Take the 2-minute tour ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

I'm trying to evaluate data taken from fluorescence microscopy with antibody staining, and am wondering whether there is any standard way to evaluate the specificity of the antibody for such measurements? In particular, I am interested in evaluating whether there is specific binding of the antibody to the cells in the experiment, or only unspecific binding.

By leaving out the primary antibody or taking a control antibody without specific binding, I can compare fluorescence intensity, and an increase in intensity in the actual experiment towards this control will indicate qualitatively that the antibody binds specifically. But I wonder whether there is a standard statistical test or something to get such a result in a more quantitative way?

share|improve this question

1 Answer 1

up vote 3 down vote accepted

When testing an antibody for an imaging application, it is almost always a good idea to test it in another application like ELISA or Western blotting to see if it binds the target of interest. For example, try to find high- and low-expressing cell lines, load equal amounts, and see if there is a difference in signal at the expected molecular weight. Check the company's literature or contact their technical support to see if they have tested the antibody by immunofluorescence (IF), and if so what their results were and how they validated them. Check the literature to see if other people have published with this product. http://highwire.stanford.edu is a pretty good full-text search engine, but they don't have all journals in their database. Google Scholar can help, too.

You can also use your high- and low-expressing cell lines to see if there is a stronger IF signal in the high-expressing line. Look at your images to see if the signal is localizing where you expect it to, and do co-staining with antibodies to other proteins that should co-localize with your target of interest. For example, if your target is expected to localize to mitochondria, co-stain with COX-IV. A quick Google search found this Cellular Localization IF Antibody Sampler Kit that has controls for several different organelles and other specific locations.

Make sure you are following the manufacturer's recommended protocol. Don't crank the gain up too high on the scope, because non-specific background goes up, too. Look at the induction of signal between high- and low-expressing cell lines. Generally you would want at least a 2-3 fold shift, especially if you see a strong change in signal intensity by Western. Try using siRNA or similar silencing technology to knock down the signal in your cells, and validate using Western and IF.

Ultimately, you can never be absolutely 100% certain that an antibody is binding a certain target, and only that target, but you can accumulate a lot of evidence to show that it is likely. Try to put yourself in the mindset of a reviewer, and objectively convince yourself that the signal is real. If you can't, then shop around and try something else. Good luck!

share|improve this answer
Don't really know if the first point is really so useful... Most of the time antibodies work in IHC but not in WB or vice-versa and not having a signal in WB does not mean that your IHC signal is aspecific. Also, if you are working on tissue (rather than cell cultures) you have many different cell types which may complicate things a lot. –  nico Jan 15 '13 at 17:38
@nico - In my experience (at a certain antibody company) it is perfectly possible to have a great IF or IHC antibody work well in WB, and vice versa. It all depends on how the Ab is screened, and the nature of its epitope. I agree that not having a WB signal would not necessarily disqualify its staining capacity, so that's why I put in the other options like differentially-expressing cell lines and siRNA. It seemed like the questioner was talking about cells, and my mind differentiated that from tissue, which may not have been a correct assumption... –  MattDMo Jan 15 '13 at 19:20
Hmmm... in my experience (in the lab) that is not that common when working with tissue. I don't work on cell culture, so maybe that is different. –  nico Jan 15 '13 at 20:50
Tissue makes for dirty WBs in general, which is why we tried to do as much validation on cell lines (and paraffin-embedded cell pellets for IHC) so we could be sure before moving on to tissue. We wanted to be as damn sure as possible we were staining the right thing before releasing something to market. That meant we didn't quite have the product diversity that some other companies might have had, but the quality and level of trust were a lot higher, overall. If you haven't used their stuff before, I highly recommend it, even though I'm not there anymore. –  MattDMo Jan 15 '13 at 21:58

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.