Peptide-Functionalized Electrospun Meshes for the Physiological Cultivation of Pulmonary Alveolar Capillary Barrier Models in a 3D-Printed Micro-Bioreactor

In vitro environments that realize biomimetic scaffolds, cellular composition, physiological shear, and strain are integral to developing tissue models of organ-specific functions. In this study, an in vitro pulmonary alveolar capillary barrier model is developed that closely mimics physiological functions by combining a synthetic biofunctionalized nanofibrous membrane system with a novel three-dimensional (3D)-printed bioreactor. The fiber meshes are fabricated from a mixture of polycaprolactone (PCL), 6-armed star-shaped isocyanate-terminated poly(ethylene glycol) (sPEG-NCO), and Arg-Gly-Asp (RGD) peptides by a one-step electrospinning process that offers full control over the fiber surface chemistry. The tunable meshes are mounted within the bioreactor where they support the co-cultivation of pulmonary epithelial (NCI-H441) and endothelial (HPMEC) cell monolayers at air–liquid interface under controlled stimulation by fluid shear stress and cyclic distention. This stimulation, which closely mimics blood circulation and breathing motion, is observed to impact alveolar endothelial cytoskeleton arrangement and improve epithelial tight junction formation as well as surfactant protein B production compared to static models. The results highlight the potential of PCL-sPEG-NCO:RGD nanofibrous scaffolds in combination with a 3D-printed bioreactor system as a platform to reconstruct and enhance in vitro models to bear a close resemblance to in vivo tissues.

equilibrate to 37°C within the incubator, resulting in hydrostatic pressure changes at the cultivation area. Since the hydrostatic pressure is required as the base value for the pressure oscillation to maintain a stable air-liquid interface, an equilibration period is necessary in case the pressures and temperatures are not monitored and controlled. We recommend an equilibration period of 1 day to ensure achieving equilibrium conditions. However, a more advanced temperature control system could render the equilibration time obsolete.
Trouble Shooting. Most of the challenges regarding the proposed bioreactor setup are related to the generation of the correct pressure distribution, which is heavily dependent on the air tightness of the dampening system and a correct heating protocol.
The air tightness of the dampening system can be tested before the experiment by heating the system to 37°C. If all clamps are closed, the increase in temperature leads to a visible rise of the medium level within the tubing system due to the expansion of the gas phase and the corresponding increase in gas pressure. In case both bottles are sufficiently air-tight, the medium level should reach equal distances measured from the dampening bottle caps. If the distances are significantly different, one of the dampening bottles leaks air and hence cannot build up the necessary gas phase pressure. In this case, the caps of the dampening bottles should be tightened again, or the PTFE sealing tape needs to be reinforced or exchanged.
Another possible cause can be loose connections or leaky tubing.

S2
After checking the air tightness of the system, it is important to ensure equal temperatures in all bottles before venting the system. If the bioreactor system is placed inside the incubator and the bottle temperatures significantly deviate from each other, the gas pressure changes occurring during the equilibration period will also differ leading to a shift in hydrostatic pressure at the cultivation area. This change in hydrostatic pressure can lead to changes in the liquid level and hence can cause flooding or drying up of the model. In case experimentation fails due to flooding or drying up, ensure that the heating procedure is performed for a proper duration and that all bottles exhibit close to identical temperatures. Other reasons could be weak pump cassette springs or the wear of pump tubing.     A) The water droplet spreads and displays a reduced angle of 26 • ± 1.2 • on meshes spun with sPEG-NCO which imparts hydrophilic properties to the PCL membranes. B) As control, pure PCL nanofibrous meshes display less spread and a higher contact angle of 110 • ± 14 • which is consistent with the inherent hydrophobic property of the PCL meshes. Fig. S11: Investigation of epithelial to mesenchymal transition by the study of α-smooth muscle actin (αSMA)in NCI-H441 at cultivation day 10 for different experimental procedures including static, exposure to shear stress, and a combination of shear stress and cyclic strain. In static, 5.4 ± 1.5% of NCI-H441 cells displayed αSMA, under shear stress, 5.6 ± 1.1% of the NCI-H441 cells displayed αSMA, and under a combination of shear stress and cyclic strain 6.9 ± 1.4% displayed αSMA. n = 5.  To acquire even PDMS surfaces necessary for tight channel sealing the excess solution was scraped off the mould before being cured on a leveled surface. In early development, stages various height differences were tested. While smaller height differences between the groove and PDMS channel led to leakage due to insufficient force, larger height differences resulted in bioreactor failure due to the deformation of the housing component during operation. Considering that even most high-temperature 3D-printing materials tend to deform over time under continuous tension at elevated temperatures, the housing component was fabricated with an extensive height of 10.5 mm, while the base plate was manufactured from 2 mm thick aluminum.

S8
Fig. S14: Overview of the seeding process on PCL-sPEG-NCO:RGD fiber meshes, and transfer of cell layer-containing fiber meshes to the bioreactor. A) Aluminium foil holding a fiber mesh directly after electrospinning. B) Wetting of the fiber mesh in PBS (1X). C) Detachment of the fiber mesh using tweezers. D) Transfer of the fiber mesh to the 3D printed transwell insert. E) Transwell insert holding the fiber mesh. F) UV sterilization of the fiber mesh. G) -H) HPMEC seeding. A cell-containing medium droplet is pipetted onto the lid of a microtiter plate. The transwell insert holding the fiber mesh is inverted and placed into the corresponding microtiter plate. The microtiter plate is closed using the lid resulting in contact between the cell-containing droplet and the fiber mesh. The microtiter plate is subsequently incubated for 2h at 37°C enabling cell sedimentation and attachment. Finally, the transwell insert is inverted again and 1000 µL of medium are pipetted in the basal, and 200 µL in the apical compartment. I) After 1 day of HPMEC cultivation at 37°C, NCI-H441 cells are seeded into the apical compartment of the transwell insert. The NCI-H441/HPMEC model is cultured for 4 days at 37°C in RPMI1640/EGM medium to produce confluent cell layers. J) The confluent fiber mesh is cut loose using a scalpel. K) The fiber mesh is detached from the transwell insert in sterile PBS (1X). L) The membrane is transferred to the membrane mount of the cultivation chamber using tweezers. M) Membrane mount holding an electrospun fiber mesh. N) The membrane mount is plugged into the bioreactor. O) Fully assembled upper part of the bioreactor holding an electrospun membrane.