The glomerular filtration barrier is composed of fenestrated endothelial cells, the glomerular basement membrane (GBM), and podocytes, terminally differentiated epithelial cells connected by interposed slit diaphragms between actin-rich foot processes (1). Seminal studies by Farquhar and Palade highlight the role of the GBM in excluding molecules of the size and negative charge of albumin from the urine under normal conditions (2). Though proteinuria can result from injury to any of these components, experimental nephrosis was initially described as mainly affecting the visceral epithelium with loss of podocyte foot processes, reduction and modification of the “urinary slits” (interposed slit diaphragm), and intracellular accumulation of vacuoles and protein absorption droplets (3, 4). Positional cloning of the NPHS1 gene encoding nephrin, a vital component of the slit diaphragm, was a landmark event that initiated the era of molecular discovery of the complex molecular architecture of the podocyte (5). Wiggins et al. performed elegant studies showing that in a diphtheria toxin rat model, podocyte loss of more than 20% results in a low-level-sustained proteinuria. Loss greater than 40% directly causes a high-level proteinuria, a decreased renal function, and lesions of focal segmental glomerulosclerosis (FSGS) (6). Podocyte loss triggers a cascade involving tuft adhesions of bare areas of GBM to Bowman’s capsule, a nidus for sclerosis development that can progress to segmental sclerosis (7). Rather than a disease, FSGS should be considered a histologic pattern of glomerular injury resulting from podocyte loss. It describes not only primary podocyte injury but also a lesion that occurs as a secondary process such as hypertensive and diabetic nephropathy (DN). Clinically and genetically heterogeneous, FSGS is characterized by segmental sclerosis of the glomerular capillary tuft, with or without deposition of IgM and complement C3 (8). The pathogenesis of FSGS is multifactorial, thereby complicating classification efforts and therapeutic approaches. The contribution of circulating permeability factors has been suggested for four decades with the recurrence of FSGS after kidney transplantation potentially occurring within hours (9, 10). Separately, many genes identified in patients with familial and sporadic FSGS encode essential podocyte proteins (11). Viral nephritides, HIV in particular, has been associated with collapsing FSGS (12).

This review discusses the knowledge to date of the underlying pathogenesis of podocyte injury leading to proteinuria and FSGS. We focus on the contributions of putative circulating factors, gene mutations, and viral infections.

The high recurrence rate of FSGS post transplant (estimated at 30–40%) suggests the presence of a circulating factor in recipients. A pivotal study by Savin et al. showed that plasma from patients with recurrent FSGS could increase the glomerular permeability to albumin in an in vitro assay. Plasmapheresis was associated with a reduced proteinuria as well as a reduced glomerular permeability (13). Gallon presented an interesting case of a kidney-transplant patient with recurrent FSGS (14). The allograft kidney regained function with a reduction in proteinuria after the graft was removed and re-transplanted to another patient whose ESRD was due to diabetes mellitus. In addition, an infant born to a mother with primary FSGS developed transient proteinuria, suggesting a circulating factor crossing the placenta leading to proteinuria (15). These studies well illustrate the existence of a circulating factor, but the nature and source of such a factor remained unclear. Several putative factors have been proposed including soluble urokinase-type plasminogen activator receptor (suPAR), CLC-1, Apolipoprotein A1, active protease, and anti-CD40.

Beginning with nephrin, the positional cloning of patients and relatives with familial FSGS has enabled the identification of numerous disease-causing podocyte genes (Table 1). They encode podocyte proteins localized to the cell membrane (TRPC6), nucleus (WT1), mitochondria (COQ2 and COQ6), lysosomes (LIMP2), and cytosol (PLCE1) (34–43). The most common mutations encode actin cytoskeletal (INF2, ACTN4, MYO1E) and slit diaphragm (NPHS1, NPHS2, CD2AP) (5, 44–49) proteins, thereby highlighting the essential role of these structures in the maintenance of the glomerular filtration barrier. Mutations in the ACTN4 gene are all in the actin-binding domain of the encoded protein and have a much higher affinity for actin filaments than of the wild-type protein (50). This gain of function produces a rigid cytoskeleton more susceptible to stress and actin network breaking. INF2, a member of the formin family, also regulates the actin cytoskeleton but in a different way. INF2 variants occur in the diaphanous inhibitory domain that is essential in the inhibition of Rho activation (51). Unchecked Rho signaling has therefore been suggested as a potential etiology for podocyte injury in INF2 mutation (51). Indeed, the activation of RhoA in podocytes has been shown to lead to albuminuria, podocyte foot process effacement, and histologic lesions of FSGS (52).

