Research ethical board (REB) approval was obtained from local ethics boards at all collection sites (CHU de Québec, #A13-2-1096; CHU de Montréal, #2015-5705; Cambridge Ethics Committee, REC #03/303 and #08/H0306/26; Cambridge University Hospitals Foundation Trust Research and Development Department, R&D #A085170 and #A091246) and all participants provided informed written consent. All subjects also completed a questionnaire about health issues and medication on the day blood was collected. Both HD patients and CTRL were recruited from HD clinics in Quebec City, Montreal and Cambridge. In total, plasma samples were collected from 64 gene carriers and 61 age-matched CTRL (Table 1); however, only a subset of patient samples had sufficiently concentrated PBMC (n = 52 HD gene carriers, n = 45 CTRL) or platelets (n = 33 HD gene carriers, n = 29 CTRL) available (Table 1). The HD group included premanifest gene carriers (plasma n = 16, PBMC n = 9, platelets n = 10) and manifest patients (plasma n = 48, PBMC n = 43, platelets n = 23) according to clinical evaluations [5, 40]. Of the manifest patients, a subset was at a more advanced stage of disease (plasma n = 17). This population was paired with healthy control samples of similar age and gender to assess plasma tau levels in more advanced disease stages (HD n = 11 females, n = 6 males, average age = 59.7 ± 11.4; CTRL n = 10 females, n = 6 males, n = 1 unknown, average age = 57.6 years ± 14.7). Comorbidities were identified based on answers provided on the questionnaire, while blood cell counts were obtained from a routine blood work which was concurrently performed for each individual (Table 1) from blood retrieved in dipotassium ethylenediaminetetraacetic acid-coated tubes (BD Vacutainer, Cat#367,861). Any subjects with abnormal blood counts (i.e., blood cell count, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, and mean platelet volume), as identified by values outside of the reference range [16], were excluded from the study. The mean pathological CAG repeat length within HD patients was 42 ± 2–23. Clinical evaluations, including the total score on the Unified Huntington’s Disease Rating Scale (cUHDRS) [26] and total functional capacity (TFC) [26], were conducted within 6 months of blood collection, with the majority of subjects having this done on the same day as blood sampling (Table 1).

All plasma was processed within 2 hours (h) of sampling. For each patient, two citrate blood collection tubes (BD Vacutainer, no. 369714) were centrifuged at 2500 × g at room temperature (RT) for 2 × 15 min (min) prior to harvesting the supernatant, aliquoting, and storing at − 80 °C.

Two citrate-coated tubes (BD Vacutainer, no. 369714) were centrifuged for 10 min at 282 × g. The supernatant was collected and 1/5 of the volume of acid citrate dextrose [ACD, 0.48% (w/v) C6H8O7, 1.32% (w/v) Na3C6H5O7, and 1.47% (w/v) C6H12O6)] and 1/50 of the volume of ethylenediamine tetraacetic acid (EDTA, 0.5 M) was added before the complete suspension was centrifuged twice for 2 min at 400 x g and 5 min at 1300 × g. The subsequent platelet pellet was dissolved in 100 µl of Tyrode (7 mM NaHCO3, 135 mM NaCl, 3 mM KCl, 0.4 mM NaH2PO4, 1 mM MgCl2, 4.5 mM glucose, and 18.7 mM HEPES buffer) pH 6.5, 900 µl of Tyrode pH 7.4, 200 µl of ACD, and 20 µl of 0.5 M EDTA. The final platelet pellet was centrifuged for 5 min at 1300 × g and dissolved in 100 µl of lysis buffer (Thermo Scientific™, Pierce™ IP Lysis Buffer, no. 87788) with protease and phosphatase inhibitors (Thermo Scientific™, Halt™ Protease and Phosphatase Inhibitor Cocktail 100X, no. 78440). All samples were stored at -80 °C.

Heparin-coated tubes (BD Vacutainer, no. 367880) were centrifuged at 282 × g for 10 min at RT. Subsequently, the cell pellet was washed in phosphate-buffered saline (PBS) containing 2% fetal bovine serum and isolated using SepMate™ according to the manufacturer’s instructions (StemCell™, no. 15460). PBMC were then homogenized in 200 µL of the same lysis buffer as described for platelets with protease and phosphatase inhibitors and stored at − 80 °C.

