Presentation description

Acetyl-coenzyme A (AcCoA) is a core metabolite in cells, but its functions go far beyond just metabolism. Prior to our genetically encodable acetyl-CoA biosensor (PancACe), AcCoA detection methods were limited to cell lysates. PancACe offers an alternative to existing methods, enabling live cell imaging while requiring a fraction of the cells. The sensor was engineered by inserting cpGFP into PanZ, a known acetyl-CoA binding protein, with substantial optimization going into its insertion site and linkers. PancACe's dynamic range (20-3,000 µM) corresponded closer with AcCoA levels in E. Coli (20-600 µM). However, we sought to shift the sensor's dynamic range to include 1-30 µM, which corresponds to AcCoA levels in mammalian cells. To improve quantifiability and affinity, mCherry was added to PancACe's N-terminus and a single residue was mutated in the PanZ binding site. This second iteration, PancACe 2.0, considerably lowered the KD of acetyl-CoA to ~25 µM with an optimal dynamic range of 1-100 µM but saw decreased specificity against propionyl-CoA (PropCoA) as the KD of both acyl-CoAs are nearly equal. Our protein modeling and previous literature illustrated that affinity, specificity, and responsivity of biosensors are largely affected by the linkers between the cpFP (cpGFP) and the binding protein (PanZ). PancACe's N-terminal linker lies very close to its binding pocket, potentially facilitating interactions with acetyl- and propionyl- groups and thus a possible solution for inducing specificity. Therefore, further iterations of PancACe seek to first address cpGFP's N-terminus linker, followed by the C-terminus linker and other random mutations to refine the sensor's function. Through screening of a 65-member N-terminal linker library, which was designed using molecular dynamic simulations, one variant exhibited low fold-change but >50x higher affinity for AcCoA over PropCoA. This lead will therefore guide further investigation with similarly structured linkers.