Podocyte-Parietal Epithelial Cell Interdependence in Glomerular Development and Disease

Abstract

Podocytes and parietal epithelial cells (PECs) are among the few principal cell types within the kidney glomerulus, the former serving as a crucial constituent of the kidney filtration barrier and the latter representing a supporting epithelial layer that adorns the inner wall of Bowman's capsule. Podocytes and PECs share a circumscript developmental lineage that only begins to diverge during the S-shaped body stage of nephron formation–occurring immediately before the emergence of the fully mature nephron. These two cell types, therefore, share a highly conserved gene expression program, evidenced by recently discovered intermediate cell types occupying a distinct spatiotemporal gene expression zone between podocytes and PECs. In addition to their homeostatic functions, podocytes and PECs also have roles in kidney pathogenesis. Rapid podocyte loss in diseases, such as rapidly progressive GN and collapsing and cellular subtypes of FSGS, is closely allied with PEC proliferation and migration toward the capillary tuft, resulting in the formation of crescents and pseudocrescents. PECs are thought to contribute to disease progression and severity, and the interdependence between these two cell types during development and in various manifestations of kidney pathology is the primary focus of this review.

Introduction

Podocytes are terminally differentiated, arborized visceral epithelial cells that cover the outer surface of the glomerular basement membrane (GBM). Along with the fenestrated endothelium and basement membrane, the pedicles of the podocytes comprise the glomerular filtration barrier. Podocyte dysfunction as a result of either genetic and/or environmental insults is a leading cause of CKD worldwide. In proliferative glomerulopathies, such as rapidly progressive GN (RPGN) and collapsing and cellular subtypes of FSGS, podocyte injury contributes to the pathogenic activation and proliferation of neighboring parietal epithelial cells (PECs).^1–4 Along with this agglomeration of activated PECs in Bowman's space, capillary injury leading to immune cell infiltration and necrosis comprise the pathogenic pleiomorphic crescent observed in RPGN.^1–4 PECs are the predominant cellular constituent of these early cellular crescents, and as interstitial fibroblasts and extracellular matrix proteins invade the crescents, they become fibrocellular and eventually fibrotic structures compromising glomerular filtration. Although several immune disorders underlie crescentic GN, PEC activation and proliferation is common to all, yet there is no unifying mechanism to explain such aberrant cell growth. Similarly, PECs make up most of the abundant cellular material that occupies Bowman's space in the collapsing and cellular forms of FSGS.^1–4 Although primary podocyte injury leading to severe proteinuria is the underlying pathology in these disorders, just how this leads to PEC activation and proliferation remains poorly understood. Furthermore, we observe a reactivation of signaling pathways involved in development in these activated PECs. Therefore, it is critical to understand not only the mechanisms mediating podocyte-PEC interdependence in these proliferative glomerulopathies but also the pathways that enable differentiation of podocytes and PECs in normal kidney development. Here, we review the developmental processes that give rise to distinct podocytes and PECs as well as the critical mediators involved in activation and proliferation of PECs as a result of podocyte injury in proliferative glomerulopathies.

Developmental Origins of Podocytes and Parietal Epithelial Cells

Developmental trajectories of the nephron, the functional unit of the kidney, have been elucidated over the course of the past several decades. Mammalian nephrogenesis reaches its apotheosis as an arborized filtration structure that is produced through a sequential cascade of two morphologic event types: branching morphogenesis of tubular structures and mesenchymal to epithelial transitions (METs).^5,6 Although much is known about the early organization of the metanephros (permanent kidney in reptiles, birds, and mammals), there remains a gap in knowledge in the processes mediating the terminal cell differentiation in the nephron, specifically that of the elusive PECs lining Bowman's capsule.⁷ As such, highlighting the developmental processes that give rise to distinct podocytes and PECs will be critical in informing our understanding of their contribution to various glomerular diseases.

Interactions of the Ureteric Bud and Metanephric Mesenchyme

A primary driver of early kidney development is the zinc-finger odd-skipped related transcription factor 1 (Osr1).^8–10 It is localized to mesenchymal precursors within the metanephric mesenchyme (MM). The MM is closely associated with the ureteric bud (UB), arising from the Wolffian duct. It is these two structures working in tandem that produce all the cell types within the nephron.¹¹ Osr1+ progenitor pools within the MM are responsible for generating both the epithelial and interstitial cell fates, with the dual expression of Paired box 2 (Pax2) and Sine Oculis Homeobox Homolog 2 (Six2) specific to the epithelial fates of podocytes and PECs (Figure 1).^9,12 The MM coalesces around the tip of the UB to produce the capping mesenchyme, which is defined by Six2+/Cited1+/Wnt4− progenitor populations, and is the source for the entire nephron-specific stem cell milieu.¹³ Continual Pax2 expression in the cap mesenchyme during its interaction with the UB is crucial for endorsing nephron-specific epithelial fates, leading eventually to PECs and podocytes, while repressing nonepithelial fates (Figure 1).¹⁴

Figure 1.

Specific transitions during stages of kidney development leading to podocyte and PEC differentiation. The molecular crosstalk between the UB and cap mesenchyme is coordinated by Pax2 and Osr1 expression, respectively. This initial stage of kidney formation gives way to two stages–formation of the pretubular aggregate which subsequently coalesces into the appearance of the renal vesicles. Both of these steps represent a window in kidney glomerular development characterized by wholesale METs. The comma-shaped body then forms at the lateral edges of what was previously the cap mesenchyme, with Notch signaling driving predominantly glomerular fate via this structure. The S-shaped body stage which appears next in orderly developmental timing finally sees the molecular and functional divergence of podocyte and PEC fates, with the former working to erect a mature GBM and the latter upregulating Wnt signaling to ensure proper final positioning on the inner wall of Bowman's capsule. The S-shaped body stage, therefore, progresses into the formation of individual, fully mature, nephron segments.

