Pluripotent Stem Cells to Rebuild a Kidney: The Lymph Node as a Possible Developmental Niche

Abstract

Kidney disease poses a global challenge. Stem cell therapy may offer an alternative therapeutic approach to kidney transplantation, which is often hampered by the limited supply of donor organs. While specific surface antigen markers have yet to be identified for the analysis and purification of kidney stem/progenitor cells for research or clinical use, the reprogramming of somatic cells to pluripotent cells and their differentiation into the various kidney lineages might represent a valuable strategy to create a renewable cell source for regenerative purposes. In this review, we first provide an overview of kidney development and explore current knowledge about the role of extra- and intrarenal cells in kidney repair and organogenesis. We then discuss recent advances in the 1) differentiation of rodent and human embryonic stem cells (ESCs) into renal lineages; 2) generation of induced pluripotent stem cells (iPSCs) from renal or nonrenal (kidney patient-derived) adult cells; 3) differentiation of iPSCs into renal lineages; and 4) direct transcriptional reprogramming of adult renal cells into kidney progenitor cells. Finally, we describe the lymph node as a potential three-dimensional (3D) in vivo environment for kidney organogenesis from pluripotent stem cells.

Introduction

The kidney is a multifunctional organ that performs several essential physiological functions, including excretion of metabolic wastes, regulation of water and ion content in the blood (osmoregulation), and secretion of hormones such as erythropoietin (EPO), calcitriol (vitamin D₃), and renin.

There are nearly 400,000 end-stage renal disease (ESRD) patients in the US and approximately 2,000,000 worldwide, increasing at a rate of 4–5% annually (16). All patients with ESRD need dialysis or transplantation for survival.

The severe shortage of donor organs has driven a search for alternatives to organ transplantation. Stem cell-based therapies are being investigated as an attractive alternative approach to restore structural and functional integrity of injured tissues. Stem cell identity and function have been fairly well defined for organs with a high cell turnover, such as bone marrow (17), intestine (100), and skin (9), and stem cell-based treatment of diseases affecting these organs shows promise. However, in organs such as the liver and kidney, where cell turnover is low and stem cell identity and function are poorly defined, cell-based therapies remain a challenge. Indeed, although renal cells with stem cell characteristics have been identified in rodents and humans, their roles in tissue repair and potential therapeutic applications remain elusive. Moreover, to date, there have been no reports of a postnatal kidney stem/progenitor cell capable of giving rise to all the cell types of the mammalian nephron, thus hampering the development of specific cell therapies for renal failure. Rather, very recent studies have revealed that, at least during the post-mesenchymal–epithelial transition developmental stage, the adult mammalian kidney undergoes continuous tubulogenesis via expansion of fate-restricted clones that may function as unipotent stem/progenitor cells (81). In other words, mature nephrons are polyclonal, and several types of stem/progenitor cells are likely responsible for kidney organogenesis (81).

The reprogramming to pluripotency and the directed differentiation of human cells may offer a renewable source of replacement cells for several diseases, including different types of kidney disease. The kidney is a highly complex organ consisting of more than 14 different cell types organized into a very specific architecture. Therefore, a detailed understanding of the spatial and temporal distribution of signals that drive nephrogenesis, as well as determination of stem/progenitor cell identity during kidney development, are of utmost importance for the successful differentiation of pluripotent stem cells (PSCs) into the different renal lineages. Here we review recent advances in this rapidly evolving field. In light of our recent observation that the lymph node can serve as an in vivo factory to generate complex structures, including the kidney (26–28,37,52), we will also discuss how this three-dimensional (3D) in vivo environment might be exploited for kidney organogenesis from PSCs.

A Brief Overview of Kidney Development

The first morphologic sign of gastrulation is the appearance of the primitive streak, an elongated region that forms through coordinated proliferation and movement of epiblast cells. The presence of the primitive streak establishes bilateral symmetry and initiates germ layer formation. The meso-endoderm (ME), which later separates to form mesoderm and endoderm, is the first structure to appear. The mesoderm of a neurula stage embryo can be divided into five regions: the chordamesoderm, the paraxial mesoderm, the intermediate mesoderm (IM), the lateral plate mesoderm, and the head mesenchyme. In all vertebrates, the kidneys and ureters will develop from the IM. The kidney develops in three progressive phases: the pronephros, the mesonephros, and the metanephros. The pronephros and mesonephros eventually degenerate in females, whereas a portion of the mesonephros in males contributes to a drainage system for the future testis, including the epididymis (93). The definitive kidney develops from the embryonic metanephros, which is formed by reciprocally inductive interactions between the metanephric mesenchyme (MM) and the ureteric bud (UB). MM secretes signals essential for the formation of the UB from the Wolffian duct. Among these, glial cell-derived neurotrophic factor (GDNF) is apparently the most important inducer (18). In turn, UB-derived wingless-type MMTV integration site family, member 9b (Wnt9b) induces a fraction of MM cells, called the cap mesenchyme (CM), to condense and generate pretubular aggregates (PTAs) (13). PTA cells express Wnt4, which acts in an autocrine fashion, promoting mesenchymal-to-epithelial transformation to form the nephron (91). Several well-defined stages of nephron morphogenesis can be distinguished: the renal vesicle, the comma-shaped body, the S-shaped body, the capillary loop stage nephron, and finally the mature nephron. As the mesenchyme differentiates, it induces the growth and repeated branching of the UB, eventually giving rise to the collecting duct system (Fig. 1).

Figure 1.

Reciprocal signaling between the metanephric mesenchyme (MM) and the ureteric bud (UB) triggers the differentiation of the cap mesenchyme (CM) into a nephron. MM secretes signals essential for the formation of the UB from the Wolffian duct. Among these, glial cell-derived neurotrophic factor (GDNF) is apparently the most important inducer. UB-derived Wnt9b induces a fraction of MM cells, called the cap mesenchyme (CM), to condense and generate pretubular aggregates (PTAs). PTA cells express Wnt4, which acts in an autocrine fashion, promoting epithelialization (mesenchymal-to-epithelial transformation) into circular renal vesicles. These vesicles undergo extensive morphogenesis (tubulogenesis) developing into comma-shaped bodies, S-shaped bodies, and capillary loop stage nephrons, finally differentiating into mature nephrons.

Mature nephrons have two major components, the glomerulus and the renal tubule, which is organized into the proximal tubule, loop of Henle, and distal convoluted tubule. The mature glomerulus contains four cell types: parietal epithelial cells (PECs), podocytes, endothelial cells, and mesangial cells (98). By the S-shaped body stage, there is a distinct and separate cellular morphology for cells fated to be podocytes and those that will form the PECs of Bowman's capsule (59). The podocytes separate from PECs forming what will become Bowman's space.

