(160t) Leveraging Computational Stability and Ancestral Sequence Reconstruction As a Platform for the Site-Wise Diversification of the Haloalkane Dehalogenase Ligand Binding Pockets

Self-labelling fusion tags provide versatility in a myriad of protein screening, isolation, and manipulation techniques. Despite their utility and irreversible interaction with a number of tag-specific ligands, the range of photophysical and kinetic, structurally dependent properties, and applications of these ligands vary greatly. For example, the Janelia Fluor (JF) small molecule fluorophore (SMF) JF646 exhibits excellent visualization of small characteristics in cellular environments, but exists largely in the inactive form, conferring low bioavailability with a K_L-Z value between 0.0012-0.0014. When bound to a fusion tag, the low background, high absorbance localization is substantial enough to visualize and distinguish between miniscule cellular features when utilizing multiple JF options. In practice, however, the delicacy of ligand-fusion tag interaction under large, diverse, and in many cases unstable microenvironments, such as those in multi-cellular organisms, is detrimental to the efficiency of in vivo imaging with these systems. Consequently, there is a need to streamline fluorescence imaging towards improved performance under in vivo conditions, altering ligand functionalities without disrupting the fine-tuned structure of SMFs. In this case, effort must be redirected towards engineering the fusion tag itself. The fine, balanced photochemical properties of SMFs experiences sharp trade-offs when engineering improved functions. Instead, engineering ligand binding regions in fusion tags has the potential to simultaneously improve the ligand interaction for a variety of SMFs. Using a haloalkane dehalogenase, engineered through directed evolution for covalent interactions with JF646 and other SMFs, we deploy computational stability calculation and ancestral sequence reconstruction (ASR) to locate residue positions and amino acids that favor structural integrity and SMF interaction within this binding pocket. Using multiple high-resolution haloalkane dehalogenase structures, comparative analysis of stability changes upon mutation have enabled the design of a degenerate codon library that implements mutually stabilizing amino acids at amenable positions from a collection of computational datasets. The complementary analysis of computational stability datasets from FoldX and Rosetta stability prediction software enables refined design for site-wise diversification, correlating more closely with stability measurements in experimental data. For ASR, phylogenetic predictions of haloalkane dehalogenase multiple sequence alignments facilitate an examination of distinct evolutionary trajectories based on functional ancestral sequences. These stable ancestors, containing high node support values, are characterized by variable posterior probabilities of amino acids. When applied to directed evolution principles, our predictions provide a stable sequence and amenable positions to aid in library design. Even more, by inheriting the modification scheme of engineered HaloTag haloalkane dehalogenase that enable covalent interactions with ligands, we can explore how highly similar sequences tolerate the success of previous directed evolution endeavors. Using yeast surface display, a naïve library of up to 10⁹ individual clones can be screened and sorted by flow cytometry, followed by further maturation and characterization of lead variants. Here, we initially construct a library diversified at 7 sites using 8 amino acids predicted to be stabilizing at those sites, thoroughly avoiding predicted-to-be destabilizing sites and amino acids. By tapping into a rich source of high-resolution structural data and stability prediction software, free-energy-based, bottom-up library optimization allows much greater flexibility in exploring structure-function relationships while engineering these systems. By leveraging computational platforms for library design, we explore the haloalkane dehalogenase binding pocket through a site-wise, ancestral interrogation of functional diversity. This work will provide useful insight into the active site residues and specific sites that influence evolutionary diversity and conservation of this protein. With a focus on optimizing the active site, which has enabled advanced molecular imaging with SMFs, we hope to discover the mode by which compact interactions are enhanced by proximal, structure-altering pocket optimization within this system.