I am trying to establish an ELISA to measure reactivity of serum from infected patients towards various viral proteins. Currently, my set up is on a neutravidin coated plate. I add whole cell lysate from HEK293T cells transfected with a plasmid encoding the protein of interest (POI). The POI has a biotin acceptor peptide (BAP) and an HA tag on its C-terminus, and we biotinylate the BAP tag with the idea that this will allow our POI to specifically bind to the neutravidin on the plate (as of right now we do not have access to purified viral protein, so this is our indirect way of partially purifying). After the whole cell lysate is added to the plate I add a 1:50 dilution of either infected patient serum or negative control human serum, following up with an HRP-conjugated goat anti-human IgG secondary.

The infected patient serum reacts strongly (O.D. ~ 0.8-1.2) to the viral proteins with very little non-specific reactivity to the non-transfected whole cell lysate control (O.D. ~ 0.05-0.08), indicating a strong and specific recognition between the serum IgG and the viral proteins. However, what has been strange, is that the serum from our negative control patients has very broad and strong reactivity (O.D. ~ 0.3-0.6) to all the viral proteins, the non-transfected whole cell lysate, and even the no lysate control.

I am wondering if anyone has had this experience before? I know that human serum can have issues with being generally "sticky" and causing high background noise in ELISAs, but if this were the case, I would expect the same issue to be happening in my infected patient serum samples. Could it be how the sera is processed? Thank you!