Nephrin is a 180 kDa protein encoded by NPHS1. Mutations in nephrin result in congenital nephrotic syndrome of the Finnish type [26]. Nephrin is comprised of a cytoplasmic domain, a short transmembrane domain and an extracellular domain with eight IgG-like motifs and a fibronectin type 3 repeat [27]. Nephrin interacts in a homophilic manner with other nephrin molecules and in a heterophilic manner with neph1 proteins via cis- and trans-interactions (see below) ^[28][29][28,29]. Together, the extracellular domains of nephrin and neph1 form the zipper-like meshwork which creates a sieve-like network with pores with an estimated radius of 12.1 nm each. Nephrin has been shown to be critical for SD formation as without nephrin, the mammalian SD fails to form [30].

The cytoplasmic portions of both nephrin and neph1 play a major role in signaling and are also important for localizing these proteins to the SD. The cytoplasmic tail of nephrin interacts with podocin (NPHS2, see below) which localizes nephrin to lipid rafts within the SD [31]. Extracellular signals are transmitted to the actin core of the each foot process by tyrosine phosphorylation of nephrin. Tyrosine phosphorylation of nephrin occurs in two distinct domains: group A and group B tyrosines [23]. Phosphorylation of either group A or group B tyrosines regulate cytoskeletal organization. Group A tyrosine phosphorylation also modulates cell survival whereas group B mediates nephrin trafficking. In cultured podocytes, tyrosine phosphorylation of nephrin on group A residues mediates lamellipodia formation whereas group B phosphorylation regulates actin polymer extension. Lamellipodia formation in cultured podocytes is regulated through the p85/PI3K/Akt/Cas/Crk pathway ^[21][32][21,32]. Foot process spreading in vivo in response to podocyte injury, a correlate of lamellipodia formation in vitro, is blocked by podocyte-specific knockout of CT 10 regulator of kinase 1/2 (Crk1/2) [21], underscoring the importance of this pathway in FP actin cytoskeletal dynamics.

Actin extension in podocytes in vitro involves tyrosine phosphorylation and recruitment of non-catalytic region of tyrosine kinase adptor protein (Nck) ^[33][34][33,34]. After nephrin phosphorylation, Nck binds to phospho-nephrin and to neuronal Wiskott–Aldrich syndrome protein (N-WASP) ^[33][35][33,35]. N-WASP then activates the actin-related proteins 2/3 (Arp2/3) complex, a complex of seven proteins including the actin-related proteins Arp2 and Arp3.The Arp2/3 complex is intrinsically inactive because the Arp2 and Arp3 proteins are held apart from each other. Binding of N-WASP to Arp2/3 moves Arp2 and Arp3 closer to each other which promotes de novo actin nucleation and filament branching [36].

In addition to its regulation by tyrosine phosphorylation, the cytoplasmic tail of nephrin also contains a consensus sequence for serine/threonine phosphorylation and can be phosphorylated by protein kinase C alpha (PKCα) [37]. Phosphorylation of nephrin by PKCα results in clathrin and dynamin-mediated endocytosis of nephrin [38], a pathway that is activated in diabetic nephropathy ^[37][39][37,39].

Nephrin signaling is also regulated by tyrosine phosphatases [40]. The Src homology region 2 domain-containing phosphatase-1 (SHP-1) phosphatase which modulates group A tyrosine residues is upregulated in diabetes ^[41] [41] and another phosphatase, Cl-Ten (Tensin2) modulates podocyte hypertrophy through mammalian target of rapamycin complex 1 (mTORC1) activation [42].

Work from animal models has shown that nephrin tyrosine phosphorylation is required for maintenance of the slit diaphragm. Conversion of all three group B residues in nephrin to phenylalanine which prevents phosphorylation results in mice that rapidly develop foot process effacement and proteinuria after birth [43].

