Hello Yang Zhang

If permeabilization step is omitted before antibody staining (only possible by fixation with chemical crosslinkers), you can specifically mark extracellular plasma membrane-bound antigens in order to distinguish them from the intracellular antigen pool.

For instance, when you use aldehyde-based fixation methods, permeabilization is performed as a second step and this two-step fixation-permeabilization process, such as paraformaldehyde fixation followed by methanol permeabilization is effective in allowing access to intracellular molecules.

So, you should use aldehyde-based fixation method to fix the cells and omit the permeabilization step. Avoid using organic solvent like methanol to fix the cells as it can both fix and permeabilize the cells.

Best.

You should stain with PFA without detergent (i.e., Tween or Triton X). Make sure to block in regular PBS without Tween and this would be staining the surface membrane. You should also stain cells with methanol fixation or PFA with Triton X as well to detect the cytoplasmic and compare the expression. You do not need to omit the Tween in this case.

I would analyse the isolate PM portion of the image you have and quantfy that. if you use thesholding in imageJ to segement the cells and make a mask, you can duplicate this mask, erode it 5-10 times and then use image caculator to subract out the small mask from the big mask. this will leave a mask that is just a ring around the cell. use the math function to divide this ring mask by 255 (the whole image is no 0 or 1 where the PM ring is. then multiply the channel you want to analyse with this ring mask using image caclulator. The result will be an image that only contains the PM. you can then analyse these images as you normally would for flourescence intencity.