Advancements in Therapeutic Development: Kidney Organoids and Organs-on-a-Chip

Abstract

The use of animal models in therapeutic development has long been the standard practice. However, ethical concerns and the inherent species differences have prompted a reevaluation of the experimental approach in human disease studies. The urgent need for alternative model systems that better mimic human pathophysiology has led to the emergence of organoids, innovative in vitro models, to simulate human organs in vitro. These organoids have gained widespread acceptance in disease models, and drug development research. In this mini-review, we explore the recent strides made in kidney organoid differentiation and highlight the synergistic potential of incorporating organ-on-a-chip systems. The emergent use of microfluidic devices reveals the importance of fluid flow in the maturation of kidney organoids and helps decipher pathomechanisms in kidney diseases. Recent research has uncovered their potential applications across a wide spectrum of kidney research areas, including hemodynamic forces at stake in kidney health and disease, immune cell infiltration, or drug delivery and toxicity. This convergence of cutting-edge technologies not only holds promise for expediting therapeutic development but also reflects an acknowledgment of the need to embrace innovative and more human-centric research models.

Editor’s Note:

The kidney is a highly vascularized organ that receives 20% of the cardiac output. Every day, the kidneys of a healthy adult generate approximately 180 L of primary filtrate, most of which was reabsorbed while only 2 L of urine was excreted. Most cells within the kidney either directly encounter fluid flow or are indirectly exposed to mechanical stress induced by fluid flow. Human pluripotent stem cell derived kidney organoids manifest limited physiological function, at least partially attributed to the lack of the native tissue microenvironment, such as fluid flow. The synergy between kidney organoids and organ-on-a-chip (OOC) methodology successfully endowed kidney organoids with physiological characteristics that were not realized in the organoids grown in a static culture condition. In this illuminating review ‒ the first article in our regenerative medicine and nephrology series ‒ Tabibzadeh and Morizane summarized various types of mechanical stress that exist in the human kidney, followed by a comprehensive overview of the OOC technology and how this technology could incorporate these physiological features into kidney organoids. The authors discuss several successful examples of using OOC to promote the configuration, maturation, and functionality of kidney organoids, while highlighting existing challenges that await further investigation.

Introduction

Human pluripotent stem cells (hPSCs) possess unique qualities such as unlimited self-renewal and the ability to generate cells from all three germ layers of the embryo. These attributes make hPSCs an excellent choice for deriving functional human kidney cells and tissues. In particular, induced pluripotent stem cells (iPSCs) can be easily obtained from patients with kidney disease, paving the way for the development of immunocompatible tissues and patient-specific models of the disease.¹ Over the past decade, significant strides have been made by leveraging our understanding of kidney development to differentiate hPSCs into cells that belong to the kidney lineage. Through directed differentiation approaches, research groups have successfully created kidney organoids that replicate key aspects of kidney development in vitro.^2–8 These kidney organoids consist of epithelial nephron-like structures that express markers associated with podocytes, proximal tubules, loops of Henle, and distal/connecting tubules. Notably, these structures are arranged in an organized and continuous manner, resembling the segmented features of the nephron. Given their ability to study kidney pathophysiology in 3D tissue and multicellular constructs, kidney organoids have gained widespread adoption among both academic and industry researchers. They serve as valuable tools for investigating a range of conditions, including genetic kidney diseases, tubular and podocyte injuries, renal fibrosis, and kidney development.^{4, 5, 8–21}

Building upon the success of organoid applications in renal disease studies, there is an ambitious aspiration to create more faithful models that can accurately replicate pathophysiological processes within human kidney organoids. Given the fact that it is currently impossible to culture the entire organs of animal and human kidneys ex vivo, significant efforts are required to develop ideal models that encompass all physiological aspects, including renal filtration function, using kidney organoids. Nonetheless, researchers have been making substantial progress in enhancing organoid disease models by integrating organ-on-chip technologies. These endeavors have led to the development of kidney organoid-on-chip models, which hold great promise for advancing our understanding of renal physiology and disease.²² In this review, we summarize the recent advancements in organoid disease models employing organ-on-chip systems and discuss their potential, limitations, and future directions to expedite research in kidney disease for therapeutic development.