Besides regulation through phosphorylation, nephrin also binds the adaptor protein CD2-associated protein (CD2AP, see below). CD2AP can interact directly with actin [44] as well as cortactin [45] and synaptopodin [46] (see below). This complex provides a direct link between the signal receptor nephrin and the FP actin cytoskeleton.

The neph1 protein shares some homology with nephrin but contains only five extracellular IgG-like domains [47]. Besides maintaining slit diaphragm integrity through its interactions with nephrin, neph1 also transduce signals from the podocyte exterior. Tyrosine phosphorylation of neph1 results in recruitment of growth factor receptor bound protein 2 (Grb2) and actin polymerization ^[48][49][48,49].

Podocin is a 383 amino acid protein with a hairpin-like structure with both cytoplasmic N- and C-terminal domains ^[50][51][50,51]. Podocin oligomerizes and is found in lipid rafts, specialized microdomains that are enriched in sphingolipid and cholesterol. Assembly of these lipid rafts clusters nephrin at the SD [51]. Podocin is also required to transport nephrin to the SD [52]. Mutations in NPHS2, the gene encoding podocin, are a significant monogenic cause of nephrotic syndrome in children (see below). Besides binding nephrin, the C-terminus of podocin also binds the adaptor protein CD2AP [51]. Podocin also plays a role in formation of tight junctions between neighboring podocytes by associating with and clustering the tight junction proteins coxsackievirus and adenovirus receptor (CAR) and zonula occludens-1 (ZO-1) [53].

Adaptor proteins play a critical role in linking nephrin and podocin to the actin cytoskeleton of the foot processes [54]. Adaptor proteins are non-catalytic molecules that contain specific domains which allow for protein–protein interactions. Several adaptor proteins have been identified as necessary for maintenance of the FP actin cytoskeleton. CD2AP binds both nephrin and podocin and links these proteins to actin [55]. Knockout of CD2AP in mice results in massive proteinuria and death by 6–7 weeks of age, indicating that CD2AP is necessary for maintenance of the SD [56]. Nck is another adaptor protein that has been extensively studied. Nck binding to nephrin also plays an important role in SD maintenance (see above). Inducible knockout of Nck in mice leads to a reduction in nephrin tyrosine phosphorylation and decreased actin recruitment to the SD [57]. ZO-1 is another adaptor protein that is important for maintenance of SD integrity. ZO-1 interacts directly with neph1 ^[58] [58] and this interaction is lost when foot processes efface in response to injury [59].

The ras homology (Rho) family of small GTPases (RhoA, Ras-related C3 botulinum toxin substrate 1 (Rac1) and cell division control protein 42 homolog (Cdc42)) play a pivotal role in actin dynamics, cell shape and motility. Traditionally, Rac1 and Cdc42 were thought to promote podocyte motility at the leading edge of a migrating cell through formation of lamellipodia (Rac1) and filopodia (Cdc42) whereas RhoA was thought to promote formation of contractile actin and myosin containing stress fibers in the cell body and rear of the cell. In accord with this view, an excess of RhoA activity compared to Rac1/Cdc42 had been shown to result in podocytes with a more stationary phenotype and intact foot processes whereas Rac1 and Cdc42 activation had been shown to promote FP motility and effacement ^[60][61][62][60,61,62]. However, more recent work has demonstrated that the balance of RhoA to Rac1/Cdc42 activity is central to maintenance of the FP actin cytoskeleton, as both constitutive overexpression and inactivation of RhoA cause FP effacement and proteinuria ^[63][64][63,64]. A variety of factors can modulate RhoA, Rac or Cdc-42 activity in podocytes including calcium influx via transient receptor potential cation channel subfamily C member 6 (TRPC6) [65].

