Dear Teena,

I personally do not work with endothelial cells, so it would be helpful if you could share an example image from your experiment. However, judging from your description (cells in very close proximity) I would suggest the following steps to get the cell area: 1. A manual delineation of the area of each cell (ROI) using selection tools (freehand selection or polygon selection). 2. Run Analyze -> 'measure' function. It will give you the area of your selection, which would be the area of your cell. The only thing you should keep in mind is that your image should have a proper scale as otherwise your area will be measured in pixels instead of micrometers. I would also always save your ROI's using ROI manager for future reference. Also, you can use a macro to speed up step 2.

For the tight-junction measurement, there are many options, but it all depends on how the staining looks like in your case (continuous or rather dotty) and what scale you would like to measure (single-cell vs the whole image).

For continuous staining, I would think about measuring mean intensity ('measure' function), while for dotty staining you could binarize the picture and also use a 'particle analyzer' to calculate the density or/and 'measure' function to quantify % of the area that is occupied by the positive signal.

Good luck!

Dear Joanna Komorowska-Müller thanks a lot for your reply. I am attaching two images for the reference. In image 2 most of the staining on plasma membrane is reduced (it is continuous staining) and as can be judged from both images 1 and 2, the cytoplasmic staining is dotty. What would be the best way to quantify these two distributions? Should I do them independently as you explained in your answer?

Also for the cell area measurement, some of the cells in image 2 are circularised. Since there is no exact boundaries for these cells, do you think manual delineation is the best way?

Thanks again for your help.

Dear Teena,

I used: "Scion Image"

Is very-very useful.

Best regards

Peter

Teena Bhakuni Thank you for attaching the images. I have just looked at them. One thing I noticed is that when I opened your .tiff file is that the image was in RGB. For the analysis image has to be saved as 8-bit.

Your staining in general is a "dotty" staining. However, as you also noted, the density of the dots is very high in the plasma membrane so it makes it look more continous.

Unfortunately, if your goal is to compare cell membrane to cytoplasm, then you would need to use the same approach. In this case I would propose to binarize the image (keep the treshold constant for all the images) -> deliniate the compartments (save ROI's for future reference) -> use measure function to measure the % of area covered by the staining.

However, to do the deliniation properly, it would be helpfull to have a second channel in which you stain the membrane of the cells as a reference and do the delination of membrane and cytoplasm ROI's based on that channel. This way you can be sure that you are capturing the whole cell membrane.

Regarding cell area measurment: Yes. However, I would strongly propose additional staining of cell membrane to be more accurate. Moreover, by looking at the pictures I have an impression that you mostly do not have full cells within your field of view. Manual delination only makes sense if you can visually see the cell and its limits are within your image. Thus, I would think that for measuring cell area you should take images with smaller magnification.

PS: When you will be reporting the results, since there also seems to be quite some variability when it comes to the cell size, I would advice to additionaly report distrubution (histogram) of cell areas.

Good luck!

Dear Péter Degrell thank you for your useful suggestion. I will look into the software and see if it can help me.

Thanks again.

Dear Joanna Komorowska-Müller sorry for my late reply. Thanks a lot for your useful suggestions.

I have tried taking pictures with lower magnification. bEnd.3 cells are elongated and even at lower magnification there is no set boundary for these cells. The only difference with lower magnification is that number of cells are more in the field of view. I wanted to ask is it OK if I perform comparison of protein distribution (membrane vs. cytoplasm) with these images? You have mentioned in your previous reply that deliniation of compartments is needed before binarizing the images.

Thank you again.

Dear Teena Bhakuni,

If you visually cannot distinguish single&full cells, even with lower magnification, then I am afraid that you need additional staining that evenly stains the membrane to quantify cell surfaces (as I have mentioned above) so the boundaries between cells are clear. Also, nuclei staining would also help to distinguish single cells. PS: General rule: one can only quantify things that are visible.

Regarding the delineation of membrane and cytoplasm compartments - same rule. You can only measure/delineate if you see it, so I am afraid that lower magnification than what you showed above, would not allow that. Also, as already mentioned you have no marker for the membrane, so the delineation will not be very accurate. You would need additional staining of the membrane on top of your protein of interest to be accurate.

Furthermore, is there a way to seed the cells at lower density? This could greatly help your analysis as maybe then you can easier distinguish single cells.

Good luck!

Dear Joanna Komorowska-Müller ,

Thanks a lot for your reply. I understand the procedure for delineation now. bEnd.3 cells form a monolayer and so even if I seed the cells at lower density, analysis can be performed only once monolayer has been formed. But still I will try what you have suggested.

Thanks again for your help.