Glomerular Cell Biology

The glomerulus or renal corpuscle is comprised of the glomerular tuft surrounded by Bowman’s capsule and space. The tuft is a specialized microvascular bed which contains three cell types including the fenestrated or sinusoidal glomerular endothelial cells, the visceral epithelial cells known as podocytes, and mesangial cells. The glomerular filtration barrier is made up of the endothelial cells and podocytes together with an intervening glomerular basement membrane, and is the site of formation of the primary urinary filtrate. In the average adult human with normal renal function, 180 L of primary urinary filtrate is formed each day. The filtration barrier must permit free passage of water and small solutes into the urine, while retaining larger macromolecules in the blood. The demands of such “high flux” filtration require that the cell types of the glomerulus exhibit many specialized features, which will be discussed in this chapter.

Cell Biology of the Glomerulus

Structure and Function of the Glomerulus (Renal Corpuscle)

The glomerulus or renal corpuscle is comprised of the glomerular tuft surrounded by Bowman’s capsule and space. The tuft is a specialized microvascular bed which contains three cell types including the fenestrated or sinusoidal glomerular endothelial cells, the visceral epithelial cells known as podocytes, and mesangial cells. The glomerular filtration barrier is made up of the endothelial cells and podocytes together with an intervening glomerular basement membrane, and is the site of formation of the primary urinary filtrate. In the average adult human with normal renal function, 180 L of primary urinary filtrate is formed each day. The filtration barrier must permit free passage of water and small solutes into the urine, while retaining larger macromolecules in the blood. The demands of such “high flux” filtration require that the cell types of the glomerulus exhibit many specialized features, which will be discussed in this chapter.

In all mammals, the renal corpuscle appears as a spherical structure, although its diameter varies to some degree with the size of the organism. In humans, glomeruli are approximately 200 μm in diameter, in elephants 300 μm, in rats 120 μm, and in rabbits 150 μm. A normal human kidney contains approximately one million individual glomeruli.

The central component of the renal corpuscle is composed of a plexus of sinusoidal or fenestrated capillaries that extend into Bowman’s space where the primary urinary filtrate accumulates. The capillary loops are held together by the mesangial cells, and they are covered by a continuous layer of podocytes ( Figure 22.1 ). At the vascular pole, the podocyte layer is continuous with the parietal epithelium of Bowman’s capsule. Cells with intermediate phenotypes (between podocyte and parietal epithelium) can be observed at the transition zone. At the urinary pole of the renal corpuscle, the parietal epithelium is continuous with the epithelium of the proximal tubule ( Figure 22.2c ).

Figure 22.1

Scanning electron micrograph of a normal mouse glomerulus.

Glomeruli are spherical bundles of capillary loops, covered by structurally unique podocytes and their foot processes (SEM×1900).

(Courtesy of Dr. Marie Jeansson, Mount Sinai Hospital, Toronto, ON.)

Figure 22.2

(a): Histology of normal human glomerulus. Hematoxylin-eosin (HE) staining showing patent capillary loops (CL) and vascular hilum (VH) (LM×600). (b): Histology of normal human glomerulus. Periodic acid Schiff (PAS) staining (LM×600). (c): Rat glomerulus sectioned through the vascular pole and the urinary pole. The afferent arteriole (AA), the efferent arteriole (EA), the extraglomerular mesangium (EGM), and the macula densa (MD) can be observed in this section. The orifice of proximal tubule (P) can be seen at the urinary pole (PE: parietal epithelial cells of Bowman’s capsule; US: urinary space; LM×490). (d): Immunostaining shows the three cell types found within the glomerular tuft (mouse glomerulus is shown) (Green: Zo-1=podocytes; red: CD31=endothelial cells; yellow: desmin=mesangial cells). The capillary loops are outlined by podocytes. Mesangial cells are located within the capillary tuft and connect capillary loops with each other (×400).

(a and b: Courtesy of Dr. Paul S. Thorner, The Hospital for Sick Children, Toronto, ON.)

At the site of transition between parietal and visceral epithelium, the afferent and efferent arterioles enter and exit the glomerulus respectively, this is known as the glomerular hilum. After entering, the afferent arteriole branches to form a complex plexus of fenestrated capillaries with loops at the urinary pole. The mesangium is required for proper structure and formation of this plexus, and in its absence only a single ballooned capillary loop forms. The capillary loops come into direct contact with the mesangium at discrete points in a small region known as the juxtamesangial portion. However, the majority of the loops are found within Bowman’s space and are covered entirely by the glomerular basement membrane and podocyte foot processes. This is the surface area across which filtration occurs. The branches of the afferent arteriole give rise to individual vascular lobules within the glomerular tuft; each of these lobules contains its own afferent and efferent capillary with some connections between lobules. After looping at the urinary pole, the efferent capillaries join to form the larger efferent arteriole, which exits the tuft at the glomerular hilum. Along the length of the efferent arteriole, the extraglomerular mesangium is gradually replaced by typical smooth muscle cells ( Figures 22.3 and 22.4 ).

Figure 22.3

Schematic diagram sshow the branching pattern of glomerular capillaries.

As soon as the afferent arteriole enters the glomerulus, it divides into two to five branches to form the glomerular capillaries. The capillaries run toward the urinary pole, continuing to branch, and then loop back to the vascular pole to unify and form the efferent arteriole within the glomerular tuft. The efferent arteriole possesses a significant intraglomerular portion ( stippled ), whereas the afferent arteriole does not.

(From Winkler, D., Elger, M., et al. (1991). Branching and confluence pattern of glomerular arterioles in the rat. Kidney Int. Suppl. 32 : S2–8, with permission.)

Figure 22.4

Scanning electron micrograph of a glomerular vascular cast.

Afferent (A) and efferent (E) arterioles can be seen.

Glomerular Endothelial Cells

General Description

Glomerular endothelial cells are large, highly flattened cells that form the innermost layer of the glomerular capillary. Peripherally these cells are extremely thin, and the cell body contains the nucleus and all the cell organelles. The peripheral portions of the endothelial cells contain numerous fenestrae, which are 50–100 nm pores that penetrate the cytoplasm ( Figure 22.5 ). The luminal side of endothelial cells is covered by a thick layer consisting of glycoproteins that form “sieve plugs” in the fenestrae and the glycocalyx.

Figure 22.5

(a): Scanning electron micrograph shows the inner surface of a healthy glomerular capillary and its beautifully fenestrated endothelium. (B): Endothelial fenestrae are largely reduced after 14 days treatment of mice with soluble Flt-1 (SEM×18,000).

(From Kamba, T., and McDonald, D. M. (2007). Mechanisms of adverse effects of anti-VEGF therapy for cancer. British Journal of Cancer 96 : 1788–1795, with permission.)

Formation of Glomerular Capillaries

Several transplantation studies have demonstrated that the glomerular capillaries are formed largely by vasculogenesis: endothelial cells are derived from angioblasts believed to be intrinsic to the metanephric mesenchyme. At E12.5, in mice before the formation of immature vasculature, kinase insert domain receptor (Kdr, also known as vascular endothelial growth factor receptor 2 (Vegfr-2) or Flk-1) positive angioblasts are present in the metanephric mesenchyme. In the S-shaped stage of the developing nephron, immature podocytes start expressing vascular endothelial growth factor-A (Vegfa), thereby attracting Kdr-expressing endothelial cells to migrate to the vascular cleft. Transforming growth factor-β1 (Tgfb1) induces apoptosis of the endothelial cells and opens the capillary lumens.

Fenestrae and Diaphragms

Ultrastructural analysis demonstrates that diaphragms are often observed in fenestrae of glomerular endothelial cells in rodent embryos. Diaphragmed fenestrae are formed in the S-shaped stage, and then the diaphragms disappear from the capillary loop stage onwards. The main component of the fenestration diaphragm is type II transmembrane glycoprotein plasmalemmal vesicle-associated protein-1 (Pv-1), but its precise function is still unknown. Fenestrae in adult glomerular endothelial cells do not have diaphragms ( Figure 22.5 ); however, fenestrae bridged by diaphragms can be found along the intraglomerular segment of the efferent arteriole and its derivatives. Diaphragms can also be observed in a drug-induced nephritis model, suggesting that diaphragms are required in the development and remodeling of fenestrations, thereby compensating for the immaturity of the barrier function in these settings.

Glycocalyx/Sieve Plugs

The luminal membrane of endothelial cells is covered by a highly negatively-charged layer called the endothelial surface layer (ESL). The relatively dense, membrane-associated part of this layer is called the glycocalyx, and the larger, less compact component is known as the endothelial cell coat ( Figure 22.6 ). The main components of the ESL are glycoproteins, glycoaminoglycans (GAGs), and membrane-associated and secreted proteoglycans. Ultrastructural examinations with sophisticated specialized fixation techniques have revealed that this layer also fills the fenestrae with slit diaphragm-like “sieve plugs”. The thickness of the glycocalyx is estimated to be 50–100 nm, and that of the loose endothelial cell coat is considered to be 200–400 nm. The relative importance of the ESL and sieve plug in the glomerular filtration barrier is still controversial.

