Huntington's disease (HD) is a progressive, familial neurodegenerative disease triggered by the expansion of a polyglutamine (polyQ) track in the protein huntingtin (htt). PolyQ sequences up to Q36 in htt are not known to be toxic, while polyQ lengths above Q36 almost invariably lead to increased disease risk and decreased ages of onset. The large number of physical states (monomers, dimers, tetramers, non-β oligomers, nanofibrils, and clustered amyloid fibrils) on the self-association landscape, with their overlapping kinetics of formation, have greatly complicated identification of the molecular species responsible for HD toxicity, drawing attention to the need for innovative approaches.After reports of HD-associated intraneuronal htt inclusions in 1997, we elucidated aggregation mechanisms of both simple polyQ sequences and the more complex polyQ-containing "exon1" fragment of htt (htt-ex1). Grounded in this work, the more recent results described here were made possible by breakthroughs in the molecular design of diagnostic polyQ derivatives and in fluorescence applications for characterizing amyloid assembly intermediates. Thus, insertion of β-turn-promoting mutations into relatively short, disordered polyQ sequences created "pro-β-hairpin" polyQs (βHPs) that exhibit amyloid formation rates comparable to the enhanced rates seen with expanded polyQ peptides. Introduction of "β-breaker" mutations into these βHP polyQ sequences created molecules that are blocked from aggregating into amyloid and also can inhibit amyloid formation by other polyQ proteins. These mutational effects were then successfully transferred into more complex htt-ex1 sequence backgrounds. Insights into the aggregation properties of htt-ex1 derivatives-as well as into the nucleation process itself-were obtained using fluorescence correlation spectroscopy (FCS) and a novel thioflavin-T (ThT) protocol that allows quantitation of htt-ex1 assembly intermediates.Using these tools, we quantified physical states of htt-ex1 at different growth times in mammalian PC12 cells engineered for inducible expression of both normal and expanded polyQ repeat length versions of htt-ex1. For expanded polyQ versions, we found tetramers, oligomers, and fibrils (but no monomers) all populated in these cells at a time when the first indication of toxicity (nuclear DNA damage) was observed. These experiments provided a strong hint that monomeric forms of htt-ex1 are not involved in toxicity, but we were otherwise unable to implicate a specific toxic self-assembled state because of the overlapping kinetics of formation. To gain a more intimate focus and control over the timelines of htt-ex1 self-assembly and the resulting toxic response, we engineered various htt-ex1-βHP molecules-with and without added β-breaker mutations-that could be expressed in rat neuronal and Drosophila models of HD. In both models, novel htt-ex1-βHP analogues exhibiting strong aggregation in spite of their very short polyQ repeat lengths proved to be toxic, dramatically breaking the "repeat length paradigm" and strongly suggesting that the toxic species must be some kind of aggregate. In both models, β-breaker analogues of htt-ex1-βHP that are slow to make amyloid-instead favoring accumulation of non-β oligomers-were nontoxic. In contrast, htt-ex1-βHP analogues that rapidly progress to amyloid states were toxic, suggesting that an aggregate possessing the fundamental amyloid folding motif is very likely the major toxic species in HD.