Organoids meet Organ-on-a-Chip (OOC) Technologies

The kidney is a highly vascularized organ that receives a significant portion of the cardiac output for urine production. The resulting dynamic blood and urinary flows create a fluidic microenvironment around nephron epithelia and vessels, subjecting them to mechanical stress. Blood filtration in glomeruli imposes pulsatile mechanical stress on podocytes, known for their sensitivity to mechanical stimuli.²³ As the glomerular filtrate passes through the tubular lumen, it generates circumferential stretching and shear stress on tubules, activating mechanosensitive signaling pathways through cellular stretching and shear stress. Approximately 99% of the glomerular filtrate is reabsorbed through the tubules that traverse the interstitial space. This results in mechanical stress, secondary interstitial fluid flow, and a hydraulic pressure gradient across the tubular epithelium, with a higher pressure on the basal side.^24–26 Notably, both the mechanosensitive ion channel Piezo1 and the stretch-activated channels are expressed on basolateral membranes in renal tubules, suggesting potential roles for basal mechanical stress in tubular function and pathophysiology.^{27, 28} Additionally, the medulla undergoes rhythmic stretching due to the contraction of smooth muscle cells along the renal pelvis.²⁹ The kidney is encapsulated by a tough fibrous layer known as the renal capsule, which is not highly elastic. Consequently, arterial pulsatile flow and smooth muscle contraction are presumed to cause compression within the confined kidney tissue, resulting in stretching mechanical stress at the cellular level (Figure 1). Such mechanical factors of shear stress, cyclic stretching, and compression are not typically simulated in kidney organoids cultured under static conditions. Recreating such a microenvironment may enhance the fidelity of these models for the study of renal physiology and diseases.

Figure 1. Mechanical forces in the kidney.

The figure illustrates the three biomechanical forces: shear stress, stretching, and compression. The kidney is surrounded by an incompressible capsule, indicating that variations in flow may impact the compression and stretching of the tissue. Compression and stretching also occur during contractions and relaxations of smooth muscle cells found in different regions and segments (such as juxtaglomerular specialized smooth muscle cells and pericytes), including the renal pelvis, leading to rhythmic mechanical stress. Shear stress affects not only endothelial cells but also tubular cells and interstitial cells, as evidenced by the presence of in vivo interstitial flow.

Organ-on-a-chip (OOC) technologies have emerged as a promising platform for simulating cellular microenvironments with greater accuracy. OOC models utilize microfluidic devices that enable precise control over fluid flow, creating a controlled and dynamic environment.^{30, 31} These microphysiological systems are designed to replicate specific aspects of organ structures and offer several advantages over traditional cell culture methods. One of the key advantages of OOC models is the use of human cells, which enhances their clinical relevance and translational potential. By employing human cells, OOC models better replicate human physiology and may allow for more accurate predictions of human responses.

The principle of OOC relies on a triad of components: a chip, a flow system, and cells. Microfluidic chips are typically composed of synthetic materials like polymers or polydimethylsiloxane (PDMS), allowing for optimal culture conditions and transparency. Some chip systems also incorporate stretchable materials to induce mechanical stretching of cells, mimicking physiological conditions.³² Additionally, 3D bioprinting can be used to engineer biological scaffolds that replicate specific organ structures, such as tubular structures.³³ The second component of the triad is the perfusion system, which relies on motorized equipment to deliver culture media and potentially blood cells through the chip at controlled output and pressure. This controlled inflow and outflow of the culture medium provides an experimental environment for modulating nutrient delivery and mechanical forces.

The third component of the triad is the cells. Typically, immortalized cell lines and primary cell culture are used, and various cellular interactions can be studied by seeding different cell types in each perfusion channel. Instead of cells, organoids can be cultured in such chip devices, providing more cellular and tissue complexity for disease modeling and drug screening (Figure 2).

