(575a) Length and Location: What Matters Most in PolyQ Protein Aggregation

Proteins with abnormally long polyglutamine (polyQ) domains arise in nine inheritable progressive neurodegenerative disorders in which the trinucleotide CAG is abnormally expanded. These diseases are characterized by loss of motor function and, eventually, cognitive function. A broad polymorphism of polyQ length is observed; the length is inversely correlated with the age of onset and severity of symptoms. In all cases the proteins aggregate into intracellular inclusions; furthermore, aggregation is believed to confer a toxic gain-of-function that leads to death of motor neurons. There remains controversy over the nature of the toxic aggregates: specifically, whether soluble non-fibrillar oligomers or mature fibrils are toxic. The proteins associated with each disorder are distinct and have no sequence or structural homologies other than the expanded polyQ domain. They range in size form 200 to over 3000 residues; several are large multi-domain proteins whereas others are stable subunits in protein complexes. Most are predominantly alpha-helical but also have a substantial amount of disorder, with the polyQ stretch often located in an unstructured region. In large measure, X-ray or NMR structures for the full-length proteins are not available, and in many cases the disease proteins are difficult or impossible to produce recombinantly. Our aim is to examine systematically the influence of polyQ domain length and location, within the context of a folded protein, on structure, stability, and aggregation. Because of the aforementioned difficulties in expressing disease proteins and the lack of structural information, we chose a generic host protein, apomyoglobin, for our studies. This protein is readily expressed and folded, its native structure is a mix of helix and disordered and thus reflects the major secondary structural elements of many disease-associated protein, and much is known about its structure and folding kinetics. We generated a library of mutant apomyoglogin proteins containing variable-length polyQ inserts either at the N-terminus (Qx-NT, where x varies from 8 to 38) or in a disordered region between the C and D helices (Qx-CD). We carried out extensive comparative characterization of the folding, stability, and aggregation properties of the mutants. Surprisingly, we found that although both length and location influenced these properties, location mattered more. When the shortest polyQ domain was placed internally (Q8-CD), there was no change in the native apomyoglobin secondary structure and the inserted polyQ was disordered. With the longer polyQ domains placed in the loop region (Q16-CD or larger), the native helices adjacent to the loop were lost, and the resulting insert was mixed disordered/β-sheet. For all Qx-CD, we observed partial loss of the heme binding pocket and a modest decrease in stability at low urea concentrations. All Qx-CD aggregated over time, independent of length, and the aggregates formed possessed a similar morphology: semiflexible linear aggregates that were nonfibrillar. Interestingly, when a control protein was produced in which polyQ was replaced with a glycine-serine repeat (GS8-CD), we observed nearly identical changes in secondary and tertiary structure, folding stability, and aggregation. These results suggest that the dominant factor driving aggregation is not specifically polyQ-driven but is attributable to perturbation of the native structure. In sharp contrast, when polyQ domains were placed at the N-terminus, we observed an increase in helical content and no loss of tertiary structure or stability in urea. Furthermore, some aggregates were observed but the aggregates were clearly fibrillar in morphology, with subtle differences caused by length of the polyQ domain. Taken together, these results demonstrate the dominant effect of location over length of polyQ domains on protein folding and aggregation.