Cell culture
In this study, human neuroblastoma cells (SH-SY5Y cells) stably expressing human tau 3 repeat isoform 2+3-10- (designated as SH-Tau3R cells, generously provided by Luc Buée) were used. The SH-Tau3R cells were cultured as previously described 74. Briefly, the cells were grown in DMEM/High glucose medium (11995-065, ThermoFisher), supplemented with 10% bovine growth serum (BGS, heat inactivated, F1051-500ML, Sigma-Aldrich), 1% glutamine (25030081, ThermoFisher), and 1% penicillin/streptomycin (15140-122, ThermoFisher). The cell cultures were maintained in a humidified incubator with 5% CO2 at 37°C. The cells were grown either in 10 cm Petri dishes, 6-, 12-, or 96-wells plates.
Primary culture of neurons
For the primary neuronal culture, cortices of mouse embryos at embryonic day 16 (E15-E17) were used from transgenic mice B6.129S2Emx1tm1(cre)Krj/J, where Emx1-Cre mice were crossed with Red Fluorescent Protein-Lox mice (Jackson Laboratories). Briefly, brains embryos were dissected out, meninges, choroid plexus and hippocampus were removed to avoid contamination and cortices were mechanically and enzymatically disrupted in the presence of trypsin-EDTA 0,25% (Gibco) for 20 min at 37°C. The cell suspension was filtered through a 70 µm cell strainer and plated onto 6-well plates (200,000 cells/well), which were pre-coated with 50 µg/ml poly-D-lysine (A3890401, ThermoFisher), or on coverslips pre-coated with 1 µg/mL polyethylenimine (043896.03, ThermoFisher) and 50 µg/mL poly-D-lysine in 24-well plates (150,000 cells/well). The cells were firstly grown for 2 hours in DMEM/High glucose medium, supplemented with 10% BGS and 1% of streptomycin/penicillin antibiotics in a 5% CO2 humidified incubator at 37°C. Then, the culture medium was changed per a growth medium (NeurobasalTM medium (21103-049, ThermoFisher), 1% glutamine (25030081, ThermoFisher), 2% B-27 supplement (17504044, ThermoFisher), 1% N-2 (17502-048, ThermoFisher) and 1% penicillin/streptomycin). The cultures were maintained at 37°C in a humid atmosphere with 5% CO2 and a growth period of 4 days was allowed before any experimental treatment was administered.
Temperatures exposure and cell treatments
Prior to initiating any treatment, the cell culture medium was replaced with fresh DMEM/High glucose medium (without BSA) or Neurobasal medium, for SH-Tau3R or primary neuronal culture, respectively. Then, the cells were placed in dedicated CO2 incubators set to 35, 37, 38 or 39°C for a duration from 6 to 72 hours (Fig.1a, h). To inhibit caspase-3 activity, cells were treated for a period of 72 hours with the selective caspase-3 inhibitor z-DEVD-FMK (A13503; Adooq Biosciences) at a concentration of 20 µM 9 dissolved in a vehicle solution (phosphate-buffered saline (PBS) containing 0.1% of DMSO). Transfection of small interfering RNA was carried out using LipofectamineTM RNAiMAX transfection reagent (13778075, ThermoFisher) according to the manufacturer’s instructions. Briefly, for each transfection, cells were cultured for 72 hours in 1 ml of Opti-MEM (ThermoFisher) containing 40 µl of LipofectamineTM, and 100 nmol of respective siRNAs. The following siRNAs were used: Silencer® Pre-designed EXT1 siRNA (ID116802, ThermoFisher), Stealth RNAiTI SDC3 siRNA (HSS145253, ThermoFisher), and SignalSilence® Caspase-3 siRNA (6466S; Cell Signaling). SilencerTM select negative control siRNA (4390843, ThermoFisher) was used as the scrambled negative control.
