Engineering Bacterial Organelles for the Sequestration of Synthetic Metabolic Pathways

Bacteria were long thought of as bags of enzymes with little internal structural organization. This idea has been challenged in recent years, particularly with the discovery of bacterial microcompartments (MCPs), which are polyhedral protein structures within the cytoplasm that house specific cellular reactions for functions such as carbon fixation. These MCPs offer metabolic engineers a solution for biosyntheses that require spatial control over the intracellular flow of proteins and small molecules. Yet much is still unknown about MCPs, particularly with respect to controlling their formation and functional properties. To this end, we develop tools for the control of microcompartment formation, enzyme encapsulation, and small molecule transport across the compartment boundary. Of particular interest is the encapsulation of enzymatic pathways that synthesize fuels, pharmaceuticals, and other products which are challenging to produce in the cytosol.

Bacterial MCPs consist of protein shells that sequester native enzymatic reactions and are thought to concentrate private cofactor pools, sequester toxic intermediates, and insulate the compartmentalized pathways from other cellular processes. Two well-characterized systems from Salmonella enterica use MCPs for 1,2-propanediol utilization (Pdu) and ethanolamine utilization (Eut). In both cases it is thought that the protein shell segregates the toxic aldehyde intermediate of the pathways from the cytosol, and that pores in the shell selectively mediate the passage of substrates and products into the shell lumen. In both the Pdu and the Eut MCP systems, short N-terminal extensions from native Pdu and Eut enzymes are sufficient to compartmentalize heterologous proteins in the protein shell. In order to tune the catalytic activity of the MCPs, we orchestrate MCP formation at the transcriptional level to control when the organelles are formed and how much heterologous cargo is encapsulated. We also create new localization signals that allow fine control over the loading of multiple enzymes to the MCPs simultaneously. We show the important role that pore residues play in small molecule transport across the shells of virus-like particles and the MCPs themselves. Lastly, we use computational tools to model MCP function in order to facilitate the selection of pathways for encapsulation. Together, these experiments provide a framework and toolkit for the creation of custom biosynthetic organelles with user-defined enzyme content and transport properties, enabling production with biosynthetic pathways which otherwise fail to function.