Hi Diego,

You have pulled up an interesting paper about the sucrose aCSF protective cutting solutions. Indeed in the Kuenzi et al paper they show that LTP can't be induced with the normal protocol in slices prepared using sucrose aCSF, but when GABA-A receptor antagonist was applied the LTP could then be efficiently evoked--i.e. the inhibition in the slices was strong enough to constrain LTP induction. While at the time it might have been tempting to conclude something was wrong with sucrose aCSF, in hindsight I think most people would agree that the most logical interpretation is that interneurons are more vulnerable to deterioration in acute slices, and that the sucrose aCSF preserves neurons better than standard aCSF (prevents Na and Cl influx and subsequent swelling during the slicing procedure). You may also wish to see the original paper from Aghajanian lab (Aghajanian and Rasmussen 1989 Synapse) describing the sucrose aCSF method. Indeed the sucrose aCSF can preserve both excitatory and inhibitory neurons better, but perhaps the effect is more pronounced for inhibition--which can account for the finding in the Kuenzi et al study. It is arguable which situation is more physiological, but in my view you must already assume a non-physiological state once the brain is removed from the rest of the body and cut into slices. It is a model system to explore synaptic function and that allows better access for interrogation than in vivo techniques. Since it is not particularly helpful to study a few healthy cells in a sea of dead or dying cells, most people are happy to implement techniques that can provide better preservation and healthier neurons. Modified strategies offering better preservation are particularly relevant when comparing two groups (e.g. WT vs KO) since all steps with the procedure are equal. However, it is a tough call when you are actually trying to make measurements that you wish to claim are as close to 'physiological' as possible.

I would only offer up my personal view from many years of work on acute brain slices. Sucrose aCSF is not rare in the field and very likely is the most common method being implemented around the world. It has been used successfully in many contexts including for virtually all brain regions and cell types accessible for recordings in slices, and for juvenile and adult animals of many species. In addition to sucrose aCSF there are many other alternatives that can be highly effective at neuronal preservation (e.g. NMDG, Choline, Tris, glycerol...), and not all sodium substitutes are equal. Agreed that one should investigate in detail what effects (if any) are notable when switching solutions or preparation method, but perhaps it is most reasonable to think about this as a panel of options and not a one size fits all. What works for one person may not work for someone else--experimental context matters. If you are interested to explore more on the topic, I made a website a while back that goes into great detail on the topic and describes my methods for improving preservation of acute brain slices from adult and aging animals (namely transgenic mice). There are many relevant citations listed on this site including some recent ones from my own work. www.brainslicemethods.com

I hope this helps you. --Jonathan

Jonathan's comments are an excellent summary of this issue. Many labs (including ours) now routinely use either sucrose-substituted or NMDG-substituted cutting solution. This seems especially critical for older animals. From personal experience, I have seen reasonably good LTP in slices cut with sucrose, although I haven't made a direct comparison.

In the end, the most important thing you can do is keep everything constant within a given set of experiments. Even if you use young animals now, if you plan to use older animals in the future, you should probably optimize everything for that. If you have good healthy slices, stick with that approach. All other things being equal, that will yield the most reliable data!

This is a very interesting question. I am doing slice recording now, and I am wondering about this too. It is said that some experienced labs used the surose ACSF

Hi, slicing procedure is not physiological at all. Neuroinflamation, energy metabolism deficit are starting during slicing and incubation following thereafter. If you manage to obtain a good electrophysiological response with sucrose solution use this approach. Many people replace NaCl with Choline chloride, others use high K-gluconate solution. Take a look on http://www.brainslicemethods.com/ I adopted some of proposed ideas and it helped me to obtain nice slices from adult mice. Good luck.

Try substituting the NaCl for choline-Cl. check in Mongiat et al PlOS One 2009.

