I have a set of cells double stained in this manner: first the cells has been live-stained to visualize the protein of interest at the membrane (named: SURFACE). Then, cells have been fixed and permeabilized. A second staining was then performed for the same protein of interest (named: TOTAL).
The two fluorophores have different colors (green and red). So far I've analyzed them by normalizing the SURFACE mean intensity to the TOTAL mean intensity (SURFACE/TOTAL), but not fully happy with this way of analysis and data representation...
Does anyone knows a better way to correlate the two signals?
(PS: the main problem is that the two signals have different intensity. Red is generally more faint than green, and even if it should mark the TOTAL protein the ratio between the two is always around 1 for control...)
Im curious as to why you don't stain the cells post fixation with both sets of markers. A live cell will indeed react differently to fixed cells. If you did a serial staining procedure you can still image total protein and surface:
1) block the cells and stain with your surface marker and it's corresponding secondary. (no permeabilization)
2) Permeablization step and stain other primary and secondary.
What antibodies are you using? Different antibodies will have different levels of effectiveness in staining (as I am sure you are aware) and could cause a discrepancy in your results. If they are the same with a different fluorophore attached, you may be able to correlate better. There are lots of kits that allow you to conjugate antibodies. You could then compare intensities of staining for surface and total with each color to normalize and make a comparison. What is the colocalization signal like? Could you quantify the intensity of the yellow?
Good luck!
Hi,
I understand that the two dyes have different intensities: there are many factors affect observed intensity.
Have you thought of staining (2 aliquots of) fixed cells (=TOTAL) with each colour separately? Then you would have some idea how intensities of these two colours compare. For example, you may find like "intensity value of 100 for red" corresponds to "intensity value of 500 for green".
Im curious as to why you don't stain the cells post fixation with both sets of markers. A live cell will indeed react differently to fixed cells. If you did a serial staining procedure you can still image total protein and surface:
1) block the cells and stain with your surface marker and it's corresponding secondary. (no permeabilization)
2) Permeablization step and stain other primary and secondary.
What antibodies are you using? Different antibodies will have different levels of effectiveness in staining (as I am sure you are aware) and could cause a discrepancy in your results. If they are the same with a different fluorophore attached, you may be able to correlate better. There are lots of kits that allow you to conjugate antibodies. You could then compare intensities of staining for surface and total with each color to normalize and make a comparison. What is the colocalization signal like? Could you quantify the intensity of the yellow?
Good luck!
Hi Dario, you could stain using the first fluorophore then quench it and restain with the same or a different fluorophore.
Article Multi-Phenotypic subtyping of circulating tumor cells using ...
Try to get two fluorophores with similar intesity and make sure the surface one is not affected by your fix/perm step. Also, I would use the median fluorescent intensity not the mean as most distributions are not equal.
If I understand you correctly, you are bothered by the numerical value of your surface/total intensity ratio. Unless you relate the fluorescence intensities to actual numbers of proteins, the surface/total ratio, if only an intensity ratio, and not a ratio of numbers of actual proteins, will carry the same information if it has for the control a numerical value of 100, 1, or 0.01. I think it sounds great that it is about one for control or larger than one for the control, because then no one will err to think that it in any way reflects the ratio of actual number of proteins on surface and in the cell. I would suggest that in fact you normalize your surface/total signal to value = one for your control cell population. Then value two for your normalized surface/total ratio will reflect twice as large a fraction on surface, and 0.5 half as large a fraction on surface.
Note that Karin G. Hermans probably meant that you should use median value of your surface/total intensity ratio for a cell population. It is of course meaningless to calculate (median voxel surface intensity)/(median voxel total intensity) for a given cell, since because of surface segregation and intracellular segration (i.e. if most surface voxels a this could have any value from 0 to infinity for almost any possible surface/intracellular distribution (if the surface intensity = 0 everywhere, then you can get only ratio 0 or undefined ratio, if the median total intensity = 0 because of most pixels being dark, i.e. having intensity 0).
This leads to one beauty problem in your analysis. it would be better to use the summed or integrated intensity of surface and total fluorescence for each cell. Of course if you define your cell to include the same set of voxels/pixels, and then calculate the mean voxel/pixel intensity for the surface signal and for the total signal, then for X number of voxels/pixels and with the integrated intensity S for the surface stain and integrated intensity T for the total stain the ratio on mean intensities (S/X)/(T/X) = S/T, but you can dismiss the means and just calculate S/T for each cell. Then it is easy to compare your cell populations, using either the raw data, or after normalizing the average (or median) S/T to 1 for the control population.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology