The ribosome is the most complex molecular machine known to man. It is an ancient behemoth, a twisted megadalton mass of RNA and protein, whose molecular motions underly the translation of the genetic code into language of proteins. Like a leviathan, that great mythological sea monster, the ribosome looms large as the central figure in the imagination of the cellular stew. Huge expenditures of energy are dedicated to its critical task: sequence-defined polymerization of proteins with remarkable speed and fidelity via amide-bond formation. Proteins, composed of combinations of a mere twenty-two α-amino acids, give rise to the myriad biological functions we ascribe to life. What if those polymers were instead composed of other monomers? What if the ribosome and its adjacent machinery could be convinced to catalyze other types of bonds, such as carbon-carbon (C-C) bonds? What functions and materials could follow from an expanded molecular repertoire?

The first chapter of this dissertation surveys a body of work surrounding the translation of noncanonical backbone monomers and formation of non-amide bonds by the ribosome. I explain the key technologies that enable a sequence-defined polymerization of unusual monomers and novel bond formation in the ribosome, while also exploring current challenges, such as C-C bond formation, and important applications of the field, such as peptide drug discovery.

In Chapter 2, I describe efforts towards C-C bond formation which include monomer synthesis, tRNA-acylation, and in vitro translation. In order to explore the fundamental features of acyl-transfers onto carbon nucleophiles, a chemical model system was developed an a series of monomers screened.

In Chapter 3, I provide an extended account of a chemical and ribosomal strategy for the synthesis of peptides bearing embedded heterocycles that support atropisomerism and macrocyclization. This includes the discovery of new carbonyl-containing monomers for translation and chemistry for their transformation into diversified, atropisomeric, and macrocyclic peptides with embedded quinolines.

Escherichia coli ribosome engineering has emerged as a powerful approach to expand protein sequence-space to include monomers that are not α-amino acids. However, current ribosome engineering efforts are complicated by the enormous size and complexity of the ribosome—a 2.8 megadalton complex composed of two subunits (50S and 30S subunits), three ribosomal RNAs, and at least 54 ribosomal proteins. Peptide bond formation is catalyzed within the peptidyl transferase center (PTC), located within the 50S ribosomal subunit. The generation of a miniaturized PTC that recapitulates the native PTC fold and function will aid a rapid screening of ribosomal mutants that support new polymerization chemistry beyond the peptide bond. As a first step toward this effort, we seek to address a fundamental question: what is the smallest element of RNA capable of both binding acylated tRNA and catalyzing bond formation? The development of a minimal PTC that recapitulates many features of an intact ribosome is described in chapter 2 of this dissertation.In addition to the minimization approach, ribosome engineering efforts can be accelerated by the development of aminoacyl-tRNA synthetase (aaRS) variants that efficiently generate acyl-tRNA with non-α-amino acids in vivo. Traditional selection experiments to evolve aaRS mutants require ribosomal translation, making this selection system unsuitable if the desired non-L-α-amino acids are poorly tolerated by the ribosome. Therefore, a new selection strategy independent of ribosomal translation is required to ensure the rapid selection of new aaRS variants that recognize novel monomers. T-box riboswitches are RNA regulatory elements that sense tRNA acylation to regulate downstream gene expression in gram-positive bacteria. We envision that T-box riboswitches could be used to regulate gene expression in the selection of aaRS variants instead of ribosomal translation. Initial efforts toward T-box engineering are described in chapter 3 of this dissertation.

True molecular understanding of cellular and chemical biology requires the characterization of cellular structures with high spatial resolution in living cells. Over the past two decades, super-resolution microscopy (also called ‘nanoscopy’) has emerged as a powerful tool for the visualization of cellular structures with nanoscale resolution. The efficiency of nanoscopy; however, is limited by the fuel—the chemical tools that support the vast lists of requirements for stimulated emission depletion (STED) and single molecule localization microcopy (SMLM). This dissertation describes worked pursued during my graduate degree towards the development of new and unique chemical tools to expand the capabilities of live cell nanoscopy and accordingly advance cell and chemical biology. Chapter 1 provides an introduction to and in-depth discussion of relevant concepts (e.g., nanoscopy, spontaneously blinking fluorophores), ultimately concluding with an overview and analysis of a class of dual component chemical tools developed by our lab termed high density environmentally sensitive (HIDE) probes. Chapter 2 describes the application of HIDE probes to multicolor imaging via simultaneous development of a mutually orthogonal bioorthogonal labeling platform and development of a novel stimulated emission depletion (STED)-compatible rhodamine fluorophore. Chapter 3 describes the development of novel bright and near-IR emitting spontaneously blinking fluorophores for two-color live cell single molecule localization microscopy (SMLM) without chemical additives and with only a single 642 nm laser. Chapter 4 briefly summarized a series of projects that were designed, but not executed during my time in the lab.