MET

A critical moment for the development of the nephron is the MET of epithelial fate-committed progenitors arising from the cap mesenchyme. The Wilms tumor 1 (WT1) transcription factor (TF) is an important regulator of MET, with its absence resulting in renal agenesis within mouse embryos (Figure 1).^15–17 This process begins with the UB signaling the overlying mesenchyme to aggregate and undergo MET through the activation of specific genetic elements (e.g., Lhx1, Fgf8), leading to the simple epithelium of the renal vesicles (Figure 1).¹⁸ In addition, although canonical Wnt signaling plays a role later in differentiation of PECs, Wnt-4 specifically has been found to be an early potent inducer of METs during nephrogenesis.¹⁹ Functional nephrons eventually arise via three critical developmental transitions and set the stage for podocyte and PEC differentiation: the MET which originates the renal vesicle from the cap mesenchyme, the segmentation and patterning of the renal vesicle, and finally the fusion of the renal vesicle with the tubular ureteric structures of the fetal kidney which finally results in a functional contiguous uriniferous tubule.²⁰

Patterning of the Renal Vesicle

The Notch pathway is critical to patterning of the renal vesicle, with evidence for Notch-specific morphogen polarity (Figure 1).^21,22 The renal vesicle undergoes a morphological shift from a simple tubular structure through a comma-shaped body and an s-shaped body, finally arriving at a fully differentiated nephron.²³ During comma-shaped body formation, the proximal tubule (PT) versus glomerular (PEC and podocyte) fates are established through an intricate interplay of Notch ligands (Dll1, Lfng, and Jag1) which lie medial and distal, repressing all proximal cell fates (Figure 1).^20,24–26 Interestingly, a reactivation of Notch signaling contributes to PEC activation and proliferation in proliferative glomerulopathies.^27,28 In addition, fully formed differentiated glomeruli are maintained by active WT1 repression of Pax2 activity in the medial and distal tubules of the s-shaped body.²⁹

Divergence of Cell Fate in Glomerular Epithelial Cells

It is at the comma-shaped body to S-shaped body transition where PECs and podocytes, who share a common lineage until this point, begin to diverge and express their cell-specific markers. In PECs, the expression of CK8, PAX2, CRIM1, FGF2, PGP9.5, SSeCKS (Akap12), claudin-1 and claudin-2 are upregulated while WT1 is downregulated (Figure 2).²⁹ In podocytes, podocalyxin, WT1, nephrin, and podocin are upregulated (Figure 2).^30,31 Vasculogenesis is initiated by Vascular Endothelial Growth Factor A, which induces endothelial cell proliferation and migration into the vascular cleft of S-shaped bodies. This causes the formation of glomerular capillaries where neighboring epithelial cells surround the capillary loop and differentiate into podocytes.^32–36 While Vascular Endothelial Growth Factor A promotes endothelial cell survival and functions as a chemoattractant, guiding endothelial cell migration into the vascular cleft of the nephron during the S-shaped body stage, Sema3a functions as a chemorepellent.³⁷ Sema3a is not only expressed in the developing S-shaped body³⁸ but is necessary for podocyte differentiation.³⁹ Interestingly, although Sema3a is visualized in both PECs and podocytes in E17 mouse glomeruli, expression is retained only in podocytes and regions of the collecting duct and distal tubule in the adult kidney (Table 1).³⁸

Figure 2.

Divergence and spatial reorganization of PECs and podocytes across the comma-shaped/S-shaped stage boundaries of kidney development. During the early stages of comma-shaped periods of kidney development, podocytes and PECs begin to acquire their respective fates, through the podocyte-specific upregulation of (1) WT1, (2) nephrin, and (3) podocin–and the PEC-specific increase in claudin-1 and Pax2. During the transition from the late comma-shaped body stage to the S-shaped body stage, PECs reorient themselves around an early version of Bowman's capsule, thereby assuming a more fully developed glomerular posture.

Table 1.

Summary of key molecules critical for podocyte and parietal epithelial cell development

Gene	Murine Model	Cell Culture Model	Methods	Key Findings	Reference No.
Sema3a	Loss of function: Flk1-lacZ+/−; Sema3a+/− Gain of function: podocin-rtTA; tet-O-Sema3a	—	Sema3a loss and gain of function mouse models	Sema3a is an essential negative regulator of endothelial survival during glomerulogenesis, and overexpression resulted in endothelial apoptosis and abnormal podocyte development	38
Notch1	—	Mouse metanephric organ culture	Cultured embryonic mouse metanephros in the presence of a y-secretase inhibitor	Inhibition of γ-secretase resulted in impairment of podocyte formation	28
Ctnnb1	Pax8-Cre; β-cateninfl/fl	—	Pax8-Cre driven genetic knockout mice resulting in β-catenin deficiency	PEC-specific knockout of Ctnnb1 resulted in developmental abnormalities and caused podocytes to be in an abnormal parietal position, replacing normal flat PECs	40
Akap12	E18 mouse kidneys and global Akap12 genetic knockout mice	—	Prenatal mouse kidneys and Akap12 genetic knockout mice	At the S-shaped stage, Akap12 expression is only found in embryonic PECs. Akap12 knockout mice exhibited PEC hyperplasia, PEC proliferation, and glomerular expression of nuclear cyclin D1	41
Cldn1, Pax8	Embryonic mouse kidneys	—	Anti-GBM model of crescentic nephritis	Cldn-1 and Pax8 expression was present in both, PECs and podocytes during the S-shaped body stage but limited to PECs by the capillary loop stage	42

PEC, parietal epithelial cell; E18, embryonic day 18; Cldn-1, claudin-1; Akap12, SSeCKS; GBM, glomerular basement membrane.

