non-specific antibody binding problem with mouse blood

RICE,LORI P lrice at
Wed Jan 30 17:08:00 EST 2008

We are trying to identify the small number of circulating mature 
and progenitor endothelial cells in mouse blood.

These cells are CD45(-), so we initially gate on that. In 
addition, we gate on TER119, an erythroid marker, to help 
eliminate any RBCs that show up in our CD45(-) population. The 
antibodies of interest are against VEGF and CD117.

After many tries with various protocols, we are back to trying a 
simplified version of a Ficoll separation method recommended for 
mouse blood. We tried staining pre- and post- Ficoll spin. We 
expected this sequencing not to matter much.

But, if we Ficoll spin first, and THEN do antibody staining, we 
see CD45(+)-TER119(+) cells which should not happen, and way too 
many VEGF, CD117, and dual stained cells.

If we stain FIRST, then Ficoll, we see much fewer CD45(-) cells to 
start with, and end up with proportionately fewer cells of 
interest, which is more expected.

However, I am wondering if our results staining pre-Ficoll are 
real or if the ficoll spin is somehow damaging either the antibody 
or clipping off the fluorochrome.

Also, we have no explanation as to why we would see so many cells 
bound to VEGF and CD117 antibodies even with rat serum added to 
the whole blood for blocking prior to antibody staining.

We thought either of these methods would be the gentlest way of 
getting rid of RBCs without making the cells "sticky" as seems to 
happen with any of the ammonium chloride lysis protocols.

Any suggestions would be greatly appreciated.

Lori Rice, Ph.D.
University of Florida
lrice at

More information about the Cytometry mailing list