[Histonet] Demonstration of antigen on cell surface - question?

Andrea T. Hooper anh2006 <@t> med.cornell.edu
Mon Jun 20 17:57:50 CDT 2005

Good points .... Also, in addition to background issues which are 
often the culprit, in my experience membrane antigens when highly 
expressed - such as say EGFR in certain cancers or in skin - are 
often detected in both cytoplasmic and membrane staining patterns 
mainly due to two reasons I can think of:

(1) The antigen truly is in the cytoplasm on it's way to the cell 
membrane therefore the staining you detect is "real"
(2) The antibody staining reaction is so strong you are getting a 
fuzzy deposition of DAB (this is especially the case in high 
amplification kits when used on overexpressed antigens)

(finally I just want to add that sometimes what is membrane can 
appear cytoplasmic depending on the shape of the cell and the cell 
type - for example when staining VEGF receptors on endothelial cells 
it is often VERY hard to tell where on the cell they are located 
using IHC).

Perhaps with a little more info about the cell type and antibody you 
are using more info can be provided in this specific case.

Generally speaking though I would suggest that you not use IHC/DAB 
etc for determining where in a cell a specific protein is expressed. 
It is more useful perhaps to use immunofluorescence and confocal 
microscopy or at minimum epifluorescence. Then you will most 
certainly get the answer you need. In addition, you can take a 
non-staining approach and do either of the two following:

(1) Make differential cell lysates of membrane and cytosolic 
fractions and do westerns
(2) Make a GFP fusion protein and express it in your cell and view 
where it is located via confocal

Good luck and I hope this makes sense!

At 4:34 PM -0400 6/20/05, Luis Chiriboga wrote:
>I think a bit more info is needed but here's a few suggestions....
>If your working in frozen or formalin fixed tissue, serially dilute both
>primary and secondary antibodies independently to arrive at your best signal
>to noise.
>If you are performing antigen retrieval (formalin fixed), run a time series
>of different HEIR times or various concentrations of enzyme.
>This should help you improve the S/N to an extent that allows easier
>If your working with a novel protein...
>If you have access to cell culture, perform cell fractionation and check
>localization via western blot.
>Have you checked in the literature? for similar class or family of protein
>and it's localization?
>If you have/know the target protein/antigen check for protein localization
>signal (NCBI-protein database & genbank), or protein sequence homology to
>other known proteins (NCBI-Blast)
>Luis Chiriboga Ph.D.
>NYU Cancer Institute and
>Bellevue Hospital Center
>New York University School Of Medicine
>Department Of Pathology 4W27
>462 First Avenue
>New York, N.Y. 10016
>W(212) 562-4667.
>F(212) 263-2041
>-----Original Message-----
>From: histonet-bounces <@t> lists.utsouthwestern.edu
>[mailto:histonet-bounces <@t> lists.utsouthwestern.edu]On Behalf Of GT Hebert
>Sent: Monday, June 20, 2005 3:49 PM
>To: Histonet <@t> lists.utsouthwestern.edu
>Subject: [Histonet] Demonstration of antigen on cell surface - question?
>How would one go about determining if your specific antigen (say for IHC) is
>located on the cell surface?  Right now it seems that the antigen is
>staining everywhere. (I am running a sheep anti-mouse 'antigen' and using a
>rabbit anti-sheep biotin conjugated with ABC + DAB for detection.  (My
>negative is normal sheep IgG -- is negative).  Is there a way through immuno
>or other that I can specifically say that the antigen is present on the cell
>surface?  Are there general cell surface markers that I can use along with
>my primary?  Thank you all.


More information about the Histonet mailing list