PECs and podocytes likely require fate specification from signaling pathways at this stage of development. Notch signaling plays a critical role in podocyte fate specification.^40,43 In cultured mouse metanephros, Notch expression was initially visualized in comma-shaped and S-shaped bodies during development using an antibody that specifically targets the γ-secretase–cleaved intracellular domain of Notch. Treatment with N-S-phenyl-glycine-t-butyl ester blocked Notch signaling by inhibiting γ-secretase activity. The γ-secretase complex is a key player in the Notch signaling cascade as it mediates the cleavage of the extracellular domain of the Notch receptor, thereby permitting the nuclear translocation of the intracellular domain and eventual expression of Notch target genes.⁴⁴ Inhibition of γ-secretase resulted in impairment of podocyte formation, evident by a markedly lower level of positive-WT1 staining in cell clusters compared with the control group.⁴⁰ Together, these data also suggest that podocyte formation may require γ-secretase activity at two distinctive stages: initially, at the S-shaped body stage duration for formation of podocyte precursors and later at the capillary-loop stage for podocyte maturation.

As podocytes deposit the constituents of the GBM, with help early on from glomerular endothelial cells,⁴⁵ the cells fated to become PECs are induced to express their characteristic genes. PECs require β-Catenin/Wnt signaling to specify their fate (Figure 1, Table 1),⁴⁶ which is, interestingly, reinduced in activated PECs in proliferative glomerulopathies.⁴⁷ Furthermore, mice with a conditional Pax8-Cre–driven knockdown of β-Catenin at the late S-shaped body stage and developing collecting ducts demonstrate a lineage switch during late nephrogenesis from PECs to differentiated podocytes in Bowman's capsule. These parietal podocytes develop directly from the parietal layer of S-shaped bodies in β-Catenin–deficient mice. This highlights a close relationship between PECs and podocytes. Podocytes are terminally differentiated and thought to be unreplenishable. However, in some conditions of podocyte injury, PECs have the potential to transdifferentiate into podocytes. This renal stem-cell-like property of PECs, where they are able to undergo cell cycle arrest and re-entry multiple times is why they might serve as podocyte precursors in mature glomeruli.^41,48 This unique capability was further investigated, where PEC-specific expression of SSeCKS was responsible for regulating the activity of cyclin D1 (Table 1).⁴⁹ Interestingly, SSeCKS is expressed in PECs, but not podocytes, starting at the S-shaped body stage and is constitutively expressed postnatally by PECs. SSeCKS expression and binding of cyclin D1 in the cytoplasm seemed to be induced via cell-to-cell contact inhibition in cultured mouse PECs. Collectively, these data demonstrate a pivotal role of SSeCKS in mediating the proliferative capacity of PECs, albeit this may ultimately depend on specific subtypes(s) of PECs in question (addressed in the next section). In summary, the PEC and podocyte lineage split begins in earnest during the early periods of comma-shaped body development and reaches its spatial apotheosis during the S-shaped body stage of nephron development (Figure 2). Although much is known regarding the genetic determinants of PEC and podocyte specification, a large knowledge gap remains in the field concerning the precise molecular cue responsible for their initial lineage divergence. A clearer view of this process is essential for informing stepwise progression of pathophysiological interplay between these two glomerular cell types.

Following the divergence of PEC and podocyte fates, once the differing cell types settle into their specific anatomical positions and discrete functional roles, there remains evidence of physical interactions that may underlie their interdependence in both healthy and diseases glomeruli.⁵⁰ Hackl et al. have clearly demonstrated the presence of podocyte/PEC connections either at the vascular pole or spanning Bowman's space without the presence of injury.⁵⁰ These bridges were also visualized in podocin-Green Fluorescent Protein mice after unilateral ureteral obstruction, a model used to study obstructive nephropathy, and presented often as very thin nanotube structures (<350 nm in diameter).⁵⁰ By contrast, broader and thicker adhesions can be appreciated between the capillary tuft and Bowman's capsule in the Munich Wistar Frömter rat model of progressive glomerular injury, a strain that demonstrates spontaneous development of systemic hypertension, proteinuria, and glomerulosclerosis.^51,52 The stated evidence and numerous other studies detailing physical links between PECs and podocytes in normal and pathological states serves to enhance our understanding of their interdependence.