The cutting solution worked pretty well in hippocampal slices fron mice and rats... Is critical to work with chilled solutions and propperly oxigenated with carbogen gas. Also adding kynurenic acid to the cutting solution will prevent excitotoxcicity, and the ascorbic acid will prevent from redox reactions in the ionic channels.

Hi，@Anton Ivanov，I am interested in the K-gluconate solution. Why some people used high K-gluconate solution? What roles does it play?

I absolutely agree with the above statements and I have to stress once more that brain slices are always non-physiological, but they have certain advantages as discussed (visibility of cells, labelling of neurons, recordings from small structures as presynaptic buttons or dendrites etc., see Bischofberger et al., Nat Protoc 2006; Davie et al., Nat Protoc 2006).

I would like to add here that also the way slices are stored is of importance, for example, to record oscillatory response pattern. Slices, which been maintained in an interface-type recording chamber, very robustly show oscillations, but slices stored in submerged-type chambers interestingly do not. It is has been demonstrated that fast-superfusion techniques storage can help to increase the oxygen supply stabilizing neuronal response pattern (see Hajos et al., EJN 2009; and Maier et al., PloSOne, 2009). Not only the oxygen supply, but also the extracellular space, the tripartite synapse, glia coverage etc. might be different in submerged vs. interface-stored slices.

We have used high glucose solution plus AMPA receptor antagonists to cut retinal slices and this has helped a lot with mammalian tissue. If you want to be physiological use 1.2 mM Calcium in your bath and do things at higher temperatures (at least 34 degrees Celcius). Use myo-inositol, ascorbic acid and pyruvate in the external solution (this helps with mammalian tissues). We also like to put phosphocreatine (5-10 mM) in the internal solution of the patch pipette.

Using sucrose or other Na replacers is expected to disturb many Na-dependent cellular transporter mechanisms. This will probably alter neuronal responses. The excessive activation of ionotropic glutamate and GABA receptors due to over release of glutamate and GABA during slice cutting is probably a main contributor to the deterioration of the slice condition (therefore signal quality). Have you tried to add GABAa, NMDA (use D-AP5,not MK-801)and AMPA (as mentioned by Henrique) receptor blockers that can be easily washed out into your regular physiological saline and compare the results with those obtained with the current method you are using?

Christian Wozny brings up very good points for consideration. I would like to add another related paper (Hajos et al, 2009 J Neurosci Methods) describing the dual-superfusion slice chamber (above and below slices) that helps improve oxygenation for slices maintained in submersion chambers. This would be ideal for whole cell patch clamp recordings, whereas interface chambers are more ideal for extracellular recordings like fEPSPs and pop spikes. The study clearly shows that higher flow rates are important for increased oxygenation in the slices and generation of oscillations in submersion conditions. The data also nicely shows how interneurons in particular are negatively effected by low oxygenation and these cells can't sustain firing levels required for maintaining oscillations (e.g. carbachol induced oscillations) unless oxygenation is greatly improved. Those who pay attention to key factors like slice angle, oxygenation, slice solutions, and consistency of technique will be the most successful in the long run.

I can say from my expirience that cutting with sucrose-based aCSF tremendously improved the quality of the hippocampal slices I used to work with while I was patching from DG granule cells of 9 month old mice. As additional trick in a way to improve quality of the slices I can also suggest using of kynurenic acid along with substitution of NaCl with sucrose.

Many important aspects have been contributed to the initial question raised by Diego, already. I just would like to add that Elke Edelmann and I found that sucrose preparation (compared to ACSF prep) changed dopamine content of cortico-hippocampal slices, thereby dramatically changing the efficiency of spike timing-dependent plasticity (STDP) in CA1 of rats. Here, sucrose preparation negatively influenced spike firing properties of CA1 pyramids, leading to impaired STDP. Spike firing and STDP were recovered completely within minutes by brief application of dopamine. Neuronal survival was not affected by ACSF or sucose prep, respectively. Therefore, at least in our hands ACSF prepared slices seem to show better STDP than sucrose prep. This cannot be explained by interneuron survival, because STDP was assessed under inhibition of GABAergic transmission under both conditions.

