I've been trying to set up a NSC differentiation protocol for a while using the 46C mESC, which contains a GFP/Puro reporter under the control of the Sox1 NSC marker gene, as well as with other cell lines such as R1.
At first I had several problems with cell attachment, as non of the cells would attach after plating in RHB-A medium (used several brands including Nunc plates). I could resolve this by increasing the 0,1% bovine gelatin coating time, from 10 minutes to overnight. Also, reducing the trypsin concentration from 0,25 to 0,1%. But the final solution came when I stopped washing the trypsinyzed cells inactivated with FBS containing medium with PBS 1x. I originally thought this was necessary, as traces of serum containing BMP4 could reduce neural commitment. But this was avoiding cell attachment in the RHB-A serum free medium.
Once that solved, I tried several seeding densities ranging 0,5-4x10*4/cm2, which I was told was a crucial parameter, and changing medium every other day. According to bibliography, by day 3 I should have more than 60% of cells expressing Sox1, and by day 4 more than 80%. Those proportions are very far from what I get, only sporadically seeing green cells (I take them in PBS to avoid the medium autofluorescence). Here I attach a couple of pictures of the few cells I managed to obtain in different experiments.
I've been told that I should passage cells by day 4 to laminin coated plates and add bFGF, in order for the NPCs to reorganize into rosettes. But as my GFP levels are so low by day 4 (and up to day 7), I think there might be something I am doing wrong in some part of the protocol.
I also analyzed the gene expression pattern by RT-qPCR using several reference genes (GAPDH, PGK1) with a 1x10*4/cm2 in a R1 mESC line. (Here I attach a graph). Although cells seem to be differentiating as seen by the downregulation of Nanog and Oct4 and upregulation of FGF5, I see no increase in the expression of Sox1 and Nestin. As upregulation of FGF5 came a little later than expected now I started thinking there might be something going on with the state of cells prior to neural differentiation. This led me to analyze the culture conditions for mESC propagation. I noticed a few differences with other publications, most importantly that we use 15% GIBCO FBS in stead of 10%. (maybe more traces of BMP4?)
Also, we use DMEM in stead of GMEM (although there are some papers in this subject that use DMEM). And lastly, we use 0,1mM 2-mercaptoethanol, and I've seen some papers using 1mM.
Next time I do it I'll propagate the cells with 10% FBS before starting the protocol... do you think this might be the solution? Any recommendation / tips are very welcome!!
Thanks a lot,
Ariel
Hi Ariel, what is the reason to grow your cells with monolayer differentiation protocol? I mean, if you just want to expand the neural precursor proportion you could try to keep them in proliferative conditions for some days and then put in differentiative conditions to keep on with your experimental goals. I'm not surprised that you don't see big increase of Sox1-expressing cells under medium containing 15 or 10% FBS...
cheers
Carmen
Hi Carmen! My goal is not to aisolate the NPCs, but to study the transition between the pluripotent state and a very efficient and directed differentiation protocol, such as this one. That's way I need a better efficiency. The fact that it is in monolayer makes it much easier to analyze cells by microscopy.
The actual differentiation protocol is under a serum free medium called RHB-A (http://www.stemcellsinc.com/Tools-and-Technologies_SC-Proven-Product-Catalog/RHB-A). At first glance it seems quite "easy" and straightforward... but appearances can be deceiving!
Regarding the FBS, my fear is that prior to starting the protocol, maybe culturing cells with that extra 5% FBS either makes the cells less "open minded" to begin differentiation as reported, or that it leaves some extra BMP traces that reduces effiency... But I don't honestly think that would make such a difference, right? I think there might be something else going on....
Here I attach the Austin Smith's publication of 2003 that set up this protocol with a slightly different medium, called N2B27, and the Abranches paper in which they used the second generation medium RHB.
Thanks a lot for your kind answer!!
Ariel
Article Ying, Q.L., Stavridis, M., Griffiths, D., Li, M. & Smith, A....
Article Neural Differentiation of Embryonic Stem Cells In Vitro: A R...
Ok, I understood better now. I'll have a look on the papers to see if something else comes up to my mind..but did you try to expand cells in a completely serum free medium before starting the differentiation protocol? I have the sensation too that just 5 % difference in FBS won't make a significant difference. Instead trying to growing them without it maybe could give you dramatic changes to the situation...
