synapse appears as the thin, dark area near the bottom center in this electron microscope picture of a section through cerebellar cortex of a rat. To the left of the synapse, an axon cut in cross section is filled with tiny round synaptic vesicles, in •which neurotransmitter is stored. To the right a dendritic process (called a spine) can be seen coming off of a large dendritic branch, which runs horizontally across the picture near the top. (The two sausage-like dark structures in this dendrite are mitochondria.) The two membrane surfaces, of the axon and dendrite, come together at the synapse, where they are thicker and darker. A 20-nanomctcr cleft separates them.
A segment of nerve axon at rest. The sodium pump has expelled most sodium ions and brought in potassium ions. Sodium channels are mainly closed. Because many potassium channels are open, enough potassium ions have left relative to those entering to charge the membrane to 70 millivolts positive outside. Bottom: A nerve impulse is traveling from left to right. At the extreme right the axon is still in the resting state. In the middle section the impulse is in full swing: sodium channels are open, sodium ions are pouring in (though not in nearly large enough amounts to produce any measurable changes in concentration in the course of one impulse); the membrane is now 40 millivolts, negative outside. At the extreme left the membrane is recovering. The resting potential is restored, because more potassium channels have opened (and then closed) and because sodium channels have automatically closed. Because sodium channels cannot immediately reopen, a second impulse cannot occur for about a millisecond. This explains why the impulse when under way cannot travel back toward the cell body.
IMPULSES, SYNAPSES, AND CIRCUITS
A large part of neuroscience concerns the nuts and bolts of the subject:
how single cells work and how information is conveyed from cell to cell
across synapses. It should be obvious that without such knowledge we are in
the position of someone who wants to understand the workings of a radio
or TV but does not know anything about resistors, condensers, or transistors.
In the last few decades, thanks to the ingenuity of several neurophysiologists,
of whom the best known are Andrew Huxley, Alan Hodgkin, Bernard Katz,
John Eccles, and Stephen Kuffler, the physicochemical mechanisms of
nerve and synaptic transmission have become well understood. It should be
equally obvious, however, that this kind of knowledge by itself cannot lead
to an understanding of the brain, just as knowledge about resistors, condensers,
and transistors alone will not make us understand a radio or TV, or knowledge
of the chemistry of ink equip us to understand a Shakespeare play.
In this chapter I begin by summing up part of what we know about nerve conduction and synaptic transmission. To grasp the subject adequately,
it is a great help to know some physical chemistry and electricity, but I think
that anyone can get a reasonable feel for the subject without that. And in
any case you only need a very rudimentary understanding of these topics to follow
the subsequent chapters.
The job of a nerve cell is to take in information from the cells that
feed into it, to sum up, or integrate, that information, and to deliver the integrated
information to other cells. The information is usually conveyed in the form
of brief events called nerve impulses. In a given cell, one impulse is the same
as any other; they are stereotyped events. At any moment a cell's rate of firing
impulses is determined by information it has just received from the cells
feeding into it, and its firing rate conveys information to the cells that it
in turn feeds into. Impulse rates vary from one every few seconds or even slower to
about 1000 per second at the extreme upper limit.
What happens when information is transferred
from one cell to another at the synapse? In the first cell, an electrical signal, or
impulse, is initiated on the part of an axon closest to the cell body. The impulse travels
down the axon to its terminals. At each terminal, as a result of the impulse,
a chemical is released into the narrow, fluid-filled gap between one cell and
the synoptic cleft—and diffuses across this 0.02-micrometer gap
to the second,
postsynaptic, cell. There it affects the membrane of the second cell
in such a way as to make the second cell either more or less likely to fire impulses.
That is quite a mouthful, but let's go back and examine the process in detail.
The nerve cell is bathed in and contains salt water. The salt consists
not only of sodium chloride, but also of potassium chloride, calcium chloride,
and a few less common salts. Because most of the salt molecules are ionized,
the fluids both inside and outside the cell will contain chloride, potassium,
sodium, and calcium ions (Cl , K+, Na+ and Ca 2+.
In the resting state, the inside and outside of the cell differ in electrical
potential by approximately one-tenth of a volt, positive outside. The precise
value is more like 0.07 volts, or 70 millivolts. The signals that the nerve conveys consist of transient changes in this resting potential, which travel
along the fiber from the cell body to the axon endings. I will begin by describing
how the charge across the cell membrane arises.
The nerve-cell membrane, which covers the entire neuron, is a structure
of extraordinary complexity. It is not continuous, like a rubber balloon
but contains millions of passages through which substances can pass
from one side to the other. Some are pores, of various sizes and shapes. These
are now known to be proteins in the form of tubes that span the fatty substance
of the membrane from one side to the other. Some are more than just pores;
they are little machine-like proteins called pumps, which can sieze ions of one
kind and bodily eject them from the cell, while bringing others in from the outside.
