Could be also the general culturing conditions (serum, medium, supplements). The density of 10^6 could be too high. Try to plate 100-200K on the coverslip. Sometimes replacing the medium only once a week is better since by exchanging it too often you are also removing the factors that produce glial cells.
I used standard ACSF (2mM Ca/3mM Mg) as extracellular solution. To record sEPSCs, I used a gluconate-based internal solution and hold the cell at -60mV, the reversal potential for Cl. A Cesium-based internal solution was used to record mEPSC in the presence of 0.5 uM TTX + 50uM picrotoxin.
I agree with you that it could be a general culturing issue. Very occationally we did get cortical cultures with slightly normal mEPSCs frequency (0.5-1Hz) around 14DIV, but most of times the frequency is low. And I have no idea about what I did differently. In another word, I am not vey clear about what contributes the most to the health of the culture. The dissection? Plate coating? Trypsin digestion? or some other factor you suggested, like density and the frequency to replace medium.
Sometimes, we did plate cells in low density, 2.5x10^5 on coverslip, mainly for immunostaining. However, I find that these low density cultures are weaker in our lab. Sometimes they died after one-two hours lipofectamine transfection. I guess this indicates the culture is not very healthy. I also tried to record from these low density neurons, and found they were similar to the high density neurons from the same culture. I will try to change medium once a week next time.
Is this possible that only the miniature glutamate releasing is underdeveloped in our culture?
Are you sure that your extracellular solution is ACSF? Also, I would reduce the Mg to 1 mM or even less than 1 mM. There is an old paper of Dodge and Rahamimoff about the effect of these ions on neurotransmitter release (although in neuromuscular junction but could be right for neurons too - http://www.ncbi.nlm.nih.gov/pubmed/6065887). Addition of Mg could reduce the amplitude of mEPSCs.
Regarding the culturing, many factors could affect it. Sometimes switching to other protocol and using other reagents could help. For example, try papain instead of trypsin, collagen/PDL or PEI coating, other density or medium. We have seen weird effect of different serum batches even from the same manufacturer. Do you add any glial cell proliferation inhibitor? If so, at which stage? The healthy glial layer is essential. That's why I suggested exchanging the medium only once a week. Also, better results could be obtained from the cultures in which the neurons are plated on the glial (astrocyte) supporting layer.
Yes, I'm convinced that my recording condition is fine, including ASCF, since with same setup we can record roburst mEPSCs in pyramidal neurons from 2 weeks old mouse acute cortical slices. I will try to reduce Mg concentration, to see if it will increase miniature releasing probability. Can I also increase Ca at the same time to further increase releasing?
As to the culture conditions, we did try to replace a few reagents in the past couple of months, and some of them (such as new FBS) did improve the general health condition of the culture, although nothing increased the mEPSCs frequency. I compared a lot of cortical culture protocals, and found that people digest the tissure with trypsin for different durations. In our lab, we digest the tissue in trypsin in 37 degress for 20-30min. Is this too long, which will have long term effects on miniature neurotransmission?
No, we never put any glial proliferation inhibitor in our medium, so it is always a neuron/glial mixed culture. But it becomes a problem for us if we want to keep the culture for longer, like more than one month, because of massive glial proliferation consuming a lot of nutrients from medium. A glial supporting layer is a good suggestion and I will try it if I can't fix the issue with the current protocol in the near future.
I would keep the Ca at "physiological" level, for example at 2 mM. The 20 min of trypsinization sounds OK, but I would use papain instead. Also, the cultures could benefit from using the trypsin inhibitor. Also, the glial proliferation inhibitor (AraC or FuDr) could help, when added a 5-7 days after the plating. It is difficult to say what is wrong with the cultures since there are many parameters that could be critical. In general, the neurons plated on glial layer are much healthier than the high density mixed cultures and display good sEPSCs and mEPSCs.
Usually, I did sEPSCs recording first using gluconate internal, then switched to a cesium internal to record mEPSCs in the presence of TTX and picrotoxin, from different cells on the same coverslip. I should have done the sEPSCs and mEPSCs in the same cell, to get my point more solid, which, like you suggested, is "my cells have a low glutamate release probability in the absence of network activity"
I think it is interesting to see certain factors only affect miniature neurotransmission, but not the evoked neurontransmission. I'm not very familiar with neurotransmission release, but does this indicate that two glutamate pool develop independently in vitro? Or evoked release comes first which is followed by mini release?
As to the glial layer, is it possible that I plate glial on the bottom the plate and add a coverslip before I plate cortical neurons? I'm also doing gene expression experiments and don't want to add extra glial into my samples. In another word, does glial have to form direct connection with neuron to support growht? Or they just provide certain growth factor?
There is a method in which the neurons are plated onto the coverslip and then the coverslips are transferred to the dishes or wells with glia. There should be some kind of support to keep the coverslips above the glial layer; usually 3-4 drops of melted paraffin are used. This way you will not add extra glial cells to the coverslip. There are different modifications of this method, I think the original method was developed by Gary Banker; you could try to see the following paper: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3308589/. Also, regarding the effect of glia on mEPSC frequency, look at the paper of Pfrieger and Barres: http://www.sciencemag.org/content/277/5332/1684.1.full
In general, I agree with James about reduction in frequency of spontaneous synaptic events after the TTX addition. By adding TTX you are turning off the network firing and the majority of sEPSCs are basically dependent on the presynaptic action-potential firing. On the contrary, mEPSCs are action potential-independent events and their appearance is essentially random. So, I would do the experiment suggested by James. The above paper of Pfrieger and Barres also reports such a decrease in frequency of spontaneous events after the TTX addition.