The Epidermal Growth Factor Receptor (EGFR), also known as ErbB1 or HER1 in humans, is a member of the ErbB family. EGFR is a receptor tyrosine kinase (RTK) responsible for communicating cellular signaling across the plasma membrane related to cell proliferation, metabolism, and migration. EGFR mutations and overexpression are known to be related to cancer. Thus, understanding EGFR biology and the molecular mechanism of EGFR activation can lead to important therapeutic strategies. The importance of the juxtamembrane (JM) domain of the EGFR has gathered attention as evidence showed that its intermolecular and intramolecular interactions are crucial for EGFR function.In this dissertation, I describe my graduate work in four chapters detailing the study of the EGFR JM function and its role in regulating the downstream biology of the EGFR. Chapter 1: This chapter builds on the literature review and previous efforts, introducing the role and importance of EGFR in cellular biology, and the role of the JM domain in EGFRactivation. In this chapter, JM structure and the manipulation to affect EGFR signaling were described. Chapter 2: This chapter describes peptides mimicking the JM domain dimer in activated EGFR used for high-resolution structural studies. The designs of these JM-mimicking peptides were detailed. Peptide synthesis, purification, and characterizations that reveal the secondary structure were discussed. Chapter 3: Looking beyond the JM domain, the structural study described in this chapter focused on peptides containing both the transmembrane domain (TM) and the JM domain segment to investigate the allosteric relationship within EGFR. Preparation and characterization of membrane mimics that stabilize TM-containing peptides were reported. Chapter 4: In the final chapter, I report a collaborative work I partake that reveals the effect of JM structure and conformation on EGFR biology in terms of cellular trafficking and receptor degradation. Another preliminary work using APEX2-based proximity labeling toward mass spectrometry proteomics was reported. Finally, this chapter summarized the outlook on future directions following the groundwork laid out in the scope of this dissertation.

Despite the growing interest in protein-based therapeutics, one of the major limitations towards their development is the fact that proteins cannot simply diffuse into the cell interior. Instead, most proteinaceous cargo must be taken up via the biological process of endocytosis. Proteins become engulfed by the endocytic vesicle that travels into the cell and many of these protein cargoes remained trapped within endosomes. While some protein therapeutics, such as Cerezyme and Fabrazyme, take advantage of this entrapment to exhibit their functions within the endocytic pathway, the inability to access the cytosol and other organelles within the cell prevents development of intracellular-targeting proteins. A number of delivery vehicles have been proposed, including liposomes, nanoparticles, polymers, and cell-penetrating peptides (CPPs). Unfortunately, each of these techniques must be optimized for their respective cargo and target and may still suffer from overall low cytosolic delivery. Cell-permeant miniature proteins (CPMPs) provide a promising solution as they reach the cytosol intact at high concentrations, have a defined mechanism, and can deliver protein cargo.

Here, I describe progress from the evaluation of CPMPs alone to delivery of therapeutically relevant proteins and enzymes in both mammalian cells and mice. First, I present a summary of CPMP development, from designing miniature proteins that installed structural elements of traditionally flexible CPPs, to describing mechanistic details of endosomal uptake and escape, to delivering enzymatic cargo to the liver of mice. Next, I describe fluorescence correlation spectroscopy (FCS) in conjunction with flow cytometry experiments for monitoring proteins in live cells and evaluating endosomal escape. This methodology is followed by FCS studies that establish diffusion properties of free dye and CPMPs to assess whether intact CPMPs reach the cytosol of Saos-2 cells. The following chapter illustrates that among a panel of CPPs and CPMPs, the CPMP ZF5.3 is a superior delivery vehicle for a model cargo in multiple cell lines. Next, I describe the development and use of ZF5.3 conjugated to argininosuccinate synthetase (ZF-AS), an enzyme involved in an inborn error of metabolism. ZF-AS was the first ZF5.3-containing molecule to participate in animal studies to evaluate clinical utility of CPMPs. To continue improving cellular delivery of therapeutically relevant proteins, I describe functional and cellular uptake studies in which ZF5.3 is appended to the well-studied endonuclease Cas9 and initial development towards ZF5.3 conjugates of NS1, a monobody that binds Ras proteins with high affinity. Finally, I conclude with remarks on CPMP-mediated delivery. Together, these projects demonstrate that the CPMP ZF5.3 is a superior delivery vehicle compared to canonical and cyclic CPPs, illustrate ZF5.3-mediated delivery of three protein cargos, and support further ZF5.3 clinical development as a tool for protein-based therapeutics.

