Inhibiting the nucleation of amyloid structure in a huntingtin fragment by targeting α-helix-rich oligomeric intermediates

Abstract

Although oligomeric intermediates are transiently formed in almost all known amyloid assembly reactions, their mechanistic roles are poorly understood. Recently, we demonstrated a critical role for the 17-amino-acid N-terminus (htt(NT) segment) of huntingtin (htt) in the oligomer-mediated amyloid assembly of htt N-terminal fragments. In this mechanism, the htt(NT) segment forms the α-helix-rich core of the oligomers, leaving much of the polyglutamine (polyQ) segment disordered and solvent-exposed. Nucleation of amyloid structure occurs within this local high concentration of disordered polyQ. Here we demonstrate the kinetic importance of htt(NT) self-assembly by describing inhibitory htt(NT)-containing peptides that appear to work by targeting nucleation within the oligomer fraction. These molecules inhibit amyloid nucleation by forming mixed oligomers with the htt(NT) domains of polyQ-containing htt N-terminal fragments. In one class of inhibitors, nucleation is passively suppressed due to the reduced local concentration of polyQ within the mixed oligomer. In the other class, nucleation is actively suppressed by a proline-rich polyQ segment covalently attached to htt(NT). Studies with D-amino acid and scrambled sequence versions of htt(NT) suggest that inhibition activity is strongly linked to the propensity of inhibitory peptides to make amphipathic α-helices. Htt(NT) derivatives with C-terminal cell-penetrating peptide segments also exhibit excellent inhibitory activity. The htt(NT)-based peptides described here, especially those with protease-resistant d-amino acids and/or with cell-penetrating sequences, may prove useful as lead therapeutics for inhibiting the nucleation of amyloid formation in Huntington's disease.

Figure 1

Kinetic effects of htt^NT on aggregation of polyQ-containing peptides by the sedimentation assay. (a) Aggregation of 23.3 μM htt^NTQ₃₇P₁₀K₂ ( ); 25.2 μM SWQ₃₇P₁₀K₂ ( ); 25.6 μM K₂Q₃₅P₁₀K₂ (▴); 7.5 μM K₂Q₄₁K₂ alone ( ); and 7.0 μM K₂Q₄₁K₂ plus 17.4 μM htt^NT ( ). (b) Aggregation of 18.8 μM of the peptide ESLKSF-Q35-PPPSKETAAAKFERQHMDS incubated alone (●) or with 29.6 μM htt^NT ( ).

Figure 2

Htt^NT inhibition of aggregation. (a, b) Incubation of 9 μM htt^NTQ₃₀P₆K₂ with different molar ratios of htt^NT to htt^NTQ₃₀P₆K₂: 0.05:1 (○); 0.16:1 (▴); 0.43:1 (▵); 0.89:1 (◆). (a) Soluble htt^NTQ₃₀P₆K₂ after centrifugation; control incubation of htt^NTQ₃₀P₆K₂ without inhibitor (●); (b) Soluble htt^NT inhibitor after centrifugation; control incubation of htt^NT alone (●); (c) Effect of higher ratios and multiple additions of htt^NT on htt^NTQ₃₀P₆K₂ aggregation: 7.0 μM htt^NTQ₃₀P₆K₂ alone (●); 5.33 μM htt^NTQ₃₀P₆K₂ with either 7.5 μM htt^NT (i.e., ~ 1.4 fold molar excess) (○) or 13.3 μM htt^NT (i.e., ~ 2.5 fold molar excess) (▵) added at t = 0; 5.33 μM htt^NTQ₃₀P₆K₂ plus 7.5 μM htt^NT reaction in which fresh, additional aliquots of htt^NT were added at 7 hrs (11.2 μM), 24 hrs (7.0 μM), 31 hrs (7.3 μM), and 68 hrs (6.4 μM) (■). Peptides were incubated at 37 °C in PBS, with aggregation monitored by the sedimentation assay.

Figure 3

The SEC profile of an incubated mixture of htt^NTQ₃₀P₆K₂ (F17W) and htt^NT was analyzed by LC-MS. The peak eluting at ~ 32 mins gave the HPLC profile shown in part B, and the mass spectrum at the intensity peak (inset) gave a mass of 6697 Da consistent with htt^NTQ₃₀P₆K₂ (F17W). The peak eluting at ~49 mins gave the HPLC profile shown in Part C, and the mass spectrum at the intensity peak (inset) gave a mass of 1972.8 Da, consistent with htt^NT. The absence of a SEC peak containing both peptides indicates that there is no significant low molecular weight complex formation under these conditions.

Figure 4

Electron micrographs of aggregates of htt N-terminal fragments. Aggregates from Htt^NTQ₃₀P₆K₂ alone incubated in PBS at 37 °C and sampled at 0 hrs (a), 15 min (b), 5.5 hrs (c), 24 hrs (d), 48 hrs (e), and 100 hrs (f). Htt^NTQ₃₀P₆ incubated with htt^NT in 1:1 ratio in PBS, 37 °C and sampled at 3 hrs (g, h), 5.5 hrs (i, j), 24 hrs (k, l), 48 hrs (m) and 72 hrs (n). htt^NTPGQ₉P^1,2,3 incubated alone for 120 hrs (o), htt^NTQ₃₀P₆K₂ and htt^NTPGQ₉P^1,2,3 co-incubated in 1:4 ratio in PBS, 37 °C for 120 hrs (p). Scale bar represents 50 nm.

