My student is recording voltage-gated chloride currents and is using it to subtract the large passive chloride currents. Some of her traces, especially the ones where she has pharmacologically blocked the majority of the chloride current, look a bit odd. Is there a "noise" level below which it is inadvisable to use this technique?
The major issue with leak subtraction is that it accumulates a few traces (usually responses to a small and reversed step), averages them, and then subtracts them from your main track. Which means that any spontaneous activity (for example) that happens during these leak pre-pulses gets subtracted from your main track as well. For example, if your cell happened to spike during the leak pre-pulses, you'll see very small (because of averaging) inverted (because of the algorithm) spikes on your main trace. And it can be very annoying.
Without seeing your recording I cannot tell what happens there, but try to mentally reverse and increase those strange currents that you see. Do they look familiar?
Alternatively, you could build the "leak-subtraction-like" protocol manually, record it several times, and just see what's happening.
As an electronic device engineer, you question excited me, would you please define what do you mean by passive chloride current. Do mean the noise current caused by the thermal random motion of the chloride ions?.
An important point to consider the nature of the thermal agitation changes its nature with the concentration of the chloride ions may from diffusive nature to molecular flow nature! like the gas molecule behavior with changing pressure. May this thought help!
Thanks Arseny, yeah I know what you mean, I sometimes get the same when I use leak subtract on spiking neurons. However, my student is recording from epithelial cells so they don't do anything at rest, I had therefore assumed that the subtraction would be accurate. I've attached a picture of the trace. It just doesn't look physiological. This is using pharmacology to block most channels (Na, K, Ca and Cl) so it shouldn't have much (if any) current flowing.
Abdelhalim, sorry, not as exciting as you think. By passive currents I mean current flow through channels that are not voltage gated. I am applying voltage steps to induce these currents.
As Arseny, I would suggest to record the P/N traces manually, and check if they evolve in one way or the other - and if so, check if increasing the delay between repetitions suppresses this evolution. If the passive currents are large, the steps to negative potentials may transiently modify ECl. I don't know much about V-gated Cl channels, but steps may also affect where the channels are standing in the activation / inactivation distribution.
Here are some things I've learned about P/N subtraction in my time at the rig:
- the higher the "N" in P/N, the more noise you introduce into the corrected trace
- typically it is best to run the P/N sweeps after the protocol sweep and not before
- the holding potential between P/N sweeps should be the same as the holding potential in your protocol
- typically you want the polarity of the P/N test pulses to be opposite to the polarity of the protocol pulses (to avoid opening channels)
- however, if you have state-dependent blockers, for example a blocker that unblocks at hyperpolarized potentials, you could see unblock during the P/N sweeps which would contaminate your corrected sweep
- therefore sometimes you might want the polarity of the P/N pulses to be the same as the protocol, and you can increase the "N" number to ensure you don't get close to voltages where channels open (then again if everything is blocked this might not matter)
Maybe some of that will help, or maybe you know all of that already, I don't know. Just thought I'd share. One last thing, though - if you are looking at passive currents, might that not be disguised as "leak" and be subtracted through a P/N protocol? I am not familiar with passive currents so I'm just curious.
Dear Morven,
No matter how people is using the P/N protocol, it was originally designed to subtract the linear component of a capacitive transient so you can discriminate gating currents out of it. The basic principle is that you record a fraction of the current evoked on the oposite polarity, then add those traces and subtract to the recording. One important thing then is to set the holding of the P/N protocol within a range where channels you want to record are in the close state.
If a fraction of your macroscopic ionic current is ohmic, like perfect leak, you will subtract that "leak" using the P/N protocol as well. Are your "passive conductances" linear over the whole range of voltages you'll perform the P/N protocol? otherwise you'll be adding current to your recording.
You should check Bezanilla's old articles. As pointed out by Arseny and Rhianna, P/N can be really tricky.
best
Some great answers there. I would always record traces both with and without leak subtraction (Pclamp 10 will save both the leak subtracted and the raw trace). The danger here is that (usually) one uses hyper-polarising pulses to avoid activating channels by depolarisation. If you are using very positive depolarising pulses. (e.g. 100 mV depolarising pulse from -60 mV would be preceded or followed by four 25 mV depolarising pulses that could take the membrane potential to -35 mV and could be enough to activate say a calcium or K conductance. Subtracting that from your trace would give some nonlinear effects. So most folks use hyperpolarizing pulses. However BEWARE if there are significant hyperpolarizing activated inward rectifiers (which there are in many cells) this would also screw up the resultant leak subtracted traces. Finally if your whole cell access resistance and leak currents are poor quality you will get nonlinear effects at very large voltage steps. I think its always good practice, to record both as this will allow you to properly evaluate the quality later and reject poor quality recordings. Finally I would never use the leak subtraction dial on any amplifier as again you can never evaluate the recording quality post hoc. As a rough rule we always reject recordings that have a holding current of greater than 100pA in whole cell.
- Badges
- Science topic
- Similar topics
- Biological Science
- Anatomy
- Tissue