Protein-protein interactions (PPIs) have been a central focus of numerous studies over the past decades, as they play a critical role in regulating essential cellular mechanisms. Interactions between proteins can initiate signaling pathways, serve as on/off switches, lead to the formation of stable complexes, metabolic reactions, developmental controls, etc. Furthermore, understanding these protein-protein interactions can create an opportunity to develop and expand the field of systems biology. This is evident in the advent of high-throughput and in-silico methods, which have produced comprehensive information on biological networks and structure-function relationship predictions, leading to the discovery of druggable targets like enzymes, receptors, and other molecules of therapeutic interest. Nevertheless, many proteins are still understudied, warranting additional techniques/technologies to fill this gap. However, attempts to analyze and characterize these interactions have primarily been conducted in in vitro or under buffer-defined conditions, and to some extent, in other cellular compartments that are non-natural and thus not favorable, thereby limiting the physiological insights gained. As observed in nature, many protein-protein interactions occur within the intracellular compartment of cells, particularly in the cytoplasm. The homeostatic redox balance maintained by the cytoplasm’s reducing environment promotes protein-protein interactions and is where a vast array of cellular molecules can be found. This presents an opportunity to analyze these interactions using a cytoplasmic screening platform. A potential route to empower protein co-localization and empower subsequent screening is to fuse each protein of interest (POI) to the C-terminal helical domains of yeast amino acid permeases (AAPs), which are known to dock in the plasma membrane through a palmitoylated motif, thereby bringing the proteins into proximity for interaction. Essentially, the objective is to develop a platform that closely mimics the natural environment of proteins interacting in the cytoplasm for high-throughput screening purposes. As a proof of concept, the localization of the C-terminal helical domains of AAPs to the yeast plasma membrane was studied using superfolder green fluorescent protein (sfGFP). To further test the system's capabilities, the split sfGFP construct was employed to study the interaction of two proteins, as the complementation of the GFP1-10 fragment with the GFP11 fragment is necessary for chromophore maturation and fluorescence production. We extensively studied ten different construct architectures, using a bicistronic DNA design. To achieve this, the C-terminal domains of Gap1 (Gap1C) and Alp1 (Alp1C) were selected for study. Gap1C contains an FWC tripeptide sequence, where the cysteine is palmitoylated with a fatty acid to facilitate anchoring to the phospholipid bilayer, whereas Alp1C lacks this feature. Initial designs aimed to mimic protein-protein interaction behavior in cell signaling, where a membrane-anchored or membrane-proximal protein recruits a cytoplasmic protein to interact. To do this, one fragment of split sfGFP was fused to the Gap1C domain while the other fragment was allowed to diffuse in the cytoplasm. Under these conditions, minimal fluorescence was observed with flow cytometry and confocal microscopy, suggesting that pure diffusion of an overexpressed protein in the cell cytoplasm alone is not sufficient to generate membrane-proximal binding interactions. However, when both split sfGFP fragments were fused to separate Gap1C domains, fluorescence was robust and localized mainly to the cell membrane. We next sought to study whether protein-protein interaction at the cell membrane could be achieved by fusing one fragment of sfGFP to the Gap1C domain and the other to the Alp1C domain, which lacks the FWC motif for palmitoylation but is known to localize to the cytoplasmic side of the cell membrane without incorporation. This architecture did not demonstrate robust fluorescence, but appending an FWC motif to the C-terminus of the Alp1C domain rescued fluorescence in this context, resulting in the most sensitive co-localization of split sfGFP fragments when one fragment was fused to the Gap1C domain and the other was fused to the Alp1C+FWC domain, potentially due to their behavior in lipid rafts. Expansion of this system into low-affinity dimeric fluorescent protein pairs and ongoing efforts in library screening using this system will also be discussed.
