What medium are you giving them? EGF and FGF2 is used for keeping progenitors. Once you withdraw these factors they will differentiate into e.g. astrocytes. I suspect you might selecting for astrocytes through your coating. Neurons are difficult to adhere to glass and usually require laminin or fibronectin to project properly. Astrocytes produce their own laminin resulting in selection.

Also you could try matrix-gel.

Hi,

first of all: what kind of neurospheres are you using? Derived from embryo, postnatal, adult? How long did you keep them in culture? In case you keep them for repeated passages, the neurospheres change their properties. The longer you keep them, the less neuronal differentiation they provide. Did you try to differentiate them actively, i.e. retinoic acid. This works nicely with neuropheres derived from the embryonic, postnatal and adult sub ventricular zone. The suggestion made by Staffan is also worth trying. We do differentiate our neurospheres on extracellular matrix coatings (similar to marigel).

I attach a figure (low quality) as an example (ßIII-Tubulin and GFAP)

Good luck

in addition to the suggestions above - some anesthetics block neuronal differentiation capacity of neural stem cells so avoid their use if you want to generate in vitro cultures.

Without growth factor, you should get very few astrocytes from my experience. What the components of the culture media you used to culture them? I just wondering may be have something to do with that.

Some major points to consider (and a few people before me have touched on them already)

1.  What protocol are you following?  If it's an ES/iPSC differentiation protocol (I infer based on some of your other questions) then ICC markers alone may not be sufficient for defining the cell types you've produced, especially a single marker for each cell fate you are interested in interrogating.  Here is a link to one of the staple protocols in the field http://www.ncbi.nlm.nih.gov/pubmed/19252484 using dual-SMAD inhibition, I suggest following some of their more robust characterization steps and considering purifying your progenitor population.

2.  Even the best iPS/ES differentiations are moderately heterogeneous and populated with many representatives of the mature fates available to the original progenitors produced.  If your major checkpoint is Nestin/Sox2 then that is a very early and non-specific progenitor population (and it can still produce glia!).  http://www.cellsignal.com/common/content/content.jsp?id=stem-cell-lineage-markers this is a wonderful tool that may help you choose more specific markers to define the populations you are producing.  Similarly http://www.rndsystems.com/Pathway.aspx?p=21706&r=15440 can help you adjust your growth factors and medium-inclusions to enrich for the progenitors you require.

3.  If you use gelatin as a substrate in this context (stem cell differentiation, dual-SMAD, etc.) it wouldn't surprise me if you select for glial progenitors over time.  If your goal is to produce neurons then PLO(10ug/cm^2) / Laminin(1-2ug/cm^2) coating is a good coating for proliferative neuronal progenitors. 

Sorry if I was a little general, please ask away if anything else crosses your mind.

Thank you guys, I culture them from cortex, and use F12+EGF+fgf2+N2+b27 to culture them, then I seed them on coated glass cover slip with medium only F12+10%fbs+ngf. as we thought ngf will help neuron growth. but now we didn't get any neuron at all.....even just seed 2nd sphere on coverslip, still no neuron...

And also thank you James Parker  to see my other question, actually ES cell is belong to my another project, this one is totally culture from mouse cortex, and currently we use p2-p4 pups. I do use gelatin at the beginning, then change to PLO, but also no neurons....kind of confused

Hi,

is there any need for the cortex? Why don't you use sub ventricular zone. This is very easy, also for the postnatal situation, and allows to get huge quantities of neurospheres (and neurons after differentiation)

You did not mention the age.

Best

Hi the age is Day3-Day4 after birth, and why not use sub ventricular zone, because we try to analysis cortex, not the ventricular zone.... 

Obvious.

well, there are several papers that address neurosheres approaches from different regions. That works equally. Why not trying the simple approach with EGF, FGF first(DMEM/F12, B27) and cultureSVZ and Cortex. Both should give rise to neurospheres If not, there is a major bug in your approach. We do these cultures often by bachelor students. They are able to get decent results. The method is stable. Cortex should not be the problem.

Best

in additon to the earlier replies

Try untaking immuno to make sure the neurospheres you are starting with are really neuronal as projenitors should also be present. This can be done by embedding the neurospheres in OCT and sectioning at 4um onto slides. You could determine levels of all your markers at this "undifferentiated" stage

For the differentiation stage obviously the glass coverslips will need to be coated otherwise your cells will not stick try both Poly-D-Lysine or gelatin initially as your cells may react differently to the coatings. For the differetiation it self compare growth factor withdrawal with growth factor withdrawal plus addition serum and retanoic acid. The addition of retanoic acid has been proposed to produce neuronal cells.

Good luck

Hi all, in addition information, the cortex we culture was 2-3 days after mice birth, and for Karl's question, it seems SVZ can yield lots neuron, but that's not the specific zone we want to research. we culture cortex first in F12, Egf, fgf2, also with N2 and B27, I can observe sphere after about 1 week, then I passing them, and they form sphere again, this sphere has been examed with Nestin, Sox2, GFAP,all positive.  and now adays I observed some differention in our culture dish, which means if the sphere is big and I didn't passing them, they will attached and differeation, also I observed some neuron-like cell, as without IF, I can't say it's neuron, but it looks like neuron. sounds like good news, huh, but  problem is on glass cover slip, i try gelatin, PDL, PLO,  and seed sphere, sphere was attached after 24h, then it didn't giving neuron-like cell anymore, all astrocytes....I really feel confused, because they have some neuron-like cell in culture dish, but not on coverslip, and I can't do IF in culture dish... 

Hi,

you can do Immunostaining in a dish. Use parafilmsheets. Cut it in the size of the dish, put the staining solutions on your cells, cover them with the parafilm, so you need only 80-100µl solution. That works very well. I did a lot of stainings like this, prior to using coverslips. You can also dissociate the spheres i.e. accouters and paste them on ECM coated coverslips. Use retinoid acid.

Best 

KH

Oh Karl, that's a good idea. I'll try to do IF in the dish and see what can I get, thanks

Hi Monica

you could also do the differentiaion as free floating neurospheres and then OCT embedd and cryosection the neurospheres. This would allow you to keep the neurospheres whole during your differentiation step and then section through the speheres prior to immuno. I found some antibodies didn't penetrate through the whole sphere when thrying to keep them whole, giving variable results.

Good luck