To a casual observer, a microscopic pond organism such as Paramecium behaves with apparent intent and coordination that suggests the involvement of a sophisticated nervous system. If it bumps into an object, it stops, swims backward, and then moves off in a new direction (Figure 5.1). This simple behavior minimally requires a sensory system to detect touch, an integrator to process information from the sensor, and a motor pathway to effect a response. Paramecium does not possess a nervous system, however. It is a single cell.
The human brain contains over a trillion neurons. It has evolved sophisticated structures and networks that allow for self-awareness, creativity, and memory. Yet the human nervous system’s basic organizational principles share many similarities with our unicellular cousins. Unicells and neurons both use changes in membrane potential (Vm) to integrate and respond to divergent and, sometimes, conflicting inputs. On an organismal level, humans, like unicells, have sensory systems to inform them about their immediate environment, integrators to process sensory data, and motor systems to effect an appropriate response.
Figure 5.1 Sensory response in Paramecium.
Modified from Jennings, H.S. Behavior of the Lower Organisms. The Columbia University Press, 1906.
II. Nervous System
In discussing how the nervous system works, it is useful to define three partially overlapping subdivisions.
- The central nervous system (CNS) includes the neurons of the brain and spinal cord. The CNS is the nervous system’s integrative and decision-making arm.
- The peripheral nervous system (PNS) collects sensory information and conveys it to the CNS for processing. It then directs motor commands from the CNS to the appropriate targets. The PNS includes neurons that originate in the cranium and spinal cord and extend beyond the CNS.
- The autonomic nervous system (ANS) is central to many discussions of human physiology because it regulates and coordinates visceral organ function, including the gastrointestinal system, lungs, heart, and vasculature. The distinction between the ANS and the other two divisions is functional rather than anatomic. The ANS can be further subdivided into the sympathetic nervous system and parasympathetic nervous system. Both divisions function largely independently of voluntary control.
The nervous system comprises a network of neurons. Although their shape may vary according to function and location within the body, the basic principles of neuronal design and operation are universal. Their role is to transmit information as rapidly as possible from one area of the body to the next. In the brain, the distance involved may be a few micrometers, but, in the periphery, it can exceed a meter. Because speed is achieved using electrical signals, a neuron can be thought of as a biologic wire. Unlike a wire, however, a neuron has the ability to integrate incoming signals before transmitting information to a recipient.
A neuron can be divided into four anatomically distinct regions: the cell body, dendrites, an axon, and one or more nerve terminals (Figure 5.2).
Figure 5.2 Neuronal anatomy.
1. Cell body
The cell body (soma) houses the nucleus and components required for protein synthesis and other normal cellular housekeeping functions.
Dendrites are branched projections of the cell body that radiate in multiple directions (“dendrite” is derived from dendros, the Greek word for tree). Some neurons have dense and elaborate dendritic trees, whereas others may be very simple. Dendrites are cellular antennae waiting to receive information from the neural net. Many tens of thousands of nerve terminals may synapse with a single neuron via its dendrites.
An axon is designed to relay information at high speed from one end of the neuron to the other. It arises from a swelling of the soma called an axon hillock. An axon is long and thin like a wire. It is often wrapped with an insulating material (myelin) that enhances signal-transmission rate (see below). Myelination begins some distance distal to the axon hillock, leaving a short initial segment that is unmyelinated.
The axoplasm (axonal cytoplasm) is filled with parallel arrays of microtubules and microfilaments. They are partly structural, but they also act like railway tracks in a mineshaft. “Ore carts” (vesicles) filled with neurotransmitters and other materials attach to the tracks and then motor along at relatively high speed (~2 ?m/s) from one end of the cell to the other. Movement away from the cell body toward the nerve terminal (anterograde transport) is powered by kinesin. The return trip (retrograde transport) relies on a different molecular motor (dynein).
4. Nerve terminal
The nerve terminal is specialized to convert an electrical signal (an action potential) into a chemical signal for dispatch to one or more recipients. The junction between the terminal and its target is called a synapse. The presynaptic and postsynaptic cell membranes are separated by a ~30–50 nm synaptic cleft. Facing the terminal across the cleft may be any of a number of different postsynaptic effector cells, including myocytes, secretory cells, or even a dendrite extending from the cell body of another neuron.
