(59f) Engineering and Understanding Enzyme and Pathway Localization to ER and Lipid Droplet Membranes

Many essential and industrially important metabolic pathways are assembled on cellular membranes: proteins are glycosylated through a series of membrane-bound reactions; in plants isoprenoids are synthesized in well-organized multienzyme structures bound to intracellular membranes; and, the key step in acetate ester production in yeast is localized to the membrane surface of lipid droplets. Such localization is advantageous. High local concentrations of enzymes and reaction intermediates can increase pathway reactions rates and prevent undesired side reactions from consuming pathway intermediates, thus leading to increased pathway yields. Intracellular scaffolding technologies (e.g., protein and nucleic acid scaffolds) have re-created these advantages in engineered pathways by co-localizing enzymes in the cytosol. While these technologies have produced substantial gains in pathway yields they cannot readily accommodate membrane-bound enzymes. This gap in synthetic biology tools significantly limits our abilities to access the benefits of enzyme co-localization with membrane-bound pathways.

The goal of this work is to identify protein parts that can be used to direct the assembly to biosynthetic pathways comprised of natively membrane-bound and cytoplasmic enzymes on intracellular membranes. Specifically, we aim to develop synthetic biology tools that can sort endoplasmic reticulum (ER) enzymes to lipid droplets (LDs) and that can direct cytoplasmic enzymes to LDs. In this way, we aim to create co-localized reaction centers in the intracellular space. The first step in engineering such a system is to understand and control localization to LDs. For example, alcohol acetyltransferase-1 in S. cerevisiae, which catalyzes the condensation of simple alcohols and acetyl-CoA to the corresponding ester, sorts from the ER to lipid droplets where it remains during stationary phase. Through a series of structure-function studies we have identified dual N- and C-terminal amphipathic helices that are responsible for this intracellular localization. Our studies of LD localization extend to a series of protein tags known to function in mammalian and plant cells. We demonstrate that the terminal hydrophobic domain from a putative methyltransferase AAM-B can localize cytoplasmic proteins to LDs via the ER. Under fermentation conditions, we demonstrate that the LD protein olesin can be used as a terminal fusion to localize proteins to LDs in yeast. Building from these localization studies we aim to organize LD-localized multistep reaction pathways for the synthesis of acetate and ethyl esters.