Podocytes are postmitotic cells that cannot undergo a complete cell division and are therefore unable to regenerate themselves. Several studies have reported the presence of podocytes on Bowman's capsule in normal and diseased kidneys, near the vascular pole (30–32). These cells, which are analogous to visceral podocytes (6), have been named parietal podocytes, and several hypotheses have been advanced to explain their origin. Although direct evidence has not yet been provided, one hypothesis is that PECs, which proliferate throughout life, acquire an intermediate phenotype between PECs and the podocytes at the vascular stalk. At a later time, they migrate onto the glomerular tuft, thereby contributing to podocyte turnover (4). Fully differentiated podocytes may also leave the glomerular tuft and migrate onto Bowman's capsule (86). Unraveling the mechanisms that allow PECs to differentiate into podocytes may provide a novel approach to minimize the consequences of podocyte loss during aging and kidney disease.

As for glomerular endothelial cells, conflicting views exist regarding their origin. Although they have mostly been assumed to be derived from metanephric precursor cells that express the vascular endothelial growth factor (VEGF) receptor fetal liver kinase-1 (Flk-1) (1,2) cross-species transplantation studies have suggested that extrarenal cells might also contribute to glomerular vasculature (23,98). According to these authors, the induced metanephric cells presumably begin production of chemoattractants for endothelial cells at an early stage of differentiation (85).

After the initial influx of endothelial cells, mesangial cells enter the glomerulus and play a key morphogenetic role in capillary tuft formation (98). Similar to endothelial cells, the origin of the mesangial cells remains to be elucidated, although their predominant origin in cross-species transplant experiments is the donor (96).

Mature nephrons have developed a vascularized renal corpuscle, and their tubules have differentiated into distinct tubular segments. From the outermost to the innermost layer, the podocytes, the glomerular basement membrane, and the fenestrated endothelial cells constitute the glomerular filtration barrier, which allows fluid and small solutes to move from the capillary lumen into Bowman's space. While the glomerulus filters the blood, the tubules modify the filtrate by reabsorbing and secreting solutes. Each distal convoluted tubule delivers its filtrate to a system of collecting ducts. The so-formed urine then flows down the ureter into the bladder where it will be temporarily stored before released through the urethra.

Kidney Repair and Organogenesis: A 14-Year Search for Bona Fide Stem/Progenitor Cells

The adult mammalian kidney has been classically defined as a nonproliferative and nonregenerative organ, although the first evidence of kidney repair dates back to the late 19th century (7). Even though the adult kidney cannot make new nephrons, it can regenerate and recover in some circumstances. Indeed, tubular regenerative capacity changes widely from acute kidney injury (AKI) to chronic kidney disease (CKD). Acute renal insults are handled with successful regeneration but chronic injuries lead to ineffective or even more damaging cellular responses (56). More precisely, nephron tubule epithelium is regenerated after AKI, but in CKD tubular damage is not repaired and is accompanied by a sustained inflammatory response and myofibroblast activation, resulting eventually in interstitial fibrosis, tubular atrophy, and nephron loss. Regeneration can occur through multiple distinct mechanisms, such as stem cells, lineage-committed progenitors or dedifferentiation of mature lineage-committed cells into stem or progenitor cells. In the context of AKI, it is still unclear whether nephron tubular epithelium regeneration is mediated by bone marrow-derived cells (BMDCs) or adult/organ-specific stem/progenitor cells. Evidence of translineage differentiation of BMDCs, including hematopoietic and mesenchymal cells, into tubular epithelial cells (46,57,76), mesangial cells (41,45), and podocytes (77) has been proposed. Further analyses, however, indicated that BMDCs do not make a significant contribution to the restoration of epithelial integrity after ischemia–reperfusion (I/R) injury (22). It is likely that these cells release anti-inflammatory cytokines, such as interleukin-10 (IL-10), facilitating repair (65). Thus, proliferation of tubular cells rather than BMDCs might account for replenishment of the tubular epithelium after an ischemic insult (40). However, whether tubules regenerate from any surviving tubular cell or a specific tubular cell subpopulation with high regenerative potential remains controversial. Attention has recently been focused on “scattered tubular cells” (STCs) (35,58,87). These cells, which can literally be found scattered throughout the mammalian proximal tubule during I/R injury and subsequent regeneration, express almost the same markers as PECs (83,87). According to some authors, no cells express STC markers in healthy rat (87) or mouse kidneys (8). Rather, upon I/R injury, proximal tubular cells transiently acquire an STC phenotype and regenerate the tubule (8). This finding led researchers to exclude the possibility that STCs could represent “fixed” progenitor cells; indeed, some authors suggested that STCs could arise from dedifferentiation of any proximal tubular cell (8,87). However, other studies suggested the existence, at least in humans, of a subpopulation of proximal tubular cells that may be endowed with a more robust phenotype, allowing increased resistance to AKI, and enabling rapid repopulation of the injured tubules (35,58). It is therefore likely that humans and rodents use different strategies for tubular regeneration; in humans tubular progenitors likely preexist, while in rodents differentiated tubular cells likely achieve a transient tubular progenitor phenotype, which promotes the regenerative process (84). Another possibility is that both regenerative strategies act in concert. Regardless of their origin, STCs show a stem cell-like phenotype, expressing CD24 and CD133 in humans (35,58,87) and vimentin and CD44 in rodents (87). However, while STCs have been analyzed for expression of stem cell surface markers and aldehyde dehydrogenase (ALDH) activity, the signaling pathways responsible for the acquisition of the STC phenotype in response to I/R in the tubular cell population still need to be determined. Elucidation of the fundamental mechanisms underlying physiological regeneration and their experimental reactivation might represent not only an interesting and revolutionary area of basic research but also provide an exciting platform to complement existing treatment procedures for kidney function restoration. Indeed, a specific understanding of the delicate network of signal transduction pathways that cause a tubular cell subpopulation to revert from a differentiated to a dedifferentiated state may open new approaches to bolster renal tubular regeneration after AKI. More precisely, once the molecular mechanisms leading to proximal tubule regenerative responses have been elucidated, this information might be used to pharmacologically direct renal proximal tubular epithelial cell reversion to a less developed state in which they are able to proliferate and regenerate the damaged tubule.