Figure 22.6

Schematic diagram showing the endothelial surface layer (ESL).

The relatively dense part of the layer close to the endothelial cells form the glycocalyx, which consists of membrane-bound proteoglycans (PG), including syndecan and glypican. Syndecan carries both chondroitin sulfate (CS) and heparan sulfate (HS) side chains, and glypican carries HS side chains. The ESL is comprised of secreted proteoglycans such as perlecan (mainly HS) and versican (mainly CS), as well as secreted glycosaminoglycans (GAG) including hyaluronan. It also traps some plasma proteins such as albumin.

(Modified from Haraldsson, B., Nystrom, J., and Deen, W. M. (2008). Properties of the glomerular barrier and mechanisms of proteinuria. Physiol. Rev. 88 : 451–487, with permission.)

Functional Maintenance of Glomerular Endothelial Cells: Insights from Studies of Angiogenic Factors

A number of factors are involved in the maintenance of glomerular endothelial structure and function, and coordinate an elaborate cross-talk between endothelial cells and other cell types of the glomerulus ( Figure 22.7 ).

Figure 22.7

Soluble factors involved in the maintenance of endothelial cells and the glomerular capillary structure.

Schematic diagram shows the location of secretion of soluble factors, and where the associated receptors are expressed. Vegfa from podocytes is required for the recruitment, survival, and maintenance of the endothelial cells, and binds to its receptor Kdr. sFlt-1 works as a decoy, but its precise function is still unknown. Angpt-1 is expressed in podocytes and binds to the endothelial Tie-2/Tek receptor in a paracrine fashion, whereas its antagonist, Angpt-2 is secreted from endothelial cells and binds to the Tie-2/Tek receptor in an autocrine fashion. Cxcl12 is secreted from podocytes and interstitial cells, and acts on Cxcr4 in endothelial cells to regulate vascular development and function. Pdgfb secreted by endothelial cells signals to the Pdgfrb receptor expressed by mesangial cells, and is a critical factor for their migration and maintenance. Endothelial cells also express some vasoactive factors including NO and endothelins. Expression of growth factors such as Tgfb, Ctgf, Igf, and Fgf are increased in disease conditions such as diabetic nephropathy, but their precise functions under normal conditions are still unclear.


Platelet-derived growth factor B (Pdgfb) is secreted from endothelial cells and binds its receptor (Pdgfrb) on mesangial cells. Pdgfb or Pdgfrb knockout mice have a defect in mesangial migration and a single, dilated glomerular capillary loop ( Figure 22.8 ). Endothelial specific deletion of Pdgfb results in the same phenotype as that seen in mice with a systemic knockout. Also, this paracrine system requires retention of the ligands in the pericellular space, because mutants with deletion of the Pdgfb retention motif demonstrate delayed mesangial migration and, later on, proteinuria and glomerulosclerosis.

Figure 22.8

PAS stain of glomeruli from E17.5 mouse embryos.

(a): Normal mouse glomerulus. Arrows show normal fold of the basement membrane. (b): Pdgfb −/− glomerulus shows a single open aneurysm-like capillary loop without any mesangium (failure of mesangial migration). There is no fold of basement membrane (arrows).

(From Betsholtz, C. (1995). Role of platelet-derived growth factor in mouse development. Int. J. Dev. Biol. 39 : 817–825 with permission.)


Vegfa facilitates the formation of fenestrae in cultured glomerular endothelial cells. Podocytes produce large amounts of Vegfa that can bind to Kdr on endothelial cells. Cell-selective deletion of Vegfa from podocytes demonstrates that Vegfa signaling is required for formation and maintenance of the glomerular vasculature, its fenestrated phenotype and the filtration barrier. Mice treated with soluble fms-related tyrosine kinase-1 (sFlt-1, discussed below), a decoy receptor of Vegfa, show striking attenuation of endothelial fenestration, highlighting the necessity for Vegfa in the maintenance of fenestration ( Figure 22.5 ). In mutant mice that carry a podocyte-specific gene deletion of Vegfa , a few endothelial cells migrate into the developing glomeruli but they fail to develop fenestrations and rapidly disappear, causing renal failure and neonatal death. The deletion of one allele of the Vegfa gene from podocytes leads to a glomerular defect known as endotheliosis, characterized by endothelial swelling and loss of fenestrations – a universal feature found in thrombotic microangiopathies. Overexpression of the major angiogenic Vegfa 164 isoform in podocytes results in collapse of the glomerular tuft. Additionally, patients receiving anti-VEGF therapy may develop proteinuria due to thrombotic microangiopathy (TMA) of the glomerulus with prominent endotheliosis ( Figure 22.9 ). Indeed, deletion of Vegfa in mature podocytes of adult mice leads to TMA. Taken together, these results indicate an indispensable role for Vegfa in the development, maintenance, and function of the glomerular vasculature and filtration barrier. They also highlight the importance of Vegfa paracrine signaling from the podocyte to Kdr on glomerular endothelial cells.

Figure 22.9

A glomerulus showing thrombotic microangiopathy from a patient who received anti-VEGF therapy, showing fragmentation of erythrocytes (FE: arrows) and foamy change (FC: arrow) of endothelium.

(Courtesy of Dr. Laura Barisoni, New York University School of Medicine, New York, NY.)

sFlt-1 is an alternatively spliced soluble form of VEGF receptor 1 (VEGFR-1)/Flt-1, and binds to VEGF as a decoy, thereby acting as a potent inhibitor of VEGF activity. Treatment of mice with adenoviral-induced sFlt-1 leads to a massive reduction of endothelial fenestrae ( Figure 22.5b ). sFlt-1 blood levels are elevated in patients with pre-eclampsia and administration of sFlt-1 to pregnant rats causes hypertension and proteinuria with histological glomerular endotheliosis. The endothelium is the most common glomerular region affected in pre-eclampsia, suggesting a functional role for sFlt-1 in the function of the glomerular endothelium. A recent study also implicated sFlt-1 in the pathogenesis of PR3-ANCA-associated vasculitis affecting the glomeruli.


Infusion of a neutralizing antibody against transforming growth factor β1 (Tgfb1) to neonatal rats causes a delay in glomerular capillary formation, including the development of fenestrations. In the choroid plexus, neutralization of both Tgfb1 and Vegfa leads to decreased cerebral perfusion, vascular thorombi, and a defect of fenestration, which does not occur when either of these treatments are administered individually. Thus, Tgfb1 also plays a crucial role in the development and maintenance of glomerular endothelial cells, and may work in concert with Vegfa. Additionally, blood levels of soluble endoglin, an antagonist of Tgfb1 are elevated in patients with pre-eclampsia, and correlate with disease severity. Administration of soluble endoglin in combination with sFlt-1 to pregnant rats results in severe pre- eclampsia, including HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets).


Another family of angiogenic factors required for the development and homeostasis of glomerular endothelial cells is the Angiopoetin–Tek signaling system. Angiopoietin 1 (Angpt1) and Angiopoietin 2 (Angpt2) are ligands for Tek tyrosine kinase (Tek/Tie-2). Angpt1 binds to the Tek receptor expressed on endothelial cells, and causes its phosphorylation. This signal leads to enhanced survival of endothelial cells, stabilization of the endothelial cell-to-cell connection, and reduced permeability. Angpt2 is considered to be a competitive antagonist of Angpt1 by binding Tek but not activating any intracellular signaling. There is some data, however, that suggests that Angpt2 can activate Tek signaling under certain conditions. Angpt1, Angpt2 and Tek are all expressed in developing kidneys. Angpt1 is expressed widely in the condensing mesenchyme in the developing kidney and its derivatives, and in mature podocytes. Angpt2 shows a more restricted expression pattern, localizing to endothelial cells, pericytes, smooth muscle cells of cortical and large blood vessels, and immature mesangial cells. Tek is expressed both in mature and immature glomerular endothelial cells. Angpt1 conventional knockout mice die at embryonic day 12.5, thus precluding any analysis of its role in the glomerular vasculature. In mouse metanephric organ culture, recombinant Angpt1 enhances the growth of interstitial capillaries. A recent report demonstrated that Angpt1 treatment of isolated rat glomeruli reduced vascular permeability and increased the depth of the glycocalyx layer. Cell type-specific/inducible knockout approaches, however, have revealed a crucial role for Angpt1 in glomeruli. Deletion of Angpt1 at E10.5 leads to a single, dilated glomerular capillary loop without mesangial migration in a portion (~10%) of glomeruli that is reminiscent of the Pdgfb/Pdgfrb mouse mutant phenotype ( Figure 22.10a,b ). Although deletion of Angpt1 after E13.5 doesn’t cause any immediate vascular phenotype, streptozotocin-induced diabetic mice with global or glomerular-specific Angpt1 deletion develop increased urinary albumin excretion, severe mesangial expansion, glomerular sclerosis, and early mortality ( Figure 22.10c,d ). Another report also demonstrated that treatment of diabetic mice with Angpt1 recombinant protein is protective for renal function. Angpt1 is therefore dispensable in quiescent, mature glomeruli, but is essential in development and in the vascular response to injury.

