Summary Document
This document is a summary of message from the Purdue University Cytometry Laboratories Cytometry Discussion list.
http://www.cyto.purdue.edu/hmarchiv/index.htm
Thank you for your replies. Here are the answers I received concerning receptor
internalisation analysis
Julie Bertout
cytometry lab
Institut Pasteur de Lille
1 rue du professeur Calmette
59800 Lille
France
my question :
I have a researcher interested in quantitating receptor internalisation
after different treatments.
Here is the way she used to do it :
- label receptors of interest with primary antibody and secondary antibody
- do the treatment she wants to test
- remove label from non-internalised receptor with a slightly acid
treatment (so only internalised receptors will still be fluorescent)
- analyse her samples fluorescence to see a difference.
the problem is that, as antibody/receptor afinity is different from one
receptor to another, she can't do her positive and negative controls...
for example, the receptor which she used as control (which
internalisation shouldn't change with/without treatment) has such a high
affinity with its antibody that she can't remove it with the acid treatment.
So I would like to know if there is any way to quench fluorescence for
the receptor that are not internalised or if anybody has a protocol to
quantitate internalised receptor?
I thought of comparing non-permeabilised vs permeabilised samples but
PFA fixation permeabilise cells a little, doesn't it?
Thank you for your answers,
Julie Bertout)
the answers :
----------------------------------------------
**
The alternative to acid stripping is to use an unlabeled quenching antibody
to the fluorochrome on your secondary (following your initial pulse). This
also allows any recycled label to quench if and when it re-appears on the
surface. Quenching efficiency is quoted as over 90% for most fluorochromes.
(equivalent to acid techniques) Mol. Probe's list is at:
http://probes.invitrogen.com/handbook/tables/0395.html
A recent example of this technique (with the ErbB2 receptor) is:
Mol Biol. Cell (vol15, 5268-5282) December 2004
.and the math behind the internalisation kinetics is detailed in:
JBC Vol257, no. 8, April 25th pp4222-4229, 1982
JBC Vol265, no.26 Sept 15th pp15713-15723, 1990
Cordially,
Simon McCallum
----------------------------------------------
**
If you check the archives, there is a great deal of discussion about the use
of Trypan Blue for quenching fluorescein. Of special note is Alice Givan's
response regarding the source of Trypan Blue being critical. If you search
the archives with the keywords Trypan Blue quenching you should be able to
follow the discussion. If anyone on the list has come up with a better way
to do it I would be most interested as well.
Regards-
Joanne Lannigan, MS
----------------------------------------------
**
Do a second color to the antibody that is still outside
Single color will tell you everything got internalized, the ratio of
one color to the other in the double stained cells over time can tell
you how many and how much you still have outside.
A problem will appear if re-expression of the outside receptor is so
fast in your system that someway keeps the receptor constant outside.
Then you will need to do laser scanning cytometry or use the Amnis
cytometry system to identify translocation
Carmen
----------------------------------------------
**
I would also worry whether the primary and secondary antibody labeling of
the receptor didn't activate the receptor and initiate internalization
independently of the treatment she is interested in testing. In some cases,
binding of a ligand to a receptor can be mimicked by binding of an antibody
to the receptor with similar effects.
John Kaptein
----------------------------------------------
**
The other problem you have is the temperature at which you stain / internalise.
If you use an FITC label use the quenching solution from the phagotest.
It is a proprietary mix of reagents but should do better than anything
else. If you contact Orpegen.de they might send you a bottle of it to
try.
Good luck
Gerhard
----------------------------------------------
**
some ideas that crossed my mind:
1) Maybe you can use a directly conjugated 1ry antibody and then come with
an anti-fluorochrome directly conjugated Ab, that way you'd be testing
single vs double stained fluorescence. As a bonus, if you have some way of
determining the F/P ratios of the antibodies you could measure correlations
of external to internal receptors which would give you more info. This would
need only 2 washes. Actually, if you don't wash after the 1st Ab you can
make do with only one wash. Washes can be a significant source of cellular
damage and artifacts, as well as reducing the number of cells available for
analysis. As you say, there is uncertainty in the acidic treatment's removal
of the 2ry Ab, which I assume is difficult to ascertain how well it worked
in a given experiment; if you do get determine it and discover it worked
only partially then you are out of luck. The approach I propose would remove
the need for acid treatment. I would strongly think of using some form of
viability probe to increase the reliability of your assay, dead cells will
be there and they can skew things.
