Cell profiler Nucleus orientation assay ,,,Need some advice

Hello to all,

I am a masters student from Cologne, Germany and currently working on my thesis. One of my associate project involves to quantify the phenotype of nuclear orientation in some xyz gene knockdown in HEK293 (HUMAN EMBRYONIC KIDNEY cells) cell line. We have observed a phenotype of round or oval shape of nucleus oriented towards one particular pole in monolayers of cells during the knockdown. This phenotype is distinctive from the normal HEK293 cells. Now we wanted to quantify the nuclear orientation, size, during the knock downs, and a good data could help us, me, to propose a theory that nuclear polarization, is due to the specific knockdowns.
I am a new user to cell profiler, I am going though the instructions, but yet I would like to know the suggestions of this community in the following issues. 1. How many pictures should be considered ( > 1000 ) ? 2. What could be the best pipeline order for analysis, in according to your view ? 3. Could you suggest any controls? 4. What is the ideal Mammalian cell splitting ratios for especially cell profiler analysis, would you recommend any change in growth conditions ?
I would be glad to know your suggestions, and I think you will be the only community I could look for any discussion on this study.

Thank you


Hi Arun,

From the image processing side of things, I can address questions 1 and 2. Questions 3 and 4 are better answered by those who are more knowledgeable about the biology involved.

Question 1: The number of images depends on the type of experiment you are carrying out. The differences between your knock-downs and your controls may be large enough that < 1000 images may be needed, provided your phenotype is sufficiently penetrant. If you performing a larger screen (e.g., RNAi), naturally the number of images will increase as the number of gene targets increase. Also, you will need to consider the number of images per-well that you will need to obtain a representative example of the phenotypes of interest. Again, the details will depend on the biology involved. In any case, CellProfiler is designed to handle both small-scale and large-scale screens.

Question 2: The pipeline will most likely consist of a LoadImages modules to input the images, a IdentifyPrimAutomatic module to identify the nuclei from the DAPI channel, an IdentifySecondary module to identify the cell body from the actin stain based on the nuclei, a MeasureObjectAreaShape module to measure the morphological characteristics of the nuclei and cells, then an ExportToExcel module to output the measurements. The MeasureObjectAreaShape module will give you information about the nuclei shape but to compare the orientation of the nuclei with respect to the cell, MeasureObjectAreaShape will give you the orientation of the nuclei and the cell, which you might then subtract using a CalculateMath module to find the relative alignment; however, the reliability of this approach depends on there being a relatively well-defined axis to the cell as well.