Sounds like a strategy that will not work. Either your first antibody is from another species or you are using another experimental animal. With your current strategy the goat antibody labels the plate. Alternatively you could think of plating the Fab fragment of your purified antibody, but it has to be real pure.

It's the same problem as for the detection of human ADA.

I suggest you reverse the order of your ELISA.

First: fix an anti mouse (goat) on the ELISA plate.

Then applying mouse serum dilutions tested.

Then: block free remaining sites with a negative mouse serum.

Finally: to reveal with the antibody A which is coupled either with biotin or with peroxidase.

This should reduce the associated background.

I also suggest to purify the sera to test for eliminating IgMs possibly increasing background noise.

Last solution: develop a monoclonal antibody against the antibody A which do not cross react with mouse Igs.

Good luck in this field we all have the same problem.

Agree with Uwe that this will probably not work.  The labelled goat anti-mouse Ig is going to detect the mouse anti-mouse Ig and also the antigen (mouse Ig).  Coat with mouse anti-mouse Ig antiserum directly.  Monitor with a labelled mouse Ig added to this.  You cannot use the goat anti-mouse Ig in your screening system.

If the antibody concentration is not limiting, you already saturate the serum compounds (antigen A) that might bind to antibody A with your pre-treatment. There's nothing left that want to bind to you coated antibody then. 

It might work if you get a labelled version of antibody A (there are kits available on the market to label antibodies yourself) and use this as secondary antibody to detect the bound protein A:

You coat with unlabelled Mouse-anti-A and block with a mouse-free blocking compound, incubate with your untreated mouse serum and detect with labelled mouse-anti-A.

You can detect antibody-bound mouse-anti A-antibody with a different set-up (but not antigen-bound + antibody-bound-antibody): You coat your plate with an unlabelled goat-anti-mouse-antibody and block the plate with mouse-free compounds. Then you pre-incubate your mouse-anti-A antibody and the mouse serum. Now you add this mixture to your plate and the mouse-anti-A-antibodies will bind. Finallly you detect the mouse-anti A-antibodies with a labelled anti-mouse antibody. This way all components are available commercially, but be aware that you detect all bound mouse-anti-A antibodies and not only those that are bound to antigen A. And I guess you are interested in antigen A finally. So you don't get any relevant information at all, since you already knew how much anti-A antibody you added.

Don't forget the controls, especially you should look for the signal if you omit the serum-incubation step (so see what the labelled anti-A antibody does with your coated plate).

Coating with the Fab is a valid option, but you have to make sure that your anti-mouse IgG conjugate is based on an anti-Fc conjugate, without cross-reactivity to the constant domains present on Fab fragments. If you fulfill that requirement, you should be able to measure the anti-idiotypic response to antibody A.

Is antibody A an igG? If not, the ELISA can work with a very specific anti-IgG conjugate.   

Fully agree with Jean-Jacques Pin from Dendritics; I was first to propose this approach over 25 years ago (cf. my manuscript on Rhematoïd factors).

Hi all,

there is a quite simple solution. On the the antibodies has to be labelled, either directlely (by an enzyme) or indirectely by e.g. biotin. Then you have many assay desigsn open.