Bio one 655209

50 × 10³/cm² hCMEC/D3 cells were seeded on a transwell filter (Greinier, 665,641) precoated with 150 μg/ml rat tail collagen type‐I and grown for 5 days to confluency. Culture medium was replaced every other day, as described previously (De Jong et al., 2018). On the fifth day, the basolateral medium was replaced with 1 ml of pre‐warmed EBM‐2, and 500 µl EBM‐2 containing 20 µg/ml DiI‐labeled exosomes was added to the apical compartment. After incubation for 18 hours at 37°C, the apical and basolateral media were collected. The filters with cells were cut out and soaked in 1 ml water for 5 minutes. Apical, basolateral, and cellular fractions were transferred into black flat‐bottomed microplates (Greiner Bio‐One 655209) in triplicate and fluorescence intensities were quantified using a Fluostar‐Optima microplate reader (BMG Labtech) with 485 nm excitation wavelength and 520 nm emission wavelength, respectively. After subtracting the respective background fluorescence (serum‐free medium for apical and basolateral and water for cellular fractions), the percentage fluorescence associated with the apical, cellular, and basolateral fractions was calculated relative to the total fluorescent content of the apical, basolateral, and cellular fractions together.

Immobilization of the His-tagged P protein from T. maritima at 0.05 mg/ml (3.4 μM) was performed on Ni-NTA biosensors. For the initial binding tests to the P protein, a compound concentration of 1 mM was used. Periods for each step of the assay were as follows: immobilization (400 s), wash (100 s), baseline establishment (100s), association (250 s), dissociation (250 s). For each titration curve, six points were used corresponding to the 125, 62.5, 31.25, 15.6 and 7.8 μM compound concentrations. Control biosensors were used to subtract non-specific binding of compounds to the biosensor surface. All assay steps (immobilization, wash, baseline, association, and dissociation) were performed in 20 mM HEPES pH 7.4, 100 mM ammonium acetate, 10 mM MgCl₂, 5% DMSO. The assays were performed on black bottom 96-well microplates (Greiner Bio-One 655209) in a total volume of 200 μl, at 30°C with orbital shaking at 1000 rpm. Experiments were controlled with the software Data Acquisition 8.2 (ForteBio, Inc.). Kinetic binding parameters were calculated using Data Analysis 8.2 (ForteBio, Inc.). After subtraction of reference biosensors, the binding curves were aligned to the X- and Y-axis and the association-dissociation inter-step curve in order to get a common baseline for the association and dissociation phases.