Resting "leak" ion channels

Matt Jones jonesmat at
Fri Oct 29 15:12:03 EST 1999

There are two sources of leak, which can be hard to sort out from each
other because of the way leak is defined. In an experiment, one usually
does an I-V curve, and then maybe does one again in the presence of a
cocktail of toxins to block known voltage-gated channels. When you
subtract these two I-Vs, you get something relatively Ohmic, which is
defined as "leak". The problem is that an Ohmic conductance is what you
would expect from just not having a very good seal with the membrane, so
it's not obvious that there are in fact ANY channels at all underlying
this Ohmic leak. However, if you plot resting potential versus various
ion concentrations, you usually find that the membrane's resting
potential is dominated by potassium conductance, with some smaller
contributions from other selectivities.

There are potassium channels and chloride channels that have been
identified and cloned and visualized in single channel studies, that are
relatively Ohmic in their I-V responses, so these would be prime
candidates for "leak" channels. Note that these channels are NOT always
open (if they were, you probably couldn't identify them easily in single
channel studies). However, another way the leak can arise through
channels is through something known as a window current. Here's what that

Voltage-gated channels are both "activated" and "inactivated" by voltage.
That is, if you step from a constant holding potential to various
activating potentials, you can get a conductance vs voltage curve that
starts at zero conductance and increases sigmoidally. If instead, you
start from various "test potentials" and jump to a fixed activating
voltage, you get an "inactivation" curve (or "h-infinity curve") that
starts at a high level of availability and DECREASES sigmoidally. These
two curves cross over each other at some potential, and the area under
this intersection is a region where the channel would be open with some
finite probability at steady-state. The window current region for the
Hodgkin-Huxley k-channel is in fact near the resting potential for the
squid axon, and if you take the HH equations and remove the "leak" term,
you still wind up with a similar resting potential because of this window
in the potassium channel activation/inactivation curves.

Hille's book and Johnston's book must both have some discussion of this
window current effect.


Matt Jones

In article <4JXR3.2431$u3.148177 at> Richard
Norman, rsnorman at writes:
>Austin So (Hae Jin) wrote in message <38175D80.96AC5698 at>...
>>Since no one but that guy ken collins replied...
>Actually Ken Collin's reply, for once, was on topic and actually relevant!
>>"leak" currents are due to things like transporters (in either direction,
>>may or may not involve ATP) which generally co-transport ions through the
>Since leak currents are responsible for electrotonic potentials, they are
>directly proportional to the driving force.  That is usually not the case
>with transporters, especially the active ones involving ATP.  Those usually
>show saturation kinetics.  The pumps, in particular, are controlled by other
>factors and do not readily respond to changes in potential.  And, besides,
>the leaks continue in the absence of metabolic energy.
>>And yes...ion channels do not "leak".
>No, in the closed position they do not.  But at resting potential some may
>remain open and therefore be responsible for the "resting" or "leak"
>>Any good neurophysiology text should give you good information about them.
>The specific reason I ask is that no good neurophysiology text gives the
>I requested.  Of course, they all describe the leak and they all describe in
>the molecular structure of the gated channels.  And some have long lists of
>varieties of gated channels.  But NO text I checked indentifies specifically
>ion channel is reponsible for the resting leak.

More information about the Neur-sci mailing list