Figure 22.10

Conditional deletion of Angpt1 results in glomerular developmental defects and enhanced diabetic glomerular injury.

(a): A normal glomerulus at embryonic day 17.5 (E17.5). (b): Early conditional deletion of Angpt1 gene at embryonic day 10.5 (midgestation) results in some abnormal glomeruli with single open capillary loops similar to that of Pdgfb-null mouse at E17.5. (c) and (d): Late deletion of Angpt1 at E16.5 doesn’t lead to any immediate phenotype. However, after 20 weeks of streptozotocin-induced diabetes, the diabetic mutant mice that carry Angpt1 deletion show an increase in mesangial matrix expansion and sclerosis (d) compared with that of diabetic controls (c) (LM×1000).

(Courtesy of Dr. Marie Jeansson, Mount Sinai Hospital, Toronto, ON.)

Angpt2 knockout mice are briefly viable in the post-natal period, and exhibit increased pericyte coverage of peritubular capillaries. The mice die soon after birth however, precluding analysis of the role of Angpt2 in more mature capillary beds. Angpt2 overexpression in podocytes causes proteinuria and podocyte apoptosis in formed glomeruli.

Angiopoietin ligands seem to function in concert with Vegfa. In Vegfa-rich conditions, Vegfa and Angpt2 work together to promote sprouting. The precise degree of cross-talk between these pathways is still under investigation.

Ephrin-Eph Family

Ephrin-Eph molecules are another family of tyrosine kinase signaling factors that are widely expressed in the developing kidney. In other organs, they are involved in the specification of arteries and veins, as well as in neural development. In the developing kidney, the Ephrin B2 ligand is expressed in podocyte precursors, but later it is expressed by endothelial and mesangial cells. In adults, the EphB4 receptor is mainly expressed in podocytes. Overexpression of the EphB4 receptor causes defects of glomerular arteriolar formation. Deletion of Ephrin B2 from Pdgfrb-expressing pericytes and the mesangial cell population results in abnormal glomerular tuft formation. Pharmacologic inhibition of EphB4 leads to delayed recovery and extended injury of endothelial cells and podocytes in a rat mesangial injury model. Therefore, it seems this family is involved in glomerular tuft formation and maintenance, but the precise mechanism is still unclear.


Chemokine CXC motif receptor type 4 (CXCR4), CXCR7, and Chemokine CXC motif ligand 12 (CXCL12) are also critical factors in the development and maintenance of glomerular vasculature. Cxcr4 and Cxcr7 are seven-transmembrane G-protein coupled receptors, and Sdf-1/Cxcl12 is their cognate chemokine ligand. A deficiency of Cxcr4 or Cxcl12 leads to a failure of vasculogenesis and embryonic lethality. During renal development, Cxcr4 −/− or Cxcl12 −/− knockout mice show a ballooning of the glomerular capillary loops which is reminiscent of Pdgfb/Pdgfrb knockout mice. On the other hand, activation of the CXCR4/CXCR7/CXCL12 axis appears to underlie some glomerular diseases, such as rapidly progressive glomerulonephritis (RPGN), diabetic nephropathy, and hemolytic uremic syndrome (HUS). It appears that Cxcl12 is secreted from podocytes or interstitial cells, and acts on Cxcr4 expressed by the endothelial cells to regulate vascular development and function.

Glomerular Endothelial Cells as a Source of Vasoregulators

Glomerular endothelial cells produce both nitric oxide, a vasodilator, and endothelin-1, a vasoconstrictor. Nitric oxide is produced by NO synthases. Both endothelial and inducible NO synthases are expressed by glomerular endothelial cells in vitro and in vivo . eNOS expression and activation is at least partially influenced by Vegfa. In many rodent disease models and human patients with kidney diseases, overproduction of NO and its derivatives has been observed. It is considered that excessive NO production generated by inducible NO synthase (iNOS) results in glomerular injury, whereas NO generated from endothelial NO synthase (eNOS) is protective by preserving endothelial survival and function. In diabetic patients, eNOS expression is increased in renal endothelial cells, whereas iNOS expression is preferentially upregulated in inflammatory cells. The degree of eNOS expression is related to the severity of the glomerular lesion and proteinuria. Diabetic eNOS knockout mice develop more severe glomerular lesions and greater albuminuria. In addition, excretion of NO-related products is often reduced in diabetic patients with nephropathy.

Endothelin-1 is a potent vasoconstrictor which binds to one of two receptors: endothelin receptor type A (Ednra); and endothelin receptor type B (Ednrb). Binding of endothelin-1 to Ednra results in vasconstriction, while binding to Ednrb causes vasodilation. In glomerular endothelial cells, Ednrb is dominant, whereas Ednra is expressed by mesangial cells.

Glomerular endothelial cells are also involved in the renin–angiotensin–aldosterone system (RAAS). They express angiotensin-converting enzyme (ACE) and produce angiotensin II. The relative contribution of the glomerular endothelium compared to the systemic endothelium with regard to angiotensin II production, however, is still unknown. Furthermore, the potential role of angiotensin receptors in endothelial cells is also unknown.


Glomerular endothelial cells form the first perm-selective barrier in the glomerulus. The relationship between their elaborate, fenestrated cell shape and function is maintained by a network of various angiogenic factors. Glomerular endothelial cells also contribute to the charge-selective barrier through the negatively-charged glycocalyx.

Mesangial Cells


Mesangial cells are irregularly shaped cells which extend processes from their cell body towards the glomerular basement membrane (GBM). The “mesangium” refers to the mesangial cells together with the mesangial matrix they produce ( Figures 22.11 and 22.12 ). Mesangial cells provide structural support to the glomerular tuft, produce and maintain mesangial matrix, communicate with other glomerular cells by secreting soluble factors, and may contribute to the glomerular capillary flow via their contractile properties.

Figure 22.11

Transmission electron micrograph showing all components of the glomerular tuft.

The glomerular tuft is marked by the GBM, which includes the pericapillary GBM (cGBM; between podocytes and the endothelium) and the perimesangial GBM (mGBM; between podocytes and mesangium). Glomerular capillaries (C) are covered by podocytes at the periphery, and connected to the mesangium, proximally. The mesangial cell body (M) possesses several processes (some are marked by stars ) which extend toward the peripherally located capillaries. In the mesangial area, abundant mesangial matrix ( triangles ) can be seen (US: urinary space; TEM×5500).

Figure 22.12

(a): Transmission electron micrograph showing a mouse glomerular capillary. A mesangial cell (MC) extends its processes to a capillary loop (C). Microprojections ( arrowheads ) from the primary process run toward the glomerular basement membrane (GBM). As shown by arrows, the endothelial cells are directly connected to the mesangium. Podocyte foot processes (FP) and fenestrated endothelium (E) are shown (TEM×13,500). (b): High magnification of the juxtamesangial part of the capillary loop showing direct contact of mesangium to endothelium. Endothelial cells are attached to the mesangial cell processes (MP) that connect opposing mesangial angles. (arrows)

(CL: capillary lumen; US: urinary space; MM: mesangial matrix; FP: foot processes; TEM×23,000).

Mesangial Cells Provide Structural Support to the Glomerular Tuft

The mesangium forms the central core of the glomerular tuft. The processes which they extend towards the GBM are densely populated by bundles of actin, myosin, and β-actinin microfilaments. These processes attach directly or by interposition of microfibrils to the GBM. They also extend underneath glomerular endothelial cells toward the mesangial angles of the GBM, anchoring two opposing mesangial angles together through their microfilament bundles ( Figure 22.12b ). These microfilament bundles cross the mesangial cells to tether opposing parts of the GBM through α3β1 integrin and the basal cell adhesion molecule (BCAM) glycoprotein, which bind laminin α5 in the GBM. These structures are believed to supply protection from hydraulic pressure by providing inward-directed tension.

Mesangial Cells Produce and Maintain Mesangial Matrix

The mesangial matrix fills the remaining spaces between the mesangial cells and the perimesangial glomerular basement membrane (GBM, for review see ). This matrix is composed of a diverse array of common matrix proteins including collagens type III, IV, V, and VI; heparan sulfate proteoglycans including biglycan and decorin ; and the elastic fiber proteins fibronectin, laminin, entactin, and fibrillin-1, among others. Fibronectin is the most abundant of these, and is associated with microfibrils which network to form the basic ultrastructure of the matrix. These microfibrils are unbranched and non-collagenous with a diameter of 15 nm, and form a dense three-dimensional network that contributes to the anchoring of mesangial cells to the GBM. It is thought that these microfibrils allow for the transmission of mesangial cell contractile forces to the GBM.

Mesangial phenotypic changes are a hallmark of certain glomerular diseases such as diabetic nephropathy. This condition is characterized by glomerular sclerosis due to an accumulation of mesangial matrix and thickening of the GBM. The sclerotic lesion contains an abundance of type IV collagen normally present in the glomerulus, but also contains types I and III collagen which are usually absent but are produced by injured mesangial cells.

