Sandwich ELISA Help!

[e5dy7]

Hi

I am trying to get an ELISA to work for me that has worked in the past for a labmate who is no longer in the lab.

It is a sandwich ELISA. Capture ab = rabbit polyclonal purified via affinity column
against protein from parasite (Bm)
ag = recombinant protein (Bm)
secondary = human anti protein (protein in parasite and
humans has 90% identity
tertiary = mouse anti human..hrp

in gel protein is not degraded

everything works in a western
- capture binds protein
- secondary binds protein

---also looked at this by doing direct elisa and everything interacts yet get negative results with sandwich, even with using very high conc of ab and ag


HELP!

[e5dy7]

oh and using carbonate bicarbonate to coat and skim milk to block

[Elisadeveloper]

A few more details would be helpful.

From what you say, both antibodies bind to the antigen in isolation, but in the experiment that you are performing as a sandwich, you can't get a signal.

The coating conditions and blocker seem appropriate, assuming the concentration of coating antibody is in the low microgram per mL range.

How big is the parasite protein? If it is small, the epitope recognized by the human mono detection reagent may not be accessible after binding by the poly. Have you run the assay the other way round?

Is the anti-human mAb appropriate? Make sure it is human specific and doesn't cross react with rabbit or bovine IgG (in the blocker).

If it does, you will get high background.

Let us know more details and maybe we can figure something out.

[e5dy7]

Yes in isolation both antibodies bind the protein.

The protein is 27KD. I`ve definitely thought about this, that the epitope may be blocked. What puzzles me is that this exact protocol worked for a labmate in the past. Following his protocol exactly we are still unable to get signal.

I have not run the assay the other way around. I actually haven`t thought about doing that. It would be interesting, I`ll have to try that. But even if this works I`m not sure what info this lends to fixing the current problem...

The antihuman is polyclonal

I`m really not sure where to go from here. I know that all Ab and ag bind in isolation. I ran the indirect ELISAs to test this at the same time as the sandwich. Indirect was positive, sanwich was negative.

The capture antibody was purified by affinity chromatography against a specific epitope of the parasite protein different from the human (to separate it from the antibodies that would bind human protein) ...

[e5dy7]

Oh geez, looking back at the first post i wrote the Ab incorrectly. To clarify:

secondary antibody is mouse anti human protein
tertiary is goat anti mouse*HRP

Sorry!

[Elisadeveloper]

If the assay really has worked before with these reagents, it appears that there may be a component failure

One possibility is that only a small fraction of the coating reagent is actually functional antibody, either due to some protein contamination, leaching of the antigen from the column, or due to denaturation during elution or freeze/thaw. This would not be such a problem when the antibody is used to detect the immobilized antigen, but if 95% of your antibody is inactive, and you coat at 1 ug/mL, the effective coat is 50 ng/mL.