Antibodies and antibody-like molecules are broadly used as biochemical reagents and therapeutics. They are highly valued as probes that can distinguish between protein targets with extraordinary molecular detail. A deep understanding of natural antibody structure and function has enabled the development of fully synthetic human antibody libraries for in vitro display. These synthetic libraries recapitulate the sophistication of molecular recognition in antibody complementarity determining regions (CDRs) and in vitro display permits exquisite control over selection conditions. Thus, protein engineers can tailor-make antibodies to bind challenging and non-conventional targets.Here, we describe three examples of protein engineering that leverage synthetic antibody libraries and in vitro display technologies to generate binders to novel epitopes. In Chapter 1, we utilize recombinant antibody pairs to target the post-translational modification (PTM) phosphotyrosine (pY) in folded protein epitopes. In Chapter 2, we develop and deploy a single-domain antibody (sdAb) library to rapidly identify inhibitors to SARS-CoV-2 viral entry. In Chapter 3, we utilize chemo-epitope specific sdAbs to create small-molecule-dependent switches and further engineer them to control cellular therapies.

Caspases are a family of dimeric thiol proteases central to the inflammatory and apoptotic pathways. Structural studies of the major inflammatory caspase, caspase-1, reveal that the enzyme exists in two states: an on-state when the active site is occupied; or an off-state when the active site is empty or when the enzyme is bound by allosteric ligand at the dimer interface. Determining which residues and regions of the enzyme are responsible for determining enzyme conformation and for regulating transitions between states would elucidate the molecular basis for allostery in caspase-1.

A network of 21 hydrogen bonds from 9 side chains connecting the active and allosteric sites change partners when going between the on- and off-states. Alanine-scanning mutagenesis of these 9 side chains shows that only two of them, Arg286 and Glu390, which form a salt bridge, have major effects causing 100- to 200- fold reductions in catalytic efficiency (k_cat/K_M). More detailed mutational analysis reveals that the enzyme is especially sensitive to substitutions of the salt bridge: a homologous R286K substitution causes a 150-fold reduction in k_cat/K_M. Thus, side chains from only a small subset of the larger H-bonding network are critical for activity. These form a contiguous set of interactions that run from one active site through the allosteric site at the dimer interface and on to the second active site. This subset constitutes a functional allosteric circuit or "hot-wire" that promotes site-to-site coupling.

Further experiments demonstrate the utility of using site-directed mutagenesis and chemical ligands to select for and examine enzyme conformation. Targeting specific glycine and proline residues in caspase-1 for mutation selects for different conformational states of the enzyme. We present a method of using covalent inhibitors for caspases to create half-labeled hybrid constructs that trap the conformational state of caspases in solution and allow us to understand how allostery plays a role in caspase activity. We present a model that provides a mechanism by which the very conformational state of the various caspase family members recapitulates their role in specific biologic pathways and provides an additional level of regulation.

Cancer immunotherapies, or treatments that utilize a patient’s own immune system to fight cancer, have been met with a great deal of success. Although most current immunotherapies have focused on boosting the adaptive immune system, there has been a growing interest in targeting components of the innate immune system. An important member of the innate immune system is the natural killer (NK) cell. Like T cells, NK cells are cytotoxic effector cells with the ability to distinguish and eliminate virally infected or transformed cells. They play a key role in the anti-tumor response and are able to identify target cells by recognizing stress-induced ligands that are frequently overexpressed on cancer cells. NK cells are best known for their ability to perform antibody-dependent cellular cytotoxicity (ADCC), a mechanism that is used by several therapeutic antibodies to eradicate tumor cells. Whereas all T cells can be targeted through the T cell receptor (TCR)/CD3 complex, NK cells have multiple activating, costimulatory, and inhibitory receptors that can be targeted to modulate their activity. This dissertation provides a strategy used to identify antibodies that can target and stimulate NK cells and the development of bispecific antibodies to redirect NK cell-mediated cytotoxicity towards tumor cells of interest. Chapter 1 discusses the background of NK cells and their role in anti-tumor immunity. Chapter 2 describes a mass spectrometry-based screening approach to identify cell surface proteins on NK cells. Chapter 3 provides the method designed to identify antibodies that are able to stimulate NK cell-mediated cytotoxicity and the generation of bispecific NK engagers.

