The Japanese mutant Aβ (ΔE22-Aβ(1-39)) forms fibrils instantaneously, with low-thioflavin T fluorescence: seeding of wild-type Aβ(1-40) into atypical fibrils by ΔE22-Aβ(1-39)

Abstract

The ΔE693 (Japanese) mutation of the β-amyloid precursor protein leads to production of ΔE22-Aβ peptides such as ΔE22-Aβ(1-39). Despite reports that these peptides do not form fibrils, here we show that, on the contrary, the peptide forms fibrils essentially instantaneously. The fibrils are typical amyloid fibrils in all respects except that they cause only low levels of thioflavin T (ThT) fluorescence, which, however, develops with no lag phase. The fibrils bind ThT, but with a lower affinity and a smaller number of binding sites than wild-type (WT) Aβ(1-40). Fluorescence depolarization confirms extremely rapid aggregation of ΔE22-Aβ(1-39). Size exclusion chromatography (SEC) indicates very low concentrations of soluble monomer and oligomer, but only in the presence of some organic solvent, e.g., 2% (v/v) DMSO. The critical concentration is approximately 1 order of magnitude lower for ΔE22-Aβ(1-39) than for WT Aβ(1-40). Several lines of evidence point to an altered structure for ΔE22-Aβ(1-39) compared to that of WT Aβ(1-40) fibrils. In addition to differences in ThT binding and fluorescence, PITHIRDS-CT solid-state nuclear magnetic resonance (NMR) measurements of ΔE22-Aβ(1-39) are not compatible with the parallel in-register β-sheet generally observed for WT Aβ(1-40) fibrils. X-ray fibril diffraction showed different D spacings: 4.7 and 10.4 Å for WT Aβ(1-40) and 4.7 and 9.6 Å for ΔE22-Aβ(1-39). Equimolar mixtures of ΔE22-Aβ(1-39) and WT Aβ(1-40) also produced fibrils extremely rapidly, and by the criteria of ThT fluorescence and electron microscopic appearance, they were the same as fibrils made from pure ΔE22-Aβ(1-39). X-ray diffraction of fibrils formed from 1:1 molar mixtures of ΔE22-Aβ(1-39) and WT Aβ(1-40) showed the same D spacings as fibrils of the pure mutant peptide, not the wild-type peptide. These findings are consistent with extremely rapid nucleation by ΔE22-Aβ(1-39), followed by fibril extension by WT Aβ(1-40), and "conversion" of the wild-type peptide to a structure similar to that of the mutant peptide, in a manner reminiscent of the prion conversion phenomenon.

Figure 1

Thioflavin T fluorescence time course. Peptides were dissolved in neat DMSO and then diluted into aqueous buffer (NaP) so that the final peptide concentration was 100 µM and the DMSO concentration was 2% (v/v). At various intervals, aliquots were removed from the slurry, and thioflavin T fluorescence was measured as described in Materials and Methods. For the mixtures, peptides were dissolved and mixed in neat DMSO so that the final total peptide concentration was 100 µM and the DMSO concentration was 2% (v/v): (circles) WT Aβ_1-40, (squares) ΔE22-Aβ_1-39, and (diamonds) 1:1 (molar) mixture of WT Aβ_1-40 and ΔE22-Aβ_1-39.

Figure 2

Electron micrographs of ΔE22-Aβ_1-39 fibrils. Peptide was dissolved in DMSO and then diluted into NaP so that the peptide concentration was 100 µM and the final concentration of DMSO was 2% (v/v). A precipitate appeared instantly. Panels A–C show peptide immediately after addition of NaP to the peptide in neat DMSO. Magnifications are 15000×, 98000×, and 149000×. The images in panels A–C were taken immediately after the peptide had been transferred from neat DMSO to buffer and were viewed within 5 min of making the sample. Panels D–F are images of the same slurry after incubation for 7 days at 37 °C. Panels G and H are electron micrographs of fibrils of a 1:1 (molar) mixture of WT Aβ_1-40 and ΔE22-Aβ_1-39. Peptides were dissolved in neat DMSO and then diluted with NaP so that the final concentration of each peptide was 50 µM (i.e., total peptide concentration of 100 µM) and the final DMSO concentration was 2%. A precipitate formed immediately. Magnifications of 15000× and 49000× for panels G and H, respectively. For all images, the CCD camera added a magnification of 1.4×. (I) Histograms for ΔE22-Aβ_1-39 and WT Aβ_1-40 fibril diameters, based on 200 and 400 diameter measurements, respectively. For both peptides, the histograms showed a single peak; means ± standard deviations were 11.0 ± 2.44 nm for WT Aβ_1-40 fibrils and 9.44 ± 2.56 nm for ΔE22-Aβ_1-39 fibrils.

