Human brain samplesET cases

The analyses included brains from 18 ET cases derived from the Essential Tremor Centralized Brain Repository (ETCBR), a longstanding collaboration between investigators at the University of Texas Southwestern Medical Center and Columbia University [4, 22, 23]. Established in 2003, the ETCBR banks brains from ET cases throughout the United States. ET diagnoses were carefully assigned by a neurologist specializing in tremor (E.D.L.), using three sequential methods, as has been our practice for 20 years, and as employed in over 50 publications [3, 22, 24]. First, the clinical diagnosis of ET was initially assigned by the treating neurologists, and second, confirmed by E.D.L. using semi-structured clinical questionnaires, medical records, and Archimedes spirals with the following criteria: (i) moderate or greater amplitude kinetic tremor (rating of 2 or higher [25]) in at least one of the submitted Archimedes spirals; (ii) no history of PD or dystonia; and (iii) no other etiology for tremor (e.g., medications and hyperthyroidism) [22, 24]. Third, a detailed, videotaped, neurological examination was performed, from which action tremor was rated and a total tremor score assigned (range 0–36 [maximum]) [22, 24]. These data were used to assign a final diagnosis of ET [22, 24], reflecting published diagnostic criteria [moderate or greater amplitude kinetic tremor (tremor rating ≥ 2) during three or more activities or a head tremor in the absence of PD or other known causes] [25], of demonstrated reliability and validity [26, 27]. No ET cases reported a history of traumatic brain injury, exposure to medications associated with cerebellar toxicity (e.g., phenytoin, chemotherapeutic agents), or heavy ethanol use [28]. Every 6–9 months, a follow-up semi-structured telephone evaluation was performed, and hand-drawn spirals were collected; a detailed, videotaped, neurological examination was repeated if there was concern about a new, emerging movement disorder.

Other tauopathy cases

The tauopathy cases used for comparison encompassed AD (N = 8), PART (N = 2), CBD (N = 8), PSP (N = 5), and CTE (N = 2) cases. The selected cases had a single confirmed histopathological diagnosis of either AD, PART, CBD, PSP, or CTE based on standard neuropathological evaluation using accepted neuropathologic consensus criteria [29,30,31,32,33]. The AD, PART, CBD, and PSP human brain tissue samples were obtained through the University of Texas Southwestern Medical Center Neuropathology Brain Bank with Institutional Review Board (IRB) approval. Select AD, PSP, and CBD cases were also obtained from the Alzheimer’s Disease Research Center (ADRC) at Washington University in St Louis. The CTE cases were obtained from the CTE Center at Boston University. A summary of these tauopathy cases is detailed in Table 1S. The frontal gyrus was used for analysis in the AD, CBD, CTE, and PSP cases, except for three AD cases in which the parietal lobe was used (AD5, AD6, and AD7) and one case in which the temporal pole was used (AD8). The temporal lobe was used in the PART cases. Of note, the PART cases were analyzed and processed concurrently with the ET cases using the methods detailed below. The AD, CBD, PSP, and CTE tissue samples had been already prepared and analyzed in our prior work following the same protocols [21]. The results from these analyses were incorporated in this study for comparison with the ET samples.

Tissue processing and neuropathological examination of ET cases

ET brains from the New York Brain Bank had a complete neuropathological assessment with standardized measurements of brain weight and postmortem interval (hours between death and placement of brain in a cold room or upon ice) [26, 27]. Standardized blocks were harvested from each brain and processed, and 7 μm-thick formalin-fixed paraffin-embedded sections were stained with Luxol fast blue/hematoxylin and eosin (LH&E) [22, 34]. Additionally, selected sections were stained by the Bielschowsky method, and with mouse monoclonal antibodies to phosphorylated tau (clone AT8, Research Diagnostics, Flanders, NJ) and β-amyloid (clone 6F/3D, Dako, Carpenteria, CA) [22, 34]. All tissues were examined microscopically by a senior neuropathologist blinded to clinical information [22, 34]. Alzheimer’s disease staging for neurofibrillary tangles (Braak NFT) [35], Consortium to Establish a Registry for Alzheimer’s disease (CERAD) ratings for neuritic plaques [26, 36], and Thal β-amyloid stages were assigned [37, 38]. For Braak NFT staging, stages I and II indicated NFTs confined to the entorhinal region, III and IV limbic region involvement (i.e., hippocampus), and V and VI moderate-to-severe neocortical involvement [35]. An additional classifier (PSP) was added for subjects that had AT8 immunohistochemical tau deposits including neurofibrillary tangles or pretangles in at least two of three regions (globus pallidus, subthalamic nucleus, and substantia nigra), and tufted astrocytes in at least one of two regions (peri-Rolandic cerebral cortex and putamen), according to the most recent diagnostic consensus criteria for PSP reported by Roemer et al. in 2022 [32]. The level of AD neuropathologic change (ADNC) was rated according to National Institute on Aging and Alzheimer’s Association (NIA-AA) guidelines using an “ABC” score as none, low, intermediate, or high [37].

