Technical Papers and Presentations

Introduction: Discovering therapeutic drugs that target renal disease is a challenge due to the adverse effects not detected until late in drug development [1]. Preclinical drug studies in animals and static human cell cultures fail to characterize in vivo renal interactions. Organ on a chip technology can overcome these challenges by generating a microenvironment with physiologically realistic fluid flows, shear stresses, and cell interactions [2]. Within the human nephron, a kidney functional unit, the glomerulus and proximal convoluted tubule (PCT) operate together to filter, osmoregulate and reabsorb compounds from blood [3]. In this study, COMSOL Multiphysics^® simulation software was used to optimize flow patterns in a nephron-on-chip to create a glomerular filtration and proximal tubule environment. The device houses three cell types (human endothelial, proximal tubule, and podocytes) in a microfluidic environment for 7 days, and can filter human serum albumin (HSA) in a physiologically accurate manner.

Methods: The nephron-on-chip system is shown in Figure 1A. The glomerulus holding immortalized human podocytes (CIHP-1) and human umbilical vein endothelial cells (HUVECs) is connected to the PCT polycarbonate microfluidic chip with human-kidney proximal tubule cells. A 2D model of the nephron-on-chip was developed with COMSOL^® to optimize the model design and operation based on tubing internal diameter (ID), tubing length, and flow rate from the pump. Design parameters consisted of achieving a passive fluid flow split at the glomerular filter T-junction and obtaining a fluid shear stress of 0.4-1.5 dyne/cm² [4] across the PCT top chamber. Domains were generated using geometry tools and discretized using tetrahedral and boundary layer elements with extremely fine meshing. The Free and Porous Media Flow interface defined the fluid and matrix properties of porous membranes in the system. Simulations were validated experimentally in the nephron-on-chip system using a peristaltic pump and 18 MΩ water for 24 hours. FITC-HSA filtration by the glomerulus was simulated through the system and compared with experimental results.

Results/Discussion: Velocity and concentration simulations (Figure 1B) optimized the experimental design and operation of the microfluidic device for 2 hours. Experimental validation tests determined that with a total pump flowrate of 40.5 µL/min, there was 16.2 µL/min (± 5.5 µL/min) exiting the waste stream and a shear stress over the proximal tubule cells within the physiological range of 0.4 dynes/cm²- 1.5 dynes/cm². Filtration of HSA in the system over 12 hours was simulated to be over 90%. The computational model demonstrates an optimized nephron-on-chip system that has been experimentally validated.

Figure 1. (A) Schematic of Nephron-on-Chip System. Red = T-Junction; Purple = Glomerulus; Orange = PCT. (B1-B2) T-Junction (Red) and Glomerulus (Purple) Velocity Simulations from time 0-120 min. (B3-B4) T-Junction (Red) and Glomerulus (Purple) FITC-HSA Concentration Simulations from time 0-120 min. (B5-B6) PCT (Orange) Velocity Simulations from time 0-120 min. (B7-B8) PCT (Orange) FITC-HSA Concentration Simulations from time 0-120 min.

Acknowledgements: Funded by Alternatives Research & Development Foundation

References: [1] Redfern, W. S. et al, 2010. [2] Jang, K.-J. et al, 2013. [3] Kriz, W. & Lehir, M, 2005. [4] Raghavan, V., et al, 2014.