antibodies working in indirect ELISA but not in inhibition ELISA - (Oct/04/2008 )

Hallo all,
I have an urgent problem to solve:
I produced monoclonal abs against a 22kD protein, purified them and tested several time by indirect ELISA, of course using the target protein as coating antigen. They work great. Later I tried to use the same abs in an inhibition ELISA (plate coated with the antigen, incubation of different amount of the antigen together with a fixed amount of abs in microtubes for a while and then incubation of these solution in the wells of the plate, secondary abs, etc...), tryed lot of different combination of abs and antigen concentration, but I always obtain a flat curve, which means that the signals related to the various antigen dilution are more or less identical and just like the one of the ZERO standard. That's not caused by a too high background.
I observed this strange event also with other monoclonal abs, raised not only against this protein, but also against a small toxin of about 300 kD.
Could some of you help me to interpretate all this in order to solve the problem?
Thank you very much

Hi,

I think you need to tell us about your detection method before we can deduce anything. One possibility is that you actually coat your microtubes when you incubate your antibody and peptide. I.e. you might not have any antibodies in the solution you transfer to your wells.

I use secondary abs anti-Ms HRP conjugated to detect my monoclonal bound onto the plate. I think the tubes are not coated with the abs during the first incubation because:
1) I always use tween, which strongly reduce the chemical interaction between proteins and "plastic"
2) the flat curves I mentioned have not a null value, but the level of the relative absorbance is someway proportional to the concentration of the abs in the solutions.
3) I've always used the same procedure and it works well.
However, all this means that some abs seem to be able to perfectly bind the antigen only if absorbed onto the plate, but not if in solution. That's unbelivable!!! I thought it could be determined by the fact the proteins, when absorbed onto the plate, may change a bit their structure and undergo a little denaturation and so, maybe, exposing better some epitopes, but how to justify the fact that I observed the same occurrence even in other clones (which I'm sure do not recognize the same epitope (I don't explain why cause it would take to much time)) and even in abs recognizing small toxins!?

Hmmm... That is indeed weird. If the antibodies do have a much higher affinity towards antigen bound to plastic as opposed to free antigen things become tricky. Did you use a carrier protein when you produced the original antibody in an animal? Might the antigen have bound itself to a serum protein (e.g. albumin) or something else that may have distorted its shape a bit? This could explain why you see it with the small toxin, since it probably has been bound to some support structure when the antibody has been developed.

In any case you might want to try to play around with salinity, pH, and Tween conc. just to see if you can find conditions in which the antibody-(free) antigen interactions are more favourable. How long, and at what temperature do you incubate your tubes? You might want to try to do it over night in the fridge. I don't know if lowering the tumbling speed of the proteins will do anything, but it could be worth a shot.

Best of luck.

I developed my abs using a peptide reproducing the N-terminal part of the protein (which is a random coiled appendix in the natural protein) and conjugate it to KLH as a carrier. In the inhibition test I tryed also the whole conjugate, but I obtained only slightly better results. I usually incubate my samples 2 hours RT and then i put them in the well and incubate for 1 hour. I already try to vary pH and other conditions, but did not get any differences. However the conditions for the binding must be the same for the coated antigen and for the one in solution!!! that's more than tricky!!!

....and I use a recombinant protein as antigen....

The Devil is in the details as always.

Do you have access to a polyclonal sera against your antigen? Changing the last step into a sandwich ELISA (using the polyclonals as coating) might improve things. I have had problems using direct ELISAs for quantification purposes, producing all kinds of weird curves.

yes I have, but I have to purify it. However I will try. If things will not work as well, it means that interaction protein-plastic makes the difference!!

Let us know how it turns out. This is all quite intriguing.

I tried a sandwich with polyclonal Rb anti-my antigen as presenting abs and then I used my "non-inhibitable" monoclonals as primary abs and it worked, even if the performance was not great. I tried again inhibition test and previous results were confirmed. Thus the hypotesis of proteins (antigens) that, when absorbed onto the plate may change a bit their structure and undergo a little denaturation and so, maybe, expose better some epitopes...falls!!!
So, the question remain: why do this two clones of monoclonals work well in indirect tests or in sandwich ELISA but not in inhibition ELISA?
This makes me crazy!!!!!!!