You may try make pictures on lowest possible power and connect them digitally like some people make panorama photos. I'm sure there is software for this.

You can try stain the cells with a fluorescent probe and try to use a gel analysis station if you have some by the hand

you can also use the crystal violet and try to scan on flat high-resolution scanner,

you may also try to stain the cells prior to plating with non-toxic dye and ....

just popping out ideas

try to take at least 15-20 pictures per well and count with imageJ. this is acceptable by journal reviewers. we fixed and stain them as well. we did both cryo section and staining on whole spheroids. to begin with using 1ml tips take media and spheroids in eppendorf and allow them to setttle down, care fully suck the media with 1ml tips only, do not use centrifugation or vacuums. All steps have been performed at room temp, using room temp reagents. I have been removing the media (and all subsequent washes/incubations) using a pipette (not an aspirator), fixing cells in 3% PFA for 30 min (3X PBS wash), permeablising in 0.5% T-X100 for 10 min (3X PBS wash) and blocking in 8% BSA-PBS + 0.01% Tween20 for 30 min (3X PBS wash). This is followed by RT Ab incubation, primary 3 hours RT, secondary 1 hour RT, before removing plastic wells and covering slide with a glass coverslip. for cryo section please find the protocol published in 

http://onlinelibrary.wiley.com/store/10.1111/cas.12475/asset/cas12475.pdf?v=1&t=i2g81l0b&s=192874e97de3a38eee0b549939c7a47af95f8047

Good Luck

Both cell lines grow in methylcellulose (approx. 0.8%) media.  Colonies formed in methylcellulose tend to settle into a single plane, which facilitates counting. If you use non-adherent plastic dishes, the spheres should not attach and spread.

Some investigators prefer inverted microscopes for counting, others prefer dissecting microscopes.  If you have an inverted microscope (preferably one with long working distance objectives), use it in the bright field mode and partially close down the aperture diaphragm to achieve some depth of field.  I prefer this approach to using phase contrast optics.  If you want to go for a dissecting microscope, be aware that not all dissecting microscopes are equally suited for the counting of colonies.  The optical axes should be at an angle (as opposed to being parallel), and you need to be able to "play" with the illumination system.  Olympus used to make one that was really well suited for this purpose.  I don't know if it is still available.  As far as "playing" with the illumination system is concerned:  it's easier to show than to describe.  Look for a colleague at your institution who uses a dissecting microscope to count hematopoietic colonies.

By the way, don't be surprised if plating efficiencies are a bit low in methylcellulose.

Hi Tiago, 

I generally seed spheres that I want to enumerate into 24 well trays. I have drawn up plate templates with grids on Powerpoint that I've then printed onto clear transparency sheets, and hence I can place the sheets under the tray and simply enumerate using a light microscope. No staining required. You could also do this with 6 well trays but it would take longer (depends what you are trying to assess and whether you need to use the spheres for a secondary passage etc).

If you are growing up large numbers of spheres, you can always take a set (smaller) volume and move to a 96 well tray, and count them using a light microscope (if you check out Stem Cell Technologies website, I think they have a video on there for human spheres), then extrapolate that to estimate the total colonies in your culture. 

If you want to stain for anything, you can cyto-spin sphere cultures onto poly-lysine coated slides and stain for H&E or do IHC/IF on the cells. This procedure requires a bit of optimisation as the spheres have a tendency to break up when put under this force. Not many people working with primary sphere cultures publish with this method but you might have more luck with cell lines. 

Good luck!

I am not sure about overcoming any problems you may encounter with the cells being in suspension but if you are able to stain and image these spheres, which it sounds like you can then I think it would be perfectly acceptable to image and count several fields of view (say 10 on low magnification 10x) and then quantify an average number of colonies from these images.