Synaptopodin is an actin-associated protein expressed in podocyte foot processes [66]. Synaptopodin interacts with α-actinin and this interaction results in bundling and elongation of actin filaments [46]. Synaptopodin also induce stress fibers in cultured podocytes by blocking the Smurf1-mediated ubiquitination and proteasomal degradation of RhoA, resulting in increased RhoA activity [67]. In vitro, synaptopodin suppresses filopodia formation by binding to IRSp53 and blocking the binding of Cdc42 and Mena to insulin receptor substrate p53 (IRSp53) [68]. Treatment of mice with a Mena inhibitor, FP(4)-Mito, reduces proteinuria in a lipopolysaccharide model of podocyte injury [68].

α-actinins are a family of proteins that cross-link actin filaments and serve as scaffolds for the assembly of large protein complexes [69]. α-actinin-4 is highly enriched in podocyte foot processes and mutations in this protein are found in FSGS. In addition, dysregulation of α-actinin-4 occurs early in several other forms of nephrotic syndrome including minimal change disease and idiopathic membranous nephropathy ^[70][71][70,71]. This protein has also been shown to be required for podocyte adhesion to the GBM via interactions with integrins [72]. Loss of α-actinin-4 disrupts interactions between the actin cytoskeleton and integrins, decreasing the overall strength of podocyte attachment to the GBM [16].

Transient receptor potential channel 6 (TRPC6) is a non-selective cation channel that is expressed in podocytes at the SD ^[73][74][73,74]. At the SD, TRPC clusters with several other proteins ^[74][75][76][74,75,76]. Mutations in TRPC6 are associated with FSGS and upregulation of TRPC6 activity has been seen in several proteinuric kidney diseases including autoimmune glomerulonephritis [77]. TRPC6 localization and activity at the SD is modulated by interactions with both nephrin and podocin [78]. In podocytes, TRPC6 also complexes with RhoA [79]. Overexpression of TRPC6 in podocytes results in derangement of the actin cytoskeleton, retraction of podocyte processes, increased cytosolic calcium, activation of RhoA and downregulation of nephrin and synaptopodin [65]. Channel activation in podocytes in vitro or isolated glomeruli has been shown to be mediated by angiotensin II ^[80][81][80,81], diacylglycerol [80], and reactive oxygen species [82]. TRPC6 has also been shown to interact with calpain 1 and 2. The calpains are a family of calcium-dependent cystine proteases that regulate the actin cytoskeleton and cell motility [83]. TRPC6 binding to calpain 1 and 2 induces cleavage of proteins which mediate podocyte adhesion to the GBM, leading to a more motile phenotype. Conversely, knockdown of TRPC6 decreases calpain-initiated cleavage of the podocyte anchoring proteins talin-1, caldesmon-1 and FAK, resulting in podocytes with increased adhesion, decreased motility and increased actin reorganization [83].

Given the critical role that nephrin plays in maintenance of the SD, it is not surprising that mutations affecting the nephrin gene, NPHS1, are associated with the severe kidney disease congenital nephrotic syndrome of the Finnish type ^{[26][33][35][84][85][86][87]}[26,33,35,84,85,86,87]. Additionally, mutations that alter nephrin trafficking or signaling have also been implicated as causing disease. Examples include mutations in the genes GTPase activating protein and VPS9 domains 1 (GAPVD1) and ankyrin repeat and FYVE domain containing 1 (ANKFY1) which were identified to cause a recessive type of nephrotic syndrome by impairing nephrin trafficking^[88] [88].

Podocin, another crucial protein of the slit diaphragm that interacts with nephrin and CD2AP, has also been implicated in human disease. Mutations in NPHS2, which codes for podocin, and LIM homeobox transcription factor 1 Beta (LMX1B), a regulator of podocin expression, have been found to cause nephrotic syndrome ^{[89][90][91][92][93]}[89,90,91,92,93]. Since podocin expression is limited to the glomeruli, NPHS2 mutations typically present without significant extrarenal manifestations ^[92][93][94][92,93,94]. However LMX1B mutations are causative in the syndromic disease nail–patella syndrome characterized by hypoplastic nails and patella, skeletal deformities and varying degrees of nephropathy ^[89][90][95][89,90,95]. The underlying pathophysiology caused by mutations affecting podocin relates to impaired nephrin signaling due to the inability of nephrin to associate with lipid rafts as well as podocin’s effects on TRPC6 ^{[52][82][96][97][98]}[52,82,96,97,98].

