Transthyretin preparation and aggregate formation
Recombinant WT and L55P TTR were expressed and purified from Escherichia coli BL21 (DE3) as described previously39. A 0.4-mg/mL sample of transthyretin in a HEPES buffer (50 mM HEPES, 150 mM NaCl, pH 6.8) was mixed with an acetate buffer (200 mM sodium acetate, 50 mM NaCl, pH 4) at a ratio of 20:20 μL and kept at 37 °C for 3 days. The pH 4.5 and pH 5 acetate buffers had the same composition, but a pH 7 solution was composed of a 50-mM NaCl solution. A 40-µL sample of aggregate was centrifuged at 20,000g at 4 °C for 5 min. After removing the supernatants, sediment aggregates were dissolved in a triple-diluted HEPES buffer (17 mM HEPES, 50 mM NaCl, pH 6.8) for pulse application.
1.26 MV/cm PEF generation
We used a nanosecond pulse to prevent plasma formation in the PEF exposure chamber. Our high-voltage nanosecond pulse generator is similar in structure to the one described in the reference40 and consists of a spark gap-driven 10-stage Marx circuit with an output capacitance of 94 pF; and a pulse-sharpening section, including a coaxial capacitor with a capacitance of 12 pF, a spark gap as the output switch, and a tail-cut switch to reduce the pulse to 1 ns. The Marx circuit was placed in a polycarbonate container pressurized to 0.57 MPa with nitrogen gas and charged to 15 kV using a high-voltage DC supply (HAR-50R0.6, Matsusada). The pulse-peaking capacitor and the tail-cut switch were placed in an aluminum container pressurized to 0.5 MPa with sulfur hexafluoride gas. The elevated high-voltage pulse was delivered to the pulse-peaking capacitor to be doubled and then quickly released to a 50-Ω coaxial cable leading to the PEF exposure chamber by closing the output switch (Fig. 1C–E; SI Fig. S2A–E).
Voltage and current sensors used a capacitive divider and a pick-up coil, both of which were integrated in the exposure chamber, respectively. The signals were acquired by a 16-GHz oscilloscope (DPO71604C, Tektronix). The voltage sensor was calibrated by a calibrated resistive divider with a range of 50 kV. Numerical simulation of the pulse delivery using CST Studio Suite (SIMULIA) validated the measurements (SI Fig. S2F–O).
The 1 ns, 126 kV pulses were delivered repeatedly to the coaxial PEF exposure chamber with a 1-mm gap, and parallel-plane electrodes 5.5 mm in diameter, to generate an electric field exceeding 1 MV/cm (SI Fig. S2P–R). Deviations of maximum voltage amplitude during the repetitive operation were approximately 5%. The electrodes were made of stainless steel (SUS316), and the anodic electrode was gold coated to minimize the chemical influence of metal ions from the electrodes. The electrical resistance and the estimated capacitance of the electrodes, including samples, were 77 Ω and 17 pF, respectively, resulting in a total impedance of 34 Ω at 250 MHz. The measured voltage at the electrodes was 126 kV, which was 20% lower than the expected voltage at a load of 50 Ω, because of negative mismatching.
The pulse repetition rate was fixed at 1.7 Hz to prevent a significant temperature rise during PEF exposure. Negative controls (NCs) were aggregate or tetramer transthyretins dissolved in a 1.26 MV/cm nsPEF-treated triple-diluted HEPES buffer (17 mM HEPES, 50 mM NaCl, pH 6.8). The conductivity of the solution was 0.578 S/m (LAQUAtwin-EC-33, HORIBA).
H2O2 detection, pH measurement, and temperature measurement
We used an Amplite Fluorimetric Hydrogen Peroxide Assay Kit *Near Infrared Fluorescence* from CosmoBio to measure H2O2 concentrations. We prepared a working solution (Amplite IR Peroxide Substrate, 0.8 U/mL peroxidase) and mixed the solution with miliQ at a ratio of 100:100 μL. The mixed solution was exposed to 1.26 MV/cm for two pulses and collected. Because the maximum treatable amount was 25 μL, we combined nine samples of two-pulse solutions to prepare 200 μL of the treated sample. A Quantus fluorometer measured the sample fluorescence with a red fluorescence filter to calculate the concentration of H2O2.
A LAQUAtwin compact pH meter was used to measure the pH of pre- and post-treated liquids. Because at least 100 μL is required for a measurement, we prepared five samples of 1,000-pulse solutions for the 100 μL treated samples.
