voltage clamp with high holding potential

[xiaosage]

Hello,

I built a multi-compartment neuron model with active channels. I want to measure the inhibitory synaptic current at the soma using Voltage clamp. I first used VSClamp to hold the voltage at the soma at 0mV (excitatory reversal potential) starting from 0ms to 500ms. Then I gave the neuron an inhibitory input on the dendrite starting at t=400ms. When I checked the current measured by the electrode, I found that there was small-amplitude periodic fluctuation even before t=400ms. In addition, the synaptic current was not smooth.

The amplitudes of the fluctuation will decrease as the time step dt decreases, and will disappear when the holding voltage is under the firing threshould. based on these facts, I guess it is caused by the numerical error related with the active channel dynamics, is that true? Thanks!

[ted]

VSClamp? What's a VSClamp? Did you mean SEClamp?

I found that there was small-amplitude periodic fluctuation even before t=400ms.

Do you have any idea about the degree of voltage control in your model? Does your clamp, whatever it is, have a series resistance that prevents adequate voltage control at the point of electrode attachment?

the synaptic current was not smooth.

Poor space clamp can cause all kinds of strange phenomena. Have you examined the time course of membrane potential at locations distant from the clamp electrode?

The amplitudes of the fluctuation will decrease as the time step dt decreases

Can't say anything about this without knowing a lot more about the biophysical mechanisms that are present and the characteristics of the oscillation. Are the biophysical mechanisms implemented with the ChannelBuilder or with NMODL? I wish I had a nickel for every poorly written mod file I have ever seen; some of those errors affected stability. Anyway, the null hypothesis in modeling is like the null hypothesis in EEG or, for that matter, any wet-lab experimentation: all results are artifact (rubbish) until proven otherwise.

I guess it is caused by the numerical error related with the active channel dynamics, is that true?

It's a guess. As of this moment, there isn't enough evidence to say whether it arises from errors in the model specification (especially the NMODL files), or whether it is attributable to channel gating, or to some other aspect of the equations that describe the model. Models that involve persistent inward sodium current, or other currents that may cause a phenomenological negative slope resistance, are prone to generating low amplitude oscillations. However, reducing dt will not affect such oscillations, unless dt is comparable to the time constants of the processes that are responsible for the oscillation; that is why I wonder if there is some problem with the mod files.

[xiaosage]

Dear Ted,

Thank you so much for your reply!

VSClamp? What's a VSClamp? Did you mean SEClamp?

Yes, I mean SEClamp. Sorry for the typo.

Are the biophysical mechanisms implemented with the ChannelBuilder or with NMODL?

They are implemented with NMODL.

Models that involve persistent inward sodium current, or other currents that may cause a phenomenological negative slope resistance, are prone to generating low amplitude oscillations.

You are so right! I have four types of channels in my model: voltage-dependent Na channel, delayed rectifier K channel (Kd), A-type K channel (KA) and hyperpolarization-activated cationic current (Ih). I blocked them one by one to figure out the cause. When I blocked the sodium channel, the oscillation disappeared. However, if I only left the sodium channel alone while blocking the rest channels, the oscillation also disappeared. It seems that the interaction between the sodium and potassium channels gave rise to the oscillation. Can you explain more about how these channel currents cause a negative slope resistance? And how to solve this problem?

Thanks again for your help!

[ted]

It seems that the interaction between the sodium and potassium channels gave rise to the oscillation.

Done exactly the way an experimentalist would have used channel blockers to dissect the role of particular currents in a biological phenomenon.

Can you explain more about how these channel currents cause a negative slope resistance?

"They" don't. Depolarization-activated sodium current does it. See chapter 8 Nonlinear properties of excitable membranes in
Jack, J.; Noble, D. & Tsien, R.
Electric Current Flow in Excitable Cells
Oxford University Press 1983

You may also find this paper informative:
Ermentrout, G. B.
The mathematics of biological oscillators.
Methods Enzymol, 1994, 240, 198-216

And how to solve this problem?

If you have a particular combination of channels with particular channel densities and reversal potentials, oscillations may occur if space clamp is not perfect. Either change the channel densities (in particular, reduce sodium channel density), and/or make sure that all parts of the model cell are voltage clamped and that the clamp's series resistance is low.

[xiaosage]

I set the density of the sodium channel a little smaller and it works now. Ted, thanks a lot for your helpful answer!