Transport of Extracellular Vesicles across the Blood-Brain Barrier: Brain Pharmacokinetics and Effects of Inflammation

Abstract

Extracellular vesicles can cross the blood-brain barrier (BBB), but little is known about passage. Here, we used multiple-time regression analysis to examine the ability of 10 exosome populations derived from mouse, human, cancerous, and non-cancerous cell lines to cross the BBB. All crossed the BBB, but rates varied over 10-fold. Lipopolysaccharide (LPS), an activator of the innate immune system, enhanced uptake independently of BBB disruption for six exosomes and decreased uptake for one. Wheatgerm agglutinin (WGA) modulated transport of five exosome populations, suggesting passage by adsorptive transcytosis. Mannose 6-phosphate inhibited uptake of J774A.1, demonstrating that its BBB transporter is the mannose 6-phosphate receptor. Uptake rates, patterns, and effects of LPS or WGA were not predicted by exosome source (mouse vs. human) or cancer status of the cell lines. The cell surface proteins CD46, AVβ6, AVβ3, and ICAM-1 were variably expressed but not predictive of transport rate nor responses to LPS or WGA. A brain-to-blood efflux mechanism variably affected CNS retention and explains how CNS-derived exosomes enter blood. In summary, all exosomes tested here readily crossed the BBB, but at varying rates and by a variety of vesicular-mediated mechanisms involving specific transporters, adsorptive transcytosis, and a brain-to-blood efflux system.

Keywords: adsorptive transcytosis; blood–brain barrier; diapedesis; exosome; extracellular vesicles; neuroinflammation; pharmacokinetics.

Figure 1

Exosome characterization: (A) Exosomes were isolated from supernatants of Kasumi-1, a human leukemia cell line using mini-SEC as described in Materials and Methods. Exosomes in Fraction #4 were harvested and characterized by tunable resistive pulse sensing (TRPS) to determine their diameters and concentrations (left); by Western blots to show the presence of exosome markers and absence of cytoplasmic proteins (middle); and by TEM to illustrate vesicular morphology and the size range (right). (B) The TEM images, size distribution profiles and Western blot profiles of Mel526 exosomes (left) and exosomes produced by human T cells (middle). (a) The size distribution profile of circulating exosomes in CD-1 mouse plasma; (b) comparisons of total exosomal protein levels of circulating exosomes in the plasma of CD-1 mice and of normal human donors. The data are means ± SD from 3 independent measurements.

Figure 2

On-bead flow cytometry of exosome populations isolated from supernatants of 4 different cell lines, captured on streptavidin beads coated with biotinylated anti-CD63 mAbs and stained with the labeled detection Abs specific for CD46, αVβ6, αVβ3 and ICAM1 as described in Methods. Data are relative fluorescence intensity (RFI) values calculated as mean fluorescence intensity (MFI) of an experimental sample/MFI of an isotype control Ab. [White peaks = experimental samples; hatched peaks = isotope controls]. Note that exosomes produced by non-malignant cells (HaCaT and primary T cells) appear to differ from exosomes produced by malignant cells by the absence of αVβ3 and lower expression levels of αVβ6. Representative results are from 1 to 3 detection experiments performed with each exosome population.

Figure 3

Clearance of exosomes from blood demonstrated two major patterns: phase decay, as exemplified by SCCVII (mouse tumor) exosomes and no change with time as exemplified by MEL526 (human melanoma) exosomes.

Figure 4

Examples of patterns for BBB passage and brain accumulation. Brain/serum ratios have been corrected for vascular space by subtracting albumin brain/serum ratios. Left upper panel illustrates continuous linear uptake of SCC-90 exosomes. Right upper and left lower panels illustrate a plateau pattern as seen with primary T cell-derived and SCCVII-derived exosomes. Only the HaCaT exosomes (right lower panel) did not show an increase in the brain/serum ratio with time. The closed circles represent data points describing the linear portion of the curve; the open circles represent data points that departed from linearity. See Table 3 for statistics.

Figure 5

Variation in the uptake of different exosome populations by brain regions. Four exosome populations showed a significantly greater uptake by the olfactory bulb (OB) than by whole brain (WBr), cortex (Cx), or cerebellum (Cb). * p < 0.05, ** p < 0.01, **** p < 0.0001.

Figure 6

Effects of LPS on exosome uptake by various brain regions. Upper left: uptake of primary T cell-derived exosomes by whole brain (WBr) and cortex (Cx), but not by OB (olfactory bulb) or cerebellum (Cb), was increased by LPS. Upper right: for MDA-MB-231-derived exosomes; LPS increased uptake by the OB and the WBr. Lower left: for PCI-30-derived exosomes, LPS increased uptake into all the assessed regions. Lower right: for SCCVII-derived exosomes, LPS decreased uptake by the OB and the CX. * p < 0.05, *** p < 0.001.

Figure 7

Effects of WGA and M6P on exosome uptake by whole brain. Shown are data for 5/10 tested exosome populations in which WGA significantly enhanced uptake. M6P inhibited uptake of 1/10 exosome populations tested (see J774A.1). Neither WGA nor M6P had an effect on primary T cell-derived exosomes, ** p < 0.01, **** p < 0.001.

Figure 8

Brain-to-blood efflux of HaCaT-derived exosomes after ICV administration. Upper panel: open circles show that exosome clearance from brain had a half-life of 7.85 min; open circles show that exosomes reach an equilibrium after about 10 min. Lower panel: Appearance of HaCaT exosomes in blood was stable at a mean of 3.9%.