Natural and synthetic prion structure from X-ray fiber diffraction

Abstract

A conformational isoform of the mammalian prion protein (PrP(Sc)) is the sole component of the infectious pathogen that causes the prion diseases. We have obtained X-ray fiber diffraction patterns from infectious prions that show cross-beta diffraction: meridional intensity at 4.8 A resolution, indicating the presence of beta strands running approximately at right angles to the filament axis and characteristic of amyloid structure. Some of the patterns also indicated the presence of a repeating unit along the fiber axis, corresponding to four beta-strands. We found that recombinant (rec) PrP amyloid differs substantially from highly infectious brain-derived prions, both in structure as demonstrated by the diffraction data, and in heterogeneity as shown by electron microscopy. In addition to the strong 4.8 A meridional reflection, the recPrP amyloid diffraction is characterized by strong equatorial intensity at approximately 10.5 A, absent from brain-derived prions, and indicating the presence of stacked beta-sheets. Synthetic prions recovered from transgenic mice inoculated with recPrP amyloid displayed structural characteristics and homogeneity similar to those of naturally occurring prions. The relationship between the structural differences and prion infectivity is uncertain, but might be explained by any of several hypotheses: only a minority of recPrP amyloid possesses a replication-competent conformation, the majority of recPrP amyloid has to undergo a conformational maturation to acquire replication competency, or inhibitory forms of recPrP amyloid interfere with replication during the initial transmission.

Fig. 1.

Fiber diffraction patterns from SHaPrP amyloids. Black arrows indicate cross-β meridional diffraction at close to 4.8 Å resolution. Color tables (Fig. S5) vary, to show clearly the positions of diffraction maxima of very different intensities. (A) SHaPrP 27–30(Sc237). The small inset in the center of the pattern uses a different color table to show the intense 63 Å reflection. (B) RecSHaPrP (90–231) amyloid. White arrow: broad equatorial diffraction at about 10.5 Å resolution, not seen in A.

Fig. 2.

Fiber diffraction patterns from PTA-precipitated MoPrP amyloids. Black arrows indicate cross-β meridional diffraction at close to 4.8 Å resolution. White arrows indicate second and third orders of meridional 19.2 Å diffraction. Color tables are given in Fig. S5. (A) MoPrP 27–30(RML). Inset uses a different color table to show the weak, broad 4.8 Å diffraction. (B) The synthetic prion strain, MoSP1, derived from recPrP (89–230) amyloid, passaged twice through Tg9949 mice and purified as in A.

Fig. 3.

Comparison of observed and calculated diffraction patterns for β-helical and stacked-sheet amyloid models. (A) Experimental diffraction pattern from SHaPrP 27–30 (Fig. 1A). (B) Calculated diffraction from a disordered non-crystalline trimeric β-helical model. (C) Model used to calculate data in B. (D) Experimental diffraction pattern from recSHaPrP amyloid (Fig. 1B). (E) Calculated diffraction from a stacked-sheet model. (F) Model used to calculate data in E. In C and F, the filament axis is perpendicular to the figure plane.

Fig. 4.

Negative stain electron microscopy and filament diameters of PrP 27–30 prion rods and recPrP amyloid filaments. (A) SHaPrP 27–30(Sc237) prion rods had an average diameter of 57 ± 11 Å. (B) RecSHaPrP (90–231) amyloid filaments were very heterogeneous in structure with an average diameter of 79 ± 38 Å. (C) MoPrP 27–30(RML) prion rods were slightly smaller than in A, with an average diameter of 48 ± 8 Å. (D) RecMoPrP (89–230) amyloid filaments were heterogeneous in structure with an average diameter of 76 ± 21 Å. (E) The diameter of brain-passaged synthetic prions (PrP 27–30, MoSP1 strain) was very similar to that of natural prion isolates with an average value of 53 ± 13 Å. (F) MoPrP 27–30(RML) prion rods prepared by the PTA precipitation method. Several of the prion rods (black arrows) were decorated by apparent lipid-detergent micelles (white arrow heads). Samples A, B, and D were analyzed in the absence of Sarkosyl, while samples C, E, and F contained 0.2% Sarkosyl. (Scale bar, 100 nm in all panels.)