PEC Subtypes

The population of PECs lining the walls of Bowman's capsule is not a monolith. Although made up of epithelial cells arising from the same developmental lineage, adult PECs can be stratified along differing morphological, molecular, and positional planes.^53–55 A recent review by D'Agati et al. illustrates that the subtypes of PECs (mouse and human) comprising the inner covering of Bowman's capsule can be oriented centrally or closer to either the vascular or tubular poles of the glomerulus and, therefore, directly interface with podocytes or proximal tubular cells.⁵³ In the adult normal kidney, quiescent PECs are flat squamous epithelial cells that uniquely express several key markers, such as Pax-2 and claudin-1, after the developmental divergence from their shared lineage with podocytes.^56,57 In the vascular pole transition zone that delineates the border between PECs and podocytes, intermediate cells can be found which express markers of both lineages.^31,58 Although these cells have been termed parietal/ectopic podocytes because of their expression of markers for both podocytes (WT-1) and PECs (Pax-2), ultrastructurally they resemble podocytes–with prominent foot processes and less flattened cell bodies than quiescent PECs.^7,31,59 In addition to these misplaced podocytes, both peripolar regions contain cells with progenitor class molecular signatures, which display intermediate gene expression patterns between podocytes and PECs.⁷ The tubular pole transition zone contains cells that stain for glycCD133+, CD24+, and podocalyxin. These cells have been referred to as adult parietal epithelial multipotent progenitors, with clonal analysis showing them capable of regenerating both lost podocytes and tubular epithelial cells.⁶⁰ The vascular pole also contains a population of cells termed podocyte committed progenitors that express the podocyte marker podocalyxin, as well as the general progenitor cell marker nestin.^27,60 These putative podocyte progenitor cells depend on Notch activation for maintenance of the cell cycle, and its downregulation leads to the terminally committed podocyte lineage.^27,60 Single-cell/nuclear genomics (RNA, assay for transposase-accessible chromatin) has recently been at the forefront of uncloaking hitherto underappreciated PEC populations in various kidney disease states.^61–64 Although few of these studies focus on PECs specifically, they are useful bystanders in the analysis of large single-cell/nuclear RNA-seq (gene expression) data sets. However, there are very limited examples of single-cell/nuclear RNA-seq studies that specifically describe PEC subcluster analysis as this will be crucial to delineating novel roles for PEC subtypes in vivo. One such recent study from Melica et al. describes a subset of putative human CD133+ PEC/renal progenitors that, after specific pathological cues during development of crescentic GN, can differentiate into functionally mature podocytes.⁶⁵ A gray area in single-cell studies of myriad kidney pathologies lies in the overlap between activated PEC gene expression (e.g., Vcam1, Havcr1, Cd44, and Slc34a1) and that of PT subsets related to injury, repair, and fibrosis.^62,66 These seeming inconsistencies need to be sorted out to truly delineate subtypes of PECs from those of injured PT.

PECs in Health and Disease

Adult podocytes are critical, terminally differentiated constituents of the glomerular filtration barrier, which do not undergo self-renewal after injury in adulthood.⁶⁷ Numerous studies have demonstrated that podocyte loss contributes to CKD, as a result of glomerulosclerosis and reduction in total functional glomeruli.^68–70 Given this, the search for a bona fide podocyte progenitor population has been an intense focus of many research laboratories worldwide. PECs have been theorized to be key players in podocyte regeneration and in pathophysiological processes; however, our understanding of PECs is incomplete.^67,71 Despite being identified almost two centuries ago, the functional role of PECs under basal conditions remains unclear. Some evidence suggests that PECs form an auxiliary filtration barrier to help prevent escape of albumin outside of Bowman's capsule.⁷² This is consistent with PEC expression of tight junction proteins, including claudin-1 and zonula occludens-1, and occludin. The expression of these tight junction markers was also reduced in anti-GBM disease and associated with increased permeability of the PEC-Bowman's basement membrane.⁷²

Despite limited evidence highlighting the functional roles of PECs in health, numerous studies have begun to investigate PEC response to injury. In proliferative glomerulopathies, a PEC response is typically secondary to outside forces, either from other glomerular cell types or systemic factors. Although a PEC response can be observed in an assortment of human kidney disease, including RPGN and the cellular and collapsing subtypes of FSGS, there are many types of glomerular disease that lack an obvious PEC response, such as minimal change disease. This suggests that PEC responses are not a generic reaction to any type of podocyte or glomerular injury but instead are triggered under specific conditions. For proliferative glomerulopathies, such as RPGN, in contrast to podocytes, which typically become injured and eventually lost, PECs acquire a proliferative and activated phenotype in which they accumulate within the urinary space, compromising glomerular filtration. PECs also become migratory and encroach onto the glomerular tuft, further perpetuating glomerular injury and sclerosis. From a clinical viewpoint, it is well accepted that a percentage of glomeruli with crescents correlates with decline in kidney function⁷³ and effort over the recent years have been aimed at understanding the pathways that inhibit PEC activation with an ultimate goal of using these pathways to identify novel therapeutic targets. The new targets in question potentially emanate from multiple disease-associated molecular pathways, as will be outlined in the forthcoming sections.

Histopathologic Consequences of PEC Activation

Proliferative glomerulopathies are characterized by unchecked cell division in Bowman's space, with bidirectional cellular and molecular inputs from both the capsule and capillary tuft.^74,75 Crescentic GN is one such kidney pathology, named for the crescent-shaped accumulation of activated cuboidal PECs and a variety of potential cellular immune mediators. T-cell and macrophage infiltration from both the glomerular capillaries and a newly denuded Bowman's capsule, as well as fibrin extravasation into the interstitium through that degraded outer glomerular shell, are hallmarks of RPGN.^74,75 The bulk of the crescentic cellular content, however, is represented by the aforementioned activated PECs. By contrast, in cellular/collapsing subtypes of FSGS, pseudocrescentic glomerular lesions are often identified which represent histopathologically limited versions of RPGN crescents, with a predominance of PEC proliferation after podocyte loss, yet without capsular breach allowing for immune cell infiltration and fibrin loss.^54,76 In both RPGN and cellular/collapsing variants of FSGS, the extent and rapidity of podocyte loss has been clearly linked to the progression of glomerular lesion severity.^70,76,77 This effect was most clearly demonstrated by a study that used transgenic rats in which human diphtheria toxin could specifically ablate podocytes, where only after passing a threshold of 20% podocyte loss did the animals develop sustained proteinuria and FSGS with PEC migration.⁷⁰

Molecular Determinants of PEC Activation

Cluster of Differentiation (CD44)

