Protein engineering to modulate dynamics of molecular machines

Proteins are macromolecules that drive nearly every biological process, including cell metabolism, signaling, and structural support. While sequencing genomes has helped decipher the primary amino acid sequences of countless proteins, it is ultimately the three-dimensional fold that determines a protein’s function. Experimental methods and, more recently, computational tools enabled to solve or predict the folds of many proteins, contributing to our understanding of the relationship between protein fold and function. Altering protein function or creating entirely new functions requires generating proteins with new folds, for which the encoding amino acid sequences are unknown. Protein engineering is the discipline that seeks to discover new sequences that fold into proteins with desired function. As methodologies have evolved, moving from evolution-inspired approaches to rational and de novo design, protein engineering has gained significance for fundamental research and therapeutic applications. This thesis explores protein engineering as a tool, both for dissecting bacterial cellular machinery and for probing potential antibacterial mechanisms. It comprises two independent, yet complementary projects that highlight the power and versatility of protein engineering.
The first project centered on the nanomachine Type VI secretion system (T6SS), a dynamic protein complex found in many diderm bacteria that functions as a molecular speargun to inject protein effectors into target cells. T6SS relies on a contractile sheath structure that undergoes significant conformational changes to deliver effectors, followed by disassembly and recycling of the sheath by an unfoldase enzyme. This recycling is highly selective: the unfoldase only recognizes and disassembles the sheath when it is contracted. While it was assumed that this selectivity relies on T6SS structural changes exposing a known recognition motif for the unfoldase, the precise mechanism enabling the unfoldase to distinguish sheath conformations remained unclear.
To address this, I analyzed structural differences between the sheath in its different conformations. These studies pinpointed specific residues of the sheath to interact upon contraction, potentially exposing previously hidden sites necessary for unfoldase recognition. By targeted mutagenesis, I engineered functional T6SS variants that either prevented or allowed unfoldase binding independently of the sheath’s conformation. Using fluorescence microscopy, I tracked the nanomachine’s dynamics in real time, observing whether the unfoldase colocalized with the T6SS sheath. Functional assays provided insights into how these mutations impacted effector delivery efficiency and recycling, while structural analyses indicated specific structural changes essential for unfoldase recognition. Together, these studies allowed me to develop a mechanistic model explaining that certain structural changes beyond exposure of the known unfoldase recognition motif regulate the specific interaction between contracted sheath and unfoldase that allows recycling of sheath.
The second project focused on the design of protein-based binders targeting two critical bacterial proteins that play essential roles in maintenance of cell envelope integrity. The first of these is Penicillin-Binding Protein 2a (PBP2a), a target for antibiotic development in Methicillin-resistant Staphylococcus aureus (MRSA). PBP2a, which enables bacterial resistance to certain β-lactam antibiotics, has traditionally been targeted at its active site. However, my goal was to develop a protein binder that could inhibit the protein by interacting with alternative, previously unexplored sites. Using modern computational protein design methods, I generated and refined binders with nanomolar affinities for PBP2a. In vitro competition assays with traditional antibiotics revealed only minimal competition for the active site, confirming a new mechanism of target binding. In cell-based assays, these binders significantly reduced MRSA viability, indicating a promising treatment approach that differs from traditional active site inhibition.
The second protein target, BamA, is essential in Gram-negative bacteria and is involved in outer membrane protein biogenesis. BamA is challenging to target effectively due to its structural flexibility, which allows it to adopt multiple conformations essential for its function. To inhibit BamA, I computationally designed protein binders intended to stabilize BamA in a non-functional state, which should result in bacterial lysis. Experimental evaluation of these binders is pending.
Taken together, by explaining regulatory mechanisms of bacterial nanomachines and exploring alternative antibacterial strategies, this thesis provides insights into the utility of protein engineering as a research tool to study complex fundamental processes and probe therapeutic strategy. These and other emerging protein engineering methods will provide novel tools to re-engineer native proteins but also design new proteins and protein complexes with advanced functions enabling applications across medicine, biotechnology, and materials science.