coupled to second messenger systems, which then phosphorylate target proteins or bring about other slower metabolic changes in the postsynaptic neuron (and also can act back on the presynaptic terminal as well). Acetylcholine acting on the heart and epinephrine acting as a hormone are examples, as are a wide range of neuropeptides. These include such molecules as hypothalamic factors (e.g., luteinizing hormone - releasing hormone, corticotropin), opioids (e.g., enkephalins and endorphins), and numerous types found also in the gut and other organs (vasoactive intestinal polypeptide, cholecystokinin, substance P, etc.; see NEUROENDOCRINOLOGY). These molecules, acting as neuromodulators, set behavioral states (e.g., the role of acetylcholine, norepinephrine, and serotonin in waking, sleeping, and levels of consciousness, or the role of neuropeptide Y in feeding behavior and anxiety and stress responses). Another role is in learning and memory; LONG-TERM POTENTIATION (LTP) and long-term depression (LTD) of synaptic responses involve second messengers, calcium, and cyclic nucleotides, and actions on the genome that may implement associative (Hebbian) learning. These latter actions overlap with the activation of receptors controlling growth and differentiation during development.
In summary, a better understanding of the effects of neurotransmitters and neuromodulators on different types of neurons and neuronal circuits is the necessary foundation for an understanding of normal cognition and changes underlying psychotic states.
Analysis of neuronal function has been aided enormously by the development of modern techniques. With the advent of DNA engineering in the 1970s, the molecular basis of neuronal structure and function is being increasingly elucidated. Modern physiological research employs a variety of methods in analyzing neuronal function. These include single- and multiple-neuron recordings in awake behaving animals (SINGLE-NEURON RECORDING); patch pipette recordings from neurons in slices taken from different brain regions (this includes slices of human cerebral cortex in tissue obtained in operations for relief of chronic epilepsy); patch recordings of single membrane channels in isolated cells grown in tissue culture; and different types of functional imaging (including movement of calcium ions, glucose uptake, and voltage-sensitive dyes). A long-term goal is to relate these changes at the neuronal level to changes in blood flow revealed by brain imaging methods (POSITRON EMISSION TOMOGRAPHY, functional MAGNETIC RESONANCE IMAGING, f MRI) at the systems level. This will enable an integrated view of neuronal function and the neural basis of cognition and cognitive disorders to begin to emerge. Databases to support this effort are becoming available on the World Wide Web. Examples are membrane receptors and channels (www.le.ac.uk/csn), canonical neurons and their compartmental models (http://senselab.med.yale.edu/neuron), and brain scans (human brain project).
Experimental approaches to the study of the neuron are being greatly aided by the development of computer models and the emergence of the new field of COMPUTATIONAL NEUROSCIENCE. This began in the 1950s with the pioneering model of the axonal action potential by Hodgkin and Huxley. In the 1960s Rall showed that complex dendritic trees could be modeled as chains of compartments incorporating properties representing action potentials and synaptic potentials. With the rise of powerful modern desktop computers it is now possible for neuroscientists to construct increasingly accurate models of different types of neurons to aid them in analyzing the neural basis of function of a given neuron in a given region. This work is reported in mainstream neuroscience journals as well as in new journals such as Neural Computation and the Journal of Computational Neuroscience. In contrast to this approach, connectionist networks reduce the soma and dendrite of a neuron to a single node, thereby excluding most of the interesting properties of neurons as summarized above. A merging of neuron-based compartmental models with NEURAL NETWORKS will therefore be welcome, because it will provide insights into how real brains actually carry out their system functions in mediating cognitive behavior. It is not unreasonable to expect that this will also provide the philosophical foundation for a more profound functional explanation of the relation between the brain and the mind.
Shepherd, G. M. (1991). Foundations of the Neuron Doctrine. New York: Oxford University Press.
Snyder, S. H. (1996). Drugs and the Brain. New York: Scientific American Books.
Copyright © 1999 Massachusetts Institute of Technology