Of all that is understood about the molecular changes associated with activated PECs, CD44 remains the most robust and widely accepted marker for activated PECs (Figure 3). The protein encoded by the CD44 gene is a cell surface glycoprotein which regulates a wide range of cellular processes, including inflammation, proliferation, cancer, and migration.^78–80 CD44 binds several target molecules, including Extra‐Cellular Matrix proteins, such as hyaluronate, osteopontin, laminin, and fibronectin.⁸¹ Interestingly, CD44 is barely detectable in the kidney under normal conditions but is expressed de novo in human kidney biopsies and murine models of cellular and collapsing subtypes of FSGS.^1–4,82,83 In a mouse model of FSGS induced by an injection of sheep antipodocyte antibodies, global Cd44 knockout confers significant protection against proteinuria, glomerulosclerosis, and PEC activation.⁸⁴ In cultured mouse PECs, overexpression of Cd44 led to increased collagen IV expression and cell migration.⁸⁴ These studies identify a significant role for CD44 as a crucial driver of PEC activation and migration. The authors also showed that activated extracellular-related kinase (ERK) 1 and 2, pERK1/2, often colocalizes with CD44 and is increased in activated PECs.⁸⁵ They found that ERK1/2 signaling regulates expression of CD44 and migration of PECs in culture, implicating ERK1/2 as an upstream regulator of CD44. However, the upstream signals triggering activation of ERK1/2 in activated PECs are unknown. An interesting point to note is the finding of similar ERK1/2 activation in mice with and without deletion of CD44.⁸⁵

Figure 3.

Graphic of healthy versus injured glomerulus, implicating specific pathways in kidney disease development. On the left the healthy glomerulus displays physiological proportions of the four key glomerular cell-types: (1) PECs, (2) podocytes, (3) endothelial cells, and (4) mesangial cells. The graphic on the right displays the progression of PEC activation, specifically in terms of an increased proliferative state, where overlapping cascades of cells grow toward the glomerular tuft–forming crescents and pseudocrescents, depending on the patient's specific disease course. A number of pathways have been shown to participate in this process of PEC activation, including but not limited to: (1) Notch, (2) Mif/Cd74, (3) Cd44, (4) EGFR, (5) MCP-1/CCR2, (6) JAK/STAT, (7) SDF-1/CXCR4, (8) AngII/AT1, (9) Cd74, and (10) Cd9. These disease initiating/modifying factors are described at length, in earlier individual subheading on kidney pathophysiology. JAK/STAT, Janus kinase/signal transducer and activator of transcription; MCP-1, monocyte chemoattractant protein-1.

Notch

The Notch signaling pathway in mammals comprises four Notch receptors (notch 1–4) and a variety of ligands, including Jagged1–2 and delta-like 1, 3, and 4. Binding of a ligand to a Notch receptor promotes proteolytic cleavage, liberating the intracellular Notch domain which can act as a TF and regulate target gene expression (Figure 3).⁸⁶ A role for Notch signaling in PEC proliferation was first reported by Ueno et al. (Table 1).²⁸ They used a NEP25 mouse FSGS model in which human CD25 is specifically expressed in podocytes and after injection of anti-Tac (Fv)-PE38 (LMB2), which is an immunotoxin that specifically binds to human CD25, causes progressive injury.²⁸ This pathological transition presents with marked PEC proliferation and the group uncovered upregulation of several Notch pathway–related genes in glomerular hyperplastic lesions. Such proteins included Notch 1, Jagged1, and Hes1, and upregulation was confirmed at both the mRNA and protein level. Notch signaling is recognized for its role in promoting epithelial-to-mesenchymal transition and could potentially be driving a similar phenotypic switch during the activation of PECs.^87,88 The authors used TGFβ-1 treatment to activate Notch signaling in cultured mouse PECs and found that pharmacological inhibition of Notch signaling significantly impaired PEC migration. However, in the NEP25 mouse model of FSGS, treatment with the same Notch signaling inhibitor, dibenzazepine, appeared to reduce PEC lesions, but worsened proteinuria and histopathology.²⁸ These results suggest a cell-specific role for Notch signaling that may warrant further exploration. In addition, because treatment with the Notch inhibitor reduced the preponderance of PEC lesions yet worsened proteinuria, one could speculate that PEC activation/proliferation may in fact be protective at the micro level, by plugging the protein leak at the expense of limited glomerular function. In recent human cancer clinical trials, Notch antibodies, such as anti-NRR1 and brontictuzumab, have been effectively used, signaling a potential off-label use in proliferative glomerulopathies.^89,90

Epidermal Growth Factor Receptors

The epidermal growth factor receptor (EGFR) family is part of a larger class of receptor tyrosine kinases (Figure 3).⁹¹ Also referred to as ErbB receptors, four transmembrane receptors make up this family and include EGFR/ErbB1, ErbB2, ErbB3, and ErbB4, all of which share conserved protein domains mediating extracellular ligand binding, a single membrane spanning domain, tyrosine kinase domain, and a C-terminal tail domain which is subject to several means of post-translational modifications.⁹¹ Several different types of ligands are capable of binding ErbB receptors, promoting receptor dimerization, kinase activation, and signal propagation. Heparin-binding epidermal growth factor (HB-EGF), a ligand for ErbB receptors, is not expressed in glomeruli at baseline but is upregulated in the setting of glomerular injury, such as puromycin amino nucleoside nephropathy and passive Heymann nephritis.^42,92 In these models, in situ hybridization and immunohistochemistry confirmed expression of HB-EGF in adhesive lesions of the glomeruli and, particularly, in both podocytes and PECs. Similar findings have been reported in an anti-GBM antibody-induced nephritis model (Table 1).⁹³ Interestingly, expression of HB-EGF in glomeruli correlated with a decline in kidney function and pretreatment with an anti–HB-EGF antibody before anti-GBM antibody administration limited this decline.⁹³ However, these studies focused on short-term changes and did not explore effects on the development of crescentic lesions and PEC activation and proliferation. In a mouse model of crescentic GN using nephrotoxic serum (NTS), both global deletion of Hbegf and podocyte-specific deletion of Egfr improved survival and conferred protection against glomerular injury.⁹⁴ In addition, pharmacological inhibition of EGFR by treatment with erlotinib led to similar protection. However, the upstream signals promoting Hbegf expression in these models and how PEC-specific EGFR signaling contributes to disease progression remains unclear.⁹⁵ Nonetheless, Erlotinib has recently been shown to attenuate the progressive chronic kidney injury associated with 5/6 nephrectomy in Sprague Dawley rats through the inhibition of Akt and Erk1/2 signaling,⁹⁶ suggesting a potential therapeutic role in proliferative glomerulopathies.

