Hi Daisy,

You need to stain the cell membrane and the nucleus. Pixels inside the membrane circumference will constitute the intracellular space. If you subtract the nuclear area from this, you obtain the cytoplasm.

You can use DAPI staining to stain nuclei. Since it gives very bright staining on a dark background, you can use simple thresholding to generate the binary nuclear mask. As far as cell membrane staining and segmentation are concerned, it is somewhat more challenging. There are membrane specific stains from ThermoFisher (e.g. CellMask, CellLight), or you can also apply fluorescent antibodies that will stain the plasma membrane of your cells (this of course depends on your cell type). Simple thresholding will most likely generate poor membrane circumferences, but manually-seeded watershed segmentation works nicely in our hands. You could also consider some machine-learning based approaches, e.g. WEKA in FIJI, or some more advanced tools, e.g. in Matlab, ilastik.

In this paper, we did something very similar, but I'm sure there are many other, similar protocols:

Article Alterations in the properties of the cell membrane due to gl...

All the best,

Peter

Hi there,

Can you share any picture in order to see whether we can help with the automatization process?

Cheers,

J

I describe a general method for doing this in the advanced session of this workshop: https://www.youtube.com/watch?list=PL00PFxcH85LPWLwYxAt6RlUPyCylpDyrv&t=2&v=LbIcDPeqGTE&feature=youtu.be