Hi Tommy, AFAIK your reasoning is correct: there shouldn't be any antigen specific effects. Especially when you're staying within the range of brightness that you observe in your target cells.

Some obvious points that you probably thought of already

- Have you looked at your compensation matrix to see if the values are plausible?

- Are the PMT voltages the same as when you made the compensation matrix? Same goes for the laser powers.

- Is the device calibrated correctly (e.g. CST in BD Cantos) ?

- Are the cells autofluorescent (e.g. cultured)

When you use cells for compensation antigens may indeed have an influence. Are you using the brightest condition for each fluorescence during compensation? I mean if you are working with a marker that is increased upon stimulation, than your stimulated cell should be the sample used for compensation. But that shouldn't be a problem when you use beads.

I've never used eFluor780 but I read that it's excited by the red laser and detected by the 780/60 band filter. So I believe it might have some spill on APC (and vice versa). The same as for FITC/PE (but considering blue laser excitation).

What comes to mind is:

Haha, hey, good timing.

My hat off to Henddrik and Andre for your answers.

To answer Hendrick's questions:

1. I did look at the compensation matrix and the values just look fine.

2. The matrix does come with suggested values but the voltages are not suitable for my experiments.

3.I haven't checked the calibration.

4. My cells are autoflourescent near FITC channel.

To Andre:

1.The instrument is BD FACS CANTO II. 

2.There is suggested values but they are all too high for my antigens.

3.I use exactly the same fluorochromes in both compensation and experiments.

4.I am doing visual compensation instead of MFI.

5.I do use FMO for my gating control.

Hi Tommy,

Looks like you did a lot of thinking on this already. I don't get your reply point 2 (both to mine and Andre's points). How are the compensation values too high for your antigens, but the matrix still look good? Do you mean the PMT values are too high?

And if you're doing visual compensation, well you can remove almost any bleed-though you want, if you don't mind massively overcompensating :-). Also: 6 fluorophors by hand is a ton of work, are you sure you don't want go automatic?

Could you show your compensation matrix? Usually they are similar across devices. I think you should see FITC->PE of like 10%,  and PE->fitc of ~15%. APC to efluor780 of maybe 8% or so, I'd have to look it up. The rest

Ha, I dug up an old experiment with almost the same fluorophores (AF700 instead of eF 780), maybe you can look at this as reference. FITC - PE bleedthrough was less than I thought, but I used super strong antigens here.

I agree with Hendrik. Manual compensation can be very tricky and would be easier if you started with suggested compensation values and made some minor adjustments. Manually, you have to check MFI because visual can lead to under or overcompensation. As he said, have you matched antigen density with fluorochrome intensity (for exemple PE for lower expressed antigens and FITC to higher expressed ones)?

Hendrick (if you read this). Have you worked with this eFluor780? I've never used it and I don't know if it bleeds into APC since they are both red laser excited and their detection filters are not as far apart. 

Hi Andre, no I haven't used it either, so I can only infer from similar fluorochromes. I've only seen it as tandem with APC as it hard to excite with the 633 laser. But the peak emission is at 780 nm, whereas for APC the peak is ~660nm, so I think the overlap should be manageable. It depends a little on the exact filters in the canto II; usually it's 660/20 and 780/60, then it *should* be fine, I think

I agree with you. I tried BD Spectrum Viewer but they don't have this fluorochrome in their chart. I don't know if Tommy is using the APC conjugated form but if he is then tandem loss is also a possibility right? I used a APCCy7 once that was a bit old and it was a living hell to compensate (APC, APCCy7). People from BD told me old APCCy7 can lose some "Cy7" and that stands in the way for proper compensation.

Hi Tommy,

did i get the Point, that all your 5 Panels have differnt PMT Voltages correctly?

If yes, you can not use 1 comp Matrix for all Panels.

best

André

Sorry for the late reply.

To adress the question both Andre and Hendrik had on point 2, here is the answer.

When I said the voltage was too high it actually meant the preset voltages by the FACS software were too high for each of the channels. I had to instead manually compensate by tweaking the voltage in each channel and defining the positive peak.

In my compensation matrix, the values for voltages are consistant through out my compensation matrix, regardless to which colour I was compensating for.

 In addition the elfour780 is an quivalent to equivalent to APC-eFluor® 780 or APC-Alexa Fluor® 750. It is conjugated to the fixable viability dye from eBioscience, a dye that would allow me to discriminate the dead population during flow experiment.

Regards,

Tommy

Hi Tommy,

Lets make an easy example to understand your Problem:

Does your PMTVs in your compensation Matrix look like this?

PMTV FITC Channel Panel 1 = 500V

PMTV FITC Channel Panel 2 = 500V

PMTV FITC Channel Panel 3 = 500V

PMTV PE Channel Panel 1 = 500V

PMTV PE Channel Panel 2 = 500V

PMTV PE Channel Panel 3 = 500V

or more like this:

PMTV FITC Channel Panel 1 = 500V

PMTV FITC Channel Panel 2 = 400V

PMTV FITC Channel Panel 3 =300V

PMTV PE Channel Panel 1 = 500V

PMTV PE Channel Panel 2 = 400V

PMTV PE Channel Panel 3 = 300V

2 additional Infos:

The PMTS Voltages of your Compensation has to be the same in your Experiment. Only then the compansation triangle is valid.

A dead cell marker wich is gated out before further analyes need not be compensated.  

Best

André

Hi Tommy,

Adjusting the PMT values by hand is usually no problem, you might not be in the super optimum performance range of the laser and have larger CVs, but the effects are minimal.

From your answer I suspect you might be confusing pmt settings and compensation, so sorry if the following seems pedantic.

Adjusting the PMT values cannot be used for "compensation" of spectral overlap. The emitted light signal remains the same regardless of sensor setting. Adjusting laser power could be used to minimize spectral overlap a little, but you can't do that on a Canto II. Alternatively you can optimize staining conditions, but that is also limited.

In other words you have to digitally/mathematically compensate, I'm not sure you are actually doing this (which is the actual meaning of the term compensation). In compensation you subtract a percentage of the signal measured in one channel from the signal measured in another channel, as you assume this fraction is due to spectral overlap. So it's a mathematical correction of you actually measured values.

Again for or clarity there are two matrices: the PMT settings list and the compensation matrix :-) The latter is for a % or fraction signal subtraction of the measured values, it should be a 8x8 matrix of values between 0 and 1 on your system.

Are you sure you have been adjusting that one?

- Hendrik

P.S. As it's a mathematical thing, compensation can also be done after the measurement, esp. if you do it by hand. So if you didn't do it correctly at first you can still use your data without any problems.

Hi Hendrik,

No I haven't touch the compensation matrix setting at all. All I did was to adjust PMT settings. Thanks for clarification.

-Tommy

Hi André,

My PMTV setting is similar to your first example with constant setting across each panel.

Regards,

Tommy