Search

Five different brands of nitrocellulose (NC), each of pore size 0.45 micron and without adsorbed antigen, bound different amounts of two labelled antisera and labelled protein A. Experiments with some non-ionic surface active agents and proteins showed that milk powder and bovine serum albumin were the most effective agents for blocking non-specific binding of labelled protein to NC. With some of the NCs, Nonidet P-40 (NP-40) and Tween 20 were almost as effective as milk powder. The protein-binding capacity of unblocked NC and the level of protein binding after blocking were found to be inversely proportional to the pore size of the NC. A comparison of blocking agents in an immunoassay with pollen proteins adsorbed to NC discs revealed that the highest specific uptakes of antiserum occurred with NP-40 and Tween and not with any of the protein blocking agents such as milk powder. Hence, for the detection of proteins using NC-based assays (but not necessarily following electroblotting), the best choices would appear to be: NC of pore size 0.45 micron; a brand of NC that provides a suitable balance between protein binding capacity and non-specific uptake of protein after blocking; a non-ionic detergent such as NP-40 or Tween 20.

The resolution and detection of individual components in complex extracts by protein blotting have been investigated. By probing nitrocellulose transfers with monospecific and multispecific antisera, it was demonstrated that dissociating conditions were required for the maximum resolution of antigens by polyacrylamide gel electrophoresis, a conclusion reinforced by results from 2-D electrophoresis. The dissociating and reducing treatments employed, however, were both shown to be responsible for some loss of total antigenicity and included the complete loss of at least one important antigen. Assays with nitrocelluloses of different pore sizes demonstrated that both higher protein-binding capacities and higher backgrounds were associated with the use of the smallest pore size, while the sensitivity of the assay was greatest when a non-ionic detergent, and not proteins, were used for blocking. Nitrocellulose-bound proteins may be stained with amido black, India ink, toluidine blue, Ponceau S or a gold sol, but these agents do not always give identical staining patterns. While detection of components with immuno-enzyme staining methods had some advantages, problems with non-specific binding were encountered. These did not occur with affinity purified radiolabelled second antibodies, which in combination with scanning of autoradiographs allowed a quantitative approach to be adopted.

A selection of different membranes commonly used to bind proteins in blotting and dot binding assays were investigated for a range of properties which would influence their performance. Large differences were observed in the membranes' ability to bind increasing amounts of protein, the effect of incubation times on the quantity of protein bound and the loss of proteins from the membranes following their incubation with different detergents or protein blocking agents. These differences could only partially explain the observed performance of the membranes when used as protein adsorbants in immunoassays and when different buffers were used for the electro-transfer of several different proteins to a range of membranes.