A microperfused incubator for tissue mimetic 3D cultures

Abstract

High density, three-dimensional (3D) cultures present physical similarities to in vivo tissue and are invaluable tools for pre-clinical therapeutic discoveries and development of tissue engineered constructs. Unfortunately, the use of dense cultures is hindered by intra-culture transport limits allowing just a few layer thick cultures for reproducible studies. In order to overcome diffusion limits in intra-culture nutrient and gas availability, a simple scalable microfluidic perfusion platform was developed and validated. A novel perfusion approach maintained laminar flow of nutrients through the culture to meet metabolic need, while removing depleted medium and catabolites. Velocity distributions and 3D flow patterns were measured using microscopic particle image velocimetry. The effectiveness of forced convection laminar perfusion was confirmed by culturing 700 µm thick neural-astrocytic (1:1) constructs at cell density approaching that of the brain (50,000 cells/mm³). At the optimized flow rate of the nutrient medium, the culture viability reached 90% through the full construct thickness at 2 days of perfusion while unperfused controls exhibited widespread cell death. The membrane aerated perfusion platform was integrated within a miniature, imaging accessible enclosure enabling temperature and gas control of the culture environment. Temperature measurements demonstrated fast feedback response to environmental changes resulting in the maintenance of the physiological temperature within 37 ± 0.2°C. Reproducible culturing of tissue equivalents within dynamically controlled environments will provide higher fidelity to in vivo function in an in vitro accessible format for cell-based assays and regenerative medicine.

Appendix

$$ Q_1 = U \cdot A \cdot \Delta T $$

(1)

$$ U = \left( {{t \mathord{\left/ {\vphantom {t k}} \right. } k} + {1 \mathord{\left/ {\vphantom {1 h}} \right. } h}} \right)^{- 1} $$

(2)

$$ Q_2 = \rho \cdot \dot{V} \cdot \left( {h_o - h_i } \right) $$

(3)

$$ \dot{V}_a = {{\left( {Q_1 + Q_2 } \right)} \mathord{\left/ {\vphantom {{\left( {Q_1 + Q_2 } \right)} {\left[ {\rho_a \cdot c_p \cdot \left( {T_{a,h} - T_{a.,o} } \right)} \right]}}} \right. } {\left[ {\rho_a \cdot c_p \cdot \left( {T_{a,h} - T_{a,o} } \right)} \right]}} $$

(4)

$$ Q_h = \rho_a \cdot \dot{V}_a \cdot c_p \cdot \left( {T_{a,i} - T_{a,h} } \right) $$

(5)

$$ L = {{Q_2 } \mathord{\left/ {\vphantom {{Q_2 } {\left( {\bar{h} \cdot \pi \cdot D \cdot \Delta T_{lm} } \right)}}} \right. } {\left( {\bar{h} \cdot \pi \cdot D \cdot \Delta T_{lm} } \right)}} $$

(6)

$$ \Delta T_{lm} = {{\left( {\Delta T_o - \Delta T_i } \right)} \mathord{\left/ {\vphantom {{\left( {\Delta T_o - \Delta T_i } \right)} {\ln \left( {{{\Delta T_o } \mathord{\left/ {\vphantom {{\Delta T_o } {\Delta T_i }}} \right. } {\Delta T_i }}} \right)}}} \right. } {\ln \left( {{{\Delta T_o } \mathord{\left/ {\vphantom {{\Delta T_o } {\Delta T_i }}} \right. } {\Delta T_i }}} \right)}} $$

(7)

$$ \bar{h} = N_u \cdot {k \mathord{\left/ {\vphantom {k D}} \right. } D} $$

(8)

$$ l = 0.05 \cdot R_{eD} \cdot P_r \cdot D $$

(9)