(I-354) Development of a Blood-Brain Barrier on a Chip Model for Drug Delivery Studies

Introduction::

The blood-brain barrier (BBB), which separates the blood stream and brain parenchyma, is extremely important for the healthy functioning of the brain. BBB disruption is implicated in many neurological diseases. Due to the biological complexity and tightness of this barrier, which restricts most molecular drugs from reaching the brain, neurological diseases are difficult to research and treat. Rodent models are the gold standard for research in this area, but phenotypic and functional differences between the rodent and human BBB limit the utility of some findings toward clinical translation. Researchers have previously developed in vitro systems that allow for the incorporation of human cells, extracellular matrix, and basement membrane proteins to create a simple but physiologically accurate BBB model. The organ-on-a-chip design is increasingly popular for its fluidic flow, as well as the ability to mimic human physiology with distinct organ chips in series. Here, we implemented a microfluidics-based BBB model with human brain microvascular endothelial cells and human primary astrocytes. We report on model fabrication and optimization. Several key BBB features, including tight junction protein expression, macromolecular transport, cell viability, and cell morphology were measured to characterize our model system and establish its utility as a tool in brain drug delivery studies.

Materials and Methods::

The blood-brain barrier model was established utilizing commercial SynBBB 3D chips (SynVivo 102005-SB3) containing two apical channels for endothelial cells and an inner chamber for astrocytes. Each device was coated with 200 µg/mL fibronectin in the apical chambers for hCMEC/D3 cells and a 1:20 dilution of matrigel in the central chamber for primary human astrocytes. The chips were maintained at 37°C and 5% CO₂ and daily media changes were conducted for two days before conditioning the endothelial cells to flow, which implemented a linear ramp profile of 10 nL/min – 2 µL/min over a period of 12 hours followed by 6 hours at 2 µL/min. Assays were conducted immediately following this process. Cytoskeletal staining was conducted utilizing ActinRed 555 and ActinGreen 488 and Hoechst 33258 nuclear staining following fixation by 4% PFA and permeabilization with 0.2% Triton X-100. Cell viability was assessed utilizing the LIVE/DEAD™ Cell Vitality Assay Kit, C12 Resazurin/SYTOX™ Green (Thermofisher L34951). To analyze the efficiency of the endothelial cell, a permeability assay was conducted where 70 kDa FITC dextran was injected at a concentration of 250 µg/mL into the apical chamber and fluorescent images were acquired at 4X at 30 second intervals for two hours utilizing a EVOS M7000 microscope equipped with an enclosure to maintain 37°C and 5% CO₂. ImageJ was used to assess the average intensity in each chamber for each frame and any frames collected before the apical chamber reached an equilibrium intensity were omitted.

Results, Conclusions, and Discussions::

Daily monitoring of cell health and growth through morphology and confluency revealed the astrocytes proliferated rapidly while the endothelial cells were more sensitive to varying environmental conditions, such as high flow rate ( >15 µl/min) and exposure to ambient temperature. Regular media changes (i.e., every 4 hours) were implemented and flow rate was restricted to 5 µL/min and lower in order to maintain endothelial cell morphology and viability (95.35% viability for 473 counted cells by LIVE/DEAD). Phase microscopy and cytoskeletal staining allowed for clear visualization of cells within the chip and there were high concentrations of cells observed at the membrane interface for both cell types. The astrocytes appeared to form a thick layer against the membrane and not only interacted with each other through complex endfeet networks seen within the central chamber, but also with the endothelial cells in the apical chamber given the observation that there were astrocyte endfeet visible within the pores of the membrane in close contact with the neighboring cells. The permeability assay data showed that there was a barrier formed by these cells capable of restricting some paracellular transport, where the permeability was calculated to be 5.75 x 10^-8 cm/s. Taken together, our BBB model has several promising features, astrocyte-endothelial cell interaction across a porous membrane, high cell viability, and biomimetic macromolecular transport properties. Further studies could leverage the BBB-on-a-chip properties to investigate BBB functional changes following chemical exposure, BBB function in disease, or brain drug delivery.

Acknowledgements (Optional): :

 We thank the other lab members and those in Dr. Michael Detamore’s lab for their support and help whenever necessary. 

Affiliations of Authors:

Trinity Hoff¹, Mia Gonzales¹, John R. Clegg^1,2,3,4

[1] Stephenson School of Biomedical Engineering

[2] Harold Hamm Diabetes Center

[3] Stephenson Cancer Center

[4] Institute for Biomedical Engineering, Science, and Technology

University of Oklahoma, Norman, OK 73019, USA.

References (Optional): : I. Wilhelm and I. A. Krizbai, "In Vitro Models of the Blood–Brain Barrier for the Study of Drug Delivery to the Brain," Molecular Pharmaceutics, vol. 11, no. 7, pp. 1949-1963, 2014, doi: 10.1021/mp500046f.