Neuronal Modelling

Matt Jones jonesmat at ohsu.edu
Thu Jun 11 11:52:31 EST 1998


In article <357FDF14.5E3369CB at idi.ntnu.no> Gerthory Toussaint,
toussain at idi.ntnu.no writes:
>   1) I read that a same neurotransmiter can bind different receptor
>(for exam GABA ==>         GABAa, GABAb). Does a neurone can have
>different type of connexion with other neurones ? I mean, if neurone A
>have connexion with B,C and D. Does it possible for A to excite B with
>GABA, C with Ach and D with NE ?
>

Many neurons can release more than one neurotransmitter substance. For
example, some neurons may release one of the amino acid transmitters
(glutamate, gaba) and also a peptide transmitter (somatostatin, substance
P, enkephalin, etc). But usually, they release both substances at the
same terminals, so that *all* postsynaptic neurons will be exposed to
both. However, since the classical neurotransmitters (amino acids,
catecholamines, etc) are stored in "small clear vesicles" but the
peptides are stored in "large dense core vesicles", they can be released
under different circumstances. Peptides may require high frequency
stimulation for release, whereas the others require only single action
potentials.


>    2) Some neurones have the possibility to firing without any input,
>does this kind of neuron in involved in "rhythmiciti" ? (Does someone
>have a good reference about rhythmiciti in the brain?)

Yes, these neurons may be involved in rhythmic circuits, or as "pacemaker
cells". But there are also rhythmic circuits that *do not* rely on
spontaneously firing neurons, but rather have rhythmicity due to
reciprocal synaptic interactions. Sorry, don't have a reference. But try
looking up work from Eve Marder's lab, or do a search on "fictive
swimming" and "lamprey". These should turn up quite a few references.


>    3) Does with know how many neurones are usually involved in a firing
>process ? (I mean if a neuron has 100 connexions, does it possible that
>only 5 exitatory inputs lead to an EPSP ?)

A typical cortical neuron needs a lot of convergent inputs arriving
relatively close together in time before it fires. Check out the most
recent J. Neurosci paper by Shadlen and Newsome for some calculations
that bear on this issue.

Cheers,

Matt Jones



More information about the Neur-sci mailing list