Clinical Application 5.1: Polio
Retrograde transport is believed to be the mechanism by which polio and many other viruses enter the central nervous system from the periphery.1 Poliovirus is an enterovirus distributed by fecal–oral contact that causes paralytic poliomyelitis. After infecting the host, the virus enters and spreads through the nervous system via nerve terminals. After fusing with the surface membrane and entering the axoplasm, the viral capsid (a protein shell) attaches to the retrograde transport machinery and motors to the cell body.
Here, it proliferates and, ultimately, destroys the neuron. The result is a flaccid paralysis of the musculature, classically affecting the lower limbs, but it can also cause fatal paralysis of the respiratory musculature. Polio has been largely eradicated in North America and Europe but is endemic in many other regions of the world. Defects in axonal transport are also believed to have a role in precipitating the neuronal death that accompanies Alzheimer disease, Huntington disease, Parkinson disease, and several other adult-onset neurodegenerative diseases.
From the Centers for Disease Control and Prevention, Public Health Image Library (photo credit: Dr. Fred Murphy, Sylvia Whitfield, 1975).
The speed with which neural nets process and output data is limited by the rate at which signals are transmitted from one component to the next. Extracting maximal speed from a neuron is achieved by using a fast action potential, by optimizing axonal geometry, and by insulating the axons.
1. Action potentials
Axons that convey signals over long distances typically display action potentials that have a very simple form and function as binary digits on the neural information net. Neuronal action potentials are mediated primarily by voltage-gated Na+ channels, which are very fast activating (Figure 5.3). When Na+ channels open, Na+ flows into the neuron down its electrochemical gradient, and Vm depolarizes rapidly toward the equilibrium potential for Na+ (see 2?II?B).
It is the rapidity of Na+-channel opening (“gating kinetics”) that allows electrical signals to propagate at high speeds down an axon’s length. Membrane repolarization occurs largely as a result of Na+-channel inactivation. Voltage-dependent K+ channels activate during a spike also, but their numbers are small and, thus, their contribution to membrane repolarization is limited.
Figure 5.3 Time course of ion channel events during a neuronal action potential.
2. Axon diameter
The rate at which electrical signals travel down an axon increases with axonal diameter (Figure 5.4). This is because internal resistance, which is inversely proportional to diameter, determines how far passive current can reach down the axon’s length before the signal decays and needs amplifying by an active current (i.e., a Na+-channel–mediated current). The amplification step is slow compared with transmission of passive current, so wide axons transmit information over long distances much faster than do thin ones.
Figure 5.4 Myelin and diameter effects on axonal conduction velocity.
The passive currents that flow during excitation dissipate with distance because the membrane contains K+ “leak” channels that lose current to the extracellular medium (see Figure 2.11). Leak channels are always open. Conduction velocity is improved significantly by insulating the axon with myelin to prevent such leak (see Figure 5.4B). Myelin is formed by glial cells and comprises concentric layers of sphingomyelin-rich membrane (see Section V below). Insulation increases conduction velocity up to 250-fold.
4. Saltatory conduction
An axon’s myelin sheath is not continuous. Every 1–2 mm is a 2–3-?m segment of exposed axonal membrane known as a node of Ranvier. Nodes are tightly packed with Na+ channels, whereas the internodal regions (the areas lying hidden beneath the myelin sheath) have virtually no channels. In practice, this means that an action potential leapfrogs from one node to the next down the length of the axon, a behavior known as nodal, or saltatory, conduction (see Figure 5.4C).
CNS neurons are a diverse group of cells, and there are many ways of classifying them. Morphologically, they can be grouped on the basis of the number of neurites (processes, such as axons and dendrites) extending from the cell body.
Pseudounipolar neurons are usually sensory. The cell body gives rise to a single process (the axon) that then splits into two branches. One branch returns sensory information from the periphery (the peripheral branch), whereas the other branch projects and conveys this information to the CNS (central branch).