Signaling Molecules Involved in Mesangial Cell Biology


In addition to α3β1 integrin observed in connections between the GBM and mesangial processes, α1β1, α2β1, and fibronectin receptors α5β1 and α8β1 integrins are expressed in the mesangial cells, and are able to activate integrin-linked cell signaling. In the absence of α1 integrin, cultured mesangial cells show decreased proliferation, increased matrix production, and altered intracellular signaling. Mice lacking α1 integrin exhibit more severe Adriamycin and diabetes-induced glomerular injury, highlighting the importance of matrix-related signaling. On the other hand, removal of α2 integrin in mice leads to amelioration of glomerular damage in an Adriamycin and partial renal ablation model. α8 integrin seems to promote adhesion, but inhibits migration and proliferation of mesangial cells in vitro .


Pdgfb is secreted from endothelial cells and binds to Pdgfrb expressed in mesangial cells, where it exerts a crucial role in mesangial migration and glomerular tuft formation. Without mesangial cell migration, the looping of glomerular capillaries does not occur. This interaction is discussed further in the endothelial section.

Ephrin B2

As mentioned in the section on the glomerular endothelium, deletion of Ephrin B2 from mesangial cells and pericytes leads to abnormal glomerular tuft formation and reduced numbers of capillary loops. It’s receptor, EphB4 is expressed in podocytes, suggesting an interaction between the mesangium and podocytes.

Other Secretary Molecules and Receptors

Mesangial cells produce, and are also influenced by, many growth factors including Tgfb1, connective tissue growth factor (Ctgf), insulin like growth factor (Igf), fibroblast growth factor (Fgf), and hepatocyte growth factor (Hgf). Among these factors, Hgf antagonizes the pro-fibrotic actions of Tgfb1, whereas the other factors are upregulated by mesangial cells in disease conditions including diabetes or Thy1.1 nephritis, and facilitate glomerular matrix accumulation. Vasoactive factors such as angiotensin II and endothelins promote mesangial proliferation, and this effect may be mediated by transactivation of the Egf receptor. However, because of the lack of effective tools for deleting genes specifically from mesangial cells, the precise functions of these factors in normal physiology remain to be determined.

Contractile Ability of Mesangial Cells

Because of the microfibrils and the contractile ability of cultured mesangial cells, it has been assumed that the mesangial cells regulate glomerular filtration by controlling the capillary surface area, but concrete in vivo evidence is lacking.

The Glomerular Basement Membrane


The glomerular basement membrane (GBM) is a specialized extracellular matrix that sits between the podocytes and vascular side of the filtration barrier. During glomerulogenesis, components of the GBM are synthesized by both the glomerular endothelial cells and podocytes, forming a bilayered structure that later fuses. Compared to basement membranes in other tissues the GBM is unusually thick, measuring 240–370 nm in adults. Ultrastructural analysis of the mature GBM shows a trilaminar structure made up of a lamina densa surrounded by the lamina rara interna and externa, which appear less dense.

Similar to other basement membranes, the GBM is primarily made up of laminin, collagen type IV, nidogen, and heparan sulfate proteoglycans. However, the GBM contains different members of some of these families compared to other basement membranes, including laminin-521, collagen a3a4a5(IV), and agrin. Laminin and collagen type IV appear to be particularly important for function of the GBM, as mutations in these factors are associated with glomerular filtration defects and renal disease. Components of the GBM are continuously “turned over”, but it is not yet clear how new components are added and old ones removed.

Given the location of the GBM between the relatively “open pores” of the endothelium and the podocyte foot processes and filtration slits bridged by the slit diaphragms, a major role for the GBM is to restrict the passage of plasma proteins into Bowman’s space. From the time of classic electron micrographic studies of the GBM in the 1950s attempting to define the specific characteristics of the GBM that impart its perm-selective properties, to the current molecular era, it is still hotly debated which component represents the major barrier. Current models suggest that all three layers of the filtration barrier are likely important (including the glycocalyx of the endothelial layer), but relative contributions remain unknown. However, it is clear that mutations in genes encoding specific GBM proteins are sufficient to cause proteinuria and renal failure, underscoring the importance of this layer to the barrier.


Laminins are secreted as heterotrimers, which are stabilized by disulfide interchain bonding. Each laminin heterotrimer is composed of α, β, and γ chains that combine with each other in nonrandom combinations to form at least 15 different heterotrimers. The laminins are named by their composition; for example: α5β2γ1 is named laminin-521 or LM-521. The heterotrimer structure appears as a “cross” with one longer and three shorter arms. At the COOH end of all α chains is a laminin globular chain (LG) that extends beyond the long arm. This LG domain interacts with cell surface receptors such as integrins and dystroglycans that are expressed by the podocytes.

Interactions between the laminin heterotrimers themselves are mediated by subdomains found in the shorter arms of the “cross.” Laminin also binds to the network of type IV collagen via nidogen, and to agrin, a heparan sulfate proteoglycan of the GBM. The combined network of type IV collagen (see below) and laminin provides mechanical properties to the basement membrane, serving as a scaffold for the placement of other matrix components.

During glomerular formation, laminin trimer compostion changes from LM-111 to LM-511, and finally to LM-521. In the adult, laminins continue to be produced by both endothelial cells and podocytes, as the protein can be identified in the endoplasmic reticula of both cell types.

A series of reports confirm an important role for laminins in GBM function. A mutation of laminin β2 results in Pierson syndrome in humans that is characterized by ocular and neurological symptoms, and congenital nephrotic syndrome. This observation was confirmed using knockout mice of laminin β2 that develop proteinuria even before the onset of visible ultrastructural changes to the podocytes. Recently, the phenotype of the laminin β2 knockout mouse was rescued by overexpression of laminin β1, suggesting redundancy between laminin β1 and β2. Also, deletion of laminin α5 in mice prevents the transition from LM-111 to the mature LM-521 that leads to breakdown of the GBM and failure of glomerular vascularization. A mouse carrying a hypomorphic mutation of laminin α5 also shows glomerular proteinuria, hematuria, and cystic kidneys, suggesting the gene dosage of the laminin α5 chain is crucial for the maintenance of the GBM.

Type IV Collagen

Collagen IV is another major component of the GBM. Similar to other collagens, collagen IV is a trimeric extracelluluar matrix component made up of α chains that are rich in Gly-X-Y amino acid repeats. The type IV collagen family includes six genetically distinct α chains that trimerize with each other in specific combinations to make three types of trimers with each other: (α1) 2 α2; α3α4α5; and (α5) 2 α6. Each of these trimers is referred to as a protomer. Protomers are secreted into the extracellular matrix, where they self-polymerize and are subsequently cross-linked by specific enzymes.

Similar to laminin, the composition of type IV collagen trimers undergoes a developmental switch in the GBM. Early in glomerulogenesis (S-shaped stage), the GBM is composed of α2/α1 chains, which are replaced at the capillary loop stage by α3α4α5. Although a basement membrane can form in the absence of type IV collagen its structure is compromised, resulting in variable defects in the filtration barrier depending on the specific genetic mutation. In the adult kidney, the type IV collagen α3α4α5 network is produced only by podocytes.

Mutations in any of the α3, α4 or α5 genes are associated with glomerular disease in patients, emphasizing the key role that type IV collagen plays in glomerular barrier function. The defects can be minimal, as observed in patients with thin basement membrane disease (also known as benign familial hematuria). This disease is inherited in an autosomal dominant fashion, and patients present with isolated microscopic hematuria. As suggested by its name, the GBMs show thinning. Thirty to forty per cent of the patients exhibit mutations in COL4A3 or COL4A4. Although originally assumed to be benign, FSGS has been reported in Cypriot families with this disease.

The same mutations in COL4A3 and COL4A4 inherited in a homozygous fashion as an autosomal recessive disease result in a more severe disease known as Alport’s syndrome, a hereditary basement membrane disease associated with progressive glomerulopathy that leads to renal failure, and is associated with deafness and ocular abnormalities. The commonest form of Alport syndrome is X-linked, caused by mutations in the α5 chain ( Figure 22.13 ). Goodpasture syndrome, an autoimmune disease characterized by glomerulonephritis and lung hemorrhage, is mediated by antibodies against the type IV collagen α3 chain.

Figure 22.13

Transmission electron micrograph of the GBM in a patient with Alport syndrome.

Thickening, splitting and lamination of the GBM can be seen (TEM×14,900).

(From Rumpelt, H. J. (1987). Alport’s syndrome: Specificity and pathogenesis of glomerular basement membrane alterations. Pediatr. Nephrol. 1 : 422–427, with permission.)

Intriguingly, recent studies have shown that a mutation of the type IV collagen α1 chain causes hereditary angiopathy with nephropathy, aneurysms, and muscle cramps (HANAC syndrome); the renal manifestations are characterized by hematuria and large bilateral cysts. A rodent model of this mutation leads to defects of the GBM, suggesting a functional role for the α1 chain in the GBM, despite the prominence of a3a4a5 trimers in mature type IV collagen.