CD2AP is scaffolding protein which links nephrin and podocin with the actin cytoskeleton ^{[56][99][100]}[56,99,100]. Mutations of CD2AP gene have resulted in nephrotic syndrome and FSGS lesions in humans ^{[56][101][102]}[56,101,102]. The mechanistic cause for disease is still being studied but is believed to be due to CD2AP’s interactions with F-actin through CAPZ as well as its interactions with Rac1 and synaptopodin ^{[15][55][100][103]}[15,55,100,103]. Similar to CD2AP, membrane-associated guanylate kinase 2 (MAGI2) is another scaffolding protein that interacts with the nephrin complex and helps maintain the functional structure of the slit diaphragm ^[104][105][104,105]. Mutations in MAGI2 have been found to cause nephrotic syndrome in mice and humans [106]. MAGI2-associated disease is caused in part due to its interaction with tachykinin 2 (TAC2), which interacts with the Rho GTPase deleted in liver cancer 1 (DLC1), as mentioned above, as well as its interaction with α-actinin-4 and synaptopodin ^{[15][107][108]}[15,107,108].

As demonstrated in the LMX1B mutations causing kidney disease, diseases can arise if transcriptional regulators of slit diaphragm proteins are mutated. MAF BZIP transcription factor B (MAFB) is an important example of this fact as it is a transcription factor that regulates podocin, nephrin, and CD2AP expression [109]. Two mutations in the MAFB gene have been identified as causative in two different syndromic diseases where proteinuria is a feature, multicentric carpotarsal osteolysis ^[110] [110] and focal segmental glomerulosclerosis with Duane retraction syndrome (FSGS-DRS) [111].

Synaptodopodin is intriguing from a disease modulation standpoint since it regulates actin activity through its effects on α-actinin-4 and through RhoA activation and Cdc42 inhibition [67]. Synaptopodin has been shown to have anti-proteinuric effects and its absence has been noted in several different disease states ^{[68][112][113][114]}[68,112,113,114]. Furthermore, mutations affecting synaptodopodin expression have been noted in patients with proteinuric kidney disease [115]. A significant amount of research is aimed at examining how to modulate synaptopodin for treatment benefit since synaptopodin has TRPC6 modifying properties and part of the effectiveness of the anti-proteinuric effects of cyclosporin A is due to modification of synaptopodin ^[113][116][113,116].

α-actinin-4 is critical for bundling F-actin but also links integrins that adhere to the basement membrane to the cytoskeleton ^[16][18][16,18]. Mutations affecting the ACTN4 gene, which codes for α-actinin-4, clinically present as nephrotic syndrome with FSGS but have a wide age range in which disease manifests ^{[115][117][118]}[115,117,118]. Notably, different mutations in the ACTN4 genes have demonstrated that the absence of α-actinin-4 as well as mutations causing increased affinity for F-actin, both lead to disease ^{[117][118][119][120]}[117,118,119,120].