Bipolar neurons are usually specialized sensory neurons. Bipolar neurons can be found in the retina (see 8?VII?A) and olfactory epithelium (see 10?III?B), for example. Their cell body gives rise to two processes. One conveys sensory information from the periphery, and the other (the axon) travels to the CNS.
Multipolar neurons have a cell body that gives rise to a single axon and numerous dendritic branches. Most CNS neurons are multipolar. They can be further subcharacterized based on the size and complexity of their dendritic tree.
D. Neurons as integrators
The unicell mentioned in the introduction is capable of integrating multiple sensory signals (e.g., mechanical, chemical, thermal) through changes in Vm. For example, a noxious signal that depolarizes Vm and increases the tendency to turn might be ignored if an attractant signal indicating nearby food hyperpolarizes the membrane and negates or overrides noxious signal input. A paramecium is not capable of conscious thought, yet it makes a decision that affects behavior based on the summed effect of multiple stimuli on Vm. The dendritic trees of higher cortical neurons receive tens of thousands of competing inputs. The likelihood that neuronal output (spiking) will be modified on the basis of these signals is similarly determined by their net effect on Vm.
1. Incoming signals
Neurons hand off information to each other via dendrites. When a presynaptic neuron fires, it releases transmitter into the synaptic cleft. If the neuron is excitatory, transmitter binding to the postsynaptic dendritic membrane causes a transient depolarization known as an excitatory postsynaptic potential (EPSP) as shown in Figure 5.5A. Inhibitory neurons release transmitters that cause transient hyperpolarizations known as inhibitory postsynaptic potentials (IPSPs). EPSP and IPSP amplitudes are graded with incoming signal(s) strength.
Figure 5.5 Summation.
EPSP = excitatory postsynaptic potential; IPSP = inhibitory postsynaptic potential; Vm = membrane potential.
Much of the information being received by neurons at their dendrites represents sensory “noise.” Isolating the strongest and most relevant signals is accomplished using a noise filter that takes advantage of a dendrite’s natural electrical properties. Postsynaptic potentials (PSPs) are passive responses that degrade rapidly as they travel toward the cell body (see Figure 2.11). Degradation is enhanced by a dendrite’s inherent electrical leakiness and its lack of myelin. In practice, this means that a small PSP may never reach the cell body. PSPs generated by strong presynaptic activity activate voltage-gated ion currents along the length of the dendrite (see Figure 2.12). These enhance the signals and, thereby, increase their likelihood of reaching the cell body.
Signal integration also begins at the dendritic level. PSPs may meet and combine with PSPs arriving from other synapses as they travel toward the soma. This phenomenon is known as summation and is reminiscent of the way in which waves (e.g., sound waves and ripples spreading across a pond surface) interfere constructively and destructively. There are two types of summation: spatial and temporal.
a. Spatial summation
If EPSPs from two different dendrites collide, they combine to create a larger EPSP (see Figure 5.5B). This is known as spatial summation and applies to IPSPs also. EPSPs and IPSPs can also summate to yield an attenuated membrane response (see Figure 5.5C).
b. Temporal summation
Two EPSPs (or IPSPs) traveling along a dendrite in rapid succession can also combine to produce a single, larger event. This is known as temporal summation (see Figure 5.5D).
The net effect of multiple PSPs on Vm determines the likelihood and intensity of neuronal output. If a depolarization is sufficiently strong, it may elicit a train of spikes. Spikes arise from the initial segment (also known as the spike initiation zone) and travel down the length of the axon toward the presynaptic terminal.
Action potentials are all-or-nothing events, so neurons must pass on information about signal strength using digital encoding. Weak stimuli may yield one or two spikes. Strong stimuli elicit spike trains (volleys) that travel in rapid succession down the axon’s length. There is tremendous variability in the size, shape, and frequency of spikes generated by different neurons. As a general rule, the number of spikes in a volley reflects incoming stimulus strength (Figure 5.6).
Figure 5.6 Digital encoding by neurons.
Vm = membrane potential.