An electrophysiology quesiton

Matt Jones jonesmat at
Wed Feb 11 16:17:18 EST 2004


Xiaoshen Li <xli6 at> wrote in message news:<c0ddhn$b6e at>...
> Matt,

> Now I understand that, when somebody shows an action potential curve or 
> a neuron is bursting, most of time that was obtained by current clamp 
> experiment.(Voltage clamp holds the neuron at certain membrane potential 
> below the AP threshold, therefore the neuron cannot go through 
> Hodgkin-Huxley cycle).

Yes. But actually you can also use voltage clamp to hold at *any*
potential, not just below threshold. Much useful information can be
gained by studying how ion channels respond to *different* (clamped)
potentials. This is what Hodgkin and Huxley did to identify the
voltage-dependence of the sodium and potassium channels that underlie
the action potential. Remember that as long as you are clamped, the
membrane can't go through the HH cycle, because that cycle requires
regenerative *changes* in voltage.

> To study synaptic transmission, I am still unclear why sometimes an 
> electrophysiologist wants to measure EPSC and sometimes EPSP. (I read a 
> paper which talks about EPSP data for a while, then EPSC data for a 
> while. My eyes were searching "P" or "C" and my mind cannot follow the 
> logic behind it).

Ultimately, it is the changes in potential (EPSPs, IPSPs and spikes)
that the *cell* cares about, because the cell is not normally in
voltage clamp. So if you want to study the physiologically relevant
response, you use current clamp. However, in current clamp it is often
very difficult to dissect out the contributions of specific channel
types (because the voltage is being pushed around by all of them at
the same time). So if you want to study the biophysical basis of the
voltage changes (i.e., what conductances are responsible and how they
work), you use voltage clamp. In the studies where they switch from
EPSPs to EPSCs, they probably first identified some phenomenon by
looking at EPSPs, then studied the underlying mechanism by looking at

> By the way, you said that current clamp inject a fixed amount of current 
> into the neuron and doesn't change it even when the channels are open 
> and the current flows in. Originally I thought the amplifier circuit 
> will adjust its battery to increase its injecting current so the 
> resulted net current be the constant fixed value and this battery 
> adjustment gives the reading of EPSP. Could you explain it a little 
> more? Thank you again.

Actually, no. As Christian said, the name is misleading. You are not
clamping the current of the cell, you are just clamping the current
injected by your amplifier (unlike voltage clamp which is attempting
to actualy clamp the voltage of the cell). And as Richard said, a
current clamp amplifier is essentially just a "voltage follower".
Basically, the electrode is attached to the base of a transistor so
that changes in electrode voltage drive a current into the base.
Changes in this current modulate the conductance of the
collector/emmitter pathway, so that the voltage at the collector is an
amplified version of the base voltage. In reality, operational
amplifiers are used because they have a lot of desireable proerties
that simple transistors don't (e.g., high input impedance, better
stability, better signal/noise ratio, higher gain, etc), but the
principle is as described above.

Xiaoshen Li <xli6 at> wrote in message news:<c0arfk$iul at>...
> Dear Everybody:
> Thank you so much for your help. I greatly appreciate it.
> Please correct me if I am wrong:
> (1)In voltage clamp mode, we usually clamp the neuron's membrane 
> potential at its resting membrane potential (say -65mv) by using some 
> electric engineering amplifier device. When presynapse fires, the 
> postsynaptic channels open, in an excitatory case, the current will flow 
> into the neuron (EPSC) which intends to depolorize the neuron membrane. 
> So the V-clamp circuit will draw the current outward to mantain the 
> neuron Vmem unchanged. The drawing out current in the amplifier is what 
> is measured. It is equal to the current flowing at the synapse(EPSC). 
> What is called "EPSC" is actually the amplifer reading.

Correct. Actually what is called "EPSC" is really meant to refer to
the current flowing at the synapse. If everything is "well clamped",
this will be the same as the current flowing through the amplifier.
However if things are badly clamped, then EPSC *should* still refer to
the synaptic current, not the amplifier current (which is in error in
the bad clamp case). However, most electrophysiologists do continue to
refer to the signal they see on the oscilloscope as the EPSC, even in
unclamped cases where it is definitely not the actual synaptic

> (2)In current clamp mode, I am very unclear now. I guess, we clamp the 
> neuron at a constant current through an amplifier too. 

We do not clamp the neuron at all. We simply clamp the current *we*
are injecting through the elecrode. The current through the neuron's
membrane will do whatever it wants to, causing the voltage to
fluctuate "naturally". However, we can influence the voltage by
injecting our own current (e.g., in order to trigger depolarizations
and spikes, etc). This really should not have been named "current
clamp". It is confusing.

> (How much current 
> should an electrophysiologist choose usually?) Say we choose to inject a 
> small constant current(Io) into the neuron. So, neuron membrane 
> potential will rise from its resting potential in the beginning with 
> certain time constant. After a period, Vmem reaches a plateau, steady 
> state (Vmem=Io*Rmem). Since Io is small, Vmem will not be high, so no 
> action potential firing.


The current to inject depends on the type of neuron, and in
particular, the input resistance. If a cell has a large resistance at
rest, then very small currents (e.g., 10 pA) will cause large
depolarizations (why?). If the cell is leaky, then much larger
currents are required to cause the same voltage change (why?).

> When presynapse fires, postsynaptic channels open. In an excitotary 
> synapse case, the driving force will push current flow into the neuron 
> through the opened channels. This current cancels part of Io. 

This current will ADD to Io, not necessarily cancel it. If they are of
different sign, then they will cancel.

>To keep 
> the current constant, the amplifier device will increase its voltage 
> potential to inject more current. 

No such adjustment occurs in current clamp. It should never have been
called current clamp. The ampifer just sits there, injecting the same
amount of current as before, completely ignoribng what the cell is
doing (except that it *reports* to you the changing voltage, but does
nothing to compensate for it).

> (3)In real life situation(no V-clamp, no I-clamp), a neuron sits there 
> with Vrest=-65mv. When an excitotary synapse fires and the postsynaptic 
> channels open, due to the driving force difference, the current will 
> flow into the neuron. This current will charge the neuron membrane and 
> depolarize it. Is this current equal to EPSC measured? Is this membrane 
> potential change equal to EPSP measured?

In real life, the voltage change during an EPSP should correspond
approximately to what you would see in current clamp (because current
clamp is essentially just observing the "natural" voltage change).

In real life, the *current* flowing through the synapse follows this

Isyn = Gsyn * (V - Esyn),

where Esyn can be taken as some constant (say, 0 mV). 

Therefore, Isyn will also be changing during an EPSP in real life. In
fact, if the cell fires a spike at the top of the EPSP, then V will
actually change sign during the spike, and Isyn will also change sign.
This means that the flow of ions actually reverses direction through
the synaptic channels during a spike (cool, huh?).

The purpose of using voltage clamp is to separate what the channels
are doing from what the cells membrane wants to do. That way, you can
learn how the conductances behave without worrying about all these
changes in current direction.

I might add that there've been some fairly silly comments in this
thread about whether currents and conductances are the same thing and
statements about conductances being better because they have
"direction" and so forth.

Current is a vector. It has a direction and it has a sign. Conductance
is a scalar. It has neither direction nor sign. Conductance is
*defined* as the ratio of current to voltage, and is therefore a
derived quantity. Voltage and current can both exist without


More information about the Neur-sci mailing list