Mechanism of prion propagation: amyloid growth occurs by monomer addition

Abstract

Abundant nonfibrillar oligomeric intermediates are a common feature of amyloid formation, and these oligomers, rather than the final fibers, have been suggested to be the toxic species in some amyloid diseases. Whether such oligomers are critical intermediates for fiber assembly or form in an alternate, potentially separable pathway, however, remains unclear. Here we study the polymerization of the amyloidogenic yeast prion protein Sup35. Rapid polymerization occurs in the absence of observable intermediates, and both targeted kinetic and direct single-molecule fluorescence measurements indicate that fibers grow by monomer addition. A three-step model (nucleation, monomer addition, and fiber fragmentation) accurately accounts for the distinctive kinetic features of amyloid formation, including weak concentration dependence, acceleration by agitation, and sigmoidal shape of the polymerization time course. Thus, amyloid growth can occur by monomer addition in a reaction distinct from and competitive with formation of potentially toxic oligomeric intermediates.

Figure 1. NM Is Predominantly Monomeric

(A) Oligomeric state of NM prior to assembly was assessed by equilibrium AUC. Above, raw plot of absorbance versus radial position for 5.8 μM NM at equilibrium, with best-fit line for a single species. Below, residuals from the fit shown above. (B) Equilibrium AUC data from 1.5, 3.8, 5.2, 5.5, and 5.8 μM samples of NM were fit to a single-species model. Shown is fitted molecular mass versus concentration. For reference, dashed lines are shown at the calculated monomer mass (29.6 kDa) and dimer mass (59.2 kDa). (C) Distribution of absorbance versus sedimentation coefficient at the indicated concentrations was obtained from velocity sedimentation data using Sedfit (Schuck et al. 2002). Inset is a magnification of the main peaks. (D) A small amount of a larger complex could be resolved by velocity sedimentation. Distribution of absorbance versus sedimentation coefficient for each of 3 μM NM and 2.7 μM NM plus GroEL (with absorbance equivalent to that of 0.3 μM NM) is shown. Inset is a magnification for sedimentation coefficients between 15.6 S and 28.4 S.

Figure 2. Kinetics of NM Fiber Growth Support a Monomer Addition Model

(A) Initial rate of polymerization versus concentration of NM. Soluble NM at the indicated final concentrations was mixed with sonicated fibers (2.5 μM fibers at 5% of final volume) and polymerization was followed by a continuous thioflavin T assay. Rates shown were determined by the initial slopes of polymerization curves. (B) Initial rate of polymerization versus concentration of soluble NM in the presence of sonicated seed (1% of final volume) measured by a discrete thioflavin T binding assay. Error bars throughout represent the standard deviation of at least three measurements. (C) Polymerization of NM labeled with Alexa-647 at a C-terminal cysteine was monitored by the quenching of Alexa-647 fluorescence. The indicated concentrations of soluble NM were mixed with sonicated fibers (4% of final volume), and the initial rate of polymerization was measured. (D) Initial rate of polymerization versus concentration of seed. Soluble NM (2.5 μM final concentration) was mixed with the indicated quantity of sonicated seed. Highly fragmented fibers were used to maximize the absolute rate. Rates were measured as in (A). (E) Initial rate of polymerization versus concentration of seed as measured by discrete thioflavin T binding assay. Initial soluble NM concentration was 2.5 μM. (F) Initial rate of polymerization versus concentration of seed as measured by Alexa-647 fluorescence quenching. Initial soluble NM concentration was 200 nM.

Figure 3. Evidence that a Conformational Conversion following NM Monomer Binding to Fiber Ends Becomes Rate Limiting at High [NM]

(A) NM is predominantly monomeric at concentrations up to 20 μM in the presence of a moderate level of denaturant (100 mM urea and 100 mM GuHCl). Oligomeric state was assessed by velocity AUC as in Figure 1C for NM at 10, 15, and 20 μM. (B) Soluble NM was mixed with sonicated fibers in the presence of 100 mM urea and 100 mM GuHCl to minimize off-pathway aggregation at higher [NM], and the initial rate of polymerization was measured as in Figure 2A. Inset, rate of polymerization of 20 μM soluble NM versus seed percentage (v/v). The continued linear dependence of rate on fiber ends indicates that at high [NM], fiber ends remain limiting.

