Biomarker changes preceding symptom onset in genetic prion disease

Importance. Genetic prion disease is a universally fatal and rapidly progressive neurodegenerative disease for which genetically targeted therapies are currently under development. Preclinical proofs of concept indicate that treatment before symptoms will offer outsize benefit. Though early treatment paradigms will be informed by the longitudinal biomarker trajectory of mutation carriers, to date limited cases have been molecularly tracked from the presymptomatic phase through symptomatic onset. Objective. To longitudinally characterize disease-relevant cerebrospinal fluid (CSF) and plasma biomarkers in individuals at risk for genetic prion disease up to disease conversion, alongside non-converters and healthy controls. Design, setting, and participants. This single-center longitudinal cohort study has followed 41 PRNP mutation carriers and 21 controls for up to 6 years. Participants spanned a range of known pathogenic PRNP variants; all subjects were asymptomatic at first visit and returned roughly annually. Four at-risk individuals experienced prion disease onset during the study. Main outcomes and measures. RT-QuIC prion seeding activity, prion protein (PrP), neurofilament light chain (NfL) total tau (t-tau), and beta synuclein were measured in CSF. Glial fibrillary acidic protein (GFAP) and NfL were measured in plasma. Results. We observed RT-QuIC seeding activity in the CSF of three E200K carriers prior to symptom onset and death, while the CSF of one P102L carrier remained RT-QuIC negative through symptom conversion. The prodromal window of RT-QuIC positivity was one year long in an E200K individual homozygous (V/V) at PRNP codon 129 and was longer than two years in two codon 129 heterozygotes (M/V). Other neurodegenerative and neuroinflammatory markers gave less consistent signal prior to symptom onset, whether analyzed relative to age or individual baseline. CSF PrP was longitudinally stable (mean CV 10%) across all individuals over up to 6 years, including at RT-QuIC positive timepoints. Conclusion and relevance. In this study, we demonstrate that at least for the E200K mutation, CSF prion seeding activity may represent the earliest detectable prodromal sign, and that its prognostic value may be modified by codon 129 genotype. Neuronal damage and neuroinflammation markers show limited sensitivity in the prodromal phase. CSF PrP levels remain stable even in the presence of RT-QuIC seeding activity.


Supplementary Methods
Genotyping.Whole blood was frozen hemolyzed and genomic DNA was extracted.All samples were genotyped by two orthogonal methods.DNA was submitted for targeted capture using a custom set of probes (Twist Biosciences) directed against ~150 kb of genomic sequence in and surrounding PRNP, then enriched DNA was subjected to deep short-read sequencing (Illumina) at the Broad Institute's Genomics Platform.Data were aligned to the hg38 reference genome and processed using Dragen 3.7.8 to yield multi-sample VCF files.In parallel, DNA also underwent a previously described 24 protocol implemented by Genewiz, combining Sanger sequencing to detect SNPs and short indels with gel sizing to detect octapeptide repeat insertions (OPRI).Briefly, the primers utilized are: Int5: 5′-TgCATgTTTTCACgATAgTAACgg-3′, DG2: 5′-gCAgTCATTATggCgAACCTTggCTg-3′, and 3′Sal: 5′-gTACTgAggATCCTCCTCATCCCACTATCAggAAgA-3′.The prodcut of the DG2/3′Sal reaction is subjected to Sanger sequencing; the product of the DG2/Int5 reaction is run on a 2% agarose gel (the wild-type product is 464 bp).Genotypes obtained by the two different methods were in agreement for all samples.Determination of haplotypes was accomplished by molecular phasing of codon 129 to pathogenic variants by paired-end Illumina sequencing reads using a custom Python 3 script run on Terra (Terra.bio);source code is available in the study's online GitHub repository.Our study includes individuals who are at risk for inheriting a PRNP mutation but have not undergone predictive testing; genotypes were determined for research purposes only and were not disclosed to participants.

