Microfluidics-based tools for culture and multi-functional assessments of three-dimensional pluripotent stem cell derived tissues

Abstract

Due to their capacity for self-renew and differentiation, pluripotent stem cells (PSCs) are used as a cell source in applications including tissue engineering, regenerative medicine, and studies of human development and disease. A common format for PSC differentiation is as three-dimensional (3D) cell aggregates. Differentiation of PSCs as aggregates mimics biochemical and biophysical aspects of in vivo embryonic development that promote differentiation and complex cell behaviors. Aggregates are used as a format both in directed differentiation approaches and in generating complex human tissue models such as organoids. Key challenges in generating and studying PSC aggregates include the low throughput and time-consuming nature of culture techniques, the lack of precise control over the culture microenvironment, and the inability to perform a variety of in situ functional assays. This thesis focuses on addressing some of these challenges through the development of microfluidics-based technologies for culture and assessment of stem cell aggregates. In Chapter 2 of this thesis, I developed a microfluidic platform for longitudinal monitoring and multi-modal phenotypic analysis of individual stem cell aggregates. This platform uses a hydrodynamic loading principle to capture pre-formed aggregates in independent traps, which enables physical isolation and tracking of aggregates during culture. I demonstrated that culture of aggregates in individual traps reduces heterogeneity in parameters such as size and gross morphology. Additionally, culture and analysis steps such as immunostaining and imaging could be performed sequentially in the same platform, enabling correlation of multiple modes of analysis for individual samples. In Chapter 3 of this thesis, I applied the microfluidic platform developed in Chapter 2 in exploring how microfluidic culture parameters influence stem cell behavior and differentiation in the context of motor neuron differentiation. Transport modeling and experimental studies were used to assess how media exchange frequency and asymmetrical device geometries modulated the biochemical environment and differentiation. These studies demonstrated that discontinuous media perfusion was effective at supporting PSC aggregate growth, but there was a balance between sufficient media exchange frequency and allowing accumulation of required cell-secreted. Finally, work in this chapter showed how asymmetrical microscale geometries can be used to generate gradients of cell-secreted factors and induce asymmetric differentiation events. In Chapter 4 of this thesis, I developed a culture platform for forebrain organoids. This aim built upon technologies presented in the first two aims and extended these for much larger organoid tissue models. The platform developed in this aim allows culture of individual organoids in an array of chambers with controlled delivery of media and reagents, the ability to perform imaging-based assays, and the ability to retrieve organoids for end-point based assays. I developed and validated two iterations of the platform design. With each design, computational fluid dynamics and qualitative transport modeling were used to validate design choices and optimize culture conditions. I validated culture of forebrain organoids in the platforms for up to 42 days through characterization of organoid size, morphology, cell types, and structural features. Results from these analyses revealed that organoids cultured in the microfluidic platform developed comparably to conventional methods. Together, the work performed in this thesis establishes a set of technologies for higher throughput, robust formation of stem cell derived tissues of multiple size scales, coupled with highly informative assays. These tools provide new capabilities that can be broadly applied by other researchers in studies of 3D stem cell differentiation and in generation of stem cell derived tissue models.