Apoptosis, a type of programmed cell death, is a universal and essential cellular function. There are numerous homeostatic biological roles of apoptosis and many diseases have or cause mys- or dis- regulated apoptosis, most notably cancer’s escape from apoptotic signals. Apoptosis has many distinctive characteristics including membrane blebbing, chromatin condensation, and most importantly for these studies, the activation of a class of protease proteins called caspases. Caspases are cysteine aspartic proteases with a unique preference to cleave hundred of substrates after acidic residues, especially aspartate. These cleavage events can activate, modify, or inhibit the substrate’s function, which leads to the dismantling of the cell during apoptosis. This process is well conserved throughout metazoan evolution, with parallel pathways in coral, fish, flies, worms and mice.

To study apoptotic protease activity, the Wells Lab has developed a proteomic-based technique. This technique is an unbiased positive enrichment labeling mass spectrometry protocol. While the protocol was developed for studying caspase substrates during apoptosis in human cell culture, it is very versatile, allowing for use in almost any protein sample like perturbed cellular lysates, different species and primary samples. Chapter 1 covers the protocol specifics and uses, summarizing a decade’s worth of optimization and application.

The DegraBase is the compilation of 44 different experiments using the N-terminal labeling method to examine apoptotic proteolytic activity described in Chapter 2. While much of the experimentation work was completed before I started the project, I completed compilation and standardization of raw data, and worked with Emily Crawford on the analysis. This global analysis reveals there is a large increase in proteolytic activity after apoptotic induction, and caspases account for 25% of the newly created fragments. Within caspase substrates, there is no single biological process or sub-cellular location that is targeted, as caspases appear to cleave substrates throughout all the different pathways of the cell. Additionally, this database is also a good resource for endogenous proteolysis, including free methionines, and signal and transis peptide processing.

Analysis of the DegraBase also reveals evolutionary and biological discoveries. As the apoptotic pathway to activate caspases is highly conserved throughout metazoans, we wanted to investigate the conservation of caspase substrates. In Chapter 3, the comparison of caspase substrates in worm, fly, mouse and human reveals a hierarchal structure. My contribution includes the murine dataset and comparison analysis in collaboration with Emily Crawford. We found caspase cleavage is highly conserved at the pathway level, while individual targets and sites are not as well conserved the more distant the animals.

The unbiased and large size of the DegraBase also reveals broader caspase activity than previously described. Caspases had been assumed to have absolute specificity for aspartate and no activity for glutamate. However, with the large dataset, in Chapter 4, I reveal significant activity after glutamate and even potential activity after phosphoserine. This activity is verified through biochemical assays and x-ray crystallography, and expands the number of apoptotic caspase substrates by 15%.

As many chemotherapeutics induce apoptosis, apoptotic, especially caspase, proteolytic substrates may be good biomarkers of treatment efficacy. The development of mouse models and a positive control are discussed in Chapter 5. These models utilize the DegraBase and labeling technology to identify peptides enriched in the blood and tumor specific to treatment responding mice as potential biomarkers.

Over the last two decades, a major focus in the biomedical research community has been placed on understanding the interaction between the human immune system and cancer. This focus has led to a considerably improved understanding of how cancer can evade immune detection, and as a consequence, the emergence of a novel class of cancer therapeutics that can trigger the immune system to recognize and eliminate previously undetected cancers. These new therapies, known as cancer immunotherapies, have resulted in unprecedented and remarkably durable outcomes in the clinic. However, there are key limitations that continue to hinder how broadly immunotherapies can be applied to treat human cancers. The research in this thesis describes efforts to apply antibody engineering as a tool to address some of these limitations.

Antibodies that recognize tumor specific antigens can serve as guides to target the immune system directly to cancer cells. However, for many cancer types, there are currently no known tumor specific antigens. This paucity of available targets continues to present a major challenge in the immunotherapy field. Chapter 1 describes the development of a technological pipeline that coupled proteomics to antibody discovery in order to discover and begin to validate novel tumor specific antigens on the cell surface proteome of Ras driven cancers. Ras driven cancers represent nearly a third of all human malignancies. The work in this chapter presents a new series of molecular targets that could potentially be leveraged to target the immune system to this large class of difficult to target cancers.