2) If you think of it, you can use a directly conjugated 1ry antibody and
then come with a 2ry Ab with a different label. small fluorochromes
shouldn't make much steric hindrance, but I'm only assuming. maybe someone
in the list with experience on this can enlighten us?
3) Don't forget to check the pH stability of the fluorescent probe used
since internalization will probably change the pH environment of the probe.
Also, check what does PFA do to the probe, it might also affect it.
4) In my experience in this kind of dynamic experiments reproducibility is
king, and getting procedures to be as simple as possible greatly reduces
sources of noise. If directly conjugated Abs are unavailable I would go as
far as conjugating my own 1ry Abs to make it faster, simpler, less washes,
less time waiting on ice. These days its no big deal to do your own
conjugates, and probably even cheaper than buying direct conjugates.
Uriel.
----------------------------------------------
**
An ImageStream flow cytometer from Amnis could do this with no additional procedures...
Regards
Vladislav Rozenkov
Hi V.,
could you please elaborate on that?
This is the type of discussion I continuously see when critically
evaluating data re: cell membrane binding versus diffuse
internalization. True sceptics will dismiss classical fluorescence
microscopy images showing cytoplasmic staining (diffuse or punctate),
arguing you really need confocal to decide whether you're seeing surface
stain versus intracellular - and I do see their point.
The way I understand the Amnis device works, confocal is really not an
option. Seeing the individual cell's staining pattern may therefore be
nice to have above and beyond the dotplots/histo's and whatnot outputs
as from a conventional cytometer, but shouldn't be used as a definitive
argument for internalization.
I would therefore argue you can use the Amnis to gather more data from
your cells, but for this particular application having the macrofocal
image shouldn't make the decisive difference. Doing the "stripped cells"
experiment, using fluor-specific secundaries, quenching and/or pH
sensitive dyes will therefore be as necessary as when using a
non-imaging device.
Guy Hermans, PhD
=-=-=-=-=-=-=-=
Hi Guy,
Since you asked...
You're correct that the ImageStream system is configured like a
traditional widefield microscope, not a confocal scope. Our 0.75 NA
objective results in an optical depth of field of about 1 micron. This,
combined with hydrodynamic focussing of the cells with a positional
accuracy of ~1.5 microns, results in a very consistent equatorial
"coffee ring" image of membrane-bound fluorescence signals.
I would say that the change from a sharp "coffee ring" staining pattern
to one that is uniformly distributed over a relatively wide band of
cytoplasm is pretty good evidence of internalization, and in my opinion
it can be done well with our system. However, I would also agree that
any methodology relying on an uncontrolled assessment of fluorescence
distribution alone would benefit from a decreased depth of field like
that provided by confocal imaging.
In my opinion, it's preferable to use an internal control for surface
localization. The way we generally assess internalization is to label
each cell with a known surface-bound control marker in one color along
with the experimental marker which is being assessed for internalization
in a second color. We then perform a cross-correlation of the two images
of each cell and look for a decrease in image correlation as evidence
that the experimental marker is redistributing. This methodology detects
capping and/or internalization because any relative signal
redistribution between the two images that exceeds 0.5 microns (our
pixel size) is reliably detected. It won't replace FRET if you need
Angstrom or nanometer-scale co-localization (like detecting the inner
versus outer leaflet membrane positioning), but by the same token, it is
very easy to implement and has a much wider "spatial dynamic range" than
FRET. The cross-correlation methodology also be used to determine
compartmentalization if you label the endosomes, lysosomes, or golgi.
Nuclear translocation of transcription factors is assessed by labeling
the nucleus with a DNA binding dye and cross-correlating the nuclear
image to the image of the transcription factor.
There are a number of studies published and in press that utilize this
methodology. If you're interested, I would recommend the following for a
more detailed description of the cross-correlation approach itself:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=16563425&query_hl=1&itool=pubmed_docsum
Best Regards,
David
----------------------------------------------