Sample processing.
Blood was collected in purple top K+ EDTA tubes, inverted gently, and centrifuged at 1,500 g for 10 minutes to retrieve plasma, aliquoted, and frozen at -80°C.20 mL of CSF was collected via gentle aspiration lumbar puncture using a 24G atraumatic Sprotte needle into 4x 5 mL syringes.Because PrP in CSF is highly sensitive to polypropylene adsorption, we followed the protocol described in Vallabh 2019 Figure S8, where 2 of the 4 collected 5 mL aliquots were ejected into tubes pre-loaded with the zwitterionic detergent CHAPS (3% wt/vol stock solution at 1% volume to yield a final 0.03% CHAPS concentration).All CSF were centrifuged at 2,000 g for 10 minutes to remove cells, and then aliquoted and frozen at -80°C.In instances where the LP yielded only a limited volume of CSF, CHAPS aliquots were prioritized.CHAPS aliquots were used for PrP and NfL quantification.Neat aliquots were used for RT-QuIC.For T-tau and betasynuclein, neat aliquots were used where available, while CHAPS aliquots were used when these were the only available samples; an assessment of the effect of 0.03% CHAPS on these assays is provided in Figures S3-S4.To minimize bias, technicians processing samples and performing biomarker assays were blinded to genotype.
RT-QuIC.Real-time quaking-induced conversion (RT-QuIC) was performed according to the protocol of Orru et al 2015, widely referred to as the IQ-CSF protocol 10 .The substrate was recombinant N-terminally truncated Syrian hamster PrP (SHaPrP90-230) expressed in E. coli and produced in-house according to a published protocol 25,26 and filtered by centrifugation at 3,214 g through a 100 kDa filter (PALL OD100C33).Final concentration in the reaction was 300 mM NaCl (Broad Institute SQM), 10 mM sodium phosphate pH 7.4 (Molecular Toxicology; Thermo C790B91), 1 mM EDTA (Broad Institute SQM), 10 µM thioflavin T (Sigma T3516-5G), 0.002% sodium dodecyl sulfate (SDS) (Invitrogen 15553-035), and 0.1 mg/mL recombinant PrP, all diluted into distilled water (InvitroGen UltraPure 10977-015).80 µL of a 1.25x concentrated master mix was loaded into each well of a 96-well plate (Nunc; Thermo 265301) and then 20 µL of CSF was added.Plates were sealed with adhesive film (VWR 37000-548).The assay was run at 55°C for 24 hours on a BMG FLUOStar OPTIMA platereader with alternating cycles of 1 minute rest and 1 minute 800 rpm shaking, with thioflavin T fluorescence measurements obtained via bottom read at 45-minute intervals with 450 nm excitation and 480 nm emission.Fluorescence kinetic curves were normalized so that 0% represents the baseline fluorescence value at first reading and 100% represents the instrument's maximum value of 65,000 fluorescence units.We committed to the pre-specified criteria of Orru et al 10 : a CSF sample was called positive if at least 50% of technical replicates (e.g.2/4) yielded at least 10% normalized signal within 24 hours.In practice, when screening undiluted CSF, all our positive samples were positive in 4/4 replicates while all negatives were positive in 0/4 replicates.For endpoint titration, 3-fold serial dilutions were run by adding 20, 6.7, 2.2, or 0.7 µL of CSF and then 0, 13.3, 17.8, or 19.3 µL of distilled water (InvitroGen UltraPure 10977-015).Titers were determined by Spearman-Karber analysis 27 ; the source code is available in this study's online GitHub repository.
PrP ELISA.PrP enzyme-linked immunosorbent assay (ELISA) was performed according to an in-house protocol previously published and described in detail 9 .The assay uses antibodies EP1802Y (Abcam ab52604) for capture and 8H4 (Abcam ab61409), biotinylated in-house, for detection.The standard curve is recombinant full-length mouse PrP (MoPrP23-231) produced in house, plated at concentrations from 0.05 ng/mL to 5 ng/mL.CSF was run at a dilution factor of 80, at which the lower limit of quantification (LLOQ) is 4 ng/mL.For longitudinal analysis (Figure 1C, Table S4), each individual was normalized to their own baseline.For comparison across mutations (Figure S2, Table S3), all individuals were normalized to the mean value in mutation-negative subjects, which was 70.6 ng/mL.GFAP.Plasma GFAP was quantified using Simoa (Quanterix) according to manufacturer instructions at a dilution factor of 4, yielding an LLOQ of 2.744 pg/mL.Samples were run in technical duplicate with a mean CV of 6.0%.
NfL. Plasma and CSF NfL were quantified using Ella by ProteinSimple (Bio-Techne) at a dilution factor of 2 yielding an LLOQ of 5.4 pg/mL.CSF aliquots containing CHAPS were used.For all Ella assays, samples were plated onto cartridges in singlicate; each sample is then run in technical triplicate with three glass nanoreactors (GNRs).
T-tau.CSF T-Tau was analyzed both by ELISA (Fujirebio) and by Ella (Bio-Techne).For Ella, samples were run at a dilution factor of 2 (except for N=6 samples run at a dilution factor of 3 due to limited volume), with an LLOQ of 1.68 pg/mL.For ELISA, samples were run at a dilution factor of 4, with an LLOQ of 39.5 pg/mL.Ella results are reported in Figure 1G, while a comparison of the two assays is given in Figure S3.
Beta-synuclein.CSF beta-synuclein was analyzed by Ella (Bio-Techne) at a dilution factor of 2 for CSF (LLOQ: 15.9 pg/mL) and either 4 or 8 for plasma depending on available sample volume (LLOQ: 31.8 pg/mL or 63.7 pg/mL respectively).As shown in Figure S4, all plasma samples from study participants were at LLOQ.