The Rho family small GTPases (RhoA, Rac1, and Cdc42) are essential in actin cytoskeletal dynamics, cell morphology, motility, and adhesion. While overactive RhoA can be deleterious, there is an emerging consensus that a relative predominance of RhoA activity relative to Rac1/Cdc42 favors a more stationary podocyte with intact foot processes (53, 54). Conversely, Cdc42/Rac1 activation is associated with a more motile podocyte phenotype and foot process retraction. Loss of function mutations in the Arhgap24 gene associated with familial FSGS cause increased levels of active Rac1 and Cdc42 and increased podocyte motility (55). Likewise, ARHGDIA mutations in patients with steroid-resistant nephrotic syndrome increase active GTP-bound Rac1 and Cdc42 again resulting in increased podocyte motility reversed with Rac1 inhibitors (56). The slit diaphragm is also intricately connected to podocyte actin cytoskeletal dynamics. CD2AP and nephrin are direct binding partners with CD2AP interacting with actin, cortactin, and the alpha-actinin modulating protein synaptopodin (57, 58). Nck adaptor proteins link nephrin to the actin cytoskeleton (59). The cytoplasmic domain of nephrin contains six conserved tyrosine residues that when phosphorylated facilitates binding to the SH2 domains of Nck, leading to actin polymerization (60). Upregulated nephrin tyrosine phosphorylation has been described in glomerular injury (61), but the involved tyrosine kinase remains undefined.

The onset of proteinuria is variable in genetic disease with autosomal-recessive etiologies generally manifesting clinically in childhood and autosomal-dominant traits in adulthood. Commercially available genetic testing is most likely to identify known disease-causing variants in infants and patients with familial and syndromic disease (62). Exomic and genomic sequencing have the potential to uncover rare novel gene variants even in sporadic disease, but confirming causation typically requires experimental modeling with cell- and animal-based models. Identification of disease-causing gene mutations in FSGS has prognostic consequences since these patients are less likely to respond to steroid and calcineurin inhibitor therapy (63). They should perhaps be prioritized for kidney transplantation since disease recurrence is lower (64). Caution is necessary in the clinical application of genetic data, however. It is estimated that only 2.9% of US patients with nephrotic syndrome have a monogenic form of the disease and there is a risk in attributing a pathogenic role to identified variants that could be noncausal (65, 66).

Podocytes can be injured by viruses either directly or by inflammatory cytokine-mediated targeting of podocyte receptors. In this context, HIV has been best studied. The renal expression of HIV genes has long been recognized as a central role in promoting HIVAN pathogenesis (79). HIV infects podocytes, tubular epithelial cells, infiltrating lymphocytes, as well as macrophages. Early studies of animal models, including the classic transgenic Tg26 mouse lines, studies on reciprocal transplantation between Tg26 and wild-type mice as well as podocyte-specific expression of HIV genes have confirmed the pathogenic role for local HIV gene expression in the kidney (80). Podocyte-predominant infection has been associated with podocyte injury, dedifferentiation, and more rapid loss of kidney function, suggestive of an HIV-driven podocyte injury as a primary pathogenic pathway in HIVAN (81). Moreover, the kidney has been shown to be one of the reservoirs for HIV, allowing active viral replication in this compartment apart from blood (82, 83). In HIVAN and collapsing forms of primary FSGS, injured podocytes were found to revert to a developmental program that includes downregulation of cyclin kinase inhibitors, entry into the cell cycle, upregulation of proliferation maker, and loss of mature phenotypic makers, including CD10/CALLA, C3b receptor, GLEPP-1, podocalyxin, synaptopodin, and importantly, WT1 (84–88). This process, termed podocyte dysregulation or podocyte dedifferentiation, is not observed in other proteinuric kidney disease marked by podocyte injury along with foot process effacement, such as minimal change disease and membranous nephropathy.