This pumping requires energy, which the cell ultimately gets by metabolizing glucose and oxygen. Other pores, called channels, are valves that can
open and close. What influences a given pore to open or close depends on what
kind of pore it is. Some are affected by the charge across the membrane; others
open or close in response to chemicals floating around in the fluid inside or
outside the cell.
The charge across the membrane at any instant is determined by the concentrations of the ions inside and out and by whether the various pores
are open or closed. (I have already said that pores are affected by the charge,
and now I am saying that the charge is determined by the pores. Let's just say for
now that the two things can be interdependent. I will explain more soon.) Given
the existence of several kinds of pores and several kinds of ions, you can
see that the system is complicated. To unravel it, as Hodgkin and Huxley did
was an immense accomplishment.
First, how does the charge get there? Suppose you start with no charge across the membrane and with the concentrations of all ions equal inside
and outside. Now you turn on a pump that ejects one kind of ion, say sodium,
and for each ion ejected brings in another kind, say potassium. The pump
will not in itself produce any charge across the membrane, because just as many
positively charged ions are pumped in as are pumped out (sodium and potassium ions both having one positive charge). But now imagine that for some
reason a large number of pores of one type, say the potassium pores, are opened.
Potassium ions will start to flow, and the rate of flow through any given
open pore will depend on the potassium concentrations: the more ions there are
near a pore opening, the more will leak across, and because more potassium
ions are inside than outside, more will flow out than in. With more charge leaving
than entering, the outside will quickly become positive with respect to the
This accumulation of charge across the membrane soon tends to discourage further potassium ions from leaving the cell, because like charges repel
one another. Very quickly—before enough K+ ions cross to produce a
measurable change in the potassium concentration—the positive-outside charge
builds up to the point at which it just balances the tendency ofK+ ions to leave.
(There are more potassium ions just inside the pore opening, but they are repelled
by the charge.) From then on, no net charge transfer occurs, and we say
the system is in equilibrium. In short, the opening of potassium pores results
in a charge across the membrane, positive outside.
Suppose, instead, we had opened the sodium pores. By repeating the argument, substituting "inside" for "outside", you can
easily see that the result would be just the reverse, a negative charge outside. If we had opened
both types of pores at the same time, the result would be a compromise. To
calculate what the membrane potential is, we have to know the relative concentrations of the two ions and the ratios of open to closed pores for each
ion—and then do some algebra.
When the nerve is at rest, most but not
all potassium channels are open, and most sodium channels are closed; the charge is consequently
positive outside. During an impulse, a large number of sodium pores in a
short length of the nerve fiber suddenly open, so that briefly the sodium
ions dominate and that part of the nerve suddenly becomes negative outside, relative
to inside. The sodium pores then reclose, and meanwhile even more potassium pores have opened than are open in the resting state. Both events—the
sodium pores reclosing and additional potassium pores opening—lead to
the rapid restoration of the positive-outside resting state. The whole sequence
lasts about one-thousandth of a second.
All this depends on the circumstances that influence pores to open and
For both Na+ and K+ channels, the pores are sensitive to the charge
across the membrane. Making the membrane less positive outside—depolarizing
it from its resting state—results in the opening of the pores. The effects
are not identical for the two kinds of pores: the sodium pores, once opened, close
of their own accord, even though the depolarization is maintained, and are then
incapable of reopening for a few thousandths of a second; the potassium pores
stay open as long as the depolarization is kept up. For a given depolarization,
the number of sodium ions entering is at first greater than the number of
potassium ions leaving, and the membrane swings negative outside with respect
to inside; later, potassium dominates and the resting potential is restored.
In this sequence of events constituting an impulse, in which pores open,
ions cross, and the membrane potential changes and changes back, the number
of ions that actually cross the membrane—sodium entering and potassium leaving—is miniscule, not nearly enough to produce a measurable
change in the concentrations of ions inside or outside the cell. In several minutes
a nerve might fire a thousand times, however, and that might be enough to change
the concentrations, were it not that the pump is meanwhile continually ejecting sodium and bringing in potassium so as to keep the concentrations at
their proper resting levels. The reason that during an impulse such small
charge transfers result in such large potential swings is a simple matter of
the capacitance of the membrane is low, and potential is equal to charge
transferred divided by capacitance.
A depolarization of the membrane—making it less positive-outside
than it is at rest—is what starts up the impulse in the first place. If,
for example, we suddenly insert some sodium ions into the resting fiber, causing a small
initial depolarization, a few sodium pores open as a consequence of that depolarization but because many potassium pores are already open, enough potassium