[Cytometry] Summary of advice on intracellular flow cytometry
anonwums1 at gmail.com
Wed Jan 13 12:13:37 EST 2010
Several people have asked me to summarize my conclusions on how to get
intracellular flow to work. So here you go.
*People say that intracellular flow cytometry isn't that hard but there are
a lot of variables to work out. There are several good books written on the
topic such as *Current Protocols in Immunology* or *Flow Cytometry Protocols
*, although no single protocol is comprehensive and often glosses over some
points stressed in other protocols.
Monoclonal antibodies are best but polyclonals can work. Generally you want
an antibody generated in a host that isn't your target tissue (if you're
staining mouse tissues, you don't want a mouse monoclonal), but you can make
due with mouse-on-mouse issues by directly labeling your antibodies (or use
Zenon kit) and diligent use of Fc block.
Direct detection using a Zenon kit or directly conjugated antibodies is
best, but indirect methods can work. Indirect titering is complicated, but
you titer the primary using a fixed secondary concentration and once you
have the primary titer, you fix it, and then titer the secondary.
*PE and APC are big molecules, but they also are bright. Some people want to
shy away from these dyes and go for smaller AlexaFluor conjugates because
they feel that bigger molecules have difficulties getting into the cells.
However, some people note that even if bigger molecules don't get into cells
very well, they make up for that by being so much brighter.
4% paraformaldehyde is best, and it seems to maintain scatter properties of
cells decently. However but it's a pain because it's toxic and you have to
make it up fresh (or make it fresh and store it at -20C before use). Some
people use diluted formaldehyde, which is much easier to work with but often
has methanol as a preservative, and methanol sometimes screws up antigens
and even at dilute concentrations, may permeablize cells.
Less fixative is often better. Some people suggest using 2% or even 1% PFA.
This is where I learned the most. There seem to be three common
Saponin seems to be really popular. It is a reversible agent, so you need it
in every perm, staining, or wash buffer. But you can wash it out in the end.
You generally want to use 0.1% to 0.5% saponin, although at the lower end,
leftover volume after washes may dilute it below 0.1% so you may want to err
on the higher end of the spectrum. You don't need long to permeablize, only
5 mins or so at room temperature.
However, saponin isn't a single chemical. It's actually a mixture of
different molecules, although the active ingredient seems to be saponingen.
Generally, you want > 25% saponingen. I called a couple of vendors to find
out their saponingen content, and they either didn't know, or gave me a very
large range (10% - 50%). Apparently, the content varies from lot to lot and
you have to revalidate each lot you get for your assay.
As a consequence, several people suggested that you don't try to use
homemade saponin solutions and buy commercially available ones which lets
you outsource the QC to the companies. People have suggested BD Perm/Wash
buffer, BeckmanCoulter Intra-Prep, and Caltag.
Use 0.1% to 0.5% TritonX-100. Unlike saponin, TritonX-100 is irreversible
and seems to make larger pores. So you perm once and then don't worry about
it. The downside is that it is probably a bit harsher. Although it's
supposed to be soluble in most aqueous buffers, based on my personal
experience, it's often very difficult to dissolve.
Using ice-cold methanol also permeablizes. This doesn't seem to be very
There are different blocking agents too
- 5% milk - one person said this is the best, although it doesn't work
for phosphospecific antibodies
- 0.1% to 0.5% BSA
- 2% - 10% FCS
- 300 ug / mL of IgG of the primary antibody host, which should also
block Fc receptors
PBS + 10% FCS seems to work. If you use a saponin containing buffer, you
need to use that.
Titering intracellular antibodies is exactly the same as for extracellular
antibodies, except often the leeway you have is much less. You usually titer
on 1 - 5 x 10^6 cells in 100 uL of staining buffer but many people argue
persuasively that it's really the concentration of antibody per volume that
matters until you get to > 50 million cells. For titration, it's best to
avoid this argument and pick a cell concentration and fix it for all
titrations. Then titer the antibody using serial dilutions. In the end, you
will have both a per cell and per volume antibody concentration. Once you're
done titering and get to your experimental conditions, you can vary either
the cell concentration or the antibody concentration a little to see if it
Start at the manufacturers suggested concentration and do two-fold titers
down from there. Titers should be in the ballpark of 5 ug / mL to 0.5 ug /
*Washes / Spins*
Any washes with a saponin based protocol need saponin in them. Don't spin
your cells too fast once they're fixed. People seem to recommend 400 - 800
g. Washes are a major source of cell loss. Someone suggested you stain in 96
well plates and then dump off your liquid to avoid using any aspirator
during serial washes. Someone else suggested doing one big wash in a large
volume (5 mL) and then spin only once.
Sadly, there seems to be no consensus on how to do that. Some people use
isotypes, but sometimes the isotypes are sticky, which is exacerbated by
fixation and you end up with too much background. Other people use a FMO
strategy. Some people incubate with purified antigen of interest to
neutralize the antibody. Still others use two different nonreacting
monoclonal antibodies to the same antigen at once and show double staining
Well, that's it. If anyone has anything else to add, I'd love to hear it.
More information about the Cytometry