Among other surface markers, c-kit has been suggested to mark a population of neonatal rat kidney cells with stem cell properties, including self-renewal, clonogenicity, and multipotentiality (79). Moreover, ex vivo expanded c-kit⁺ cells integrated into several compartments of the kidney, including tubules, vessels, and glomeruli, and contributed to functional and morphological improvement of renal function following acute I/R injury in rats (79). Whether a similar stem cell population is present in human kidneys remains to be determined.

The neural cell adhesion molecule (NCAM) 1 has also been indicated as a candidate kidney stem/progenitor marker. This glycoprotein, which is strongly expressed in the condensed mesenchyme, nephrogenic zone, and in Wilms' tumor (WT) progenitor blastema, is not expressed in mature kidney epithelia after nephron differentiation (12). Interestingly, once placed in culture and released from quiescence, adult human kidney epithelial cells, uniformly positive for CD24/CD133, reexpress NCAM1 in a specific cell subset that attains a stem/progenitor phenotype (12).

Much attention has also been devoted over the past decade to the isolation and/or creation of an unlimited supply of cells equivalent to the CM cells. These cells self-renew throughout kidney development and supply the cells required for nephron formation. Intensive research has successively come up with several cardinal transcription factors defining kidney stem/progenitor cells, including sal-like 1 (Sall1), Cbp/P300-interacting transactivator 1 (CITED1), SIX2, and odd-skipped related 1 (OSR1). Osafune et al. used knock-in Sall1-GFP mice to isolate cells strongly expressing Sall1 and demonstrated via colony-forming assays that the renal stem cell pool is contained within this population (72). CITED1 (10) and SIX2 (50) were then shown to mark a population of self-renewing CM progenitor cells. However, unlike SIX2, CITED1 is not expressed in the CM until some days after activation of SIX2, after nephrogenesis has commenced (50). Thus, SIX2, but not CITED1, could mark the nephron progenitor (NP) population throughout nephrogenesis. In further studies, OSR1 was shown to mark an even earlier lineage in the IM, capable of giving rise to all metanephric cell components, including the SIX2⁺ epithelial NPs, renal vasculature, and smooth muscle cells (67). In contrast, more recent evidence suggests that OSR1 might actually act downstream of SIX2 in the CM (104). Indeed, SIX2 is required for maintenance of OSR1 mRNA expression in the CM, and in turn OSR1 is indispensable for maintenance of the progenitor state of the CM cells during kidney organogenesis, since tissue-specific inactivation of OSR1 in the CM causes premature NP depletion and severe renal hypoplasia (104). Coimmunoprecipitation analysis of OSR1 and SIX2, using cotransfected HEK297T human embryonic kidney cells, indicated that OSR1 and SIX2 interact with each other (104). Both OSR1 and SIX2 can interact with TCF proteins. However, only OSR1 stabilizes the interaction between TCF/Lef and the Groucho/transducing-like enhancer-of-split (Tle) family of transcriptional corepressors. This interaction prevents the activation of Wnt/β-catenin target genes, consequently preventing premature differentiation of NPs (71). SIX2 also acts independently of OSR1 to activate expression of many CM genes, including itself, GDNF, and OSR1 (104). As UB-derived Wnt9b directs a subset of CM cells to form PTAs (13), TCF might be converted from repressor to activator and activate expression of differentiation genes, including Wnt4 (104).

In addition to the NPs, the developing kidney also comprises forkhead box D1 (FoxD1)-positive stromal progenitors, which give rise to interstitial cells and pericytes residing within the adult kidney (39), and the Scl⁺ and Flk1⁺ vascular progenitor cells that differentiate into renal vessels (19).

Even though specific surface antigen markers identifying kidney stem/progenitor cells have yet to be confirmed, the artificial reactivation of SIX2 and OSR1 in the adult kidney could be a valuable strategy to create a regenerative cell source in the treatment of kidney disease.

Differentiation of Embryonic Stem Cells into Renal Lineages

Embryonic stem cells (ESCs) are derived from in vitro culture of the blastocyst inner cell mass (25). Several studies have addressed ESC differentiation into renal lineages.

Kim and Dressler first reported that a combination of retinoic acid (RA), activin A, and bone morphogenic protein (BMP) 7 promotes the differentiation of mouse ESCs into IM-like cells and ultimately into renal epithelial cells able to contribute to tubular epithelia when injected into developing kidney rudiments (47). Similarly, Vigneau et al., using a Brachyury-GFP knock-in mouse ESC reporter line, showed that activin A enhances ESC differentiation toward a mesoderm phenotype (99). Importantly, fluorescence-activated cell sorting (FACS)-selected Brachyury⁺ cells were capable of integrating into proximal tubules when injected into mouse embryonic kidneys.

Other authors stably transfected Wnt4 cDNA into mouse ESCs. Embryoid bodies (EBs) derived from these cells expressed aquaporin-2 (AQP2), whose levels were enhanced by hepatocyte growth factor (HGF) and activin A. Notably, when transplanted into mouse renal cortex, Wnt4-EBs assembled into tubular-like formations (51).

Among other factors, BMP4 was also used to achieve primitive streak gene induction in mouse EB cultures and subsequent mesoderm development (11). Moreover, factors secreted by fetal renal cells were shown to further stimulate mouse ESCs, pretreated with RA and activin A, to differentiate into renal precursor cells in vitro (80).

In these studies, ESC differentiation was obtained using exogenous cytokines and EB formation. Other authors have investigated the possibility of inducing differentiation of ESCs into IM by cultivating these cells in a monolayer supplemented with a combination of small molecules. In particular, a combination of Janus-associated tyrosine kinase inhibitor 1 (JAK-inh), the phosphatidylinositol 3-kinase (PI3K) inhibitor LY294002, and the Rho/serum response factor (SRF) pathway inhibitor CCG1423 was found to induce expression of BMP7 in mouse ESCs (60). When these cells were cultured with RA, expression of OSR1 and paired box gene 2 (PAX2) became enhanced.

