I see a hook effect when measuring antibody binding to protein by ELISA using TMB substrate but not when using ABST substrate. I am measuring antibody that has been deglycosolated with PNGase. Does anyone know of an explanation for this? Thanks.
It is somewhat surprising that you would observe hook effect only with the more sensitive reagent. This would exclude the possiblity for the classical reason for the hook effect which is the incompatible amount of coating-detection antibodies on the assay and analyte concentration in the sample.
I wonder do you stop the TMB reaction with acid (H2SO4 or similar?). I have noticed that strongest reactions can to start precipitate if there's a lag between stopping and measurement. There might also be issues with the TMB itself, which can result in precipitation and leading to loss of signal.
Which format you are using competitive or sandwich?. Hook effect is generally found in sandwich type of format. One can get hook effect in competitive formate if antibody used in the assay has positive cooperativity. This can be found out by Scatchard plot wherein binding sites in the antibody will be one or below one in IgG. Substrate those are able to give sensitive assays are prone to hook effect.