Angiotensin II Receptor Type 1

It is well established that a dysregulation of the renin-angiotensin-aldosterone system (RAAS) is linked to kidney dysfunction, and as such, numerous drugs targeting this system are commonly used in patients with kidney disease. Binding of angiotensin II (AngII) to the angiotensin II type 1 (AT1) receptor can result in generation of cytokines, chemokines, reactive oxygen species, and adhesion molecules leading to altered cell migration and proliferation (Figure 3).⁹⁷ Glomerular upregulation of AngII has been identified in proteinuric models and attempts to target this pathway have shown promise, especially through the use of RAAS blockade, which exhibit both antiproteinuric and antiproliferative effects, the latter of which may be harnessed to block the downstream effects of PEC activation.^98,99 Using an angiotensin-converting enzyme (ACE) inhibitor to block this pathway, one study demonstrated a reduction in glomerular injury in a rat model of FSGS with PEC proliferation, in part due to upregulation of a cell cycle inhibitor.⁹⁹ Another study found that treatment with an AT1 receptor blocker (ARB) protected against glomerular injury and crescent formation in a similar rat model.¹⁰⁰ ARB treatment also reduced AngII-induced proliferation and collagen secretion in a coculture model (mesangial cells, PECs, and macrophages). Interestingly, they also showed that cotreatment with a CC-chemokine receptor 2 (CCR2) antagonist further enhanced these renoprotective effects, which suggests a concomitant role of monocyte chemoattractant protein-1/CCR2 signaling in crescent formation. These findings seem to be a potentially relevant therapeutic strategy in humans, given the increase in AT1 expression in activated PECs in human GN.¹⁰¹

Stromal Cell–Derived Factor 1/C-X-C Motif Chemokine Receptor 4

Stromal cell–derived factor 1 (SDF-1) (also known as C-X-C motif chemokine ligand 12) is a secreted ligand which binds C-X-C motif chemokine receptor 4 (CXCR4), a seven-pass G-protein–coupled transmembrane receptor (Figure 3).¹⁰² Interestingly, this signaling pathway regulates stem/cell progenitor cell trafficking through modulation of cell migration, chemotaxis, and adhesion.¹⁰² In the kidney, CXCR4 upregulation has been observed in crescentic lesions with simultaneous induction of SDF-1 in human and rat podocytes.¹⁰¹ Because treatment with an ACE inhibitor normalized CXCR4 protein levels in a rat model of crescentic GN, it is possible that there is cooperation between AngII/At1 and SDF-1/CXCR4 signaling. In addition, podocyte-specific deletion of the Von Hippel-Lindau (Vhlh) gene leads to formation of RPGN in mice with upregulation of glomerular CXCR4.¹⁰³ Similarly, overexpression of Cxcr4 in rat podocytes alone is sufficient to induce glomerular injury with crescentic lesions. A recent study reported that upregulated SDF-1 was first observed in injured podocytes and then in activated PECs.¹⁰⁴ In cultured mouse PECs, recombinant SDF-1 increased Cd44 and Cxcr4 at the mRNA level and promoted migration, which was attenuated by small‐interfering RNA -mediated Cd44 knockdown. However, in a mouse model of proliferative GN, treatment with an SDF-1 inhibitor failed to show any improvements in kidney function and had no effect on CD44 expression. Nonetheless, the authors suggest that sufficient amount of inhibitor may not be reaching the podocytes and PECs to show a protective effect. Regardless, the SDF-1/CXCR4 signaling axis remains of interest with respect to PEC proliferation and potential means of podocyte-PEC interdependence, deserving of more attention in future studies.

