I'm wondering if the gluconate is doing something weird to your cells when Cs- is present. You could try for CsMeSO4 instead. There are advocates for and against using either gluconate- or methansulfanate-based solutions. I would suggest trying CsMeSO4 before making any big changes

Are you blocking K+ channels externally as well (e.g. TEA)? This might help. When you depolarize to 0 mV, what kind of currents are you getting?

I've recorded GABAergic neurons using 140 mM CsCl internally for 30 min+ at -70 mV, so I don't think the interaction between Cs+ and Cl- is doing anything.

These two solutions have the same total number of ions, and the dissociation of the gluconate salts are meant to be very similar,,. However there is the possibility that the osmotic pressure may still differ between the two solutions. this is worth a check. It the cs-soultion is hypo-osmotic try adding sucrose to bring it back to that of the bath solution.

If you are using Ag/AgCl electrodes, you should increase the internal chloride concentration (at least 5 mM).

Hope it helps.

I agree with Marcelo on this point

Hello Matheus!

I am having a similar problem when patching relatively large pyramidal neurons in slices or after acute isolation: prolonged recordings with Cs-containing pipette solution at holding voltages lower than -50 mV lead to unstable baseline. Such problem exists only with Cs-containing solutions (Cs-MeS, Cs-F, Cs-Citrate, Cs-Cl). The same cells are doing fine when being recorded at negative voltages with potassium-containing solutions (K-Gluconate, K-Cl, K-Citrate). If I encounter this problem I am trying to keep the holding voltage relatively high (-25 - -50 mV) most of the time during the recording. If a more negative voltage is required I try to minimize the time of the recording or use protocols which allow the cell to rest for some time at -25 mV between the recordings.

I think that this problem might have the following explanation: when potassium is substituted for cesium inside the cell the reversal potential for leak current shifts to 0 mV and lead to stronger inward currents at negative voltages, which is not so good in the long run (it is always better to keep the holding current as low as possible). Smaller cells with high input resistance tend to have much more stable baselines under the same conditions.

Could you please share a solution to your problem if you find one?

Good luck with your experiments!

Are you using a salt-bridge for your ground?

J. Wesley Maddox I'm not using TEA, because we assumed Cs+ was enough. Maybe I should give it a shot if I go back to this configuration. I'm not using a salt-bridge. Just a chlorinated silver coil going in the solution.

Marcelo A Catalán Thank you, I will try this next time. What would be the rationale. though?

Thank you, Paul A Smith. I'll make sure to check my osmolarities again.

Dmitry Amakhin, thanks for reaching out. I thought my K+ leak currents were essentially nullified by Cs+ inside of the cell no? Maybe there is still some resilient current that isn't blocked and J. Wesley Maddox 's suggestion of combining Cs with TEA could be worth a shot.

I’m assuming you’re recording a neuron. Are you blocking voltage-gated sodium channels and glutamate receptors? I have successfully recorded GABAergic mIPSCs by making the reversal potential for chloride 0 mV using a high CsCl internal solution and holding the cell at -70 mV. I’m not sure if this will help you in your configuration.

You may consider using a 3M KCl agarose salt bridge and a AgCl pellet electrode for your recordings (Or just the pellet electrode). In my experience, using a chlorinated silver during continuous perfusion ephys experiments will be fine at the start of your recordings, but after a while, the leak will begin to fluctuate tremendously unless you change to a fresh chlorinated silver wire.

Good luck!

Matheus Macedo-Lima , because silver chloride electrode is a redox electrode, where there is an equilibrium between AgCl and Ag. As you can infer from this, Cl- concentration affects the equilibrium.

Hope it helps.

Hi, Matheus! Thank you for starting this thread! I have the same issue that you and Dmitry report when I record hippocampal and cortical pyramidal neurons. I see the increase in holding current initially anywhere from -30 to -90pA, which doesn't happen with K-Gluconate. My holding potentials range from -65mV to -75mV, depending on my experiment. I notice that the holding current needs approximately 20 minutes to truly level off and stabilize.

I see this issue with both Cs-Gluconate and with Cs-Methanesulfonate. I use TEA in my cesium-based internal recipes. The phenomenon still exists. Phenomenon also still exists despite verifying internal solution osmolarity between 285-295 mOsM.

My thoughts were the same as what Dmitry stated, although I'm not 100% positive. My thoughts are that the cesium is blocking the leak potassium channels, and thus I need to inject more negative current to achieve the desired holding potential. My "input resistance" is also changing (decreasing) over time, which is also puzzling because I was expecting it to increase with cesium-based internal solutions.

Any ideas? Do you see similar things? I am doing LTP recordings with whole-cell, and my inclusion criteria have needed modification to accommodate the strange changes in holding current and Rinput. Any input would be greatly appreciated!