Nidogen-1 and -2 (previously known as entactin-1 and entactin-2) are virtually ubiquitous basement membrane glycoproteins. Nidogen-1 binds both laminin and type IV collagen, suggesting that it is involved in forming the GBM. However, it was demonstrated that a single knockout of either nidogen-1 or nidogen-2 in mice results in no GBM phenotype. Although deletion of both nidogen genes results in perinatal lethality, suggesting some overlapping roles of the two genes, the glomerular basement membrane still forms apparently normally. The exact role and requirement of nidogen in the GBM is not yet known.


Heparan sulfate proteoglycans of the GBM are concentrated in the laminae rarae internae and externae, and provide an electronegative charge to the GBM. The major proteoglycans of the GBM are heparan sulfate proteoglycans; most prominent is agrin, but perlecan is also present. Although classical studies suggested that the negative charge of the GBM was essential to retard passage of neutral and negatively-charged macromolecules across the barrier, this model has been challenged due to results obtained from genetically modified mice. Podocyte-specific knockout mice for agrin and/or perlecan, either in isolation or together, does not result in proteinuria or overt filtration defects. Current thoughts on the glomerular charge barrier are discussed later in the chapter.


Podocyte Morphology

Mature podocytes are highly-differentiated, polarized epithelial cells that sit on the glomerular basement membrane in Bowman’s space. They function as vascular support cells, wrapping around the underlying glomerular capillaries, providing growth factors necessary for endothelial health and survival. Podocytes are characterized by a highly arborized and unique cytoskeleton. They have a large cell body ( Figure 22.14a ) that gives rise to a complex network of processes including primary, major processes, which then continue to branch as they extend around the glomerular capillary loops until they form terminal foot processes. The foot processes are the only point of contact between the podocyte and the GBM ( Figure 22.15a ). The foot processes of neighbouring podocytes interdigitate ( Figure 22.14b ), and are connected to one another through a specialized intercellular junction known as the slit diaphragm (SD). The SD bridges the porous filtration slits; these structures have been recently visualized at high magnification ( Figures 22.14b , and 22.15b ). In disease, these structures can be dysregulated, causing a disorganization of foot processes and breakdown of the GFB ( Figure 22.15c,d ).

Figure 22.14

Scanning electron micrograph shows podocytes wrapping around glomerular capillary loops.

(a): Mouse podocytes have a large, smooth-surfaced cell body (CB) which branches into primary processes (PP), which continue to branch as they wrap around the capillary loop forming the actin-based foot processes (SEM×5600). (b): Scanning electron micrograph of rat podocytes. The filtration slits are created between the interdigitating foot processes, the terminal ends of the branching processes (FS) (SEM×5670). (Courtesy of Dr. Monika Wnuk, Mount Sinai Hospital, Toronto, ON.)

Figure 22.15

Transmission electron micrograph of a normal and abnormal human podocytes.

(a): Coronal section of a podocyte intimately associated and lining a glomerular capillary loop (CL) cell body (CB) can be seen. Note the space (*) between the cell body and the basement membrane as the foot processes are the only point of contact between the podocyte and the GBM. (b): High power transmission EM shows podocyte foot processes (FP) attached to the glomerular basement membrane (GBM). Foot processes contain actin bundles (AB), and adjacent foot processes are attached by their slit diaphragms (SDs). (c): Transmission electron micrograph of podocyte foot process effacement in human focal segmental glomerulosclerosis (FSGS). (d): Scheme of podocyte foot process flattening. Left panel shows normal podocyte foot processes, actin cytoskeleton supports its elaborated shape. Once the actin cytoskeleton is disorganized, podocytes are no longer able to keep the foot process assembly, which results in fusion and flattening of foot processes (right panel).

((a)–(c): Courtesy of Dr. Dontscho Kerjaschki, Medical University Vienna, Vienna, Austria; (b): From Ronco, P. (2007). Proteinuria: Is it all in the foot? J. Clin. Invest. 117 : 2079–2082, with permission.)

While the podocyte is terminally differentiated and largely post-mitotic, the podocyte cell body contains a number of organelles that are crucial to maintain its high metabolic activity and secretory function, including a large nucleus, abundant lysosomes, and many mitochondria. In addition to producing factors necessary for maintaining their own specialized architectural cytoskeleton and protecting adjacent endothelial cells, podocytes must synthesize many of the components of the GBM in the mature glomerulus (reviewed in ).

The structure of the cell body and the primary major processes is determined by microtubules and intermediate filaments (vimentin, desmin, and nestin), while the shape (width and length) of the foot processes is largely determined by actin microfilament bundles. The actin bundles form continuous loops that run longitudinally in the foot processes and end at the sole plates, connecting adjacent foot processes associated with a single, primary, major process ( Figure 22.15b ). The loops of actin bundles are tethered to the microtubules and intermediate filaments of the major processes. This connection is thought to be mediated by Tau, which is known to form connections between microtubules and microfilaments and is concentrated at these areas. In the foot processes, a complex system links the slit diaphragm proteins, GBM receptors on the basolateral side, and the actin cytoskeleton to regulate the functional morphology of podocytes; this will be discussed at greater length later.

Podocyte Development, Transcription Factors, and Notch

Podocyte precursors first appear in the “S-shaped body” phase of glomerular development as a columnar epithelium attached along their lateral membrane by a cadherin junction containing P-cadherin. This nascent glomerulus is also populated by a thin epithelial layer called the Bowman’s capsule, and a capillary loop that begins to enter the glomerular cleft. As glomerular development progresses, the primitive podocytes extend themselves around the capillary loop by an unknown mechanism and differentiate into mature podocytes. A number of transcription factors have been identified that are expressed in the immature and mature podocytes, including Wt-1, Pod1 (Tcf21), Lmx1b, Kreisler (Maf1), and Foxc2.


The transcription factor Wt-1 is first expressed in the metanephric mesenchyme, which contains progenitors of the tubular and glomerular epithelium. As renal development continues, its expression becomes restricted to the renal vesicle, comma, and the S-shaped body. Later on, its expression is observed in immature and mature podocytes, which persists throughout life.

Wt-1 is a transcription factor with four zinc fingers that can bind both DNA and RNA. There are four major splice variants of Wt-1 mRNA, which may be responsible for its variety of functions in development and normal physiology. Embryonic loss of Wt-1 in mice leads to complete renal and gonadal agenesis, defects in the epicardium, lack of sub-epicardial mesenchymal cells, and adrenal agenesis, suggesting its important role in multiple organs.

Several mouse models underscore the importance of Wt-1 function during nephrogenesis and in podocytes. Postnatal deletion of the Wt-1 gene results in glomerulosclerosis with loss of podocyte foot processes and molecular markers of differentiation such as nephrin, atrophy of the exocrine pancreas and spleen, severe reduction in bone and fat mass, and failure of erythropoiesis. In Wt-1 knockout mice, restitution of Wt-1 expression using a human derived Wt-1-YAC transgene leads to a rescue of cardiac defects, but incomplete glomerular development. Furthermore, Wt-1 haploinsufficient mice develop adult onset mesangiosclerosis and glomerulonephritis.

In humans, Wt-1 mutations are associated with two glomerulopathies: Denys–Drash syndrome (DDS) and Frasier syndrome (FS), which can both present early in life and cause abnormal glomerular development. DDS is caused by heterozygous mutations of the Wt-1 gene that predominantly affect the zinc finger regions in exons 8 (zinc finger II) and 9 (zinc finger III), and directly interfere with the DNA-binding capacity of Wt-1. The characteristic clinical picture of DDS is a triad of congenital nephrotic syndrome, XY pseudohermaphroditism, and Wilm’s tumor. The characteristic renal lesion in these patients is diffuse mesangiosclerosis due to increased matrix deposition on the vascular side of the glomerular basement membrane. Expression of the DDS mutation in mice, either in podocytes or systemically, leads to various glomerular lesions similar to those of humans. Wt-1 is expressed in podocytes, which are found on the opposite side of the glomerular basement membrane (GBM) to the mesangium. The glomerular lesion in DDS therefore highlights the importance of interactions that occur between podocytes on one side of the GBM and endothelial cells or mesangial cells on the other side of the GBM. Frasier syndrome (FS) is characterized by focal segmental glomerular sclerosis, male-to-female sex reversal, but no tumors. It results from a dominant mutation that causes an inability to include an alternatively spliced lysine-threonine-serine (KTS) sequence after the third zinc finger of Wt-1 . Interestingly, mice engineered to exclusively express the variant that is +KTS or −KTS display normal induction of the metanephric mesenchyme by the ureteric bud, but have malformed glomeruli. These findings suggest different roles for the alternatively spliced variants of Wt-1 in glomerular development and maintenance.

While the contribution of Wt-1 to renal development and maintenance is undeniable, its precise role is still being elucidated. Human and animal models seem to indicate it plays a role in regulating the expression of a number of developmental genes. During early renal development the regulatory gene Pax-2 is expressed in the induced metanephric mesenchyme and ureteric bud, but is absent in the mature renal epithelium. In podocytes, the expression of Wt-1 coincides with a downregulation of Pax-2, suggesting a role for Wt-1 in transcriptional repression of this gene. Indeed, in patients with DDS, downregulation of Wt-1 was associated with increased Pax-2 expression ; however, it is unclear if this increased Pax-2 expression is pathogenic. Recent advances in high-throughput technology for assessing transcription factor activity have elucidated a number of other Wt-1 regulated genes that are known to play a role in renal development, including Six2, Bmp7, Sall1, and HeyL.

