Method Profile

BBB Chip Drug-Exposure Monitoring Coupled to Norepinephrine Secretion

Human cell-line barrier–neural cultures pair pramipexole concentration measurements with evoked norepinephrine after chemical injury.

Brain & Neural ModelsOrgan Chips
Green neuronal processes and blue nuclei in a human neural stem-cell-derived culture.
Fluorescence micrograph of neurons differentiated from human neural stem cells. (Yirui Sun, CC BY 4.0)

Research Applications

Relate trans-barrier drug arrival to a functional secretory response and compare microfluidic with static transfer.

Model & Design

hCMEC/D3 human endothelial cells and undifferentiated SH-SY5Y human neuroblastoma cells on opposite collagen-I/fibronectin-coated PET faces; no astrocytes/pericytes.

Experimental Format

120 μL/min dynamic culture; 24-hour 3 mM 1-methyl-4-phenylpyridinium (MPP+) injury then 24-hour pramipexole exposure; static Transwell comparator.

Model Components

  • Component

    Cells and exposure

    Description

    7.5×10^5 endothelial and 1×10^5 SH-SY5Y cells. Vascular pramipexole 750 μM; selected cell-response benchmark 200 μM. Injury-free, injured/no-drug and nutrient-replenishment controls.

  • Component

    Measurement system

    Description

    Electrochemical concentration sensor with regeneration and LC–MS drug comparison; separate LC–MS evoked norepinephrine assay. The drug sensor does not measure norepinephrine online.

Measurements & Endpoints

  • Neural-side pramipexole concentration and permeability
  • Evoked norepinephrine secretion
  • CCK-8 metabolic activity and apoptosis
  • ZO-1 and 4/40-kDa dextran permeability
  • TNFα, IL-1β and IL-6

Analysis

Norepinephrine secretion and CCK-8 metabolic activity are separate response measures. Reported ± values are mean ± standard deviation (SD). Total clock times of 37 and 32 hours include the preceding 24-hour injury period, corresponding to approximately 13 and 8 hours after drug dosing.

Published Evidence

  • Study Design

    Dynamic versus static barrier transfer, with orthogonal drug concentration detection and chemical-injury response comparisons.

    Study Sample

    At least three independent experiments in source statistics; barrier Figure 5 n = 6–8 by condition; no donor cohort.

    Comparison

    Chip versus Transwell, injured/no-drug versus pramipexole, normal versus injured cells and nutrient controls.

    Finding

    Pramipexole transfer was slower in chip than Transwell (Papp 7.28×10^-7 versus 1.22×10^-6 cm/s). Norepinephrine improved after dosing; apoptosis 19.95 ± 1.06% fell to 13.95 ± 1.63%. TNFα declined but IL-1β/IL-6 were unchanged; 6 mM MPP+ injury resisted rescue. 40-kDa permeability did not increase significantly with the selected injury.

    Interpretation & Limits
    • Undifferentiated neuroblastoma cells are not normal noradrenergic or dopaminergic neurons; evoked product was predominantly norepinephrine.
    • High experimental pramipexole concentrations and acute MPP+ injury are not clinical exposure predictions.
    • Drug sensing is in situ but functional secretion requires offline LC–MS sampling.
    • Sensor fouling reduced the fifth uncleaned signal to 47.6%; regeneration was necessary for repeat measurements.

Sources

  1. Micro pharmacokinetics-pharmacodynamics monitoring of anti-Parkinson’s disease drugs using a microphysiological BBB-brain organ-on-a-chip Microsystems & Nanoengineering. 2026. doi:10.1038/s41378-025-01116-w ↗