Human 3D In-Vitro Model of Neurovascular Unit to Study Mechanisms of Traumatic Brain Injury

Introduction:: Introduction: Traumatic Brain injury is the leading cause of long-lasting mental and physical disabilities affecting all ages, races, and socioeconomic classes worldwide 1. TBI is characterized as a primary injury event inflicted by mechanical impact and followed by secondary damage resulting from cellular and molecular responses to direct stimulation. Recently long-lasting pathological changes in mild TBI survivors were discovered, with a clinical picture similar to an affliction called chronic traumatic encephalopathy (CTE) 1. CTE damage includes severe neurodegeneration, demyelination, accumulation of phosphorylated Tau protein, and neuroinflammation 2. Glial and endothelial cells play crucial roles in these pathological changes; however, the mechanisms that drive CTE after TBI are still poorly understood 3. One of the main reasons these questions remain unresolved is the lack of a relevant system to enable the study of CTE affliction. The golden standard model of CTE is the post-mortem brains that only show the late stages of the disease without any indication of its onset. Thus, there is a critical need to develop a model of chronic traumatic encephalopathy to understand the mechanisms that drive neurodegeneration and for therapy discovery.

We aimed to create a physiologically relevant, human in vitro 3D tissue model of neuro-vascular (NVU) to study the pathobiology of damage progression post-TBI.


Materials and Methods:: Methods: Human neurons (N), primary astrocytes (A), and a microglial (M) cell line were seeded in silk protein porous scaffolds (d=6mm) at 2:0.5:0.1 million cells 4, respectively. The following day the scaffolds were infused with collagen gel containing brain microvascular endothelial cells (BMEC) (E) and pericytes (P) at a ratio of 1:0.3 million cells, respectively. The injury was inflicted using a controlled cortical impactor (CCI) (piston d=3mm, 6m/s speed, and 0.6 mm penetration). All assays were performed following the manufacturer’s protocol, and differences between groups with p< 0.05 were considered statistically significant.


Results, Conclusions, and Discussions::
Results and Discussion: We addressed the neurovascular model fabrication sequentially; initially, we created a neuro-glia model to ensure neuronal cells' homeostatic and functional growth. At six weeks post-seeding, the neuro-glia unit reached a mature state: neurons expressed maturation markers (MAP2 and Syn1) and formed dense networks (Tuj1); astrocytes and microglia expressed homeostatic cell-specific markers (ALDH1A1 and TMEM119, respectively). Next, we optimized the conditions for the vascular unit, with final culture conditions being E-A-P (1: 0.5: 0.3 mln respectively) and embedded in collagen gel infused with matrigel. Unlike any other condition, these tricultures showed extended viability (over two months, ongoing) and dense vascular-like network formation, with all cells expressing their homeostatic cell-specific markers (ZO-1, CD31, Aquaporin-4, and α-Smooth Muscle Actin). At 24 hours after CCI, the neuroglial unit showed 80% degradation of Tuj1 networks (n=12, p< 0.001) in mitochondria fission-dependent manner (Fig.1).

We successfully created neuro-glia (over two years) and neuro-vascular units (over two months ongoing). We have discovered that extracellular matrix proteins drastically affected the network formation at early time points (up to 4 weeks), while at later points, the effect began to diminish. Co(di)-culture and tri-culture models combining different cell types in the NVU significantly boosted network formation that remained distinguished beyond two months. Molecular and structural changes associated with mild injury-induced CTE are under study in cultures of combined neuroglial and neurovascular units.




Acknowledgements (Optional): : We thank the University of Cincinnati for its financial support.

References (Optional): :
1. DeKosky ST, et al. PMID: 23558985; 2. Scheid R, et al. PMID: 20386669; 3. Cherry JD, et al. PMID: 27793189; 4. Liaudanskaya V, et al. bioRxiv preprint. 2022.