If you always check the pH, osmolarity, and other physical parameters such as temperature (room, ACSF solution), micropipette tip and internal solution before patch, then you have to see if the carbogen aeration of ACSF is OK (in terms of the gas ratio and bubbling).

Also, working at lower temps (19-22 degrees Celsius, near temp shock) decrease excitability in almost all settings compared to 33-36 degrees. Try to increase the recording temperature to something around 35-36 degrees.

It may seem strange but manually check the pressure and shake your carbogen gas tank by slightly rolling the bottom of the tank on the ground for a few times.

Steady ACSF ( pH~7.4) ventilation (with 95% O2 and 5% CO 2) is also crucial, and under no circumstances this should vary.

Regularly bubbled bicarbonate extracellular solution with carbogen can be helpful.

If nothing happens, and everything in the setting seems solid, try to test an increase in Ca and decrease in Mg concentrations.

Hi Akihiro,

What are the ionic concentrations of your intra- and extracellular solutions?

For example, low K+ in the extracellular solution may cause hyperpolarization.

Best wishes,

Paul

Do you see signs of spontaneous activity when you are on cell?

Bahman Sadeghi

Paul G. Morris

William M Connelly

Thank you all for your replies.

Dear Bahman, I of course ventilate ACSFs with 95%O2+5%CO2, I don't know how to confirm the carbogen is certainly dissolved though. According to your suggestion, I will try warmer conditions.

Dear Paul, I used to use 2.5 mM for Kout and 140 mM (with additional KOH for titration) for Kin. I also suspected the imbalance of the potassium, and I actually modified Kout/in to 3.5/120(+addition). It slightly depolarized neurons (~10 mV), but they were still deeply hyperpolarized...

Dear William, Thank you again for your comment. I usually get quick tight seals (

It would be worth performing some on-cell recordings to confirm you get spiking on-cell. If you don't, then your "problem" isn't anything to do with your recording solution.

It is possible that your cells appear depolarized when you break in because of the K+ you have bathed the cell in from your patch pipette. So that were at -70mV before you electrode came along, then get depolarized by K+, and then you see them returning to -70 mV after you break in.

William M Connelly

I appreciate your help. I will try some on-cell recordings from them.

For future reference, could I ask some questions? First, I'm not familiar with on-cell recording, and it is difficult to me to get adequate, somewhat loose, seal. (My seal is usually higher than 5 GOhm ) Do you have any tips to get soft seals?

Second, if I can't find spikes under on-cell recordings, do you think I should suspect the recording ACSFs?

Third, every time I wait for two or three minutes after seal to wash out the pipette solution nearby. How long do you wait for in your experiments?

--- they die.....

Albert Gjedde

Thank you for your reply.

Does that mean the cells are already dead before patching, or the cells are dying when they are getting hyperpolarized? Of course I avoid to use neurons in which their nuclei can be found. I can see spotaneous EPSPs during recording, and their shapes and dendritic structures (revealed by biocytin) seem OK...

At least their ability to maintain a membrane potential is lost and passive ion distribution is lost.

> it is difficult to me to get adequate, somewhat loose, seal. (My seal is usually higher than 5 GOhm ) Do you have any tips to get soft seals?

At least in my understanding, you dont need to (or really want to) get loose seal recordings to perform on-cell recordings. The exact nature of the currents you record will vary depending on whether you have a gigaseal, a loose seal, or are just extracellular, but the facts you are trying to gather from the recording wont change: Are the cells spiking prior to you going whole cell? For what it's worth, to perform most easy to interpret recordings, you want to get a gigaseal on the cell, be in voltage clamp, with Vcmd = 0. You also want your pipette to be filled with aCSF.

However, if you really want a loose seal recording, the method is easy: don't apply positive pressure to the pipette while searching for a cell.

> Second, if I can't find spikes under on-cell recordings, do you think I should suspect the recording ACSFs?

On one hand, that does seem like the logical first step, but on the other hand, could you really have screwed up the aCSF? It's not complicated to make, and by the looks of things, you are an experienced scientist. What could you have done? It seems unlikely to me that that is the problem. I would start to wonder whether perhaps you are in the wrong location. The Hypothalamus is an extremely diverse nucleus (as I know you know). Are you certain that all neurons in the hypothalamus are spontaneously active? Is there any chance you have gotten into the wrong area.

> Third, every time I wait for two or three minutes after seal to wash out the pipette solution nearby. How long do you wait for in your experiments?

Difficult to say: In vivo, the effect of the ejected pipette solution lasts for several minutes. However, in slices, it only lasts for a few seconds. I've never done a rigorous study. I would have thought a minute would be more than adequate in vitro.

William M Connelly

Thank you again for the information. I have tried on-cell recordings. To come to the point, I could find spikes from some neurons, but they were still hyperpolarized... As you can see in the attached, some cells showed rhythmic bursts (aCSF in pipette, w/o synaptic blockers). I noticed that their membrane potential (estimated by adjusting the currents to zero) seemed OK just after seal, but they went deep over time (from -40~-50 to -100 mV). I felt weird because they kept firing even in such deeply-hyperpolarized states. After break in, they went up to around -10 mV.

>could you really have screwed up the aCSF?

No. To be honest, I don't suspect solutions. I had suspected wash-in/out effects. But, as long as I see the data of on-cell recordings, wash-in/out may not be the cause...

> I would start to wonder whether perhaps you are in the wrong location.

I use hypothalamic neurons expressing promoter-driven GFP. So I believe I'm in the right place. GFP-negative neurons also showed strong hyperpolarization. I used to patch mouse CA1 neurons with blind techniques. I have got so many depolarized (basically, dead) neurons there, but I've never met such hyperpolarized neurons. I will check the membrane potential in well-known regions such as hippocampus and neocortex.

Beautiful Rhythmic Activity!

> I use hypothalamic neurons expressing promoter-driven GFP. So I believe I'm in the right place.

Okay, good. I just wanted to cover all the options.

So that x axis is in seconds and minutes correct? i.e. that who trace is 2 minutes and 40 seconds?

I would have liked to have seen a longer recording. It looks to me like you aren't quite perfectly following the membrane potential, i.e. I think the cell was at -50 mV at 50 seconds, then became more hyperpolarized, but it never got down to -100 mV. It's also worth pointing out that the method you are using to estimate Vm is dangerous, as it relies on the fact that the resistance of the patch of membrane under your electrode is much much less than your seal, which isn't necessarily true.

Still, you see spontaneous activity. The same activity that you loose when it whole cell current clamp, correct?. I believe this suggests that either your pipette solution is bad, your electrode is bad, or what I now suspect: your amplifier is passing a some kind of current even when you think you are not. The EPC7 is a very old amplifier. In fact, even when it was new HEKA knew it would develop bias currents. And you need to regularly correct for this. There is a potentiometer you can access on the front panel called "CC zero" (in the top left corner) which I think you need to look into how to correctly set. I believe you just connect a 100kOhm resistor across the headstage, and confirm that you read zero millivolts... but I can't remember exactly. So try to find a manual for the amplifier. Even contact HEKA directly.