One ET case with pathologically confirmed high ADNC was examined in this study as a positive control for detection of AD-type tau seeds (Table 1, case 1). The remaining ET cases were chosen based on the following criteria: (1) presence of no (Thal A0, CERAD C0) or few amyloid deposits (Thal A1 or CERAD C1); (2) higher Braak NFT stage (5–6), reflecting at least some extension into frontal association cortices and/or temporal pole; and/or (3) presence of PSP-type changes. Frozen samples from cerebral cortex regions including frontal (BA9), motor (BA4), parietal (BA7), occipital (BA31), temporal (BA37), and temporal pole (BA38) were assayed according to sample availability. In cases with PSP-type changes detected, cerebellar cortex and dentate nucleus were assayed, as available. Cases that had only focal PSP-type pathology not meeting current Rainwater criteria [32] were designated as “PSP-like.”

Tissue homogenization

Frozen brain was suspended in Tris-buffered saline (TBS) containing cOmplete mini protease inhibitor tablet (Roche) at a concentration of 10% weight/volume. Homogenates were prepared by probe homogenization using a Power Gen 125 tissue homogenizer (Fisher Scientific) in a vented hood. The homogenates were then centrifuged at 21,000×g for 20 min. The supernatant was collected as the total soluble protein lysate. Protein concentration was measured using a Pierce 660 assay (Pierce). Fractions were aliquoted into low-binding tubes (Thermo Fisher) and frozen at − 80 °C until future use.

Tau RD biosensor cell lines

The following tau repeat domain (RD) biosensor cell lines were used for tau seeding assays in this study.

Tau RD 3R-Cer/4R-Ruby (3R/4R) cells

HEK293T cells containing the 3-repeat (3R) version of the wild-type (WT) human tau RD (residues 246-408) C-terminally fused to mCerulean (Cer), and the 4R version fused to mRuby fluorescent protein (Rub) were used for tau seeding assays and the 3R/3R tau Ala scan. These biosensor cells exhibit high seeding after exposure to brain homogenates from certain tauopathies (e.g., AD) but not others (e.g., CBD) [21].

Tau RD 4R-Cer/4R-Ruby (4R/4R) cells

HEK293T cells containing the 4R version of the WT human tau RD (residues 246-408), C-terminally fused to monomeric Cerulean3 (Cer) and monomeric Ruby fluorescent protein (mRuby3) were used for tau seeding assays and the 4R/4R tau Ala scan. These biosensor cells exhibit high seeding after exposure to brain homogenates from certain tauopathies (e.g., CBD, PSP) but not others (e.g., AD) [21].

Tau (246-408) Ala scan library

The tau-(246-408)-Ala-mEOS3.2 point mutant library used in this study was previously generated and described in detail [21]. Twist Biosciences synthesized gene fragments encoding the human 2N4R tau sequence from residues 246 to 408 with Ala codon substitutions (GCC) at each position. These gene fragments were then conjugated to a common lentiviral plasmid harboring the mEOS3.2 fluorescent protein, thus generating an arrayed library of plasmids with the different sequential tau Ala point mutants fused to mEOS3.2. All plasmids were verified by Sanger sequencing.

