Re: PI staining w/ PBMC

William George Telford (
Wed, 11 Jun 1997 21:44:27 -0400 (EDT)

Hi Ryan...

We've had the same problem as you when attempting to analyze apoptosis in
human cells using ethanol fixation and staining with PI or other DNA
binding agents, namely the accumulation of an apoptotic peak with a very
low fluorescence intensity. It is usually impossible to distinguish
apoptotic cells from debris using a single parameter linear scale; log
scaling is often only somewhat successful for distinguishing apoptotic
cells from debris. It appears that the amount of DNA extruded by
apoptotic cells is quite variable from cell type to cell type; this
appears to be particularly true for human cells, and is undoubtedly
related to the type of DNA damage experienced during apoptotic death. We
generally limit our use this technique in mouse lymphoid
tissues, since they give a nice, clean apoptotic peak at an index of about
0.5 of the diploid G1 peak.

If you really want to use single DNA dye staining to look at apoptosis,
analyzing on a log scale for fluorescence with forward scatter as a second
parameter is probably the best way to go. Apoptotic cells will show
somewhat reduced FSC signal but will be distinguishable from debris (which
should be quite small). Also be aware that some apoptotic stimuli can
cause extensive cell blebbing, which will reduce your apparent dead cell
yield (since these blebs will likely appear as debris). Most importantly,
make sure that the population you think is apoptotic actually is - if
dying cells constitute the majority of objects in your culture, simple
examination under a fluorescence or confocal microscope (or EM) may be
sufficient. If it is a smaller number, it would probably be best to sort
the cells and then check by microscopy (I frankly think everyone who does
any sort of flow assay for apoptosis should do this at least once as an
essential control). You should also probably do another type of cell
death assay in parallel (at least a couple of times) to confirm the
presence of cell death (such as two-dye exclusion, TUNEL, FITC-annexin,
etc.). If you have continued difficulties with single-dye methods, it
might be best to switch - they are being used less as of late since they
are not as good at distinguishing dead cells from debris in some cell
types, as well as the variable position/appearance of the dead cell

Also check Dr. Darzynkiewicz's protocol for single-dye staining using
adjusted phosphate-citrate buffers prior to staining - this technique
attempts to control the amount of DNA extrusion and thus the position of
the apoptotic peak.

Hope this helps. If you want, FAX me a copy of your data and I can
compare it to what we've seen for apoptotic cells.


Bill Telford
Hospital for Special Surgery/
Cornell University Medical College
Flow Cytometry Core Facility
New York, NY 10021

phone: (212) 606-1925
FAX: (212) 717-1192

On Mon, 9 Jun 1997, Ryan Hung wrote:

> I've just started using FACS analysis for analyzing apoptosis in PBMC a
> couple of months ago. I figured that I would first try the inexpensive
> technique of PI staining. However, to date, having tried both direct
> staining with PI (with sodium citrate, Triton X-100, RNAase A) and
> ethanol-fixation prior to PI staining, I have not been able to get either
> to work reliably.
> My model system itself consists of PBMC extracted from healthy human
> donors at a concentration of 3x10^6 cells/ml in 1ml of RPMI growth medium
> in 24 well plates. I'm using camptothecin in a concentration of 5 ug/ml
> as my positive control of apoptosis. I incubate for between 14-18h at
> 37oC in 5% CO2.
> With direct staining with PI, I've typically obtained two closely-spaced
> high fluorescent intensity peaks when reading FL3 on a histogram plot. I
> see virtually no lower fluorescent intensity peaks.
> With ethanol fixation, I've had problems with gating when using a linear
> scale on FSC, so I've tried using a log scale. I understand from previous
> discussion that this may be incorporating significant amounts of cellular
> debris. What I observe in terms of histograms are a single
> high-fluorescence peak, with a lower fluorescent peak at about 1/5 the
> intensity of the high peak, rather than the expected 1/2.
> Since my results from both techniques don't seem to correspond well with
> what I've found in literature, I've been finding it difficult to analyze.
> Any ideas or suggestions would be much appreciated.
> Ryan.
> _/ \__/ \__/ \__/ \__/ \__/ \__/ \__/rhung@vcn.bc.ca__/ \__/ \__/
> \_Apoptosis=programmed cell death/ \__/ \ \__/ \__
> _/ --you can't live without it!/ \__/ \
> \__/ \__/ \__/ \__/ \__/ \__/ \__/ \My words Copyright (C) 1997 \__

Home Page Table of Contents Sponsors E-Mail Archive Web Sites

CD-ROM Vol 3 was produced by Monica M. Shively and other staff at the Purdue University Cytometry Laboratories and distributed free of charge as an educational service to the cytometry community. If you have any comments please direct them to Dr. J. Paul Robinson, Professor & Director, PUCL, Purdue University, West Lafayette, IN 47907. Phone: (765)-494-0757; FAX(765) 494-0517; Web , EMAIL