Macromolecular therapeutics like proteins and RNAs have immense potential to impact human health. Relative to small molecules, their architecture is significantly more complex and structurally diverse, enabling high-affinity and high-specificity interactions in the crowded cell interior. However, macromolecules (or biologics) are also typically charged and hydrophilic, and thus cannot directly cross a lipid bilayer. This results in cellular uptake through active pathways such as endocytosis which result in significant degradation unless the macromolecule can escape an endosome. Endosomal escape has been a bottleneck to the clinical application of biologics for decades, spurring the development of a multitude of delivery vehicles and assays to optimize them.

Here, we begin by providing a detailed review of the current state of the protein and RNA delivery field. We discuss techniques to establish absolute quantitation of exogenously delivered macromolecules in discrete locations within the cell, as well as an overview of delivery modalities that promote cytosolic and nuclear delivery of a variety of biologics. Next, we present a thorough interrogation of the mechanisms governing delivery of a cell-permeant miniature protein designed by the Schepartz lab called ZF5.3. Using a highly quantitative technique called fluorescence correlation spectroscopy (FCS), we establish molecular design rules for optimal cargos that expand the therapeutic scope of ZF5.3 and provide insights into its endosomal escape pathway. These findings are followed by the development of a highly sensitive tool to track the intracellular trafficking of ZF5.3. We exploit the increasing acidity of the endocytic pathway to design an assay that uses fluorescence lifetime as a readout for the pH microenvironment of ZF5.3. We apply this assay toward tracking both endosomal maturation and cytosolic localization of ZF5.3 to gain insights into the relationship between pH and intracellular delivery. Finally, we apply FCS to the nuclear delivery of a high-impact biologic, the Cas9 ribonucleoprotein complex (RNP), to establish a quantitative relationship between nuclear concentration and gene editing outcomes in multiple mammalian cell lines. We conclude with an outlook on the future of macromolecular delivery and the challenges that lie ahead. Together, the projects presented in this dissertation will contribute to the growing understanding of intracellular delivery pathways and provide a framework to optimize and quantify delivery strategies for biologics

Cells biosynthesize sequence-defined polymers of 20 canonical α-amino acids known as proteins, while chemists synthesize polymers of diverse monomer structures but with limited control over sequence and length. Expanding the monomer repertoire of the cellular translation system beyond α-amino acids could produce novel sequence-defined polymers with never-before-seen properties. Towards this end, various components of the translation apparatus require engineering to recognize non-native monomers for polymerization. In this dissertation, I describe my work to identify new aminoacyl-tRNA synthetase variants that acylate tRNAs with non-α-amino acids, the first step in cellular translation, and employ the variants to translate peptides containing new monomers. Screening known variants of pyrrolysyl-tRNA synthetase for in vitro acylation activity against a panel of non-α-amino acid monomers led to the discovery of the first α-thio acid, N-formyl-α-amino acid, and α-carboxyl acid synthetase substrates. The crystal structure of a pyrrolysyl-tRNA synthetase variant, MaFRSA, bound to an α-carboxyl acid substrate revealed a defined network of hydrogen bonds used to recognize the second substrate carboxyl group. The pyrrolysyl-tRNA synthetase variants supported translation of peptides bearing α-thio acid, N-formyl-α-amino acid, and α-carboxyl acid monomers at the N-terminus in vitro and translation of a model protein containing an α-hydroxy acid in vivo. To develop a T-box-based translation-independent selection strategy to evolve aminoacyl-tRNA synthetases for non-α-amino acids, I tested orthogonal tRNA recognition by several candidate T-boxes. Translation-dependent selection of pyrrolysyl-tRNA synthetase variants identified initial hits for acylation of several α-hydroxy acids and additional rounds of selection are ongoing.