Protein misfolding is associated with numerous devastating neurodegenerative disorders including amyotrophic lateral sclerosis, Parkinson’s disease, and Alzheimer’s disease. Each of these disorders is associated with the aberrant folding and consequent aggregation of specific proteins, which accumulate into aggregates or amyloid fibrils. No therapies are available that prevent or reverse the formation of disordered aggregates, toxic pre-amyloid oligomers, or amyloid fibrils. Amyloid is a highly stable protein conformation that is resistant to degradation and denaturation. While amyloid is often associated with disease, surprisingly amyloid has also been harnessed for beneficial purposes such as adaptation to cellular stresses and genetic inheritance. For instance, yeast employ amyloid for beneficial purposes and thus amyloid assembly and disassembly is tightly regulated in yeast. We are interested in better understanding factors that regulate amyloid formation, how protein misfolding contributes to disease, and how protein misfolding can be reversed with protein disaggregases. We are also exploring these ideas in the context of liquid-liquid phase separation, and have demonstrated that disaggregases can counter pathological phase separation.
Engineering Protein Disaggregases
Amyloid is a highly stable protein conformation that is resistant to degradation and conditions that would denature most other proteins. Yet, protein disaggregases are capable of disassembling proteins from these highly stable states and restoring proteins to their native folds. In yeast, beneficial amyloids are regulated by the AAA+ protein Hsp104. Amyloid is a highly conserved secondary structure, and so the structures implicated in human disease closely resemble those Hsp104 naturally regulates in yeast. Thus, Hsp104 has weak disaggregase activity against several substrates implicated in disease including α-synuclein, polyglutamine and Aβ. We have used protein engineering techniques to further enhance the activity of Hsp104 against these and other substrates. We have developed a large set of potentiated Hsp104 variants that suppress the toxicity of these proteins, restore their proper localization, and clear preformed aggregates. We are now interested in further engineering Hsp104 to develop variants with improved properties. Just as many proteins are believed to have evolved from their roles as generalists to specialists over many years of evolution, we hypothesize that we can evolve Hsp104 from its role as a generalist that regulates the yeast proteome to a specialist tuned to recognize substrates implicated in human disease. We have recently determined the structure of Hsp104 with near-atomic resolution using cryo-EM, enabling us to also apply rational design approaches. Recently several novel protein-remodeling factors have been identified in humans, and so we are also interested in studying and engineering them.
Understanding Hsp104 Potentiation
We have uncovered numerous diverse mutations that potentiate Hsp104. These mutations are present throughout a coiled coil region of Hsp104, and the introduction of both conservative and non-conservative mutations at these positions potentiate the protein. This is surprising, as typically diverse mutations would lead to a loss rather than a gain of function. We are interested in better understanding how so many distinct mutations are capable of activating Hsp104.
Modeling Protein Misfolding
Protein misfolding has been implicated in numerous neurodegenerative and cardiovascular disorders. To better study these disorders, model systems have been generated in diverse organisms including C. elegans, Drosophila, and cultured neurons. We are particularly interested in using S. cerevisiae as a model system to study aspects of human proteinopathies. S. cerevisiae benefits from unparalleled genetic tractability, and the fact that many key cell biology pathways are conserved from yeast to humans. Indeed, many proteins that misfold and are implicated in human disease also misfold and cause toxicity in yeast. Thus, yeast models of protein misfolding implicated in ALS and Parkinson’s disease have been harnessed to conduct small molecule and genome wide screens. We are interested in developing yeast models for several other protein-misfolding disorders. These models can then be applied to conduct genetic or small molecule screens.