Directed evolution of a genetically encoded contrast agent for ultrasound

Abstract

A major challenge in the field of biol. imaging and synthetic biol. is noninvasively visualizing the function of natural and engineered cells inside opaque samples such as living animals. One promising technol. that addresses this limitation is ultrasound, with its penetration depth of several cm and spatial resoln. of tens of mm. Recently, the first genetically encoded ultrasound contrast agents-gas vesicles (GVs)-were developed to link ultrasound to mol. and cellular function via heterologous expression in both commensal bacteria and mammalian cells. GVs are air-filled protein nanostructures derived from buoyant photosynthetic microbes, in which they serve as cellular flotation devices. GVs are encoded by operons of 8-14 genes, and most of their mol. makeup comprises the structural protein GvpA. The air inside GVs allows them to scatter ultrasound. Just as the discovery of the first fluorescent proteins was followed by the engineering and evolution of their properties, we are working to engineer the properties of GVs as acoustic reporters. Here, we pursue this goal using directed evolution by both devising a strategy for high-throughput acoustic screening of GVs in bacterial colonies and validating its ability to identify new phenotypes in mutant libraries. We generated scanning site satn. libraries for two homologs of GvpA and screened them in E. coli using a custom-built robotic ultrasound plate scanner. Custom imaging pulse sequences were used to assess the acoustic phenotypes of each colony, including total backscattering, nonlinear scattering, and collapse pressure. Using this technique, we identified mutants of GvpA with >150x higher acoustic signal than their parents. These techniques will enable directed evolution to play as big a role in the engineering of acoustic biomols. as it has in the development of their fluorescent counterparts.