Janus Kinase/Signal Transducer and Activator of Transcription

The Janus kinase/signal transducer and activator of transcription signaling pathway regulates cellular response to different signaling ligands and is increasingly recognized in recent years for its importance in kidney disease (Figure 3).¹⁰⁵ Although there are many different signaling ligands and receptors in this pathway, canonical activation of this pathway includes IL-6-type cytokines binding to gp130 receptors and triggering Janus kinase activation. STAT3 is a central member of this signaling pathway and transcriptionally regulates numerous downstream target genes involved in proliferation, migration, and inflammation.¹⁰⁶ Aberrant STAT3 activation occurs in various human kidney diseases and kidney disease models, including collapsing FSGS and NTS-induced nephritis.^105,107,108 Although podocyte-specific deletion of Stat3 is protective in NTS-induced nephritis and global Stat3 deletion has also conferred protection in a murine model of HIV-1–associated nephropathy, studies aimed at investigating the role of PEC-specific Stat3 have not been performed.^109,110 In addition, we previously reported that podocyte-specific deletion of Kruppel-like factor 4 (Klf4) exacerbates podocyte loss, PEC activation, and proliferation, leading to FSGS after NTS treatment.⁸³ In these injured glomeruli, STAT3 signaling was activated in both podocytes and activated PECs, and treatment with a small-molecule inhibitor of STAT3 reduced proteinuria. Interestingly, we postulated that key ligand-receptor interactions, such as fibronectin 1 (FN1)–αVβ6, between podocytes and PECs contributed to PEC activation in the setting of Klf4 knockdown specifically in podocytes.¹¹¹ Finally, our group has recently demonstrated that the inhibition of STAT3 in these podocyte-specific Klf4 knockout mice significantly abrogated podocyte loss and albuminuria.¹¹¹ In addition, when KLF4 is induced specifically in podocytes, we observed a renoprotective effect not only in terms of a reduction in STAT3 activation and podocyte injury but also in the FN1–αVβ6 interaction and PEC activation.¹¹¹ These studies highlight the potential therapeutic utility of deploying STAT3 inhibitors as well as targeting FN1–αVβ6 interaction in the treatment of proliferative glomerulopathies. Although there are a number of pharmacological inhibitors of STAT3 signaling that have been developed in recent years for multiple diseases (various cancers and rheumatoid arthritis), concerns of specificity and systemic toxicity have limited their potential use in kidney disease.^111,112 Nonetheless, repurposing of small molecules targeting the SH2 and DNA binding domains of STAT3 and the use of antisense oligonucleotide technology (AZD9150) might serve as potential viable approaches to target STAT3 signaling in proliferative glomerulopathies.^111,113

Macrophage Migration Inhibitory Factor/Cluster of Differentiation 74

Macrophage migration inhibitory factor (MIF) is typically known for its role in immunity and regulating immune responses (Figure 3).¹¹⁴ It is known to bind several target receptors, including CXCR2, CXCR4, and cluster of differentiation 74 (CD74), leading to alterations in expression of chemokines, cytokines, and adhesion molecules and changes in proliferation and cell survival.^115,116 In the healthy human kidney, MIF is expressed is tubular epithelial cells and some glomerular cells.¹¹⁷ In a rat model of crescentic GN, MIF is expressed de novo in both podocytes and PECs.¹¹⁷ For CD74, de novo expression in mesangial and PECs has been reported in both humans and mice with proliferative GN (NTS mouse model).¹¹⁸ NTS treatment upregulated MIF secretion from cultured podocytes and recombinant MIF stimulated proliferation of both mesangial cells and PECs in-vitro, which was attenuated by coincubation with anti-CD74 antibody. Although both global Mif or Cd74 deletion¹¹⁹ protected mice from crescentic GN, it remains unknown what other cell types are contributing to these salutary effects because of the lack of cell specificity of the deletion. Future studies may determine whether similar protection can be achieved through PEC-specific Cd74 deletion or podocyte-specific Mif deletion. Such information may be vital to thoroughly map out MIF/Cd74 signaling. Previous studies also suggest that MIF-CD74 signaling might require CD44, which suggests that other concurrent signals are required to induce CD44 expression during initial PEC activation.¹²⁰

Wnt/β-Catenin

The Wnt/β-catenin signaling pathway is widely studied for its role in regulating fundamental processes at both the cellular and organ level. Canonical Wnt signaling is centered on regulation of the transcriptional activity of β-catenin, which regulates several target genes, including c-Myc, Vegf, Ctgf, and cyclin-D1.¹²¹ In glomeruli, Wnt signaling is known for its regulation of developmental pathways, with reported roles in regulating podocyte dedifferentiation and PEC activation.^122,123 A critical early study of Wnt signaling in kidney disease demonstrated the first established link between this pathway and a genuine murine recapitulation of collapsing FSGS with prominent PEC proliferation.⁴⁷ However, the detrimental role for Wnt signaling has also been demonstrated in a mouse model of diabetic kidney disease, in the absence of PEC proliferation, suggesting a degree of nonspecificity.¹²⁴ Consequently, more work is required to investigate the specific mechanisms mediating Wnt/β-catenin signaling in models of proliferative glomerulopathy.

Cluster of Differentiation 9

Cluster of differentiation 9 (CD9) is a tetraspanin protein that has recently been identified as a key regulator of PEC activation (Figure 3).¹²⁵ As a tetraspanin, CD9 has four transmembrane domains and regulates various cellular processes through formation of tetraspanin-enriched microdomains, spatially regulating function of adhesion receptors, growth factor receptors, and more.^126,127 According to Lazareth et al., CD9 is expressed de novo in PECs during crescentic GN in both mice and humans. Temporally, Cd9 was induced early in crescentic GN and before induction of Cd44. PEC-specific Cd9 was shown as a key driver of crescentic GN progression. Interestingly, they also report a link between Cd9 and Itgβ1, EGFR, and platelet‐derived growth factor receptor, suggesting specific involvement of Cd9 in regulating proliferative and migratory responses. They also demonstrated that knockdown of Cd9 in cultured mouse PECs inhibited CD44, thereby demonstrating that CD9 might be the trigger for PEC activation. However, the question remains: What regulates CD9 expression? A recent report documents that mechanical stress may induce CD9 expression, but further investigation is warranted.¹²⁸

Limitations to Consider

Although many of these pathways have been shown to be important for PEC activation, several limitations and questions remain. A limitation to remember is that many of these studies may not be able to entirely recapitulate the true nature of human PECs because they are conducted in rodent models. It is clear that human PECs are distinct from mouse PECs, including the expression of CD133 and CD24, two stem cell markers which are unique to humans.^55,129 Other genetic differences between the two should not be ignored. Another layer of complexity is added because of emerging evidence to suggest considerable heterogeneity among PEC subtypes between mice and humans.^53,130 Such factors should be considered in future investigations. Arguably the biggest question that remains is a lack of understanding of the upstream signals that serve as the initial triggers for PEC activation in the setting of podocyte injury.