The experiments are described in the 2 papers:

Dopamine Modulates Spike Timing-Dependent Plasticity and Action Potential Properties in CA1 Pyramidal Neurons of Acute Rat Hippocampal Slices.

Edelmann E, Lessmann V. Front Synaptic Neurosci. 2011 Nov 3;3:6. doi: 10.3389/fnsyn.2011.00006. eCollection 2011.

Dopamine regulates intrinsic excitability thereby gating successful induction of spike timing-dependent plasticity in CA1 of the hippocampus.

Edelmann E, Lessmann V. Front Neurosci. 2013 Mar 6;7:25. doi: 10.3389/fnins.2013.00025. eCollection 2013.

Best, Volkmar

There have been some great thoughts already shared by many regarding various types of preparations. I just need to add one other perspective. We use normal ACSF for hippocampal preparations and sucrose ACSF for ventral tegmental area slices usually. Our slices in the ventral tegmental area seem to really benefit from the sucrose. Therefore, the brain region you experiment on could be important in your decision of solutions. However, what I really want to highlight is the fact that many of us get great LTP while using normal NaCl ACSF. So while sucrose would probably be okay to use, because you are not having success in normal ACSF it suggests something else in your slicing protocol should be fixed. For example, if the brain dissection is not completed within about 1 1/2 minutes or less after decapitation or if you don't use cold 'slushy' ACSF to slice the brain in or if the slicing process itself takes too long in the cutting chamber then you won't get good LTP as the slices won’t be as healthy, as I can attest. So there are many other variables to consider in addition to the solution type you choose. So I would suggest before worrying about only the cutting solution, going back and making sure your entire slice procedures are maximized to be the best they can be is critical. After that then you can examine whether to use normal or sucrose ACSF, so if you do this first, then using either solution your experiments will be better. So I would recommend revisiting your slicing techniques and make sure they are ideal first. Best of luck and again great thoughts by everyone.

Good luck,

Jeff

PS. Submerged slice holding chambers do seem to work best for field EPSP LTP as was mentioned above.

Anton and others,

Choline Cl is an agonist of alpha-7 nAChR

http://www.ncbi.nlm.nih.gov/pubmed/17522313

We compared Choline Cl to sucrose and normal ASCF for slicing. It does affect slices. After this paper we stopped using it. I would not recommend it

Although I also heard that people who previously published using choline-chloride aCSF slicing solution (e.g. late 1990s Dan Johnston and colleagues come to mind) stopped using it because of concerns over potential alterations to neuronal function, there are still many other prominent labs that have consistently continued to use this Choline chloride aCSF solution for slicing (Karel Svodoba and colleagues for example). I think it is fair to say that certain formulations are appropriate in particular experimental contexts but not others, as was nicely illustrated by Volkmar above concerning STDP and sucrose aCSF. However, I would urge people not to make blanket statements to strongly argue against some particular formulation or procedure.

If you are particularly interested in preserving morphology of pyramidal neurons in brain slices (particularly adult animal slice), it is quite obvious to me that choline chloride produces the best morphological preservation of all substitutes I have tried....and I have tried most everything. I would also add that one might also consider the fact that kynurenic acid is an antagonist of ionotropic glutamate receptors, yet many people include it in their slicing solutions and later record activity of these same ionotropic glutamate receptors (after a suitable wash out and recovery time). Thus, the fact that choline itself can act on some receptors is not disturbing in and of itself so long as it washes out after slicing and the 'normal' state can be adequately recovered prior to collecting the experimental data. I am not advocating for use of choline chloride aCSF necessarily, but just providing a slightly different perspective than what was discussed above. It is a very interesting topic for conversation. Although I wish there could be one golden standard way to do slice physiology in all contexts, in my experience it just isn't the case. As a general rule for my slice work I try to do things according to the widely accepted practice first, then if this method proves inadequate I would be more than happy to implement alternative approaches that could allow me to make further progress on an otherwise stalled study. In addition, once I stumble onto a clear improvement in methodology I rarely look back.