Figure 5

Spectroscopic probes of aggregate structure. (a) Tryptophan fluorescence spectra of isolated and resuspended aggregates from the following reactions: monomeric htt^NTQ₃₀P₆K₂ (F17W), λ_em = 355 nm (black); aggregated htt^NTQ₃₀P₆K₂ (F17W) alone, λ_em = 347.5 nm, (green); htt^NT plus htt^NTQ₃₀P₆K₂ (F17W), λ_em = 349 nm, (magenta); htt^NTQ₃ (F17W) inhibitor plus htt^NTQ₃₀P₆K₂, λ_em = 337 nm, (blue); htt^NTPGQ₉P^1,2,3 (F17W) plus htt^NTQ₃₀P₆K₂, λ_em = 347.5 nm, (orange). (b) Time course of the Trp fluorescence peak shift in isolated aggregates when the htt^NTQ₃ (F17W) inhibitor is incubated with htt^NTQ₃₀P₆K₂ at a 1.1:1 ratio; (c) Second deriviative FTIR spectra of aggregates isolated at about 100 hrs from incubation reactions of htt^NT (black), htt^NTQ₃₀P₆K₂ (red), and a reaction with a 1:1 mixture of htt_NT and htt^NTQ₃₀P₆K₂ (blue); (d) primary FTIR spectra of aggregates generated upon incubation of either all-L htt^NT (blue), all-D htt^NT (red), and retro-inverso htt^NT (green).

Figure 6

Abilities of approximately equimolar htt^NT analogs to inhibit aggregation of htt N-terminal fragments in PBS at 37 °C, as determined by a sedimentation assay. (a) inhibition of 7.6 μM htt^NTQ₃₀P₁₀K₂ aggregation by all-L htt^NT (15 μM), all-D htt^NT (11.1 μM) and retroinverso-htt^NT (9.5 μM) (b, c) inhibition of 5–10 μM htt^NTQ₃₀P₆K₂ aggregation by 1:1 molar amounts of wild type and scrambled htt^NTQ sequences.

Figure 7

Scatterplot analyses of inhibitory activities of wild type htt^NT and scrambled peptides (Table 2). (a) aggregation inhibition vs. AGADIR α-helix propensity (linear fit; R² = 0.80); (b) aggregation inhibition vs. hydrophobic moment (μH) score (exponential fit; R² = 0.88); (c) aggregation inhibition vs. normalized μH values determined by multiplying the hydrophobic moment times the measured % α-helix as determined by CD spectroscopy (three parameter exponential fit; R² = 0.70) (see Table 2). In each plot the position of the htt^NT wild type sequence is in red.

Figure 8

Properties of htt^NTPGQ₉P^1,2,3. (a) inhibition htt^NTQ₃₀P₆K₂ aggregation by equimolar htt^NTPGQ₉P^1,2,3; (b) DLS of htt^NTPGQ₉P^1,2,3 incubated alone for three hrs; (c) SEC analysis of oligomer assembly and disassembly of htt^NTPGQ₉P^1,2,3. 300 μM peptide was incubated for 24 hrs, then separated by SEC in PBS running buffer (dashed line). The oligomer peak (~18 mins) was collected and incubated at 37°C, then analyzed at the times shown. Between 44 and 72 hrs a substantial portion of oligomer (~18 mins) dissociates to monomer (~42 mins).

Figure 9

Properties of cell penetrating peptide versions of inhibitors. (a) Inhibition of htt^NTQ₃₀P₆K₂ aggregation by htt^NTG₅K₈ and controls; (b) Inhibition of htt^NTQ₃₀P₁₀K₂ aggregation by htt^NT PGQ₉P^1,2,3 K₈ and controls; c) Cytosolic delivery of htt^NTG₄CK₈-Alexa633 into SHSY5Y cells (Methods). After 10 mins exposure to peptide plus 1-pyrenebutyric acid, cells were washed extensively with PBS and imaged.

Figure 10

α-Helix rich molecules in htt N-terminal fragment aggregation and its inhibition. (a) A model for nucleated growth in htt N-terminal fragment aggregation (modified from Jayaraman et al., Ms. submitted). When the htt N-terminal fragment monomer (a1: htt^NT, green; polyQ, orange; Pro-rich domain, black) assembles into the oligomeric intermediate (a2), the htt^NT segment takes on α-helix structure whose packing with other htt^NT segments into a helical bundle stabilizes the oligomer. As a consequence, the disordered polyQ chains are brought together closely in space within the oligomer, which lowers the barrier to formation of nuclei with amyloid structural elements (a3, polyQ extended chain, red). Monomers present at this stage facilitate elongation into growing amyloid fibrils (a4, a5, a6). In the final stages of amyloid growth, as the monomer pool is depleted, oligomers that have not undergone nucleation dissociate (a2 -> a1) to generate more monomers to support fibril elongation. (b) structural model for the inhibition complex (oligomer) between htt^NT (blue cylinders) and an htt N-terminal fragment; (c) structural model for the inhibition complex between an htt N-terminal fragment and a C-terminally extended htt^NT inhibitor, such as htt^NTPGQ₉P^1,2,3 or htt^NTPGQ₉P^1,2,3K₈ (blue cylinders = htt^NT helices; blue lines = PGQ₉ sequence; red circles = additional Pro residues (see Table 1)).