I just started doing electrophysiology since last year and now interest in learning about the zinc sensitivity of AMPAR mediated mini EPSC. For this goal, I am trying to record from dissociated cultured embryonic hippocampal neurons at DIV 14-17. We use Banker configuration to keep the neurons fairly healthy (co-cultured with rat astrocyte). My recording solutions are: tyrodes (hepes based buffer ~ 250 mOs) for the bath and cesium gluconate (~ 240 mOs) for the pipette. I also have PTX, TTX and APV in the bath to isolate AMPA mediated currents from the others.
In doing so, I have several challenges below. It would be great to have your advice/ recommendation of how to overcome them. Or, just to hear that someone also experienced these difficulties and how they dealt with that.
- Unstable baseline: I could record some decent EPSC traces from normal neurons, with no perfusion. However, when I tried to record on either transfected cells or to perfuse the ZnCl2 during recording, the baseline just looks really not so nice. The combination of the two (transfected + perfusing) is even worse. Do you have any strategy to keep the baseline stable, or to keep the transfected neurons healthy? I am using calcium phosphate to transfect neurons at DIV 9. The protocol is pretty much identical to the one in this paper.
http://www.ncbi.nlm.nih.gov/pubmed/23100419
- Distinguish between an increase in frequency versus amplitude: In my luckiest day, I got a few recording on the untransfected neurons with zinc perfusion and maybe saw more frequent events during zinc treatment. Please see the attached file for an example. I wanted to analyze whether the zinc affects frequency and amplitude of the AMPA EPSC. What softwares or codes would you recommend for this kind of analysis? I am thinking about mini analysis but it is kind of expensive. The clampfit did not work really well (or maybe I am not so good at using it). Also, how do you know if the drug actually increase the frequency or it just increases the amplitude of the events that were small?
- Inconsistent response: Among the cells that I recorded from, some of them seemed to respond to zinc and some of them do not. I am wondering if this is a cell type dependent effect. Other than doing post-hoc staining, how would you try to narrow down the neuron population to patch on? I am trying to only patch on only pyramidal neurons, but there shape of the neurons are so diverse and there are not many typical pyramidal-looking ones.
- Low density: last but not least, it looks like my culture is quite sparse. This is shown by the fact that there are only a few cells (usually less than 10) in an 40x field of view. We plated 900 000 cells on a 100 mm petri dishes that contain 6 of 22 mm coverslips. We later flip the coverslips into the 60 mm dishes coated with astrocytes. I also found that the frequency of the EPSC is quite low. Usually less than 1 event every second. I am a bit hesitant to increase the density since it could affect the transfection efficiency, and sometimes the denser neurons just require too frequent media change. What is the ideal density that one should aim for to have nice and active culture?
Hi Vladan,
Thanks so much for the help! :)
According to your suggestion, it looks like I am already at the top of the density range (55490 cells/ coverslip). Maybe the viability of the cells is not so good since at the time of patching (around DIV 14), I do not have much cells per 40 x field of view. Usually less than 10 neurons if looking directly by eye, and less than 3-4 if looking through the camera. Do you have any experience of how to improve neuronal viability?
I also heard lots of people recommended using Mini Analysis. It is just a bit costly. I'll figure out a way to convince my PI to get it then... :D
Yes, I definitely see how unhealthy the transfected neurons in compare with the untransfected ones. We also tried viral infection and electroporation but these methods require me to do it at early time point. I want one that could be use when the neurons are already at DIV 9 - 10. Lipofectamine is even worse than calcium phosphate in my hand. Did you have experience with any other method?
Thanks for pointing out the noise problem. It is actually not that bad if I zoom in... It is about 5 - 10 pA. But I definitely should work to optimize the recording condition. Ah, the 200 pA peaks are just the response of the cell to my access test. :D I want to constantly check the quality of the recording. And I am holding the neurons at - 65 mV. :) With PTX (100 uM), TTX (1 uM) and APV (50 uM) in the bath, there aren't any activity driven events seen. We also have bicuculine and QX 314, I can definitely try those. Regarding creatine, I have not heard of it. What is it for? Yes, I believe that the seal configuration is still not ideal but not very sure how to make it better. :(
Thanks again for your detail answer, Vladan!
Have a good new week!
Huong
Hi Huong,
I've been recording from rat hippocampal neurons for many years. I find that using E17 provides me with the best cultures and have been able to record out to 6 weeks or longer. I've also found plating 2E5 cells per 35mm plate (using one 22mm coverslip or multiple 12mm slips within the plate) maintains very healthy cultures.
http://www.ncbi.nlm.nih.gov/pubmed/10493110
Another note is to try Lipofectamine 2000 Reagent by Invitrogen for transfecting. I've had great success with this reagent in multiple cell types.
Thank so much, Ann! :) Glad to know another very experienced physiologist. Could you say a few things regarding how you see the physiology of the neurons change over time? I looked up in the literature to find a good systematic characterization of neuronal culture but have not found any that looks promising.