Wt-1 also plays a role in maintaining podocyte homeostasis beyond development. The essential podocyte slit diaphragm protein nephrin has been identified as a transcriptional target of Wt-1, and is downregulated in the glomeruli of mice lacking the Wt-1(−KTS) splice variant. Furthermore, Wt-1 modulates the podocyte glycocalyx through transcriptional regulation of Podocalyxin and 6-O-endosulfatases Sulf1 and Sulf2, which play a role in maintaining the charge of the glomerular filtration barrier, podocyte foot process separation, and bioavailability of essential growth factors such as vascular endothelial growth factor-α (Vegfa) and fibroblast growth factor (Fgf)-2. In addition to post-translational regulation of Vegfa bioavailability, Wt-1 can also directly regulate expression of Vegfa, which is essential to the maintainenance of the glomerular vasculature. Given this wide range of regulatory functions, it is still unclear what is responsible for the pathology observed in DDS and FS patients, but it is likely a convergence of a number of these signaling networks.


Lmx1b encodes a Lim-domain protein that is mutated in Nail–Patella syndrome. In mice lacking Lmx1b, podocytes retain their immature cuboidal phenotype, fail to form foot processes or endothelial fenestrations, and have a split glomerular basement membrane. These findings are associated with a concomitant downregulation of Vegfa, Synaptopodin, Nphs2, type III collagen α4 (Col3a4), and type IV collagen α4 (Col4a4). Lmx1b-binding elements in the Nphs2 and Col4a4 promoter regions suggest that this occurs through direct transcriptional regulation.


Tcf21 (also known as Pod1/Capsulin/Epicardin) encodes a basic helix-loop-helix transcription factor that is highly-expressed in the developing kidney, lung, intestine, and pancreas at sites of mesenchymal–epithelial interaction. In the developing mouse kidney, it is expressed in the condensing mesenchyme, and knockdown of Tcf21 expression in renal explants causes decreased mesenchymal condensation and ureteric bud branching. Subsequently, Tcf21 expression is restricted to the primitive podocytes of the S-shaped bodies, and genetic deletion of Tcf21 in mice leads to failure of podocyte terminal differentiation ( Figure 22.16 ). Tcf21 −/− knockout mice have a marked reduction in glomerular number, and the remaining glomeruli are developmentally arrested between S-shaped body and capillary loop stages.

Figure 22.16

Transmission electron micrograph of podocyte foot processes at E18.5.

Normal mouse podocytes show organized foot processes assembly (a), whereas Tcf21 null mouse podocytes demonstrate defects of foot process development (b).


Kreisler (Mafb) encodes a basic domain leucine zipper transcription factor that is expressed in podocytes at the capillary loop stage of glomerular development. Mice with a homozygous enu mutation of Kreisler show a similar but milder phenotype to those with a Tcf21 deletion. They are born with glomeruli arrested at the capillary loop stage, and podocytes that adhere to the GBM but fail to form foot processes. Kreisler mutants, however, express Tcf21, suggesting that Tcf21 is either upstream of Kreisler in podocyte development or acts by a different mechanism.


Foxc2 is a winged helix transcription factor that was identified during a screen of enriched genes in mRNA isolated from glomeruli at different stages of development. Foxc2 is first expressed in putative podocytes during the comma shaped body stage of glomerular development, and as such is the earliest known podocyte marker. Glomeruli from Foxc2 −/− mice display aberrant podocyte foot process formation, mesangial cell clustering at the base of the glomerular tuft, and swollen endothelial cells lacking fenestrae, similar to Tcf21 −/− mutant mice. In these mice, the endothelial and mesangial defects are thought to be secondary to the podocyte defects. Gene array of Foxc2 −/− glomeruli identified over 700 differentially regulated genes. Notably, however, there was a strong downregulation of Nphs2, Col4α3, and Col4α4, which are fundamental to slit diaphragm assembly and GBM formation.


The Notch family includes four well-conserved genes that encode transmembrane receptors involved in cell fate specification and development from invertebrates to mammals. Notch is fundamental in the segmentation of the metanephric mesenchyme into 20 functionally-distinct cell types segregated into different compartments along the nephron.

During normal development, Notch2 is expressed in condensing structures of the metanephric mesenchyme, such as the metanephric vesicles, comma-shaped bodies, and S-shaped bodies, but is eventually restricted to differentiating podocytes in more mature glomeruli. Jag1 is the Notch2 ligand thought to be important for glomerular differentiation. During development, it is expressed in renal vesicles, comma-shaped bodies, and S-shaped bodies. In more mature glomeruli it then localizes to the inner region of the glomerular tuft to endothelial and/or mesangial cells. Mice homozygous for a hypomorphic mutation in the Notch2 gene exhibit hypoplastic kidneys, and an arrest of glomerular development prior to the capillary loop stage. A conditional inactivation of Notch2 from nephron progenitor cells results in a more severe “distal tubule only” phenotype, with a complete failure of glomerulogenesis and proximal tubule formation.

In humans, mutations in Jag1 or Notch2 cause Alagille syndrome, an autosomal dominant disorder characterized by the presence of cholestatic liver disease, cardiac disease, ocular abnormalities, skeletal abnormalities, and characteristic facial features. A large proportion of Alagille patients also develop renal disease characterized by glomerular lesions, cystic kidneys, and ultimately renal failure. Taken together these studies suggest an important role for Notch signaling in establishing the proximal–distal orientation of the nephron, and segmentation of the proximal structures.

Recent evidence also reports a crucial role for the Notch pathway in the maintenance of mature podocytes. In patients with a variety of proteinuric nephropathies, including diabetic nephropathy and FSGS, activation of the Notch pathway was observed in podocytes. Furthermore, upregulation of Notch expression in podocytes of transgenic mice results in severe proteinuria and glomerulosclerosis with dedifferentiation of podocytes.

Podocyte Slit Diaphragm Assembly

The attachments between foot processes are comprised of the “slit diaphragm” (SD). SD assembly is an important part of glomerular development, as it is integral to the correct interdigitation of podocyte foot processes. The SD, which is visible by high power electron microscopy, is a structure that connects adjacent foot processes. It consists of a complex of transmembrane proteins and cytoskeletal adaptor proteins that link adjacent foot processes to the complex actin cytoskeleton of the podocyte and make up a component of the protein barrier (reviewed in ). While the relative importance of the endothelial layer versus the glomerular basement membrane versus the podocyte SD in maintaining the perm-selectivity of the glomerular filtration barrier is debatable, studies in humans and mice have shown that SD components are essential to this function.

In 1974 Rodewald and Karnovsky described the slit diaphragm as rod-like units connected in the center to a linear bar forming a zipper-like pattern with pores. They hypothesized that because of the size of these pores (40 Å×140 Å), the slit-diaphragm was the principal filtration barrier to plasma proteins in the kidney.


While this first description of the SD shed light on the filtration function of podocytes, the molecular composition of the SD remained poorly-defined. However, the discovery that a mutation in the NPHS1 gene causes Congenital Nephrotic syndrome of the Finnish variety suggested that dysregulation of SD structure can cause glomerular disease. This disease is characterized by massive proteinuria in utero , lack of a slit diaphragm, and abnormal foot process formation. The NPHS1 gene encodes a 180 kDa transmembrane protein of the immunoglobulin superfamily called nephrin that is expressed in the glomerular podocyte, and localizes specifically to the SD. These findings defined the importance of nephrin to the formation and maintenance of a normal SD, and led to the postulation of a “zipper-like model” of nephrin assembly in the SD. Accordingly, it is considered that the SD is the principle structure of the glomerular filtaration barrier and nephrin is its main component. Subsequent work, however, has also elucidated an important role for nephrin as a mediator of actin cytoskeletal organization by binding Src homology domain SH2/SH3 containing Nck adaptor proteins. The cytoplasmic tail of nephrin contains three tyrosine-aspartic acid-x-valine (YDxV) residues which, when phosphrylated by Src family kinases, recruit the SH2 Nck adaptor proteins and induce local actin polymerization ( Figure 22.17 ). Following the identification of nephrin, intensive research has led to the discovery of several other transmembrane proteins that participate in the formation of the slit diaphragm ( Figure 22.18 ).

Figure 22.17

Transmission electron micrograph of distinct mouse podocyte foot-processes present at 4 days of birth (a), but absent in Nck knockout mice (b). (c)–(g): Cellular immunostaining showing co-localization of Nck2, nephrin at the actin tail. Nck-nephrin interaction is required for nephrin-dependent actin reorganization (Green: nephrin; purple: Nck2; red: palloidin=actin). (d), (e), and (f) show actin, nck2, and nephrin staining, respectively.