Figure 5. Effect of Fiber Fragmentation on Polymerization Kinetics

(A) De novo NM polymerization (1.0–12 μM) followed by a continuous thioflavin T binding assay. (B) Lag time (measured as time to 5% completion of polymerization) versus NM concentration for the polymerizations shown in (A). (C) Schematic of nucleated polymerization with fragmentation. At low concentrations fiber growth is limited by monomer binding, whereas at high concentrations conformational conversion after binding becomes limiting. (D) The de novo polymerization data shown in (A) were fit with a linearized model (Ferrone 1999) that includes nucleation, fiber growth by monomer addition, and fragmentation, assuming the indicated sizes for the smallest stable amyloid species (nucleus size). Plotted are the residuals (best-fit value minus observed data). Each block (labeled at the left by the nucleus size used) represents the residuals from simultaneously fitting all of the data. Each line within the block displays the individual residuals from a single concentration (1 μM [top] to 12 μM [bottom]). Residuals are color coded according to the color key at the right, with red and green indicating large errors and yellow indicating a good fit. Time varies from time zero (farthest left) to the time of 3% completion of polymerization. Below (NP) are residuals from fitting the same data using a simple nucleated polymerization model. Note the systematic deviations from the model for large nucleus sizes and for NP.

Figure 4. Agitation Causes Fiber Fragmentation

(A) Dependence of de novo polymerization reactions on degree of agitation. Polymerization in a microplate with shaking every minute (line only) was followed by thioflavin T fluorescence. Polymerization in a test tube disturbed only by pipetting to take samples for measurement (diamonds) was measured by Congo Red binding. Polymerization in a microplate with absolutely no agitation (multiple samples were started in parallel and no sample was measured more than once) (squares) was measured by Congo Red binding. (B) Effect of agitation on elongation rate. Identical reactions with 5 μM soluble NM and 2% (v/v) seeds were grown with (triangles) or without (circles) agitation. The seeds were made fresh and sheared by passing through a 25-gauge needle ten times. Polymerization was assayed by discrete thioflavin T measurements. Inset, identical reactions followed for 100 min. (C) Effect of agitation on fiber lengths. Long fibers were grown from sonicated seeds in the absence of agitation and then subjected to agitation (end-over-end rotation in a 2-ml tube). AFM images were taken after 0, 1, and 15 h of rotation. Each image is 5 μm by 5 μm. (D) Effect of agitation on seeding efficacy. Fibers from the samples imaged in (C) were used to seed polymerization of 10 μM soluble NM, and the initial rate of polymerization was measured as in Figure 2A. (E) De novo NM polymerization was measured by continuous thioflavin T fluorescence, and relative fiber number was computed as a function of time (see Materials and Methods). Inset, fiber number versus time was fit to an exponential curve for time = 0 to time = 100 min.

Figure 6. Fundamental Unit of Addition for NM Fiber Growth Determined by TIRF

(A) Schematic of experimental setup. Cy5-labeled fibers (red) were attached to the microscope slide via biotin-streptavidin linkage. Cy3-labeled NM (green) was added in solution. (B) TIRF image of fibers (red) grown by addition of 200 nM NM (67% labeled with Cy3 [green]) for approximately 1 h. Cy3 fluorescence at fiber ends in this image represents multiple addition events. (C) Schematic of expected results if oligomers (top, trimer model shown for example) or monomers (bottom) are added to fiber ends. Graphs at the right show simulated data of fraction of additions versus fluorescence intensity of addition. Note that for the monomer model, the simulated intensities are the same whether 9% or 67% of the soluble NM is labeled. (D) Observed data: fraction of observed events versus fluorescence intensity of events. Intensities of Cy3 spots appearing at fiber ends were measured. (E) Intensities of Cy3 spots appearing at fiber ends (circles) and intensities of Cy3 spots that bleached in a single step (squares).