Method Profile
Microelectrode-array stimulation and offline decoding quantify separability of evoked cortical-organoid responses under neural inhibitors and disease-related perturbations.
Add a network-level functional endpoint to neural pharmacology and disease-model comparisons.
Human stem-cell-derived cortical organoids, including a familial-Alzheimer’s APP-duplication induced-pluripotent-stem-cell (iPSC) line and a healthy-donor iPSC comparator.
Three reported spatial stimulation patterns, repeated microelectrode-array (MEA) recording and logistic-regression decoding; spontaneous activity and connectivity measured alongside evoked responses.
Neural tissues
Human embryonic-stem-cell and iPSC resources; two-to-four-month developmental comparisons. APP and healthy-donor organoids compared at three months.
Pharmacological perturbations
Twenty-minute exposures: NMDA and AMPA/kainate receptor antagonists AP5/CNQX at 25 μM each, bicuculline at 20 μM and tetrodotoxin (TTX) at 1 μM in the reported comparisons.
Recording and decoder
MEA surface contact, three spatial patterns with thirty trials each, top twenty active channels, ten-millisecond bins, initial ten-millisecond artifact exclusion and a 2:1 recording-level train/test split.
Interpret accuracy as the ability of the offline classifier to separate recorded evoked patterns. Link inhibitor effects to firing and connectivity without equating trial-level test data with independent donor/organoid prediction.
Measure response classification before and after neural inhibitors.
Figure 3: five organoids across three independent experiments.
Baseline versus twenty-minute AP5/CNQX, bicuculline or TTX exposure.
Accuracy changed from 89% to 55.5%, 88.5% to 78% and 85.5% to 32.5%, respectively. Bicuculline increased firing despite reduced accuracy; TTX eliminated activity/connectivity.
Compare APP-duplication and healthy-donor cortical organoids.
Functional n = 15 organoids and amyloid n = 6, each from three experiments.
Three-month patient-derived APP versus healthy-donor organoids.
APP organoids had more amyloid and reduced accuracy/dissimilarity despite similar spontaneous firing and global efficiency.
Brainopheno combines electrical stimulation, microelectrode recordings and an offline decoder to measure response separability in human cortical organoids. Neural inhibitors reduced classification accuracy, while disease and monocyte comparisons revealed changes missed by some spontaneous-activity measures.