Animals
In this study, three-months-old C57BL6 (males and females) and 18-month-old hTau (males) mice or their littermate control tau knockout (TKO; males) 75 were used. The hTau mice were generated by crossing mice expressing the 6 isoforms of nonmutated human tau (known as 8c mice) 76 with murine TKO mice 77. The founders of hTau and TKO colonies originated from a C57BL6 background (B6.Cg-Mapttm1(EGFP)Klt-Tg(MAPT)8cPdav/J, Jackson Laboratories). The animals were handled according to procedures endorsed by the “The Animal Care Committee of Université Laval (CPAUL-3, approbation number: CHU-22-1027)” under the guidelines of the Canadian Council on Animal Care. All mice had access to water and food ad libitum. The mice were housed in a 12 h light/12 h dark cycle, with the lights being turned on at 7:15 am. At the end of each experiment, mice were euthanized through decapitation without anesthesia, as anesthesia leads to tau hyperphosphorylation 78,79. The brains were promptly removed and cortices were dissected on ice, frozen on liquid nitrogen and stored at −80°C for further analysis
Sleeping vs. awake mice
Mice were subjected to a continuous period of darkness lasting for 3 days. The determination of subjective day was determined as previously described 20. Briefly, sleeping C57BL6 mice (n=5 males and n=5 females) were euthanized between 10:30 and 11:30 am local time (at Circadian Time 4 (CT4), 16 h after the onset of activity) and active mice (n=5 males and n=5 females) were euthanized between 10:30 and 11:30 pm local time (at CT16, 4 h after the onset of activity) (Fig. 5a). Furthermore, the sleeping criterion corresponded to mice in the nest, in a “resting posture”, as elucidated by Thoman and Carroll: absence of locomotor activity, absence of movement, absence of erect posture 80. The core BT of mice was assessed just before euthanasia with a rectal probe (RET-3, Brain Tree Scientific Inc) connected to a digital thermometer (Thermalert TH5; Physitemp).
Sleep deprivation
As previously described by our group 20, a subset of C57BL6 mice was intentionally kept awake for the first 6 hours of the light period (sleep deprivation (SD) group, n=9, males and females). Naive mice (n=7, males and females) were allowed to sleep without any disturbance. All mice were euthanatized by decapitation at the end of SD period (Fig. 4e). Prior to SD experiment, a subset of five mice of both groups was abdominally implanted with telemetric probes (BodyCap, Anipill) enabling continuous monitoring of their BT. The baseline BT was assessed the day preceding the SD protocol for the same set of animals.
Cold and heat exposures
On the day preceding the experiment, hTau mice were individually housed to prevent any mutual heating. For the entire duration of the study, the naive group (n=5) and the negative control TKO (n=3) remained at the standard temperature of the animal facility (22°C). As previously described by our group 21,81, the two other groups of mice underwent a 4-hour exposure period either at temperature of 4°C (n=3) or 38°C (n=5). The core BT of mice was assessed just prior to euthanasia utilizing a rectal probe (RET-3, Brain Tree Scientific Inc) connected to a digital thermometer (Thermalert TH5; Physitemp).
CSF collection
The mice were anesthetized with isoflurane and positioned on a stereotaxic instrument. To maintain core BT, a water heating pad was used. Under the observation of a dissection microscope, the subcutaneous tissues and muscles (m. biventer cervicis and m. rectus capitis dorsalis major) were gently separated via blunt dissection utilizing forceps. This separation facilitated the exposure of the dura mater of the cisterna magna. A capillary tube was introduced through the dura mater into the cisterna magna in order to induce the CSF flow into the capillary tube.
Protein extraction
The samples (cell lysates or mice cortices) were homogenized by sonication in Radioimmunoprecipitation assay (RIPA) buffer, then centrifuged for 20 min at 20,000g at 4°C. The resulting supernatant was collected, and the total protein concentration was assayed (Pierce™ BCA Protein Assay Kits, 23225, ThermoFisher). The samples were diluted in sample buffer (NuPAGE LDS; Invitrogen) containing 5% of 2-β-mercapto-ethanol, 1 mM Na3VO4, 1 mM NaF, 1 mM PMSF, 10 μl/ml of Proteases Inhibitors Cocktail (P8340; Sigma-Aldrich). The samples were then subjected to denaturation for 10 min at 95 °C.
Western blotting
Western blot analysis was conducted as previously described 82. 10-20 μg of the samples were separated on an SDS-10% polyacrylamide gel and transferred onto nitrocellulose membranes (Amersham Biosciences). The membranes were saturated, hybridized with the appropriate antibodies, and revealed as described in 82. For immunoblots targeting phospho-tau epitopes, the signal was normalized to the total tau protein. Used as a loading control, other proteins were normalized to β-actin. Representative lanes from the immunoblots were exhibited for each specific experimental condition. The dashed lines indicate segments where certain lanes from the same blot were excluded, and the remaining lanes were combined. Brightness levels were adjusted as necessary to enhance visualization and accuracy.