Figure 2. Accelerating disease studies through integration of organoids and organ-on-chip technologies.

Key features of organoids, microfluidic chips, and perfusion systems are listed near each scheme.

The combination of organoids with OOC systems is expected to provide new insights into pathophysiological studies. The mechanical stress controlled by the OOC system enables investigation of its impact on organoid development, cellular differentiation, and its roles in renal pathophysiology. The modulation of the local environment by the architecture and extracellular matrices may also facilitate organoid maturation and contribute to disease studies. Another notable feature of OOC models is the ability to monitor cellular responses and metabolism in real-time during mechanical stimuli in organoids.³⁴ This real-time monitoring capability provides valuable insights into dynamic changes at the cellular level, which can be challenging to capture in animal models. By capturing these dynamic cellular responses, organoid-on-a-chip models may offer a more comprehensive understanding of mechanobiology and cellular behavior (Figure 3).

Figure 3. Anticipated impacts of organoid-on-chip technologies.

(A) By integrating organoids with organ-on-chip systems, the effects of the local environment and mechanical stresses can be explored to enhance the physiological and pathophysiological relevance of organoid models. (B) Left: A brightfield image of organoids cultured on a perfusion chip. A scale bar: 500 μm. Right: Live imaging of an organoid tubule stained with live-staining dyes for cellular membranes (green) and nuclei (red). A scale bar: 15 μm. (C) Illustrations of the kidney organoid-on-chip showing anticipated impacts and potential applications.

Effects of fluid flow on kidney organoid vascularization and maturation

Several groups have embarked on an exciting new avenue by combining organoid and OOC technologies.^35–38 Homan and Gupta et al. cultured kidney organoids in perfusion chips, which subjected the whole organoids to flow-induced mechanical stress. Although it is not an exact replication of the in vivo microenvironment, this approach has yielded encouraging results. Notably, under fluidic culture, the study observed a significant enhancement in vessel formation originating from the intrinsic endothelial progenitors.³⁶ Furthermore, there was a substantial increase in vascular invasion into glomerular structures. This enhancement in vascularization was accompanied by a significant elevation in vascular endothelial growth factor-A (VEGFA) expression within the perfused kidney organoids, and inhibiting VEGF using bevacizumab led to a decrease in vascularized glomeruli. Intriguingly, when VEGF was added under static conditions, it did promote vascular structures within kidney organoids, but failed to enhance vascular invasion into glomeruli. These findings underscore the crucial role of endogenous VEGFA production in the formation of vascularized glomerular structures necessary for proper kidney function and blood filtration. Additionally, the tubules within perfused kidney organoids exhibited enhanced cellular polarity and increased transporter expression, indicative of epithelial maturation. The visualization of fluid flow using fluorescent beads revealed a pulsatile flow profile and the presence of interstitial and vascular luminal flow in this perfusion chip setting, suggesting the involvement of shear stress, compression, and cellular stretching in vascular development and nephron maturation.

Vascular nephropathies are among the most common causes of AKI and CKD, and they may benefit from vascularized organoid models as well.^{39, 40} Hemodynamic changes such as glomerular hyperfiltration and hyperpressure play a crucial role in diabetic nephropathy, especially during its early stages.⁴¹ Despite its prevalence, the pathophysiology of diabetic nephropathy remains poorly understood. Therefore, replicating specific circulatory phenomena observed during the course of diabetic nephropathy in organoids-on-a-chip could significantly enhance our understanding of its multifaceted pathophysiology. Additionally, the versatility of organoid-on-a-chip design holds promise for the development of innovative approaches to studying diseases such as ischemia-reperfusion injury and hypertensive nephropathy.

