Action Potential bandwidth

Iain McClatchie iain at
Wed Aug 23 15:24:22 EST 2000

Richard> All the studies [...] indicate that ions move in a
Richard> dissipative flow by electrochemical diffusion through
Richard> essentially water filled channels, although with some local
Richard> interaction with ion selectivity sites in the channel.
Richard> There is no opporltunity for the energy released to be
Richard> coupled to any other process without significantly
Richard> interfering with measured flow rates.

If, as you claim, there is no energy coupled into the ion channel,
then there can be no computation.

"Simulations of ion channels - watching ions and water move"
Mark S.P. Sansom a mark at, Indira H. Shrivastava a,
Kishani M. Ranatunga b and Graham R. Smith a
Trends in Biochemical Sciences 2000, 25:368-374

   The motion of ions within pores seems also, in general, to be
   restricted relative to their motion in bulk solution. [....]
   (e.g. an approximately threefold reduction for K+ ions versus an
   ~12-fold reduction for water for a hexameric Alm helix-bundle

Ions, water, and channel are all coupled:

   In the narrow selectivity filter of the bacterial K+ channel
   (i.e. the region of the channel that discriminates between
   different species of ions) a column of water molecules and
   K+ ions moves in a concerted fashion. 

Now I'm not arguing that the ion channels in this bacterium are
trading photons, but it does appear that there is ample coupling
between ions and channel protein, giving opportunity to couple energy
into the protein.  Given that the channel is greatly slowing the
diffusion of ions relative to bulk water, it would appear that a
fair bit of the ions' potential energy is being dissipated into the
channel structure somehow.  In the case of the bacterium and the
human peripheral nervous system, I would expect this energy to be
dissipated as heat with no computation.  But in the human CNS, with
its difficult-to-model complicated ion channels, the energy may be
exchanged between ion channels before dissipation, which would lead
to computation.

I think Dr. Skaggs is onto a bigger problem with this screwy idea:
I don't have any easily testable predictions yet.  Any ideas?

-Iain McClatchie                            650-364-0520 voice                       650-364-0530 FAX
iain at                             650-906-8832 cell

More information about the Neur-sci mailing list