Antibodies
All antibodies used in this study, in addition to their dilution, are listed in Extended Data Table 3.
Dot blotting
The cell medium was harvested following appropriate treatments and centrifugated for 10 min at 20,000g at 4°C to remove cell debris. In order to assess extracellular content of proteins by dot blotting, 100 μl of cell medium were deposited onto nitrocellulose membranes (Amersham Biosciences), utilizing a microfiltration blotting apparatus (Bio-Dot Apparatus 1706545, Bio-Rad). The membranes were saturated, hybridized with appropriate antibodies (Extended Data Table 3) and revealed as described in 82. For dot blots targeting phospho-tau epitopes, the signal was normalized to the total tau protein. In the case of other proteins, the normalization was performed relative to the respective extracellular LDH value (CytoTox 96® Non-Radioactive Cytotoxicity Assay, Promega). Representative dots signal was exhibited for each specific experimental condition. The dashed lines indicate segments where certain dots from the same blot were excluded, and the remaining dots were combined. Brightness levels were adjusted as necessary to enhance visualization and accuracy.
Co-immunoprecipitation
Co-immunoprecipitation (co-IP) analyses were performed to determine interactions between PIP2 total tau and TauC3, following manufacturer’s instructions (Pierce™ Classic Magnetic IP/Co-IP Kit, 88804, ThermoFisher Scientific). Briefly, SH-Tau3R cells were harvested using lysis buffer, incubated at 4 °C for 5 min, and centrifuged at 13,000g to pellet cellular debris. The supernatants were collected, proteins levels were adjusted to 500 µg and primary antibodies (Extended Data Table 3) were added to samples, except for the negative control (NC) sample. The samples were then incubated overnight at 4 °C on a rotating device. Following this step, protein A/G magnetic beads (25 μl) were added to each sample and incubated for 1 hour with agitation at room temperature. The antibody-bound beads were extracted using a magnetic device and washed three times. The beads were dissociated using the elution buffer and separated magnetically. To neutralize the low pH environment, 10 µl of neutralization buffer were added to the supernatant. The resulting sample was diluted with sample buffer (NuPAGE LDS; Invitrogen) containing 5% of 2-β-mercapto-ethanol, 1 mM Na3VO4, 1 mM NaF, 1 mM PMSF, 10 μl/ml of Proteases Inhibitors Cocktail, and finally boiled at 95°C for 5 minutes. The proteins were analyzed using Western blot analysis.
Immunocytochemistry
Neurons from primary culture were fixed in PBS 1X (311-010-CL, Multicell) /4% paraformaldehyde (19210 Electron Microscopy Sciences, ThermoFisher)/10% sucrose (S53, ThermoFisher) for 20 min at room temperature. After washing 3 times with PBS 1X, cells were permeabilized in 0.2% Triton X-100 (T8787-100ML, Millipore) in PBS 1X for 30 min at room temperature and blocked with 5% Goat Serum Heat Inactivated (G6767, Millipore) in PBS 1X for 1 hour at room temperature. Then, cells were incubated with the primary antibodies against TauC3 and SDC3 (Extended Data Table 3) in 5% goat serum heat inactivated in PBS 1X at 4°C overnight. After washing 3 times with PBS 1X, the secondaries antibodies (anti-mouse Alexa Fluor 488 diluted at 1:1000 (#A-11029, ThermoFisher) and goat anti-rabbit IgG Alexa Fluor 633 diluted at 1:1000 (#A-21070, ThermoFisher)) were added for 2 hours. After 3 washes with PBS 1X, DAPI (4′,6-diamidino-2-phenylindole, (ThermoFisher) 3,5µL of DAPI in 25 mL PBS 1X) was used for nuclei staining and coverslips were mounted with Fluoromont-G (00-4958-02, Invitrogen). Sections were imaged on a Zeiss LSM800 confocal microscope system equipped with 405, 488, 561 and 640 nm lasers. Confocal images were acquired and mosaics created using the Zen Blue Edition software (v. 2.3, Carl Zeiss).
Caspase-3 activity assay kit
The SH-Tau3R cells were cultured in 96-well plates and treated according to appropriate experimental conditions (Fig. 1a). The colorimetric caspase-3 Assay Kit (ab39401, abcam) was used to determine the activity of caspase-3, and following manufacturer’s instruction.