Figure S4. Quality control analyses on the Ella beta-synuclein assay. A) Parallelism (also called dilution linearity) tested on 6 CSF samples (3 suspected prion disease and 3 normal pressure hydrocephalus) and 2 plasma samples (2 suspected prion disease). Suspected prion disease patients are symptomatic individuals seen clinically at Massachusetts General Hospital outside of our study. All plasma samples from participants in our study were at the lower limit of quantification (LLQ). B) Mean coefficient of variation among 2 technical replicates per sample for the samples shown in (A). Note that this refers to plating the same sample twice, in separate wells, on the Ella cartridge; the measurement in each well is in turn the average of 3 replicate measurements. C) Comparison of 5 CSF samples from study participants
analyzed both with and without the addition of 0.03% CHAPS, a detergent shown to reduce loss of PrP to plastic.Mean 27% higher reading in CHAPS samples, P = 0.057, paired T-test.

Table S2. Means, standard deviations, and ranges of biomarker values from all visits by participants who did not develop active disease, by mutation status.
These summary statistics exclude all visits from the 4 participants who converted to active disease.In each cell, the top row shows mean±SD, while the bottom row shows range (min-max).CSF RT-QuIC positive shows the number of CSF samples that yielded an overall positive call.Each RT-QuIC reaction was run in quadruplicate; in this study, every positive sample was positive in all 4/4 replicates, while every negative sample was positive in 0/4 replicates.

Table S3. Mean CSF PrP concentration (ng/mL) by mutation.
These are the numeric values for the data shown in Figure S2.Results were grouped first by individual to determine mean CSF PrP concentration across longitudinal CSF samples, then grouped by mutation to determine mean and SD across individuals.N is the number of individuals in each group.

Other information Funding
Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based 5 Figure S1.Flow chart of participant recruitment.At launch in July 2017, the study was open to known mutation carriers, those at risk, and known controls.From November 2021 new enrollment restricted to only known carriers, but already-enrolled individuals were invited to continue to participate regardless.

Figure S2 .
Figure S2.CSF PrP concentration by PRNP mutation.Each point represents the mean of all available CSF samples for one study participant.Data are normalized to the mean of the mutation-negative controls ("none").P values are for differences from the control group in a linear model (lm in R, equivalent to Type I ANOVA).

Figure S3 .
Figure S3.Quality control analyses on the Ella T-tau assay.A) Comparison of CSF T-tau concentrations in pg/mL for N=151 CSF samples determined by Fujirebio ELISA (x axis) versus Ella (y axis).The red line shows the best fit linear regression which is ella = 16.3 pg/mL + 67% × elisa.The Pearson's correlation is r = 0.94, P = 5.3e-70.B) Comparison of mean CSF T-tau values per individual by Ella in study participants vs. 3 symptomatic patients with suspected prion disease.C) Mean testretest CV for longitudinal LPs from the same individual: Ella 9.5%, ELISA 12.7%.D) Paired analysis of N=5 CSF samples analyzed by T-tau Ella both with and without the addition of 0.03% CHAPS.Mean value with CHAPS is 3% higher, P = 0.55 by paired T test.

Table S4 . Long-term test-retest reliability of CSF PrP.
Summary of data from Figure1C.Test-retest mean CV summarized for all non-converting study participants with ≥3 years of longitudinal CSF data.