Human ESCs were also used to generate kidney structures. Takasato et al. described a three-stage framework for the differentiation of human ESCs from primitive streak [MIXL1⁺ (mix paired-like homeobox 1), LHX1⁺ (LIM homeobox 1)] to IM (OSR1⁺, PAX2⁺, LHX1⁺), and finally to MM [SIX2⁺, WT1⁺, GDNF⁺, HOXD11⁺ (homeobox D11)] (95). Induction of the posterior primitive streak from ESCs was first achieved using high BMP4/low activin A concentrations, or high concentrations of CHIR99021 (CHIR), a potent glycogen synthase kinase 3 β (GSK-3β) inhibitor/Wnt pathway agonist. Fibroblast growth factor (FGF) family members were then used to induce IM from posterior primitive streak. Specifically, ESC-derived posterior primitive streak populations, cultured with FGF2 or FGF9, but not FGF8, coexpressed OSR1, PAX2, and LHX1 with >80% of cells expressing PAX2, suggesting differential IM induction. Next, a combination of FGF9, BMP7, and RA after an initial induction using BMP4/activin A, or addition of RA/FGF9 after an initial CHIR induction, induced ureteric and metanephric progenitors, as indicated by formation of E-cadherin⁺ epithelial structures surrounded by clumps of mesenchyme positive for MM markers. The induced MM formed what seemed to be early nephrons/renal vesicles in vitro. In reaggregation assays, ESC-derived cells induced using the three-stage CHIR-FGF9 protocol integrated into all major cellular compartments of the developing kidney, demonstrating the feasibility of differentiating ESCs to a self-organizing kidney.

Transcriptional Reprogramming of Adult Renal or Nonrenal (Kidney Patient-Derived) Cells to Induced Pluripotent Stem Cells (iPSCs) and Differentiation of iPSCs into Mature Renal Structures

It is likely that adult kidney-derived cells are more amenable to reprogramming than any other cell type, as they are already situated within the renal lineage.

Song et al. (89) first provided proof-of-concept for the direct nuclear reprogramming of primary normal human mesangial cells (NHMCs) using retroviral delivery of the classic Yamanaka factors (94) octamer-binding transcription factor 3/4 (OCT3/4), sex-determining region Y-box 2 (SOX2), Krüppel-like factor 4 (KLF4), and v-myc avian myelocytomatosis viral oncogene homolog (c-Myc), hereafter referred to as OSKM. Twelve days after transfection, an average of 40 iPSC colonies were observed in cultures seeded with 5 × 10⁴ NHMCs. These colonies exhibited morphology similar to human ESCs and were positive for alkaline phosphatase (ALP), a phenotypic assessment of undifferentiated cells. Moreover, mesangial cell markers, including megsin, Thy-1 (CD90), desmin, α-smooth muscle actin (α-SMA), and Runt-related transcription factor 1 (RUNX1), were downregulated following reprogramming to iPSCs. Conversely, iPSCs expressed OCT3/4, NANOG, SOX2, FGF4, reduced expression protein 1 (REX1), telomerase reverse transcriptase (hTERT), developmental pluripotency-associated (DPPA)-2, 4, and 5. No relevant genomic aberrations were detected by karyotype analysis. Next, iPSC differentiation potential was investigated both in vitro and in vivo. Spontaneous in vitro differentiation was assessed by EB assay, which led to the appearance of cell types from the three different germ layers. Similarly, injection of undifferentiated iPSCs into immunodeficient mice resulted in the formation of encapsulated cystic teratomas, further confirming their pluripotent potential.

KLF4 and c-Myc are well-known oncogenes and are thereby a cause for concern when used for cell reprogramming. For this reason, Montserrat et al. investigated the possibility of creating iPSCs using just OCT4 and SOX2 on primary renal proximal epithelial cells isolated from patients undergoing tumor nephrectomy (66). Reprogrammed cells were observed as early as 13 days after viral transduction. A downregulation of the epithelial marker CD13 was observed upon reprogramming, while expression of surface markers associated with kidney development and/or progenitor cells such as CD10, CD24, and CD133 was retained. Apart from pluripotency markers, these cells also maintained occludin and vimentin expression, indicating that iPSCs retain an “epigenetic memory” of their cell type of origin.

Zhou et al. attempted generation of iPSCs from urine (108,109). Between 2,000 and 7,000 renal tubular cells are shed into the urine each day (44). These cells are undamaged and fully functional, making urine a valuable cell source for generating pluripotent cells. The urinary cell colonies routinely show two types of morphology, named type I and II. Type I colonies have a more regular appearance with smooth-edged contours and cobblestone-like cell morphologies, whereas type II colonies are more randomly arranged. Type II cells tend to grow more quickly than type I cells. Both types of colonies can be present in the same sample collection, but seldom in equal abundance. Moreover, cultures enriched in type I cells are slightly more frequent than type II cells. Type II cells may arise from partial epithelial dedifferentiation. In the Zhou et al. study, urine cells were subjected to OSKM retroviral infection. The reprogramming efficiency varied among donors, but in general was between 0.1% and 4%. Urine iPSCs were characterized by standard procedures including ALP staining, immunofluorescence for human ESC markers, quantitative polymerase chain reaction (qPCR) for endogenous ESC genes, and EB and teratoma assays.

Having iPSCs derived from kidney patients may help shed light on the cause and progression of disease. However, being able to establish iPSC lines from kidney patients might not be feasible, as the kidney itself is not readily accessible. The efficient production of adult iPSCs from urine is therefore of utmost importance, permitting cell collection at any time and eliminating the need for invasive procedures or cell banks. Moreover, since iPSCs derived from different sources retain a transient epigenetic memory of their somatic cells of origin (74), urine-derived iPSCs could prove to be a better source for obtaining renal tissue.

iPSCs have also been derived from keratinocytes of patients affected by autosomal-dominant polycystic kidney disease (ADPKD), systemic lupus erythematous (SLE), or Wilms tumor (WT), using lentiviral vectors expressing the OSKM factors (97). Approximately 5 to 20 iPSC-like colonies formed from 10⁵ infected cells. Similar to the above-described studies, patient-specific iPSCs expressed pluripotency markers, spontaneously differentiated in vitro through EB cultures, and formed teratomas in vivo.

While in all these studies further iPSC differentiation into mature renal lineages was not reported, Song et al. reported a method for differentiation of kidney-derived iPSCs into podocyte progenitors (90). To initiate differentiation, iPSC colonies, previously maintained with mouse embryonic fibroblast (MEF) feeders, were mechanically cut into pieces and plated in a suspension culture containing activin A, BMP7, and RA. After 3 days, the iPSCs were transferred to gelatin-coated plates and cultured for another 7–8 days in the same medium. These culture conditions favored iPSC conversion into cells possessing the typical arborized phenotype consisting of a main cell body with elongated processes characteristic of cultured human podocytes. Reprogrammed cells expressed markers of induced MM, including WT1 and PAX2, in addition to specific markers of mature podocytes such as synaptopodin and nephrin. Furthermore, iPSC-derived podocytes showed contractile ability and permeability and integrated into reaggregated metanephric kidney explants where they incorporated into developing glomeruli.