Membrane fluidity
The SH-Tau3R cells were cultured within 96-well plates and subjected to treatment as outlined in the experimental groups (Fig. 1a). The membrane fluidity (ab189819, abcam) was assessed according to manufacturer’s instructions. Briefly, the cells were incubated 1 hour at temperatures of 35°C, 37°C or 38°C in a cell medium supplemented with 5 µM of Fluorescent Lipid Reagent and 0.08% Pluronic F127. The fluorescence intensity was then measured (Infinite F200, Tecan) at wavelengths of 400nm and 470nm, using the appropriate filter for excitation at 350nm. The recorded fluorescence values were corrected by subtracting the corresponding blanks from each sample, and the fluorescence ratio of excimer emission (470nm) to monomer emission (400nm) was calculated.
ELISA assays of extracellular tau
Total and phosphorylated tau concentrations within the cell medium were quantified using ELISA kit: Tau (total) Human KHB0041; Tau [pS199] Human KHB7041; Tau [pT231] Human KHB8051; Tau [pS396] Human KHB7031; Tau (total) Mouse KMB7001 (ThermoFisher). Prior to analysis, the samples were suitably diluted in diluent buffer (1:50 for human tau, 1:2 for mouse tau and phospho-tau). The ELISA assays were performed in accordance with the instructions provided by the manufacturer.
Quantitative PCR
Total RNA was isolated from SH-Tau3R cells using TrizolÒ reagent (Life Technology) in accordance with the manufacturer’s instructions. The quantification of RNA was conducted, and 1 mg of total RNA was used for cDNA synthesis using the iscriptTM cDNA Synthesis Kit (Biorad), containing an optimal blend of oligo-dT and random primers. For subsequent PCR amplification, 1 µl of the resultant cDNA was used as template. The primer sequence used for the PCR amplification are reported in Extended Data Table 3. The qPCR mix was formulated with 18 µL per 2 µL of 20 ng cDNA. The mix consisted of 0.5 µL of both the forward and reverse primers, 10 µL of SYBR Green PCR Master Mix (Applied Biosystems), and 7.5 µL of nuclease free water. The qPCR program began with a hot start at 95°C for 3 minutes, succeeded by 40 cycles at 95°C for 15 seconds, followed by 60°C for 1 minute, using a LightCycler 480 II apparatus (Roche). The melting curves were evaluated to ensure a single PCR product. To quantify cDNA levels, the comparative 2ΔΔCt method was employed. Ct values corresponding to the target gene were normalized to the Ct values of the house-keeping gene GAPDH (glyceraldehyde 3-phosphate dehydrogenase). The results were expressed as n-fold differences relative to the experimental control.
Human Studies
CSF temperature correlations: Detailed information about participants, CSF collection, and study design can be found in Lucey et al. 83. Thirteen participants who completed the placebo group of a recently published clinical trial had 6 ml of CSF collected every 2 hours for 36 hours via an indwelling lumbar catheter 83. All participants were cognitively unimpaired and in good general health except for poor sleep efficiency <85% measured by actigraphy. Body temperature was recorded every four hours with a temporal forehead thermometer (Adc Adtemp 427, American diagnostic Corp, United-States). CSF tau forms (T181, S202, T217) were measured by immunoprecipitation/mass spectrometry as previously described 83. NfL protein levels were quantified using the NF-lightTM ELISA kit (UmanDiagnostics, Umea, Sweden) following the manufacturer’s protocol. The assay’s measurement range is 100 pg/ml to 10,000 pg/ml with a detection threshold of 33 pg/ml. CSF samples were prepared through dilution with an equal volume of Sample Diluent, achieving a 1:100 dilution ratio, to ensure a volume suitable for analysis. The quantitation process entailed the enzymatic conversion of a colorless substrate into a colored product indicative of the NfL concentration in the samples. Absorbance readings were taken at 450 nm with a reference wavelength of 620-650 nm. To ensure consistency, samples with known high levels of NfL (“bloody CSF”) were utilized as positive controls on each plate diluted to 1:1000. CSF tau-181, tau-202, and tau-217 concentrations were averaged at 8AM, 4PM, and 8PM. Differences between temperature, CSF tau levels, and CSF NfL levels were then calculated for use in the analyses, and detailed data for each participant are provided in (Extended Data Table 1). NfL was selected as control protein because its soluble concentration is not affected by sleep-wake activity 14. We selected time points of 8 AM vs 4 PM for post-sleep vs post-wakefulness tau levels, which also corresponded to the minimum and maximum BT, respectively. These intervals were selected based on the following rationale. First, we assumed little delay between tau secretion and appearance of tau in the CSF, meaning that BT taken at the time of CSF collection would roughly reflect brain temperature at the time of tau secretion. Second, we selected 8 am vs 4 pm as the interval that maximized that difference in sleep vs wake temperatures, given that BT was not recorded during sleep, and for the majority of participants, had already begun to drop between 4 pm and 8 pm (Extended Data Table 1).
