*3.2. Kidney-on-Chip Models*

In this view, the kidney-on-chip approach tries to mimic the dynamic, flowing environment more closely and consists of a 2D/3D cellular model developed in a microfluidic device. This whole system finally consists of a 3D architecture built up by renal cells grown on an ECM interface or next to perfusable microchannels where the media and/or other body fluids such as blood or urine can flow across the cell surface [140] (Figure 2). Using these models, studies have shown that fluid shear stress (FSS) has a major influence on renal tubular cells' phenotype [141], hence their importance in nephrotoxicity and other renalfunction-related studies. FSS can induce inflammation through immune-cell-mediated activation and monocyte adhesion [142]; it increases the expression of certain genes such as ABCG2, RBP4 (a marker for tubular function loss) [143], CYP1A1, and SLC47A1, affects the cytoskeleton organization, and upregulates the formation of tight junctions [144]. Consequently, kidney-on-chip models are designed for high-throughput screening of drug toxicity by delivering essential output data regarding the glomerular filtration processes, the drugs pharmacokinetics with impact on drugs validation, and relevant dose determination. In this view, several types of kidney-on-chip models have been developed targeting different segments of the renal unit.

#### 3.2.1. Glomerulus-on-Chip

The glomerulus is the filtering unit of the kidney, consisting of a capillary network and podocytes, highly differentiated epithelial cells that are responsible for the actual filtration process of the blood. Considering this, there is an obvious interest in developing a functional system recapitulating the glomerular function to support preclinical stages of new drug development or to sustain research in the kidney disease field, including (but not limited to) kidney neoplasms. Despite the efforts made in the past few years to recreate a glomerulus-on-chip system, the main challenge in developing an in vitro glomerulus model is the lack of functional human podocytes. Various studies have been performed to create a glomerulus-on-chip model by obtaining human-induced pluripotent stem-cell (iPSCs)-derived podocytes and placing them in a microfluidic device together with endothelial cells (primary or secondary iPSC-derived). These microfluidic devices were designed to create the two compartments separated by a porous membrane. Musah et al. showed that podocytes extended their processes through the membrane when exposed to constant flow and mechanical strain, and selective filtration was proven by the presence of excreted inulin and retention of albumin on each side of the membrane [145]. Roye et al. also tried to build a glomerulus-on-chip with hPSC-derived podocyte and endothelial cells, and succeeded in obtaining essential functionality and structure [146]. Both models were carried out to study the effect of adryamicin as a chemotherapeutic agent, and showed that the treatment induced podocyte detachment and endothelial-barrier disruption, leading to albuminuria. This glomerulus-on-chip model could be a promising start for an in vitro study of proteinuria and glomerular kidney disease and chemotherapeutic drug-nephrotoxicity assessment. Moreover, in 2017 Wang et al. developed a glomerulus-on-chip model using rat cells to create a diabetic nephropathy model that also served to assess the hyperglycemic pathological response [147].

**Figure 2.** Complex kidney-on-a-chip for personalized medicine: (**A**) kidney and nephron, (**B**) glomerulus-on-a-chip, and (**C**) tubule-on-a-chip.
