You Must Remember This

Bob bbruner at
Sun Oct 21 22:20:14 EST 2001

On Mon, 22 Oct 2001 02:33:02 +0200, "Brian" <zhil at> wrote:

>"Bob" <bbruner at> skrev i melding
>news:3bd334e9.18829718 at
>> On Mon, 15 Oct 2001 01:56:23 +1000, Michael Jameson
>> <m.jameson at> wrote:
>> >The paper that is the main focus of the article is "DNA recombination as
>> >a possible mechanism in declarative memory" by Sandra Pena de Ortiz and
>> >Yuri Arshavsky, Journal of Neuroscience Research, vol 63, p 72 (2001).
>> Ok, I have looked at that paper. It is rather vague, but intriguing.
>> Those who want to consider the questions the poster raised should read
>> it.
>> The basic idea they suggest is that memory storage may involve somatic
>> recombination, as with immunoglobulin genes. They provide evidence,
>> both from the known literature and from their new work, that somatic
>> recombination is "possible" or even "likely" in the nervous system.
>> The proposal does lead to some predictions that are presumably open to
>> testing, so it does seem like a constructive proposal. One could
>> rephrase the proposal... what is the role of enzymes of somatic
>> recombination in the nervous system?
>> Some of the objections raised in this thread are irrelevant. For
>> example, is reverse transcription (RT)  a problem. No, nothing in
>> their proposal requires RT (and if it did, I'm sure the cell could
>> come up with some). Discussions of roles for junk DNA are essentially
>> irrelevant, since there is no clear need for any particular amount of
>> DNA.
>> bob
>Thank you Bob,
>That cleared things somewhat.
>So, basically the long term meories are stored as molecules in the cytoplasm

It is not clear. They really do not propose _how_ it all occurs. (or
if they do, it exceeds my understanding of the neuro issues, which is
not hard  :-) ).

Their key point is the proposed role for somatic recombination (of the
general type used for immunoglobulin gene rearrangements). The hints
for this relate to findings that some of the required gene products
are found in the nervous system, and that knockout mice for these
genes show specific neuro defects.

That leaves a big gap between what is known and their proposal that
DNA, via somatic recomb, may "store memories". But it does open up
some questions for experimental study. Example... Turn off one of
these enzymes (using your favorite turn-off-gene technology, such as
RNA(i)). Does that affect memory acquisition, or any other
identifiable process?

Their "big idea" is certainly speculative. But it does have something
behind it, and leads to expts. So it may be a useful idea, even if the
answer isn't exactly what they suggested. There are tantalizing hints
that somatic recomb is doing _something_ in neuro system.

(Brian... Did you get my private note?)


More information about the Neur-sci mailing list