Plasma temperature correlations: Data were collected from 24 older adults 68.39±5.25 years of age, 17 of whom were female. Subjects were enrolled in cross-sectional study examining the relationship between core BT and plasma and PET AD biomarkers. Subjects were cognitively normal (n=21) or had mild cognitive impairment (n=3) as determined by the clinical dementia rating scale (CDR), and were medically healthy with only mild, or no sleep apnea. Prior to the study, participants were screened with interviews and one week of home actigraphy for sleep-wake disorders including sleep less that 6 hours per night, significant phase advance or phase-delay. Additional exclusion criteria are detailed in 62 and included AD dementia (CDR > 0.5), medical comorbidities and the use of medications that might affect sleep or thermoregulation, major psychiatric disorders and moderate-severe substance use disorders, shift work within the last 6 months, or traveling across 1 or more time zones within 2 weeks of study participation.
The study design was a semi-naturalistic protocol fully detailed in 62. Briefly, participants underwent continuous measurement of core body temperature using an ingestible telemetric device (Cortemp, HQInc) that sampled temperature every 15 seconds with an accuracy of 0.2°C for a minimum of 36 hours spanning 2 nights. During this time, 2 in-lab nocturnal polysomnograms were measured, and participants were free to behave as they chose during the intervening day between the lab nocturnal recordings. The goal of this design was to capture data that most closely represented the typical BT for each participant. Blood draws for plasma tau were collected on four occasions, in the mornings (7:00 am) and evenings (7:00 pm) on both mornings and nights (Extended Data Table 2). Prior to analysis, temperature data were preprocessed to exclude gaps and artifacts as detailed in 62. Data presented for BT-tau correlations comprised tau levels from night 2 and morning 2, given BT was not always measured prior to blood draw on night 1 (Table S3). Paired tau levels and BT data were obtained for 15 subjects (Extended Data Table 2). For BT–tau correlations, the difference between the average BT between 6–7 pm and 1–2 am was calculated. This interval was chosen because these times represented the sample average minimum and maximum BT, and as such their difference maximized the diurnal BT difference, or ∆BT. Plasma sampling times were selected to maximize efficiency in collecting data. Food intake was not regulated, but the morning sample was typically before the morning meal, whereas the evening sample was typically before the evening meal. Concentrations of plasma tau were measured using the neurology 3-PLEX kit and Simoa HD-X instruments (Quanterix, Billerica, MA, USA) at the NYU Alzheimer’s Disease Center Biomarker Core according to the manufacturer’s instructions. Plasma extraction was performed as described previously. Assays were run in duplicate to obtain inter-assay coefficient of variations (CVs). The inter-assay CV was under 20% for all samples.
Statistical analysis
A minimum of two distinct experiments were carried out for each experimental condition. Prior to conducting each analysis of variance, an assessment of Gaussian distribution was performed, and its validity was confirmed through a Kolmogorov-Smirnov test (utilizing GraphPad Prism 9.0). Depending on the specific analysis, two-tailed t-tests (or Mann-Whitney tests), as well as one- or two-way ANOVAs (or Kruskal-Wallis tests), were applied. Post hoc analyses, involving either Tukey’s or Dunnett’s tests, were subsequently employed. A significance threshold of P < 0.05 was employed to determine statistical significance. The presentation of data incorporated either box and whisker plots (illustrating the range from minimum to maximum values, encompassing the median) or mean ± standard error of the mean (s.e.m). The scatter plots depicted on each graph provide an indication of the number of data points, and detailed statistical information is provided in Extended Data Table 4.