Q about artificial and biological neural-nets

Douglas Fitts dfitts at carson.u.washington.edu
Sun Feb 28 10:52:02 EST 1993


In article <1moshmINNgcp at shelley.u.washington.edu> lamontg at u.washington.edu writes:
[beginning omitted ...]

>recently, i was having a disucssion with a friend of mine over a 
>hypothetical model i have of a small amount of "neurotoxicity" effecting
>specifically axons, possibly being actually functional in terms of
>producing a more efficient, and well-adjusted neural net.  the hypothesis
[rest of post omitted...]

I'm not sure if you intend some kind of exogenous toxicity or not.
As far as I know, toxins tend to affect either the entire neuron, such as
lipid soluble toxins (alcohol), or else act at one or other end if the fiber 
where there are macromolecules other than ion channels it can attack (the 
channels are also not specific to the axon).  Excitotoxins attack receptors 
on cell bodies, thus gradually causing the whole cell to die, and these 
can be used to distinguish whether the effect of a gross lesion
such as an electrolytic lesion works because of damage to cell bodies or
to fibers of passage.  That is, neurons having their cell bodies
in the area will bind the toxin, become hyperexcited, and then osmotically 
break open because of too great an influx of ions.  Cells having only a fiber 
passing through the area are not attacked.  I don't know of any toxins that 
specifically attack axons without attacking the entire cell.  Anybody want to
jump in here?  I'd like to know if the complementary tool to excitotoxins 
is available, especially if usable in vivo.


Doug 



More information about the Neur-sci mailing list