I prefer making my own buffers controlling exactly the concentration of salts and not having a ton of additive things that aren't normally in CSF.

ACSF is the standard. I use it because I want my solutions to be the same across different preps. I find that my culture neurons recording are very similar in some cases to what I see in slice and in vivo. I don't think you will have any problems with using regular ACSF.

ACSF is fine. We also use extracellular recording solution (140 mM NaCl, 5.4 mM KCl, 25 mM HEPES, 1.3 mM CaCl2, 33mM glucose, 0.001mM glycine, and 0.001mM tetrodotoxin, pH7.35, 310–320 mOsm osmolarity) for NMDAR-mediated whole-cell currents etc.

I disagree with the 2 posts above, ACSF is not fine, it is, as you know, a bicarbonate buffer and needs C02 bubbling. This is why, most of us, when recording cultures use a HEPES based solution. I have never read a paper using ACSF to record cultures because it would be difficult to bubble the solution.

It also depends what currents/ channels you are recording.

Hibernate E is not a recording medium but a growing medium I thought.

The reason for using ACSF is historical and practical (junction potential calculation, can keep a while without degradation). "ACSF" however may mean very different composition in papers from different groups, with additives and buffer alternatives. We routinely add pyruvate to our external (Hepes-based) solution because we found our cells to be healhier this way. To me, using a medium close in composition to a culture medium makes perfect sense since 1) your cells grew happily in the culture medium for a week and apparently like it and 2) you want to avoid stressing them during recording. Starving them of AAs may not be a good idea if you are doing long recordings of prot-synth dependent plasticity. Reviewers and colleagues may complain that by using HibE you make comparison with earlier works harder but if you document your evidence that your cells are healthier, then go for it and others will follow you :-) I hope this helps  

For physiological investigations I prefer culture media (with mo serum) buffered with HEPES. The recordings are more stable and there are no possible effect of extracellular solution change on the cells. (and I am also lazy to make the ACSF) For pharmacological investigations I usually make my own recording solution (ACSF with HEPES buffering, not bicarbonate) because the unknown factors in the culture media might influence drug effects. I just do not know and I do not want to risk it.

I agree with Ann Marie, Olivier and Peter that it is better to use a solution with known constituents and pH controlled by HEPES in recordings. If you are lazy to prepare your own use Tyrodes or HBSS supplemented with HEPES. Acidification could be a cause of higher spiking activity when using bicarbonate. Shorter lifetime of cells in acute recording with Tyrodes seems to be strange and should not be entirely related to the nature of external solution per se

Thank you everyone for a very informative discussion! :) I appreciate the advice and thoughts from you. Very interesting to see a wide range of opinions. 

@Ann, Nathan and Ke: The sad thing is that I do have some trouble having long recording using tyrodes, the modified version of ACSF that has HEPES as buffer. Maybe I could try using the standard ACSF (with bicarbonate) instead? Would you mind sharing your ACSF and internal solution recipes? My lab has the carbogen supply and it is possible to do this. 

@Olivier: Would you please talk a bit more what you meant when saying "It also depends what currents/ channels you are recording."? I am recording AMPAR mEPSC at the moment but plan to do some more plasticity experiments in the near term. Do you only include normal salts in your recording solution, or is there anything else that you would add to keep the cells happy for longer? 

@Peter: Does you mean that if you need to look at an effect of drug on the cells firing properties, synaptic activity, etc, then you will use ACSF with HEPES? And if you only need to compare these physiological properties between two genotypes, you will go for the culture media? What is mo serum? 

@Victor: The hibernate E or the tyrodes both do not use bicarbonate as buffer. They have either MOPS or HEPES. So I hope it is not due to acidification that make the cells fire more. So, in your experience, what else would ditate the shorter lifetime of cells in acute recording? 

@Thomas: I very much like your rationale and ideas. These thoughts have a lot in common with mine. :) I do not mind providing evidence showing the cells are much happier in HibernateE. My professor is a bit hesitating to use a brand new solution for ephys though. I guess reviewers sometimes could be more nick-picky than we predict.... 

Thanks again everyone!

Huong, I am sorry for my potential misreading of your initial question on the side of HEPES content. I have done quite a bit of work with cultures of rat hippocampal (targeting Nav channels) and cortical (targeting NMDR) neurons in the past. In my experiments I was using home made or commercial Tyrodes-like solutions with few (if only) problems of longevity in acute recording. Cells were lasting in WCR for 30-40 min easily. The most critical contributor to recording stability in my experience was the osmolality difference between external and internal solutions which should not exceed 5-7 mOsm negative inside. I believe you do measure this for both of your experimental arrangements. Obviously, general health of culture does matter 

Hi Victor. Thanks for the answer! Yes I do measure the osmolarity before every recording. And my internal exceed the external only 5 - 7 mOsm as you said. :) It is reassuring to know that you can record up to 40 min even when using tyrodes. I guess the only two differences here are the culture health and also our patching techniques. I would greatly appreciate if you could speak more of these! 

My thinking is the followings: Physiological conditions is not a simple question. For cultures the culture medium is physiological. If we are interested in how they behave in physiological conditions (undisturbed as much as possible) we should do the recording in the medium. For example if I do non-invasive chronic recordings of network activity with multielectrode arrays, I would do it in the medium. Similarly, those questions targeting drug effects on the physiology of the cells or the network I would use medium with HEPES buffer instead of the CO2 buffering. Logical error, but I would not use serum for the recordings, even if the cells are growing in the presence of serum. In this logic serum required for the differentiation of the cells and should have less effect on acute processes. If I am interested in the pharmacological analysis of a single ion channel or receptor I would use ACSF, because usually those experiments are not physiological, we are setting the experimental conditions to optimize the recorded currents (blocking other channels with toxins or metal ions, etc.). In this case ACSF gives more control. Back to the question of physiological. The assumption is that cells in the culture mimics the   behaviour of the cells in the tissue. It is true more or less in the medium during culture (differentiation, ion channel expression). In ACSF cells can not be cultured. Thus, for me the culture medium is more physiological than ACSF for cultured cells.  

