Cells, if you have them, are usually the best compensation controls, as long as they are staining brighter than your samples.  For example, I use anti-CD3 stimulated cells stained with single colors for markers such as PD-1, Fox-P3, or Ki67.  However, some markers downregulate upon activation, like CD127, and often CD3 is well (this will depend on whether or not your CD3 stain is intracellular or surface).  However, if cells are a limited resource, and you need to use compensation beads, you should titrate down the the antibody on the beads so the MFI of the positive population on your beads matches that of the positive population on your single color stained cells.  PMT voltages should not be changed to accommodate brightly staining beads.

I think you want a really clear distinction and brightness of your compensation beads, so that you can compensate quite more efficiently. Later the cells will not be so bright because obviously the antigen-fluorochrome conentration on the cell surface is much lower that on the bead surface, but having a good compensation will help getting rid of any doubt about the separation between the cell populations even with lower signals.

Thanks Davide,

I appreciate your answer a lot.

You are right, the fluorochrome concentration on my cells was much lower and that's why I did not lower the voltage as much as I did for beads. And still the data I receive when compensating with cells are much better than with beads. This is what confuses me. I have the impression the beads were too bright and I had to use a voltage with diminishes the detection of my actual samples. Could this be?

Well, if I didn´t get this wrong, the voltage to which you refer to is meant as a treshold, intended to cut off from your measurement all the aspecific and background signals. So, try to increase the voltage and see if there´s any difference. I suggest you to test the voltage first on your cells just to be sure that after setting up the right voltage your cells don´t disappear from the plot, and then have a proper compensation with the microbeads.

Thanks again Davide for taking the time to answer and try to help! You were the only one who bothered ;)

I guess I will just stick to my cells as a compensation control, at least for this specific colour. Because too low voltages when using beads just squeeze my populations into a too narrow area.

That could not be a problem, you just want to compensate the colours and not to quantify your beads. Anyway, compensating with your cells is not much of a problem if you know how to do it. Good luck!

Hi Sven,

Let me give you a run down on voltages and compensation.

COMPENSATION

The essential requirement for successful compensation is to have the same fluorophore labelling an object (cells/beads) with a signal that is as bright, or brighter than your sample. The reason that beads are often better, is that some cellular antigens which have very low density (i.e. they are expressing at low levels), and so they don't provide a good enough signal. COMP beads capture antibody at high levels, and so provide a strong signal which is ideal for COMP.

VOLTAGES

Here is what happens as voltages increase: at low voltage the negative and low-end signal will be taken over by electronic noise, and low-end signal (dim signal) collapses into the negative population. As voltage increases, a point is reached where true signal is detected above electronic noise, and the relationship between negative-dim-bright is linear (or proportional). Voltages that are too high mean that the highest signal goes off scale.

What is probably happening in your case, is that your PE-CF594 signal was really bright, and so the voltage had to be reduced to keep it on scale. As a result, the voltage was low enough that your dim (CD56) population collapsed into the negative.

I would like to know, when you used your PE-CF594 COMP beads, where (i.e. 1x10^5, etc) did the positive peak end up (i.e. in the graph where your CD56 pop is very low)? This will be helpful in diagnosing your specific problem.

- Tom

Probably your initial PMT voltages are too high. First, you should use the min PMT voltage for each PMT/filter combination in order to use the linear measuring area of the PMTs. At that point your compensation values determined by beads are usable as well for your sample. Here is a reference:

"Establishing Optimum Baseline PMT Gains to Maximize Resolution

on BD Biosciences Digital Flow Cytometers" by Joe Trotter.

Cells, if you have them, are usually the best compensation controls, as long as they are staining brighter than your samples.  For example, I use anti-CD3 stimulated cells stained with single colors for markers such as PD-1, Fox-P3, or Ki67.  However, some markers downregulate upon activation, like CD127, and often CD3 is well (this will depend on whether or not your CD3 stain is intracellular or surface).  However, if cells are a limited resource, and you need to use compensation beads, you should titrate down the the antibody on the beads so the MFI of the positive population on your beads matches that of the positive population on your single color stained cells.  PMT voltages should not be changed to accommodate brightly staining beads.

Hi Arielle,

This is exactly what I thought, my beads were just too bright so that I had to accommodate the voltage in order to fit the positive population into the graph. This in turn diminished the separation of my populations, right?! Titrating down the staining for the beads is probably a good idea, didn't think of this yet.

Thank you for your answer!

Just remember it is the sample that is important. In an initial set-up single colour cell controls are run, compensated and the impact on the sample is reviewed, settings may need optimizing etc. Once this has been done to satisfaction the beads are then run on the settings that you judge to be optimal for your SAMPLE. If beads go offscale you could try adding less antibody. The beads are a control that help to maintain consistency and can help to detect issues with antibodies etc.

Thank you Ann for your contribution. So by doing the initial set-up with single stained cells and then the compensation with beads, you get the best out of both. That's a very nice way of doing it, involves a little bit more work and cell material though.

Great response, Arielle. I would just add that you are compensating the fluorophores, not the antigen. So, you can use another antibody to a more prevalent antigen, e.g., CD3, on cells to do your compensation with cells, if the antigen for your original experiment is low. It may help to titrate down the staining for your beads, as the beads tend to have a much higher affinity for the antibodies than cells. You can compensate with a combination of beads and cells (just adjust your forward and side scatter, but keep your other channels the same). I have these issues sometimes, and the issues vary across panels/experiments.

YW

Thank you Yohann!

Hi Petersen,

I concur with Arielle, Yohann and others who have indicated that you titrate the antibody on your beads to get it on scale and not increase or decrease voltage because ideally voltage should be set up with your sample. I have also just been watching Tim Bushnell videos through his Excyte Mastery class and found them very useful in optimising flow cytometry assays. Especially on the importance of getting your compensation right which is fundamental to all multicolour flow cytometry assays. Perhaps you might want to consider signing up. 

It should be noted that balance between channels is as important as brightness--if not more-so. As long as one can discriminate between populations, it is not necessary for each marker to be as bright as possible. This, in fact, leads to unnecessary spillover and ultimately a matrix that can no longer handle the n-dimensional interactions of the combined cross-signals. Leave a bit of room at the top, in other words.