iPSCs obtained from human foreskin and dermal fibroblast cells through transduction with OSKM factors were used by other authors to generate structures expressing kidney proximal tubular markers (53). In this study, CHIR induced iPSC differentiation into ME-like cells. Gene expression profiling of CHIR-induced cells revealed that expression of primitive streak genes, including Brachyury, MIXL1, eomesodermin homolog (Xenopus laevis) (EOMES), forkhead box A2 (FOXA2), and Goosecoid (GSC), was rapidly upregulated within 24 h of treatment, peaked between 36 and 48 h, and decreased by 72 h, a pattern consistent with the transient expression of these genes during gastrulation. In parallel, the induction of ME genes was accompanied by the loss of pluripotency, as reflected by a reduction of OCT4 and NANOG mRNA. Importantly, CHIR induction resulted in a sustained upregulation of the epithelial-to-mesenchymal transition-specifying factor SNAI1 and downregulation of E-cadherin expression, suggesting CHIR is highly effective in induction of iPSC differentiation into ME-like cells via a program that mimics normal development in vivo. Further investigations highlighted the critical importance of timing and duration of signaling factors with regard to cell fate determination in this differentiation system. Withdrawal of CHIR after 24 h resulted in the reversion of incompletely differentiated cells back to the pluripotent state. High doses of activin A could divert CHIR-induced cells into definitive endoderm, but only if the drug was administered 24 h, and not 48 h, after CHIR treatment. Importantly, CHIR induction for 36 h, followed by the addition of FGF2 and RA, was effective in generating putative IM cells. Upon growth factor withdrawal, these cells, double positive for PAX2 and LHX1, formed polarized, ciliated tubular structures that expressed markers of kidney proximal tubular cells, including Lotus tetragonolobus lectin (LTL), kidney-specific protein (KSP), and N-cadherin. Moreover, in recombination explant cultures, these putative IM cells formed chimeric laminin-bounded structures and stochastically incorporated into clusters of SIX2⁺ mouse metanephric cells, although they did not express SIX2 protein by immunofluorescence. Importantly, SIX2 expression together with other markers of NPs could be promoted using a combination of FGF9 and activin A, further demonstrating the potential of PAX2⁺ LHX1⁺ population to give rise to IM derivatives.

These studies, however, did not quantitatively monitor IM differentiation. To this purpose, Mae et al. generated an OSR1-GFP knock-in iPSC reporter line using bacterial artificial chromosome (BAC)-based vectors (61). Treatment with CHIR and activin A at Stage 1, and CHIR and BMP7 at Stage 2, differentiated iPSCs into OSR1⁺ IM cells at an induction rate >90%. These cells could be further differentiated into cells of the adrenal cortex and gonad, as well as the kidney, following an additional 7-day treatment with BMP7 and Wnt3a, or CHIR. By performing high-throughput screening of approximately 1,800 chemical compounds, the same group successively identified two retinoic acid receptor (RAR) agonists, AM580 and TTNPB, which, when combined with CHIR, efficiently induced the differentiation of human iPSCs/ESCs into IM cells in only 5 days, without going through the ME step (5). These IM cells could further differentiate into 3D renal structures both in vitro and in vivo. Thus, a small molecule method can be used to rapidly, efficiently, and consistently produce IM cells at a relatively low cost.

Another group reported the differentiation of iPSCs into UB-committed kidney progenitor-like cells (102). iPSCs were derived from human fibroblasts using a non-integrative method based on the use of p53 suppression and nontransforming L-Myc with episomal plasmid vectors, first described by Okita et al. (71). The so-generated iPSCs were subsequently primed toward a UB progenitor-like cell fate using a 4-day protocol. Briefly, BMP4 and FGF2 rapidly specified iPSCs into mesoderm-committed cells as demonstrated by the significant induction of Brachyury. Subsequent exposure to RA, activin A, and BMP2 committed cells toward an IM fate, leading to renal progenitor-like cell generation, as suggested by upregulation of IM markers together with the upregulation of markers typical of UB lineages including homeobox B7 (HOXB7), RET, and GDNF family receptor α-1 (GFRA1). Interestingly, these differentiated cells integrated into the developing mouse UB structures and continued to mature.

Redefining Temporal and Spatial Transcriptional Programs in Metanephric Kidney for the Direct Differentiation of PSCs toward the Metanephric Mesenchyme

It seems unlikely that a single reprogramming strategy will generate all the different cell types comprising the adult kidney. Ideally, the creation of a NP cell population would circumvent this problem, as subsequent differentiation would replenish the full spectrum of renal cells, leading to regeneration. However, in order to achieve this goal, the spatial, anatomical, and temporal origin of NPs needs to be fully elucidated.

The mechanisms underlying how the nascent mesoderm becomes committed to the IM and how the MM and UB lineage segregate from one another have only recently been clarified. Osr1 is one of the earliest markers for the IM (E8.5–E9.5) and also marks MM (E11.5–E15.5). By using Osr1-GFP knock-in mice, Taguchi et al. confirmed that the Osr1-GFP⁺ population sorted from E11.5 and E15.5 kidneys contains colony-forming NPs (93). Upon closer examination, the authors found that the expression of Integrina8 (Itga8) and the absence of platelet-derived growth factor receptor α (Pdgfra) could further enrich for colony-forming cells. Colony-forming abilities of Osr1-GFP⁺ cells were therefore examined at earlier stages. At E8.5, no overlap of GFP with Itga8 was detected, and the GFP⁺ population did not form any colonies. However, at E9.5, the GFP⁺ population possessed the ability to form colonies. Since these cells were only detected in the anterior IM, authors hypothesized that they may represent mesonephric NPs. It was also thought that the non-colony-forming cells located in the posterior IM could be a precursor population for the metanephric NPs. When this posterior section of the E9.5 embryos was plated and treated with specific growth factors, metanephric NPs were derived. Interestingly, many combinations of growth factors failed to induce metanephric NPs from OSR1⁺ cells at E8.5, but colonies did form in the OSR1^– fraction. This led the authors to hypothesize that the posterior immature mesoderm, which was negative for OSR1, contained prospective NPs. Brachyury represents a marker of the primitive streak and posterior nascent mesoderm. Using an inducible Brachyury Cre transgenic mouse line, authors found that at E8.5, the precursor of the UB is located anteriorly in the Brachyury^– population and is already segregated from the MM, which is localized in Brachyury⁺ posterior nascent mesoderm. These findings argue against the conventional concept that the entire kidney is derived from the anteriorly formed IM, which extends caudally. Thus, the precursor of the MM is maintained and posteriorized in the Brachyury⁺ state until the E8.5 postgastrulation stage. This information was later used to derive MM from PSCs. A three-step protocol was applied to induce metanephric NPs from E8.5 Brachyury⁺ posterior mesoderm. In the initial phase of the induction, high concentrations of CHIR in combination with BMP4 were required. In the following step, lower CHIR concentrations, together with activin A and RA, were effective in increasing the expression of nephric genes. Subsequently, a further reduction of CHIR, addition of FGF9, and removal of BMP4, activin A, and RA led to the appearance of the metanephric NPs. Metanephric NPs were therefore generated from mouse ESC-derived EBs mimicking the protocol for the embryonic posterior mesoderm. When induced EBs were cocultured with embryonic spinal cords, either in vitro or in vivo, 3D renal tubules and glomeruli were observed. Interestingly, transplanted glomeruli could become connected to the host circulation, an essential requirement for glomerular function as a filtration apparatus. The protocol for mouse ESC differentiation was then successfully applied to human iPSCs, demonstrating the possibility of generating the kidney's complex organization by mimicking the proper temporal combinations of cell signaling that occur in vivo.