Thank you all for your contributions. It's great to know the different views and experiences about this issue.

Thank you very much Jonathan for your response, it has perfectly gone to the point about the viability of the sucrose-ACSF cut slices.

It is very useful to know about the other several improvements you all have proposed and the discussion about the different effects that these modifications can exert onto the results, depending on what one want to study. I was well surprised by the fact that different types of storing chambers can affect the recordings of oscillatory response patterns and other phenomena, mentioned by Christian Wozny.

I do agree and understand that the slice preparation is per se far to be something physiological. However, I just wondered if we can assume that the better preservation of both principal and GABAergic neurons in the sucrose-treated slices, contributes to have a healthier slice, which potentially would give us a closer interpretation to what would be happening in the intact brain. While I am aware of how risky is this assumption, I think this is our goal, that in spite of the slice preparation which is by itself not physiological, we try to achieve something that resembles as much as possible the physiological behaviour of neuronal networks. So, e.g. focusing onto the LTP issue, could it not be that the less LTP seen in the sucrose-treated slices is actually a better approach to what is going on in the intact neuronal network? More straight, would sucrose treated slices constitute a more adequate way to study synaptic transmission, because of its healthier state?

However, at the same time, I find Volkmar’s and Cun-Jian’s observations very meaningful, pushing me back in using sucrose-treated slices. Having this in mind, I completely agree with Jeffrey’s comment, about that I should be able to obtain suitable and healthy slices using regular ACSF. I try to follow the basic steps in order to get and preserve good slices, trying to do it fast and carefully, though (I try to do my best, after one and half year of experience). However, what I find very interesting, is the fact that when I work with mouse, I get way better slices than from rats (both rats and mice 6 weeks old), even performing exactly the same procedure (using regular ACSF for cutting). Do you have any idea why this is happening?

Finally, after one week working with this sucrose-based procedure in my rat slices. I definitely find a higher quality signal. However, the stability of my experiments seems to be somewhat compromised. Have you realised before of something similar?

Thank you again for such a nice discussion. Your responses are really helpful to me.

Diego

[Finally, after one week working with this sucrose-based procedure in my rat slices. I definitely find a higher quality signal. However, the stability of my experiments seems to be somewhat compromised. Have you realised before of something similar?]

You would have to define what you mean by 'stability'---are your responses running down over time? This can be due to several factors---including temperature, mechanical drift of the electrode assembly, instability of the slice (e.g. if it is well submerged and weighted down), and also leakage of solution from the recording electrode tip (if you use aCSF in your recording electrode, it is less of a problem than if you are using 3 M NaCl). It is probably more likely that one of these issues is causing instability, as opposed to the sucrose per-se.

I work in rat spinal cord slices. The gold standard there is to cut the slice in sucrose-aCSF, then incubate them for 1 hr in aCSF with a slightly higher K+ concentration, 5 mM instead of 1.9 mM. Spinal cord slices seem to be much more delicate than hippocampal slices. Most people use them to study dorsal horn physiology in relation to pain processing. The dorsal horn contain many small cells that seem to die quickly if conditions are not optimal. Another issue is the preservation of the axons and presynaptic terminals of primary afferent fibers, which are cut-off from their cell bodies in the dorsal root ganglion. In my experience, primary afferent terminals leak more substance P into the slice if 5 mM K+ is omitted in the 1 hr "recovery" incubation.

I think some systematic studies are needed to fully understand these issues.

I guess I'll contribute by plugging my own paper:

West, P. J., Saunders, G. W., Remigio, G. J., Wilcox, K. S. & White, H. S. Antiseizure drugs differentially modulate theta-burst induced long-term potentiation in C57BL/6 mice. Epilepsia 55, 214–223 (2014).

