Fig 1.
Nomination of colon as a tissue for peripheral PrP quantification.
A) PRNP RNA expression in transcripts per million (TPM) in human tissues according to GTEx v8 public data. Each sub-tissue (e.g., brain – cerebellum) is represented by one point as the median TPM across all samples for that tissue, and each tissue (e.g., brain) is represented by one bar as the median of those medians. B) Prnp RNA expression in TPM in mouse tissues according to Söllner et al [22]. Each animal is represented by one point for individually measured TPM, and each tissue is represented by one bar as the median of the animals. C) Western blot (top) and Coomassie (bottom) of organs all from the same 1 WT and 1 KO animal, 6D11 anti-PrP antibody, see Methods for details. D) Organs tested by PrP ELISA at a 1:100 final dilution (10% homogenates at a further 1:10). PrP ELISA as reported except using double the detection mAb concentration (0.5 µg/mL instead of 0.25 µg/mL). E) Organs tested by PrP ELISA at a 1:25 final dilution (10% homogenates at a further 1:2.5). PrP ELISA as reported except using double the detection mAb concentration (0.5 µg/mL instead of 0.25 µg/mL). See S1 Fig for further assay development.
Fig 2.
Determination of PrP half-life by targeted mass spectrometry.
A) Diagram of mature, post-translationally modified mouse PrP. Adapted from a CC BY licensed diagram we have previously released (https://github.com/ericminikel/prp_mrm/), see Minikel & Kuhn et al for details [23]. Indicated are the positions of peptides measured here in mass spectrometry, the epitopes of antibodies used in our in-house ELISA as well as D13, the antibody used for Western blot PrP quantification and conformation-dependent immunoassay by Safar et al [14]. B) Accumulation of 13C6 label from chow in the VVEQ peptide in mouse brain and colon. C) The best fit to the proportion of plasma free lysine empirically found to be 13C6 labeled, as reported by Fornasiero (maroon), and the proportion of peptide expected to be labeled over time as a function of half-life (shown in days on the right side). See Methods > Labeled peptide accumulation models for details. D) The ratio of proportion labeled (C) for a peptide of each half-life compared to a peptide of 5-day half-life. Ratios of 0.5 and 1.5 are arbitrary landmarks highlighted to orient the eyes to a straight horizontal line. E) The proportion of PrP peptides VVEQ and GENF that are 13C6 labeled after 8 days as a function of mouse genotype. All differences are non-significant at Bonferroni-corrected P > 0.05, 2-sided T-test. F) Relative isotope abundance (RIA), in other words, the mean ratio of heavy to light lysine inferred to have been available over the 8-day labeling period, by tissue. Each point is one animal (N = 3 wild-type C57BL/6N per tissue). Segments represent means and error bars 95% confidence intervals per tissue. G) Proportion labeled in brain (y axis) versus brain half-life previously reported by Fornasiero for all measured non-PrP peptides (circles); black = measured peptide signal above LLQ, gray = below LLQ. The maroon line represents the expected proportion labeled after 8 days as a function of half-life, based on the plasma free lysine model from (C). The horizontal blue lines represent the proportion labeled observed for the two PrP peptides, and their vertical projection from the maroon curve down to the x axis represents the estimation of half-life from those proportion labeled measurements. G) As in (F), but for colon. The maroon line uses the plasma free lysine model from (C) adjusted based on the ratio of empirical RIA in colon compared to brain from (F).
Table 1.
Mouse lines used in this study. All PrP expression levels are measured in whole hemisphere, either by our in-house ELISA [17] or by the IQ Proteomics MRM assay described herein, and are shown normalized to wild-type animals (1.0x = wild-type epxression). *Maintained on a background of homozygous ZH3/ZH3 PrP knockout. **MoPrP-A refers to the mouse reference genome PrP sequence as found in C57BL/6N and most other commonly used mouse strains (as opposed to the MoPrP-B allele, containing the two substitutions L108F and V189T, found in certain strains [24]). ***The 3F4 epitope is two amino acid changes from the MoPrP-A sequence: L108M + V111M. ****Mean value obtained for 2 peptides in young and aged ki-3F4-FFI mice in Fig 4 of this study.
Fig 3.
Determination of PrP half-life by ASO administration and timed sacrifice.
A) Residual Prnp RNA and PrP protein (y axis), normalized to the mean of saline controls, at various timepoints (x axis) for wild-type mice after a single ICV dose of 500 µg active ASO 6. Each point represents a whole brain hemisphere from one animal. All measurements in saline controls (both RNA and protein) are shown in gray. For each timepoint, line segments represent means and error bars represent 95% confidence intervals. The green curve represents linearly interpolated residual RNA concentration. The dashed blue curve represents a single-rate exponential decay model fit to the data. The solid blue curve represents a mixture-of-rates fit to the data. B) As in (A) but for Ki817 human PrP 129V knock-in mice after a single dose 118 µg of ASO N. C) As in (A) but for wild-type mice infected with RML prions and treated with 300 µg ASO 6 at 105 dpi. D) Residual PrP in Sprague-Dawley rats treated with 1 mg of ASO 6 at day 0. Each point represents one animal, and for each timepoint, line segments represent means and error bars represent 95% confidence intervals. Long lines connect means of different timepoints.
Fig 4.
PrP turnover in mice with the equivalent of the D178N pathogenic variant by targeted mass spectrometry.
A) Peptide abundance across all young animals. Each point is one peptide included in this targeted MS assay. The x axis position is the log2 fold change in total intensity (light + heavy peptide) in ki-3F4-FFI animals (N = 16) versus combined ki-3F4-WT and C57BL/6N animals (N = 32 total). The y axis is -log10(P), with the P value from a 2-sided T-test. B) Half-life across all young animals. Each point is one peptide included in this targeted MS assay. A half-life is fit to the labeled chow days and proportion heavy peptide across all ki-3F4-FFI animals (N = 4 per timepoint times 4 timepoints) and compared to that fit to the combined control groups (N = 8 per timepoint times 4 timepoints). The x axis position is the log2 fold change in this estimated half-life. The y axis is -log10(P), with the P value of the main effect of genotype (ki-3F4-FFI vs. combined controls) in a beta regression model for proportion labeled peptide as a function of chow days. C) Label accumulation in the GENF peptide by genotype. Each point is one animal. The x axis position is the number of days of labeled chow, the y axis position is the proportion of the GENF peptide that is heavy lysine labeled. Solid lines represent mixture-of-rates decay model fit to the data. D) As in C but for the VVEQ peptide. E-H) As in A-D but for aged, 63-week old mice.
Table 2.
Antisense oligonucleotides used in this study. Black: unmodified DNA (2′H). Orange: 2′ methoxyethyl (2’MOE). Blue: 2′-4′ constrained ethyl (cEt). Unmarked backbone linkages: phosphorothioate (PS). Linkages marked with o = phosphodiester (PO). mC: 5-methylcytosine.