I am also very happy to know that you could maintain the culture for that long since we are a bit struggling with this. How did you actually maintain them for that long? We in fact also use E17 - E18, but the neurons usually do not go pass 21 day in vitro. A few dishes could survive for up to 4, 5 or 6 weeks, but very rare. We have tried a bunch of different cell densities, neuronal media, feeding schedule... At the moment, the banker configuration seems to be best among all, but I am still not super happy (cell density is too low and there are still high variability across coverslips in the same prep). I am not sure if I have written enough description about my protocol here (to avoid endless message), but if you need any further information in order to give me specific suggestion, I'd be happy to write more. :)
I did try the L2K from Invitrogen but the neurons in my hand seemed to not like it. Did you use the protocol recommended on their website or do you have other modification?
Thanks again for spending time!
Huong,
I never use any primary culture until at least DIV14. I've done some characterization and found that many of the channels I have interest in don't start to manifest themselves fully until then but start to show around DIV10. So to ensure expression, I've always waited until at least DIV14. I've found tremendous differences in the type of coating used on the coverslips depending on the neuronal population you want to grow. With hippocampal neurons, I use rat tail collagen/laminin. Aspirate off the laminin directly prior to plating for best results. Serum free media as in the paper I attached keeps the cells happy with only a half media change once a week. Manual trituration of the tissue to disperse the cells for plating vs enzymatic is the gentlest but make sure the tip of the pipette is firepolished so not to shear the cells.
I saw that there is an updated lipofectamine protocol. I use the media the cells are grown in. 120ul media + 5ul lipofectamine, let stand 5 minutes. In another tube place 121ul of media + 4 ul of cDNA. Mix tube 1 and tube 2 together and let stand for 20min. Add to cells dropwise around the plate (this is for a 35mm dish)
Hope this helps!
Hi Huong,
I used this protocol for years and we are getting good co-cultures with it. Check it out.
Goetze, B., Grunewald, B., Kiebler, M., & Macchi, P (2003). Coupling the iron-responsive element to GFP--an inducible system to study translation in a single living cell Science's STKE : Signal Transduction Knowledge Environment, 2003, PL12. doi:10.1126/stke.2003.204.pl12
There are also some old protocols for calcium precipitation transfections out there.
They worked well for me but Lipofectamine might be better today.
Good luck
Thank you Oliver! I am checking out the paper you suggested. It looks really promising and has lots of helpful details! :) I will try the lipofectamine protocol that Ann suggested, yet if you have any good old protocols for calcium ones, would you please share with me? Ideally, our lab prefers the a solution that has the highest efficiency but reasonable cost... We switched from lipofectamine to calfectin quite a while ago since lipofectamine is a bit quite expensive. :D
Hi Vladan,
Thanks so much for the advices and info. I have not tried PEI before but will ask our lab manager to see if we have this reagents. Regarding creatine, my internal has a small amount of ATP and GTP, so I guess this is quite comparable to creatine.
I am super thrilled to hear that you could transfect neurons older than DIV 15. Do you usually use transfected neurons for electrophysiology or imaging? Anyway, this is quite amazing! Please forgive my stupidity, but there are a few points I would like to confirm with you to make sure that I understand the protocol completely.
- Are you growing neurons in Neurobasal + B27 + glutamine without antibiotics? And when doing transfection, you use Neurobasal with glutamine only? Our lab uses neurobasal + B27 + glutamax to grow neurons. We also have astrocytes co-cultured with neurons as in the Banker configuration.
- Usually, people add CNQX and APV in the transfection media to avoid toxicity. Would you recommend doing this?
- When mixing the DNA and the lipofectamine mixtures together, do you have any preference of how to do this? When doing calcium phosphate transfection, I found that bouncing the microfuge tubes to the hood really helps with the efficiency in compare to vortexing or pipetting!
Thanks so much for your time!
Best,
Huong
Hi Huong,
try this protocol: http://www.lamondlab.com/pdf/CaPTf.PDF
I think it’s for HEK cells but setting up the calcium precipitation is the same.
Personally I used (for 6cm cell culture dish):
12 µl CaCl
+106µL H2O
+ 6µg plasmid (2µl volume)
After mixing
+ 120 µl 2xHBS
Most important is that you add the calcium/DNA-mix dropwise to the HBS (better droplets: 3 drops , mixing, 3 drops and so on). The more careful you are doing this the finer will be the precipitate (check it with a microscope). Then add the precipitation mix dropwise to the cells.
Test different protocols to find the best for your cells.
Best,
Oliver
To Vladan: Thanks for addressing my questions in great detail. :) I am very excited to try the new lipofectamin transfection that you suggested. It is quite interesting that you did not use CNQX/APV and the cells still survive in great shape for electrophysiology. Other people have always told me to include these drugs in transfection media. So impressive! Is there a reason why you left B27 out during transfection? I have not heard of this strategy before and really curious about it. By the way, yes, I do use endotoxin-free maxi preps (with the Maxiprep Kits from Quiagen or Invitrogen).
