Work in my laboratory covers a broad range of topics unified by one theme: the protein-folding problem. We use several organisms and a variety of techniques to investigate different aspects of this problem: folding mechanisms, impediments to correct folding, and the many consequences misfolding has on biological systems.
Perhaps the greatest revelation from our recent work is that protein folding has double-edged consequences. The same types of structural changes that cause dreadful diseases with some proteins can create beneficial effects with others. For example, the self-perpetuating conformational transitions of the mammalian prion protein cause fatal spongiform encephalopathies. Yet, very similar transitions in other prions serve as mechanisms for the inheritance of new, often beneficial, traits in fungi and may help to maintain stable synapses in higher organisms. Moreover, genetic and environmentally induced variations in the folding of key cellular proteins can cause cancer and contribute to neurodegenerative disease, but they can also help fuel the pace of evolutionary change. In retrospect, these revelations make perfect sense. Nature has been coping with the protein-folding problem for a few billion years; it is reasonable that it would occasionally, perhaps even often, turn it to advantage.
Prion Assembly and Inheritance
Prion proteins exhibit an unusual ability to exist in self-perpetuating structural states with altered functions. Unlike in mammals, in yeast cells prions are generally not toxic. Instead, they act as protein-only elements of inheritance. The prion state involves conversion of one domain of the protein into an amyloid fold, and this conversion occurs through an unusual molten oligomeric intermediate. We identified the critical nucleating contacts involved in the conversion of one intensely studied protein, Sup35, a yeast prion involved in translation termination. Using peptide arrays, we determined that this region also controls two previously mysterious properties of prion biology: the species barrier and the formation of distinct prion strains.
The remarkable species specificity of these interactions, and the fact that they map to regions of intermolecular contact, suggest a powerful new technology for identifying the critical contacts in a wide variety of proteins. We intend to analyze proteins associated with misfolding diseases to define residues involved in nucleation, the nature of specificities at contact sites, interacting proteins, chaperone-binding sites, and potential drug targets.
In collaboration with Kausik Si and Eric Kandel (HHMI, Columbia University College of Physicians and Surgeons), we have found that a regulatory protein that plays an important role in synaptic plasticity behaves as a prion in yeast. Cytoplasmic polyadenylation element–binding protein (CPEB) maintains synapses by promoting the local translation of mRNAs. We postulate that the self-perpetuating folding of its prion domain acts as a molecular memory. Efforts are under way to identify the structural core of the CPEB amyloid and to pinpoint regions involved in nucleating prion formation.
We have also performed a bioinformatics search, similar to that which identified CPEB, for new prions. We are now investigating the ability of candidate prions to form amyloids, to show heritability and infectivity, and to alter their functional state when their conformational state changes. What we learn from this work will enable us to extend our analysis to advanced species to open a global new perspective on the previously mysterious and arcane biology of prion proteins.
Protein Folding and Evolution
When Sup35 switches to its prion state, ribosomes start to read through stop codons. This produces a wide variety of phenotypes that differ in every genetic background. Thus, this prion uncovers hidden genetic variation that can help cells survive fluctuating environments and might provide a route to the evolution of new traits. We are studying these prion transitions in wild yeast strains, and we will use the natural genetic variation they contain to identify and genetically map alleles responsible for new traits, and the mutations that fix them.
Many important cellular proteins would fail to fold properly without assistance from molecular chaperone proteins. Among these, Hsp90 (heat-shock protein 90) is unusual in that it is specialized to promote the maturation of signal transducers, proteins regulating a multitude of processes controlling life, death, growth, and development. Hsp90 is thereby uniquely positioned to couple environmental contingency to the evolution of new traits.
Our previous work has helped to define three mechanisms by which Hsp90 might influence the acquisition of new characteristics. First, by robustly maintaining signaling pathways, Hsp90 can buffer the effect of small mutations, allowing the storage of cryptic genetic variation; when Hsp90 function is compromised, new traits appear and can be assimilated by enrichment of the underlying genetic variation in subsequent generations. Second, in the evolution of many cancers, Hsp90 chaperones mutated cell regulators that have acquired new functions but are prone to misfolding; when Hsp90 function is compromised, new traits are lost. Third, fungal mutations that confer drug-resistance phenotypes create new dependencies on Hsp90-chaperoned signal transduction pathways. When Hsp90 function is compromised, drug resistance is lost. We have discovered these powerful evolutionary mechanisms in a variety of organisms largely by serendipity, but expect them to operate in all eukaryotes. We are now investigating Hsp90's role in the rapid evolution of cancer cells and in the life cycle of the malarial parasite.
Protein Folding and Human Disease
Neurodegenerative diseases have been linked to protein misfolding and the accumulation of protein aggregates. We are using yeast models to investigate complex diseases such as Parkinson's (PD) and Huntington's (HD). Using a yeast model of PD expressing inducibly toxic levels of human α-synuclein, we have performed high-throughput chemical, genetic, and cyclic peptide screens for modifiers of toxicity. Working with the laboratories of Guy Caldwell (University of Alabama), Nancy Bonini (HHMI, University of Pennsylvania), and J.-C. Rochet (Purdue University), we have found that modifiers discovered in yeast can specifically reverse toxicity in neuronal cell cultures and animal models of PD.
To model HD, we express fragments of the human huntingtin (Htt) protein that contain long stretches of polyglutamine (polyQ) of different lengths. Expression of these fragments is toxic if the polyQ stretch exceeds a certain length, with their toxicity increasing with polyQ length beyond that point. We are using this model in screens for chemical and genetic modifiers as well, and are generating additional yeast models of protein-folding diseases. Our results suggest that this approach of expressing a toxic human disease protein in yeast will be broadly applicable to a range of neurodegenerative diseases.
Activation of the heat-shock response in a cell induces the expression of a variety of heat-shock protein chaperones that assist in primary protein folding and guard against disease-associated aggregation. Yet high expression of Hsps in tumors correlates with poor prognosis of the disease and resistance to chemotherapy. Through its chaperone activities, Hsp90 can allow cancers to evolve invasive, metastatic, and drug-resistant phenotypes. We have found that eliminating HSF1, the dominant regulator of the heat-shock response, protects mice from tumors induced by oncogene mutations. Ironically, activation of the heat-shock response is a double-edged sword in the prevention of deadly diseases. Although it can prevent the protein aggregation associated with degenerative diseases of aging, it also puts tissues at risk for cancer.
Yeast provides many advantages for high-throughput screening—speed, robustness, low cost, and genetic manipulability. This raises the exciting possibility that the power of yeast biology can be harnessed to hasten our understanding of complex disorders and to discover new therapeutic interventions.