Another major challenge in the immunotherapy space is that the treatments often work too well. The result is rapid hyperactivation of the immune system, often leading to horrible side effects faced by patients including severe fevers, neurotoxicity, and sometimes death. A prevailing thought in the community is that a more gradual activation of this immune response might ameliorate side effects while maintaining efficacy. Chapter 2 describes the development of a new class of molecular switches called Antibody-based Chemically Inducible Dimerizers that enable immune cell activation to be precisely regulated by administration of small molecule drugs.

The development of engineered biologics as both tools and therapeutic molecules have transformed the scientific landscape tremendously. Due to their increased specificity, potency, bioavailability and decreased immunogenicity, antibodies have become a popular scaffold for drug development. In 2023, antibodies accounted for five of the top ten selling drugs in the US. In addition to their therapeutic purposes, the high specificity of antibodies allows them to be powerful tools for studying and modifying biological systems. Antibodies are now regularly used to modify cellular signaling pathways, block protein interactions, detect a specific antigen of interest, among many other applications.Though our body naturally produces antibodies to fight off infection, the ability to generate synthetic antibodies against a chosen antigen of interest gives scientists the opportunity to expand the scope of druggable targets. A library consisting of up to 1011 fully synthetic antibodies can be displayed on systems such as phage, yeast, or mammalian cells, then screened for binding against an immobilized antigen. If binders are identified from the pool, they are isolated and sequenced to obtain information about the antibody’s amino acid combination for further production and characterization. These antibodies can then be optimized and applied in various formats for downstream biological experiments. Here, I describe three approaches for using antibody and protein engineering to uncover novel information on biological systems at the protein or cellular level. First in chapter 1, we use naïve antibody discovery to identify functional antibody modulators of protein arginine deiminase (PAD4), a protein highly implicated in the development of autoimmune diseases. In chapter 2, we develop a PAD4 fusion protein to target the enzyme to HER2+ expressing cancer cells and identify the substrate scope of PAD4 at the cell surface. In the last chapter, we create an antibody-drug conjugate using site-specific bioconjugation for targeted degradation of challenging cell surface proteins.

Immunotherapies such as adoptive cell therapies and checkpoint inhibitors are curing previously lethal cancers, however only a small fraction of patients are responsive. A major predictor of the response to immunotherapy is the extent of tumor infiltrating lymphocytes (TILs) in the tumor microenvironment. Unfortunately, most solid tumors exclude TILs through various immunosuppressive mechanisms and are classified as immune ‘cold’ tumors. In this thesis research, we studied how the expression of dominant oncogenes broadly modulates immune evasion phenotypes across cancer models. We translated those findings into therapeutic antibodies that stimulate innate immune pathways to turn immune ‘cold’ tumors ‘hot’ and improve response rates to immunotherapies.In Chapter 1 we used discovery proteomics to characterize isogenic models of driver oncogenes (Myc, KRAS, BRAF, MEK, AKT, HER2, EGFR) and identify common axes of dysregulation. From an unbiased approach, the most ubiquitous and dramatic molecular and functional phenotype across oncogenes was the suppression of proteins regulated by type 1 interferon, particularly antiviral dsRNA sensors. These effects were validated in patient- and cancer-derived tumor cells driven by KRAS and/or Myc oncogenes. These results have broad implications for standard therapies like radiation, epigenetic and cytotoxic drugs, and immunotherapies that require type 1 interferon and antiviral pathways for their cytotoxic effects, and suggest opportunities for oncolytic and gene-therapy viruses. In Chapter 2 we used in vitro display technologies and protein engineering to build therapeutic antibodies that re-activate innate immune signaling in the tumor microenvironment. ENPP1 is an extracellular phosphodiesterase that hydrolyzes cGAMP, an immunostimulant that is transported between cells to activate type 1 interferon. ENPP1 is frequently up-regulated on tumor cells, and small molecule drugs have shown inhibiting ENPP1 stimulates type 1 interferon signaling and sensitizes immune ‘cold’ tumors to radiation and immunotherapy. Here, we generated first-in-class variable heavy (VH) single-domain antibodies that bind and allosterically inhibit ENPP1. The VH domains were recombinantly engineered into multivalent formats and immunotherapies that improved their selectivity and potency. A cryo-EM structure of the VH-ENPP1 complex revealed its allosteric inhibitory epitope and a novel mechanism of substrate-selective inhibition.