Biosensor cell transduction with brain homogenates and tau seeding assay

Tau RD 3R-Cer/4R-Ruby (3R/4R) and Tau RD 4R-Cer/4R-Ruby (4R/4R) cells were plated at a density of 20,000 cells per well of a 96-well plate. 24 h later, the cells were transduced with a complex of 10 μg of 10% w/v brain homogenate, 0.75 μl of lipofectamine 2000 (Invitrogen), and 9.25 μl of Opti-MEM for a final treatment volume of 20 μl per well. 72 h after transduction, cells were harvested using 0.25% trypsin digestion for 5 min at 37 °C, quenched with Dulbecco’s Modified Eagle Medium (DMEM), and transferred to 96-well U bottom plates. They were then centrifuged for 5 min at 1500 rpm, the supernatant was aspirated, and the cells were fixed in 2% paraformaldehyde in PBS for 10 min. The cells were then centrifuged and resuspended in 150 μl of PBS for analysis using flow cytometry.

Ala scan

For the Ala scan incorporation assay, 3R/4R or 4R/4R biosensor cells were plated at 20,000 cells per well in 96-well plates and treated with sonicated brain homogenates as described above. After 48 h, once seeded aggregates were observed, the cells were replated at a concentration ratio of 1:6 into 6 new 96-well plates to generate technical replicates. The cells were then treated with a library of tau-(246-408)-Ala-mEOS3.2 lentivirus. One tau Ala point mutant was added per well. 72 h after transduction, the cells were harvested and fixed in paraformaldehyde as described above. The Ala scan assay determines the tau residues that are critical for monomer incorporation into seeded aggregates in each tauopathy. If a residue in tau-(246-408) is critical for incorporation into a particular tauopathy fold, substitution of that residue with Ala will block monomer incorporation, resulting in no or low detectable seeding assessed by FRET between Cer and mEOS3.2. Conversely, if a residue is non-critical for incorporation, the Ala substitution at that position will have no effect, and FRET signal will be detected. This assay generates a barcode map of critical residues for each tau fold, which can be matched to the cryo-EM structure [21].

Flow cytometry

The BD LSRFortessa flow cytometer was used to perform Fluorescence Resonance Energy Transfer (FRET) analysis. To measure CFP or Cer signal, and FRET, cells were excited with the 405 nm laser, and fluorescence was captured with a 488/50 nm and 525/50 nm filter. To measure YFP and mEOS3.2, cells were excited with a 488 nm laser and fluorescence was captured with a 525/50 nm filter. To measure Ruby, cells were excited with a 561 nm laser and fluorescence was detected using a 610/20 nm filter.

Data analysis

Seeded 3R/4R or 4R/4R biosensor cells transduced with the tau-(246-408) Ala scan library were analyzed by gating for homogeneous side-scatter and forward-scatter. The FRET signal between Cer and Ruby, measured in the Pacific Blue and Qdot605 channels, was used to gate for cells that contained seeded aggregates in these biosensors. Within this population, a narrow gate for cells positive for FITC (non-photoconverted mEOS3.2) was selected, which corresponded to cells expressing the tau-(246-408)-Ala-mEOS3.2 point mutants. The FRET signal between Cer and mEOS3.2 was plotted within this population, and a narrow subset of bright cells were selected to calculate the median fluorescence intensity (MFI) in the AmCyan channel. All gates were kept constant across the incorporation assay and across technical replicates. The fluorescence intensity of the FRET signal in the AmCyan channel was subsequently used for the incorporation assay analysis as it served as a marker of the degree of incorporation of each tau-(246-408) Ala point mutant to the seeded aggregates within the biosensor cells. Of note, Ala point mutants K274A, T386A, and P397A were eliminated from the scan analysis due to low transduction efficiency.

Data processing

FRET median fluorescence intensity (MFI) values between Cer and mEOS3.2 measured in the AmCyan channel were normalized by plate to prevent batch variation. Within our analysis, most Ala residues in the N- and the C-termini of the sequence did not impact incorporation. Thus, we used these values, and the lowest value in each plate to normalize the data and obtain the Incorporation Value for each residue position, following this formula:

$$Incorporation Value x = \frac},$$

where Residue MFI x is the FRET MFI in the AmCyan channel for each position, Minimum MFI is the minimum FRET value in the scan, and Average MFI is the Average MFI for the first 20 residues (on the first plate) or the last 10 residues (on the second plate). The average of three technical replicates was used for downstream analysis.