Parietal epithelial cells (PECs) are resident kidney cells that line the inner Bowman’s capsule [1]. They originate from the metanephric mesenchyme and undergo a mesenchymal-to-epithelial transition during early nephrogenesis, similar to podocytes. Recently, in the 21st century, increasing evidence suggests that PECs have crucial functions in health and disease ^[2][3][4][5][2,3,4,5]. Podocyte loss is known to drive glomerulosclerosis [6], a unifying mechanism for end-stage kidney disease in all glomerular diseases. In addition to podocyte loss, PEC activation is now known to be associated with severe glomerular diseases. Lineage tracking research has shown that PECs comprise extracapillary proliferation in animal models of podocytopathies and crescentic glomerulonephritis (CrescGN) ^{[7][8][9][10][11]}[7,8,9,10,11]. Activated PECs are shown to participate in focal segmental glomerulosclerosis (FSGS) lesions [12], the most common primary glomerulopathy leading to end-stage kidney disease [13]. PEC activation may link podocyte loss and progressive glomerulosclerosis. Although the evidence on this pathogenic role of PECs ^{[7][8][9][10][11][12]}[7,8,9,10,11,12] seems contradictory to their proposed reparative role ^[2][3][4][5][2,3,4,5], these may represent two sides of PEC behavior based on increased or decreased adverse or conducive milieus and mediators. Recent research on regulators of gene expression has signaled the role of microRNA-193a (miR193a) as a mediator in the pathogenesis of FSGS and CrescGN ^{[14][15][16][17][18]}[14,15,16,17,18]. The modulation of miR193a has been reported to alter PEC behavior in vitro [16]. Further research to understand the effect of miR193a or its downstream mediators on PECs, especially animal models, will provide insight into this connection.

2. PECs and Their Phenotypes

PECs are attenuated squamous cells with a condensed nucleus and express specific markers such as paired box gene (PAX) 2, PAX8, annexin A3, Claudin-1/2/16, and Ubiquitin C-terminal hydrolase-L1 (UCH-L1) [8]. Unlike podocytes, PECs proliferate at a low level throughout life [19].

2.1. Types of PECs in Human Kidneys

2.1.1. Based on Location and Expression of Molecular Markers

Bariety et al. [20] used the term “parietal podocytes” for the PECs at the vascular pole of glomeruli that co-express PAX2 (PEC marker) and WT1 (+other podocyte markers) and “other PECs” for those located away from the vascular pole that expressed cytokeratin without podocyte proteins. The origin and function of parietal podocytes are not yet clear. They are similar to visceral podocytes in every aspect except the presence of PAX2, which implies their ability to divide. Ronconi et al. [3] have classified PECs based on stem cell markers (CD133/CD24) and podocyte marker podocalyxin (PDX) expression: the PECs at the vascular pole (CD133-CD24-PDX+) are differentiated cells resembling podocytes, the PECs in between the vascular and tubular poles (CD133+CD24+PDX+) are transitional cells, i.e., podocyte-committed progenitors with a low regenerative potential, and the PECs at the tubular pole are progenitor cells (also called “adult parietal multipotent progenitors” expressing CD133+CD24+PDX-) with a high regenerative potential. The progenitor cells at the tubular pole express only the stem cell markers and not the podocyte markers, however, they were capable of transforming into both podocytes and proximal tubular cells in vitro and in vivo. Hence, they were called bipotent progenitors. These bipotent progenitor cells were shown to reduce albuminuria and glomerular damage when injected into diseased mice.

2.1.2. Based on Morphology

Kuppe et al. [21] have described PECs as flat, intermediate, and cuboidal. Cuboidal PECs have brush border-mimicking proximal tubular cells and line the Bowman’s capsule over the tubular pole. Intermediate PECs express the phenotype of injured proximal tubule cells that are trying to regenerate, i.e., they do not have a brush border, are triangular in shape with less cytoplasm, and express markers such as keratin 7 and keratin 19. They are seen at the junction between the cuboidal PECs and flat PECs in glomerular disease. Intermediate and cuboidal PECs had a higher activation potential, i.e., they preferentially express CD44 and Ki-67, and were shown to be involved in sclerosis/crescent formation in mouse models of FSGS and CrescGN, respectively, whereas flat PECs had lower activation potential. Intermediate PECs were shown to be the predominant PEC subtype forming synechiae between the glomerular tuft and Bowman’s capsule in human kidney biopsies of early FSGS recurrences that predominantly had tip lesions. Intermediate PECs were less and flat PECs were more abundant in kidney biopsies with FSGS not otherwise specified (NOS). Whether the abundance of intermediate PECs in tip lesion and the abundance of flat PECs in FSGS-NOS reflects different stages of injury and results in favorable prognosis of tip lesion remains to be studied.

