An electrophysiology quesiton

Matt Jones jonesmat at physiology.wisc.edu
Thu Feb 12 12:50:16 EST 2004


Hi,


Xiaoshen Li <xli6 at gmu.edu> wrote in message news:<c0ebjc$ktr at portal.gmu.edu>...
> I have another question related with voltage clamp. This question is 
> following:
> 
> "An interesting side issue worth mentioning here is that a voltage clamp 
> applied to one end of a cable is mathematically equivalent to a short 
> circuit at that end. 
> The use of a voltage clamp will thus fundamentally 
> alter the electronic structure of a neuron..." (Foundations of Cellular 
> Neurophysiology by Johnston and Wu, 1995, page 78)
> 
> Could you explain why a voltage clamp on a neuron is adding a short 
> circuit to the neuron? 


If you have a circuit with different voltages at different locations,
and you create a short circuit by soldering a wire between two points
that are normally at different voltages, then those two points will be
"clamped" at the same potential. The reason is that, by definition, a
conductor (your wire) is isopotential (has the same potential
everywhere). So if there were to exist a voltage difference between
the two points, current would flow from one point to the other through
your wire until the potentials were equalized.

This is similar to what happens when ions flow down their gradients
through ion channels. Ion channels also implement a form of voltage
clamp themselves. Current flows in whatever direction tends to take
the membrane to equilibrium, and then stops flowing when equilibrium
is reached. The effect is to clamp the cell at the equilibrium
potential for that channel. By the way, this principle is the essence
of dynamic clamp, which I'll discuss later.

A voltage clamp also makes current flow to maintain a certain voltage.
However in this case, the voltage is not the equilibrium potential, it
is the user-specified potential.

So that is why a) a copper wire, b) an ion channel, and c) a voltage
clamp can all be considered a short circuit-like thing. I wouldn't
really consider a voltage clamp a real short circuit though, since a
short (and a channel) is allowing the system to reach an equilibrium
and the vclamp is imposing a non-equilibrium condition (which requires
that external energy be injected into the circuit).


> Since voltage clamp changes the electronic 
> structure of a neuron, all the data obtained is actually from a changed 
> neuron. 

Yes, but all data obtained by anyone about anything is really from a
changed object. It isn't possible to observe anything without changing
it. In vclamp, we're preventing the cell from following the
equilibrium of its channels. Even in passive cclamp, we've stabbed the
cell with a gigantic piece of glass and altered its resistance. In
whole cell patch recording, we've dialyzed the cell with some alien
solution. In any sort of electrical recording, we're robbing charge
from the cell somehow so that we can measure something or other. In
optical recordings, we've filled the cell with some dye that alters
somethingorother (e.g., buffers calcium, changes membrane fluidity,
what have you).


> Maybe there is a need to reinterpret the published voltage clamp 
> data if the original authors were not aware of this issue.

The original authors ARE aware of these issues (in most cases).
Presumably the authors are smart people. And I KNOW that the people
who developed these technologies are bloody smart people. The data has
already been interpreted with this in mind (in most cases).
Reinterpretation is not necessary as long as you understand the
techniques being used and their advantages/limitations. Richard gave a
good summary of the problems involved in voltage clamp. While you may
have a switch on your amplifier that can turn on the voltage clamp
circuit, that's a far far cry from beng able to actually voltage clamp
anything. Understanding how the thing works is a lot more important
than knowing where the switch is.


By the way, I strongly suggest reading this:

http://www.axon.com/mr_Axon_Guide.html

This is a practical handbook on all manner of electrophysiology
issues, put out by Axon Instruments. It's easy to read, its packed
with important information, and its free.



Now, about dynamic clamp.  First, again I suggest reading before
asking. The Dorval et al paper (posted by Doktor DynaSoar) is a good
thing to read, but it's not the original paper, and it's focus is on
one specific software implementation of dynamic clamp. The two
original papers are:

Sharp AA, O'Neil MB, Abbott LF, Marder E.
Dynamic clamp: computer-generated conductances in real neurons.
J Neurophysiol. 1993 Mar;69(3):992-5. 

Robinson HP, Kawai N.	
Injection of digitally synthesized synaptic conductance transients to
measure the integrative properties of neurons.
J Neurosci Methods. 1993 Sep;49(3):157-65. 


Note that it was developed independently by two labs in the same year,
signifying that it is probably an idea whose time has come. I also
suggest reading the article on dynamic clamp in this Axon Instruments
technical newsletter:

http://www.axon.com/axobits/AxoBits40.pdf



First, Francis Burton said he heard that dynamic clamp was injecting
voltage clamp waveforms that look like action potentials.  This
actually ISN'T what  dynamic clamp means, but it is a very common
confusion.


Second, and I apologize for stooping this low, but this is really
priceless:


"k p  Collins" <kpaulc@[----------]earthlink.net> wrote in message news:<mfzWb.634$hm4.620 at newsread3.news.atl.earthlink.net>...
> I've never heard of it elsewhere, but, if I
> understand correctly, it's =part= of what
> I've been discussing.


Huh? 
"I've never heard of it before, but that's what I've been talking
about all along" (?)



Ok. Moving right along.....





Here's what dynamic clamp is:

You use a special circuit (or computer program) to make your current
clamp amplifier inject the current that would normally flow through an
ion channel:

I = G * (V-E).



So instead of just stupidly injecting a fixed current, as cclamp
amplifiers would normally do, your circuit monitors the cell's
voltage, it monitors the conductance command that you provide, you
also provide it with what you want the reversal potential for that
channel to be, and then the circuit computes how much current should
flow through that conductance, at that voltage, through a channel with
that reversal potential. It sends this current value to the amplifier,
and the amplifier injects that current.


That's it.


Now why is this so cool? Because now you have a current THAT BEHAVES
AS IF IT WERE COMING FROM AN ACTUAL ION CHANNEL. It is as if you
actually inserted an ion channel into the membrane that had the
conductance, kinetics, voltage dependence  and reversal potential that
you specified. The ONLY differences between this and a real ion
channel is that no real ions are flowing, and that you can only insert
this "channel" at places accessible to your electrode. So it cannot
actually let in calcium like NMDA receptors would do, and you normally
cannot place it at the most distal tips of dendrites. But otherwise,
it's identical to a real channel in terms of its electrical effect on
the cell.

Here are a few things that arise from using dynamic clamp that you
CANNOT achieve with either voltage or standard current clamp:

1) Your stimulus tends to draw the voltage towards a specific reversal
potential (like an ion channel does).

2) Your stimulus shortens the membrane time constant (because it
increases cell conductance, like an ion channel does).

3) Your stimulus can mimic shunting inhibition, rather than merely
hyperpolarization (like an ion channel does).

These effects are all simple consequences of the equation above,
coupled with the equation governing how a cell's voltage changes:

I = C dV/dt


Finally, you can implement the dclamp computation either in analog
circuitry or with a computer. Both have their advantages and
disadvantages. I believe the Sharp paper above used a computer and the
Robinson paper used an analog circuit. The analog circuit can be built
for about 40 bucks.




Matt



More information about the Neur-sci mailing list