Directed evolution of proteins with novel, highly specialized, or optimized function(s) requires physical association among genes, gene products, and the effects/activities of these products. Whereas this association is maintained by cellular compartmentalization in vivo, genotype-phenotype linkage can also be achieved via transcription/translation of PCR libraries in an oil:water emulsion. The latter approach, referred to as in vitro compartmentalization (IVC), obviates some of the practical difficulties related to compartmental access, system complexity, and genetic selection associated with in vivo evolution, and has been successfully used to evolve enzymes with improved functionality or altered specificity. This project 4 uses IVC to evolve HIV-1 RT variants resistant to RNase H inhibitors that we identified by high-throughput screening but are toxic in cell culture. A second adaptation of IVC aims at evolving zinc finger proteins that recognize the PPT-U3 junction and is based on mass spectrometry and NMR investigations indicating that abnormal duplex geometry at this junction may be common to retroviruses and LTR-retrotransposons. [Corresponds to Le Grice Project 4 in the April 2007 site visit report of the HIV Drug Resistance Program]
