ELISA methods for the analysis of antibody responses induced in multiple sclerosis patients treated with recombinant interferon-beta

Upon treatment with protein therapeutics, a subset of patients will typical ly develop antibodies against the drug. These anti-drug antibodies can be o f concern because they have the potential to alter the drug's therapeutic a ctivity. In the case of relapsing-omitting multiple sclerosis (RRMS) patien ts receiving recombinant interferon-beta (IFN-beta), those receiving BETASE RON(R) (IFN-beta-1b; E. coli expressed, non-glycosylated, des-Met-1, Cys17S er recombinant IFN-beta) have a higher incidence of IFN-beta specific antib odies compared to those receiving AVONEX(R) (IFN-beta-1a; mammalian cell-ex pressed, natural sequence, glycosylated recombinant IFN-beta). The current study reports the development and characterization of ELISAs that detect di stinct components of the anti-IFN-beta response in patients' sera, and ther efore can potentially be used to characterize the composition of the anti-I FN-beta antibody response. ELISAs were developed using a constant detecting reagent but a variety of IFN-beta-derived test antigens (e.g., native IFN- beta, biotinylated IFN-beta, IFN-beta peptides) and capture methods. Assays were characterized using serum samples from a small number of patients tre ated with recombinant IFN-beta (either BETASERON(R) or AVONEX(R)). Assays i n which IFN-beta was captured via a specific mAb, or in which biotinylated IFN-beta was captured via streptavidin, detected serum antibodies that reco gnize IFN-beta in its native structural state. In contrast, assays in which IFN-beta was coated directly onto the assay plates detected antibodies tha t recognize forms of IFN-beta possessing a folded structure distinct from t he native structure. Certain epitopes present on native IFN-beta were not r epresented in these assays in which the test antigen was directly coated on plastic. Antibodies specific for linear epitopes could be detected using l inear peptides as test antigens; the locations of these epitopes were mappe d by reference to the X-ray crystal structure of IFN-beta-1a. Together, the se data show that the mode of antigen presentation employed in IFN-beta ELI SAs determines which antibody specificities are detected, and can affect wh ether or not a given serum sample is identified as positive for anti-IFN-be ta antibodies. As a consequence, screening samples in a single ELISA format presenting IFN-beta in a non-native form may lead to underestimation of th e incidence of IFN-beta treated MS patients that have generated antibodies specific to the native, active form of the drug. (C) 1999 Elsevier Science B.V. All rights reserved.