Additional Methods for Direct Reprogramming of Adult Cells: Use of Lineage-Instructive Transcription Factors and Cell-Free Extracts

Understanding the complex regulatory network of transcription factors that establish and maintain cell fate during development is crucial for the conversion of one somatic cell type to another through direct reprogramming. A combinatorial screen for lineage-instructive transcription factors was applied to identify the conditions under which adult renal cells could be transcriptionally reprogrammed to NPs (36). To this purpose, the adult proximal tubule cell line human kidney 2 (HK2) was transduced with 15 different factors, either alone or in combination. The kidney-specific transcription regulators eyes absent homolog 1 (EYA1), SIX1, SIX2, PAX2, c-Myc, n-Myc, OSR1, Mu-crystallin (CRYM), homeobox A11 (HOXA11), WT1, and mesenchyme homeobox 2 (MEOX2), SNAI1, SNAI2, and the stemness-conferring genes high mobility group AT-hook 2 (HMGA2) and OCT4, were included in the study. The histone deacetylase inhibitor valproic acid (VPA) was also used during reprogramming to assist in chromatin relaxation. The combination comprising SIX1, SIX2, OSR1, HOXA11, EYA1, and SNAI2 (pool 8) was the most effective in inducing epithelial-to-mesenchymal transition and in activating the NP gene regulatory network. Indeed, pool 8-reprogrammed cells became elongated and spindle-shaped by day 7 of reprogramming and showed decreased levels of E-cadherin and increased levels of matrix metalloproteinases (MMP) 9 and 2. Furthermore, NP markers like CITED1, SIX2, PAX2, EYA1, OSR1, Sall1, and FoxD1 were increased in pool 8-reprogrammed cells. A stringent assay for NP potential, in which E12.5 mouse kidneys were dissociated to single cells and mixed with GFP-labeled test cells, was then performed. Through this recombination assay, pool 8-transduced cells were found to differentially integrate into endogenous renal structures ex vivo. This indicates reinitiation of kidney development from a population of adult cells by generating embryonic progenitors may be feasible. Despite these promising results, further differentiation of the generated NP-like cells into mature renal lineages was not reported. Moreover, while the creation of a NP cell population could be of clinical importance, it cannot be excluded that progenitor cells with a more limited differentiation potential may also suffice as a therapeutic tool, since some pathologies are limited to specific cell types, such as podocyte loss seen in many glomerular diseases. Similarly, it is plausible that kidney endocrine functions such as hormone production could be replaced by transplanting only one particular renal cell type.

In parallel with these developments, an intriguing technology for direct cell reprogramming by exposing reversibly permeabilized somatic cells to cell-free extracts has emerged. Recently, human bone marrow stromal cells have been directly reprogrammed into cells that closely resemble renal proximal tubular epithelial cells using an extract from HK2 cells (73).

Assessing Therapeutic Effect of PSCs in Kidney Injury

Despite recent progress in iPSC research toward regenerative nephrology, most studies did not analyze the induced cells in functional, in vivo models. Little evidence of the therapeutic effect of iPSCs in animal models of kidney disease has been reported to date. In the study by Lee et al. (55), iPSCs were first generated by reprogramming mouse embryonic fibroblasts with three transcription factors, OCT4, SOX2, and KLF4. In kidney tissues with I/R-induced injury, transplantation of 5 × 10⁵ undifferentiated iPSCs via an intrarenal arterial route resulted in antioxidative, anti-inflammatory, and antiapoptotic responses. Production of nitrate/nitrite and malondialdehyde (MDA), as well as other proinflammatory cytokines, was substantially decreased, suggesting that iPSCs might be a potential resource for stem cell-based therapy against I/R-induced AKI. However, a larger number of iPSCs (5 × 10⁷ per animal) failed to reduce AKI. Microscopic examination indicated the high cell dose failed to resolve kidney injury as iPSC clumping and aggregation impaired renal circulation. Thus, the transplanted iPSC dose and concurrent renal blood perfusion monitoring are extremely crucial elements of cell therapy for the treatment of AKI.

Human bone marrow stromal cells reprogrammed using HK2 extracts (see above paragraph) were also shown to reduce renal damage in an experimental murine model of cisplatin-induced AKI (73). The same animal model was also used to test the functional activity of iPSC-derived renal progenitor cells (43). Undifferentiated iPSCs were exposed for 6 days to RA and to the small molecules RhoA inhibitor (CCG1423) and PI3K inhibitor (LY294002). Activin A was added for 2 days starting on day 2. From day 6 until 19, iPSCs were exposed to a cocktail of the nephrogenic factors FGF2, BMP7, and GDNF to commit cells toward renal progenitors. These cells robustly engrafted into damaged tubules, restoring renal function and structure in cisplatin mice with AKI. Although preliminary, these results indicate iPSCs can serve as an alternate and renewable source of engraftable cells for kidney regeneration, creating the basis for future applications in stem cell-based therapy.

Limitations of iPSC Technology and New Avenues for Cellular Reprogramming: The Chemically Induced PSCs

Although iPSCs hold great promise for regenerative medicine, there are still several issues surrounding this strategy that need to be solved, including the low efficiency of iPSC induction and differentiation, which may reflect the existence of genes or pathways acting as barriers to pluripotency reprogramming (78). By analyzing the transcriptional profiles of primordial germ cells (PGCs, the embryonic precursors to the gametes), PSCs and somatic cells, Qin et al. (78) recently found that the Hippo pathway, which regulates principal developmental processes including apoptosis, stem cell maintenance, differentiation, and organ size control (106), constitutes a barrier to cellular reprogramming. The authors found that large tumor-suppressor kinase 2 (LATS2), a kinase of the Hippo pathway as well as a tumor suppressor, was highly expressed in PGCs, but not in PSCs. Importantly, knockdown of LATS2 increased the efficiency of human iPSC generation. Further investigation revealed that LATS2 repressed reprogramming in human cells by posttran-scriptionally antagonizing TAZ, but not YAP, the major downstream effectors of the Hippo pathway.