In this paper, we directly compare the effects of anticonvulsants on TBS- and HFS-induced LTP in both ACSF and Sucrose-ACSF prepared slices. The only caveat to this conversation is that all but one experiment were performed on tissues from C57BL/6 mice (which may yet be another variable to consider). I went into these experiments very aware of the Kuenzi paper (and had been using ACSF for slice prep for a number of years as a consequence). However, my data shows that (at least in slices from mice) LTP in sucrose-prepared slices is equivalent to ACSF-slices. Furthermore, anticonvulsants that are believed to work primarily through a GABAergic mechanism (phenobarbital and tiagabine) do not modulate LTP in ACSF-slices but do attenuate LTP in sucrose-slices.

Take home message: (in our hands, in tissues from mice, using TBS for LTP induction in area CA1) sucrose preparation is superior to ACSF.

To follow up Alexey Semyanov's comment: Choline is also a weak muscarinic agonist. We have noticed that when we use the Choline substituted Ringer for animals past weaning muscarinic M1 responses in cortex are dampened in comparison to those seen in slices prepared using Hepes enriched Ringer. Other GPCR mediated responses (e.g. 5-HT2A serotonin) do not seem to be affected. Other that that the choline substituted solution works great for us when dealing with animals past about a month of age.

In our hand the sucrose aCSF cutting solution is OK. You may want to check into the problem of its pH, temperature or other paramaters as well, such as the age of animals, thickness of slices, minimal "resting" time for slices spent in some control solution after cutting - before your measurements start.

We have used sucrose-ACSF routinely now for many years, examaining the role of inhibition in hippocampal plasticity. In our hands the Kuenzi observation is correct, however I agree with the Jeffrey Edwards in that overall slice quality will be an issue regardless of the ACSF used. Particularly in regard to FS-INs, which receive a far glutamatergic input than principal cells, you will see greater excitotoxicity and cell loss than for the humble CA1 pyramidal cell. And I would go further to say that dissection times of over ~40 seconds already lead to a more compromised inhibitory circuitry.

One other major point stands out for me, do you really want to examine LTP with a compromised inhibitory network. As sucrose-ACSF improves the survival of a greater number of neuron subtypes both excitatory and inhibitory, you are starting from a more natural state, better reflecting the intact network. Fine, network LTP is reduced in this set-up, but this better reflects the LTP profile of a dynamic network, rather than just the excitatory circuitry. If your question is not interested in the surrounding inhibitory network, I would still argue that it is better to remove inhibition in a controlled manner using selective pharmacology (GABAA blockers/antagonists: i.e. picrotoxin, gabazine, bicuculline and GABAB antagonists: i.e. CGP-55,845) rather than the more variable method of sacrificing a unknown number of inhibitory neurons, as with normal-ACSF. Thereby, fully discounting most, if not all, GABAergic mechanisms from the equation.

Furthermore, as I mention above, as some interneuron populations are more prone to excitotoxicity than others, you are not likely completely removing inhibition, rather, more likely selectively removing a population of interneurons (perhaps FS-INs) that you can observe easily at the soma. Which likely reflects large perisomatic inhibitory responses and may potentially leave a population of dendritic inhibitory responses, which you may or may not observe at the soma, intact.

I want to try choline chloride ACSF for acute brain slicing. Is it ok to prepare a stock of all the standard salts (including choline chloride) mixed together (as 1X stock) and keep it in -20 C for a period of month and use it when required after thawing it ?

In my hands, freezing ACSF at -20C is not really good. I usually make a stock solution (including choline chloride) concentrated at 10x (without glucose and carbonate, added at the last moment to avoid bacteria growth and precipitation) and I keep it at -4C for at least 2 weeks. I'm pretty sure that 3 weeks and even 1 month is ok.

Thanks for the interesting discussion. May I add a bit more mess to this argument?

What about effect of perfusion before slicing? With which solution-temperature-duration? Embedding in agar before slicing does something? Using carbon steel instead of zirconia?