Simple Polyglutamine Models

1.4.3.1 Rationale for Simple PolyQ Models As previously discussed, a wide range of seemingly unrelated proteins are affected by polyQ repeat expansion: the androgen receptor (Spinobulbar muscular atrophy), the Ca_V2.1 P/Q voltage-dependent calcium channel (Spinocerebellar ataxia type 6), the TATA-binding protein (Spinocerebellar ataxia type 17), huntingtin (Huntington’s Disease), and six others. The lowest common denominator between these proteins is their polyQ domains, which causes neurodegenerative phenotypes in a repeat length dependent manner. In each instance, expanded polyQ domains also gain a propensity to aggregate into amyloid in vitro, in vivo, and in human patients ^6–8. Additionally, expanded polyQ domains alone aggregate and are toxic when expressed in C. elegans ¹⁷⁴, Drosophila ⁷⁸, when simple polyQ aggregates are delivered to cell culture media ⁴⁷, and when CAG repeats are inserted into phosphoribosyltransferage (HPRT), which is not associated with any of the ten CAG repeat diseases ¹⁷⁵. Collectively, these data point toward a common and central role of aggregation of expanded polyQ domains in the toxic gain-of-function of each disease.

Studying the lowest common denominator, the expanded polyQ sequence, is therefore an attractive research pursuit; the properties of expanded polyQ sequences may represent a universal mechanism of aggregation and toxicity. Therefore, mechanisms of simple polyQ

aggregation are immediately relevant and therapeutics targeting simple polyQ aggregation and toxicity could prove beneficial to the polyQ repeat diseases.

Additionally, the effect that protein context (N- and C-terminal flanking sequences, expression profiles, etc.) has on expanded polyQ domains can be best understood in the context of how “naked” polyQ peptides behave. For instance, in some polyQ diseases, N- and C-terminal flanking sequences alter the aggregation mechanism of the associated polyQ domain, such as the Josephin domain of Ataxin 3 ^155,156 and the htt^NT segment of huntingtin ¹⁷⁶, which undergoes a more complex intermediate oligomer formation step ^39,176 (Section 1.4.4).

Finally, the polyQ repeat-length dependence of disease and aggregation in vivo is mirrored by simple polyQ repeat-length aggregation kinetics in vitro ^77,107. Simple polyQ peptides recapitulate a portion of the aggregation pathway by forming similar final β-sheet rich amyloid-like structures by FTIR and EM ¹⁷⁶. The final aggregates have their own toxicity associated with them, and can bestow toxicity when delivered to cell media ⁴⁷, or when expressed in Drosophila ⁷⁸.

1.4.3.2 Simple PolyQ as a Model The fibril elongation reaction of simple polyQ peptides, which undergo a single nucleation step, can be mathematically modeled to initial aggregation kinetics by applying the Eaton-Ferrone model 22,107,177,178

, represented by the Equation 1, where Kn* is the equilibrium constant of nucleation, k²₊ is the aggregate elongation constant (11,400 liters/mol*s)

177, t is time in seconds, and C is the initial monomer concentration.

Equation 1: Eaton-Ferrone Kinetics Model of Nucleation

Information about the nucleation and elongation processes can be obtained by plotting the decay in monomer concentration for early reaction time points versus t² to yield a slope of (1/2)(Kn*)(k²+)(C^n*+2), which is a representation of the initial reaction rate. A plot of the log of this rate versus the log of the starting concentration, for a series of reactions at different starting concentrations, yields a slope equal to n* + 2. Using this model, nucleus size n*, nucleus equilibrium constant kn*, and the Gibbs free energy of nucleus formation ΔGn* can be calculated experimentally ^108,177. The Eaton-Ferrone model of aggregation provides descriptive features of aggregation phenomenon that follow a simple single-step nucleation. Not only are these parameters important for describing the behavior and mechanism of polyQ aggregation, but also can be used to diagnose how future aggregation inhibitors affect the aggregation mechanism.

In HD, the age of onset decreases as polyQ repeat length increases. In vitro, the kinetics of nucleation and aggregation are enhanced as polyQ repeat length increases. PolyQ peptides with as few as 8-10 glutamine repeats may be capable of aggregating into β-rich amyloid-like structures in vitro ^38,76,179, albeit extremely slowly. These short polyQ peptides likely aggregate too slowly to escape protein quality control mechanisms in vivo. PolyQ peptides with fewer than 25 glutamine repeats receive an enhancement to aggregation kinetics with each additional glutamine repeat; however, after 25 glutamine repeats, additional glutamines drastically enhance aggregation by altering the mechanism of aggregation. Kar et al. found that longer simple polyQ peptides (Q ≥ 25) nucleate aggregation efficiently with a monomeric nucleus, n* = 1 (a single

molecule is required for nucleation, likely through intramolecular rearrangement), while shorter polyQ peptides have an inefficient tetrameric nucleus of n* = 4 (four molecules required to initiate and template amyloid aggregation) ⁷⁷. The enhancement of nucleation efficiency results in a substantial boost in kinetics of nucleation between Q ≤ 23 and Q ≥ 25, which can be rationalized as the generic difference in kinetics between inter- versus intra-molecular reactions.

The initial nucleation event is extremely rare, even for expanded polyQ sequences, resulting in a lag-phase in the kinetics profile, followed by a rapid elongation phase. Even expanded K2Q47K2 has a tremendously small portion of monomers exploring the nucleus state at any given time in solution (K_n* = 2.6 * 10^-9). The rate limiting lag-phase can be bypassed if pre-formed polyQ aggregates are added to soluble polyQ peptides, resulting in seeded elongation.

Addition of pre-formed aggregate seeds eliminates the energy barrier in the formation of the initial nucleation event, thus removing the lag-phase and proceeding directly to the rapid elongation phase. This process is relevant to disease, as seeding elongation may occur in vivo as part of a prion-like transmission and propagation of mutant huntingtin ^180–182. Simple polyQ models may be useful to screen for or validate molecules that block seeding competency, that dampen nucleation formation, or limit polyQ elongation efficiency.