Hi Ariel
I'm not working in this field now, but as a co-author of Abranches et al paper (and one of two people who actually set up the modified Quilong Yin protocol) I can give you some suggestions of what might be a bad influence in your system.
The RHB-A should be a medium of choice for what you are planning to do. In the original paper of Yin et al (2003) they used N2B27 medium, whose original composition was defined by Quilong Yin himself. When we started to reproduce this protocol, we tried a new RHB-A medium in parallel with N2B27, and found that RHB-A gave us more consistent results. However, at that time, even with N2B27 medium we were achieving 80% GFP at day 4 with 46C cells, and using other cell lines we could see neural rosettes forming after replating on day 4, meaning that both media are good for neural commitment, but some conditions have to be assured to achieve this.
1.You are absolutely right to worry about 15% FBS in your ES medium, it does influence a lot. We were routinely expanding ES cells in FBS-containing media (DMEM or GMEM, both are good). For best results, plate your cells in defined serum-free medium, like Clonal Grade or similar, for just 1 passage before neural commitment (see below)
2. A very important step before starting neural commitment is the "day before" plating at high density (1-1.5x10^6 cells per one well of 6well plate). On this high density step you allow cells to grow in close contact to each others, and I think that it sort of "synchronizes" them, so they enter differentiation in unison. If you look at neural commitment in embryo, cells first form epiblast=epithelial layer and then a part of it specializes into neural plate which is also an epithelium. So, cells establish apico-basal polarity and adherent junctions, which is very important for neural commitment. If you do this high-density plating in a serum-free medium, you achieve two goals in one step. Do not exceed 24h at this step! Ideally, plate high density in the morning of day -1 of differentiation, and next morning plate for neural differentiation (day0)
3. Always start with freshly thawed cells, maximum three-four passages before initiating neural diff. Multiple passages lower neural diferentiation efficiency, I can't tell the exact reason why
4. The media might be an issue by themselves. We were working with beta versions of N2B27 and RHB, and this is never the same thing as commercial versions. These media have very sensitive components, some of them are known, others not. When you receive a new order, be sure that the medium arrived frozen, and once thawed, aliquot it in one-use aliquots. Once the aliquot is thawed, use it quickly.
I think, you are right to reduce trypsin concentration and adjusting plating density (it is user-specific, the original plating density of Quilong never worked well for us, and we had to find our "own" plating densities).
If neither of above suggestions work, you can try to do N2B27 from individual ingredients (I can send the recipe to you), but this is the last resource and might not work well too, because N2B27 contains insulin which is very tricky to prepare and to maintain soluble. So if nothing works, better try to get a fresh batch of cells
Hope this was helpful
Good luck! Please count on me for any further help you might need in future
Evguenia
Thanks Carmen, it is definetely a good advice! And Evguenia, thanks to you too for that amazingly detailed description! A few months ago we shared some e-mails with Elsa Abranches and also very kindly pointed out a few things to improve in my protocol; she even told me her opinion about the pictures of my cells, she was veeery kind.
Hopefully I can have this protocol running shortly (I'll cross my fingers).
I am going to follow all your advices. At first we started with the home made N2B27 (Elsa gave us the recipe you use), but at that moment we had the low attachment problem. We then swtitched to RHB-A as a less variable and certified medium.
Next time we try, we will lower the FBS concentration to 10% (do you usually use it like this for all your experiments?). We will also try to culture them the day before with serum-free medium. As we don't have Clonal Grade (and it is quite expensive here in Argentina) we were thinking in using standard media with KSR, or home made N2B27 with LIF. Do you think these would do the trick?
I also have a question about trypsin inactivation with serum: when you culture the cells the day before using Clonal Grade (or on day 4 of differentiation) and then you tripsinize them... how to you inactivate the trypsin to avoid putting them in contact again with FBS? Or is it for such a short period that it doesn't influence? is it necessary to use other options like TripleEx?
Thanks to you both again! This space is really great to get in contact with people working in you same area, so eager to help!
Best,
Ariel
Hello Ariel, if you are worried about FBS, I suggest you to add BMP/SMAD1 inhibitor dorsomorphin in the medium. KSR based EB generation should give a high proportion of NPCs also.
Adnan
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Plating