LTP via NMDA in neocortex

Bill Skaggs bill at
Thu Mar 21 18:03:15 EST 1996

CHOLSCHR at TCD.IE (Christian Holscher) wrote:
  > I hate to spoil the party but surely you folks have kept up to
  > date on recent developments on learning and memory, LTP, and the
  > role of the NMDA receptor. Most interestingly enough, straight
  > from the lab of Richard Morris (the man who introduced the
  > NMDAR-learning correlation) a paper has been published recently
  > which simply states that at least spatial learning is most
  > probably not dependent on LTP in the dentate gyrus and on the
  > NMDA-receptor.

I think this needs a bit of clarification.  The paper you refer to
(Nature 379:826-829) is very nice if you read the whole thing, but a
bit misleading if you only read the conclusions.  The main source of
confusion is that the term LTP has been used in the literature to mean
two rather different things, and the authors do not carefully
distinguish between them.

Definition 1:  LTP is a long-lasting increase in synaptic efficacy,
  commonly (but not necessarily) induced by strong, high-frequency
  stimulation of input pathways.

Definition 2:  LTP is the process of applying strong, high-frequency
  stimulation to an input pathway and thereby causing a long-lasting
  increase in synaptic efficacy.

The question here is whether the word LTP applies to the process
(stimulating input pathways) or its result (an increase in synaptic
efficacy).  This distinction might sound like nit-picking, but it's
actually crucial, because the animals in the Nature paper lacked LTP
in the dentate gyrus according to definition 2, but possessed LTP
according to definition 1.  One of the results of the paper was that
synaptic enhancement *could* be produced in these animals if a GABA
antagonist was used to reduce the level of inhibition.  The authors
concluded, in fact, that their animals were probably not lacking the
molecular machinery for synaptic enhancement, but rather were subject
to stronger inhibition in the dentate gyrus.

Why does this matter?  It matters because, functionally, definition 1
(LTP as a result) is surely more important than definition 2 (LTP as a
process).  Massed high-frequency activity of input pathways is
effective for enhancing synapses in the laboratory, but it doesn't
occur under natural conditions, where synaptic enhancement, if it
occurs at all, must be produced by quite different sorts of activity
patterns.  Therefore, the fact that massed high-frequency stimulation
is no longer effective in the mutant animals does not by any means
imply that the "natural" synapse-enhancing patterns will no longer be
effective.  It is reasonable to conclude that the dentate gyrus will
not function in exactly the same way in the mutant animals as in
normal animals, but it is not reasonable, on the evidence presented,
to conclude that they lack the capacity for synaptic change under
natural conditions.

There are other quibbles, but this is the most important one.

	-- Bill

More information about the Neur-sci mailing list