Brain tissue homogenization and biochemical measures (Aβ1–42, Aβ1–40, P-tau and T-tau)

Frozen post-mortem tissue (20 mg) from the MTG and subjacent white matter representing Brodmann area 22 (BA22) was obtained for 26 participants with serial plasma measures. Each sample was homogenized in 10% (w/v) Tris-buffered saline (TBS + protease inhibitors, cOmplete™, Roche) and sonicated for 1 min using a sonicating probe at 4 °C. The tissue lysate underwent ultra-centrifugation at 100,000 x g for 1 h (4 °C) and the supernatant was collected (TBS fraction [T]). Two % (w/v) sodium dodecyl sulfate (SDS) was added to the remaining pellet that was sonicated further for 1 min at 4 °C and centrifuged at 25,000 g for 1 h (4 °C). The supernatant was collected as the SDS fraction [S]. Lastly, 70% (w/v) formic acid was added to the remaining pellet, sonicated and spun under the same conditions as the SDS fraction. Subsequently, the supernatant was collected and neutralized at 1:20 with 1 M Tris for immunoassay analysis (formic acid fraction [F]). Tissue homogenates (T, S and F) were analysed for Aβ_1–42, Aβ_1–40, P-tau and T-tau by the human Luminex 4-plex xMAP assay (Millipore; HNABTMAG-68 K) at the Maurice Wohl Clinical Neuroscience Institute, London, UK. Briefly, both fractions T and S were diluted 1:2 and fraction F was undiluted (post pH neutralization at 1:20). Median fluorescent intensity (MFI) was measured using the Luminex 200 using Xponent 3.1 software. Both raw absorbance and MFI was exported to Sigma plot (ver13; Systat software) for estimation of protein concentration using a five-parameter logistic fit. The intra- and inter-assay coefficients of variance for Aβ_1–40 (8.2% & 16%), Aβ_1–42 (6.1% & 16%), P-tau (6.7% & 7%) and T-tau (9.1% & 14%) respectively. Each individual fraction and the sum of all fractions were used for further statistical analysis.