Other epigenetic barriers for iPSC reprogramming, resulting from activity of the histone variant macroH2A (29), histone H3 lysine 9 (H3K9) methyltransferases/demethylases (15), and Bright/Arid3A transcription factor (75), have also been described. Deciphering features of somatic cell chromatin is therefore of utmost importance for the successful generation of reprogrammed cells.

Of note, the unlimited differentiation potential of pluripotent cells, if not controlled properly, could lead to the formation of unwanted tissues, even tumors. Direct in vivo cell reprogramming without reversion to a PSC could avoid the formation of unwanted tissues or tumors. To our knowledge, only a few studies to date have reported in vivo reprogramming of mature cells. Adenoviruses encoding specific transcription factors were injected into the pancreas of an adult mouse to convert exocrine cells to β-cells (107). These induced β-cells closely resembled endogenous β-cells and ameliorated hyperglycemia.

Nagy and Nagy (68) suggested that fibroblasts induced to pluripotency with the classic Yamanaka factors (94) go through different steps of development, with an uncertain fate between the “point of no return” (PNR) and the “commitment point to pluripotent stem cells” (CPSCs). If factor expression is stopped during this period, the cells do not become iPSCs nor do they return to their original state. They instead occupy an undefined state that was euphemistically termed “Area 51”. Cells in Area 51 might have properties similar to those of lineage-committed progenitors. Thus, it is likely that the Yamanaka factors, rather than imposing pluripotency, create an epigenetic state in which the culture conditions provide the key determinant of ultimate cellular phenotype. It turns out that a better understanding of the growth factors involved in defining and maintaining a required cell type could allow for lineage conversion between any two cell types, without the need to return to pluripotency, thus avoiding the risk of tumor formation. However, growth factors are expensive, and their effects can be inconsistent among different lots. Differentiation methods using small-molecule compounds would be less expensive and more consistent (5). By screening about 10,000 small molecules, researchers found a combination of seven compounds that were able to reprogram 0.2% of cells, just as in standard iPSC production protocols (38). These compounds included forskolin (FSK, a cAMP agonist), VPA, CHIR, 616452 (a transforming growth factor-β type I receptor inhibitor), tranylcypromine (a monoamine oxidase inhibitor), DZNep (an S-adenosylhomocysteine hydrolase inhibitor), and TTNPB. When injected into eight-cell embryos or blastocysts, these chemically induced PSCs (CiPSCs) were capable of integrating into organs of all three germ layers, including gonads, with transmission to subsequent generations, suggesting they were fully reprogrammed to pluripotency. Importantly, unlike chimeric mice generated from iPSCs, the chimeric mice generated from CiPSCs were 100% viable and apparently healthy for up to 6 months. Overall, these observations open new avenues in regenerative medicine, as patient-specific cells could be made without genetic manipulation, abolishing the risk of dangerous mutations or cancer.

The Lymph Node as a Potential 3D in Vivo Environment for Kidney Organogenesis from PSCs

The possibility of implanting developing nephrons beneath the renal capsule of a postnatal or an adult host kidney has been studied since the early 1990s (101). However, despite many published reports, the ability of transplanted nephrons to incorporate into the host collecting system, increasing host renal function, has never been demonstrated for isotranplants, allotransplants, or xenotransplants (33). However, very recently, Imberti et al. showed that embryonic kidneys, upon transplantation in syngeneic conditions under the kidney capsule of rats with progressive chronic nephropathy, not only undergo successful organogenesis with development into functional nephrons but also activate a local regenerative process within the host renal tissue (42). Indeed, increased vascularization and tubular cell proliferation, as well as reduced oxidative stress and apoptosis, were found in areas adjacent to the grafted metanephroi compared to distant regions of the same kidney (42). Realtime PCR and immunohistochemical analysis revealed grafted metanephroi were able to significantly increase the expression of growth factors including VEGF, FGF2, HGF, and insulin-like growth factor (IGF)-1 in host cells located in areas adjacent to the grafts compared to distant ones (42). Moreover, the activation of the senescence marker protein 30 (SMP30) in tubular cells in proximity to the grafted metanephroi, together with the appearance of NCAM, suggested the recapitulation of the tubular cell developmental program (42). Despite these promising findings, the authors did not observe an improvement of renal functional parameters nor a reduction in the degree of glomerulosclerosis and fibrosis (42). More importantly, it is not clear whether this approach may also be possible in the fibrotic and inflammatory milieu of a chronically injured human kidney. Significant differences exist between murine and human kidney anatomy. Indeed, the human kidney capsule and parenchyma cannot be as easily separated as in rodents to permit cell transplantation (64). Conversely, ectopic kidney organogenesis might be an alternative method to provide auxiliary kidney function (14,20). Unfortunately, while several extravascular and a few immunoprivileged sites have been considered as potential ectopic transplantation sites for organs like pancreas (88) and liver (69), ectopic sites for kidney reconstruction have not yet been fully examined. Hammerman's group suggested whole rat metanephroi implanted into the omentum might enlarge, become vascularized, and form mature tubules and glomeruli (82). However, other studies showed transplanted metanephroi can grow and develop for only a short time in the host omentum (21) unless an end-to-end anastomosis to the host ureters is performed (62). Only a few reports have shown it is technically feasible to microsurgically connect donor and host ureters. In these studies, ureteroureterostomy slowed progression of kidney failure in nephrectomized animals (49), prolonged short-term survival of anephric rats (62), or raised blood pressure in acutely hypotensive rats (105). Although promising, these results point to several clinical limitations, including postoperative adhesions and intestinal obstruction following omental manipulation (24,48). Transplantation sites other than the omentum might therefore be needed for successful restoration of renal function. Several other locations have been studied. Interestingly, the transplantation site itself has been shown to affect transplant growth. For example, compared with transplantation into the omentum, transplantation of rat metanephroi into the para-aortic area resulted in better renin production, although no changes in EPO production were observed (63).

