Intracellular Aβ42 Aggregation Leads to Cellular Thermogenesis

The aggregation of Aβ42 is a hallmark of Alzheimer’s disease. It is still not known what the biochemical changes are inside a cell which will eventually lead to Aβ42 aggregation. Thermogenesis has been associated with cellular stress, the latter of which may promote aggregation. We perform intracellular thermometry measurements using fluorescent polymeric thermometers to show that Aβ42 aggregation in live cells leads to an increase in cell-averaged temperatures. This rise in temperature is mitigated upon treatment with an aggregation inhibitor of Aβ42 and is independent of mitochondrial damage that can otherwise lead to thermogenesis. With this, we present a diagnostic assay which could be used to screen small-molecule inhibitors to amyloid proteins in physiologically relevant settings. To interpret our experimental observations and motivate the development of future models, we perform classical molecular dynamics of model Aβ peptides to examine the factors that hinder thermal dissipation. We observe that this is controlled by the presence of ions in its surrounding environment, the morphology of the amyloid peptides, and the extent of its hydrogen-bonding interactions with water. We show that aggregation and heat retention by Aβ peptides are favored under intracellular-mimicking ionic conditions, which could potentially promote thermogenesis. The latter will, in turn, trigger further nucleation events that accelerate disease progression.

Aβ42 and MJ040X were incubated for 24 hours before imaging. FCCP (Merck KGaA, Darmstadt, Germany) treatment was performed during the day of imaging, with FCCP added at a concentration of 10 μM (from a 10 mM stock in DMSO) and cells were left to incubate for 30 minutes. All reagents were kept at -20 o C and aliquoted to prevent freeze-thaw cycles.

AβM1-42 Purification
Recombinant AβM1-42 (also referred to as Aβ) was purified as described in Stephens et al. 1 . The plasmid pET3a containing AβM42 cDNA was transformed into Escherichia coli (E. coli) One Shot BL21 (DE3) pLysS (ThermoFisher Scientific). Liquid culture of E. coli was induced for expression for 4 hours when the OD600 reached 0.6−0.8 by the addition of 1 mM isopropyl-β-thiogalactopyranoside (IPTG). The cells were pelleted in 50 mL volumes and the supernatant discarded. The pellet contained Aβ in inclusion bodies. The inclusion bodies were washed four times in wash buffer, as detailed in 1 to obtain clean and pure inclusion bodies. The inclusion body pellet from 50 mL of culture was placed on ice with a small magnetic stir bar on a magnetic stirrer. 200 μL of 6 M guanidinium chloride (GuHCl) was added to the pellet and stirred vigorously for 30 min to solubilise the Aβ. 15 mL of ice-cold ion exchange (IEX) buffer A (10 mM Tris, 1 mM ethylenediaminetetraacetic acid (EDTA), pH 9) was added slowly to the solubilised pellet to dilute the 6 M GuHCl and to permit binding of Aβ to the ion exchange column. The solubilised Aβ was filtered through a 0.22 μm filter (Millex-GP, MillPore, Merck KGaA) before being placed on ice prior to chromatography. Aβ was loaded onto a HiTrap Q HP column (GE, Healthcare, Chicago, IL, USA) and eluted against a linear gradient of IEX buffer B (10 mM Tris, 1 mM EDTA, 0.75 M NaCl, pH 9) over seven column volumes followed by two column volumes of 100% buffer B. Purification was performed on an ÄKTA Pure Fast Protein Liquid Chromatography (FPLC) and monitored by absorption at 280 nm (GE Healthcare). The concentration of Aβ was determined by absorption at 280 nm on a NanoVue spectrometer (Biochrom Ltd., Cambridge, UK) using the extinction coefficient of 1490

Fixing of cells for the immunostaining of Aβ42
Cell media was removed and replaced with 4% paraformaldehyde (PFA, Merck KGaA) diluted in 1×PBS.
The sample was fixed for 10 minutes, before blocking with 5 w/v% bovine serum albumin (BSA, Merck KGaA) diluted in PBS for 1 hour. Between antibody incubation, three washes of 50 μM Triton X-100 (Invitrogen, ThermoFisher Scientific) in PBS (henceforth referred to as PBST) were performed. Primary and secondary antibodies used were β-Αmyloid(1-42) polyclonal antibody (Invitrogen, ThermoFisher Scientific) and Goat anti-Rabbit Alexa Fluor 647 (ThermoFisher Scientific), diluted by 1:100 and 1:400 in PBST respectively. The sample was kept at room temperature throughout all the steps and wrapped in aluminium foil to prevent any bleaching especially after the addition of the secondary antibody. For storage before imaging, the sample was kept at 4 o C in dark conditions.  where, and are the semi-major and semi-minor axes of an ellipse, respectively. Aggregate masking was achieved by an intensity threshold, following a pixel-size threshold to ensure that only aggregates with dimensions above the resolution limit (i.e., estimated at 70 nm) were included in the analysis.

ATeam1.03 ATP FRET sensor
Design of the Förster resonance energy transfer (FRET-)based, cytosolic ATP sensor, ATeam1.03 can be found in Imamura et al. 9 . Two days before imaging, 200 ng ATeam1.03 plasmid diluted in antibioticsfree DMEM (ThermoFisher Scientific) was transfected using Lipofectamine 2000 (ThermoFisher Scientific) four hours after which the medium was changed, and the cells were kept incubated for another 20 hours before imaging. TCSPC-FLIM was only performed on the CFP donor channel whilst the YFP channel was used as a guide to confirm that the fluorescence seen in the CFP channel is due to  However, it should be noted that the aggregates formed were significantly smaller than the equivalent in the presence of pre-formed seeds. 15 Moreover, in comparison to the more aggressive familial mutant, E22G-Aβ42, structures formed do not display the same degree of polymorphism (i.e., ranging from clusters to bundles and large perinuclear aggresomes). 16 Supplementary Figure    Temperature was determined by a thermocouple inserted into the cell medium, and an objective warmer (to prevent the heat sink effect by the objective) was used in conjunction with a stage top heater. To validate the ability of the system setup in terms of temperature stability and ramping, we first tested the system with Rhodamine B (Supplementary Figure 4a), a standard fluorescence dye that Calibration in live HEK293T     Exponential fitting* 5-8.5 5-7.5 Phasor plots** N/A 7.8-11.5