This dissertation focuses on quantifying protein folding stability determinants and presenting initial experiments that can guide the development of a novel assay that identifies cell-penetrating miniproteins. First, despite over a century of scholarship on protein folding stability, applying this knowledge to design proteins computationally remains limited. Usually, protein designers generate many protein structures, ranging from dozens to thousands, but only a fraction of them will successfully express in E. coli and remain soluble in solution. This suggests that we still lack a fuller understanding of the determinants of protein folding stability and incorporating this knowledge into the protein design process. Addressing this challenge could increase the success rate in designing stable proteins for various therapeutic and biomedical applications, such as creating new binders and biosensors. To better understand the determinants of protein folding stability, I used a miniprotein (ɑββɑ topology, 43-residues) that was previously difficult to design as a model system. By combining computational protein design, high-throughput experimentation, and machine learning, I designed stable ɑββɑ miniproteins with greater success than in previous work. Then, I quantified how individual biophysical forces uniquely contribute to folding stability and propose a “recipe” for designing future iterations of stable ɑββɑ miniproteins. The second focus of this dissertation is to provide preliminary work that can guide the creation of a novel high-throughput screen for cell-penetrating miniproteins. Many protein-protein interactions that are implicated in disease occur in the cell cytosol, but many small molecule drugs (currently the most common class of pharmaceutical drug) are not always effective. This is because small molecules require a deep binding pocket in a protein to bind, but this is not a characteristic feature of protein-protein interfaces. An alternative to small molecule drugs is a protein-based therapeutic, but proteins do not readily cross the cell plasma membrane given the hydrophilic surfaces of proteins and the nonpolar lipid bilayer of the cell plasma membrane. Only a few proteins with cell-penetrating capabilities have been characterized (e.g. histone proteins, PTEN, zinc-finger 5.3), but this nonetheless lends credence to the hypothesis that there exists determinants for cell-penetration. Thus, a high-throughput screen could help identify cell-penetrating proteins from which we could discover general design rules for cell-penetration. Here, I show that combining intein splicing, synthetic transcription factors, and reporter gene activation in mammalian cells can signal the presence of a protein-of-interest in the cell cytosol. I propose that this system can serve as the foundation for the design of a “reporter” cell, which expresses a fluorescent protein only if a cell-penetrating protein has entered the cell cytosol. This “reporter” cell, in conjunction with a “secretor” cell, can serve as two core components in a high-throughput droplet microfluidic screen for detecting novel cell-penetrating miniproteins. I will describe in the final chapter possible directions for making this assay a reality.

Creator

• Kim, Tae-Eun

DOI

• https://doi.org/10.21985/n2-r31r-6725

Subject

• Biology
• Cellular biology
• Biophysics

Language

• en

Alternate Identifier

• http://dissertations.umi.com/northwestern:16695
• etdadmin_upload_1009800

Keyword

• protein design
• cell-penetrating protein
• synthetic biology
• protein folding stability

Date created

• 2023-01-01

Resource type

• Dissertation

Rights statement

• In Copyright