The influence of fluid flow on drug studies has also garnered significant attention within the realm of organoid-on-a-chip models. Research groups have demonstrated that kidney organoids, when cultured under dynamic flow conditions, exhibit heightened sensitivity to drug toxicity compared to their counterparts cultured in static environments.³⁸ Although whole organoids were not cultured on the chip, Aceves et al. conducted a study showcasing increased cisplatin- and aristolochic acid-induced toxicity in proximal tubular epithelial cells derived from kidney organoids, when compared to immortalized proximal tubular cell lines within cylindrical chip models.⁴² Although the flow-enhanced drug delivery may partly explain this heightened sensitivity, the observed increase in drug transporter expression induced by flow conditions is likely the main underlying cause of this heightened sensitivity. This upregulation of transporter expression may provide a more accurate reflection of in vivo nephrotoxicity compared to conventional toxicity assays conducted on cell lines. Additionally, this model holds promise for advancing emerging drug delivery systems, including nanoparticles and genome editing-based therapies, which are increasingly gaining prominence within the pharmaceutical landscape.⁴³

Incorporation of immune cells in kidney organoid-on-chip

Kidney organoid-on-a-chip models have also been successfully employed in co-culture conditions, revealing interesting interactions between the kidney organoids and other cell types. Kroll et al. showed the infiltration of peripheral blood mononuclear cells (PBMCs) into kidney organoid-on-a-chip, circulating inside the lumen of the kidney organoid vasculature to assess T cell-mediated cytotoxicity using T cell bispecific antibodies.⁴⁴ However, whether immune cells entered the vascular lumen via perfusion or migration was unclear, primarily due to the extended timelapse between perfusion and the distribution of PBMCs within the organoids. In a kidney organoid-on-a-chip model with HUVEC co-culture incorporated into the chip channels, Bas-Cristóbal Menéndez et al. observed the migration of HUVECs into the organoids under flow conditions, forming large open lumen structures connected to the organoids proper vasculature.³⁵ These two encouraging observations suggest that, even though further research is warranted, achieving a more accurate replication of blood flow, drug delivery, and immune cell infiltration in organoid-on-a-chip models may become feasible in the near future. This perspective is of major importance in replicating the spectrum of kidney immune-mediated diseases, from kidney transplant rejection, to complement-mediated glomerulonephritis, among others.⁴⁵

Elucidation of mechanosensing pathomechanisms for therapeutic development

Various mechanosensing signals have been identified, facilitating biological processes through intracellular signaling pathways. These signals involve cellular components such as stretch-activated channels, focal adhesions, integrins, cytoskeletal rearrangement, primary cilia, ion channels, and G protein-coupled receptors (GPCRs).^46–48 Their crucial roles in kidney development are reported by previous studies using animal models: e.g. Bock et al. revealed the crucial role of Rac1 in maintaining the integrity of the kidney collecting duct by regulating actomyosin activity.⁴⁹

Mechanosensing signals are also believed to be involved in disease onset and progression, as many genes associated with genetic kidney disease are expressed in primary cilia, mechanosensing organelles. For example, fibrocystin, encoded by the PKHD1 gene responsible for autosomal recessive polycystic kidney disease (ARPKD), localizes to primary cilia, basal bodies, and cell membranes in kidney and liver epithelia.⁵⁰ Considering fibrocystin’s localization, it is reasonable to speculate that mechanotransduction plays a role in ARPKD pathogenesis. However, the lack of tools for investigating mechanosensing has hindered progress in understanding these mechanisms in kidney diseases. Recently, proof-of-concept studies using organoid-on-chip models have been conducted in ARPKD and autosomal dominant polycystic kidney disease (ADPKD). Hiratsuka and Miyoshi, utilizing CRISPR/Cas9 genome editing, created PKHD1 knockout kidney organoids. Their work showcased the development of cysts in distal nephrons within a fluidic culture system, aligning with patient phenotypes.¹⁴ This outcome stands in contrast to previous organoid studies, which showed proximal tubule dilatation in static culture with forskolin treatment.⁸ Through transcriptomic approaches and targeted drug screening, they identified mechanosensing pathomechanisms involving FOS and RAC1, a small GTPase involved in mechanotransduction and in collecting duct cell polarity and adhesion.¹⁴ Additionally, the use of a fluorescent lifetime imaging microscope (FLIM) with a fluorescent lipid tension reporter provided direct evidence of cellular stretching in organoid tubules exposed to flow. Li et al. utilized PKD1 or PKD2 knockout organoids exposed to fluid shear stress and found that cyst expansion occurs through an absorptive pathway driven by glucose transport.²¹ These recent studies provide compelling evidence for the unique utility of organoid-on-a-chip models in studying mechanosensing pathomechanisms in ciliopathies.