((a) and (b): From Jones, N., et al. (2006). Nck adaptor proteins link nephrin to the actin cytoskeleton of kidney podocytes. Nature 440 : 818–823, with permission; (c)–(g): Courtesy of Drs. Tony Pawson and Nina Jones.)

Figure 22.18

Schematic diagram of the podocyte foot process and slit diaphragm with associated molecules.


NPHS2 is a gene that is mutated in some forms of steroid-resistant nephrotic syndrome that cause early onset proteinuria, and focal and segmental glomerulosclerosis. This gene encodes a protein called podocin, which associates in podocyte lipid rafts with nephrin and another SD component, CD2AP, via its C-terminal domain. Nephrin is thought to contribute directly to the formation of the SD, while podocin and CD2AP are thought to mediate its connection to the podocyte actin cytoskeleton. Mouse models in which any of these three components are disrupted lead to a congenital nephrotic syndrome.

Neph Proteins

The C-terminal domain of podocin also interacts with another group of three immunoglobulin superfamily transmembrane proteins called Neph1, 2, and 3 that bear significant homology to nephrin. This family is defined by their well-conserved cytoplasmic tail with a centrally located tyrosine residue required for interacting with podocin. The extracellular domain of Neph1 interacts with nephrin in the SD, and is essential for the regulation of glomerular perm-selectivity. In addition to podocin, the C-terminal (intracellular) domain binds the tight junction protein-1 (Tjp-1/ZO-1), which in turn tethers it to the podocyte actin cytoskeleton. Like nephrin, Neph1 plays a role in intracellular signaling, and is tyrosine phosphorylated at the SD, particularly in certain disease models. Mutations in Neph1 result in proteinuria and perinatal lethality in mice, but the phenotype is less dramatic than that observed in nephrin knockout mice, suggesting that nephrin may be more crucial to maintain the SD. This Nephrin-centric view of the slit diaphragm was called into question by a recent study showing that chickens and developing chicks lack nephrin, but express all three Neph proteins. Ultrastructurally, however, chickens are still able to assemble a SD in the glomerular filtration barrier. These SDs lacking nephrin still express P-cadherin and the large protocadherin Fat-1, suggesting that heterophilic interactions between Neph proteins and these other cadherins may be sufficient to allow SD assembly.

Cadherin and Catenins

In addition to these atypical junctions, the slit diaphragm contains adherens junction proteins P-cadherin, and α, β, γ catenins. P-cadherin is a transmembrane protein, and the extracellular domain is thought to contribute to slit diaphragm formation, while its cytoplasmic tail connects to β or γ catenin. Linkage of this complex to the actin cytoskeleton is believed to occur through an interaction between α catenin and Tjp-1/ZO-1 or α-actinin-4, both of which can directly bind actin.

The occludens protein Tjp-1/ZO-1 is expressed specifically at the points of insertion of the slit diaphragms in mature glomeruli. Tjp-1/ZO-1 appears early on in podocyte development when the apical junctional complexes between podocytes are composed of typical tight and adherens junctions, and persists as these junctional complexes migrate to the basolateral side and ultimately form the SDs. These findings suggest that tight junction proteins play a role in podocyte SD development and function, independent of their ability to form tight junctions.

α -Actinin-4

In humans, mutations in α-actinin-4 result in an autosomal dominant familial FSGS. In vitro the mutant form of α-actinin-4 binds F-actin more strongly than wild-type. This is thought to reduce the podocytes’ ability to respond dynamically to the hydrostatic pressure required for normal glomerular filtration, and ultimately leads to podocyte injury. In addition, α-actinin-4 is thought to mediate the interaction between the actin cytoskeleton and integrins to regulate podocyte adhesion to the glomerular basement membrane.


FAT1 is another member of the cadherin superfamily expressed in the podocyte at the SD, and has 34 tandem cadherin-like extracellular repeats. With its large extracellular domain, FAT1 is a major molecular component of the SD, and is required for normal foot process formation as FAT1 knockout mice lack SDs. Because of its localization to the cell–cell contact sites and tips of cellular processes, FAT1 may be involved in the initial steps of cell–cell interaction between podocytes.

Rho GTPases

Given the actin rich nature of the podocyte cytoskeleton, it is not surprising that Rho GTPases – master regulators of cytoskeletal dynamics – are important in podocyte biology. At the leading edge, Rac1 and Cdc42 promote lamellipodia and fillipodia formation, thus enhancing cell motility. In contrast, RhoA promotes a contractile phenotype by inducing formation of actin-myosin stress fibers. In this sense, it is believed that a balance between the opposing activities of RhoA and Cdc42/Rac1 regulates podocyte cytoskeletal dynamics. Both overexpression and inhibition of these small Rho GTPases in podocytes causes glomerular injury in mouse models. Podocyte-specific deletion of Cdc42 leads to a congenital nephropathy in transgenic mice by impairing actin polymerization at sites of nephrin clustering ( Figure 22.19 ), while RhoA deletion does not result in proteinuria. By contrast, overexpression of RhoA in podocytes results in an FSGS phenotype in mice.

Figure 22.19

Scanning ((a) and (b)) and transmission ((c) and (d)) electron micrograph of murine podocytes.

(a) and (c): Normal appearance of podocytes. (b) and (d): Podocytes lacking Cdc42 show extensive effacement of foot process at postnatal day 5 ((a), (b)×11, 000; (c), (d)×27,000).

(From Scott, R. P., et al. (2012) Podocyte-specific loss of Cdc42 leads to congenital nephropathy. J. Am. Soc. Nephrol. 23 (7): 1149–1154 with permission.)

Recently, dysregulation of Rho-GTPase signaling has also been linked to human glomerular disease. ARHGAP24 is a RhoA-activated Rac1 GTPase-activating protein (Rac1-GAP); a mutation that impairs Rac1-GAP activity results in FSGS in humans. Two recent reports also showed that mutations in the Inverted Formin 2 ( IFN2 ) gene result in autosomal dominant FSGS. INF2 interacts with other diaphanous related formins such as Cdc42; disease causing mutations in INF2 result in decreased CDC42 targeting to the plasma membrane, and impaired actin polymerization and depolymerization in podocytes.

Podocyte–GBM Interaction

The podocyte actin cytoskeleton is not only required for formation of podocyte foot processes and the slit diaphragm, but is also involved in a dynamic and bidirectional cross-talk with the glomerular basement membrane (GBM). In addition to production of GBM components, podocytes also express transmembrane molecules on their basolateral surface which interact with the GBM to regulate intracellular signaling. These include α3β1 integrin, αvβ3 integrin, and α- and β-dystroglycans.

α 3 β 1 Integrin

α3β1 integrin is a heterodimeric cell adhesion receptor with specificity for collagen types I and VI, laminins, fibronectin, and nidogen. α3β1 integrin is highly-expressed on the basolateral surface of mature and developing podocytes, and plays an important role in the development and maintenance of podocyte foot processes. Mice with a targeted mutation of the α3 integrin gene lack podocyte foot processes, and have a reduced number of capillary loops. Furthermore, the podocyte-derived basement membrane fails to fuse with the endothelium-derived basement membrane and becomes fragmented and disorganized. A recent paper reported three paediatric patients who have mutations in ITGA3 , which encodes α3 integrin; they presented with congenital nephrotic syndrome and severe basement membrane abnormalities. Taken together, these findings indicate that α3β1 integrin is not only a receptor for GBM components, but is also required for its development and maintenance. Similarly, podocyte specific ablation of the β1 integrin gene in mice results in massive proteinuria, abnormal capillary morphogenesis, podocyte foot process effacement, and podocyte apoptosis.

α v β 3 Integrin

While α3β1 integrin is the primary integrin expressed in podocytes, more recent work has identified a role for αvβ3 integrin in podocyte homeostasis. αvβ3 integrin is a vitronectin receptor that localizes to podocyte foot processes, and is predominantly associated with non-lipid raft fractions of the cell membrane. Genetic deletion of β3 integrin does not result in an overt renal phenotype, but these mice are protected from LPS-induced proteinuria, suggesting that αvβ3 integrin activation may play a role in the pathogenesis of proteinuria in the setting of glomerular injury. Further work by Wei et al. showed that activation of β3 integrin by soluble urokinase receptor (suPAR) may occur in primary focal segmental glomerular sclerosis; the authors suggest that suPAR may be the circulating factor responsible for recurrence of FSGS after transplant. While these data suggest that αvβ3 integrin activation may injure the podocyte, other work shows that αv integrins act as receptors for osteopontin, which is a protective factor in stretch-induced podocyte injury.


Integrins are not the only adhesion proteins to be expressed in podocyte foot processes. The dystrophin–glycoprotein complex (DGC) is a group of proteins that includes α- and β-dystroglycan, utrophin, and dystrophin, and plays a central role in stabilizing skeletal muscle cell membranes by tethering the cytoskeleton to the basement membrane components laminin, perlecan, and agrin. α- and β-dystroglycan have also been localized to podocyte foot processes, and their expression is diminished in several mouse models of glomerular disease and human minimal change disease. However, recent studies have shown that genetic deletion of dystroglycan from podocytes in mice does not result in a glomerular phenotype or increased susceptibility to injury, suggesting that integrins may be the primary functional extracellular matrix receptors in the podocyte.