An AMOTH FL-2400 fiberoptic thermometer with a FS300-2M probe measured on-time temperature of the pulse-treated solution.
The starting temperature was a room temperature of approximately 18 °C. The head of the FS300-2 M fiber probe received a fluorescent coating with a signal intensity that depended on temperature. An AMOTH FL-2400 fiberoptic thermometer exposed the coating to a laser and detected the signal intensity. As shown in SI Fig. S3E,F, the fiber probe was placed in the sample solution (triple-diluted HEPES buffer without transthyretin) through the electrode hole, and we were able to measure the temperature in real time.
Native PAGE used a 4% stacking gel (4 w/v% acrylamide/bis mixed solution 29:1, 0.125 M Tris–Cl pH 6.8, 0.09 w/v% ammonium peroxodisulfate solution, 0.08 v/v% N,N,N′,N′-tetramethylethylenediamine) and 14% separation gel (14 w/v% acrylamide/bis mixed solution 29:1, 0.375 M Tris–Cl pH 8.8, 0.09 w/v% ammonium peroxodisulfate solution, 0.08 v/v% N,N,N′,N′-tetramethylethylenediamine). Next, 5.5 µL of transthyretin solution and 5.5 µL of the sample buffer (0.1 M Tris–Cl pH 6.8, 20 v/v% glycerol, 0.025 w/v% bromophenol blue [BPB]) were mixed, and 10 µL of mixed solutions were applied to the wells of the gel. A Mini300 electric power source obtained from AS ONE applied a constant 15 mA current for 200 min for electrophoresis.
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis
SDS-PAGE used 4% stacking gel (4 w/v% acrylamide/bis mixed solution 29:1, 0.125 M Tris–Cl pH 6.8, 0.1 w/v% SDS, 0.09 w/v% ammonium peroxodisulfate solution, 0.08 v/v% N,N,N′,N′-tetramethylethylenediamine) and 15% separation gel (15 w/v% acrylamide/bis mixed solution 29:1, 0.375 M Tris–Cl pH 8.8, 0.1 w/v% SDS, 0.09 w/v% ammonium peroxodisulfate solution, 0.08 v/v% N,N,N′,N′-tetramethylethylenediamine). Transthyretin solution and a sample buffer (0.1 M Tris–Cl pH 6.8, 20 v/v% glycerol, 4 w/v% SDS, 12 v/v% 2-mercaptoethanol, 0.025 w/v% BPB) were mixed at ratio of 1:1 and heated at 100 °C for 10 min. After cooling, the samples were centrifuged at 20,000g for 1 min. Next, 3.5 µL of mixed solutions at 0.1 mg/mL and 7 µL of mixed solutions at 0.05 mg/mL were applied to wells of the gel. A Mini300 electric power source obtained from AS ONE applied a constant 18 mA current for 110 min for electrophoresis.
Band-intensity distribution analysis
ImageJ software was used to analyze the electrophoresis band-intensity distributions. For the native PAGE and SDS-PAGE results, 10 vertical lines were drawn across each band, with ImageJ determining the intensity distribution through the lines for each band and presenting the average distributions measured through 10 vertical lines for each band.
Aggregate fluorescence analysis
We used a fluorescent microscope (Leica, DMi8) combined with a digital camera (Canon, EOS 8000D) to observe aggregate fluorescence. Aggregate (0.1 mg/mL) was dissolved in thioflavin-T-glycine buffer (25 mM glycine, 10 µM thioflavin-T) and incubated on ice for 3 min without light. A green fluorescence protein filter was used.
A protein assay BCA kit from Wako was used to measure the amount of transthyretin aggregate. Color reaction occurred at 37 °C for 60 min. An iMark microplate reader from BIO-RAD measured absorbance at 570 nm. Aggregate absorbance, without PEF application, was calculated by subtracting the raw absorbance of a 0-pulse aggregate from that of a no-pulse triple-diluted HEPES buffer. Absorbance of 1,000 pulses at 1.26 MV/cm was acquired by subtracting the raw absorbance of treated aggregate from that of an exposed triple-diluted HEPES buffer.
Data were presented as the mean ± standard error for n samples (as shown in Figs. 3E, 5E, 6D). Statistical analyses were performed using a two-tailed t test; p < 0.01 was considered statistically significant.