Promises and Limitations of Organoid-on-a-Chip Technologies

Organoids and their integration with OOC technologies hold great promise for gaining new insights into physiology and disease mechanisms. However, these technologies are still undergoing advancements, and further improvements are necessary. One major challenge in organoid-on-a-chip models lies in accurately replicating the complex microenvironment provided by the native cardiac circulation system. While the current models can induce mechanical stress, the flow pathways in native kidneys are predominantly driven by blood flow. Addressing this challenge, ongoing efforts are focused on developing chip models that incorporate vascular channels connected to organoid vessels.³⁵ The establishment of vascular inflow and outflow in the chip system would enable the simulation of physiological shear stress driven by human-level blood pressure. This improvement is particularly important for developing more faithful models of glomerular diseases that affect filtration function.

Another challenge arises from the lack of precise data regarding physiological mechanical stress at the cellular level within native human kidneys. To measure shear stress accurately, we need to calculate it by dividing the applied force by the relevant area. Yet, measuring these parameters directly in human kidneys in vivo is currently unfeasible. Consequently, hypothetical calculations and in vitro experiments are often employed to estimate “physiological” shear stress levels, which typically range from approximately 0.1 to 0.3 dyn/cm² for proximal tubular lumens.^51,52 It is worth noting that actual shear stress values in tubular lumens of native kidneys may be reduced by tubular mucus and fluctuate depending on the glomerular filtration rate. Consequently, an ongoing debate surrounds the definition of “physiological” models in organoid-on-a-chip systems. The studies mentioned earlier notably applied shear stress ranging from 0.00001 to 0.035 dyn/cm² to the entire organoids,^36,38 seemingly lower than that experienced by proximal tubular lumen in native kidneys. Nevertheless, this applied shear stress might be relevant to induce interstitial and tubular cellular stretching and compression within the perfusion chip. As there is currently no optimal tool to distinguish the specific effect of these various mechanical forces in vivo, these estimated values can serve as a foundational point for designing experiments aimed at investigating mechanosensing biological processes in organoid-on-chip models. While we anticipate further advancements in organoid and organoid-on-chip model systems, the current iterations of these models offer valuable opportunities to explore biomechanical cues in renal pathophysiology that were previously unexplored without the use of organoid models.

Concluding Remarks

Kidney organoids and organs-on-a-chip have emerged as valuable tools for enhancing our understanding of renal physiology and diseases. These advanced models offer significant improvements in evaluating the effectiveness, safety, and toxicity of experimental compounds. Notably, kidney organoids hold immense potential for reducing or eliminating the need for animal studies, aligning with ethical considerations and promoting more humane research practices. A significant breakthrough has occurred with the recognition by the American Congress and the US Food and Drug Administration of the value of ex vivo data obtained from organoids and OOC approaches, opening the door for considering such data in drug approval processes for clinical trials.⁵³ This recognition signifies a transformative shift in drug development and highlights the growing importance of these innovative models. Furthermore, kidney organoids enable precision-medicine approaches, particularly in the realm of inherited diseases. By utilizing patient-specific induced pluripotent stem cells (iPSCs) that possess naturally occurring variants or mutations, researchers can create disease models that mimic the mutation-specific phenotypes. The significance of these advancements is further emphasized by the International Society of Nephrology, which recently provided its first consensus guidance for preclinical animal studies. Notably, their guidance encourages researchers to consider the feasibility of conducting experiments using organoids or OOC, reflecting the increasing recognition of the potential of these innovative models in advancing scientific knowledge and reducing reliance on animal testing.⁵⁴ We hope that these new technologies receive support from the research communities and foster collaborative efforts to accelerate the development of novel therapies for kidney patients.