Integrin-Linked Kinase

Given the central, non-redundant role of α3β1 integrin in the development and maintenance of the glomerular filtration barrier, it is of interest to identify its interacting proteins in podocytes. Integrin-linked kinase (ILK) is a serine/threonine kinase with kinase-dependent and -independent functions that interact with the cytoplasmic domains of β1 and β3 integrins. Detailed molecular studies have elucidated a role for ILK in podocyte morphology and health. ILK forms a complex with Pinch1 and α-Parvin to regulate matrix adhesion, foot process formation, and inhibit podocyte apoptosis. ILK also forms a ternary complex with α-actinin-4 and α3β1 integrin, providing a link between the GBM, the actin cytoskeleton, and nephrin in the SD. ILK was first identified as a potential mediator of glomerular disease by two groups showing that it is upregulated in glomeruli of patients with diabetic nephropathy, congenital nephrotic syndrome of the Finnish type, and two proteinuric mouse models. Indeed, activation of ILK in vivo using a rodent model of puromycin-associated nephropathy, and in vitro by overexpression of a kinase-active ILK transgene, caused activation of β-catenin, podocyte detachment and apoptosis, and transcriptional repression of the SD components P-cadherin and Cd2ap. However, while ILK activation may contribute to podocyte injury in disease, it also plays a fundamentally important role in normal podocyte physiology. Mice with a podocyte-specific deletion of the ILK gene appear normal at birth, but develop focal segmental glomerular sclerosis with GBM thickening, and podocyte foot process effacement characterized by an aberrant distribution of α-actinin-4 and nephrin. The deletion of ILK causes an upregulation of focal adhesion kinase (FAK), a non-receptor tyrosine kinase which is involved in focal adhesion turnover, cell spreading, and motility. This result is supported by recent work showing that FAK is activated in LPS and anti-GBM models of podocyte injury, and genetic deletion of FAK protects mice from proteinuria.

Tetraspanin CD151

CD151 is a tetraspanin family protein with affinity for α3β1 integrin that is expressed abundantly in the glomerulus. While CD151 is not absolutely required for α3β1 integrin binding to the extracellular matrix, it stabilizes the active conformation of α3β1 integrin and strengthens this interaction. A nonsense mutation in CD151 causes hereditary nephropathy in patients characterized by a splitting and thickening of the GBM, along with pretibial epidermis bullosa, sensorineural deafness, and thalassemia. A podocyte specific knockout mouse model recapitulates this renal phenotype. Mechanistically, CD151 causes a redistribution of α3β1 integrin at the interface between the podocyte and GBM, increasing its binding affinity for laminin-511/521. These findings suggest that CD151 may strengthen the adhesion of podocytes to the GBM, protecting them from higher glomerular pressures.

Negative Charge on the Surface of Podocytes

Although heparan sulfate proteoglycans are key components of the GBM and endothelial cell layer (as discussed above), they are also expressed by podocytes. The two major families of heparan sulfates expressed by the podocyte are Podocalyxin and Syndecans (primarily Syndecan I and Syndecan IV). Podocalyxin and Syndecans are transmembrane proteins involved in regulation of the podocyte actin cytoskeleton through regulation of signaling pathways (see below). While heparan sulfates provide a negative charge to the surface of the podocyte, sialyation of glycoproteins (including podocalyxin) and gangliosides also imparts a negative charge. Both are important components of glomerular barrier function.


Podocalyxin is a CD34-related sialomucin protein that is highly-expressed by podocytes and also by mesothelia, vascular endothelial cells, hematopoietic stem cells, and platelets. Mice with a conventional knockout of the podocalyxin gene die within the first 24 hours of life from renal failure. Importantly, podocyte foot processes do not form, and intercellular junctions between adjacent foot processes are abnormal and appear immature. These data provide functional evidence that podocalyxin is a key molecule in podocyte development. Cell biologic experiments have shown that podocalyxin interacts with the Na + /H + exchanger regulatory factor 2 (NHERF2) and phosphorylated ezrin in a complex, connecting it to the actin cytoskeleton of the podocyte foot process. Disruption of this interaction results in nephrotic syndrome. Podocalyxin appears to regulate foot process architecture through activation of the small Rho-GTPase, RhoA, mediated by its interaction with the NHERF/Ezrin complex.

Sialyation Defects and the Podocyte

Podocytes express a number of sialyated proteins including podocalyxin (as described above), other proteoglycans, and gangliosides. Loss of sialyation of podocalyxin has been identified in various nephrotic syndrome experimental models, such as in rodents injected with sialidase, puromycin aminonucleoside or protaimine sulfate, which neutralizes negative charges. All of these compounds cause an abrupt onset of proteinuria, together with foot process effacement. Simultaneous infusion of sialic acid prevents the proteinuria and podocyte foot process fusion observed with puromycin injection, presumably due to resialylation of critical glomerular proteins. More recently, loss of sialyation of podocalyxin was observed in a transgenic rat model of minimal change disease due to overexpression of Anptl4.

Although many studies have focused on sialylation defects of podocalyxin, the podocyte expresses other sialylated glycoproteins and gangliosides. Mutations in sialylating enzymes have also been associated with glomerular defects and proteinuria. For example, a point mutation in a gene encoding one of the key enzymes needed for sialic acid biosynthesis (uridine disphospho-N-acetylglucosamine (UDP-GlcNAc) 2-epimerase/N-acetyl-mannosamine (ManNAc) kinase (GNE/MNK)) results in severe perinatal glomerular disease in mice, characterized by splitting of the glomerular basement membrane, hematuria, and proteinuria. The phenotype was partially rescued by dietary supplementation with ManNAc. Intriguingly, ManNAc supplementation also appeared to rescue the Anptl4-induced sialyation defect in rats, suggesting it might represent a new therapy for certain forms of glomerular disease.

Podocytes and Metabolism

Although the majority of podocyte studies to date have focused on its unique cytoskeletal architecture and its role as a structural component of the filtration barrier, recent studies have highlighted the importance of metabolic regulatory pathways in podocyte function.


The mechanistic target of rapamycin (mTOR) is an evolutionarily-conserved serine-threonine kinase that interacts with regulatory associated protein of mTOR (Rptor) or Rptor-independent companion of MTOR (Rictor) to form mTORC1 and mTORC2 complexes, respectively. mTORC1 is a key regulator of cellular metabolism, including protein translation, ribosomal biogenesis, cell growth and proliferation, and suppression of autophagy in response to amino acids, growth factors, and elevated cellular ATP levels. mTORC2 is regulated primarily by growth factors to promote actin cytoskeletal rearrangement, cell survival, and cell cycle progression. Rapamycin is an mTOR inhibitor that is used clinically and is thought to specifically inhibit mTORC1 function. In certain cell types including the podocyte, however, chronic inhibition of mTORC1 by rapamycin also results in downregulation of mTORC2 functions. The importance of mTOR in podocyte biology was first suggested by the clinical observation that rapamycin causes proteinuria.

Deletion of mTOR itself or R ptor in podocytes of mice results in proteinuria. Conversely, ectopic mTORC1 activation in mouse podocytes, accomplished by deletion of its suppressor Tsc1 , results in kidney disease with many of the features of diabetic nephropathy (DN), including podocyte hypertrophy and loss and proteinuria, and is attributed to endoplasmic reticulum stress. Although mTORC1 appears to be crucial for podocyte function, loss of Rictor from podocytes of mice does not result in a phentoype.

Given its central role in cellular metabolism, mTOR is likely to play multiple roles in podocyte biology. Regulation of autophagy is one pathway regulated by mTOR that appears to be crucial for podocyte function. Autophagy is a lysosomal-dependent cellular survival response to starvation or lack of growth factors in which cells degrade cellular constituents from proteins to entire organelles, such as mitochondria, in order to provide a supply of nutrients under conditions of stress. A basal level of autophagy is, however, necessary to remove damaged organelles, excessive lipids, and long-lived or misfolded proteins. The basal level of autophagy appears to be increased in podocytes, and it has been suggested that autophagy may be required to protect this long-living cell from injury. In keeping with this model, deletion of Atg5, a key component of the autophagic pathway, results in late onset of glomerular disease in mice at 20 to 24 months of age due to the accumulation of damaged organelles and ubiquitinated protein complexes. In contrast, deletion of mTOR from podocytes results in disruption of autophagic flux, with subsequent accumulation of autophagolysomes in the podocyte ( Figure 22.20 ). Clinically, dysregulation of autophagy is observed in patients with lysosomal storage diseases such as Fabry’s disease, Aspartilglucoseaminuria or Scheie’s disease, where the inability to acidify lysosomes causes a failure of lysosomal reformation ; these patients are prone to developing proteinuria, providing additional support that the autophagic pathway is clinically relevant.

Jun 6, 2019 | Posted by in NEPHROLOGY | Comments Off on Glomerular Cell Biology
Premium Wordpress Themes by UFO Themes