Membrane proteins constitute the largest class of drug targets but they present many challenges in drug discovery. Importantly, the discovery of potential drug candidates is hampered by the limited availability of efficient methods for screening drug-protein interactions. In this work we present a novel strategy for rapid identification of molecules capable of binding to a selected membrane protein. An integral membrane protein (human aquaporin-1) was incorporated into planar lipid bilayer disks (lipodisks), which were subsequently covalently coupled to porous derivatized silica and packed into HPLC columns. The obtained affinity columns were used in a typical protocol for fragment screening by weak affinity chromatography (WAC), in which one hit was identified out of a 200 compound collection. The lipodisk-based strategy, which ensures a stable and native-like lipid environment for the protein, is expected to work also with other membrane proteins and screening procedures.

Lab-on-a-chip systems with molecular motor driven transport of analytes attached to cytoskeletal filament shuttles (actin filaments, microtubules) circumvent challenges with nanoscale liquid transport. However, the filaments have limited cargo-carrying capacity and limitations either in transportation speed (microtubules) or control over motility direction (actin). To overcome these constraints we here report incorporation of covalently attached antibodies into self-propelled actin bundles (nanocarriers) formed by cross-linking antibody conjugated actin filaments viafascin, a natural actin-bundling protein. We demonstrate high maximum antigen binding activity and propulsion by surface adsorbed myosin motors. Analyte transport capacity is tested using both protein antigens and microvesicles, a novel class of diagnostic markers. Increased incubation concentration with protein antigen in the 0.1–100 nM range (1 min) reduces the fraction of motile bundles and their velocity but maximum transportation capacity of >1 antigen per nm of bundle length is feasible. At sub-nanomolar protein analyte concentration, motility is very well preserved opening for orders of magnitude improved limit of detection using motor driven concentration on nanoscale sensors. Microvesicle-complexing to monoclonal antibodies on the nanocarriers compromises motility but nanocarrier aggregation via microvesicles shows unique potential in label-free detection with the aggregates themselves as non-toxic reporter elements.