3. Are PECs Reparative?

Podocytes are lost in the urine under normal physiological conditions [5]. The continuous replacement of the lost podocytes possibly prevents proteinuria from developing. Clinical reports of the reversal of diabetic nephropathy in humans [22] and animal models ^[23][24][23,24] suggest podocyte regeneration under diseased conditions. PECs were observed to acquire podocyte phenotype and markers when cultured in conducive media containing vitamin D3, retinoic acid, and dexamethasone [3]. The evidence on “Are PECs the source of podocyte renewal during growth and during disease?” is discussed in the following section.

3.1. Are PECs Progenitors for Podocytes during Postnatal Glomerular Growth?

PECs and podocytes share a common progenitor [25]. Puelles et al. [26] observed healthy human adults with larger glomeruli to have more podocytes than children, reflecting podocyte gain after birth. They noted the PECs co-expressing stem cell markers (CD24, CD133) and podocyte markers (p57, synaptopodin) near the vascular pole in both human children (≤ 3 years) and young adult glomeruli. Appel et al. [27] observed the PECs to replace the podocytes in adolescent mice during growth. The progenitor PECs express stem cell markers CD133 and CD24. The trans-differentiation of PECs to podocytes was shown using lineage tracking of PECs in rats followed postnatally on day 5 or 10 when nephrogenesis was complete. The PECs were observed to be increased over the glomerular tuft and all the PECs co-expressed podocyte markers (podocin, nephrin, and p57) along with the PEC activation marker (CD44) [27]. Berger et al. [28] showed recruitment of podocytes from PECs during normal growth until infancy in mice and in a limited number of human kidney biopsies. These studies suggest that podocytes are generated postnatally and that PECs are the likely source in animal models; however, it is still not clear if that holds true in humans.

3.2. Are PECs Progenitors for Podocytes during Aging?

Podocyte regeneration was not observed in aging mice kidneys, which showed progressive loss of podocytes and the development of glomerulosclerosis [29]. Similarly, Berger et al. did not observe podocyte renewal from PECs in aging mice [27]. Nevertheless, Kaverina et al. have recently documented PEC migration and differentiation to podocytes in glomeruli with low podocyte count in otherwise healthy but aged mice [30]. It is not yet known if PECs that differentiate into podocytes can function as healthy podocytes and reverse aging-induced kidney changes.

3.3. Are PECs Progenitors for Podocytes during Kidney Disease?

In a mouse model of partial acute podocyte depletion, genetically labeled PECs were observed to give rise to differentiated cells resembling podocytes after four weeks [29]. This finding coincided with proteinuria reduction in the mice [29]. In a mouse model of FSGS caused by acute podocyte depletion, genetically labeled PECs were shown to migrate to the affected glomerular tuft [31]. On day 28, the same PECs were noted to co-express podocyte markers, correlating with increased podocyte number and reduced glomerulosclerosis severity. The authors later reported [32] that PECs started increasingly co-expressing podocyte markers on day 28 and day 56 after injury. Lasagni et al. [33] noted that renal progenitor cells lining Bowman’s capsule differentiate into podocytes through intraperitoneal retinoic acid, leading to disease remission in a mouse model of FSGS. Ronconi et al. [3] showed that injection of progenitor PECs in mice with adriamycin-induced nephropathy reduced albuminuria and chronic glomerular damage. While Eng and colleagues ^[31][32][31,32] provide strong evidence of the ability of PECs to differentiate into podocytes in adult mice with acute podocyte depletion, contemporary researchers argue that the complexity of transgenic tagging of PECs and podocytes might have confounded their analysis [34]. Berger et al. [28] noted that PECs did not differentiate into podocytes in response to nephron loss induced by progressive partial nephrectomies in mice. Similarly, Wanner et al. [29] observed that nephron loss causing glomerular hypertrophy was not followed by podocyte renewal. Therefore, PECs did not contribute to podocyte regeneration after gradual nephron loss. The source of podocyte renewal during the recovery of human glomerular diseases is not confirmed to be PECs. Therefore, evidence on the reparative role of PECs after podocyte injury remains controversial.

