Introduction

Biopharmaceutical research and development faces a major productivity crisis in the depreciating efforts to develop novel drugs (Woodcock & Woosley, 2008). Despite over 30 years of investment in biomedical sciences and the scientific tools used in drug discovery, few results have been well-translated in the preclinical and clinical stages (Giffin, Robinson, & Olson, 2009; Wehling, 2006). The reliance on conventional cell culture systems and animal models during preclinical testing hinders the establishment of human-related kidney predictive models (Awdishu & Mehta, 2017; Wilmer et al., 2016). Microphysiological systems (MPS) accurately model human systems in a compact, efficient fluidic tool that can introduce controlled spatiotemporal micro-environments (Kim & Wu, 2012), support accurate human responses, allow for real-time imaging, encourage cell differentiation, and pinpoint cell-cell interactions under a variety of physiological conditions (Phillips et al., 2020; C. Sakolish et al., 2018; C. M. Sakolish, Esch, Hickman, Shuler, & Mahler, 2016; C. M. Sakolish, Philip, & Mahler, 2019; Wu et al., 2020). Recent developments in stem cell research (Musah et al., 2017; Sciancalepore et al., 2014), regenerative medicine (Gomes, Rodrigues, Domingues, & Reis, 2017), biomaterials (Homan et al., 2016; van Midwoud, Janse, Merema, Groothuis, & Verpoorte, 2012), tissue engineering (Jang et al., 2013; Jansen et al., 2014; Ng, Zhuang, Lin, & Teo, 2012; Tourovskaia, Fauver, Kramer, Simonson, & Neumann, 2014), and microfluidics allow for integration into three-dimensional (3D) MPS.
In the kidney, millions of nephrons employ filtration, reabsorption, secretion, and excretion processes. Each functional unit collaborates to filter out wastes and xenobiotics, separate water, ions, and small molecules from the blood, and recycle compounds to the blood (Figure 1A-B ) (Marieb & Hoehn, 2007). Within the nephron, glomerular filtration consists of passive movement of plasma from glomerulus capillaries to the Bowman’s capsule that is freely permeable to water and small solutes (Na+, urea, and glucose), but not permeable to blood, white blood cells, platelets, or large molecular weight (>67 kDa) serum proteins (albumin) (Koeppen & Stanton, 2012). The glomerular filtrate exits the glomerulus to enter a selective barrier of highly coiled tubules. There, the proximal convoluted tubule (PCT) utilizes active and passive transport to reabsorb glucose, sodium chloride, and water from the glomerular filtrate (Marieb & Hoehn, 2007; Zanetti, 2020). During reabsorption, highly-concentrated filtrate becomes the leading site for nephrotoxin accumulation, a precursor for acute kidney injury or chronic kidney disease. Nearly 90% of renal toxicity cases are derived from both the glomerulus and PCT of the kidney (Bonventre, Vaidya, Schmouder, Feig, & Dieterle, 2010; Jang et al., 2013; C. M. Sakolish et al., 2016). Current MPS designs of the kidney have been well established for the PCT segment of the nephron, representing only a section of the renal absorption process (Jang et al., 2013; C. Sakolish et al., 2020; C. Sakolish et al., 2018; C. M. Sakolish et al., 2016; C. M. Sakolish et al., 2019; Weber et al., 2016). However, incorporating both filtration and absorption interfaces will refine the physiological relevance of the kidney in vitro barrier model, and permit rapid screening for drug toxicity in preclinical studies.
In this study a novel, physiologically realistic MPS of the proximal tubule and glomerulus that incorporates cross flow filtration and is capable of long-term operation has been developed. The tri-culture system houses conditionally immortalized human podocytes (CIHP-1) to represent the ultrafiltration processes from the fenestrations in the glomerular capsule, human umbilical vein endothelial cells (HUVECs) for the capillaries recirculating solutes in the bloodstream, and human kidney-2 (HK-2) to recreate the reabsorption processes in the PCT. Key design requirements included integration of the glomerular filtration fraction (0.15-0.2) and tubular reabsorption (0.65-0.7) (Feher, 2017; Marieb & Hoehn, 2007). The cells within the MPS were grown under dynamic flow engineered to mimic flow conditions in vivo (0.4-1.2 dyne-s/cm2)(Wilmer et al., 2016) for seven days, and then the system was challenged with fluorescein isothiocyanate- human serum albumin (FITC-HSA) to assess its filtration functional capacity. Cells within the device were imaged using confocal microscopy for attachment and cytoskeletal reorganization. MPS culture medium and imaging was completed without the introduction of animal by-products. This MPS introduces a novel definition of a PCT and glomerulus with physiological capability of blood serum protein filtration, glucose resorption, and filtrate formation capacity.