Among others, our previous findings suggest the lymph node might be a more clinically relevant transplantation site (27,28,37,52). First, there are over 500 lymph nodes in the human body, many of which are easily accessible. Second, although a single lymph node structurally limits the number of donor cells that can be transplanted, it is technically feasible to transplant more than one lymph node to gain sufficient organ or tissue function from the transplanted cells. The potential loss of function in a few lymph nodes does not seem to compromise the overall function of the lymphatic system (54). Third, lymph nodes have ready access to the bloodstream, nurturing the transplanted cells with nutrients as well as hormones and signaling agents needed for growth. Importantly, new angiogenesis occurs fast enough in this site to sustain cell survival and engraftment. All these characteristics allowed us to generate a functional kidney within a mouse lymph node (28). We showed a time-dependent maturation of embryonic kidney fragments within the lymph node. Although some degree of maturation could already be observed in 3-week grafts (Fig. 2), the first mature nephrons did not appear until the sixth week. These structures showed blood filtration capability as well as the ability to concentrate urine. Both bone marrow hematopoietic and stromal cells were found to support ectopic kidney organogenesis. Importantly, not only did the lymph node furnish the developing tissue with host cells but the lymph node also provided growth and homeostatic signals, since a decrease in native renal mass could push maturation of the ectopic graft. Although further studies are needed to understand how donor cells communicate with neighboring and distant host cells, our study suggested that the lymph node might provide a unique 3D environment for the validation of the differentiation potential and functional maturity of candidate stem cells in regenerative nephrology. We believe the lymph node may also provide a unique site for kidney organogenesis for therapeutic purposes. However, various hurdles remain before a clinical therapy can become a reality. To apply this technology clinically, a suitable cell/tissue source would be needed. Cell culture techniques to produce renal organoids starting from mouse embryonic kidney precursors have been described, but several experimental attempts to develop functional glomeruli have failed because the avascular in vitro environment does not permit glomerulogenesis. A significant breakthrough in this field has recently been provided by Remuzzi's group, which showed that renal organoids grown in vitro from single-cell suspensions derived from E11.5 murine kidneys can generate both glomeruli capable of filtering blood and tubules capable of reabsorbing filtered macromolecules when implanted beneath the kidney capsule of athymic rat hosts (103). However, these organoids were only viable for 3–4 weeks and began to involute thereafter. Moreover, for clinical use, these organoids would have to come from a human fetus, which is ethically problematic. iPSCs have the potential to overcome the various ethical problems surrounding the use of ESCs. As we have discussed, great effort has been made to induce differentiation of these cells into renal lineages. However, in most of the studies described in this review, iPSC differentiation was performed in monolayers, which may represent an adverse environment for self-organization and morphogenesis. The possibility exists that iPSCs require 3D environments or even transplantation into live animals to fully mature (92). Indeed, although these cells can partially differentiate along a given cell lineage in vitro, in most cases, their progeny survive very poorly in vivo, failing to completely mature and generate a source of functional cells required for tissue repair (34,70). However, if the iPSCs are allowed to differentiate in vivo rather than in vitro, maturation and survival of various cell types can be enhanced (3). Therefore, it is essential to provide the reprogrammed cells with a suitable niche; otherwise, the newly established stem cell phenotype will be unstable. Our studies indicate that lymph node transplantation might offer unprecedented opportunities to study artificial stem cell maturation and function, as well as provide a unique organogenesis site for therapeutic purposes. Briefly, somatic cells derived from patients could be reprogrammed into renal progenitors by utilizing our understanding of mouse kidney development and organogenesis (Fig. 3). While specific surface antigen markers have yet to be identified for kidney stem/progenitor cells, intensive research has come up with several cardinal transcription factors including SIX2, which marks nephron progenitors throughout nephrogenesis (50). In addition to nephron progenitors, the developing mouse kidney is also comprised of HOXB7⁺ ureteric epithelial progenitors, FoxD1⁺ stromal progenitors, which give rise to interstitial cells and pericytes residing within the adult kidney (39), and the Scl⁺ and Flk1⁺ vascular progenitor cells that differentiate into renal vessels (19). While these cell subsets might be easily isolated from transgenic reporter gene mouse strains, human cells with these characteristics might be obtained through reprogramming patient-derived cells using multistep differentiation protocols. The lymph node might be used later as a bioreactor to grow these cells. The expectation is that the renal progenitor-like cells will interact with each other and undergo nephron morphogenesis via vesicle, comma-shaped body, and S-shaped body stages (Fig. 3). Proof of our concept, that the derived ectopic kidney structures are functional, will be the rescue (or at least partial rescue) of animals suffering from renal failure in a model of chronic kidney disease. To this purpose, fluid wastes produced by the engrafted lymph nodes must be allowed to drain out of the animal. This might be achieved by anastomosing ectopic kidney tissue and host ureters with an artificial conduit using tissue-engineering techniques (Fig. 3).

Figure 2.

Lymph nodes are permissive sites for kidney organogenesis. Top: schematic view of transplantation of embryonic kidney at the S-shaped body stage of nephrogenesis into the jejunal lymph node, maturing into glomeruli. Bottom: tissue sections of embryonic kidney 3 weeks after transplantation into the lymph node. The sections were stained with hematoxylin and eosin, anti-collagen IV (constituent of mesangial matrix), anti-CD31 (blood vessel marker), or anti-podoplanin (podocyte-specific marker) antibodies (scale bars: 50 μm).

Figure 3.

Proposed cellular approach to rebuild a functional ectopic kidney by transplanting renal progenitors into the lymph node. Renal progenitors can be obtained from patient-derived iPSCs (e.g., reprogrammed from dermal fibroblasts or kidney cells) following exposure to combinations of nephrogenic factors. Alternatively, renal progenitors can be isolated directly from the embryonic kidney. During development, kidney rudiment comprises a ureteric bud (UB) surrounded by a cap of undifferentiated metanephric mesenchyme (CM). The progenitor populations that are likely necessary for rebuilding a functional kidney are nephron progenitors (giving rise to all nephron epithelia), ureteric epithelium progenitors (giving rise to collecting duct cells), renal stromal cells progenitors (giving rise to pericytes, mesangial cells, vascular smooth muscle, and subsets of peritubular endothelium), and vascular progenitors (giving rise to glomerular and nonglomerular endothelium). Following injection of renal progenitors into the lymph node, nephron morphogenesis is expected to occur, leading to the emergence of fully functional nephrons. Engineering techniques will be necessary to drain fluid wastes out of the ectopic graft and to ensure that ectopic renal functions support failing kidneys.

Taken together, we suggest the lymph node might be used to study functional specialization from populations with pluripotent features, and might also provide a unique site for kidney organogenesis for therapeutic purposes.