I forgot to ask: It is only the cell recorded WCPC that degrades in 10-20min or is it the whole culture (ie: can you get another good cell in the same dish after the first one). If so, the thyrode is not the culprit but the internal might be.  

@Peter: Thanks Peter. Very reasonable explantation! My hypothesis has something to do with the drug's interaction with the cell physiology, which in turn might affect the AMPA receptors in an indirect way for e.g. receptor density, receptor subunit composition. But the idea is that at the concentration I am using, the drug should not change the kinetic or activity of each subunits. So I guess this would mean using the medium will be a better choice. However, since I only want to look at AMPAR, I do have to block other channels. And you are absolutely right that the medium (like hibernate E) might compromise this goal since it has a lot more ingredient. 

@Thomas: I was able to record on another cell in the same dish after the first one when using tyrodes. Yet neither of them lasted more than 15 - 20 min. Using the same internal and on the same day, I could get 2 - 4 cells from one coverslip if I use hibernate E. Do you think the internal might still be the culprit? 

Huong,

What I meant had to do with the types of currents you record. I record VGCC using a Barium bath, TEA and CS, something the cells don't like very much. In Hepes, each cell recording generally is OK for 15-25 min tops. I don't record with cells in that solution (external) after 1 hour.

Tell us what is in your internal.

Hi Olivier, thanks for the info! Yeah, the salts in your solution sound a bit not very nice to cells. :) But that is just what you need to do for VGCC recording, I guess

I am using either CsGluconate or KGluconate as internal. Kgluconate is for when I want to see action potential. CsGluconate is for AMPAR mEPSC. The original internal solutions were made to be paired up with another tyrode external solution that has high osmolarity (~300 mOs).

Since I recognized that the osmolarity of the neurobasal media is just around 245 mOs and the cells seem to last long in hibernate E (~235 mOs), I am diluting these internals down to ~240 - 242 mOs so that they would match with the hibernate E or the low osmolarity tyrodes (~235 mOs) that I have. I attached the recipes here. We have an osmometer in the new lab so I always check the osmolarity before each recording day. 

About osmolarity of solution. You could add sucrose to the bath to raise osmolarity of the Ibernate to 300 mOsm, and that way you probably would not have to change internal osmolarity that much.

I would recommend a Na-ATP solution or Tris-ATP solution since Mg can be a blocker at NMDA currents and likely affect EPSCs. Also, you could try K methyl sulfonate instead of K-gluconate. See my original post about prior work on that.

Sorry, I got confused with a different post.

This one is the one where other potassium salts are discussed briefly. In there are the references I was pointing you to.

https://www.researchgate.net/post/What_is_the_difference_bwteeen_potassium_sources_used_in_internal_solutions_for_whole_cell_patch_clamp

Hi Olivier,

Oh yes, I can add sucrose to the bath to bring it up. This sounds way easier than adjusting the osmolarity of the internal. I was just thinking since the cells are growing in a lower osmolarity solution, they might not like that. 

Thanks for the ideas about using different ATP salt and methyl sulfonate. I will check out your post  and the papers in there! :)

You wrote above: "And my internal exceed the external only 5 - 7 mOsm as you said. :) ". This is exactly the opposite to what it should be: osmolality of internal solution should be 5-7 mOsm lower than osmolality of external one! :)

Hi Victor,

Thanks for being observant. I actually tried the internal that is lower than the external and for a long time, I could not keep the access being stable at all. :) So I had a discussion on research gate about this. And another scientist shared with me the trick of using internal higher than external in culture neurons. Sorry I could not pull out the discussion link. So far it is really working well in term of keeping the access very constant. :) 

Every lab has its own legends and myths! :)

:)) Ha ha. Yes, that is true in many cases. The good thing is forum like research gate helps us to share and decode the myths and legends. So scientists can avoid making mistake that someone else already did... 

 It's somewhat strange, Huong. It's my long-term experience with multiple living cell types (both recombinant and native cells) that internal osmolality should be lower than osmolality of external solution by 5-20 mOsm. Neuronal cells are especially sensitive to the difference which should not exceed 5-7 milli-units. If not within the range, it leads to cell swelling and preliminary death. This is also long-term experience of my lab which included six conventional and two automatic patch-clamp rigs. It's also experience shared over a large circle of my colleagues both in academia and industry.

I use oxygenated ACSF (300 mOsm) combined with internal that is about 290 mOsm. No problems with simple (up to 20 min) recordings.

Thank you Ewa for the answer! :) 

Hi Victor, Yes, neuronal cells are super sensitive to the difference. The other day I made a mistake with the osmolarity pairing and the cells just crapped out after 10 min or so. I also thought it is fairly strange that my neurons like the internal to be hyper-osmotic in comparison to the external... My new lab is also an intensive ephys lab and everyone else is doing slice recording with 300 mOs external ACSF and 280 - 290 mOs internal. I just analyzed the access resistance from several cells I recorded recently and they either stayed very stable up to 30 min or somewhat improved after the first 3 min. And the values usually below 20 mega ohm. I was never able to have an access as good as this when using hypo-osmotic internal.