Cip/Kip family of cyclin kinase inhibitors acts in both G1 and S phases, while p21 also inhibits G2/M phase complexes. In healthy podocytes, p27 and p57 are expressed but p21 is suppressed (85). However, in patients with HIVAN or idiopathic collapsing FSGS, the expression of p27 and p57 is significantly suppressed, along with an increase in pro-mitotic cyclins (cyclin A), an increase in proliferation marker ki-67, and an expression of p21 (84, 85). Interestingly, the decrease in p27 and p57 was also found in histologically normal glomeruli, suggestive of a decrease of anti-mitotic signals preceding morphological change. Moreover, administering cyclin kinase inhibitors is associated with a restoration of normal podocyte phenotype and attenuated HIVAN presentation without any suppression of HIV gene expression (89, 90). Podocyte dedifferentiation is accompanied by the proliferation of parietal epithelial cells (PECs), which express parietal cell marker such as CK8 and PAX2 and lack podocyte markers. Cell bridges to the CK8-positive parietal lining could be observed without any evidence of coexpression of podocyte markers. These findings suggest that glomerular PECs provide a niche for podocyte progenitor cells (12, 91, 92).

Novel mechanisms of HIV-mediated kidney injury have been recently uncovered. Mammalian target of rapamycin (mTOR) has been found to be critical for p53-induced oxidative cell injury with mTOR inhibition protecting against HIV-induced podocyte apoptosis (93). The protective function of mTOR inhibition in HIVAN could be due to the regulation of microRNAs since rapamycin treatment reverses the downregulation of miR99a, miR-100a, miR-199a, miR-200a, miR-200b, miR-200c, miR-429, and miR-141 caused by HIV infection of human podocytes in culture (94). HIV also compromises the podocyte actin cytoskeleton through downregulation of the vitamin D receptor with associated enhanced deleterious intracellular angiotensin II and cathepsin L expression (95).

Low birth weight individuals have long been known to be susceptible to a reduced nephron number and resulting abnormal hemodynamic stress on remaining nephrons (102, 103). Low birth weight is more commonly seen in FSGS patients than in the general population (104). Proteinuria has also been recognized in patients with unilateral renal agenesis where the remaining kidney has a reduced nephron number (105). A decreased nephron mass results in glomerular hypertension and hyperfiltration that increase mechanical stretch and injury to podocytes (106, 107). Glomerular hyperfiltration has also been associated with the pathogenesis of obesity-related FSGS (108). Here as with a decreased nephron mass, the activation of the renin–angiotensin–aldosterone system, the upregulation of TGFβ, and glomerulomegaly contribute to the vicious cycle that leads to proteinuria and FSGS (109, 110). Angiotensin blockers are the mainstay of treatment for patients with proteinuria due to hyperfiltration. There is some evidence that bariatric surgery can normalize proteinuria in patients with obesity-related glomerular disease (111).

The identification of molecules essential for maintaining podocyte homeostasis has opened the door to targeted therapeutics for FSGS. Indeed, a number of repurposed drugs currently in clinical use have been shown to act directly on podocytes. Steroids protect podocytes from puromycin aminonucleoside-induced injury (112). Rituximab protects podocytes from injury by preserving sphingomyelin phosphodiesterase acid-like 3b (SMPDL-3b) expression (113). Cyclosporine stabilizes the actin-bundling protein synaptopodin, preventing cathepsin-mediated cleavage (114). Novel target-specific agents being tested at various stages of development include GDC-0879 (115), AC1903 (116), Abatacept (117), Bis-T-23 (118), and cycloRGDfV (17) as summarized in Figure 1.

Podocyte injury due to circulating factors, gene mutations, infections, and many diverse etiologies culminate in the clinical manifestation of proteinuria and the histologic finding of FSGS. The evidence for a circulating permeability factor has been evolving in recent years, and various studies have suggested the pathogenic role of suPAR and other factors in the development of FSGS. The composition, origin, identity, and synergistic roles of circulating factor(s) remain unclear, representing an area of high unmet clinical need. Increasingly, podocyte-associated gene mutations are being identified in patients with FSGS. Expanded knowledge of the molecular architecture of the glomerular filtration barrier raises the prospect of a targeted therapeutic drug development.

HIV and other viral infections can induce podocyte injury and FSGS with an increasing interest in the enhanced susceptibility conferred by APOL1-risk alleles. The numerous divergent etiologies and pathogenic mechanisms with resulting variable response rates to therapy and recurrence post transplant speak to the need for a reclassification of the disease into more clinically relevant subtypes. Further exploration of the contributions of JAK/STAT, Rho-GTPase, and ApoL1 to podocyte injury and survival could enhance the quest for novel therapeutic agents.

YW, SS, and KC conceived the article contents, prepared the manuscript, and endorsed the final draft submitted.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.