Antibody' specifity

[Scott Coutts]

Nikolai Scherbak wrote:
 >
 > Dear colleagues! I am wondering how specific antibodies are?
 >

Antibodies are generally quite specific, but it is not unusual to see
cross-reactivity depending on your antibody and it's source. This can be 
a problem particularly with polyclonal serum.

 >
 > I am working with transgenic plants (A. thaliana) and using
 > polyclonal antibodies raised against protein that I would like to
 > overexpress in transgenes. Unfortunately, my antibody binds even for
 > non-transformed control plants. I know. That there is another protein
 > i Arabidopsis with up to 50% identity to my recombinant protein. Can
 > this be an explanation to the positive result in negative control?
 >

Yes, it certainly can.

 >
 > Or, by another words: can antibodies raised against one protein bind
 > also to another protein with the 50% similarity to the first one?
 > Mol.weight of proteins is almost the same...
 >

Definitely. But there are a few ways around this problem.

The easiest step is preabsorption. This may not work, though. What you 
do is mix your antiserum with your positive control tissue. This way, 
you're removing all of the antibodies that will cross react with your 
transgenic line's 'background'. All you will be left with is antibodes 
that _don't_ bind your positive control. Assuming that some of the 
antibodies are specific for the unique portions of your introduced 
protein, then it will still work. If your protein of interest and the 
endogenous protein do not have some unique epitopes, then this will not 
work. But it's worth a try because it's probably the quickest and 
cheapest way around your problem.

Otherwise, if you can find an area of your protein that is different 
from the one already present in A. thaliana, then you can raise 
antibodies against this small fragment. This way, you should end up with 
polyclonal serum that reacts only against that unique portion of the 
protein. This may or may not be possible depending on the protein 
sequences and whether you can raise your own antibodies.

Lastly, and probably the most reliable method would be to obtain a 
monoclonal antibody. This way, you're antibodies are binding to only a 
single epitope in your target protein. Unless you pay someone to do it 
(it can be very expensive) then it can be a lot of work. Again, this 
will not work if there are not unique epitopes in your protein when 
compared to the endogenous one.

Good luck!