Figure 3

(A) Binding of ThT to WT Aβ_1-40 (circles) and ΔE22-Aβ_1-39 (squares) fibrils. Various volumes of fibril slurries (nominal peptide concentration of 1 mM) were added to a solution of ThT, as described in Materials and Methods. After an incubation period, the mixture was centrifuged and the supernatant analyzed for ThT concentration by HPLC. (B) Congo Red absorbance in the presence of WT Aβ_1-40 (circles) and ΔE22-Aβ_1-39 (squares) fibrils; the figure also shows the spectrum of Congo Red alone (diamonds).

Figure 4

Critical concentrations of WT Aβ_1-40 (●) and ΔE22-Aβ_1-39 (■) estimated by seeding solutions of these peptides with fibrillar seeds, as described in Materials and Methods. At various times, aliquots were removed to determine the concentration of peptide in solution, using the fluorescamine assay described in Materials and Methods. The critical concentration was estimated from a fit of the data to the equation of a first-order approach to equilibrium.

Figure 5

Fluorescence depolarization for (circles) WT Aβ_1-40, (squares) ΔE22-Aβ_1-39, and (diamonds) a 1:1 (molar) mixture of WT Aβ_1-40 and ΔE22-Aβ_1-39. The line represents a nonlinear least-squares fit of the data to the first-order rate equation A = (A_∞–A0)[1-exp(−kt)] + A₀, where A, A₀, and A_∞ are the anisotropies at time t, time zero, and infinite time, respectively, and k is the rate constant.

Figure 6

Size exclusion chromatography of WT Aβ_1-40 and ΔE22-Aβ_1-39 performed using a Superdex Peptide 10/300 column. For panels A–D, the peptide was initially dissolved in neat DMSO, which was diluted with NaP so that the final peptide concentration was 100 (A and B) or 10 µM(C and D). The eluant was NaP, without DMSO. For panels E and F, the peptide concentrations were 100 and 200 µM, respectively, and the samples were prepared as described for panels A–D. The eluant, however, was NaP containing 2% (v/v) DMSO: (A) 100 µM WT Aβ_1-40, (B) 100 µM ΔE22-Aβ_1-39, (C) 10 µM WT Aβ_1-40, (D) 100 µM ΔE22-Aβ_1-39, (E) 100 µM WT Aβ_1-40 (black) or ΔE22-Aβ_1-39 (blue), and (F) 200 µM WT Aβ_1-40 (black) or ΔE22-Aβ_1-39 (blue). For panels A–D, the effluent was monitored at 220 nm; for panels E and F, the effluent was monitored at 274 nm rather than 220 nm, because of the presence of DMSO in the eluant.

Figure 7

CD spectra of WT Aβ_1-40 and ΔE22-Aβ_1-39. (A)WT Aβ_1-40 was dissolved in HFIP and then diluted to a final peptide concentration of 100 µM with NaP. Spectra were recorded immediately after dilution of the peptide into NaP (t = 0 days, red) and after incubation at 37 °C for 7 days (t = 7 days, blue). (B) ΔE22-Aβ_1-39, treated the same as WT Aβ_1-40 in panel A. (C) WT Aβ_1-40 (red) and ΔE22-Aβ_1-39 (blue) were dissolved in HFIP and then diluted with NaP as described for panels A and B. Immediately after the sample had been diluted into NaP, an initial CD spectrum was measured and the mixture was then centrifuged at 14300g, and the CD spectra of the supernatants (top 1 mL out of 1.5 mL in the centrifuge tube) were measured: (filled circles) WT Aβ_1-40 before centrifugation, (empty circles) WT Aβ_1-40 supernatant, (filled squares) ΔE22-Aβ_1-39 before centrifugation, and (empty squares) ΔE22-Aβ_1-39 supernatant.

Figure 8

Raw PITHIRDS-CT data for rehydrated ΔE22-Aβ40 fibrils with 1-¹³C labels at Val18 (A) and Val36 (B). For comparison, comparable data are provided for WT Aβ_1–40 with a 1-¹³C label at Val12 (C).

Figure 9

X-ray fiber diffraction pattern for partially aligned fibrils of ΔE22-Aβ_1–39 and for fibrils formed from a 1:1 (molar) mixture of ΔE22-Aβ_1–39 and WT Aβ_1–40. (A) One-dimensional azimuthal plots, showing intensity as a function of D spacing for WT Aβ_1–40 (black), ΔE22-Aβ_1–39 (red), and the 1:1 (molar) mixture of ΔE22-Aβ_1–39 and WT Aβ_1–40 (blue). The intensities have been normalized to the maximal intensity value of the peak of ~10 Å spacing. The one-dimensional azimuthal plot for WT Aβ_1–40 is from ref . (B) Pseudocolored plot of the X-ray diffraction pattern for ΔE22-Aβ_1–39. (C) Pseudocolored plot of the X-ray diffraction pattern for the 1:1 (molar) mixture of ΔE22-Aβ_1–39 and WT Aβ_1–40.