Organ-on-Chip Systems Integrating Human Adipose Tissues

Abstract

Adipose tissue constitutes about one fourth of a healthy adult human’s body mass and is involved in a large variety of (patho-)physiological processes. Especially in the era of ‘diabesity’, a thorough understanding of human adipose tissue has become more important than ever. Yet, research on human adipose biology is hampered by a lack of predictive model systems. Even though many valuable insights could be gained from animal models, they often fall short of predicting human physiology. Then again, unusual characteristics of mature adipocytes, such as buoyancy, fragility, and large size, make conventional cell culture approaches challenging. In recent years, organ-on-chip (OoC) technology has emerged from a synergy of tissue engineering and microfluidics approaches. OoC systems integrate engineered tissues into physiological microenvironments supplied by a vasculature-like perfusion. Yet even though OoC technology is thriving regarding other organ systems, there has only been very little focus on adipose tissue so far. Hence, the objectives of this thesis were to design, develop and characterize adipose tissue-on-chip models. To achieve this, designs, biomaterials and fabrication approaches were developed leading to three generations of microfluidic platforms specifically tailored to the needs of human adipose tissues. Moreover, protocols and logistics for sourcing, isolating, and utilizing almost all adipose tissue cell types from one donor were established. Together, this enabled the generation of white and beige adipose tissues (WAT and bAT, respectively) on chip either by injecting mature adipose cell types (in case of WAT) or by inducing adipogenesis on chip (in case of bAT). Along the way towards a highly complex, immunocompetent autologous model integrating almost all adipose-associated cell types, a mix-and-match strategy was established allowing for a flexible combination of cellular modules to fit-for-purpose models serving a specific scientific question. Moreover, a toolbox of readout methods was compiled that enabled a comprehensive characterization of on-chip adipose tissue structure and function, demonstrating functional on-chip WAT culture times beyond one month. Case studies on compound screening and immune responses highlighted the models’ suitability as tools for target identification in drug discovery or for studies on immunometabolism. All in all, the developed models hold great potential for mechanistic studies on adipose tissue biology or disease modelling in the context of obesity and diabetes, as well as for personalized or precision medicine due to its fully autologous character.