Progress in understanding kidney disease mechanisms and the development of targeted therapeutics have been limited by the lack of functional in vitro models that can closely recapitulate human physiological responses. Organ Chip (or organ-on-a-chip) microfluidic devices provide unique opportunities to overcome some of these challenges given their ability to model the structure and function of tissues and organs in vitro. Previously established organ chip models typically consist of heterogenous cell populations sourced from multiple donors, limiting their applications in patient-specific disease modeling and personalized medicine. In this study, we engineered a personalized glomerulus chip system reconstituted from human induced pluripotent stem (iPS) cell-derived vascular endothelial cells (ECs) and podocytes from a single patient. Our stem cell-derived kidney glomerulus chip successfully mimics the structure and some essential functions of the glomerular filtration barrier. We further modeled glomerular injury in our tissue chips by administering a clinically relevant dose of the chemotherapy drug Adriamycin. The drug disrupts the structural integrity of the endothelium and the podocyte tissue layers, leading to significant albuminuria as observed in patients with glomerulopathies. We anticipate that the personalized glomerulus chip model established in this report could help advance future studies of kidney disease mechanisms and the discovery of personalized therapies. Given the remarkable ability of human iPS cells to differentiate into almost any cell type, this work also provides a blueprint for the establishment of more personalized organ chip and ‘body-on-a-chip’ models in the future.

No abstract

When fluids flow through straight channels sustained turbulence occurs at high Reynolds numbers (typically $ \textrm{Re}\sim O(1000)$). It is difficult to mix multiple fluids flowing through a straight channel in low Reynolds number laminar regime ($ \textrm{Re}<O(100)$) because in absence of turbulence, mixing between the component fluids occurs primarily via slow molecular diffusion process. This letter reports a simple way to significantly enhance low Reynolds number (in our case $\textrm{Re}\leq10$) passive microfluidic flow mixing in a straight microchannel by introducing asymmetric wetting boundary conditions on the floor of the channel. We show experimentally and numerically that by creating carefully chosen 2D hydrophobic slip patterns on the floor of the channels we can introduce stretching, folding and/or recirculation in the flowing fluid volume, the essential elements to achieve mixing in absence of turbulence. We also show that there are two distinctive pathways to produce homogeneous mixing in microchannels induced by the inhomogeneity of the boundary conditions. It can be achieved either by: 1) introducing stretching, folding and twisting of fluid volumes, i.e., via a horse-shoe type transformation map, or 2) by creating chaotic advection, through manipulation of the hydrophobic boundary patterns on the floor of the channels. We have also shown that by superposing stretching and folding with chaotic advection, mixing can be optimized by significantly reducing mixing length, thereby opening up new design opportunities for simple yet efficient passive microfluidic reactors.