7-AAD COMPENSATION PROBLEM

Dear flowers,

 

Thank you very much for your replies. Here is a compilation of them:

 

 

 

Dear Flowers,

 

I am trying to perform 7-AAD staining for cell cycle with simultaneous
surface staining.

When I tried to compensate 7-AAD (linear, FL3) with PE (log, FL2),

 

 1) I wasn't able to see any event on an FL2 vs FL3 plot, while running
the PE sample and 

 

 2) When I removed the FL3 signal out of FL2 while running the 7-AAD
tube, I lost all the PE positivity, 

both in FL1 vs FL2 plot and my final sample (stained with FITC,PE and
7-AAD).

 

I also observed the same phenomenon with APC when I tried to perform
3-color analysis with 7-AAD.

 

Can someone please tell me what am I doing wrong? I've read the papers
by Rabinovitch (1986) and also Schmid (1991)

but they don't seem to worry at all about compensation problems...

 

Thanks in advance, 

 

Dr Ioannis Kotsianidis 
Clinical Research Fellow 



Dear Flower,

 

1. Did you used single-color tubes? If not, try to prepare a
single-color tube for each flurochrome you are using besides another
tube for the isotypes controls. 

2. On doing compensation with 7-AAD, try to use a log-scale instead of
linear. 

 

Greetings

 

Dr. M Elshal

Research Fellow

Flow Cytometry Core, 

NHLBI-NIH

Bethesda, MD

USA

 

Dear Ioannis,

 

I have attached our strategy for compensation. I recognize your problem,
which I also encountered the first time trying to measure FITC, PE and
7AAD stained samples. As an alternative you can use DRAQ5. However,
using DRAQ5 it is difficult to separate APC signals. Another alternative
is TO-PRO-3, which can be combined with FITC, PE and PE-Cy5. We use PI
in combination with FITC, PE and APC, but you need a highly expressed
protein for the FL2 channel in order to discriminate PE signals from PI.
Remember that the automated set-up using compensation beads is useless.
Turn all compensation settings to 0.0% and try to follow the guide I
have attached (PowerPoint).

 

Success,

Willem E. Corver, BSc, PhD
Researcher
Department of Pathology
Leiden University Medical Centre
P.O. Box 9600, Building 1, L1-Q
2300 RC  Leiden
The Netherlands
Tel.: +31 71 5266604
Fax.: +31 71 5248158

Dr. Ioannis Kotsianidis,

 

What you've described is quite normal. One way to avoid these
compensation troubles is to select a different DNA stain. It sounds like
you have at least a 2-laser system (488nm and 633nm) such as a
FACSCalibur. On a Calibur, you might try using TO-PRO-3 from Molecular
Probes, which is excited by the 633nm laser and can be measured by the
FL4 detector. It will work with FL1- and FL2-detected fluorochromes with
no FL4 compensation. If you have a cytometer with a UV laser, DAPI or
one of the Hoechst dyes could be another choice. Just be aware that at
high dye concentrations, these UV-excited dyes become slightly
488nm-excitable, and fluoresce into the "green" detector (FL1 with a
530/30 bandpass filter). However, the combination of DAPI/Hoechst, PE,
and APC will require no DAPI/Hoechst compensation.

 

Some assays that involve DNA content and another fluorochrome to be
measured do not necessarily require compensation. We typically run TUNEL
(TdT-FITC and PI) and intracellular cyclins (cyclin-FITC and PI) without
compensation. Instead of using "square" or "quadrant" regions/ markers,
we typically use a trapezoid or rhomboid shape. As long as the
populations are distinct, it does not matter if they do not "line up"
properly because no compensation was applied.

 

For simultaneous DNA content and surface antigen expression, you may
want to apply compensation. In this case it is generally more important
to compensate out the surface fluorochrome(s) from the DNA channel to
avoid "shifting" DNA peaks. The compensation technique is based on the
same principles as published on the web by Dr. Roederer, but the
mechanics are a little different.

 

1. Start with a sample labeled just with the DNA stain. (An unlabeled
sample is not as useful.) Zero all compensation settings. Set the PMT
voltage for the DNA parameter (in your case, FL3 for 7-AAD) until the
G0/G1 peak is at your target channel (we typically use channel 200 on a
1024-channel scale to allow measurement of tetraploid cells). On the DNA
histogram, make regions to select the G0/G1 population and the G2/M
population.

 

2. Continuing with the same sample, make 2 histograms for one of the
other parameters (FL1, for example). One histogram should be gated to
show only G0/G1 cells, and the other gated to show only G2/M cells.
Adjust the (FL1) PMT voltage until the G0/G1 population is at your
target channel (we typically use channel 128 on a 1024-channel scale,
with "channel" log data units). (Look at the median for the G0/G1-gated
FL1 histogram.) Now apply compensation (FL1-%FL3) until the G2/M
population is also at the target channel (look at the median for the
G2/M-gated FL1 histogram). It is likely that now the G0/G1 FL1 median
has decreased. If possible, keep increasing the compensation until both
medians are equal (but never more than 50%). If it is not possible to
make them equal, go back to the compensation setting at which the G2/M
FL1 median was at the target channel.

 

3. Repeat step 2 for any other parameters being measured (FL2, FL4).

 

4. At this point, there should be no (or very little) DNA dye
fluorescence in your other parameters: they should all look "negative",
as if you were looking at an unlabeled sample. Now you need to remove
the fluorescence from the other fluorochromes from your DNA parameter.
Make a second histogram of your DNA parameter. Remove any gates from the
histograms of your other parameters. Set markers on the two DNA
histograms around the G0/G1 peak position.

