Neurophysiology Primer: Part 1

Structure of a neuron

Structure of a nerve cell

A nerve cell consists of three main parts. The largest part is the main body of the cell, called the soma. This contains the nucleus and the structures that keep the cell alive.

From the soma come many branching fibres called dendrites. The dendrites are lined with specialised juctions, called synapses, through which a neuron recieves information from other neurons. Some dendrites also contain dendritic spines - small growths that seem to play a part in learning and memory.

The third main part is the axon. This is a single fibre thicker and longer than the dendrites. A mature cell will either have one axon or none at all, but may have many dendrites. An axon, however will often have many branchings at the end furthest from the soma. In this way, a neuron can carry information to many cells and can also have many neurons connected to itself. The axon ends in a small branching structure that attaches to other nerve cells.

The action potential

At the simplest level of operation, it can be said that a cell recieves its 'inputs' from the dendrites and sends its 'output' down the axon. This output is in the form of an electro-chemical impulse. In other words, the exchange of charged chemical particles (ions) are used to send the messages.

The cell sends an electrical impulse along the axon by exchanging ions through the cell membrane. In general, the neuron is slightly negatively charged with respect to the outside of the cell. This is known as the resting potential. Applying a small negative current to a neuron will depolarise it for a short time (ie make it closer to a neutral charge), it will quickly return to resting potential. If we raise the current very slowly, we will eventually come to a level called the threshold. Once we pass over the threshold, the current will cause gateways to open up allowing a massive, rapid flow of positive ions into the cell, this causes the potential to shoot up to a high positive level before dropping off again. This is known as an action potential.

An action potential passes along an axon because the positive charge of an area slightly depolarises adjacent areas of the membrane. This sets off an action potential in this area which polarises surrounding areas etc. Thus the potential propogates down the length of the axon. Once an area has gone through an action potential, it becomes less permeable to the ions that are part of the potential. This stops the action potential from occuring and continuing forever or from moving wildly in any direction. The time it takes for an area to recover from an action potential is known as the refractory period. The refractory period thus sets a maximum on the firing frequency of the neuron. If the refractory period is short enough, several action potentials can move down an axon at the same time.

Some (but not all) axons are also sheathed in a substance known as Myelin. Myelin covers the length of the axon except for small nodes about 1mm apart. Myelin prevents the ions from moving through the membrane, but the nodes have many of the ion gates necessary for action potentials. When an action potential depolarises a node, the charge is strong enough to depolarise the next node, skipping all of the distance in between. In this way, the action potential quickly jumps from node to node down the length of the axon, much quicker than an action poptential will usually propogate. This jumping effect is known as saltatory conduction.

Action potentials are either all-or-nothing and their size is the same (completely independant of the size of the stimulus that created it). This all-or-none law is where the analogies with computers were formed, seeming so similar to 0/1-only signals. The 'message' of a neuron is conveyed by the timing of the action potentials. there are several different ways to do this. A one-time action potential may sgnal 'something is here'. Another neuron may be constantly firing in its resting state, but change the frequency of action potentials in the presence of (or at higher intensities of) a certain stimulus. Other neurons signal by sending action potentials in clusters instead of beig regularly spaced.

The graded potential

So far I've only talked about how the output of a neuron travels. The inputs (from the dendrites and there through the soma) propogate in a different way, called a graded potential. As the name implies, the potential is different for different levels of stimulus and the signal also degrades as it passes along the dendrite. Because of this, an axon that is attached to a synapse further from the soma will have less effect than one that is attached closer.

It is in this region that the signals must gather to create the current to trigger an action potential to be generated. If the combined signals are stronger than the threshold value, the action potential starts and the neuron fires. If the cumulative electrical charge of the graded potential is below the threshold, the neuron will not fire. In addition, some nerves act to inhibit a neuron, ie they act to lower the cumulative result.

This process of adding the combined contributions from neurons is modeled in neural network software (though greatly simplified).

The synapse

But how does a neuron actually affect another neuron?

On the ends of the axon's branches are synaptic knobs. These contain vesicles (small cellular containers) of transmitter substance or neurotransmitter. When the neuron fires, this causes a change in ion levels inside the cell membrane, triggering the release of the transmitter substance into the synaptic cleft (the space between the synaptic knob and the dendrite it attaches to).

The actual chemical effect of the neurotransmitter differs from neuron to neuron. There are many different types of nuerotransmitter and it effects different neurons in different ways. I won't go into the chemistry. The most important difference is that some synapses are inhibitory (ie serve to reduce the likelihood that the next neuron will fire) and some are excitatory (serve to make it more likely that the next neuron will fire). When talking about adding up the inputs from various synapses, the excitatory synapses add to the total wheras the inhibitory ones subtract from it.

References:

Coren S, Ward L, Enns J,Sensation and Perception [4^th Ed],
Harcourt Brace college publishers, Fort Worth, 1994

Kalat, JW Biological psychology [5th Ed]
Brookes/Cole publishing Co., Pacific Grove California, USA, 1995