Download Memory Consolidation and REM Sleep

Memory Consolidation and REM Sleep
Robert Miller
Aldous Huxley's Brave New World considers the possibility of humans learning simply by
listening to recorded messages played during their sleep. Can learning really be this simple?
What is the role of sleep in learning and memory? It is known that sleep exists in two phases,
REM and Non REM sleep. It is speculated that Non REM sleep is the time that the body and
brain use to rebuild themselves after a long period of wakefulness. REM sleep, however cannot
be easily explained. Scientists have speculated that REM sleep performs many functions among
them development of the brain, synthesis of neuro-proteins, and coordination of eye
movements. This paper will explore one specific function of REM sleep: the role in memory
consolidation. There are many theories floating around the internet, but no common truths
about the mechanisms of how the brain serves to remember events of the day during REM
sleep.
REM sleep is a period late in the sleep cycle in which the brain and body become active,
increasing heart rate and blood pressure. The eyes shudder quickly back and forth, giving this
stage the name Rapid Eye Movement (REM) sleep. Electroencephalograph patterns for REM
sleep are much like those during wakefulness, and include many fast beta-rhythms. It may even
be that the brain works harder during REM sleep than when awake. REM sleep usually lasts
anywhere from 11 to 25 minutes, typically longer in the later sleep cycles of the night. REM
sleep is most often associated with dreaming, for most dreams occur during this period.
One of the first theories linking REM sleep to memory was offered in 1966 by Roffwarg, Musio
and Dement and suggested that repetitive firing of neurons during REM sleep in human fetuses
was associated with neuron growth and development, and this synaptic reinforcement
continued in adult life during REM sleep. This landmark concept lead to the theory of dynamic
stabilization which speculates that information, inherited or learned, is remembered by
repetitive use of the circuits which store the information. Dynamic stabilization is the
spontaneous firing of these neurons during REM sleep, which would effect the same result on
memory as function use. The theory of dynamic stabilization also proposes that REM sleep
increases the activity of neuron circuits which are usually dormant during wakefulness. In other
words, memories which we don't think about from day to day are relived during REM sleep (in
the form of dreams) so that we can remember them when needed.
The mechanism which allows this type of memory consolidation to occur during REM sleep
may be found in the hippocampus, a region of the brain with a well-established link to
declarative memory. By studying patients with retrograde amnesia, it was determined that the
"replay" of information in the hippocampus leads to permanent storage of information in the
neocortex. In the absence of such replay episodes, a person will forget, however continued
replay for 1 to 3 years yields memory. Several studies have shown that the hippocampal theta
EEG rhythms assist in transferring information to the neocortex. Furthermore, long term
potentiation, a possible mechanism for long term memory storage, is "optimally induced with
stimulation at theta frequency". Proponents of dynamic stabilization argue that because theta
waves occur continually during REM sleep, REM sleep is involved in the memory consolidation
process.
Replaying memories over in our heads is the nature of dreams during REM sleep. Opponents
of this theory often wonder why don't we remember our dreams if dreaming is involved in
memory? The answer is that dreaming acts as the short-term unconscious mechanism by
which memories may be delivered to the neocortex. We remember the memory, but not how
it is delivered. Remembering dreams "...would be the equivalent of keeping a record of
manifestations of each night's record keeping. Included in the record of one night's record
keeping would be a record of the record keeping of the previous night, which itself would
include a record of the record keeping two nights earlier, etc. Keeping such redundant records,
which would tend to increase exponentially, would overwhelm our brains with tremendous
amounts of useless information (records of record keeping)."
Experimental data with animals seems to agree with the functional link between REM sleep
and memory. Rats subjected to learning tasks demonstrated increased REM sleep after these
tests. This reflects that learning induces REM sleep. In addition, it was found that rats subjected
to learning fell into REM sleep faster if they had experienced more trials of the learning activity.
A possible explanation of the earlier occurrence of REM sleep may be that as more information
is acquired, it becomes more urgent to begin the memory consolidation process. In another
experiment, learning events in rats were paired with audible tones during REM sleep. When
the tones were replayed during wakefulness, the rats exhibited the learned behaviour. Thus
information processed during REM sleep can be exhibited as behaviour during subsequent
periods of wakefulness. Follow up to the above experiment found that neurons in the
hippocampus, the auditory thalamus and the amygdala experienced enhanced discharged
when rats conditioned to react to specific tone were subjected to the same tone during REM
sleep. "By showing that cellular and autonomic conditioned responses acquired in wakefulness
can be expressed in subsequent PS (REM sleep), this set of results suggests that the neuronal
networks involved in learning are in a functional state close to that in wakefulness and can be
reactivated during PS (REM sleep)."
Studies with humans have been in agreement with the animal studies. Experiments on humans
found that REM sleep deprivation leads to poor performance on a variety of recall tests or
logical tasks. It has also been found that memory loss occurs when sleep is deprived on the
same night or two nights after material has been learned. Interestingly, memory has been
affected only when subjects have been selectively deprived of one of the first two or last two
REM episodes of the night. In addition, REM density, or the number of REM's per minute,
increases for several nights after complex material has been studied.
Of course not all scientists agree that there is a link between sleep and memory consolidation.
A counter theory exists which states that the theta waves involved in memory consolidation
reaching the hippocampus do not match the theta wave exhibited during REM sleep. Evidence
to support this theory outlines that the theta waves of REM sleep are random and chaotic, not
organized and orderly, and the hippocampus has no mechanism for transferring disorganized
information.
An alternative theory suggests that REM sleep functions to dispose of unwanted memories
through a mechanism called reverse learning. Reverse learning operates during REM sleep to
prevent the brain from being overloaded with massive amounts of information stored during
wakefulness. The mechanism by which this process takes place is based on the random
stimulation of the forebrain by the brain stem, which in turn excites the neurons corresponding
to the unwanted memories. The strengths of these individual synapses will be modified during
reverse learning so that they will less likely to be activated in the future. The authors of this
theory admit that directly testing this theory is difficult, and thus it remains speculative at this
time.
If there is definite link between memory and REM sleep, it has not yet been proven with
certainty. However, I believe that such a connection exists. There are many mechanisms
through which the brain accomplishes any task. Definitively proving this connection would
imply controlling for all other mechanisms by which the memory stores information. Sleep
deprivation studies cannot accurately measure the body's memory capacity because other
mechanisms will compensate for the loss of sleep. Despite this, there are many studies which
do reflect the brain's decreased memory capacity during REM deprivation. This is a good
indication that a link exists. I know from experience that sleep effects my ability to perform
even the simplest of daily mental activities. Likewise it has not been proven that Aldous
Huxley's vision of learning during sleep is valid. Until such theories are proven, I am content in
thinking that sleep has an effect on memory.