Fig 1.
Schematic diagram of end-point dilution (ED) RT-QuIC assay for quantifying pathological αSynD seeds.
(A) Various type of PD/DLB patient samples for quantitating αSynD seeds. (B) 10-fold serial dilutions of tissue/biospecimens were prepared and seeded (4-replicates/dilution) in a 96-well plate containing reaction mix comprising buffer components, recombinant αSyn substrate and detection dye Thioflavin-T (ThT). (C) The plates were incubated under standard αSyn RT-QuIC conditions and resulting outcomes plotted as ThT fluorescence versus time. (D) Positive replicate wells per dilution were enumerated based on whether the ThT fluorescence threshold was reached by a selected cut-off time. The concentration of pathological αSynD seeds from patient biospecimen is expressed as SD50 (50% seeding dose) per volume or mass of sample, based on the dilution at which 50% of wells were ThT-positive, as estimated using the Spearman-Karber or other analytical methods. (E) Individual biospecimens tested in independent ED assays have often displayed variability in ThT-positivity, resulting in deviations of the SD50 estimates from the true value (red plus).
Fig 2.
Effect of dilution factor and replicate number on the reproducibility of αSynD seed concentration estimates in PD (A-D) and DLB (E-H) brain homogenates by ED RT-QuIC assays.
(A) & (E) Percent positive wells for triplicate assays versus sample dilution with 10-, 5-, and 2-fold dilution factors (as indicated in color legend). (B) & (F) Effect of decreasing dilution factor on consistency of log10 SD50/mg estimates from 3 independent ED assays. (C) & (G) Percent positive wells from triplicate ED assays versus sample dilution with 4-, 8-, and 12-replicates per dilution (as indicated in color legend). (D) & (H) Effect of increasing replicate numbers on consistency of log10SD50/mg estimates from triplicate assays. Data points in A, E correspond to arithmetic averages while C, G represent individual replicates of the total % positive wells from the 3 separate assays, each with quadruplicate wells at each dilution. Horizontal dotted lines in A, C, E, G denote 50% positive wells. Data points in B, F, D, H indicate log10 SD50/mg brain tissue, with horizontal and vertical bars indicating the mean and 95% CI respectively.
Fig 3.
Theoretical comparison of various SD50 estimation methods in 10F4R and 2F8R assays.
For a hypothetical sample with a seed concentration of 105 SD50/mg, one million random plates were generated, with each plate having a certain number of positive wells (ThT signals above threshold) in each dilution. For each outcome, the SK, RM, Poison or midSIN, methods were used to estimate the log10 SD50/mg αSynD seed concentration for either (A) the 10F4R assay with an initial dilution of 10−4, 5 dilutions separated by 10-fold, and 4 replicates/dilution, resulting in 121 unique readout outcomes; or (B) the 2F8R assay with an initial dilution of 10−4, 12 dilutions separated by 2-fold, and 8 replicates/dilution, resulting in 119,637 unique readout outcomes. Depending on the estimation method, different readout outcomes can result in the same estimated log10 SD50/mg. The number of unique concentration estimates obtained by each method is indicated above each graph. The true sample concentration (105 SD50/mg) (vertical solid black line), along with the mean (dash-dot), 68% CI (thick dotted), and 95% CI (thin dotted) of the estimated concentrations are indicated.
Fig 4.
Reproducibility assessment of the 10F4R and 2F8R assay in estimating αSynD seed concentrations.
(A) & (B) Replicate heterogeneity in log10 SD50/mg values obtained from ED assays performed in triplicate on PD and DLB BHs, respectively. The 2F8R assay (orange) reduces variability in SD50 estimates compared to the 10F4R assay (blue). Comparison of 4 quantification algorithms displays enhanced quantification accuracy of Poisson and midSIN (low CIs) in the 2F8R assay. Each colored circle represents one of the 3 or 4 replicate measurements, the numbers correspond to the difference between the minimum and maximum log10 SD50/mg values obtained, and the shading indicates this [min., max.] range.
Fig 5.
Stratification of samples differing in αSynD seed concentrations.
The schematic on the left shows experimental design for testing 2-fold and 4-fold differences in αSynD seed concentration in PD brain tissue. Scatter plots showing quantification efficacy of all four algorithms (SK, RM, Poisson and midSIN) in the (A)-(D) 10F4R (top panel) and (E)-(H) 2F8R (bottom panel) ED assay formats. In each plot, log10 SD50 values from 3 independent ED assays are displayed (black circles). Error bars represent arithmetic mean with 95% CI. The trend of SD50 estimates in tested sample dilutions are shown as dotted black lines in each plot. Statistical significance is denoted in blue as *p<0.05, **p<0.01 and ***p<0.005.
Fig 6.
Quantification of time-dependent αSynD seed reduction by HOCl treatment.
The schematic on the left shows experimental design for HOCl treatment of PD BH for different treatment times followed by estimation of differences in αSynD seed concentration by ED assays. (A)-(H) Scatter plots showing estimate using the SK, RM, Poisson and midSIN algorithms in the 10F4R (A-D) and 2F8R (E-H) ED assay formats. In each plot, log10 SD50/mg tissue values from 3 independent ED assays are displayed. In each case, the triplicate ED assay performed on untreated PD BH is shown as control (Ctrl). Error bars represent arithmetic mean with 95% CI. The trend of SD50 estimates with increasing HOCl treatments is shown as dotted black lines in each plot. Statistical significance is denoted in blue as *p<0.05, **p<0.01 and ***p<0.005.
Fig 7.
Quantification of αSynD seeding in patient CSF samples.
2F4R RT-QuIC quantification of αSynD seeding in patient CSFs. A-F Replicate heterogeneity (grey bars) in log10 SD50 estimates for 6 antemortem PD patient CSFs (PD1-PD6) obtained from 3 independent 2F4R ED assays. The log10 SD50 values obtained from SK (green), RM (orange), Poisson (violet) and midSIN (blue) 4 methods are shown. The error bars in midSIN estimates represent 95% CI for each replicate based on the amount of noise/confidence of individual SD50 estimates. Each colored circle represents one of the 3 replicates, the numbers correspond to the difference between the minimum and maximum log10 SD50/15 μL values obtained, and the shading indicates this [min., max.] range.
Fig 8.
Quantification of αSynD seeding activity in patient skin and olfactory mucosa (OM) samples.
Schematic showing 2F8R (A) and D2F8R (B) assay formats. D2F8R format reduces pipetting steps by directly preparing an intermediate sample dilution to minimize seed loss (see methods). Replicate heterogeneity (colored circles and bars) in log10 SD50 estimates for patient skin samples viz. (C) PD1, (D) PD2 and (E) DLB 1, and (F) OM sample from a PD patient, obtained from 3 independent 10F4R (blue) and 2F8R (orange) assay formats. Enhanced accuracy of all 4 quantification algorithms (low CIs) was obtained with D2F8R assays (red) as shown with (E) DLB1 skin and (F) PDOM samples. Numerical values above or below each bar represent log10 difference between the maximum and minimum SD50 values estimated with a given method between the 3 replicate ED assays. In (C)-(E), the mg denominators in the y-axis labels refer to weight of solid tissue and in (F), the μL denominator refers to volume of packed OM swabbing-derived pellet.