Distinguish 2 cell types differing in surface marker expression



Good afternoon,

My project consists in quantifying the levels of a transcription factor in the nucleus and cytoplasm. Because the transcription factor is also in the cytoplasm, I have been able to use its staining to define the cells and the outer limits of the cytoplasm (identify secondary objects) after having identified the nuclei (via identify primary objects). I then identified tertiary objects (cytoplasm) and got the values for the transcription factor in cytoplasm vs. nuclei.

My experimental setup will change in a way that I will have, in the same picture, 2 different cell types that I’d like to analyze as described above, with the added data that the 2 different cell types have a different surface staining (cell type A is stained with BV421 and cell type B is stained with Pe-Texas Red). How do you suggest that I build the pipeline so that it distinguishes between the two cell types (via a mask?), defines the outer edge of the cell, and quantifies the transcription factor expression in cytoplasm vs. nucleus. (Cytoplasm can still be defined as done previously using the transcription factor staining).

Thanks in advance for your help in this!


I would do the following (if I understood your needs correctly…)
1/ do your segmentations as before, based on TF staining
2/ add intensity measurement on cytoplasm for the 2 channels corresponding to BV21 and PE-TR
3/ filter objects as type A or B based on intensity measurements (mean intensity or mean intensity edge for a surface marker) by setting a threshold
4/ do measurement of TF in related cytoplasm & nucleus for both A and B cell types

Doe’s it make sense?
Good luck :wink:


Bonjour Fabien!
It makes total sense. Thanks a lot for the quick reply!
Will try it and get back you!


Hi Fabien,
I gave it a try and it worked well!
You’ll find here the pipeline and the pictures (the two cell types are distinguished by either Thy1 in c3, or CD45 in c4. DAPI is in c2 and the transcription factor is in c1). In order to get cytoplasmic and nuclear values of the transcription factor in each cell type, I’ve added relate objects, but I feel there might be an easier way. What do you think?
Many thanks for your help!
PhilippeThy1 Cd45 filtering.cpproj (140.5 KB)