Emerging Technologies

Although key questions surrounding PEC biology remain unanswered, there are a number of new experimental and analytical tools which hold great promise in addressing these knowledge gaps. Perhaps the most powerful of these new approaches is single-cell RNA sequencing (scRNA-seq), which has for the first time allowed the identification of the wealth of unique cell types present in complex tissues, such as the kidney.^61,131–135 A recent study from our group that investigated the molecular pathophysiology of podocyte-specific Klf4 knock-down used single nucleus RNA-seq, demonstrating that specific PEC subtypes can be identified in an unbiased manner.¹¹¹ These single-cell studies encompass both development and specific adult disease states, such as diabetic nephropathy and other CKDs.^61,133,135 In addition, given the sparse nature of PECs in uninjured glomeruli, two technologies present themselves as ideal additions to scRNA-seq: spatial transcriptomics and Cleavage Under Targets and Release Using Nuclease (CUT&RUN). In regard to the first, unique molecular identifiers (cellular barcodes) can be applied to cells in kidney sections that still maintain their relative positional information, providing a reference map for rare cell type (i.e., PEC) distribution, followed by next-generation sequencing of single-cell libraries. CUT&RUN can be used on low input samples (hundreds of cells) to ascertain information about the binding and local chromatin modifying properties of specific TFs potentially implicated in the scRNA-seq and/or spatial transcriptomic studies. Such information can be enormously informative when it comes to unraveling the genomic regulatory properties of cell-type specific TFs and would normally require millions of input cells (i.e., chromatin immunoprecipitation, ChIP-seq) which is a limiting factor in studies of PECs in-vivo. To specifically get at the heart of PEC/podocyte interdependence, the field can turn to a number of approaches borrowed from the field of neurobiology, specifically, growing both primary cell types under tissue culture conditions in modified Campenot chambers (devised by Robert Campenot to study isolated neuronal compartments in cultured neurons).¹³⁶ Given the PEC/podocyte bridges mentioned under a previous review heading, such devices may prove useful in isolating paracrine and physical mediators important for glomerular development and disease.¹³⁶ Finally, there are a number of online repositories housing kidney sequencing data in the context of specific kidney pathologies (e.g., Kidney Precision Medicine Project, Broad Institute Single Cell Portal); however, specific data outlining PEC gene expression and chromatin dynamics in the context of proliferative glomerulopathies are largely absent. Although some of the emerging technologies described herein have been instrumental in decoding the gene regulatory logic in other kidney diseases, to most effectively address the limitations above, a larger emphasis has to be made on exploring the gene expression and chromatin substructure of mouse models of RPGN and cellular/collapsing FSGS and their human antecedents.

Disclosures

Y. Gowthaman reports Research Funding: NIH/NIDDK and Veterans Affairs. S.K. Mallipattu reports Consultancy: L.E.K. Consulting and Wildwood Therapeutics, Inc.; Research Funding: Dialysis Clinic Inc. and Spectral Medical Inc.; Patents: Krüppel-like factor 15 (KLF15) Small Molecule Agonists in Kidney Disease, US 63/018.247, April 30, 2021; and Advisory or Leadership Role: Clinically Integrated Network Board Member (Accountable Care Organization, LLC Stony Brook Medicine). D.J. Salant reports Consultancy: Advance Medical, Pfizer, UpToDate, and Visterra; Research Funding: NIH; Honoraria: Several academic institutions and national societies; Patents or Royalties: Patent: “Diagnostics in membranous nephropathy”–Boston Medical Center; Advisory or Leadership Role: Editorial board: American Journal of Physiology and JASN and Scientific Advisory Board: NEPTUNE; and Other Interests or Relationships: National Kidney Foundation Medical Advisory Board. The remaining authors have nothing to disclose.

Funding

This work was supported by funds from NIH/NIDDK (DK112984, DK121846), Veterans Affairs (I01BX003698, I01BX005300), and Dialysis Clinic Inc. to S.K. Mallipattu.

Author Contributions

Conceptualization: Robert Bronstein, Sandeep K. Mallipattu, Jesse Pace, David J. Salant.

Data curation: Robert Bronstein, Sandeep K. Mallipattu, Jesse Pace.

Formal analysis: Robert Bronstein, Sandeep K. Mallipattu, Jesse Pace.

Funding acquisition: Jesse Pace.

Investigation: Robert Bronstein, Yogesh Gowthaman, Sandeep K. Mallipattu, Jesse Pace.

Methodology: Robert Bronstein, Yogesh Gowthaman, Sandeep K. Mallipattu, Jesse Pace.

Project administration: Sandeep K. Mallipattu, Jesse Pace.

Supervision: Sandeep K. Mallipattu, David J. Salant.

Validation: Robert Bronstein, Sandeep K. Mallipattu, Jesse Pace, David J. Salant.

Visualization: Robert Bronstein, Yogesh Gowthaman, Sandeep K. Mallipattu, Jesse Pace.

Writing – original draft: Robert Bronstein, Yogesh Gowthaman, Sandeep K. Mallipattu, Jesse Pace.

Writing – review & editing: Robert Bronstein, Yogesh Gowthaman, Sandeep K. Mallipattu, Jesse Pace, David J. Salant.