Multiscale study of the phenotypic consequences of protein folding intermediates in dihydrofolate reductase

点击翻译按钮获取中文摘要

Native atomic structures of many proteins are now known, but they are often just the tip of the iceberg: there are intermediate states on the way to the native state and off-pathway kinetic traps and oligomers. Intermediates have important biological consequences. Many are aggregation-prone and associated with protein deposition diseases, from Parkinson's and ALS to serpin deficiency. Intermediates are targeted by the cellular protein quality control machinery – which can lead to disease, as in cystic fibrosis. Mutations linked to congenital diseases (from ALS to Creutzfeldt-Jakob to cataracts) or somatic mutations in sporadically arising diseases (p53- deficient cancers) often shift specific proteins toward intermediate conformations. Conversely, intermediates close to the native state are targets of remarkably successful drugs. Thus, nonnucleoside inhibitors of HIV reverse transcriptase bind to a pocket absent in that protein's native structure; the drug itself induces or selects this off-native conformation. Finally, intermediates have a critical impact on biotechnology – for example, for stability and storage of therapeutic antibodies. Despite its medical significance, the space of protein intermediates remains largely unexplored. Most have low stability and tend to oligomerize or aggregate, making structure determination very challenging. The most fruitful approach has been to find variants (mutants) that stabilize a given intermediate enough for structural investigation, but this requires good mutational screens. Screens in vitro or in silico can reveal intermediates, yet at low throughput; in vivo screens are high-throughput, yet they can detect intermediates only indirectly. However, in vitro studies of protein-protein interactions (PPI) have advanced greatly in recent years due to high- throughput screening techniques. This project will integrate rapid in silico unfolding simulations, in vitro screening methods adapted from the PPI field, and in vivo measurements of bacterial fitness to investigate intermediates of an essential bacterial enzyme, dihydrofolate reductase (DHFR), which is the target of many antibiotics and the locus of many antibiotic-resistance mutations. This project has three goals. First, to detect populated intermediates in a large library of DHFR variants, combining atomistic Monte Carlo simulations and hydrophobicity fractionation of the protein library in vitro. This requires adapting a “display” method from the PPI field: barcoding each protein molecule with RNA so as to identify the fractionated variants via barcode sequencing. Second, to distinguish and stabilize intermediates populated by mutants (and thus perhaps inducible in the wild-type protein as well) using a library of conformationally selective binding partners, such as nanobodies. Third, to evaluate the effect of distinct DHFR intermediates on the fitness of E. coli cells. Accomplishing these goals will map accessible DHFR intermediates that could be new antibiotic targets. It will also provide a much needed case study on the role of protein folding intermediates on all scales: from atomistic details to fitness effects in populations.