Lousy Antibodies: Help!

U58563 at uicvm.uic.edu U58563 at uicvm.uic.edu
Thu Apr 6 22:10:34 EST 1995


I've had some inconsistent and frustrating results with an antibody against a
novel tyrosine kinase, and another person in my lab has done more experiments
than I (and been frustrated more!).
   When used for immunohistochemistry, the antibody gives a nuclear stain in
transfected cells, and can be amplified with gold/silver to give nuclear
staining in tissue sections.
   When used for a Western blot, however, it gives NO band of the appropriate
size, and sticks strongly to histones.  But yet, this background is -not-
competed by peptide, while the immunohistochem result -is- competed by peptide.
[I know they're histones because they show up in the Coomassie stain, and are
strongly enriched in nuclei but not nuclear salt extract]
   In massive overexpression (bacterial or mammalian) it gives a weak band on
a Western, of the proper size.
   In immunoprecipitation, it sucks, but can again be forced with difficulty
in massive-overexpression systems.

So... Is the immunohistochemistry avoiding problems all the other procedures
face?  Or can a nonspecific signal be competed by peptide?  What suggestions
can you think of?  All ideas appreciated.

Please E-mail directly to me (Mike Serfas at u58563 at uicvm.uic.edu), because
propagation of posts to and from this site is bad and getting worse.



More information about the Methods mailing list