Development of an advanced multi-cellular and dynamic flow model of the human alveolar airway

This project aimed to construct a 3D dynamic model of the human air-blood barrier. The model was an assembly of the alveolar region of the human lung, and the associated vasculature on the basolateral side, all integrated into a bioreactor under fluid flow. The alveolar model aimed to include epithelial cells, such as alveolar type I (NCI-H441 or hAELVi) and type II (A549), together with macrophages (PMA differentiated THP-1 (dTHP-1)). The biomaterial selected to mimic the alveolar region extracellular matrix was derived from the decellularization of human lung fibroblast cells, cultured into a sheet. To create the vasculature model, endothelial cells (HULEC 5a) were co-cultured with human lung fibroblasts (HLF-1) within a hydrogel. Two biomaterials were tested to support the growth of the vasculature model; (i) a gelatin and alginate mesh or (ii) a fibrin gel. Regarding the bioreactor, besides ensuring the sterility of the cellular model, the goal was to enable the culture of the alveolar in vitro model at the air-liquid interface (ALI) and the vascular model under fluid-flow dynamics. To achieve what was proposed above, the project was divided into three parts; the first focused on the alveolar model, where the development of the decellularized membrane, as well as the cell culture model are described in chapters 3 and 4. The decellularization procedure included several approaches to remove the DNA content. These consisted of freeze-thaw cycles and exposure to surfactants such as sodium dodecyl sulfate (SDS) and Triton. The decellularized extracellular matrix (dECM), was tested as a scaffold or cross-linked with riboflavin or fibrin to support the growth of NCI-H441, hAELVi, A549 and dTHP-1 cells. It was observed that only monocultures of A549 and dTHP-1 cells were able to grow and proliferate under both submerged and ALI conditions upon the dECM crosslink with fibrin (dECM/fibrin). The second part of the project addresses the vasculature model (chapter 5). The gelatin and alginate mesh was first investigated to support the endothelial cells (ECs) growth, but with a lack of success. Moving forward, the fibrin gel was investigated, where a range of concentrations of fibrinogen and thrombin were tested. After some final gel volume adjustment, the fibrin was able to support the ECs growth and proliferation. The fibrin gel was characterized regarding its rheology and porosity which, together with cell proliferation data, 2.5 mg/mL fibrinogen and 5 U/mL thrombin mix was selected to build the vasculature co-culture model within the bio-matrix. Once established, supplemented media was designed to induce angiogenesis, which was able to promote, in static conditions, cell organization into 3D geometries resembling tip cells and cord formation. Under dynamic conditions, the vasculature shows ECs with tip cell morphology, however, cord formation was lacking, as was fibroblast proliferation. Finally, the third part is comprised of the bioreactor design, fabrication and simulation. Multiple iterations were designed, created and tested (chapter 6). Initially, the first and second version of the bioreactor was constructed in acrylic or polycarbonate. Yet, this did not allow for the alveolar model positioning. Therefore, subsequent versions were manufactured in laser cut to ensure complete transparency of acrylic and included a silicone part where the cellular model was placed. Based on these findings, a complete model was created by integrating the alveolar and vasculature models into the bioreactor (chapter 7). Under fluid flow, the full model showed a complete A549 cell monolayer populated with dTHP-1 on the apical layer. Though, unfortunately, when combined with all parts together, the vasculature model showed a lack of cell viability and proliferation. In conclusion, the proposed model demonstrated the possibility to combine a diverse cell population with a physiologically relevant mechanical environment. The integrated model requires further optimizations, such as in the dECM biomaterial and a different cell source in the vasculature. However, a foundation to create a dynamic air-blood barrier system has been shown in order to gain further insight towards building anatomically and physiologically relevant non-animal models for next generation risk assessment