For all immunoblots, positive (prefrontal cortex lysate from mice expressing human tau) and negative controls (prefrontal cortex lysate from tau knock-out mice) were included. For all samples, 40 µg of protein was combined with 1 × 1% (v/v) Laemmli buffer (312.5 mM Tris–HCl, 30% glycerol [Sigma, no. G5516-4L], 12.5% β-mercaptoethanol [Sigma, no. M3148-100ML], 10% SDS, 0.025 M EDTA, and 0.01% Bromophenol blue [Sigma, no. B0126-25G]), and water to a final volume of 40 µL. Prior to loading, all PBMC samples were sonicated in a water bath sonicator for 8 × 5 s and heated at 95 °C for 5 min, while platelet samples were heated at 95 °C for 5 min. Samples were then loaded into 10% SDS polyacrylamide gels and migrated for 90 min at 100 V in running buffer (25 mM Tris HCl, 190 mM glycine [Sigma, no. G7126-5 KG], and 0.1% SDS) and then transferred onto a 0.45 µm polyvinylidene difluoride membrane (GE Healthcare Life Science: 10,600,023) for 1 h at 100 V in transfer buffer (25 mM Tris HCl, 190 mM glycine, and 20% methanol [Fisher Chemical, no. A452-4]). After completion of the transfer, total protein was detected by incubating in ponceau red for 1 min. After the removal of ponceau red by sequential PBS washes, non-specific binding was eliminated by blocking with 3% gelatin extracted from cold water fish skin (Sigma, no. G7041-500G) in PBS for 1 h at RT, followed by overnight incubation at 4 °C with the following primary antibodies: t-tau (1:10,000, Dako, no. A0024) or p-tau 1:5,000 pS199, (Invitrogen, no. 44734G) diluted in 3% fish gelatin in PBS supplemented with 0.1% Tween 20 (PBST) (Fisher Bioreagent, no. BP337-500). The following day, membranes were washed three times for 10 min, incubated for 45 min at RT with IRDye 800CW (LI-COR Biotechnology, no. 926–32,212) and IRDye 680RD (LI-COR Biosciences, no. 926–68,073) antibodies and quantified using Odyssey CLx imaging system (LI-COR Biosciences).

All tau measurements in plasma samples were performed at the Neurochemistry laboratory of the University of Gothenburg (Mölndal, Sweden) using Simoa HD-X instruments (Quanterix). Plasma tau measurements included p-tau 181, p-tau 231, and N-terminal containing tau fragments (NTA-tau). Development and validation of these assays have been described elsewhere [18, 24, 26]. Prior to the assessment of samples, assay beads were suspended in bead diluent, biotinylated detector antibodies in tau 2.0 assay diluent (#101,556, Quanterix) and the enzyme streptavidin-conjugated β-galactosidase (SBG) concentrate (#103,397, Quanterix) in SBG diluent (#100,376, Quanterix). For plasma, randomized samples were thawed, vortexed, centrifuged (4000 g, 10 min), and diluted 1:2 with tau 2.0 assay diluent. An eight-point calibrator curve using recombinant GSK-3β phosphorylated full-length tau-441 (for p-tau 181 and p-tau 231) and non-phosphorylated full-length tau-441 (for NTA-tau) was included in all plates. Two internal quality control (iQC) samples were also present on each plate, before and after the analyzed samples, to control for inter- and intra-assay variability. Repeatability and intermediate precision values in the cohort were < 15%. Calibrators and internal quality control samples were run as duplicates.

Quantification of immunoblot band intensity was performed using the Image Lab 6.1 Software (Bio-Rad Laboratories, Inc.) for ponceau signal and the Odyssey Imaging System (Odyssey; Li-Cor, Lincoln, NE) for total tau (t-tau) and p-tau. Protein signal was corrected to ponceau when p-tau and t-tau are shown individually to control for loading discrepancies. For the analysis of p-tau levels, signals were corrected to t-tau. To normalize differences between gels, each independent gel included all experimental groups, and results were calculated as a percentage of CTRL on that gel prior to pooling the results.

Statistical analyses were performed with GraphPad Prism v. 9.0.1 (GraphPad, San Diego, California, USA) or R Studio®1.3.1093. The normality of the data were inspected with the Brown–Forsythe and Bartlett’s test for one-way ANOVA’s as well as the D’Agostino and Pearson normality tests prior to completion of linear regressions. When data were non-normally distributed, non-parametric tests were used where possible, and data were transformed where not possible. Differences between CTRL and HD gene carriers were evaluated using either the Mann–Whitney U test (severe disease comparisons for plasma, and all comparisons for PBMC and platelets) or a one-way ANOVA test with Dunnett’s multiple comparison test (plasma). For all one-way ANOVA and t tests, values that were more than 2 standard deviations away from the mean were identified as outliers and removed from these analyses. In all graphs, bars represent the mean with individual data points indicating biological replicates.

Effects of demographic variables and disease metrics on tau levels were evaluated using simple linear regressions. As tau levels are known to increase with age—and we observed a relationship in our own data (supplementary Table 1)—all data were age corrected using the partial correlation method. For the platelet and PBMC data, a reciprocal transformation was also performed after which the residuals were both randomly and normally distributed. Partial Spearman correlations were performed using the pcor function in R. Outliers were not removed for any of the regression analyses.