Prioritizing Functional Goals as We Rebuild the Kidney

Remarkable progress in the generation of kidney organoids from human pluripotent stem cells has been made over the last 5 years, allowing investigators to model human kidney in vitro. All kidney organoid protocols induce pluripotent stem cells into metanephric mesenchyme, the cellular source of kidney epithelial cells with the exception of collecting duct. The metanephric mesenchyme then self-organizes, differentiating into recognizable and interconnected tubule segments, including glomeruli-like structures (reviewed in ref. 1). Typical kidney organoids contain about 100 nephron-like structures. A large number of protocol modifications and applications of organoids have already been made. They have been adapted into suspension culture, used for isolation of specific kidney cell types, implanted into microfluidic flow devices, and optimized for high content screening.¹ The increasing availability of reagents, such as well characterized pluripotent stem cell lines expressing fluorescent reporters under control of kidney cell–specific marker genes, should spur further progress.²

It is widely recognized that organoid cell types are substantially immature and in general, reflect the human second trimester stage of development.^3,4 There are signs of improvement, however, and the most promising approach to promote organoid maturation has been transplantation in vivo. In this setting, host vasculature invades the organoid and is specifically attracted to glomeruli. Sharmin et al.⁵ could show the development of slit diaphragm–like structures between podocytes in human organoids transplanted under the kidney capsule. van den Berg et al.⁶ demonstrated mouse blood flow through transplanted human organoid glomeruli by live imaging, and Bantounas et al.⁷ found intravenously injected fluorescein-labeled dextran in organoid tubule lumens, suggesting that glomerular filtration might be occurring. Subramanian et al.⁸ have also documented a reduction in nonrenal “off-target” cells after organoid transplantation.

With this progress, investigators are beginning to set their sights on function. An article in this issue of JASN by Yamanaka et al.⁹ reports a step in the right direction. The authors have shown that the vascular-epithelial interface in implanted, vascularized kidney organoids demonstrates the property of selective permeation. Injected small molecular weight dextrans appear in Bowman’s space and tubule lumens more rapidly than large molecular weight dextrans. The authors did not work out the mechanism of dextran permeation (i.e., whether it was due to simple diffusion or solvent drag due to ultrafiltration), nor did they identify the specific path that the dextran molecules take.⁹ However, as in the glomerulus, the dextrans must traverse cells, basement membranes, and epithelial cells, and it is reasonable to view the demonstrated permselectivity as being dependent on the special characteristics of the cells that make up these three barriers as in the native glomerulus. The question that we ask here is, from a functional perspective, what are reasonable goals in the development of organoids for renal repair or replacement?

If the long-term goal is to replace renal function, then it seems necessary to ask, “What does the kidney do?” Although it is commonplace to use the term “renal function” as a surrogate for glomerular filtration, renal function is really a lot more. Classically, as elucidated by Smith¹⁰ and his contemporaries in the early part of the 20th century, renal transport function consists of three components: glomerular ultrafiltration, tubular reabsorption, and tubular secretion. Replacement of these functions is challenging, not only because it is difficult to grow the right epithelia with the right transporters or because it is difficult to grow glomeruli with properly oriented substructures, but also (and critically), because of a requirement for connectivity of the epithelial lumens to the outside of the body. Steady-state ultrafiltration, reabsorption, and secretion require a conduit for steady-state excretion of urine. Thus, looking beyond the work by Yamanaka et al.⁹ and focusing on nephron structures in renal organoids, we will have to find a way to make organoids connect, presumably to existing drainage pathways or drainage pathways that we can make using tissue engineering approaches.

However, a broader view of renal function offers more room for immediate optimism. Beyond the Homer Smith view of transport function of the kidney, the kidney does a lot more. For example, there are endocrine functions (e.g., renin production, vitamin D hydroxylation, erythropoietin secretion, and klotho production), metabolic functions (e.g., arginine production and fructose clearance), and detoxification roles. These do not require a drainage pathway. These additional functions are likely to be more readily reconstituted with organoid implants than transport functions, and they may, therefore, deserve priority.

Disclosures

Dr. Humphreys is a member of the National Institutes of Health–funded ReBuilding a Kidney consortium (DK107374) and reports personal fees from Merck, personal fees from Janssen, personal fees from Medimmune, personal fees from Roche, personal fees from Celgene, personal fees from Chinook Therapeutics, personal fees from Genentech, grants from Chinook Therapeutics, grants from Janssen, and other from Chinook Therapeutics outside the submitted work. Dr. Knepper has nothing to disclose.

Funding

Dr. Knepper is funded by the Division of Intramural Research, National Heart, Lung, and Blood Institute (projects ZIA-HL001285 and ZIA-HL006129, MAK).