I am using cells that seem to have high autofluorescence. I've read that I can use one of the detectors free of staining with any antibodies to measure the amount of autofluorescence directly. My panel so far is PerCP, PE, APC, PE-Cy7,and APC-Cy7. So, I can add FITC to measure staining for autofluorescence.
What I don't understand is how to do that in terms of negative and positive population between this sample and my unstained sample (I mean, essentially they are the same sample, and they would have the same fluorescence intensity).
I've been reading the papers by Roederer, they are intense! I don't mind reading them, they are really interesting, I just get really lost.
It's my first time setting up the comp matrix for the multicolor flow and I still have a lot to learn.
I've been reading that people widely use the extra detector to eliminate autofluorescence, if somebody knows how to do this, please help me.
Hi Aleks,
What cells are you using? For example, FACS analysis of total spleen sample gives you background in every channel. To my knowledge, this is the case for any type of background, not caused by antibodies.
The best way to discriminate between real stain and background is taking control samples for each antibody individually, and the so called FMO (fluorescence minus one), thus for each channel take all channels, except the FMO (eg. leave out FITC, for FMO of FITC, but take the rest, and so on). This will give you "many" control samples, but it will provide you the real positive cells per sample, no matter the background.
Good luck, hope it helps,
Rogier
You need to include controls for each individual antibody as well as unstained cells. This will help you to determine the correct compensation and also determine how much autofluorescence your cells have.
Dear Anne,
Thank you for your response. Yes I do the single stained controls and the unstained samples. I also do the FMO controls.
The problem I have is that I work with "aged" brain cells and they have a lot of auto fluorescence and since some of my single stained controls from cells are not very bright the software compensation undercompensates some parameters and overcompensates others. I manually adjusted it but I dont know if I did it correctly. I was hoping to get rid of autofluorescence either by subtracting it or by measuring it in an extra detector. Either way I dont know how to do either
Dear Aleks,
Have you considered running your compensation controls on beads instead of cells?
As a rule, one wants as much signal as possible from your single-stained compensation controls, and beads (coated with anti-mouse-antibodies) will usually bind your antibody quite strongly (particularly useful if your antigen is weakly expressed on cells). You can buy such compensation-beads from several manufacturers, and they are fairly cheap - just look for "compensation beads".
Obviously, beads won't solve any auto-fluorescence issues, but they will certainly help you get your compensation matrix right. You could then look at FMO-controls and isotype-controls (if you can find those!) to discriminate between autofluorescence, non-specific binding, and "true" signal.
There may be some markers that are highly expressed on your aging cells, and others that are weakly expressed. Since FITC is less bright than something like PE or APC, you could play around with using FITC for all the markers you are looking at, and seeing if it works better for some markers than others, and do this with all the stains you are using and all the markers you are testing for. So for example, if you are looking at CD45+ cells, stain them FITC-CD45, PE-CD45, APC-CD45, etc. and see if there is a change in expression. Any marker that may be weakly expressed should be stained with something like FITC, other strongly expressed markers can use PE or APC.
PE-Cy7 and APC-Cy7 are also conjugated antibodies, so you may have issues with the Cy7 seperating from the PE and APC, so you may want to compensate for this and definitely use comp-beads for these and your other controls.
Good luck!
You may want to include single stains of each antibody as well as run a unstained control.
I’m a bit confused about the question. If your cells of interest are high in autofluorescence (e.g. macrophages) then eliminating the autofluorescence cells (which would be in a diagonal 45degree line extending from your negative cells in an empty FITC or V500 channel) by gating would also exclude your cells of interest. For that case it would be a question how to distinguish cells positive or negative for marker X. This has been addressed by other answers (I’m personally more critical about FMOs, I think biological controls are always preferable if feasible).
In case you talk about a contaminating autofluorescence population in your cell suspension, then yes you can clean up your gating by excluding the FITC+/V500+ cells on the diagonal (in an empty channel).
This all assuming that you already included a live/death marker to exclude dead cells. This step would be more critical.
You should try to run your analysis on a spectral cytometer.(SP6800 from Sony).
Autofluorescence can be stored as one separate chanel and substracted to all other measured parameters. Practically it allows you to set out of the game autofluorescence for your measurements so that the final signal you get is only due to the expression of your marker of interest. As signal is sotred as one autofluorescent parameter, information is not lost and you can get specifically on autofluorescence positive cells if needed
- Badges
- Science method