As outlined, regulation of the cytoskeleton dynamic is maintained by a complex balance through RhoA, Rac1, and Cdc42. Given this complexity and need for a balance between all these pathways, numerous gene mutations associated with this system have been identified as causing diseases. Mutations in Rho GDP dissociation Inhibitor alpha (ARHGDIA) and Rho-GTPase activating protein 24 (ARHGAP24) have both been shown to result in increased Rac1 and Cdc42 activity resulting in disease ^{[121][122][123]}[121,122,123]. Similarly, mutations in KN motif and ankyrin repeat domains 1/2/4 (KANK1, KANK2, and KANK4) have all been reported to be associated with nephrotic syndrome due to their interactions with proteins that interact with RhoA (KANK 1 interacts with synaptopodin and KANK2 interacts with ARHGDIA) [124]. Mutations in FAT1, intersectin 1 (ITSN1), and ITSN2 genes have all been reported to cause nephrotic syndrome due to their adverse effects on Cdc42 activation resulting in impaired podocyte migration ^[107][125][107,125]. One of the most studied mutations affecting Cdc42 are the mutations in inverted formin 2 (INF2) which codes for a formin protein and has been found to cause Charcot–Marie–Tooth disease as well as FSGS without Charcot–Marie–Tooth disease ^{[126][127][128][129][130][131]}[126,127,128,129,130,131]. Subsequently, mutations in the INF2 gene have been found to be one of the most common causes in familial nephrotic syndrome ^[130][131][130,131].

DLC1 is another example of a Rho GTPase-activating protein in which mutations cause nephrotic syndrome [107]. Since DLC1 is regulated by cyclic dependent kinase 20 (CDK20) and tensin-2 (TNS2), mutations in TNS2 and cyclin-dependent kinase 20 (CDK20) are also causative in some cases of nephrotic syndrome [107].

Podocalyxin is a cell surface sialomucin expressed by several different cell types including podocytes, vascular endothelium, hematopoietic cells and some subsets of neurons, that is important for podocyte actin cytoskeleton remodeling [132]. Podocalyxin’s C-terminal binding motif (DTHL) interacts with Na⁺/H⁺ exchanger regulator proteins (NHERF1 and NHERF2) and actin-binding protein ezrin, and can modify the cytoskeleton through these interactions [132]. When podocalyxin is knocked down, podocytes develop abnormal cell morphology. Knockout of the PODXL gene in mice results in loss of foot processes and loss of the slit diaphragm entirely causing a phenotype of anuric renal failure, omphalocele, and perinatal death [133]. A reported loss-of-function mutation in PODXL has been reported clinically with a very similar phenotype to the one observed in PODXL knockout mice [134]. More recently, podocalyxin has been studied outside the kidney for its role in other diseases like cancer, neurologic and atherosclerotic diseases ^{[132][135][136][137]}[132,135,136,137].

TPRC6 plays multiple roles in activating RhoA and Rac1 in response to mechanical stress and influences the function of nephrin, podocin, CD2AP and synaptopodin by regulating intracellular calcium [138]. TPRC6 mutations have also been noted in several cases of congenital FSGS ^{[75][139][140][141]}[75,139,140,141]. Interestingly, disease severity caused by TRPC6 mutations seems to be dependent on the degree to which they alter intracellular calcium ^[138][142][138,142]. Most of these mutations are gain-of-function mutations mediated in part through angiotensin II that causes excessive intracellular calcium ^{[77][138][139]}[77,138,139]. There have been mutations reported that result in a loss of function of TRPC6 which cause disease as well [140]. Remarkably, disease presentation can vary widely in age of onset and severity including some families with members with proteinuric disease and some without disease, despite having the same mutation [143]. As such, it has been postulated that a “second hit” is needed to manifest disease. This theory is supported by a transgenic mouse with a TRPC6 mutation that has no evidence of disease at baseline but developed more severe disease after injury than wild type animals [144]. Since TRPC6 is regulated in part by angiotensin II, a significant amount of research is being done to investigate its role in modulating non-congenital glomerular diseases such as diabetic nephropathy and autoimmune glomerulonephritis [138]. Studies looking at modulating TRPC6 in other disease states have yielded some intriguing results [138], but additional studies are needed to determine when and how to optimally alternate TRPC6 signaling in order to derive clinical benefit since both over and under activation of the channel can be associated with disease.

