"Andrew Gyles" <syzygium at alphalink.com.au> wrote in message news:<9om3lc$e05sl$1 at ID-94640.news.dfncis.de>...
> In December 2000 there was some discussion in this newsgroup of my
> hypothesis that mitochondria might act as flip-flop memory elements in
>> One counter-argument ran as follows: "Why should neurons need a flip-flop
> memory element? There's no evidence they're digital".
I remember that discussion. I hope I wasn't one of the people arguing
that neurons aren't digital. I don't think this "not digital" argument
is really a good argument against your hypothesis.
My arguments would be this: OK, mitochondria might act as flip-flops.
But a) do they really do that? b) If they do do that, is that behavior
actually -used- by the neuron to store information that's relevant to
how it acts as a signaling element within the circuit, and c) there
are lot's of other ways that a neuron could incorporate flip-flop
types of behavior, e.g., latching of CAM kinase II into an
autophosphorylated state, expression of previously silent AMPA
synapses, transcriptional regulation of genes for K channels, etc etc
etc.. Practically -any- neuronal behavior could potentially serve as
the substrate for a flip-flop role if regulated appropriately. So why
is the mitochondria-based mechanism more plausible or likely to be
useful to the neuron than any of these other, more well-established
mechanisms for generating the same sorts of behavior?