Expanding the capabilities of lateral flow assays using computationally designed affinity proteins

Thesis (Ph.D.)--University of Washington, 2019Point-of-care diagnostics have enabled clinical testing in areas previously considered challenging, specifically for underserved populations and in low resource settings. Lateral flow tests, such as the ubiquitous pregnancy test, have proven relatively successful in their implementation due to their low cost and ease of use; however their application has been limited to a select group of targets and types of assays. There is a need for novel molecular recognition elements that address some of the key limitations of antibody use in lateral flow assays. The following dissertation describes the development of lateral flow assays using novel molecular recognition elements, computationally designed proteins. We describe the first lateral flow assays using computationally designed binders, targeting the head and stem region of the influenza glycoprotein, hemagglutinin (HA). The best performing of these assays, using a head region specific HA binder, was integrated into a two-dimensional paper network that integrated enzymatic amplification. Not only did this device sensitively detect native influenza virus from a spiked patient sample, the computationally designed binders proved highly thermostable when integrated into a paper network. Lastly, we used our knowledge of lateral flow assays to use modular design to develop an Ebola glycoprotein (GP) assay using an Ebola specific computationally designed binder. While we began by investigating the use of a nitrocellulose binding protein to anchor our Ebola binder, we found that the use of a streptavidin test line with biotinylated binder led to the best performance for detection of Ebola GP. All together, this work introduces computationally designed affinity proteins as an antibody alternative for lateral flow assay development. Future work developing modular protein assembly for lateral flow assays will enable more rapid development of this novel low cost diagnostic platform for a wider range of applications than previously possible