 

5. Look at a sample labeled with the DNA stain and one other
fluorochrome (FITC for FL1, for example). It is important to use
double-labeled controls, not the usual single-color controls. For the
histogram for the appropriate parameter (FL1), make regions around both
the "negative" and "positive" populations (or at the very least the
"dim" and "bright" populations). Gate one DNA histogram to show only
(FL1) "negative" cells, and gate the other DNA histogram to show only
(FL1) "positive" cells. Adjust the compensation (FL3-%FL1) until the
medians of the G0/G1 peaks on both DNA histograms are equal. (Do not
adjust any PMT voltages, even if the G0/G1 median falls below the target
channel now, or you will have to start over.)

 

6. Repeat step 5 for any other parameters being measured (FL2, FL4).

 

Well, that's how it should work in the general sense. I just noticed
that as given, this procedure won't work the the Calibur because of the
lack of direct FL1/FL3 compensation circuits. In that case, for step 2
start with FL2 (PE) and follow the directions as written, adjusting
FL2-%FL3. Now when you repeat step 2 for FL1 (FITC), adjust FL1-%FL2
(instead of the non-existant FL1-%FL3). Similarly, for step 5 start with
FL2 (PE) and follow the directions as written, adjusting FL3-%FL2. When
repeating step 5 for FL1 (FITC), adjust FL2-%FL1 (instead of the
non-existant FL3-%FL1).

 

Also for the Calibur, although you could run FITC with 7-AAD (without
PE), you may find it difficult to set "FL3-%FL1" (because this implies
setting both FL3-%FL2 and FL2-%FL1, but you won't be able to set
FL3-%FL2). For "dim" FITC signals, it may not be a problem to have
FL3-%FL2 and FL2-%FL1 be zero, but for "bright" FITC signals -- like
AnnexinV-FITC -- you may have problems. In this case you may be able to
"fudge" the compensation by setting either FL3-%FL2 or FL2-%FL1 to a
"typical" value, and then adjust the other compensation as above. (Note
that 7-AAD overlap into FL1 is essentially zero and FITC overlap into
FL3 should be minimal, so most people would run the combination of FITC
and 7-AAD uncompensated.)

 

There should be no other special issues with other combinations of FITC,
PE, and/or APC with 7-AAD for the Calibur.

 

As a side note, I just want to point out that 7-AAD is perhaps the worst
DNA stain for DNA content/cell cycle analysis. Of all the DNA dyes,
7-AAD produces the largest %CVs (widest peaks) which can make
deconvolution questionable or even impossible. If you're able to get CVs
of around 5% with 7-AAD (or you are satisfied with the analysis), then
that should be acceptable. Otherwise, DAPI is perhaps the best DNA stain
for this assay, but not many people have access to a cytometer with UV
excitation. It is difficult to use the old standby, PI, with more than
just FITC. As I pointed out at the beginning of this message, TO-PRO-3
is an alternative when you have 633nm excitation. Although there are a
plethora of other DNA stains, I would try to avoid something like SYTOX
Green for your assay (but it is a good dye for other assays). The reason
is that the fluorescence from DNA dyes is often quite a bit brighter
than from, say, CD3-PE surface staining. That is, it will be more
difficult to remove SYTOX Green overlap from the PE detector than FITC
overlap from the PE detector. As a general rule of thumb, the DNA dye
should either have a large Stokes shift (like 7-AAD) to allow other dyes
with smaller Stokes shifts (FITC, PE) to be excited by the same laser,
or to "isolate" the DNA dye as the only fluorochrome excited by one of
the lasers (like DAPI [UV] or TO-PRO-3 [633nm]).

 

Good luck,

Eric

Hi Ioannis, compensation with very bright labelling can be tricky. It is
possible, but as you have already been advised you must use single
tubes. I found it most useful to gate on a population of PE+ve cells,
and look at those in an FL3 vs FL2 plot. Try adjusting these out of the
7-AAD first. Then look at your 7AAD+ve cells, increase compensation
until the FL2 is clear. Now look at your double stained population...it
should look OK. The other thing which may help to sort out your
populations is to exclude apoptotic cells (sub-G1 fragments) as this can
make your analysis cleaner (assuming you are not interested in these
cells). The compensation values should still hold for the apoptotic
population (under/over compensation is most noticible on the brightest

cells- see the following site, its worth going through the tutorial,
most

helpful) 

 

http://www.drmr.com/compensation/index.html 

 

 

I've attached a pic of some of the gating (but I'm afraid I no longer
have the bivariate plots of FL3 vs FL2) I used to look at CD34-PE and
7AAD, I hope it helps. You should also note that the 7AAD may go into
your FL1 a little too, on the machine I used there was no option to
compensate between FL3 &FL1.

 

regards, Sian.

 

 

Dr.Sian Rizzo

Prostate Stem Cell Laboratory

Male Urological Cancer Research Centre

Institute of Cancer Research

15 Cotswold Road

Sutton

Surrey SM2 5NG

Hi there

 

Are you gating on 7AAD positive or negative cells in your analysis?  I
am using 7AAD as an exclusion dye together with surface staining (rather
than cell cycle analysis), but hopefully the principles are the same.

 

1)  I use single colour controls for fitc and pe, and individually
remove excess FL1 and FL2 fluorescence from the FL3 channel.

 

2)  I do not remove FL3 out of FL2.  The reason is because I am using
7AAD as an exclusion dye, ie. in my analysis I gate on the 7AAD negative
population.  Since 7AAD negative cells have no fluorescence they will
not contribute excess fluorescence into other channels (and therefore do
not need to be compensated).

 

I hope this helps.

 

Lydene McArthur

Haematology Surface Markers department

Canterbury Health Laboratories

Christchurch 

New Zealand