A 3D-printed multi-organ-on-chip platform to model brain-immune interactions in neuroinflammation and neurodegeneration

Introduction::
The brain is in constant communication with cervical lymph nodes (LNs). The antigens picked up by interstitial and cerebrospinal fluid drain to local LNs through the lymphatic vessels in the meninges [1]. In instances of neuroinflammation and neuronal damage, the brain-LN communication is still poorly understood, and it is difficult to dissect in standard mouse models. Multi-organ-on-chips (MOOCs) may provide an experimentally-accessible platform to study communication by connecting multiple tissue models using biomimetic fluid flow [2]. Although blood-brain barrier and brain chips are well established, there are limited MOOCs that include brain models co-cultured with other organs. Furthermore, few include aspects of immunity such as microglia or infiltrating macrophages rather than modeling communication with primary immune organs such as the LN [3]. To address this gap, we are developing an innovative microfluidic platform to model the lymphatic connection between the brain and LN to study neuroinflammation and neurodegeneration. Our strategy builds on connecting Transwell models of the human brain, meninges, and LN paracortex recently established in the Munson lab in a 3D-printed MOOC with tubing-free fluidic control (Figure 1). Transwell models enable a modular design, where 3D cell cultures can be customized for each modeled system. Additionally, we used murine tissue slices to access spatially-structured tissue. To mimic the separation in vivo, we envisioned distinct media compartments for the brain and LN, with communication across the meninges. As a proof-of-principle, the device will be used to model amyloid-ß drainage from the brain into the LN during early Alzheimer’s disease.




Materials and Methods::
The brain-immune chip was designed to consist of two distinct channel loops for media separation, with an intersection at the meningeal barrier culture well and individual pump wells for separate fluidic control. We designed single-media devices for both Transwell models and ex vivo tissue slices for initial device characterization. The devices were drawn using Fusion 360 and printed with a CADWorks3D MiiCraft P110Y DLP printer in MiiCraft Clear resin. We expanded on an impeller pump platform previously established in our lab [4] to recirculate fluid on the brain-immune chip using a new dual-motor design. To support tissue slice culture in the single-media chips, we designed a 3D-printed removable mesh insert for easy slice insertion and removal. Similarly, to interface Transwells with the fluidic channel network and enable flow across or through the membrane, we designed a pocket for an O-ring gasket to generate a reversible seal.



We used COMSOL Multiphysics to (1) optimize culture well and plug geometry, and (2) determine channel velocity required for biologically-relevant interstitial flow for brain, meninges, and LN models. We confirmed fluid circulation on-chip by injecting dye in the single-media Transwell device. Initial tests of tissue viability and function were performed with murine lymph node slices. LN slice viability was quantified using an MTS assay. To model brain antigen drainage and uptake in the LN on-chip, we used the protein DQ-Ovalbumin (DQ-OVA), which becomes fluorescent upon internalization and proteolytic cleavage.


Results, Conclusions, and Discussions::
We utilized 3D computational modeling to determine the conditions needed to achieve biomimetic fluid velocities for each tissue model. For the Transwell models, we developed a plug system to locally modulate the fluid flow for each tissue model without altering the fluid flow in the rest of the device. The plugs reduced the total well volume while directing the flow path closer to the in vitro culture model, thus increasing velocity at the membrane (Figure 1e). With the use of plugs, we achieved our goal velocities across the meningeal barrier and through the Alzheimer’s brain (1-2 µm/s and < 1 µm/s, respectively).




The optimized brain-immune chip design provided easy insertion and removal of modified Transwells with an O-ring gasket, thus enabling asynchronous (off-chip) tissue model generation, as well as removal for imaging throughout the experiment. For user-friendly fluidic control, the device was placed on the compact impeller pump platform, with no tubing required; at least 6 pumps fit per shelf in an incubator (Figure 1g). This system was sufficient to recirculate fluid, assessed by addition of blue dye (Figure 2a). Viability of primarily cells and tissue on 3D printed devices is a significant concern, but fortunately we found that coating the chip with Parylene-C significantly improved viability in experiments with LN slices (Figure 2b,c), consistent with experiments with murine splenocytes [5]. On Parylene-coated chips, LN slices remained viable after 24 hr culture under various flow conditions (Figure 2d,e). Finally, we found that the chip was sufficient to traffic proteins from an upstream tissue compartment to a downstream LN culture, as live LN slices captured and processed DQ-OVA in greater quantities on-chip (Figure 2g) than off chip (Figure 2h), in spatially-resolved regions (Figure 2i).



In conclusion, a user-friendly, two-media MOOC platform and impeller pump was developed to model brain-LN interactions in the context of disease. In ongoing work, we are finalizing the device design for biologically-relevant fluid flow and quantifying velocity experimentally. In the future, we will co-culture brain, meninges, and LN models previously established in the Munson lab on-chip under flow to study brain-LN interactions in systems such as Alzheimer’s.


Acknowledgements (Optional): :

References (Optional): :
[1] Louveau, A.; Herz, J.; Alme, M. N.; Salvador, A. F.; Dong, M. Q.; Viar, K. E.; Herod, S. G.; Knopp, J.; Setliff, J. C.; Lupi, A. L.; DaMesquita, S.; Frost, E. L.; Gaultier, A.; Harris, T. H.; Cao, R.; Hu, S.; Lukens, J. R.; Smirnov, I.; Overall, C. C.; Oliver, G.; Kipnis, J. CNS lymphatic drainage and neuroinflammation are regulated by meningeal lymphatic vasculature. Nature Neuroscience, 2018, 21, 1380-1391.




[2] Picollet-D’hahan, D.; Zuchowska, A.; Lemeunier, I.; Le Gac, S. Multiorgan-on-a-Chip: A Systemic Approach To Model and Decipher Inter-Organ Communication. Trends in Biotechnology, 2021, 39, 788-810.




[3] Hammel, J. H.; Cook, S. R.; Belanger, M. C.; Munson, J. M.; Pompano, R. R. Modeling Immunity In Vitro: Slices, Chips, and Engineered Tissues. Annual review of Biomedical Engineering, 2021, 23, 461-491.




[4] Cook, S. R.; Musgrove, H. B.; Throckmorton, A. L; Pompano, R. R. Microscale impeller pump for recirculating flow in organs-on-chip and microreactors. Lab on a Chip, 2022, 22, 605.




[5] Musgrove, H. B.; Cook, S. R.; Pompano, R. R. Parylene-C coating protects resin 3D printed devices from materials erosion and cytotoxicity. ChemRxiv, 2023.