UCLA

Growth of engineered tissue constructs is dependent on spatiotemporally regulated signals. The optical opacity and dynamic physical properties of developing tissue present a challenge for controlling flow-induced shear distribution in thick, perfused constructs. Tools capable of applying controlled mechanical stimuli throughout engineered tissue constructs and simultaneously obtaining readouts of construct growth have not been developed. The features of magnetic resonance imaging (MRI) that make it clinically suitable; primarily its noninvasiveness, large penetration depth, number of available contrast weightings, and use of non-ionizing radiation; make it worth investigating as a tool for monitoring thick and increasingly complex tissue cultures.

This work presents an MRI compatible, multi-inlet perfusion bioreactor capable of delivering arbitrary flow and, by extension, flow-induced shear patterns throughout 3D tissue constructs by varying flowrates between twelve inlets. Multiple scaffolds were evaluated for mechanical compatibility with the perfusion bioreactor and biocompatibility with endothelial and parenchymal cell lines. Cell population distribution was compared in identical scaffolds cultured under static and patterned perfusion conditions. Diffusion, $T_2$, and magnetization transfer (MT) MRI weightings were investigated as a means to generate quantitative maps of cell density and viability.

It was found that flow induced shear maps could be calculated in multiple environments from a combination of MRI velocimetry maps, culture chamber geometry, and substrate properties. Several biopolymer hydrogels and macroporous sponges were shown to be mechanically compatible with long term perfusion while promoting sufficient endothelial and parenchymal cell growth. Flow-induced shear patterns within a tissue engineering construct were shown to influence cell distribution. Viable cell density was quantifiable within physiological ranges using diffusion-, $T_2$-, and MT-weighted MRI. Viability was independently quantified from cell density using a combination of MT- and diffusion-weighted MRI with a multivariate surface calibration.

This work demonstrates the components necessary to achieve the long-term goal of closed loop, flow and shear controlled tissue development. The tools described here can be immediately applied toward determining the relationship between cell population distribution and shear pattern in the centimeter scale, which is a critical piece of information necessary to create a tissue growth control algorithm.