Phospholipase Cε (PLCε1), a cytoplasmic phospholipase, has been found to be important in podocytes due to its function of regulating internal calcium balance by releasing of calcium from internal stores as well as through activating TRPC6 ^[65][145][65,145]. As mentioned, intracellular calcium affects nephrin and synaptopodin expression in addition to RhoA activation ^{[146][147][148]}[146,147,148]. Mutations in PLCε1 have been reported to cause SRNS because of its effects on nephrin, synaptopodin, TRPC6 and RhoA ^{[146][149][150]}[146,149,150].

Glomerular epithelial protien-1 (GLEPP1) is a tyrosine phosphatase expressed on podocyte foot processes and suggested to regulate glomerular pressure and affects nephrin content at the slit diaphragm ^[151][152][151,152]. Decreases in GLEPP-1 have been reported in several diseases including a recessive form of nephrotic kidney disease associated with a mutation on the GLEPP-1 gene, PTPRO ^{[114][151][153][154][155]}[114,151,153,154,155]. Intriguingly, PTPRO knockout mice were reported to be non-proteinuric at baseline but did have significant microscopic abnormalities in podocyte structure and spacing [152].

Integrins link the cytoskeleton to the extracellular matrix and are composed of α and ß subunits. Integrins are classified as laminin-, collagen- or arginine-glycine-aspartic acid-binding receptors based on their heterodimeric composition which allows for binding to different components of the extracellular matrix. In podocytes, the integrin most critical for maintaining the GFB is the laminin-binding integrin α3β1 [156]. In mice, podocyte-specific knockout of either α3 or β1 integrin results in severe proteinuria and death and in humans mutations in the gene encoding α3 result in severe congenital nephrotic syndrome and a highly disorganized GBM ^[157][158][157,158].

In cultured podocytes, integrin ligation results in tyrosine phosphorylation of nephrin [159]. Clustering of integrins also leads to activation of focal adhesion kinase (FAK) [160]. FAK plays an important role in cell motility and migration including lamellipodia formation and the dynamic assembly or disassembly of focal adhesion complexes by regulating Rho and Rac activity ^{[161][162][163]}[161,162,163]. Activation of integrin α3β1 also leads to actin polymerization through activation of the Arp2/3 complex, cortactin and the WASP family of proteins [13]. α3β1 integrins are linked to Arp2/3 via vinculin [13]. Podocyte-specific knockout of vinculin results in podocytes with altered FA size, mislocalization of ZO-1 and increased cell migration [164].

While integrin α3β1 is critical for normal podocyte function, podocyte injury leads to activation of the constitutively inactive podocyte integrin αvβ3. Activation of αvβ3 integrin is thought to form a complex between soluble urokinase-type plasminogen activator receptor (suPAR), apolipoprotein 1 (ApoL1) and α3βv integrin which propagates podocyte detachment from the GBM [165]. suPAR-induced activation of this integrin also increases activity of TRPC6 [13].

Integrins are linked to the actin cytoskeleton by several cytoskeletal adaptor proteins including talin1. In mice, loss of talin1 in podocytes results in significant disruption of the actin cytoskeleton [166]. In murine models of glomerular disease, there is increased calpain-induced cleavage of talin1 which also results in disruption of the actin cytoskeleton [166].

At FAs, the adaptor protein vinculin links integrins directly with protein complexes involved in actin polymerization including Arp2/3, cortactin and N-WASP ^[13] [13]. The Rho-A-associated protein kinase ROCK is also involved in stress fiber and focal adhesion formation. ROCK induces formation of stress fibers and FAs by phosphorylating myosin light chain. This promotes actin binding by myosin II and increases foot process contractility [167].

Dynamin is another GTPase which is important for FA formation and function in podocytes. In contrast to Rho, Rac and Cdc42 which are small GTPases, dynamin is a large multidomain protein. Dynamin colocalizes with actin and is involved in growth cone and filopodia formation ^{[168][169][170]} [168,169,170]. Dynamin can interact directly with actin ^[171][172] [171,172] and, after forming dynamin rings, can displace the capping protein gelsolin from the barbed ends of actin filaments promoting actin polymerization [173].