To Oliver: Thanks for sending me the protocol! Besides optimizing neuronal transfection, I actually was also looking for one that specified for HEK cells in order to make lentivirus. I also thinks that this protocol might be also quite reasonable to use for neurons. :) This is great. I found an article in Nature Protocol about calcium phosphate transfection and they also emphasized the importance of how you add the calcium/DNA mix into HBS. A combination between adding dropwise and bouncing the tube seems to give me best efficiency so far. ^ ^
@Vladan: I just want to let you know that your transfection protocol works really nicely! We transfected three synaptic constructs on neurons at DIV 11 (last Sunday - two days ago) and two of them already show nice expression + efficiency. There are about three - five cells got transfected per 10X field. :) Thanks so much for the great advice! I attach here the written protocol in case anyone wants to try.
Thanks for the advice, Moore. :) It makes a lot of sense that large Ra would interfere with mEPSC recording experiment. Usually when I start with a recording, the Ra is very reasonable, like in the range of 9 - 15 M, but most of the times, this value just gets larger and larger, which is quite bad. @.@ I am not sure what to do with it yet. I might fiddle with the osmolarity difference between external and internal solution or apply a bit of negative/positive pressure to avoid resealing. :)
A suggestion for the density issue. When you seed your cell suspension, could you deposit a drop on each coverslip rather than spread it in the dish ? This may require a higher density of the cell suspension. This way, you may keep the same total number of cell per dish but increase cell density on coverslip which should increase health and minis freq. Your RMS noise seems to be around 4-5 pA which should be fine for the clampit template matching. Creating the template is the most important part, do not be satisfied with default values. In our hands, once properly set, it is quite robust.
As for interpretation of frequency increase, indeed minis analysis is not enough. Did you test for presynaptic effects (release probability) in your culture? If there are no PPR changes after zinc application, you may assume that the apparent frequency increase is indeed a post-synaptic potentiation (The literature seems a bit inconsistent on this point).
Finally, you did not mention the ZnCl2 concentration that you are using. At high concentration, you will also interfere with VGCC and this may affect your baseline. Adding a VGCC blocker may help.
I hope this helps. Best wishes for your experiments.
My PI wrote a mini analysis program for Igor Pro that's freely available at http://alford.bios.uic.edu/Research/software.html. If you don't have Igor, the programming language is similar to R, which is free, so you could probably translate it. It works nice for our recordings.
I've also transfected with L2K in cultured neurons around DIV5-6, and they seem to be healthy afterwards. The protocol is attached.
@Thomas: Thanks so much for the suggestions! They are all very to the point and super useful. Regarding the density, we are trying to plate with twice as much cells, and it seems to help. Our lab also use the drop method before, but since we also need the neurons for imaging (like immunostaining), it does not work really nicely. One solution is to have neurons for imaging separate from electrophysiology. But then I do not know if the results from two methods would be comparable to test one common hypothesis. :-? As you may know, density would affect development. I will try template matching of Clampfit! :) It is a great idea. I agree that the frequency increase could be due to either pre or postsynaptic effect. We might use pair cell recording and should be able to address this. Also, I am manipulating a postsynaptic proteins, so that should let me know if the effect is at least partially postsynaptic or not. :) I am using around 10 - 20 uM ZnCl2. Do you know what concentration of zinc would interfere VGCC? I know that Zn also interact with GABA and NMDA as well, so I have the blockers for these in my bath solution.
@Vladan: Thanks for sending me the original reference! Do you have one from your research group that described the protocol? I am just asking so that I might cite your paper in case we publish anything using your protocol. :) Just so you know, the neurons that we transfected 11 days ago with your methods are still surviving and doing really well. I am super happy and surprised at the same time. I did not expect this method to allow such sustained expression. One thing I notice is that among the constructs that I transfected, the bigger one (around 10 kbp) is not transfected as well as the other. Do you notice this before? Would you have any recommendation of how to compensate for the size of the DNA?
@Emily: It is so nice of you to send me the protocol and the mini analysis tool. :) I am quite thrilled to know that you could transfect neurons that early. How old do you usually use them, and what for? :) I have not learnt R before, but definitely will try to learn it soon and try the tool from your PI! Thanks again for the helps!
It is supposed to be a straightforward recording. I do not see any problems from your protocol. The health condition of the neurons and your perfusion system might not be consistent or stable enough for the recording. Increase the bath temperature up to 37 degree may be helpful. Best, Ke
Hi Ke, yes, it should be a straightforward assay to do. I was also working on our cell health and tried different media plus culturing conditions. The Banker configuration seems to give us the healthiest and most active neurons. Increasing bath temperature should also help with activity. I will try this! Is the patching at high temperature a bit more challenging than room temperature? Thank you!
- Badges
- Science topic