Next: Trading memorizing for thinking
Up: Why Create Models?
Previous: Education: The Proper Balance
Index
Click for printer friendely version of this HowTo
Eric Kandel and his laboratory [11,12,13] used the
sea snail
(aplysia) to demonstrate the mechanisms that implement neuronal
learning. Exploring the siphon withdrawal reflex1.1, they found that
learning was
activity-dependent. That is, training (repeated stimuli)
that occurs frequently appeared
to result in more reliable learning than infrequent stimulation.
Since memory is not magic, then there
must be a change within a cell or within
a network of cells that reflects the learned or memorized element.
Our guess is that short term memory is probably
the result of intracellular accumulation of calcium1.2, Ca,
associated with repeated
training episodes. Long term memory probably reflects
structural changes, either an amplification or attenuation of expression of
a cellular signal receptor or channel.
If short term memory is the result of the accumulation of
intracellular Ca, then there must be some means for controlling this
accumulation.
An action potential1.3
of some duration is generated each time a cell is
excited by a suprathreshold stimulus. The action potential represents the
competition between inward (depolarizing) and outward (repolarizing)
currents and its duration reflects the magnitude of the net inward or
net outward current (net current = inward - outward).
An inward sodium ion, Na, current flowing through sodium
channels is usually responsible for initially depolarizing a neuron and
generating an action potential.
An outward K current is usually responsible
for restoring the charge balanace that is altered by the influx of Na.
Because calcium channels are open at potentials more positive than the sodium channel
activation potential, they are open during a large portion of the action
potential. The length of time the calcium channels remain open is determined by
the duration of the action potential, which in turn is determined by the
amplitude of the repolarizing potassium current(s). Large K currents will rapidly
repolarize the cell while small K currents will prolong the duration of
the action potential. Consequently, the calcium influx can be controlled by
altering the the availability of potassium channels.
Some potassium channels are activated by intracellular second
messengers. These receptor-linked channels are sensitive to
extracellular neurotransmitters (ACh, serotonin, GABA, dopamine)
and provide a way of communicating extracellular events
(presence of a neurotransmitter in the vacinity of a membrane-bound
receptor) to an intracellular process (generating an action potential).
When the time between training episodes
(activation of a channel or activation of a receptor) is greater than the time
required to restore the balance of intracellular agents reflecting
short term meory, then there will be no accumulation and thus no
learning. Its easy to understand why rereading a poem 1/year is less
likely to result in memorizing the poem than rereading it 1/hour.
If, on the other hand, the time between training episodes is less than
the restoration time of the intracellular Ca,
then the intracellular [Ca] will increase until a threshold is
reached (my guess).
The protein expression machinery is probably activated when this
threshold is exceeded - resulting in either expressing
something new, or amplifying/attenuating the expression of existing
cellular component, for example, a
receptor, such that the cell will have a permanently increased
sensitivity to a circulating neurotransmitter. We speculate that
short term memory reflects accumulation of something while long term
memory reflects a structural change in the cell or cellular network.
Given this view, its interesting to look at forgetting and unlearning.
The less something is used, the more likely it will be forgotten.
On the other hand, something that is frequently used will be reinforced.
To unlearn a frequently used habit thus requires significant expenditure
of effort - either to reverse the structural changes or to disable
the structrual changes. If someone want's a great research project - we
suggest exploring the molecular process of forgetting - to complement
Kandel's work on remembering.
This model suggests a main idea: Repetition is essential for most
types of learning. We'll simply state that Repetition is the first
law of learning!
Next: Trading memorizing for thinking
Up: Why Create Models?
Previous: Education: The Proper Balance
Index
Click for printer friendely version of this HowTo
Frank Starmer
2004-05-19