4. Are PECs Pathogenic?

4.1. PEC Activation in Classical FSGS and Collapsing Glomerulopathy

The PECs respond rapidly to extensive podocyte loss by activating themselves to migrate and secrete an extracellular matrix to seal the injured GBM. In an experimental mouse model, the severity of podocyte loss was correlated with the degree of proliferation of PECs [35]. Although various molecules are associated with PEC activation, no single cellular process has been confirmed to drive their trigger [36]. Activated PECs express CD44 and secrete the extracellular matrix of the Bowman’s capsule type, which is different from that present in the GBM. Excessive PEC activation is associated with the progression of glomerular diseases ^{[7][8][9][10][11]}[7,8,9,10,11]. In genetically labeled PEC reporter mice with FSGS, the migration of PECs towards the injured glomerular tuft was noted, and the intensity of CD44 expression by PECs was positively related to the severity of glomerulosclerosis ^[27][31][27,31]. Furthermore, activated PECs formed the cellular adhesions between the Bowman’s capsule and the glomerular tuft in human FSGS [21]. Increased CD44 expression by PECs on the glomerular tuft has been shown to correlate with poor clinical kidney function in children with FSGS [37]. Benigni et al. [38] reported the ability of angiotensin-converting enzyme inhibitors (ACEi) to normalize glomerular lesions by blocking PEC activation in a rat model of glomerulosclerosis. These studies suggest that PEC activation, marked by CD44 expression, is associated with more severe glomerulosclerosis in FSGS. The degree of PEC proliferation correlates with the severity of podocyte injury [35]. The massive (perhaps acute) loss of podocytes can invoke extensive PEC proliferation, resulting in the formation of layers of cells (pseudo-crescents) and consequent glomerular tuft collapse, termed collapsing glomerulopathy (CG) [39]. While its categorization within the ambit of FSGS is controversial ^[40][41][40,41], typical CG is associated with podocyte loss and massive albuminuria. CG typically occurs in the setting of the human immunodeficiency virus (HIV) infection in patients with APOL1 high-risk alleles (G1 and G2) [42]. The normal APOL1G0 allele is expressed in podocytes but not PECs [18] and is speculated to be protective against HIV-associated nephropathy ^[43][44][43,44]. APOL1G0 facilitates PEC transition to podocytes [18] in vitro, and its association with miR193a is described below. Interestingly, HIV has been demonstrated to induce APOL1G0 expression in PECs and, therefore, increase PEC transition to podocytes in vitro [18]. In contrast, HIV infection is also known to enhance the expression of APOL1 risk alleles [45] and trigger FSGS development by causing podocyte de-differentiation in vivo [46]. Moreover, APOL1 risk alleles are observed to cause altered podocyte function [47]. Thus, CG likely develops in HIV positive patients carrying APOL1 risk alleles when the following events occur simultaneously: (i). massive podocyte loss (direct HIV-induced and APOL1G1/G2 risk allele-induced), (ii). PEC activation following podocyte injury, and (iii). compromised podocyte renewal by PECs (due to lack of APOL1G0). On the other hand, in the pathogenesis of classical FSGS, only the loss of a critical number of podocytes acts as the trigger for activating adjacent PECs that, based on the severity of podocyte injury and the local environment, result in either progressive sclerosis (deposition of excessive extracellular matrix) or the repair of the glomerular filtration barrier (podocyte renewal and sealing of the denudation).

4.2. PEC Activation in Crescentic GN

CrescGN is characterized by extracapillary proliferation in the Bowman’s space in response to a breach in GBM and the resultant flow of inflammatory mediators, such as immune cells and phlogogenic plasma [10]. Activated PECs are the predominant proliferating cells in the mouse nephrotoxic nephritis model of inflammatory CrescGN ^[7][36][7,36]. These activated PECs are a population of CD133+CD24+ progenitor cells, as shown in human kidney biopsies [11]. De novo CD9 expression by PECs is a prerequisite for CD44 expression and activation of PECs [48]. C-X-C Motif Chemokine Receptor 4 (CXCR4) and angiotensin II/angiotensin II type-1 (AT-1) receptor pathways are reported to be instrumental in activating PECs to form crescents ^[49][50][49,50]. Treatment with ACEi resulted in the normalization of CXCR4 staining in the animal model of CrescGN [49]. Further, CXCR2 and its ligand are expressed in PECs and podocytes, respectively, in experimental FSGS [51]. Therefore, activated PECs, characterized by chemokine receptor expression, and podocytes, marked by chemokine receptor ligand expression, appear interrelated. Hence, podocyte injury could trigger PEC activation in CrescGN, similar to that seen in FSGS. Although activated PECs in both human FSGS and CrescGN show a similar staining pattern, CD44, claudin-1, CD24, and CD133 [36], the manner of PEC activation is different. Whereas adhesions of glomerular capillary tufts to the Bowman’s Capsule are early features in FSGS [21], they are not prerequisites for forming cellular crescents [52]. PECs are mainly proliferative in CrescGN in response to an inflammatory stimulus after GBM rupture. However, PECs primarily acquire a migratory and secretory phenotype in FSGS in response to podocyte loss. Interestingly, CG shares its phenotype with sclerotic lesions in FSGS and extracapillary proliferation in CrescGN. They are differentiated based on the phenotype of the adjacent glomeruli in the same section, a breach of the GBM and presence of immune cells in the Bowman’s space in CrescGN, and a collapse of the capillary loops without breach of GBM and absence of immune cells in CG. Nonetheless, distinguishing severe CG with circumferential pseudo-crescent formation from severe CrescGN with compression of the glomerular tuft (resembling collapse) can sometimes be difficult on morphology alone. The molecular basis of PEC activation in these diseases might be different.